Understanding [PDF]

Understanding

Autism

From Basic Neuroscience to Treatment

© 2006 by Taylor & Francis Group, LLC

Understanding

Autism

From Basic Neuroscience to Treatment Edited by

Steven O. Moldin, Ph.D. University of Southern California Los Angeles, California

John L.R. Rubenstein, M.D., Ph.D. University of California San Francisco, California

Boca Raton London New York

CRC is an imprint of the Taylor & Francis Group, an informa business

© 2006 by Taylor & Francis Group, LLC

2732_Discl.fm Page 1 Thursday, March 23, 2006 9:46 AM

This book was written by Dr. Steven O. Moldin in his private capacity, outside his professional position as Director of the Office of Human Genetics and Genomic Resources, and Associate Director, Division of Neuroscience & Basic Behavioral Science, National Institute of Mental Health (NIMH), National Institutes of Health (NIH), Bethesda, Maryland, USA. The views expressed in this book do not necessarily represent the views of NIMH, NIH, the Department of Health and Human Services nor of the United States Government. Published in 2006 by CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2006 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-10: 0-8493-2732-6 (Hardcover) International Standard Book Number-13: 978-0-8493-2732-2 (Hardcover) Library of Congress Card Number 2005046674 This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data Understanding autism : from basic neuroscience to treatment / edited by Steven O. Moldin, John L.R. Rubenstein. p. ; cm. Includes bibliographical references and index. ISBN 0-8493-2732-6 (alk. paper) 1. Autism. I. Moldin, Steven O. II. Rubenstein, John L.R., 1955[DNLM: 1. Autistic Disorder--diagnosis. 2. Autistic Disorder--genetics. 3. Autistic Disorder--therapy. WM 203.5 U546 2006] RC553.A88U53 2006 616.85’882--dc22

2005046674

Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com Taylor & Francis Group is the Academic Division of Informa plc.

© 2006 by Taylor & Francis Group, LLC

and the CRC Press Web site at http://www.crcpress.com

2732_C000.fm Page v Saturday, March 18, 2006 3:07 PM

Foreword Autism spectrum disorders exact a serious toll on those affected, on their families, and, as a result of behavioral problems and significant levels of disability, on society. Despite that toll, autism has not historically received adequate research attention. The lack of research reflected a mistaken belief that autism was very rare, and there was a lack of reliable, broadly accepted diagnostic standards, and until recently, a lack of scientific tools with which to make progress. As attested to by this volume, all this has changed. While we still have far to go before we possess a deep understanding of autism that translates into highly efficacious new treatments, research on autism is moving at an increasingly rapid pace and yielding promising new results. What factors have led to the exciting new progress in autism research that is summarized in this text? An important contribution, without doubt, was the tireless advocacy of the families of affected individuals, families who refused to accept the message that there was no hope. Interest in autism was also spurred by the broad recognition that it represented a spectrum of disorders, which in aggregate were more common than had been believed and more costly to society (Chapter 20). This new epidemiological information (Chapter 2) and indeed all of autism research were facilitated by improved approaches to diagnosis. Progress in phenotyping and diagnosis (Chapter 1) required not only new scientific efforts, but also a willingness on the part of many researchers to reach consensus on the assignment of individuals to the autism spectrum and to different points along the way. All this could not have created the stirrings of progress without the emergence of new scientific tools and approaches. Some, like the explosion of new tools for genetics developed entirely independently of autism research, others like social neuroscience — the investigation of the circuitry and neural mechanisms that underlie human interaction — developed in part, as a result of the new interest in autism. Social neuroscience itself depended on the development of increasingly good tools for human neuroimaging and the broad advance made by cognitive neuroscience in addressing the underpinnings of thought, emotion, and behavior. It is hard to overstate both the potential significance and also the difficulty of a genetic analysis of the autism spectrum (Chapters 3 and 4). Gene discovery should provide tools for neuroscientists trying to understand the neural basis of autism, and may provide clues to new treatments. Based on twin studies, we know that of all risk factors, genes make the largest contribution to autism spectrum disorders. As is well known, twin studies also reveal that genes are not fate — not all monozygotic twin pairs with an affected member are concordant for an autism spectrum disorder. Thus, developmental, stochastic, or environmental factors must play a role in converting genetic risk into disease. Nonetheless, genes exert a powerful effect, more powerful than for schizophrenia or bipolar disorder, for example. Despite the large effect of genes, the search for genetic risk factors has proved extremely frustrating. In recent years, it has been hypothesized that autism spectrum disorders —indeed, all common neuropsychiatric phenotypes — are genetically complex. That is to say

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page vi Saturday, March 18, 2006 3:07 PM

that while genes are highly significant in aggregate, the genetic component of risk is comprised of many genetic loci, each making a small contribution that is therefore difficult to detect. Moreover, different combinations of genetic risk factors might act in different populations. Without the recent genomic and genetic tools that derived from the Human Genome Project and its many follow-on projects, it would simply be impossible to identify risk genes involved in common neuropsychiatric disorders. Recent maps of human genetic diversity and cost-effective tools for genotyping are beginning to make gene hunting feasible. If the challenges posed by genetic complexity were not enough, we are beginning to realize that the inherited sequence of genes may not be the whole explanation for the influence of the genome on disease risk (Chapter 5). Epigenetic factors may also play a role in autism spectrum disorders and other neuropsychiatric disorders. Epigenetic influences on gene expression reflect the methylation of certain DNA sequences and other mechanisms that regulate the structure of chromatin and therefore gene expression. Epigenetic risk factors are not detected by many of the common approaches to neuropsychiatric genetics. While it is daunting to confront the complexity of identifying genetic, epigenetic, developmental, and environmental factors that interact to produce autism, it is also a hopeful sign that scientists investigating autism are incorporating sophisticated new approaches and, at the same time, adding to them (Chapter 5). If genetics represents one of the central “bottom-up” approaches to autism, cognitive and social neuroscience represent critical “top-down” approaches. Several chapters in this book reflect advances in our analysis of the cognitive (Chapter 8), the emotional (Chapter 6), and the social brain (Chapter 10) that are producing insights into autism. Beyond deeper understanding, the results that will emerge from these fields should ultimately have profound practical significance. Cognitive tests combined with neuroimaging should make an important contribution to phenotyping for genetics and other more basic investigations. Right now, the heterogeneity of individuals with autistic symptoms is a major obstacle to research. In time, these approaches, perhaps combined with genetics, may lead to more certain clinical diagnoses, perhaps predictive of treatment response. Finally, as in all disease processes, biomarkers, i.e., objectively measurable correlates of disease progression, can play an important role in the development and monitoring of treatment. Rather than relying on subjective ratings, clinical trials of both pharmacologic and behavioral treatments, and clinical practice might one day be able to benefit from objective measures. As reflected in this volume, developmental neurobiology, the creation of animal models (Chapter 12), systems neurobiology (Chapters 7, 9, 11, 13), and structural neuroimaging (Chapter 15) of the human brain are also making significant contributions to our understanding of the autism spectrum. As a disorder of the brain that impairs higher cognitive function, emotion, and behavioral control, autism can only be understood by bringing many different disciplines together. While we still need a great deal of progress in each of these areas and above all at their interfaces, the accomplishments of the past decade, as illustrated in this book, have created a palpable sense of positive movement. Steven E. Hyman, M.D. Harvard University Cambridge, Massachusetts

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page vii Saturday, March 18, 2006 3:07 PM

Preface This volume presents a comprehensive review of the state-of-the-art research on the diagnosis, treatment, and prevention of autism. It also addresses potential mechanisms that may underlie the development of autism and the neural systems that are likely to be affected by these molecular, genetic, and infectious etiologies. Autism is a devastating neurodevelopmental disorder with early childhood onset and a recently reported prevalence of as high as 73 per 10,000. Autistic symptoms continue throughout life and include significant impairments in social, communicative, cognitive, and behavioral functioning. There is strong evidence for genetic factors in the etiology of autism; however, the mode of inheritance is complex and reflects the interplay of the environment and many different genetic factors. Postmortem and MRI studies have identified abnormalities in several major cortical and subcortical brain structures in autism, and neuropsychological studies have implicated specific impairments in executive functioning and in processing social and emotional information. Distinctive neuropathological features have been reported, such as Purkinje cell loss in the cerebellum. Recent work has implicated neonatal brain undergrowth followed by rapid and excessive postnatal brain growth. Whereas these studies are consistent with the hypothesis that autism is primarily a disorder of brain functioning, no defining set of pathophysiological mechanisms has been unambiguously defined. Animal models offer a complementary approach to human studies and provide the opportunity to study in an experimental system the neural substrates of behaviors that are profoundly impaired in autistic individuals. One particularly promising avenue of inquiry has focused on the neurobiology of social behavior. Despite promising results, researchers have yet to identify a robust model of many of the core abnormalities observed in autism and their specific neurodevelopmental trajectories. This may be a limitation of studying nonhumans. The past few years have witnessed extraordinary advances in molecular genetic techniques and the accumulation of structural genomics information and resources in both human and model organisms (Moldin and Hyman, 2005). In addition, new approaches are dramatically increasing our understanding of brain development, structure, and function. With the development of new technologies and the availability of genomic resources, previously inconceivable experiments in basic neuroscience are now being routinely implemented in autism research. Research activities are being stimulated by funding from private foundations including Cure Autism Now, the National Alliance for Autism Research, the Southwest Autism Research and Resource Center, the Simons Foundation, and Autism Speaks — and multiple institutes at the National Institutes of Health (NIMH, NINDS, NICHD, NIDCD, NIEHS). Intensive autism advocacy interest resulted in the creation of the Coalition for Autism Research and Education (CARE), formed by members of the Congressional Autism Caucus, which includes nearly 200 congressmen from

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page viii Saturday, March 18, 2006 3:07 PM

almost every state. The Children’s Health Act of 2000 passed by Congress mandated the establishment of autism research centers and an Interagency Autism Coordinating Committee (IACC) to coordinate autism research and other efforts within the Department of Health and Human Services. Members include the five NIH institutes that support autism research and numerous other federal agencies, e.g., CMS, CDC, DOE, FDA, and HRSA. The availability of exciting new scientific tools and technologies, as well as intensive advocacy efforts and the availability of millions of dollars in funding from private foundations, are setting the stage for a renaissance in autism research studies in humans and in experimental systems. These burgeoning research activities will have a profound impact on generating considerable insights into the etiology and pathophysiology of autism. However, much of the current research on autism has not achieved its full potential, as human studies, guided by a fundamental grounding in neural pathways, circuits, and molecules, expand continuously with the avalanche of information being made available with the sequencing of the human genome and those of other organisms (Moldin, 2003). This volume reviews key findings on autism, ranging from the genetic and neural mechanisms that may underlie this disorder through state-of-the-art diagnosis, epidemiology, clinical neuroscience, and treatment approaches. The book concludes with a chapter that discusses the economic cost of autism and provides a biomedical and public health perspective of the impact of this devastating disease. The chapters are authored by clinical and basic researchers who are at the forefront of molecular and systems neuroscience, clinical neuroscience, genetics, and health economics. The first section (Chapter 1 and Chapter 2) describes current approaches to diagnosis of autism and autism spectrum disorders, and reviews the state of the art of the epidemiology of autism. The latter is especially important, given the recent controversies about the changing incidence and prevalence of autism. The second section (Chapter 3 to Chapter 6) covers genetic and genomic technologies that are currently being used to dissect the molecular basis of autism. Chapter 3 provides an exhaustive review of cytogenetic, linkage, association, candidate gene, and genetic mapping studies in autism. Chapter 4 describes exciting new approaches to the analysis of gene expression and also the extension of current studies to include endophenotypes in families of autistic probands. Chapter 5 examines the potential role that epigenetic mechanisms may play in the etiology of autism, through the regulation of gene expression by mechanisms that do not alter gene sequences. The analysis of specific developmental disorders that result in autistic symptoms are providing insights into mechanisms underlying autism. These include Angelman’s and Rett syndromes (Chapter 5), tuberous sclerosis, and Fragile X syndrome (Chapter 6). Fragile X syndrome is caused by reduced expression of the FMRP RNA-binding protein whose function is linked to synaptic function. Based on observations made in Fragile X and several other lines of evidence, the field is currently exploring the possibility that alterations in synapse development and signaling may underlie some forms of autism (Zogbi, 2003; Rubenstein and Merzenich, 2003; Levitt et al., 2004). This perspective has been bolstered by the identification of function-altering mutations in some neuroligin genes in a small subset of autistic

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page ix Saturday, March 18, 2006 3:07 PM

people (Jamain et al., 2003). Neuroligin proteins, together with their binding partners (neurexins), regulate the formation of synapses; different neurexin/neuroligin combinations appear to participate in specifying whether new synapses are assembled into excitatory or inhibitory synapses (Graf et al., 2004; Chih et al., 2005; Chubykin et al., 2005; Cline, 2005). Given that ~40% of autistic people develop epilepsy, a syndrome characterized by excessive excitatory tone, it is possible that autism can be caused by developmental abnormalities that alter the ratio of excitatory/inhibitory synapses in key neural circuits (Hussman, 2001; Rubenstein and Merzenich, 2003; Levitt et al., 2004). This hypothesis is currently being examined in several laboratories. For instance, Rett syndrome results in increased expression of the Dlx5 transcription factor (Horike et al., 2005). The Dlx gene family has a central role in regulating the development and function of forebrain inhibitory neurons (Panganiban and Rubenstein, 2002; Cobos et al., 2005). Recently, five nonsynonymous Dlx2 and Dlx5 mutations have been identified in autistic probands (Hamilton et al., 2005). Although it is premature to conclude that there is a causal link that connects the Rett gene (MeCP2), the Dlx genes, and autism, this line of investigation illustrates one current approach that is aimed at elucidating the molecular and cellular basis for some forms of autism. The third section of this volume (Chapter 7 to Chapter 13) focuses on behaviors, and the neural systems that underlie them, that are frequently abnormal in autistic people. These include fear and anxiety (Chapter 7), cerebellar networks (Chapter 8), language (Chapter 9), the prefrontal cortex (Chapter 10), the amygdala and other aspects of complex social behavior (Chapter 11), and the thalamus and neuromodulatory systems (Chapter 12). All of these chapters are especially important to neuroscientists and nonneuroscientists alike, given that they provide a clear explication of the molecular and cellular neuroscience of autism that serve as the bedrock for future autism research paradigms. Chapter 13 describes the considerable potential of animal models in autism for characterizing the roles of genes and environment, understanding pathogenesis, and testing potential therapeutic approaches. This approach is important, given the recent controversies regarding the potential link between autism and environmental neurotoxins, such as mercury. The fourth section (Chapter 14 to Chapter 17) describes current clinical findings from studies of autistic people. Included are reviews and perspectives that focus on neuroanatomy and neurochemistry (Chapter 14), functional neuroimaging (Chapter 15), structural neuroimaging (Chapter 16), and neurophysiology and neuropsychology (Chapter 17). Given well-replicated observations of elevated platelet serotonin in a subset of autistic probands and first-degree relatives (Chapter 14), considerable excitement has been generated by recent work (McCauley et al., 2004; Sutcliffe et al., 2005) implicating multiple rare alleles at the serotonin transporter locus (SLC6A4) in the etiology of autism. Current state-of-the-art treatments for autism are described in the fifth section. The discussion of both pharmacological (Chapter 18) and behavioral, educational, and developmental treatments (Chapter 19) point to new avenues for future research, and also suggest ways in which our understanding of underlying neural systems can inform drug development and therapeutics.

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page x Saturday, March 18, 2006 3:07 PM

Finally, this volume concludes with discussion of the health economics of autism (Chapter 20) that includes a theoretical and methodological outline of how to identify the public health costs of autism. Also included are estimates of those costs based on publicly available data. In any given year, autism can cost our society about $3.2 million per autistic child; over a lifetime each individual costs a staggering $35 billion in direct medical, direct non-medical, and lost productivity costs to care for all autistic children over all of their lifetimes. Ganz’s work is the most comprehensive cost analysis conducted to date, and this information will be useful to cost-effectiveness analysts performing economic evaluations of treatment and prevention options and to policymakers and advocates as a reference source on the costs of autism. We anticipate that this book will be a valuable resource for clinical and basic scientists, as well as legislators and research advocates, and will provide a powerful and comprehensive synthesis of current research on autism and its underlying neural substrates. This information is expected to grow and guide future basic and clinical research on autism. A key objective is to stimulate new directions for research on etiology and pathophysiology, as well as for the development of new drugs. We express our appreciation to each of the contributors who worked so diligently on their chapters, and the anonymous reviewers who generously took the time to carefully read each chapter and provide detailed feedback to the authors. Steven Moldin is grateful for all of the kind encouragement and support he received from Yelizaveta Zolotukhina. We also thank the staff of CRC Press/Taylor and Francis Group, especially Barbara Norwitz and Kari Budyk, for their help, patience, and encouragement. Most importantly, we express our deep gratitude to the caring relatives of autistic people, to the health care workers and researchers who dedicate themselves to this serious problem, and to the advocates who vigorously raise funds, raise public awareness, and fight for legislation aimed at solving this difficult problem. Their devotion to the scientific pursuit of greater understanding, better treatments — and ultimately a cure — has been an inspiration to us. It is to them, and to the people with autism whom we wish to help, that we humbly dedicate this book.

REFERENCES Chih, B., Engelman, H., and Scheiffele P., Control of excitatory and inhibitory synapse formation by neuroligins, Science 2005, 307(5713): 1324–1328. Chubykin, A.A., Liu, X., Comoletti, D., Tsigelny, I., Taylor, P., and Sudhof, T.C., Dissection of synapse induction by neuroligins: effect of a neuroligin mutation associated with autism, J Biol Chem 2005, 280(23): 22365–22374. Cline, H., Synaptogenesis: a balancing act between excitation and inhibition, Curr Biol 2005, 15(6): R203–205. Cobos, I., Calcagnotto, M.E., Vilaythong, A.J., Thwin, M.T., Noebels, J.L., Baraban, S.C., and Rubenstein, J.L., Mice lacking Dlx1 show subtype-specific loss of interneurons, reduced inhibition and epilepsy, Nat Neurosci 2005, 8(8): 1059–1068.

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page xi Saturday, March 18, 2006 3:07 PM

Graf, E.R., Zhang, X., Jin, S.X., Linhoff, M.W., and Craig, A.M., Neurexins induce differentiation of GABA and glutamate postsynaptic specializations via neuroligins, Cell 2004, 119(7): 1013–1026. Hamilton, S.P., Woo, J., Carlson, E.J., Ghanem, N., Ekker, M., and Rubenstein, J.L.R., Analysis of four DLX homeobox genes in autistic probands, BMC Genet, 2005, Nov. 2(6): 52. Horike, S.I., Cai, S., Miyano, M., Cheng, J.F., Kohwi-Shigematsu, T., Loss of silent-chromatin looping and impaired imprinting of DLX5 in Rett syndrome, Nat Genet 2005, 37: 31–40. Hussman, J.P., Suppressed GABAergic inhibition as a common factor in suspected etiologies of autism, J Autism Dev Disord 2001, 31: 247–248. Jamain, S., Quach, H., Betancur, C., Rastam, M., Colineaux, C., Gillberg, I.C., Soderstrom, H., Giros, B., Leboyer, M., Gillberg, C., and Bourgeron, T., Paris autism research international sibpair study. Mutations of the X-linked genes encoding neuroligins NLGN3 and NLGN4 are associated with autism, Nat Genet 2003, 34(1): 27–29. Levitt, P., Eagleson, K.L., and Powell, E.M., Regulation of neocortical interneuron development and the implications for neurodevelopmental disorders, Trends Neurosci 2004, 27: 400–406. McCauley, J.L., Olson, L.M., Dowd, M., Amin, T., Steele, A., Blakely, R.D., Folstein, S.E., Haines, J.L., and Sutcliffe, J.S., Linkage and association analysis at the serotonin transporter (SLC6A4) locus in a rigid-compulsive subset of autism, Am J Med Genet (Neuropsychiatric Genet) 2004, 127B: 104–112. Moldin, S.O., Neurobiology of autism: the new frontier, Genes Brain Behav 2003, 2(5): 253–254. Moldin, S.O. and Hyman, S.E., Genome, transcriptome, and proteome, in Sadock, B.J. and Sadock, V.A., Eds., Kaplan and Sadock’s Comprehensive Textbook of Psychiatry, 8th ed., Baltimore, MD: Lippincott Williams & Wilkins, 2005: pp. 115–125. Panganiban, G. and Rubenstein, J.L., Developmental functions of the Distal-less/Dlx homeobox genes, Development 2002, 129(19): 4371–4386. Rubenstein, J.L.R. and Merzenich, M.M., Model of autism: increased ratio of excitation/ inhibition in key neural systems, Genes Brain Behav 2003, 2: 255–267. Sutcliffe, J.S., Delahanty, R.J., Prasad, H.C., McCauley, J.L., Han, Q., Jiang, L., Li, C., Folstein, S.E., and Blakeley, R.D., Allelic heterogeneity at the serotonin transporter locus (SLC6A4) confers susceptibility to autism and rigid-compulsive behaviors, Am J Hum Genet 2005, 77: 265–279. Zoghbi, H.Y., Postnatal neurodevelopmental disorders: meeting at the synapse?, Science 2003, 302(5646): 826–830.

Steven O. Moldin, Ph.D. Washington, D.C. John L. R. Rubenstein, M.D., Ph.D. San Francisco, California

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page xiii Saturday, March 18, 2006 3:07 PM

Editors Steven O. Moldin, Ph.D., is Executive Director of the Washington, DC-based Office for Research Advancement in the Office of the Provost at the University of Southern California. Dr. Moldin received his B.A. (magna cum laude with distinction and Phi Beta Kappa) in psychology from Boston University in 1983, his M.A. in psychology from Yeshiva University, New York, in 1985, and his Ph.D. in clinical psychology from Yeshiva University in 1988. After working as an assistant research scientist in the Department of Medical Genetics at the New York State Psychiatric Institute, New York, and completing a clinical internship at Hillside Hospital, Long Island Jewish Medical Center, Glen Oaks, New York, from 1988 to 1991, he received postdoctoral training in quantitative genetics at Washington University School of Medicine in St. Louis, Missouri. From 1991 to 1995 he was an assistant professor in the Department of Psychiatry at Washington University School of Medicine and director of the Center for Psychiatric Genetic Counseling at the Washington University Medical Center from 1993 to 1995. Dr. Moldin joined the National Institute of Mental Health (NIMH), National Institutes of Health (NIH) in 1995 to manage an extramural research portfolio. He left public service in 2006, having led the Office of Human Genetics & Genomic Resources and having served as Associate Director of the Division of Neuroscience and Basic Behavioral Science. He is an associate editor of Genes, Brain, and Behavior and serves on the editorial board of the American Journal of Medical Genetics (Neuropsychiatric Genetics). Dr. Moldin has published over 50 papers and book chapters, including co-editing with Dr. Hemin Chin the volume Methods in Genomic Neuroscience in CRC Press’s Methods & New Frontiers in Neuroscience series. Dr. Moldin’s current position includes promoting major research initiatives at the University of Southern California on a broad range of topics that include neuroscience, genomics, and autism. John L. R. Rubenstein, M.D., Ph.D., is the Nina Ireland Distinguished Professor in Child Psychiatry at the University of California at San Francisco (UCSF). Dr. Rubenstein received his B.S. degree (Phi Beta Kappa) in chemistry in 1977 from Stanford University. He then became an MSTP student, earning a Ph.D. in biophysics in 1982. He worked with Harden McConnell and James Rothman on the effect of cholesterol on the motions of phospholipids and proteins in membranes, and on the biogenesis of plasma membrane proteins. He received his M.D. from Stanford Medical School in 1986. As a postdoctoral fellow at the Pasteur Institute with Francois Jacob and Jean Francois Nicolas (1984 to 1986) he showed that antisense RNA can inhibit gene expression, and developed retroviral vectors for gene delivery and fate mapping in mouse embryos. As a resident physician in child

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page xiv Saturday, March 18, 2006 3:07 PM

psychiatry at Stanford (1986 to 1991) he began the experiments that he continues to this date on the genetic control of brain development and function. He is particularly interested in genetic mechanisms that could contribute to causing neuropsychiatric disorders such as autism and schizophrenia. He has been on the faculty of UCSF since 1991and has coauthored over 150 scientific papers. He is currently associate editor for the Journal of Comparative Neurology, section editor for Neuroscience, and on the editorial board of Development, Cerebral Cortex, Developmental Dynamics, Thalamus, and Experimental Neurology. He is on the scientific advisory board of Merck Pharmaceuticals, Cure Autism Now, and National Alliance for Research on Schizophrenia and Depression (NARSAD).

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page xv Saturday, March 18, 2006 3:07 PM

Contributors Maricela Alarcón, Ph.D. David Geffen School of Medicine University of California at Los Angeles Los Angeles, California

Arthur L. Beaudet, M.D. Department of Molecular and Human Genetics Baylor College of Medicine Houston, Texas

David G. Amaral, Ph.D. The M.I.N.D. Institute University of California Davis Medical Center Sacramento, California

Elena Bonora, Ph.D. Unità di Genetica Medica Dipartimento di Medicina Interna Epatologia e Cardioangiologia Bologna, Italy

George Anderson, Ph.D. Child Study Center Yale University School of Medicine New Haven, Connecticut

Ruth A. Carper, Ph.D. Center for Autism Research Department of Neurosciences University of California, San Diego La Jolla, California

Anthony J. Bailey, M.D. Department of Psychiatry University of Oxford Park Hospital for Children Oxford, United Kingdom

Katarzyna Chawarska, Ph.D. Child Study Center Yale University School of Medicine New Haven, Connecticut

Gabrielle Barnby, Ph.D. The Wellcome Trust Centre for Human Genetics University of Oxford Oxford, United Kingdom

Eric Courschene, Ph.D. Department of Neurosciences University of California, San Diego La Jolla, California

Margaret L. Bauman, M.D. Department of Neurology Harvard Medical School Massachusetts General Hospital Wellesley, Massachusetts

Geraldine Dawson, Ph.D. University of Washington Autism Center University of Washington Seattle, Washington

Melissa D. Bauman, Ph.D. The M.I.N.D. Institute University of California Davis Medical Center Sacramento, California

Richard P. Dum, Ph.D. Department of Neurobiology University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page xvi Saturday, March 18, 2006 3:07 PM

Eric Fombonne, M.D. The Montreal Children’s Hospital The McGill University Health Centre Montreal, Quebec, Canada

Christopher J. Machado, Ph.D. The M.I.N.D. Institute University of California Davis Medical Center Sacramento, California

Michael L. Ganz, Ph.D. Department of Society, Human Development, and Health Harvard School of Public Health Boston, Massachusetts

Christopher J. McDougle, M.D. Department of Psychiatry Indiana University School of Medicine Indianapolis, Indiana

Daniel H. Geschwind, M.D., Ph.D. David Geffen School of Medicine University of California at Los Angeles Los Angeles, California Steven E. Hyman, M.D. Office of the Provost Harvard University Cambridge, Massachusetts Kevin S. LaBar, Ph.D. Center for Cognitive Neuroscience Duke University Durham, North Carolina Janine A. Lamb, D.Phil. Wellcome Trust Centre for Human Genetics University of Oxford Oxford, United Kingdom Joseph E. LeDoux, Ph.D. Center for Neural Science New York University New York, New York Catherine Lord, Ph.D. UMACC Departments of Psychology and Psychiatry Ann Arbor, Michigan

© 2006 by Taylor & Francis Group, LLC

Nancy J. Minshew, M.D. Department of Psychiatry University of Pittsburgh Pittsburgh, Pennsylvania Anthony P. Monaco, M.D. Wellcome Trust Centre for Human Genetics University of Oxford Oxford, United Kingdom Usha Narayanan, Ph.D. Department of Human Genetics Emory University School of Medicine Atlanta, Georgia Sally Ozonoff, Ph.D. The M.I.N.D. Institute University of California Davis Medical Center Sacramento, California Paul H. Patterson, Ph.D. Division of Biology California Institute of Technology Pasadena, California Elaine Perry, M.D., Ph.D. IAH Research Labs Newcastle General Hospital Newcastle upon Tyne, United Kingdom David J. Posey, M.D. Riley Hospital for Children Indiana University School of Medicine Indianapolis, Indiana

2732_C000.fm Page xvii Saturday, March 18, 2006 3:07 PM

Joseph L. Price, D.Phil. Department of Anatomy and Neurobiology Washington University School of Medicine St. Louis, Missouri

Peter L. Strick, Ph.D. Veterans Affairs Medical Center and Department of Neurobiology University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania

Melissa Ray IAH Research Labs Newcastle General Hospital Newcastle upon Tyne United Kingdom

Helen Tager-Flusberg, Ph.D. Laboratory of Developmental Cognitive Neuroscience Department of Anatomy and Neurobiology Boston University School of Medicine Boston, Massachusetts

Sally J. Rogers, Ph.D. The M.I.N.D. Institute University of California Davis Medical Center Sacramento, California Robert T. Schultz, Ph.D. Yale Child Study Center Yale University New Haven, Connecticut Cynthia M. Schumann, Ph.D. Department of Neuroscience University of California, San Diego La Jolla, California Sarah J. Spence, M.D., Ph.D. UCLA Center for Autism Research and Treatment David Geffen School of Medicine Los Angeles, California Mircea Steriade, M.D., D.Sc. Faculty of Medicine Laval University Montreal, Quebec, Canada Kimberly A. Stigler, M.D. Riley Hospital for Children Indiana University School of Medicine Indianapolis, Indiana

© 2006 by Taylor & Francis Group, LLC

Michael T. Ullman, Ph.D. Department of Neuroscience Georgetown University Washington, D.C. Fred R. Volkmar, M.D. Yale Child Study Center Yale University New Haven, Connecticut Matthew Walenski, Ph.D. Department of Neuroscience Georgetown University Medical Center Washington, D.C. Stephen T. Warren, Ph.D. Department of Genetics Emory University School of Medicine Atlanta, Georgia Sara J. Webb, Ph.D. Psychiatry and Behavioral Sciences BRU/Autism Project University of Washington Seattle, Washington Graham M. Wideman, B.A.Sc. Center for Autism Research La Jolla, California

2732_C000.fm Page xviii Saturday, March 18, 2006 3:07 PM

Diane L. Williams, Ph.D., CCC-SLP Department of Psychiatry University of Pittsburgh Pittsburgh, Pennsylvania

© 2006 by Taylor & Francis Group, LLC

Huda Y. Zoghbi, M.D. Department of Molecular and Human Genetics Howard Hughes Medical Institute Baylor College of Medicine Houston, Texas

2732_C000.fm Page xix Saturday, March 18, 2006 3:07 PM

Contents Chapter 1 Autism Spectrum Disorders: Phenotype and Diagnosis...........................................1 Catherine Lord and Sarah J. Spence Chapter 2 Past and Future Perspectives on Autism Epidemiology .........................................25 Eric Fombonne Chapter 3 Genetic Basis of Autism..........................................................................................49 Elena Bonora, Janine A. Lamb, Gabrielle Barnby, Anthony J. Bailey, and Anthony P. Monaco Chapter 4 Finding Genes in Spite of Heterogeneity: Endophenotypes, QTL Mapping, and Expression Profiling in Autism........................................................75 Daniel H. Geschwind and Maricela Alarcón Chapter 5 A Mixed Epigenetic and Genetic and Mixed De Novo and Inherited Model for Autism..............................................................................95 Arthur L. Beaudet and Huda Y. Zoghbi Chapter 6 Neurobiology of Related Disorders: Fragile X Syndrome ...................................113 Usha Narayanan and Stephen T. Warren Chapter 7 Fear and Anxiety Pathways ...................................................................................133 Kevin S. LaBar and Joseph E. LeDoux Chapter 8 Cerebellar Networks and Autism: An Anatomical Hypothesis ............................155 Richard P. Dum and Peter L. Strick Chapter 9 Language in Autism...............................................................................................175 Matthew Walenski, Helen Tager-Flusberg, and Michael T. Ullman Chapter 10 Prefrontal Cortex ...................................................................................................205 Joseph L. Price

© 2006 by Taylor & Francis Group, LLC

2732_C000.fm Page xx Saturday, March 18, 2006 3:07 PM

Chapter 11 The Social Brain, Amygdala, and Autism ............................................................227 Cynthia M. Schumann, Melissa D. Bauman, Christopher J. Machado, and David G. Amaral Chapter 12 The Thalamus and Neuromodulatory Systems .....................................................255 Mircea Steriade Chapter 13 Modeling Features of Autism in Animals.............................................................277 Paul H. Patterson Chapter 14 Neuroanatomical and Neurochemical Studies of the Autistic Brain: Current Thought and Future Directions.....................................................303 Margaret L. Bauman, George Anderson, Elaine Perry, and Melissa Ray Chapter 15 The Social Brain in Autism: Perspectives from Neuropsychology and Neuroimaging .................................................................................................323 Robert T. Schultz, Katarzyna Chawarska, and Fred R. Volkmar Chapter 16 Structural Neuroimaging .......................................................................................349 Ruth A. Carper, Graham M. Wideman, and Eric Courchesne Chapter 17 Neuropsychology and Neurophysiology of Autism Spectrum Disorders ...............................................................................................379 Nancy J. Minshew, Sara J. Webb, Diane L. Williams, and Geraldine Dawson Chapter 18 Pharmacological Treatments..................................................................................417 Christopher J. McDougle, David J. Posey, and Kimberly A. Stigler Chapter 19 Behavioral, Educational, and Developmental Treatments for Autism .................443 Sally J. Rogers and Sally Ozonoff Chapter 20 The Costs of Autism..............................................................................................475 Michael L. Ganz

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 1 Wednesday, February 15, 2006 11:55 AM

1

Autism Spectrum Disorders: Phenotype and Diagnosis Catherine Lord and Sarah Spence

CONTENTS The Spectrum of Autistic Disorders..........................................................................1 The Phenotype ...........................................................................................................3 Core Features......................................................................................................3 Social and Communication Deficits in ASD ................................................3 Repetitive and Restricted Behaviors and Interests........................................4 Sex Differences...................................................................................................5 Other Associated Features..................................................................................5 ASD and Cognitive Impairments ..................................................................5 Relationship to Sensory and Motor Impairments .........................................6 Relationship to Epilepsy................................................................................7 Macrocephaly in ASD ...................................................................................7 Comorbid Psychiatric Diagnoses ..................................................................8 Developmental Trajectories................................................................................9 Prognoses for ASD ...........................................................................................10 Relationship to Disorders with Known Etiology....................................................11 Neurogenetic Syndromes..................................................................................11 Diagnosis and Assessment.......................................................................................12 Standardized Diagnoses....................................................................................13 Summary and Conclusions......................................................................................14 Acknowledgments....................................................................................................15 References................................................................................................................15

THE SPECTRUM OF AUTISTIC DISORDERS Autism is a syndrome that emerges in the first three years of life and is defined by a pattern of qualitative abnormalities in reciprocal social interaction, communication, and repetitive interests and behaviors. The Diagnostic and Statistical Manual (DSM-IV)

1

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 2 Wednesday, February 15, 2006 11:55 AM

2

Understanding Autism: From Basic Neuroscience to Treatment

of the American Psychiatric Association includes five different disorders under an umbrella term of pervasive developmental disorders (PDDs). These include autistic disorder, Asperger’s disorder, pervasive development disorder not otherwise specified (PDDNOS), Rett syndrome, and childhood disintegrative disorder (CDD). Because there is great diversity in the severity of these features in affected individuals, an umbrella term of autism spectrum disorders (ASD) has been suggested to include autism, atypical autism or PDDNOS, and Asperger’s syndrome (see Volkmar et al., 2005). Disorders within the spectrum are discriminated from each other primarily by milder and less comprehensive difficulties (PDDNOS and atypical autism) or by the absence of language delay and mental retardation (Asperger’s syndrome). Family history and twin studies suggest that, at least in some cases, these disorders share genetic roots, but the degree to which different etiologies and genetic patterns account for individual differences within ASD is an open question (Piven et al., 1997). The last two disorders in the PDD umbrella are more rare and have more specific diagnostic features. Rett syndrome is included because of the phenotypic overlap with autism, at least in the preschool age group. Rett syndrome is characterized by a period of normal development followed by a regression in language and social skills usually between 6 and 18 months as well as the onset of hand stereotypies such as hand wringing (Hagberg and Witt-Engerstrom, 1986; Mount et al., 2003). It occurs mostly in girls. The presence of abnormal physical features seen in Rett syndrome, e.g., head growth deceleration, loss of purposeful hand movements, ataxia and gait abnormalities, scoliosis, and hyperventilation and breath holding (Tanguay, 2000), and the nature of the regression are the keys to differentiating this rare disorder from ASD. By definition, children with CDD must have normal development until age two and then experience a regression that affects not only social communication, but also other areas such as gross and fine motor skills (Volkmar and Rutter, 1995). Most often these regressions occur after age three. It is important to note that most researchers do not include individuals with Rett and CDD in samples of autism or ASD. Because of the lack of a clear neurobiological marker, ASDs are necessarily defined by behavior, which is both intriguing and frustrating for researchers. Nevertheless, there have been major advances in the last 20 years in the ability to reliably define and quantify the behaviors that differentiate autism and other ASDs from other disorders and from typical development (Lord and Corsello, 2005). In some ways, the field of autism is at a crossroads where categorical diagnoses, such as Asperger’s syndrome, or even not quite categorical diagnoses, such as PDDNOS, do not seem sufficient. However, reliable and valid measures of independent dimensions in ASD are not yet easily accessible. The next 10 years may see substantial progress in these areas as large and carefully documented samples become available for study. The purpose of this chapter is to provide background concerning phenotype and diagnosis for the interested neuroscientist. Having initially defined terms, various aspects of the phenotype of ASDs are discussed and general issues in diagnosis are considered.

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 3 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

3

THE PHENOTYPE CORE FEATURES Social and Communication Deficits in ASD In the last 20 years, our understanding of the aspects of communication and social interaction specific to children with ASD, compared to children with various developmental disabilities, has become increasingly refined. For example, deficits in communication in ASD go beyond language delay to include a social failure to compensate for this delay, which other children (e.g., children who are deaf) do through gesture, eye contact, and increased attention to facial expressions. Other common deficits in communication associated with autism include stereotyped language, such as delayed echolalia, reciting passages from favorite videos or commercials, pronoun reversal, and use of stereotyped phrases (such as a child who says, “Can I call you right back, sweetie?” when he does not want to answer a question). Most individuals with autism are delayed in their acquisition of both receptive and expressive language. Delays in receptive language have been proposed to be particularly associated with autism in preschool children (Philofsky et al., 2004). Formerly, it was expected that half of all individuals with autism would not use speech as their primary mode of communication. Nevertheless, in a recent study, the proportion of 9-year-olds with ASD who spoke fluently was about 40% in two independent samples, and the proportion who were nonverbal (i.e., who used fewer than five words on a daily basis) was less than 15% (Lord et al., in press), perhaps because of better intervention and also the broadening of the diagnostic criteria (see Chapter 2). Individuals with autism have difficulty with imitation, imaginative play, and nonverbal communication — three categories of behavior that are sometimes considered examples of communication deficits and sometimes social deficits (APA, 2000; WHO, 1992). The most prototypical examples of social deficits have to do with reciprocity, such as seeking to share enjoyment (e.g., coming to get a parent to see a new Lego construction), feeling genuine concern and offering comfort to another person, and forming caring friendships that go beyond classroom or parentarranged interactions. Several recent studies have suggested that, rather than considering social and communication deficits as separate, it is more parsimonious and valid to think of a single social communication factor that includes nonverbal communication and reciprocal conversation. Separate consideration would still be given to whether an individual is delayed in basic dimensions of vocabulary, syntax, and phonology in either or both receptive and expressive language (Charman et al., 2005). Many of the examples of social communication deficits in ASD involve behaviors that typical infants master in the first year or two of life, such as following another person’s shift in gaze and other aspects of joint attention, vocalizing “back” to someone who is talking to them, and smiling at someone who smiles and vocalizes to them in a positive way (Baranek, 1999). In fact, difficulties with joint attention are probably considered the most clear “schema” marking autism as different from other developmental disorders (Mundy and Sigman, 1989). Lack of social reciprocity

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 4 Wednesday, February 15, 2006 11:55 AM

4

Understanding Autism: From Basic Neuroscience to Treatment

at a higher level of abstraction also fills that role, particularly for older children and adults who may have learned the importance of attending to gaze but still may not be able to have a conversation in which they spontaneously ask about and listen to how someone else feels about a particular event. Children and adults with ASD are different from all but very young babies in basic social behaviors. On the other hand, some of the deficits in autism occur in areas that are associated with developmental change such as gestures, more complex imaginative play, and cooperative play in a group, thus representing delays as much as deviance. Repetitive and Restricted Behaviors and Interests The next defining area, restricted and repetitive behaviors and interests, contains the largest proportion of examples that represent deviance. Here we see the presence of behaviors that would be abnormal at any age (e.g., stereotyped hand and finger movements, odd ways of visually inspecting objects, and unusual intonation). There are also unusual preoccupations (e.g., with drainpipes, flags, TV show credits) and circumscribed interests so intense that they interfere with social interactions and other behaviors (e.g., a child who has to carry Disney figurines with him and will not put them down even in order to play with a new toy or pick up a cookie). They also include preoccupations with a part of an object, such as the wheels on toy vehicles, and unusual sensory responses, such as smelling toys or people, as well as repetitive behaviors such as lining up toys or spinning objects, flicking light switches, or opening and closing cupboard doors. Unusual motor behaviors, most often involving rapid movements of the hands and fingers, often in peripheral vision, or whole body movements, such as spinning or running and flapping or repetitive hopping and posturing are also common (LeCouter et al., 1989). Repetitive behaviors and interests differ from the abnormalities in social communicative behaviors in their variety across individuals with ASD, and variability within individuals across time (Charman et al., 2005). They also appear to have somewhat different trajectories, at least from early childhood to adolescence (Lord et al., in press; Richler, Bishop, and Lord, in press). ASD is also characterized by insistence on sameness which includes both the development of unusual rituals (such as lining toes up with a crack in the sidewalk on the way to school) and substantial distress when everyday routines are violated (such as having a bath earlier or later than usual), although it should be noted that these behaviors occur in substantial numbers of children and adults with other disabilities as well (Shao et al., 2003; Lord et al., 1994). There are subtle differences in the behaviors shown by children with ASD and those by children and adults with obsessive-compulsive disorder (OCD). In ASD, repetitive behaviors are often enjoyable to the child or adult and involve intense interests such as dinosaurs or flags or Japanese animé; whereas with OCD the individual is typically uncomfortable when carrying out the behavior, and the behaviors are most often common compulsive behaviors such as checking and counting (Leckman et al., 1997). There are a number of theoretical issues frequently raised in ASD that occur in consideration of a number of psychiatric disorders. Space is too limited to discuss

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 5 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

5

them here, but they include the idea of a spectrum of disorders, how delay is separated from deviance, how “core” difficulties affect experience and in turn affect learning and development, and how co-occurring conditions should be taken into account. Active research programs are currently addressing these issues.

SEX DIFFERENCES Autism and all other ASDs except Rett Syndrome are much more common in males than females, with ratios of 3 or 4 to 1, found in most epidemiological studies, and ratios of up to 10 to 1 for many research samples (Lord et al., 1982; Fombonne, in press). Sex ratios move closer to 1:1 for children with autism who are profoundly retarded. When differences in IQ are controlled for, it is not clear if females with ASD are different from males in terms of their behavioral presentation (Lord et al., 1982), but there are some suggestions from genetic-linkage studies that multiplex families with all affected males may differ genetically from families with males and females affected (Cantor et al., 2005; Lamb et al., 2005; Stone et al., 2004). Because of the rarity of females with autism and the need to control for IQ, small sample sizes have been a major limitation to many studies.

OTHER ASSOCIATED FEATURES ASD and Cognitive Impairments ASDs are associated with various degrees of mental retardation, which are often, though not always, related to the severity of autistic symptoms. As described in Chapter 2 of this volume, earlier it was believed that more than 75% of children with autism were also mildly to severely retarded. However, most recent epidemiological studies have indicated that the proportion of children with ASDs with nonverbal IQs below 70 (e.g., mildly to severely retarded) may be less than 50%. In part, this may be due to society and medicine’s increasing recognition of individuals with milder ASDs (Gernsbacher et al., 2005). It may also be due to better understanding of appropriate ways to separate nonverbal problem-solving skills from language skills in the assessment of children with limited receptive and functional expressive language skills. If carried out by an experienced clinician with appropriate tests, nonverbal IQs appear to be quite stable from age two to later school age in most children with ASDs and, if anything, they may increase up to about 20 points over time (Lord et al., 1982; Charman et al., 2005; McGovern and Sigman, 2005). There is much more variation in verbal IQ scores than nonverbal IQs in ASD. Separate measures of nonverbal intelligence and verbal ability are important in interpreting specific behaviors in ASD that are often linked to neurobiological factors, and these are also significant in selecting proper control groups. Consequently, it is critical that these assessments are done systematically and by experienced clinicians. There have also been several interesting, though not yet replicated, neurobiological findings (e.g., differences in head circumference) related to children with autism with very marked differences between nonverbal and verbal IQ (Joseph et al., 2002). Larger differences between performance IQ and adaptive scores are

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 6 Wednesday, February 15, 2006 11:55 AM

6

Understanding Autism: From Basic Neuroscience to Treatment

also one of the distinguishing characteristics of ASD compared to mental retardation without ASD. Comprehensive theories explaining the neuropsychological aspects of autism continue to be proposed but have not been able to easily address the range of findings with regard to phenotype, history and prognosis, and neurobiological factors. Various theories have attempted to account for social cognitive deficits through concepts such as a lack of central coherence (Frith, 2004), a lack of theory of mind (BaronCohen and Howlin, 1994; Lord and Richler, in press), and deficits in executive functioning (Ozonoff, 1995). A lack of central coherence is described as the inability to integrate information when a whole represents more than the sum of its parts (e.g., a picture of a face made up of typed “x’s”). Executive functioning refers to the ability to plan and organize action, including inhibiting simple responses and anticipating a progression of events. Specific deficits in joint attention, implicit learning, imitation, memory, and other aspects of information processing have all been proposed (Mundy and Sigman, 1989; Renner et al., 2000; Rogers and Pennington, 1991; Stone et al., 1997; Boucher, 1981; Ozonoff et al., 1994). When general intellectual level is taken into account, it has been difficult to show strong associations between specific cognitive deficits and the core social deficits of ASD except for joint attention. Nevertheless, the theories provide insight and working hypotheses for potentially important phenotypic characteristics and for behavioral treatments. Relationship to Sensory and Motor Impairments While not part of the diagnostic criteria, parents often report abnormal sensory behaviors in children with ASD. Both increased and decreased responsiveness to sensory stimuli in all domains have been reported (Rogers et al., 2003). Some individuals with ASD are described as tactilely defensive. They appear to not want to be touched, or they do not want to touch certain textures or surfaces. Common complaints include the inability to wear socks or shoes, an intolerance of clothes made of certain fabrics or tags in clothing, or extreme disturbance caused by brushing, washing, or cutting the child’s hair. Some individuals seek proprioceptive input by crawling under furniture or into small, cramped spaces, or seek vestibular input by spinning, swinging, or bouncing repetitively. In the auditory domain, severe behavioral reactions can be triggered by loud or unusual noises, or sometimes by common sounds such as coughing or singing. Other individuals may visually inspect objects (e.g., peering out of the corners of their eyes or examining things at very close range). Others have reported an increased pain threshold (Charman and Baird, 2002; Filipek et al., 2000). There are some data suggesting an underlying motor impairment in children with ASD. Motor milestones are delayed in up to 33% of cases (Mayes and Calhoun, 2003). Gait disturbances such as tiptoeing (Kielinen et al., 2004; Vilensky et al., 1981) and problems with balance and coordination (Ghaziuddin and Butler, 1998; Jones and Prior, 1985) have been documented. In a more detailed analysis, Minshew et al. showed the presence of significant postural abnormalities in children with ASD (Minshew et al., 2004).

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 7 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

7

A higher incidence of autism and autism-like behaviors has also been reported in individuals with primary sensory impairments, e.g., those with visual or auditory loss. It has long been recognized that congenital blindness is associated with an autism-like presentation (Carvill, 2001). Children blind from birth frequently show repetitive motor behaviors similar to the hand and finger stereotypies exhibited in autism, the so-called “blindisms.” Beyond that, others have reported impaired social and communication skills in blind children, as well (Brown et al., 1997; Hobson and Bishop, 2003). However, factors such as the etiology of the blindness, as well as methodological problems with differing diagnostic criteria and behavioral observation scales, make this a complex association (Carvill, 2001). There are fewer reports on the association between hearing loss and autism, but one study examining a large sample of hearing-impaired children found an autism rate of 4% (Jure et al., 1991). Relationship to Epilepsy The occurrence of epilepsy in individuals with autism has long been recognized; however, the reported prevalence varies widely from 5 to 44% (Tuchman and Rapin, 2002). Critical review of the literature reveals that heterogeneity in samples may play a role in this variability (see Ballaban-Gil and Tuchman, 2000 for review). Factors as simple as differences in recruitment and nonuniformity in epilepsy determination and autism diagnosis likely contributed. The age of participants in the sample may also play a role because of the bimodal age of seizure onset in autism (early childhood and adolescence). Rates may also be inflated by inclusion of nonidiopathic autism cases (e.g., those with gross brain malformations, cerebral palsy, tuberous sclerosis, or other neurogenetic disorders), which themselves have a higher rate of associated epilepsy. Finally, because some studies have found an association between lower IQ and increased risk of epilepsy (Pavone et al., 2004; Tuchman et al., 1991), differing IQ levels in the sample may impact reported rates. Epileptiform EEGs (e.g., spikes, spike wave, or sharp waves) are also reported in ASD patients with rates varying from 18 to 60% in those with seizures (Kawasaki et al., 1997; Rossi et al., 1995) and from 8 to 46% of those without seizures (Kawasaki et al., 1997; Tuchman et al., 1997; Tuchman et al., 1991). The increased risk for epilepsy and EEG abnormalities in individuals with ASD may provide an important clue to the underlying neuropathology in at least a subset of cases. Few studies have directly investigated the relationship of epilepsy and the common deficits in language, cognition, and behavior seen in this population. Over 10 years ago, Tuchman (1994) raised a series of fundamental questions regarding this relationship, including whether the occurrence of epilepsy or epileptiform EEGs together with the cognitive, language, and behavioral deficits seen in ASD were all just an epiphenomenon of the underlying neural dysfunction or were causally related to these deficits. So far these questions have gone unanswered. Macrocephaly in ASD Beginning with Kanner’s initial report (Kanner, 1943), increased incidence of large head size (macrocephaly) has been reported in autism. Postmortem studies (Bailey et al., 1998; Kemper and Bauman, 1993) as well as structural imaging studies

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 8 Wednesday, February 15, 2006 11:55 AM

8

Understanding Autism: From Basic Neuroscience to Treatment

(Piven et al., 1996; Piven et al., 1995) suggest that brain volume is increased in autism. Studies using strict definitions of macrocephaly (>97% or 2 SD or more above the mean) have confirmed this phenomenon (Fombonne et al., 1999; Lainhart et al., 1997; Stevenson et al., 1997) but report variable rates from as low as 12% (Fidler, Bailey, and Smalley, 2000) to as high as 30% of cases with autism or ASD (Woodhouse et al., 1996). In a meta-analysis of over 500 patients, a rate of 20.6% for macrocephaly in autism was calculated (Fombonne et al., 1999). The association between head circumference and gender is not well understood. One study showed significantly more females with macrocephaly (Lainhart et al., 1997); another showed the opposite (Davidovitch et al., 1996), and still others have found no association (Fombonne et al., 1999). There is also controversy regarding the age at which the head growth occurs. It is generally agreed that larger head size is common in younger children with ASD, but adult data has yielded more conflicting results. This may be a function of the fact that head growth better reflects brain growth at earlier ages. Lainhart and colleagues (Lainhart et al., 1997) found a greater percentage of older subjects with macrocephaly, suggesting that brain growth is occurring later. However, Aylward and colleagues found no overall difference in measurements of head size as a function of age, but did report significantly larger brain volumes in younger vs. older ASD children compared to controls (Aylward et al., 2002). Others have also reported the normalization of brain volume with age (Courchesne et al., 2001). These findings could be interpreted as suggesting that the earlier brain growth is associated with macrocephaly, which persists as the children age despite normalization of the brain size. Yet even the age at which the abnormal brain growth starts is unclear. Infantile macrocephaly has been associated with an increased risk of developing an autism spectrum disorder (Bolton et al., 2001). However, in a small retrospective analysis, Courchesne and colleagues reported the head circumferences of autism subjects were actually smaller at birth compared to a reference sample, and then a pattern of very early rapid head growth emerged (e.g., between birth and 14 months of age) in the autistic subjects. They further suggested that this may even be a marker for development of autism (Courchesne et al., 2003). Macrocephaly is also more common in nonautistic family members of autistic individuals (Fidler et al., 2000; Miles et al., 2000) and therefore may represent part of the broader phenotype in autism. On the other hand, it is not specific to ASD. High rates also occur in ADHD (Ghaziuddin et al., 1999) and possibly other developmental disorders. Comorbid Psychiatric Diagnoses Many symptoms present in psychiatric diagnoses can also be seen in individuals with ASD including attentional deficits and hyperactivity, anxiety, obsessivecompulsive behaviors, depression, and even psychosis. However, with the complexity of the behavioral phenotype in ASD, it is often difficult to determine the extent to which a given behavior is indicative of a separate psychiatric diagnosis or simply a manifestation of the autism. This phenomenon of comorbidity is of interest to

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 9 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

9

researchers because it may indicate important neuroanatomical, neurochemical, or genetic overlaps between the ASDs and these other disorders. Probably the most commonly recognized co-occurring symptom complex is that of attentional deficits and hyperactivity. By definition according to the DSM-IV, an ASD is an exclusionary criterion for making an attention deficit hyperactivity disorder (ADHD) diagnosis. However, studies have found that the symptomatology that would qualify for diagnosis is present in approximately one third of ASD individuals (Goldstein and Schwebach, 2004). There are many studies showing a higher prevalence of depression in individuals with ASD (for review, see Ghaziuddin et al., 2002). Depression might be more common in higher functioning patients (Ghaziuddin et al., 1998). Although this area of research is more controversial, there is also some evidence that there is an increased risk of bipolar disorder in individuals with ASD as well (Stahlberg et al., 2004; Wozniak et al., Kim et al., 1997) Anxiety (Muris et al., 1998), obsessive-compulsive symptoms (Russell et al., 2005), and even Tourette syndrome (Baron-Cohen et al., 1999) are also present in individuals with ASD. Schizophrenia has been reported in autism, but rates are extremely variable (Volkmar and Cohen, 1991; Konstantareas and Hewitt, 2001). Formal thought disorder has also been shown in individuals with autism at higher rates than those with other psychiatric diagnoses (van der Gaag, Caplan, van Engeland, Loman, and Buitelaar, 2005).

DEVELOPMENTAL TRAJECTORIES There are a number of trajectories of development that are associated with ASDs. Most parents of children with ASD, with hindsight, describe ways in which their children’s development was not quite right prior to 18 months (Rogers and DiLalla, 1990). Analyses of videotapes from the first year of life of children who were later diagnosed with autism revealed differences from typical children in response to name and in a number of different sensory behaviors even then. Differences between children later diagnosed with autism and children with other developmental disorders have been found but are less clear (Baranek, 1999; Osterling and Dawson, 1994). For example, difficulties with gaze were found to be more common in very young children who were later shown to have a developmental delay and not autism than those who received autism diagnoses. Before 12 months of age, differences between children with autism and those with typical development tend to be subtle and not clearly discriminative, whereas differences in social responsiveness are much clearer at or after age 1 (Zwaigenbaum et al., 2005). About one quarter to one third of children with ASD experience a clear loss of social and communication skills in the second year of life (Lord et al., 2004). Most often these children had already begun to fall behind in subtle ways before the regression. However, compared to children with ASD without losses, they had more frequent and more sophisticated social and communication behaviors until the time of regression, at which point they usually stopped talking and lost some if not all skills such as waving, playing peek-a-boo, and imitating sounds (Luyster et al., 2005). Children with ASDs who have had regressions do not seem to represent a

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 10 Wednesday, February 15, 2006 11:55 AM

10

Understanding Autism: From Basic Neuroscience to Treatment

discrete subgroup within ASD in terms of later symptoms or outcomes, but do show slightly lower verbal IQ scores, more gastrointestinal symptoms, and a more frequent family history of autoimmune disorders, compared to ASD children without regression (Richler et al., in press; Molloy et al., in press). As mentioned previously, regression is also a core feature of the developmental trajectory of Rett syndrome and CDD.

PROGNOSES

FOR

ASD

Prognoses for ASD are quite variable, though children with severe autism in early childhood seldom fall outside the spectrum of autism-related disorders as older children and adults (Lord et al., in press). Nevertheless, it is quite common for specific behavioral characteristics to change in many ways, including variation over time in the number and extent of repetitive behaviors (Moore and Goodson, 2003; Bishop and Norbury, 2002) and changes from social withdrawal or passivity to more active but odd interactions (Wing, 2005). Most children with early diagnoses of autism will not be completely independent as adults; many will need support in employment and residential living (Howlin, 2000). Nevertheless, a significant minority, especially of cases with less clear manifestations of the disorder in early years and with fluent language by age 5, will be able to take responsibility for many of their activities of daily living and will complete high school and even postgraduate education (Szatmari et al., 2003). The number of adults with autism who can drive, live on their own, and hold some kind of partially supported or independent employment has grown significantly in the last 10 years (Howlin et al., 2004). Social motivation and social skills frequently, though not always, improve in later childhood and adolescence and early adulthood. Nonverbal intelligence scores in the early years, particularly if a good assessment of early language skills is not available, are predictors of later prognosis, as are repetitive behaviors and the severity of social–communication deficits as measured through parent report or direct observation in younger children (Howlin et al., 2004; Lord et al., in press). Comorbid depression and generalized anxiety disorder become increasingly frequent as children with ASD enter adolescence and adulthood (Ghaziuddin and Greden, 1998). Szatmari and colleagues (2003) have suggested that overall functioning for individuals with ASDs is best characterized by two factors: one that describes the severity of autism-specific difficulties in social and repetitive behaviors, and the other that describes general level of functioning, including receptive and expressive language, nonverbal intelligence and adaptive skills, as well as the presence of comorbid disorders (Mahoney et al., 1998). As discussed in Chapter 2 of this volume, both CDD and Rett syndrome are very rare and are associated with more uniformly poor prognoses than autism and other ASDs. For many years, there have been case reports of individuals with autism who have marked deterioration in behavior and sometimes in other skills during adolescence (Rutter et al., 1994), but this is rarely reported in follow-up studies.

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 11 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

11

RELATIONSHIP TO DISORDERS WITH KNOWN ETIOLOGY Because autism is a clinical syndrome, it can and does coexist with other medical disorders. Current estimates are that 10 to 15% of ASD cases may be etiologically related to some known neurological or genetic disorder (Barton and Volkmar, 1998; Rutter et al., 1994) and 3 to 9% are thought to have detectable cytogenetic abnormalities (Fombonne et al., 1997; Wassink et al., 2001). Studying syndromes with high prevalence of autism may provide clues into the neuropathological and genetic underpinnings of autism spectrum disorders.

NEUROGENETIC SYNDROMES The most well-known associated medical conditions are Fragile X (FRAX) and tuberous sclerosis complex (TSC). A high percentage of patients with both of these disorders will show autistic features; although the occurrence of these disorders in samples ascertained for ASD is relatively small. Autism has been reported in up to 44% of patients with Fragile X syndrome (Philofksy et al., 2004), but fewer than 5% of autism patients have Fragile X (Bailey et al., 1993; Brown et al., 2002; Klauck et al., 1997; Piven et al., 1991; Wassink et al., 2001). Similarly, autism has been reported in 17 to 61% of individuals with TSC (Curatolo et al., 2004) but fewer than 3% of individuals with autism actually have TSC (Gillberg et al., 1994; Smalley, 1998) although this may be higher (8 to 14%) in autism patients with epilepsy (Smalley, 1998). Two preliminary screening studies suggested that MECP2 mutation rates (the mutation associated with Rett syndrome) may be as high as 3 to 5% in female autism samples (Carney et al., 2003; Lam et al., 2000) although other studies have not found the mutation (Lobo-Menendez et al., 2003; Vourc'h et al., 2001). A more recently described syndrome involving a maternally derived duplication of chromosome 15q overlapping the region deleted in Prader-Willi and Angelman syndromes has been reported to occur in 1 to 3% of autism cases (Cook et al., 1998; Cook et al., 1997; Schroer et al., 1998; Weidmer-Mikhail et al., 1998; Wolpert et al., 2000). Although these children frequently meet diagnostic criteria for ASD, they also exhibit significant cognitive impairment, gross motor delays, hypotonia, epilepsy, and facial dysmorphisms (C. Schanen, personal communication). There are also case reports and small case series of autism co-occurring with many other genetic disorders including neurofibromatosis (Gillberg and Forsell, 1984; Williams and Hersh, 1998), hypomelanosis of Ito (Zappella, 1993), Moebius syndrome (Gillberg and Steffenburg, 1989), Prader-Willi and Angelman syndromes (Steffenburg et al., 1996), Joubert syndrome (Ozonoff et al., 1999), Down syndrome (Fombonne et al., 1997; Kent et al., 1999), Williams syndrome (Reiss et al., 1985), Sotos syndrome (Morrow et al., 1990), muscular dystrophy (Komoto et al., 1984; Zwaigenbaum and Tarnopolsky, 2003), Cowden syndrome (Goffin et al., 2001), phenylketonuria (PKU; Baieli et al., 2003); Smith–Lemli–Opitz syndrome (SLO; Tierney et al., 2001). Autistic symptomatology has also been reported in association with chromosomal anomalies (deletions, translocations, and duplications) on the sex

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 12 Wednesday, February 15, 2006 11:55 AM

12

Understanding Autism: From Basic Neuroscience to Treatment

chromosomes and almost every autosome (see Gillberg; 1998, Lauritsen et al., 1999; Miles et al., 2005; and also Chapter 3 of this volume for reviews). These associations can be important for a variety of reasons: single gene disorders and anomalies may point to autism susceptibility genes; many genetic disorders have implications for genetic counseling in the families; knowledge of the comorbid condition may influence prognosis; and rarely, as in the case of the metabolic disorders, they point to a treatment that could impact the course of the autistic symptomatology.

DIAGNOSIS AND ASSESSMENT An experienced clinician, using standardized methods, can reliably diagnose autism starting at age 2, and sometimes even younger. However, diagnoses, especially of PDDNOS and atypical autism, at age 2 are significantly less stable than they will be at age three. Most children with less certain diagnoses at 2 will go on to have more obvious autistic features, including repetitive behaviors by age 3, but a significant minority seem to truly “grow out of ASD,” and another group may remain just in or outside ASD for many years, as measured by standard instruments or judged by experienced clinicians (Bishop and Norbury, 2002; Lord et al., in press). It has been common wisdom that the symptoms of autism are most clearly recognizable between about 4 and 5 years (Le Couter et al., 1989), but it is not clear if this is still the case when many children, at least in North America and Western Europe, receive diagnoses and begin intervention several years before this. In general, the few intervention studies indicating “recovery” from autism in a significant number of cases have not been replicated (see Dawson and Osterling, 1997; Rogers, 2000). More typical have been gains in IQ, made primarily by the children with the highest IQs at the start (Smith, 1999; Sheinkopf and Siegel, 1998). Specific behavioral interventions have been shown to result in clear improvements in specific behaviors (e.g., Goldstein, 2002; Stahmer et al., 2003; Lord and McGee, 2001), but to date there have not been sufficiently well-controlled studies of comprehensive interventions to compare their effectiveness, determine the “active ingredients” in the treatments that account for improvements, or to look at individual differences in responses to behavioral treatments (see Tsai, 1999 and McDougle, 1997 for a discussion of psychopharmacological treatments). Less classic cases of ASD tend to receive later diagnoses than children with autism (Szatmari et al., 1989). This may not reflect later onset of symptoms as much as delayed recognition that the child has core social communication deficits. Many of these children receive other psychiatric diagnoses such as ADHD or oppositional behavior or emotional disturbance either in preschool or the early school years. Although these children experience considerable difficulty in social situations and in school, their fluent language and sometimes their socially directed but odd behavior confuse diagnosticians and clinicians. From a developmental perspective, their early experiences may be very different from a more cognitively impaired child with autism. The more able child with ASD today will have had much greater opportunities for learning in ordinary school and through exposure to same-age peers and age-appropriate expectations than most children with autism in the past.

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 13 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

13

However, they also may have been teased and rejected in ways that a child with autism in specialized educational settings and who is less aware of social feedback would not have experienced. Even now, despite improved awareness and recognition of ASDs, there are some individuals who do not receive an ASD diagnosis until their teenage or even adult years. Usually these individuals were identified as having other disorders such as ADHD or anxiety disorders as children, but because of relatively strong verbal skills, the possibility of ASD was never raised.

STANDARDIZED DIAGNOSES One of the major changes in the field of ASD in the last 20 years has been international adoption of standard measures used for assignment of diagnostic status. These include a caregiver interview, the Autism Diagnostic Interview-Revised (ADI-R: Le Couteur et al., 2003), which provides information about social reciprocity, communication and repetitive behaviors and interests both currently and in the past, and the Autism Diagnostic Observation Schedule (ADOS: Lord et al., 1999), which offers structured observation in standardized contexts, and codings of social and communicative behavior, carried out by a clinician in a 30- to 45-minute office visit (Lord et al., 2000). The ADOS and ADI-R both use additive models within the three domains (social reciprocity, communication, and repetitive behaviors) to create a diagnostic algorithm. The ADI-R was intended to distinguish between autism and nonspectrum disorders, though recently an algorithm for atypical autism and PDDNOS has been proposed (Risi et al., submitted). Different cutoffs are used in the communication domain for verbal and nonverbal individuals so that these scores must be considered separately. The ADOS is a structured observation that consists of four modules with different sets of materials, tasks, and algorithms. Use of a specific module is determined by the child’s or adult’s expressive language level and chronological age. Algorithms offer classifications of autism or ASD, though many children with clinical diagnoses of ASD will fall in the autism range and vice versa. Neither the ADI-R or ADOS is intended to be used as a simple measure of severity because codes for individual items are ordinal and do not represent equal intervals. Traditionally, scores on both instruments are related in nonlinear ways to chronological age and IQ. Nevertheless, in samples of individuals with ASD who fall within relatively narrow age and IQ ranges, ADOS scores have been shown to be related to several neurobiological features (see Chapter 15 of this volume), including eye-tracking and activation of the fusiform gyrus during face processing (Klin et al., 2002; Schultz et al., 2000). When used together, the ADI-R and ADOS offer sensitivity of 82% and specificity of 86% for autism and 60% sensitivity and 88% specificity for nonautism ASD in children from age 4 through 12, excluding children with profound mental retardation. Using the ADI-R and ADOS, better sensitivity for children who had ASD but did not meet criteria for autism could only be achieved if specificity was sacrificed. It seems likely that a combination of direct observation by a clinician and a caregiver interview results in the most reliable diagnoses (Risi et al., submitted).

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 14 Wednesday, February 15, 2006 11:55 AM

14

Understanding Autism: From Basic Neuroscience to Treatment

Because of the required training and the length of time involved in the administration of the ADI-R, there have been attempts at using alternative instruments. However, replacing the ADI-R with the Social Communication Questionnaire (SCQ; see Rutter et al., 2003), a parent screening questionnaire developed using questions from an earlier form of the ADI-R, resulted in significantly less sensitivity than using the full interview (Corsello et al., submitted). It may be that other combinations of parent report and clinician observation based instruments including the Social Responsiveness Scale (SRS: Constantino et al., 2000), the Communicative Competence Checklist (CCC: Bishop, 2003), and even the Vineland Adaptive Behavior Scales (VABS: Sparrow et al., 2005), may work equally well and be more efficient than the current system; however, these have not yet been well researched. It should also be noted that with briefer scales, opportunities for more detailed phenotypic analyses may be limited or less focused on the core features of autism.

SUMMARY AND CONCLUSIONS Altogether, autism spectrum disorders are one of the most reliably diagnosable of childhood onset psychiatric disorders (Volkmar and Rutter, 1995). They are of interest to neuroscientists on many levels, from the quest to find effective treatments or methods of prevention to the search for biological bases of social behavior and early communication. Despite the reliability of the diagnosis, there is still remarkable heterogeneity in clinical presentation. As yet, there are no pathognomonic signs, symptoms, or biomarkers that are universal or specific to autism. Poor social use of eye contact is probably the most frequently identified single behavior in ASD, and it is not always present, nor does it only occur in individuals with ASD. Because there seem likely to be a number of neurobiological pathways leading to ASD, there has been interest in identifying more homogeneous subtypes that may be associated with genetic or other patterns. This approach is far less straightforward than one might think because of the interrelatedness of various aspects of the disorder — for example: mental retardation, language level, social skills, ordinary maturity, and experience. Studies of very young children offer the purest opportunity to observe the disorder “unfolding,” but are limited by the kind of neurobiological and specific cognitive measures appropriate at young ages. Studies of siblings of children with autism offer the opportunity to identify children at younger ages and earlier points in development but may not be representative of other children with ASDs (Szatmari et al., 2000). Adaptations of neuropsychological measures, such as eye tracking and eventrelated potentials (ERPs) (see Chapter 17 of this volume), are also a potential source of information. Older, higher functioning children and adults are more easily studied using traditional neuropsychological and neurophysiological measures, but are more complicated diagnostically. These measures in older children and adults also seem more likely to have been affected by different life experiences in terms of time and quality of social interactions than they would be in very young children (Schultz et al., 2000). Thus, interpreting what is the cause or the result of unusual behaviors and preferences becomes more difficult to assess. Larger samples are now available, allowing researchers to look more closely at individual differences in phenotype and to provide more specific conceptualizations of the nature of social deficits and repetitive

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 15 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

15

behaviors and interests. Developing psychometrically valid continuous measures of severity in ASD that are not confounded by age and IQ will also provide the opportunity for more interesting studies of links between genotype and phenotype.

ACKNOWLEDGMENTS This work was supported by grants NIMH R01 MH066496 and R01 MH46865 to Dr. Lord and was carried out as part of the NICHD/NIDCD Collaborative Programs for Excellence in Autism (CPEA). Dr. Spence would like to acknowledge previous grant support from the M.I.N.D. Institute and current support from the NIMH (MH64547 and MH068172).

REFERENCES American Psychiatric Association (2000). Diagnostic and Statistical Manual of Mental Disorders-Text Revision (4th ed., text revision). Washington, D.C., American Psychiatric Association. Aylward, E.H., Minshew, N.J., Field, K., Sparks, B.F., and Singh, N. (2002). Effects of age on brain volume and head circumference in autism. Neurology, 59(2), 175–183. Baieli, S., Pavone, L., Meli, C., Fiumara, A., and Coleman, M. (2003). Autism and phenylketonuria. Journal of Autism and Developmental Disorders, 33(2), 201–204. Bailey, A., Bolton, P., Butler, L., Le Couteur, A., Murphy, M., Scott, S. et al. (1993). Prevalence of the fragile X anomaly amongst autistic twins and singletons. Journal of Child Psychology and Psychiatry, 34(5), 673–688. Bailey, A., Luthert, P., Dean, A., Harding, B., Janota, I., Montgomery, M. et al. (1998). A clinicopathological study of autism. Brain, 121(Pt. 5), 889–905. Ballaban-Gil, K. and Tuchman, R. (2000). Epilepsy and epileptiform EEG: association with autism and language disorders. Mental Retardation and Developmental Disability Research Revue, 6(4), 300–308. Baranek, G.T. (1999). Autism during infancy: a retrospective video analysis of sensory-motor and social behaviors at 9–12 months of age. Journal of Autism and Developmental Disorders 29(3): 213–224. Baron-Cohen, S. and Howlin, P. (1994). The theory of mind deficit in autism: some questions for teaching and diagnosis. In Baron-Cohen, S., Tager-Flusberg, H., and Cohen, D. (Eds.), Understanding Other Minds: Perspectives from Autism (pp. 466–480). New York: Oxford University Press. Baron-Cohen, S., Scahill, V.L., Izaguirre, J., Hornsey, H., and Robertson, M.M. (1999). The prevalence of Gilles de la Tourette syndrome in children and adolescents with autism: a large scale study. Psychological Medicine, 29(5), 1151–1159. Barton, M. and Volkmar, F. (1998). How commonly are known medical conditions associated with autism? Journal of Autism and Developmental Disorders, 28(4), 273–278. Bishop, D.V. and Norbury, C. F. (2002). Exploring the borderlands of autistic disorder and specific language impairment: a study using standardised diagnostic instruments. Journal of Child Psychology and Psychiatry and Allied Disciplines, 43(7), 917–929. Bishop, D.V. (2003). Children’s Communication Checklist (CCC-2), 2nd ed., London, Psychological Corporation. Bolton, P.F., Roobol, M., Allsopp, L., and Pickles, A. (2001). Association between idiopathic infantile macrocephaly and autism spectrum disorders. Lancet, 358(9283), 726–727.

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 16 Wednesday, February 15, 2006 11:55 AM

16

Understanding Autism: From Basic Neuroscience to Treatment

Boucher, J. (1981). Memory for recent events in autistic children. Journal of Autism and Developmental Disorders, 11(3), 293–301. Brown, R., Hobson, R.P., Lee, A., and Stevenson, J. (1997). Are there autistic-like features in congenitally blind children? Journal of Child Psychology and Psychiatry, 38(6), 693–703. Brown, W.T., Nolin, S.L., Dobkin, C.S., Houck, G.S., Glicksman, A., Ding, X.D. et al. (2002). Frequency of fragile X in multiplex autism: testing AGRE families. The American Journal of Human Genetics, 71(4), 347. Cantor, R., Chen, G., Pajukanta, P., and Lange, K. (2005). Association testing in a linked region using large pedigrees. The American Journal of Human Genetics, 76, 538–542. Carney, R.M., Wolpert, C.M., Ravan, S.A., Shahbazian, M., Ashley-Koch, A., Cuccaro, M.L. et al. (2003). Identification of MeCP2 mutations in a series of females with autistic disorder. Pediatric Neurology, 28(3), 205–211. Carvill, S. (2001). Sensory impairments, intellectual disability and psychiatry. Journal of Intellectual Disability Research, 45(Pt. 6), 467–483. Charman, T. and Baird, G. (2002). Practitioner review: diagnosis of autism spectrum disorder in 2- and 3-year-old children. Journal of Child Psychology and Psychiatry, 43(3), 289–305. Charman, T., Taylor, E., Drew, A., Cockerill, H., Brown, J.A., and Baird, G. (2005). Outcome at 7 years of children diagnosed with autism at age 2: predictive validity of assessments conducted at 2 and 3 years of age and pattern of symptom change over time. Journal of Child Psychology and Psychiatry, 46(5), 500–513. Constantino, J.N., Przybeck, T., Friesen, D., and Todd, R.D. (2000). Reciprocal social behavior in children with and without pervasive developmental disorders. Journal of Developmental and Behavioral Pediatrics, 21(1), 2–11. Cook, E.H., Jr., Courchesne, R.Y., Cox, N.J., Lord, C., Gonen, D., Guter, S.J. et al. (1998). Linkage-disequilibrium mapping of autistic disorder, with 15q11–13 markers. The American Journal of Human Genetics, 62(5), 1077–1083. Cook, E.H., Jr., Lindgren, V., Leventhal, B.L., Courchesne, R., Lincoln, A., Shulman, C. et al. (1997). Autism or atypical autism in maternally but not paternally derived proximal 15q duplication. The American Journal of Human Genetics, 60(4), 928–934. Corsello, C., Lord, C., Cook, E., Leventhal, B., and Pickles, A. (submitted). The Social Communication Questionnaire (SCQ) Revisited. Courchesne, E., Carper, R., and Akshoomoff, N. (2003). Evidence of brain overgrowth in the first year of life in autism. Journal of the American Medical Association, 290(3), 337–344. Courchesne, E., Karns, C.M., Davis, H.R., Ziccardi, R., Carper, R.A., Tigue, Z.D. et al. (2001). Unusual brain growth patterns in early life in patients with autistic disorder: an MRI study. Neurology, 57(2), 245–254. Curatolo, P., Porfirio, M.C., Manzi, B., and Seri, S. (2004). Autism in tuberous sclerosis. European Journal of Paediatric Neurology, 8(6), 327–332. Davidovitch, M., Patterson, B., and Gartside, P. (1996). Head circumference measurements in children with autism. Journal of Child Neurology, 11(5), 389–393. Dawson, G. and Osterling, J. (1997). Early intervention in autism. In Guralnick, M.J. (Ed.), The Effectiveness of Early Intervention (pp. 307–326). Baltimore, MD: Paul H. Brookes. Fidler, D.J., Bailey, J.N., and Smalley, S.L. (2000). Macrocephaly in autism and other pervasive developmental disorders. Developmental Medicine and Child Neurology, 42(11), 737–740. Filipek, P.A., Accardo, P.J., Ashwal, S., Baranek, G.T., Cook, E.H., Jr., Dawson, G. et al. (2000). Practice parameter: screening and diagnosis of autism: report of the Quality Standards Subcommittee of the American Academy of Neurology and the Child Neurology Society. Neurology, 55(4), 468–479.

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 17 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

17

Fombonne, E. (in press). Epidemiological surveys of autism and other pervasive developmental disorders: an update. Journal of Autism and Developmental Disorders. Fombonne, E., Du Mazaubrun, C., Cans, C., and Grandjean, H. (1997). Autism and associated medical disorders in a French epidemiological survey. Journal of the American Academy on Child and Adolescent Psychiatry, 36(11), 1561–1569. Fombonne, E., Roge, B., Claverie, J., Courty, S., and Fremolle, J. (1999). Microcephaly and macrocephaly in autism. Journal of Autism and Developmental Disorders, 29(2), 113–119. Frith, C. (2004). Is autism a disconnection disorder? Lancet Neurology, 3(10), 577–577. Gernsbacher, M.A., Dawson, M., and Goldsmith, H.H. (2005) Three reasons not to believe in an autism epidemic. Current Directions in Psychological Science, 14(2), 55–58. Ghaziuddin, M. and Greden, J. (1998). Depression in children with autism/pervasive developmental disorders: a case-control family history study. Journal of Autism and Developmental Disorders, 28(2), 111–115. Ghaziuddin, M. and Butler, E. (1998). Clumsiness in autism and Asperger syndrome: a further report. Journal of Intellectual Disability Research, 42(Pt. 1), 43–48. Ghaziuddin, M., Ghaziuddin, N., and Greden, J. (2002). Depression in persons with autism: implications for research and clinical care. Journal of Autism and Development Disorders, 32(4), 299–306. Ghaziuddin, M., Weidmer-Mikhail, E., and Ghaziuddin, N. (1998). Comorbidity of Asperger syndrome: a preliminary report. Journal of Intellectual Disability Research, 42(Pt. 4), 279–283. Ghaziuddin, M., Zaccagnini, J., Tsai, L., and Elardo, S. (1999). Is megalencephaly specific to autism? Journal of Intellectual Disability Research, 43(Pt. 4), 279–282. Gillberg, C. (1998). Chromosomal disorders and autism. Journal of Autism and Developmental Disorders, 28(5), 415–425. Gillberg, C. and Forsell, C. (1984). Childhood psychosis and neurofibromatosis — more than a coincidence? Journal of Autism and Developmental Disorders, 14(1), 1–8. Gillberg, C. and Steffenburg, S. (1989). Autistic behavior in Moebius syndrome. Acta Paediatrica Scandinavica, 78(2), 314–316. Gillberg, I.C., Gillberg, C., and Ahlsen, G. (1994). Autistic behaviour and attention deficits in tuberous sclerosis: a population-based study. Developmental Medicine and Child Neurology, 36(1), 50–56. Goffin, A., Hoefsloot, L.H., Bosgoed, E., Swillen, A., and Fryns, J.P. (2001). PTEN mutation in a family with Cowden syndrome and autism. The American Journal of Human Genetics, 105(6), 521–524. Goldstein, H. (2002). Communication intervention for children with autism: a review of treatment efficacy. Journal of Autism & Developmental Disorders, 32(5), 373–396. Goldstein, S. and Schwebach, A.J. (2004). The comorbidity of Pervasive Developmental Disorder and Attention Deficit Hyperactivity Disorder: results of a retrospective chart review. J Autism Dev Disord, 34(3), 329–339. Hagberg, B. and Witt-Engerstrom, I. (1986). Rett syndrome: a suggested staging system for describing impairment profile with increasing age towards adolescence. American Journal of Medical Genetics — Supplement, 1, 47–59. Hobson, R.P. and Bishop, M. (2003). The pathogenesis of autism: insights from congenital blindness. Philosophical Transactions of the Royal Society of London: Biological Sciences, 358(1430), 335–344. Howlin, P. (2000). Outcome in adult life for more able individuals with autism or Asperger syndrome. Autism: Special Issue: Asperger Syndrome, 4(1), 63–83. Howlin, P., Goode, S., Hutton, J., and Rutter, M. (2004). Adult outcome for children with autism. Journal of Child Psychology and Psychiatry, 45(2), 212–229.

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 18 Wednesday, February 15, 2006 11:55 AM

18

Understanding Autism: From Basic Neuroscience to Treatment

Jones, V. and Prior, M. (1985). Motor imitation abilities and neurological signs in autistic children. Journal of Autism and Developmental Disorders, 15(1), 37–46. Joseph, R.M., Tager-Flusberg, H., and Lord, C. (2002). Cognitive profiles and socialcommunicative functioning in children with autism spectrum disorder. Journal of Child Psychology and Psychiatry and Allied Disciplines, 43(6), 807–821. Jure, R., Rapin, I., and Tuchman, R.F. (1991). Hearing-impaired autistic children. Developmental Medicine and Child Neurology, 33(12), 1062–1072. Kanner, L. (1943). Autistic disturbances of affective contact. Nervous Child, 10, 217–250. Kawasaki, Y., Yokota, K., Shinomiya, M., Shimizu, Y., and Niwa, S. (1997). Brief report: electroencephalographic paroxysmal activities in the frontal area emerged in middle childhood and during adolescence in a follow-up study of autism. Journal of Autism and Developmental Disorders, 27(5), 605–620. Kemper, T.L. and Bauman, M.L. (1993). The contribution of neuropathologic studies to the understanding of autism. Neurologic Clinics, 11(1), 175–187. Kent, L., Evans, J., Paul, M., and Sharp, M. (1999). Comorbidity of autistic spectrum disorders in children with Down syndrome. Developmental Medicine and Child Neurology, 41(3), 153–158. Kielinen, M., Rantala, H., Timonen, E., Linna, S.L., and Moilanen, I. (2004). Associated medical disorders and disabilities in children with autistic disorder: a populationbased study. Autism, 8(1), 49–60. Klauck, S.M., Munstermann, E., Bieber-Martig, B., Ruhl, D., Lisch, S., Schmotzer, G. et al. (1997). Molecular genetic analysis of the FMR-1 gene in a large collection of autistic patients. Human Genetics, 100(2), 224–229. Klin, A., Jones, W., Schultz, R.T., Volkmar, F., and Cohen, D.J. (2002) Visual fixation patterns during viewing of naturalistic social situations as predictors of social competence in individuals with autism. Archives of General Psychiatry, 59, 809–816. Komoto, J., Usui, S., Otsuki, S., and Terao, A. (1984). Infantile autism and Duchenne muscular dystrophy. Journal of Autism and Developmental Disorders, 14(2), 191–195. Konstantareas, M.M. and Hewitt, T. (2001). Autistic disorder and schizophrenia: diagnostic overlaps. Journal of Autism and Developmental Disorders, 31(1), 19–28. Lainhart, J.E., Piven, J., Wzorek, M., Landa, R., Santangelo, S.L., Coon, H. et al. (1997). Macrocephaly in children and adults with autism. Journal of the American Academy of Child and Adolescent Psychiatry, 36(2), 282–290. Lam, C.W., Yeung, W.L., Ko, C.H., Poon, P.M., Tong, S.F., Chan, K.Y. et al. (2000). Spectrum of mutations in the MECP2 gene in patients with infantile autism and Rett syndrome. Medical Genetics, 37(12), E41. Lamb, J.A., Barnby, G., Bonora, E., Sykes, N., Bacchelli, E., Blasi, F. et al. (2005). Analysis of IMGSAC autism susceptibility loci: evidence for sex limited and parent of origin specific effects. Journal of Medical Genetics, 42(2), 132–137. Lauritsen, M., Mors, O., Mortensen, P.B., and Ewald, H. (1999). Infantile autism and associated autosomal chromosome abnormalities: a register-based study and a literature survey. Journal of Child Psychology and Psychiatry, 40(3), 335–345. Leckman, J.F., Grice, D.E., Boardman, J., and Zhang, H. (1997). Symptoms of obsessivecompulsive disorder. American Journal of Psychiatry, 154(7), 911–917. Le Couteur, A., Rutter, M., Lord, C., Rios, P., Robertson, S., Holdgrafer, M. et al. (1989). Autism diagnostic interview: a standardized investigator-based instrument. Journal of Autism and Developmental Disorders, 19(3), 363–387. Le Couteur, A., Lord, C., and Rutter, M. (2003). The Autism Diagnostic Interview — Revised (ADI — R). Los Angeles, CA: Western Psychological Services.

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 19 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

19

Lobo-Menendez, F., Sossey-Alaoui, K., Bell, J.M., Copeland-Yates, S.A., Plank, S.M., Sanford, S.O. et al. (2003). Absence of MeCP2 mutations in patients from the South Carolina autism project. American Journal of Medical Genetics, 117B(1), 97–101. Lord, C. and Corsello, C. (2005). Diagnostic instruments in autism spectrum disorders. In Volkmar, F.R., Klin, A., and Paul, R. (Eds.), Handbook of Autism and Pervasive Developmental Disorders, 3rd ed. Hoboken, NJ: John Wiley & Sons. Lord, C. and McGee, J.P. (2001). Educating Children with Autism. Washington, D.C.: National Academy Press. Lord, C. and Richler, J. (in press). Early diagnosis of children with autistic spectrum disorders. In T. Charman & W. Stone (Eds.) Early Social Communication in Autism Spectrum Disorders. New York: Guilford. Lord, C., Risi, S., DiLavore, P., Shulman, C., Thurm, A., Pickles, A. (in press), Autism from two to nine, Archives of General Psychiatry. Lord, C., Risi, S., Lambrecht, L., Cook, E.H., Leventhal, B.L., DiLavore, P.C. et al. (2000). The autism diagnostic observation schedule-generic: a standard measure of social and communication deficits associated with the spectrum of autism. Journal of Autism and Developmental Disorders, 30(3), 205–223. Lord, C., Rutter, M., DiLavore, P., and Risi, S. (1999). Autism Diagnostic Observation Schedule-Generic. Los Angeles, CA: Western Psychological Services. Lord, C., Rutter, M., and Le Couteur, A. (1994). Autism diagnostic interview-revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. Journal of Autism and Developmental Disorders, 24(5), 659–685. Lord, C., Schopler, E., and Revicki, D. (1982). Sex differences in autism. Journal of Autism and Developmental Disorders, 12, 317–330. Lord, C., Shulman, C., and DiLavore, P. (2004). Regression and word loss in autistic spectrum disorders. Journal of Child Psychology and Psychiatry, 45(5), 936–955. Luyster, R., Richler, J., Risi, S., Hsu, W.L., Dawson, G., Bernier, R. et al. (2005). Early regression in social communication in autistic spectrum disorders: a CPEA study. Developmental Neuropsychology, 27(3), 311–336. Mahoney, W.J., Szatmari, P., MacLean, J.E., Bryson, S.E., Bartolucci, G., Walter, S.D. et al. (1998). Reliability and accuracy of differentiating pervasive developmental disorder subtypes. Journal of the American Academy of Child & Adolescent Psychiatry, 37(3), 278–285. Mayes, S.D. and Calhoun, S.L. (2003). Ability profiles in children with autism: influence of age and IQ. Autism, 7(1), 65–80. McDougle, C.J. (1997). Psychopharmacology. In Cohen, D.J. and Volkmar, F.R. (Eds.), Handbook of Autism and Pervasive Developmental Disorder, 2nd ed. (pp. 707–729). New York: John Wiley & Sons. McGovern, C.W. and Sigman, M. (2005). Continuity and change from early childhood to adolescence in autism. Journal of Child Psychology and Psychiatry, 46(4), 401–408. Miles, J.H., Hadden, L.L., Takahashi, T.N., and Hillman, R.E. (2000). Head circumference is an independent clinical finding associated with autism. American Journal of Medical Genetics, 95(4), 339–350. Miles, J., Takahashi, T., Bagby, S., Sahota, P., Vaslow, D., Wang C. et al. (2005). Essential versus complex autism: definition of fundamental prognostic subtypes. American Journal of Medical Genetics, 135(2), 171–180. Minshew, N.J., Sung, K., Jones, B.L., and Furman, J.M. (2004). Underdevelopment of the postural control system in autism. Neurology, 63(11), 2056–2061.

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 20 Wednesday, February 15, 2006 11:55 AM

20

Understanding Autism: From Basic Neuroscience to Treatment

Molloy, C.A., Dawson, G., Dunn, M., Hyman, S.L., McMahon, W.M., Minshew, N. et al. (in press). Familial autoimmune thyroid disease as a risk factor for regression in children with autism spectrum disorder: a CPEA study. Journal of Autism and Developmental Disorders. Moore, V. and Goodson, S. (2003). How well does early diagnosis of autism stand the test of time? Follow-up study of children assessed for autism at age 2 and development of an early diagnostic service. Autism, 7(1), 47–63. Morrow, J.D., Whitman, B.Y., and Accardo, P.J. (1990). Autistic disorder in Sotos syndrome: a case report. European Journal of Pediatrics, 149(8), 567–569. Mount, R.H., Charman, T., Hastings, R.P., Reilly, S., and Cass, H. (2003). Features of autism in Rett syndrome and severe mental retardation. Journal of Autism and Developmental Disorders, 33(4), 435–442. Mundy, P. and Sigman, M. (1989). The theoretical implications of joint-attention deficits in autism. Development and Psychopathology, 1(3), 173–183. Muris, P., Steerneman, P., Merckelbach, H., Holdrinet, I., and Meesters, C. (1998). Comorbid anxiety symptoms in children with pervasive developmental disorders. Journal of Anxiety Disorders, 12(4), 387–393. Osterling, J. and Dawson, G. (1994). Early recognition of children with autism: a study of first birthday home videotapes. Journal of Autism and Developmental Disorders, 24(3), 247–257. Ozonoff, S. (1995). Executive function in autism. In Schopler, E. and Mesibov, G.B. (Ed.), Learning and Cognition in Autism (pp. 99–219). New York: Plenum Press. Ozonoff, S., Strayer, D.L., McMahon, W.M., and Filoux, F. (1994). Executive function abilities in autism and Tourette Syndrome: an information processing approach. Journal of Child Psychology and Psychiatry, 35, 1015–1032. Ozonoff, S., Williams, B.J., Gale, S., and Miller, J.N. (1999). Autism and autistic behavior in Joubert syndrome. Journal of Child Neurology, 14(10), 636–641. Pavone, P., Incorpora, G., Fiumara, A., Parano, E., Trifiletti, R.R., and Ruggieri, M. (2004). Epilepsy is not a prominent feature of primary autism. Neuropediatrics, 35(4), 207–210. Philofsky, A., Hepburn, S.L., Hayes, A., Hagerman, R., and Rogers, S.J. (2004). Linguistic and cognitive functioning and autism symptoms in young children with fragile X syndrome. American Journal of Mental Retardation, 109(3), 208–218. Piven, J., Arndt, S., Bailey, J., and Andreasen, N. (1996). Regional brain enlargement in autism: a magnetic resonance imaging study. Journal of the American Academy of Child and Adolescent Psychiatry, 35(4), 530–536. Piven, J., Arndt, S., Bailey, J., Havercamp, S., Andreasen, N.C., and Palmer, P. (1995). An MRI study of brain size in autism. American Journal of Psychiatry, 152(8), 1145–1149. Piven, J., Gayle, J., Landa, R., Wzorek, M., and Folstein, S. (1991). The prevalence of fragile X in a sample of autistic individuals diagnosed using a standardized interview. Journal of the American Academy of Child and Adolescent Psychiatry, 30(5), 825–830. Piven, J., Palmer, P., Jacobi, D., Childress, D. et al. (1997). Broader autism phenotype: evidence from a family history study of multiple-incidence autism families. American Journal of Psychiatry, 154(2), 185–190. Reiss, A.L., Feinstein, C., Rosenbaum, K.N., Borengasser-Caruso, M.A., and Goldsmith, B.M. (1985). Autism associated with Williams syndrome. Journal of Pediatrics, 106(2), 247–249. Renner, P., Klinger, L.G., and Klinger, M.R. (2000). Implicit and explicit memory in autism: is autism an amnesic disorder? Journal of Autism and Developmental Disorders, 30(1), 3–14.

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 21 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

21

Richler, J., Bishop, S.L., and Lord, C. (in press). Repetitive behaviors and intelligence in autism. Journal of Child Neuropsychology. Richler, J., Luyster, R., Risi, S., Hsu, W.L., Dawson, G., Bernier, R. et al. (in press). Is there a regressive phenotype of Autism Spectrum Disorder associated with the measles-mumpsrubella vaccine? A CPEA study. Journal of Autism and Developmental Disorders. Risi, S., Lord, C., Corsello, C., Chrysler, C., Szatmari, P., Cook, E.H., Leventhal, B.L., Pickles, A. (submitted). Combining information from multiple sources in the diagnosis of autism spectrum disorders. Journal of the American Academy of Child and Adolescent Psychiatry. Rogers, S.J. (2000). Interventions that facilitate socialization in children with autism. Journal of Autism and Developmental Disorders: Special Issue: Treatments for People with Autism and other Pervasive Developmental Disorders: Research Perspectives, 30(5), 399–409. Rogers, S.J. and DiLalla, D.L. (1990). Age of symptom onset in young children with pervasive developmental disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 29(6), 863–872. Rogers, S.J. and Pennington, B.F. (1991). A theoretical approach to the deficits in infantile autism. Development and Psychopathology, 3, 137–162. Rogers, S.J., Hepburn, S., and Wehner, E. (2003). Parent reports of sensory symptoms in toddlers with autism and those with other developmental disorders. Journal of Autism and Developmental Disorders, 33(6), 631–642. Rossi, P.G., Parmeggiani, A., Bach, V., Santucci, M., and Visconti, P. (1995). EEG features and epilepsy in patients with autism. Brain Development, 17(3), 169–174. Russell, A. J., Mataix-Cols, D., Anson, M., and Murphy, D.G. (2005). Obsessions and compulsions in Asperger syndrome and high-functioning autism. British Journal of Psychiatry, 186, 525–528. Rutter, M., Bailey, A., Bolton, P., and Le Couteur, A. (1994). Autism and known medical conditions: myth and substance. Journal of Child Psychology and Psychiatry, 35(2), 311–322. Rutter, M., Bailey, A., Lord, C., and Berument, S.K. (2003). Social Communication Questionnaire. Los Angeles, CA. Schroer, R.J., Phelan, M.C., Michaelis, R.C., Crawford, E.C., Skinner, S.A., Cuccaro, M. et al. (1998). Autism and maternally derived aberrations of chromosome 15q. American Journal of Medical Genetics, 76(4), 327–336. Schultz, R.T., Gauthier, I., Klin, A., Fulbright, R.K., Anderson, A.W., Volkmar, F. et al. (2000). Abnormal ventral temporal cortical activity during face discrimination among individuals with autism and Asperger syndrome. Archives of General Psychiatry, 57(4), 331–340. Shao, Y.J., Cuccaro, M.L., Hauser, E.R., Raiford, K.L., Menold, M.M., Wolpert, C.M. et al. (2003). Fine mapping of Autistic disorder to chromosome 15q11-q13 by use of phenotypic subtypes. American Journal of Human Genetics, 72(3), 539–548. Sheinkopf, S.J. and Siegel, B. (1998). Home based behavioral treatment of young children with autism. Journal of Autism and Developmental Disorders, 28(1), 15–23. Smalley, S.L. (1998). Autism and tuberous sclerosis. Journal of Autism and Developmental Disorders, 28(5), 407–414. Smith, T. (1999). Outcome of early intervention for children with autism. Clinical Psychology: Science and Practice, 6(1), 33–49. Sparrow, S.S., Cicchetti, D.V., and Balla, D.A. (2005). Vineland Adaptive Behavior Scales (2nd ed.). Circle Pines, MN: AGS Publishing. Stahlberg, O., Soderstrom, H., Rastam, M., and Gillberg, C. (2004). Bipolar disorder, schizophrenia, and other psychotic disorders in adults with childhood onset AD/HD and/or autism spectrum disorders. Journal of Neural Transmission, 111, 891–902

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 22 Wednesday, February 15, 2006 11:55 AM

22

Understanding Autism: From Basic Neuroscience to Treatment

Stahmer, A.C., Ingersoll, B., and Carter, C. (2003). Behavioral approaches to promoting play. Autism, 7(4), 401–413. Steffenburg, S., Gillberg, C.L., Steffenburg, U., and Kyllerman, M. (1996). Autism in Angelman syndrome: a population-based study. Pediatric Neurology, 14(2), 131–136. Stevenson, R.E., Schroer, R.J., Skinner, C., Fender, D., and Simensen, R.J. (1997). Autism and macrocephaly. Lancet, 349(9067), 1744–1745. Stone, W.L., Ousley, O.Y., and Littleford, C. (1997). Motor imitation in young children with autism: what’s the object? Journal of Abnormal Child Psychology, 25, 475–485. Stone, J.L., Merriman, B., Cantor, R.M., Yonan, A.L., Gilliam, T.C., Geschwind, D.H. et al. (2004). Evidence for sex-specific risk alleles in autism spectrum disorder. American Journal of Human Genetics, 75(6), 1117–1123. Szatmari, P., Bartolucci, G., and Bremner, R. (1989). Asperger’s syndrome and autism: comparison of early history and outcome. Developmental Medicine and Child Neurology, 31(6), 709–720. Szatmari, P., MacLean, J.E., Jones, M.B., Bryson, S.E., Zwaigenbaum, L., Bartolucci, G. et al. (2000). The familial aggregation of the lesser variant in biological and nonbiological relatives of PDD probands: a family history study. Journal of Child Psychology and Psychiatry, 41(5), 579–586. Szatmari, P., Bryson, S.E., Boyle, M.H., Streiner, D.L., and Duku, E. (2003). Predictors of outcome among high functioning children with autism and Asperger syndrome. Journal of Child Psychology and Psychiatry, 44(4), 520–528. Tanguay, P.E. (2000). Pervasive developmental disorders: a 10-year review. Journal of the American Academy of Child and Adolescent Psychiatry, 39(9), 1079–1095. Tierney, E., Nwokoro, N.A., Porter, F.D., Freund, L.S., Ghuman, J.K., and Kelley, R.I. (2001). Behavior phenotype in the RSH/Smith-Lemli-Opitz syndrome. American Journal of Medical Genetics, 98(2), 191–200. Tsai, L.Y. (1999). Psychopharmacology in autism. Psychosomatic Medicine, 61(5), 651–665. Tuchman, R., Jayakar, P., Yaylali, I., and Villalobos, R. (1997). Seizures and EEG findings in children with autism spectrum disorders. CNS Spectrums, 3, 61–70. Tuchman, R. and Rapin, I. (2002). Epilepsy in autism. Lancet Neurology, 1(6), 352–358. Tuchman, R.F. (1994). Epilepsy, language, and behavior: clinical models in childhood. Journal of Child Neurology, 9(1), 95–102. Tuchman, R.F., Rapin, I., and Shinnar, S. (1991). Autistic and dysphasic children. II: Epilepsy [published erratum appears in Pediatrics, 90(2 Pt. 1): 264, August 1992]. Pediatrics, 88(6), 1219–1225. van der Gaag, R.J., Caplan, R., van Engeland, H., Loman, F., and Buitelaar, J.K. (2005). A controlled study of formal thought disorder in children with autism and multiple complex developmental disorders. J Child Adolesc Psychopharmacol, 15(3), 465–476. Vilensky, J.A., Damasio, A.R., and Maurer, R.G. (1981). Gait disturbances in patients with autistic behavior: a preliminary study. Archives of Neurology, 38(10), 646–649. Volkmar, F.R. and Cohen, D.J. (1991). Comorbid association of autism and schizophrenia. American Journal of Psychiatry, 148(12), 1705–1707. Volkmar, F.R. and Rutter, M. (1995). Childhood disintegrative disorder: results of the DSMIV autism field trial. Journal of the American Academy of Child and Adolescent Psychiatry, 34(8), 1092–1095. Volkmar, F.R., Paul, R., Klin, A., and Cohen, D. (2005). Handbook of Autism and Pervasive Developmental Disorders. Hoboken, NJ: John Wiley& Sons.

© 2006 by Taylor & Francis Group, LLC

2732_C001.fm Page 23 Wednesday, February 15, 2006 11:55 AM

Autism Spectrum Disorders: Phenotype and Diagnosis

23

Vourc’h, P., Bienvenu, T., Beldjord, C., Chelly, J., Barthelemy, C., Muh, J.P. et al. (2001). No mutations in the coding region of the Rett syndrome gene MECP2 in 59 autistic patients. European Journal of Human Genetics, 9(7), 556–558. Wassink, T.H., Piven, J., and Patil, S.R. (2001). Chromosomal abnormalities in a clinic sample of individuals with autistic disorder. Psychiatric Genetics, 11(2), 57–63. Weidmer-Mikhail, E., Sheldon, S., and Ghaziuddin, M. (1998). Chromosomes in autism and related pervasive developmental disorders: a cytogenetic study. Journal of Intellectual Disability Research, 42(Pt. 1), 8–12. Williams, P.G. and Hersh, J.H. (1998). Brief report: the association of neurofibromatosis type 1 and autism. Journal of Autism and Developmental Disorders, 28(6), 567–571. Wing, L. (2005). Problems of Categorical Classification Systems. In Volkmar, F.R., Klin, A., and Paul, R. (Eds.), Handbook of Autism and Pervasive Developmental Disorders, 3rd ed. Hoboken, NJ: John Wiley & Sons. Wolpert, C.M., Menold, M.M., Bass, M.P., Qumsiyeh, M.B., Donnelly, S.L., Ravan, S.A. et al. (2000). Three probands with autistic disorder and isodicentric chromosome 15. American Journal of Medical Genetics, 96(3), 365–372. Woodhouse, W., Bailey, A., Rutter, M., Bolton, P., Baird, G., and Le Couteur, A. (1996). Head circumference in autism and other pervasive developmental disorders. Journal of Child Psychology and Psychiatry, 37(6), 665–671. World Health Organization (1992). The ICD 10 Classification of Mental and Behavioral Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva, Switzerland. Wozniak, J., Biederman, J., Faraone, S.V., Frazier, J., Kim, J. et al. (1997). Mania in children with pervasive developmental disorder revisited. Journal of the American Academy of Child and Adolescent Psychiatry, 36, 1552–9; discussion 9–60. Zappella, M. (1993). Autism and hypomelanosis of Ito in twins. Developmental Medicine and Child Neurology, 35(9), 826–832. Zwaigenbaum, L., Bryson, S., Rogers, T., Roberts, W., Brian, J., and Szatmari, P. (2005). Behavioral manifestations of autism in the first year of life. International Journal of Developmental Neuroscience, 23(2–3), 143–152. Zwaigenbaum, L. and Tarnopolsky, M. (2003). Two children with muscular dystrophies ascertained due to referral for diagnosis of autism. Journal of Autism and Developmental Disorders, 33(2), 193–199.

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 25 Wednesday, February 15, 2006 4:33 PM

2

Past and Future Perspectives on Autism Epidemiology Eric Fombonne

CONTENTS Introduction..............................................................................................................26 Selection of Studies .................................................................................................26 Survey Descriptions.................................................................................................26 Study Designs ..........................................................................................................31 Characteristics of Autistic Samples.........................................................................31 Prevalence Estimations ............................................................................................32 Autistic Disorder ..............................................................................................32 Unspecified PDDs — PDDNOS (Pervasive Development Disorder Not Otherwise Specified) ..................................................................32 Asperger Syndrome (AS) and Childhood Disintegrative Disorder (CDD) ................................................................................................33 Prevalence for Combined PDDs ......................................................................36 Time Trends .............................................................................................................36 Referral Statistics..............................................................................................38 Comparison of Cross-Sectional Epidemiological Surveys..............................39 Repeat Surveys in Defined Geographical Areas..............................................40 Successive Birth Cohorts..................................................................................41 Incidence Studies..............................................................................................42 Conclusion on Time Trends .............................................................................42 Correlates of Autism................................................................................................42 Associated Medical Conditions........................................................................42 Autism, Race, and Immigrant Status ...............................................................42 Autism and Social Class ..................................................................................43 Cluster Reports ........................................................................................................43 Conclusion ...............................................................................................................44 References................................................................................................................45

25

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 26 Wednesday, February 15, 2006 4:33 PM

26

Understanding Autism: From Basic Neuroscience to Treatment

INTRODUCTION The aims of this chapter are to provide an up-to-date review of the methodological features and substantive results of published epidemiological surveys. This chapter updates our previous review1 with the inclusion of eight new studies made available since then. The specific questions addressed in this chapter are the following: 1. 2. 3. 4.

What is the range of prevalence estimates for autism and related disorders? Is the incidence of autism increasing? What are the correlates of autistic spectrum disorders? What is the role, if any, of cluster reports in causal investigations of autism?

SELECTION OF STUDIES The studies were identified through systematic searches of the major scientific literature databases (MEDLINE, PsycINFO) and from prior reviews.1,2 Only studies published in the English language were included. Surveys which relied only on a questionnaire-based approach to define caseness were excluded as the validity of the diagnosis is unsatisfactory in these studies. Overall, 43 studies published between 1966 and 2004 were selected that surveyed pervasive development disorders (PDDs) in clearly demarcated nonoverlapping samples. Of these, 37 studies provided information on rates of autistic disorder*, 3 studies only provided estimates on all PDDs combined, and 3 studies provided data only on high-functioning PDDs.

SURVEY DESCRIPTIONS Surveys were conducted in 14 countries and half of the results have been published since 1997. Details of the precise socio-demographic composition and economical activities of the area surveyed in each study were generally lacking; most studies were, however, conducted in predominantly urban or mixed areas, with only 2 (studies 6 and 11) surveys carried out in predominantly rural areas. The proportion of children from immigrant families was generally not available and was very low in 5 surveyed populations (studies 11, 12, 19, 23, and 26). Only in studies 4, 34, and 38 was there a substantial minority of children with either an immigrant or a different ethnic background living in the area. The age range of the population included in the surveys is spread from birth to early adult life, with an overall median age of 8.0. Similarly, in 39 studies, there is huge variation in the size of the population surveyed, with a median population size of 63,860 subjects (mean = 255,000) and about half of the studies relying on targeted populations ranging in size from 16,000 to 167,000.

* Studies are referred in the text by the numbers used in Table 2.1 to index each of them.

© 2006 by Taylor & Francis Group, LLC

Year of Publication

Authors

Country

Area

Age

Number of Subjects with Autism

Diagnostic Criteria

Percentage with Normal IQ

1

1966

Lotter9

U.K.

Middlesex

78,000

8–10

32

Rating scale

2

1970

Brask10

Denmark

Aarhus County

46,500

2–14

20

Clinical

—

3

1970

Treffert11

U.S.

Wisconsin

899,750

3–12

69

Kanner

—

4

1976

Wing et al.12

U.K.

Camberwell

25,000

5–14

17a

30

5

1982

Japan

Fukushima-Ken 609,848

0–18

142

6

1983

Sweden

0–20

39

1984

Ireland

County of Västerbotten East

69,000

7

65,000

8–10

28

8

1986

Germany

West Berlin

279,616

0–14

52

Kanner’s criteria Rutter’s criteria

9

1987

Hoshino et al.13 Bohman et al.14 McCarthy et al.15 Steinhausen et al.16 Burd et al.17

24-Items rating scale of Lotter Kanner’s criteria Rutter’s criteria

U.S.

North Dakota

180,986

2–18

59

DSM-III

—

10

1987

Matsuishi et al.18

Japan

Kurume City

32,834

4–12

51

DSM-III

—

15.6

— 20.5 — 55.8

Gender Ratio (M:F) 2.6 (23/9) 1.4 (12/7) 3.06 (52/17) 16 (16/1) 9.9 (129/13) 1.6 (24/15) 1.33 (16/12) 2.25 (36/16) 2.7 (43/16) 4.7 (42/9)

Prevalence Rate/ 10,000

95% CI

4.1

2.7 ; 5.5

4.3

2.4 ; 6.2

0.7

0.6 ; 0.9

4.82

2.1 ; 7.5

2.33

1.9 ; 2.7

5.6

3.9 ; 7.4

4.3

2.7 ; 5.9

1.9

1.4 ; 2.4

3.26

2.4 ; 4.1

15.5

11.3 ; 19.8

© 2006 by Taylor & Francis Group, LLC

27

(continued)

2732_C002.fm Page 27 Wednesday, February 15, 2006 4:33 PM

Study Number

Size of Target Population

Past and Future Perspectives on Autism Epidemiology

TABLE 2.1 Prevalence Surveys of Autistic Disorder

Study Num- Year of ber Publication 1988

12

1988

13

1989

14

1989

15

Country

Area

Age

Number of Subjects with Autism

Tanoue et al.19 Bryson et al.20

Japan

Southern Ibaraki 95,394

7

Canada

20,800

6–14

21

New RDC

Japan

12,263

3

16

DSM-III

135,180

3–9

61

1989

Sugiyama and Abe 21 Cialdella and Mamelle22 Ritvo et al.23

Part of NovaScotia Nagoya

769,620

3–27

241

16

1991

Gillberg et al.24

Swedend

4–13

74

17

1992

France

18

1992

Fombonne and du Mazaubrun25 Wignyosumarto et al.26

South-West 78,106 Gothenburg + Bohuslän County 4 Régions 274,816 14 départements

© 2006 by Taylor & Francis Group, LLC

France U.S.

Indonesia

1 Département (Rhône) Utah

Yogyakarita (Southeast of Jakarta)

5,120

132

Diagnostic Criteria DSM-III

13.8

11.5 ; 16.2

10.1

5.8 ; 14.4

—

13.0

6.7 ; 19.4

DSM-III-like

—

2.3

4.5

3.4 ; 5.6

DSM-III

34

2.47

2.1 ; 2.8

DSM-III-R

18

3.73 (190/51) 2.7 (54/20)

9.5

7.3 ; 11.6

13.3

2.1 (105/49)

4.9

4.1 ; 5.7

0

2.0 (4/2)

11.7

2.3 ; 21.1

Clinical ICD10-like

4–7

CARS

—

95% CI

4.07 (106/26) 2.5 (15/6) —

9 and 13 154

6

Percentage with Gender Prevalence Normal Ratio Rate/ IQ (M:F) 10,000

23.8

Understanding Autism: From Basic Neuroscience to Treatment

11

Authors

Size of Target Population

2732_C002.fm Page 28 Wednesday, February 15, 2006 4:33 PM

28

TABLE 2.1 (Continued) Prevalence Surveys of Autistic Disorder

20

1997

Fombonne et al.28 France

3 Départements 325,347

8–16

174

21

1997

Webb et al.29

3–15

53

22

1997

South Glamorgan, Wales Mölnlycke

3–6

23

1998

3–14

24 25

1999 1999

1,941 Arvidsson et al.30 Sweden (West coast) Sponheim and Norway Akershus County 65,688 Skjeldal31 U.K. North Thames •490,000 Taylor et al.32 Karlstad 826 Kadesjö et al.33 Sweden (Central)

26

2000

Baird et al.34

U.K.

16,235

—

50

27

2000

Powell et al.35

U.K.

25,377

—

62

28

2000

Kielinen et al.8

Finland

North (Oulu and 152,732 Lapland)

—

187

29

2001

U.S.

36

2001

5–15

27

31

2001

Brick Township, 8,896 NJ Angleterre et 10,438 Pays de Galles Whole island 43,153

—

30

Bertrand et al.36 Fombonne et al.37 Magnússon and Saemundsen38

5–14

57

© 2006 by Taylor & Francis Group, LLC

Japan

U.K.

U.K. Iceland

Yokohama

Southeast Thames West Midlands

8,537

73,301

5

18

ICD-10

50.0

Clinical ICD10-like DSM-III-R

12.1

9

ICD-10

22.2

34

ICD-10

47.1c

ICD-10 DSM-IIIR/ICD-10 Gillberg’s criteria (Asperger syndrome) ICD-10

— 50.0

0–16 427 6.7–7.7 6

Clinical/ICD10/ DSM-IV ICD-8/ ICD-9/ ICD-10 DSM-IV DSM-IV/ ICD-10 Mostly ICD-10

2.6 (13.5) 1.81 (112/62) 6.57 (46/7)

21.08

11.4 ; 30.8

5.35

4.6 ; 6.1

7.2

5.3 ; 9.3

3.5 (7/2) 2.09 (23/11) — 5.0 (5/1)

46.4

16.1 ; 76.6

5.2

3.4 ; 6.9

8.7 72.6

7.9 ; 9.5 14.7 ; 130.6

15.7 (47/3) —

30.8

22.9 ; 40.6

7.8

5.8 ; 10.5

49.8

4.12 (156/50)

12.2

10.5 ; 14.0

36.7

2.2 (25/11) 8.0 (24/3) 4.2 (46/11)

40.5

28.0 ; 56.0

26.1

16.2 ; 36.0

13.2

9.8 ; 16.6

—

60 —

55.5 15.8

(continued)

2732_C002.fm Page 29 Wednesday, February 15, 2006 4:33 PM

Honda et al.27

29

1996

Past and Future Perspectives on Autism Epidemiology

19

Study Num Year of ber Publication 2001

33

2001

34

2002a

35 36

2002 2005

37

2005

Country

Area

Age

Diagnostic Criteria

Chakrabarti and Fombonne39 Davidovitch et al.40 Croen et al.3

U.K. Staffordshire (Midlands)

15,500

2.5–6.5

26

ICD10/ DSM-IV

Israel

Haiffa

26,160

7–11

26

U.S.

California DDS 4,950,333

5–12

5,038

DSM-III-R/ DSM-IV CDER (Full syndrome)

Madsen et al.5 Chakrabarti and Fombonne41 Fombonne et al.42

Denmark National register 63,859 U.K. Staffordshire 10,903 (Midlands) Canada Montreal island 27,749 (Quebec)

8 4–7

46 24

4–17

61

a

ICD-10 ICD-10/DSMIV DSM-IV

Percentage with Gender Prevalence Normal Ratio Rate/ IQ (M:F) 10,000

95% CI

29.2

3.3 (20/6)

16.8

10.3 ; 23.2

—

4.2 (21/5) 4.47 (4116/9 21) — 3.8 (19/5) 5.1 (51/10)

10.0

6.6;14.4

11.0

10.7;11.3

7.2 22.0

5.0 – 10.0 14.4 ; 32.2

21.6

16.5; 27.8

62.85e

— 33.3 —

This number corresponds to the sample described in Wing and Gould (1979). This rate corresponds to the first published paper on this survey and is based on 12 subjects amongst children aged 5–14 yr. cIn this study, mild mental retardation was combined with normal IQ, whereas moderate and severe mental retardation were grouped together. dFor the Göteborg surveys by Gillberg et al. (Gillberg, 1984; Steffenburg and Gillberg, 1986; Gillberg et al., 1991), a detailed examination showed that there was overlap between the samples included in the three surveys; consequently only the last survey has been included in this table. eThis proportion is likely to be overestimated and to reflect an underreporting of mental retardation in the CDER evaluations. b

© 2006 by Taylor & Francis Group, LLC

Understanding Autism: From Basic Neuroscience to Treatment

32

Authors

Size of Target Population

Number of Subjects with Autism

2732_C002.fm Page 30 Wednesday, February 15, 2006 4:33 PM

30

TABLE 2.1 (Continued) Prevalence Surveys of Autistic Disorder

2732_C002.fm Page 31 Wednesday, February 15, 2006 4:33 PM

Past and Future Perspectives on Autism Epidemiology

31

STUDY DESIGNS A few studies have relied on existing administrative databases3,4 or on national registers5 for case identification. Most investigations have relied on a two-stage or multistage approach to identify cases in the studied populations. The first screening stage of these studies often consisted of sending letters or brief screening scales requesting school and health professionals to identify possible cases of autism. Each investigation varied in several key aspects of this screening stage. First, the coverage of the population varied enormously from one study to another. In addition, the surveyed areas varied in terms of service development as a function of the specific educational or health care systems of each country and of the year of investigation. Secondly, the type of information sent out to professionals invited to identify children varied from simple letters including a few clinical descriptors of autism-related symptoms or diagnostic checklists rephrased in nontechnical terms to more systematic screening based on questionnaires or rating scales of known reliability and validity. Thirdly, participation rates in the first screening stages provide another source of variation in the screening efficiency of the surveys, although refusal rates tended to be very low. Only two studies (studies 1 and 30) provided an estimate of the reliability of the screening procedure. The sensitivity of the screening methodology is also difficult to gauge in autism surveys. The usual approach, which consists of sampling at random from among screened negative subjects in order to estimate the proportion of false negatives, has not been used in these surveys for the obvious reason that, due to the very low frequency of the disorder, it would be both imprecise and very costly to undertake such estimations. As a consequence, prevalence estimates must be seen as underestimates of the true prevalence rates. The magnitude of this underestimation is unknown in each survey. Similar considerations (such as these about the methodological variability across studies) also apply to the intensive assessment phases. Participation rates in these second-stage assessments were generally high. The source of information used to determine caseness usually involved a combination of informants and data sources, with a direct assessment of the person with autism in 21 studies. The assessments were conducted with various diagnostic instruments, ranging from a classical clinical examination to the use of batteries of standardized measures. The Autism Diagnostic Interview (ADI)6 and/or the Autism Diagnostic Observational Schedule (ADOS)7 were used in the most recent surveys. The precise diagnostic criteria retained to define caseness vary according to the study and, to a large extent, reflect historical changes in classification systems. Thus, Kanner’s criteria and Lotter’s and Rutter’s definitions were used in the first 8 surveys (all conducted before 1982), whereas DSM-based definitions took over thereafter, with ICD-10 as well after 1990. Kielinen et al.8 have shown that a two- to threefold variation in rates of autism can result from applying different diagnostic criteria to the same survey data.

CHARACTERISTICS OF AUTISTIC SAMPLES Data on children with autistic disorder were available in 37 surveys (studies 1 to 37; see Table 2.1). The number of subjects affected with autism ranged from 6

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 32 Wednesday, February 15, 2006 4:33 PM

32

Understanding Autism: From Basic Neuroscience to Treatment

(studies 18 and 25) to 5,038 (study 34) across studies (median: 48; mean: 209). An assessment of intellectual function was obtained in 21 studies. The median proportion of subjects without intellectual impairment was 29.6% (range: 0 to 60%)*. The corresponding figures are 29.3% (range: 6.6 to 100%) for mild to moderate intellectual impairments and 38.5% (range: 0 to 81.3%) for severe to profound levels of mental retardation. Gender repartition among subjects with autism was reported in 32 studies, totaling 6,963 subjects with autism, and the mean male:female ratio was 4.3:1. Gender differences were more pronounced when autism was not associated with mental retardation. In 13 studies (865 subjects) where the sex ratio was available, for the subjects with normal intellectual functioning, the median sex ratio was 5.5:1. Conversely, in 12 studies (813 subjects), the median sex ratio was 1.95:1 in the group with autism and moderate to severe mental retardation.

PREVALENCE ESTIMATIONS AUTISTIC DISORDER Prevalence estimates ranged from 0.7/10,000 to 72.6/10,000 (Table 2.1). Prevalence rates were negatively correlated with sample size (Spearman r = 0.73; p < .01); small-scale studies tended to report higher prevalence rates. When surveys were combined in two groups according to the median year of publication (1994), the median prevalence rate for 18 surveys published in the period 1966 to 1993 was 4.7/10,000, and the median rate for the 18 surveys published in the period 1994 to 2004 was 12.7/10,000. Indeed, the correlation between prevalence rate and year of publication reached statistical significance (Spearman r = 0.65; p < .01). The results of the 22 surveys with prevalence rates over 7/10,000 were all published after 1987. These findings point toward an increase in prevalence estimates in the last 15 to 20 years. The interpretation of this trend is discussed later. In order to derive a best estimate of the current prevalence of autism, it was therefore deemed appropriate to restrict the analysis to the 28 surveys published since 1987. The average rate was 16.2/10,000 and the median rate was 11.3/10,000. Similar values were obtained when slightly different rules and time cutpoints were used, with median and mean rates fluctuating between 10 and 13/10,000 and 13 and 18/10,000, respectively. From these results, a conservative estimate for the current prevalence of autistic disorder would be somewhere between 10/10,000 and 16/10,000. For further calculations, we arbitrarily adopted the midpoint of this interval as the working rate for autism prevalence, i.e., the value of 13/10,000.

UNSPECIFIED PDDS — PDDNOS (PERVASIVE DEVELOPMENT DISORDER NOT OTHERWISE SPECIFIED) Several studies have provided useful information on the rates of syndromes that are similar to autism but fall short of strict diagnostic criteria for autistic disorder.1 Different labels have been used to characterize them, such as the triad * Study 23, which relied upon different IQ groupings, has been excluded.

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 33 Wednesday, February 15, 2006 4:33 PM

Past and Future Perspectives on Autism Epidemiology

33

of impairments involving impairments in reciprocal social interaction, communication, and imagination.43 These groups would be overlapping with current diagnostic labels such as atypical autism and PDDNOS. Fourteen of the 37 surveys yielded separate estimates of the prevalence of these developmental disorders, with 11 studies showing higher rates for the nonautism disorders than for autism. The ratio of the rate of nonautistic PDD to the rate of autism had a mean value of 1.6, which, assuming a 13/10,000 rate for autistic disorder, translates into an average prevalence estimate of 20.8/10,000 for PDDNOS. This group has been much less studied in previous epidemiological studies but progressive recognition of its importance and relevance to autism has led to changes in the design of more recent epidemiological surveys (see following text), which are now designed to include these less typical forms in their case definition. It should be clear from these figures that they represent a very substantial group of children whose treatment needs are likely to be as important as those of children with autism. Yet, PDDNOS is an ill-defined heterogeneous group of autism spectrum disorders, the boundaries of which remain uncertain with respect to both classical autism and nonautistic spectrum developmental disorders. More research on this PDD variant is required, especially as it accounts for a large proportion of PDDs in recent surveys.

ASPERGER SYNDROME (AS) DISORDER (CDD)

AND

CHILDHOOD DISINTEGRATIVE

The reader is referred to recent epidemiological reviews for these two conditions.44,45 In brief, epidemiological studies of AS are sparse, probably due to the fact that it was acknowledged as a separate diagnostic category only recently in both ICD-10 and DSM-IV. Only two epidemiological surveys have been conducted which specifically investigated its prevalence (studies 4146 and 4233). However, only a handful (N < 5) of cases were identified in these surveys, with the resulting estimates being extremely imprecise. By contrast, other recent autism surveys have consistently identified smaller numbers of children with AS than those with autism within the same survey. In 8 such surveys (studies 23 to 27 and study 32 reviewed in Fombonne and Tidmarsh,45 and studies 36 and 37), the ratio of autism to AS rates in each survey was above unity, suggesting that the rate of AS was consistently lower than that for autism (Table 2.2). How much lower it was is difficult to establish from existing data, but a ratio of 5 to 1 would appear an acceptable, albeit conservative, conclusion based on this limited available evidence. This translates into a rate for AS which would be 15 th that of autism. We therefore used for subsequent calculations an estimate of 2.6/10,000 for AS, recognizing the strong limitations of the available data on AS. A recent survey of high-functioning PDDs in Welsh mainstream primary schools has yielded a relatively high (uncorrected) prevalence estimate of 14.5/10,000 (study 4347), but no separate rate was available for the Asperger disorder specifically. Few surveys have provided data on CDD (Table 2.3). Prevalence estimates ranged from 1.1 to 9.2/100,000. The pooled estimate based on 7 identified cases and a surveyed population of 358,633 children, was 1.9/100,000. The upper bound

© 2006 by Taylor & Francis Group, LLC

Assessment Study Number

Size of Population Sponheim and 65,688 Skjeldal, 199831

24

Taylor et al., 199932 Kadesjö et al., 199933

25

27 26 32

36

Informants

Instruments

3–14

Parent, child

490,000

0–16

Record

826

6.7–7.7

Child, parent, professional

Powell et al., 200035 Baird et al., 200034 Chakrabarti and Fombonne, 200139

25,377

1–4.9

Records

16,235

7

15,500

2.5–6.5

Chakrabarti and Fombonne, 200541

10,903

2.5–6.5

Parents, child, ADI-R, psychometry other data Child, parent, ADI-R, 2 weeks professional multidisciplinary assessment, MerrillPalmer, WPPSI Child, ADI-R, 2 weeks parent, multidisciplinary professional assessment, MerrillPalmer, WPPSI

Overall

© 2006 by Taylor & Francis Group, LLC

Parental interview + direct observation, CARS, ABC Rating of all data available in child record ADI-R, Griffiths scale or WISC, AS Screening questionnaire ADI-R, available data

Diagnostic Criteria

N

Asperger Syndrome (AS)

Rate/ 10,000

N

Rate/ 10,000

Autism/ AS Ratio

ICD-10

32

4.9

2

0.3

16.0

ICD-10

42 7 6

8.7

71

1.4

6.0

72.6

4

48.4

1.5

54

—

—

3.4

45

27.7

5

3.1

9.0

ICD-10, DSM-IV

26

16.8

13

8.4

2.0

ICD-10, DSM-IV

24

22.0

12

11.0

2.0

DSM-III-R/ ICD-10 Gillberg’s criteria (Asyndromes) DSM-III-R, DSM-IV, ICD-10 ICD-10, DSM-IV

614

16

123

5.0

Understanding Autism: From Basic Neuroscience to Treatment

23

Age group

Autism

2732_C002.fm Page 34 Wednesday, February 15, 2006 4:33 PM

34

TABLE 2.2 Asperger Syndrome (AS) in Recent Autism Surveys

9

23 31 32

36

Country (Region/State)

Size of Target Population

Age Group

Burd et al., 198717

U.S. (North Dakota)

180,986

2–18

Sponheim and Skjeldal, 199831 Magnusson and Saemundsen, 200138 Chakrabarti and Fombonne, 200139

Norway (Akershus County) Iceland (whole island)

65,688

3–14

85,556

5–14

U.K. (Staffordshire, Midlands)

15,500

2.5–6.5

Chakrabarti and Fombonne, 200541

U.K. (Staffordshire, Midlands)

10,903

2.5–6.5

Pooled estimates

358,633

Assessment Structured parental interview and review of all data available — DSM-III criteria Parental interview and direct observation (CARS, ABC) ADI-R, CARS, and psychological tests — mostly ICD-10 ADI-R, 2 weeks multidisciplinary assessment, Merrill-Palmer, WPPSI — ICD-10/DSM-IV ADI-R, 2 weeks multidisciplinary assessment, Merrill-Palmer, WPPSI — ICD-10/DSM-IV

Prevalence Estimate (/100,000)

95% CI (/100,000)

N

M/F

2

2/0

1.11

0.13 ; 3.4

1

?

1.52

0.04 ; 8.5

2

2/0

2.34

0.3 ; 8.4

1

1/0

6.4

0.16 ; 35.9

1

1/0

9.2

0–58.6

7

6/0

1.9

0.87–4.15

35

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 35 Wednesday, February 15, 2006 4:33 PM

Study Number

Past and Future Perspectives on Autism Epidemiology

TABLE 2.3 Surveys of Childhood Disintegrative Disorder (CDD)

2732_C002.fm Page 36 Wednesday, February 15, 2006 4:33 PM

36

Understanding Autism: From Basic Neuroscience to Treatment

limit of the associated confidence interval (4.15/100,000) indicates that CDD is a very rare condition, with 1 case being present for every 65 cases of autistic disorder.

PREVALENCE

FOR

COMBINED PDDS

Taking the aforementioned conservative estimates, the prevalence for all PDDs is at least 36.4/10,000 (i.e., the sum of estimates for autism [13/10,000], PDDNOS [20.8/10,000], and AS [2.6/10,000]). This global estimate is derived from a conservative analysis of existing data. However, nine recent epidemiological surveys yielded even higher rates in seven instances (Table 2.4). The two surveys that did not, probably underestimated the rates. In the Danish investigation (study 35), case finding depended upon notification to a national registry, a method which is usually associated with lower sensitivity for case finding. The Atlanta survey by the CDC (study 39) was based on a very large population (which typically yields a lower prevalence rate: see preceding text) and age-specific rates were, in fact, in the 40 to 45/10,000 range in some birth cohorts.50 In most other studies, the case definition chosen for these investigations was that of a PDD as opposed to the narrower approach of focusing on autistic disorder that was typical of previous surveys. Investigators were concerned with any combination of severe developmental abnormalities occurring in one or more of the three symptomatic domains defining PDD and autism. Case-finding techniques employed in these surveys were proactive, relying on multiple and repeated screening phases, involving both different informants at each phase and surveying the same cohorts at different ages, which certainly maximized the sensitivity of case identification. Assessments were performed with standardized diagnostic measures (i.e., ADI-R and ADOS) that match well the more dimensional approach retained for case definition. The size of the targeted populations was reasonably small (between 9,000 and 28,000), allowing for the most efficient use of research resources. Conducted in different regions and countries by different teams, the convergence of estimates to around 60/10,000 for all PDDs combined is striking, especially when coming from studies with improved methodology. This estimate is now the best estimate currently available for the prevalence of PDDs.

TIME TRENDS The debate on the hypothesis of a secular increase in rates of autism has been obscured by a lack of clarity in the measures of disease occurrence used by investigators or, rather, in their interpretation. In particular, it is crucial to differentiate prevalence (the proportion of individuals in a population who suffer from a defined disorder) from incidence (the number of new cases occurring in a population over a period of time). Prevalence is useful to estimate needs and plan services, whereas only incidence rates can be used for causal research. Both prevalence and incidence estimates will be inflated when case definition is broadened and case ascertainment is improved. Time trends in rates can therefore only be gauged in investigations which hold these parameters under strict control over time. These methodological

© 2006 by Taylor & Francis Group, LLC

Study Number

Rate/ 10,000

Male/Female Ratio

PDDNOS + AS % IQ Normal

Rate/ 10,000

Male/Female Ratio

Author

Age 7 3–10 4–7

30.8 40.5 16.8

15.7 2.2 3.3

60 37 29

27.1 27.0 44.5

4.5 3.7 4.3

8 4–7

7.7 22.0

— 4.0

— 33.3

22.2 35.8

37 38 39

Baird et al., 200034 Bertrand et al., 200136 Chakrabarti and Fombonne, 200139 Madsen et al., 20025 Chakrabarti and Fombonne, 2005 41 Fombonne et al., 200542 Scott et al., 200248 Yeargin-Allsopp et al., 200349

40

Gurney et al., 20034

4–17 5–11 3–10 6–11

21.6 — — —

5.1 — — —

— — — —

42.9 — — —

26 29 32 35 36

a

All PDDs % IQ Normal

Rate/ 10,000

— 51 94

57.9 67.5 61.3

— 8.7

— 91.6

30.0 58.7

4.7 — — —

— — — —

64.9 58.3a 34.0 52.0

Computed by the author.

37

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 37 Wednesday, February 15, 2006 4:33 PM

Autism

Past and Future Perspectives on Autism Epidemiology

TABLE 2.4 Newer Epidemiological Surveys of PDDs

2732_C002.fm Page 38 Wednesday, February 15, 2006 4:33 PM

38

Understanding Autism: From Basic Neuroscience to Treatment

requirements must be borne in mind when reviewing the evidence for a secular increase in rates of PDDs. Five approaches to assess this question have been used in the literature; these are discussed in the following subsections.

REFERRAL STATISTICS Increasing numbers of children referred to specialist services or known to special education registers have been taken as evidence for an increased incidence of autism spectrum disorders. However, trends over time in referred samples are confounded by many factors such as referral patterns, availability of services, heightened public awareness, decreasing age at diagnosis, and changes over time in diagnostic concepts and practices, to name only a few. Failure to control for these confounding factors was obvious in some recent reports,51 such as the widelyquoted reports from the California educational services.52,53 Firstly, these reports applied to numbers rather than rates, and failure to relate these numbers to meaningful denominators left the interpretation of an upward trend vulnerable to changes in the composition of the underlying population. For example, the population of California was 19,971,000 in 1970 and rose to 35,116,000 as of July 1, 2002, a change of +75.8%. Second, the focus on the year-to-year changes in absolute numbers of subjects known to California state-funded services detracts from more meaningful comparisons. For example, as of December 2002, the total of subjects with a PDD diagnosis was 17,748 in the 0 to 19 age group (including 16,108 autism codes 1 and 2 and 1,640 other PDDs53). The population of 0- to 19-yr-olds of California was 10,462,273 in July 2002. If one applies a rather conservative PDD rate of 30/10,000, one would expect to have 31,386 subjects with a PDD within this age group living in California. These calculations do not support the “epidemic” interpretation of the California Department of Development Services (DDS) data. Rather, they suggest that children identified in the California DDS database were only a subset of the population prevalence pool and that the increasing numbers reflect merely an increasing proportion of children accessing services. Third, no attempt was ever made to adjust the trends for changes in diagnostic concepts and definitions. However, major nosographical modifications were introduced during the corresponding years, with a general tendency in most classifications to broaden the concept of autism (as embodied in the terms “autism spectrum” or “pervasive developmental disorder”). Fourth, age characteristics of the subjects recorded in official statistics were portrayed in a confusing manner, where the preponderance of young subjects was presented as evidence of increasing rates in successive birth cohorts (see Fombonne51). The problems associated with disentangling age from period and cohort effects in such observational data are well known in the epidemiological literature and deserve better statistical handling. Fifth, the decreasing age at diagnosis leads, in itself, to increasing numbers of young children being identified in official statistics or referred to already busy specialist services. Earlier identification of children from the prevalence pool may result in increased service activity; however, it does not mean increased incidence.

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 39 Wednesday, February 15, 2006 4:33 PM

Past and Future Perspectives on Autism Epidemiology

39

Another study of this data set was subsequently launched to demonstrate the validity of the epidemic hypothesis.54 The authors relied on the DDS data and aimed at ruling out changes in diagnostic practices and immigration into California as factors explaining the increased numbers. Although immigration was reasonably ruled out, the study comparing diagnoses of autism and mental retardation over time was impossible to interpret in light of the extremely low (<20%) response rates. Furthermore, a study based only on cases registered for services cannot rule out the possibility that the proportion of cases within the general population who registered with services has changed over time. For example, assuming a constant incidence and prevalence at two different time points (i.e., there is no epidemic), the number of cases known to a public agency delivering services could well increase by 200% if the proportion of cases from the community referred to services rises from 25 to 75% in the interval. In order to rule out this likely explanation (see preceding text), data over time are needed both on referred subjects and on nonreferred (or referred to other services) subjects. Failure to address this possibility precludes any inference drawn from a study of the California DDS database population from being generalized to the California population.50 The conclusions of this report were therefore simply unwarranted. Some evidence of “diagnostic switching” was produced in California55 and in the U.K. by Jick and Kaye56 who showed that the incidence of specific developmental disorders (including language disorders) decreased by about the same amount that the incidence of diagnoses of autism increased in boys born from 1990 to 1997. On the whole, evidence from these referral statistics is very weak, and proper epidemiological studies are needed in order to assess secular changes in the incidence of a disorder.

COMPARISON

OF

CROSS-SECTIONAL EPIDEMIOLOGICAL SURVEYS

As shown earlier, epidemiological surveys of autism each possess unique design features which could account almost entirely for the between-studies variations in rates; time trends in rates of autism are therefore difficult to gauge from published prevalence rates. The significant correlation previously mentioned between prevalence rate and year of publication could merely reflect increased efficiency over time in the case-identification methods used in surveys as well as changes in diagnostic concepts and practices.8,29,38 The most convincing evidence that method factors could account for most of the variability in published prevalence estimates comes from a direct comparison of eight recent surveys conducted in the U.K. and the U.S. (Table 2.5). In each country, four surveys were conducted at around the same year and with similar age groups. As there is no reason to expect huge between-area differences in rates, prevalence estimates should therefore be comparable within each country. However, an inspection of the estimates obtained in each set of studies (Table 2.5: right-hand column) shows a 6-fold variation in rates for the U.K. surveys, and a 14-fold variation in the U.S. rates. In each set of studies, the high rates derive from surveys where intensive population-based screening techniques were employed, whereas the lower rates were obtained from the studies relying on administrative methods for case finding. Because no passage of time was involved, the magnitude of these differences in rates can only be attributed to differences in case-identification

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 40 Wednesday, February 15, 2006 4:33 PM

40

Understanding Autism: From Basic Neuroscience to Treatment

TABLE 2.5 Impact of Study Design on Prevalence

Size

Age Group

Method

PDD Rate/ 10,000

U.K. Studies

Location

Chakrabarti and Fombonne, 200139 Baird et al., 200034

Staffordshire

15,500

2 1/2–6 1/2

Intense screening + assessment

62.6

Southeast Thames

16,235

7

57.9

Fombonne et al., 200137

England and Wales

10,438

5–15

Taylor et al., 199932

North Thames

490,000

0–16

Early screening + follow-up identification National household survey of psychiatric disorders Administrative records

26.1

10.1

U.S. Studies Bertrand et al., 200136 Sturmey and Vernon, 200157 California DDS, 199952 Hillman et al., 200058

Brick Township, NJ Texas

8,896

3–10

Multiple sources of ascertainment

67

3,564,577

6–18

Educational services

16

California

3,215,000

4–9

Educational services

15

Missouri

—

5–9

Educational services

4.8

methods across surveys. Thus, no inference on trends in the incidence of PDDs can be derived from a simple comparison of prevalence rates over time, because studies conducted at different periods are likely to differ even more with respect to their methodology.

REPEAT SURVEYS

IN

DEFINED GEOGRAPHICAL AREAS

Repeated surveys, using the same methodology and conducted in the same geographical area at different points in time, can potentially yield useful information on time trends provided that methods are kept relatively constant. The Göteborg studies24,59 provided 3 prevalence estimates which increased over a short period of time from 4.0 (1980) to 6.6 (1984) and then to 9.5/10,000 (1988), the gradient being even steeper when the rates for the urban area alone are considered (4.0, 7.5, and

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 41 Wednesday, February 15, 2006 4:33 PM

Past and Future Perspectives on Autism Epidemiology

41

11.6/10,000).24 However, comparisons of these rates is not straightforward as different age groups were included in each survey. Secondly, the increased prevalence found in the second survey was explained by the improved detection among the mentally retarded and that in the third survey by the inclusion of cases born to immigrant parents. That the majority of the latter group was born abroad suggests that migration into the area could be a key explanation. Taken in conjunction with a change in local services and a progressive broadening of the definition of autism over time (acknowledged by the authors24), these findings do not provide evidence for an increased incidence in the rate of autism. Two separate surveys of children born between 1992 and 1995 and between 1996 and 1998 in Staffordshire in the U.K. (Table 2.1; studies 32 and 36) were performed with rigorously identical methods for case definition and case identification. The prevalence for combined PDDs was comparable and not statistically different in the two surveys,41 suggesting no upward trend in the overall rates of the PDDs during the studies’ time interval.

SUCCESSIVE BIRTH COHORTS In large surveys encompassing a wide age range, increasing prevalence rates among the most recent birth cohorts could be interpreted as indicating a secular increase in the incidence of the disorder, provided that alternative explanations can confidently be ruled out. This analysis was used in two large French surveys (Table 2.1, studies 17 and 20). The surveys included birth cohorts from 1972 to 1985 (735,000 children, 389 of who had autism) and, pooling the data of both surveys, age-specific rates showed no upward trend.28 A recent analysis of special educational disability data from Minnesota showed a 16-fold increase in the number of children identified with a PDD from 1991–1992 to 2001–2002.4 The increase was not specific to autism because, during the same period, an increase of 50% was observed for all disability categories (except severe mental handicap), especially for the category including attention deficit hyperactivity disorder (ADHD). The large sample size allowed the authors to assess age, period, and cohort effects. Prevalence increased regularly in successive birth cohorts; for example, among the 7-yr-olds, the prevalence rose from 18/10,000 in those born in 1989, to 29/10,000 in those born in 1991, and to 55/10,000 in those born in 1993, which was suggestive of birth-cohort effects. Within the same birth cohort, age effects were also apparent: for children born in 1989, the prevalence rose with age from 13/10,000 at age 6 to 21/10,000 at age 9 and 33/10,000 at age 11. As argued by the authors, this pattern is not consistent with what one would expect from a chronic nonfatal condition diagnosed in the first years of life. Their analysis also showed a marked period effect that identified the early 1990s as the period when rates started to go up in all ages and birth cohorts. Gurney et al.4 further argued that this phenomenon coincided closely with the inclusion of PDDs in the federal Individual with Disabilities Educational Act (IDEA) funding and reporting mechanism in the U.S. A similar interpretation of upward trends had been put forward by Croen et al.3 in their analysis of the California DDS data.

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 42 Wednesday, February 15, 2006 4:33 PM

42

Understanding Autism: From Basic Neuroscience to Treatment

INCIDENCE STUDIES Only three studies provided recent incidence estimates.35,60,61 All studies showed an upward trend in incidence over short periods of time. In the largest study of 1410 subjects, there was a tenfold increase in the rate of first recorded diagnoses of PDDs in the U.K. general practice medical records from 1988 to 2001.61 The increase was more marked for PDDs other than autism, but the increase in autism was also obvious. However, none of these studies could determine the impact of changes over time in diagnostic criteria, improved awareness, and service availability on the upward trend.

CONCLUSION

ON

TIME TRENDS

The available epidemiological evidence does not strongly support the hypothesis that the incidence of autism has increased. As it stands now, the recent upward trend in rates of prevalence cannot be directly attributed to an increase in the incidence of the disorder. There is good evidence that changes in diagnostic criteria, diagnostic substitution, changes in the policies for special education, and the increasing availability of services are responsible for the higher prevalence figures. Most of the existing epidemiological data are inadequate to properly test hypotheses regarding changes in the incidence of autism in human populations. Moreover, due to the low frequency of autism and PDDs, power is seriously limited in most investigations and variations of small magnitude in the incidence of the disorder are very likely to go undetected.

CORRELATES OF AUTISM ASSOCIATED MEDICAL CONDITIONS Overall the proportion of cases of autism that could be causally attributed to known medical disorders was low. Tuberous sclerosis and fragile X disorder are the medical disorders with the strongest association with autism. The fraction of cases of autism with any known medical condition of potential etiological significance ranged from 0 to 16.7%, with a mean value of 5.9% (see Fombonne1 for further details).

AUTISM, RACE,

AND IMMIGRANT

STATUS

Some investigators have mentioned the possibility that rates of autism might be higher among immigrants.24,62–64 Five of the 17 children with autism identified in the Camberwell study were of Caribbean origin (study 443) and the estimated rate of autism was 6.3/10,000 for this group as compared to 4.4/10,000 for the rest of the population.43 However, the wide confidence intervals associated with the rates from this study (Table 2.1) indicate that there is no statistically significant difference. In addition, this area of London had received a large proportion of immigrants from the Caribbean region in the 1960s and, under circumstances where migration flux in and out of an area are happening, estimation of population rates should be viewed with much caution. Similarly, the findings from the Göteborg studies paralleled an increased migration flux in the early 1980s in this area.63 In the

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 43 Wednesday, February 15, 2006 4:33 PM

Past and Future Perspectives on Autism Epidemiology

43

Icelandic survey (study 31), 2.5% of the parents of the autistic children were of non-European origin compared to a corresponding rate of 0.5% in the whole population, but it was unclear if this represented a significant difference. In Norway (study 23), the proportion of children with autism and a non-European origin was marginally, but not significantly, raised as compared to the population rate of immigrants (8% vs. 2.3%), but this was based on a very small sample (2 children of non-European origin). A U.K. survey found comparable rates in areas contrasting for their ethnic composition.35 In the Utah survey (study 15), the racial distribution of the parents of autistic children showed no deviation from that in the general population of the state. The analysis of a large sample (N = 4,356) of Californian PDD children showed a lower risk of autism in the children of Mexicoborn mothers and a similar risk for children of mothers born outside the U.S. as compared to California-born mothers.3 In this study, the risk of PDD was raised in African American mothers with an adjusted rate ratio of 1.6 (95% CI: 1.5 to 1.8); by contrast, the prevalence was similar in white, black, and other races in the population-based survey of Atlanta,49 where case ascertainment is likely to have been more complete than in the previous study. Taken together, the combined results of these reports should be interpreted in the specific methodological context of these investigations. Most studies had low numbers of identified cases and especially small numbers of autistic children born of immigrant parents. Statistical testing was not rigorously conducted and doubts could be raised in several studies about the appropriateness of the comparison data that were used.1 The hypothesis of an association between immigrant status or race and autism, therefore, remains largely unsupported by the empirical results. Most of the claims about these possible correlates of autism derived from post-hoc observations of very small samples and were not subjected to rigorous statistical testing. Large studies have generally failed to detect such associations.

AUTISM

AND

SOCIAL CLASS

Twelve studies provided information on the social class of the families of autistic children. Of these, four studies (1, 2, 3, and 5) suggested an association between autism and social class or parental education. The year of data collection for these four investigations was before 1980 (Table 2.1), and all studies conducted thereafter provided no evidence for the association. Earlier findings were probably due to artifacts in the data regarding availability of services and in the case-finding methods, as already shown in other samples.62,65

CLUSTER REPORTS Occasional reports of space or time clustering of cases of autism have raised concerns in the general public. In fact, only one such report has been published in the professional literature66; this described 7 children with either autism or PDDNOS living within a few streets of each other in a small town of the Midlands (U.K.). The cluster was first identified by a parent and the subsequent analysis was uninformed with proper statistical procedures and was inconclusive as to whether or not this cluster could have occurred by chance alone.

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 44 Wednesday, February 15, 2006 4:33 PM

44

Understanding Autism: From Basic Neuroscience to Treatment

Cluster alarms are likely to represent random occurrence in most instances, as illustrated by several recent investigations of cluster alarms for other rare disorders of childhood. Cluster alarms in autism have not been investigated with scientific rigor although research strategies and ad hoc statistical procedures exist for that purpose. The approach to such cluster alarms should be to confirm the alarm in the first place, using the available techniques to assess the significance of clusters and to exclude random noise in the space and time distribution of the disorder. It is only when an alarm has been confirmed that more complex epidemiological investigations should be set up to investigate risk factors and causal mechanisms.

CONCLUSION Epidemiological surveys of autism and PDDs have now been carried out in several countries. Methodological differences in case definition and case-finding procedures make between-survey comparisons difficult to perform. However, from recent studies, a best estimate of 0.6% can be confidently derived for the prevalence of autism spectrum disorders. Current evidence does not support the hypothesis of a secular increase in the incidence of autism, but power to detect time trends is seriously limited in existing data sets. Although it appears that prevalence estimates have gone up over time, this increase most likely represents changes in the concepts, definitions, service availability, and awareness of autism spectrum disorders in both the lay public and professionals. To assess whether or not the true incidence has increased, method factors that account for an important proportion of the variability in rates must be tightly controlled. Future epidemiological studies of autism should follow three directions. First, as the hypothesis of a true increase in the incidence of PDDs cannot be definitely ruled out, surveillance programs should be implemented. Selected, well-defined, relatively stable (i.e., with no massive migrations in or out of the area) populations could be monitored over time. Reflecting modern concepts of PDD, a broad dimensional approach to case definition should be employed. Data should be collected with standardized tools, focusing on symptoms and domains of impairment rather than on predefined diagnostic categories that change frequently and sometimes lack validity. Applying diagnostic algorithms to symptom-oriented data sets can then be performed to detect trends, unconfounded by changes in case definition. Trends will be best evaluated among children of primary school age (i.e., 8- to 12-yr-olds) to avoid problems arising both from diagnostic difficulties in preschoolers and from the tendency for high-functioning presentations to be detected at a later age. Second, focused epidemiological investigations should be performed to evaluate risk associations with newly identified risk factors. Thus, claims of an association between immunizations (MMR and thimerosal) were promptly investigated by epidemiologists to be finally rejected.67,68 Similarly, time and space clustering, if reported, should be investigated with proper methods. Third, in the absence of strong leads on risk factors, large well-powered hypothesis-generating case–control studies should be performed to evaluate risk associations between PDDs and a broad range of environmental exposures, especially those occurring prenatally or in early postnatal life. In these investigations, DNA samples should be systematically collected because

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 45 Wednesday, February 15, 2006 4:33 PM

Past and Future Perspectives on Autism Epidemiology

45

gene–environment interactions and etiological heterogeneity will likely be important to develop causal models of the disorder. Taking 35/10,000 and 60/10,000 as two working rates for the combination of all PDDs, and using population estimates for the U.S. of July 1, 2002, it can be estimated that about 284,000, and up to 486,000, subjects under the age of 20 suffer from a PDD in the U.S. These figures have straightforward implications for current and future needs in services and early educational intervention programs.

REFERENCES 1. Fombonne, E., Epidemiological surveys of autism and other pervasive developmental disorders: an update, Journal of Autism and Developmental Disorders, 33(4), 365, 2003. 2. Fombonne, E., The epidemiology of autism: a review, Psychological Medicine, 29, 769, 1999. 3. Croen, L.A., Grether, J.K., Hoogstrate, J., and Selvin, S., The changing prevalence of autism in California, Journal of Autism and Developmental Disorder, 32(3), 207, 2002. 4. Gurney, J.G., Fritz, M.S., Ness, K.K., Sievers, P., Newschaffer, C.J., Shapiro, E.G., Analysis of prevalence trends of autism spectrum disorder in Minnesota, Archives of Pediatrics and Adolescent Medicine, 157(7), 622, 2003. 5. Madsen, K.M., Hviid, A., Vestergaard, M., Schendel, D., Wohlfahrt, J., Thorsen, P., A population-based study of measles, mumps, and rubella vaccination and autism, New England Journal of Medicine, 347(19), 1477, 2002. 6. LeCouteur, A., Rutter, M., Lord, C., Rios, P., Robertson, S., Holdgrafer, M., and McLennan, J., Autism diagnostic interview: a standardized investigator-based instrument, Journal of Autism and Developmental Disorders, 19, 363, 1989. 7. Lord, C., Risi, S., Lembrecht, L., Cook, E.H., Leventhal, B.L., DiLavore, P.C., Pickles, A., and Rutter, M., The autism diagnostic observation schedule-generic: a standard measure of social and communication deficits associated with the spectrum of autism, Journal of Autism and Developmental Disorders, 30, 205, 2000. 8. Kielinen, M., Linna, S.L., and Moilanen, I., Autism in Northern Finland, European Child and Adolescent Psychiatry, 9, 162, 2000. 9. Lotter, V., Epidemiology of autistic conditions in young children: prevalence, Social Psychiatry, 1, 124, 1966. 10. Brask, B.H., A prevalence investigation of childhood psychoses, in Nordic Symposium on the Care of Psychotic Children, Oslo, Norway: Barnepsychiatrist Forening, 1970. 11. Treffert, D.A, Epidemiology of infantile autism, Archives of General Psychiatry, 22, 431, 1970. 12. Wing, L., Yeates, S.R., Brierly, L.M., and Gould, J., The prevalence of early childhood autism: comparison of administrative and epidemiological studies, Psychological Medicine, 6, 89, 1976. 13. Hoshino, Y., Yashima, Y., Ishige, K., Tachibana, R., Watanabe, M., Kancki, M., Kumashiro, H., Ueno, B., Takahashi, E., Furukawa, H., The epidemiological study of autism in FukushimaKen, Folia Psychiatrica et Neurologica Japonica, 36, 115, 1982. 14. Bohman, M., Bohman, I.L., Björck, P.O., and Sjöholm, E., Childhood psychosis in a northern Swedish county: some preliminary findings from an epidemiological survey, in Schmidt, M.H. and Remschmidt, H. (Eds.), Epidemiological Approaches in Child Psychiatry, Georg Thieme Verlag, Stuttgart, 1983, pp. 164–173.

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 46 Wednesday, February 15, 2006 4:33 PM

46

Understanding Autism: From Basic Neuroscience to Treatment 15. McCarthy, P., Fitzgerald, M., and Smith, M.A., Prevalence of childhood autism in Ireland, Irish Medical Journal, 77(5), 129, 1984. 16. Steinhausen, H.-C., Göbel, D., Breinlinger, M., and Wohlloben, B., A community survey of infantile autism, Journal of the American Academy of Child Psychiatry, 25(2), 186, 1986 17. Burd, L., Fisher, W., and Kerbeshan, J. A prevalence study of pervasive developmental disorders in North Dakota, Journal of the American Academy of Child and Adolescent Psychiatry, 26, 5, 700–703, 1987. 18. Matsuishi, T., Shiotsuki, M., Yoshimura, K., Shoji, H., Imuta, F., Yamashita, F., High prevalence of infantile autism in Kurume City, Japan, Journal of Child Neurology, 2, 268, 1987. 19. Tanoue, Y., Oda, S., Asano, F., and Kawashima, K., Epidemiology of infantile autism in Southern Ibaraki, Japan: differences in prevalence in birth cohorts, Journal of Autism and Developmental Disorders, 18, 155, 1988. 20. Bryson, S.E., Clark, B.S., and Smith, I.M, First report of a Canadian epidemiological study of autistic syndromes, Journal of Child Psychology and Psychiatry, 4, 433–445, 1988. 21. Sugiyama, T. and Abe, T., The prevalence of autism in Nagoya, Japan: a total population study, Journal of Autism and Developmental Disorders, 19(1), 87, 1989. 22. Cialdella, P. and Mamelle, N., An epidemiological study of infantile autism in a French department, Journal of Child Psychology and Psychiatry, 30, 1, 165, 1989. 23. Ritvo, E.R., Freeman, B.J., Pingree, C., Mason-Brothers, A., Jorde, L., Jenson, W.R., McMahon, W.M., Petersen, P.B., Mo, A., and Ritvo, A., The UCLA-University of Utah epidemiologic survey of autism: prevalence, American Journal of Psychiatry, 146(2), 194, 1989. 24. Gillberg, C., Steffenburg, S., and Schaumann, H. Is autism more common now than ten years ago? British Journal of Psychiatry, 158, 403, 1991. 25. Fombonne, E. and du Mazaubrun, C., Prevalence of infantile autism in 4 French regions, Social Psychiatry and Psychiatric Epidemiology, 27, 203, 1992. 26. Wignyosumarto, S., Mukhlas, M., and Shirataki, S., Epidemiological and clinical study of autistic children in Yogyakarta, Indonesia, Kobe Journal of Medical Sciences, 38(1), 1, 1992. 27. Honda, H., Shimizu, Y., Misumi, K., Niimi, M., and Ohashi, Y., Cumulative incidence and prevalence of childhood autism in children in Japan, British Journal of Psychiatry, 169, 228, 1996. 28. Fombonne, E., du Mazaubrun, C., Cans, C., and Grandjean, H., Autism and associated medical disorders in a large French epidemiological sample, Journal of the American Academy of Child and Adolescent Psychiatry, 36(11), 1561, 1997. 29. Webb, E.V.J., Lobo, S., Hervas, A., Scourfield, J., and Fraser, W.I., The changing prevalence of autistic disorder in a Welsh health district, Developmental Medicine and Child Neurology, 39, 150, 1997. 30. Arvidsson, T., Danielsson, B., Forsberg, P., Gillberg, C., Johansson, M, and Kjellgren, G. Autism in 3–6-year-old children in a suburb of Goteborg, Sweden, Autism, 2, 163–173, 1997. 31. Sponheim, E. and Skjeldal, O., Autism and related disorders: epidemiological findings in a Norwegian study using ICD-10 diagnostic criteria, Journal of Autism and Developmental Disorders, 28, 217, 1998 32. Taylor, B., Miller, E., Farrington, C.P., Petropoulos, M-C., Favot-Mayaud, I., Li, J., Waight, P.A., Autism and measles, mumps, and rubella vaccine: no epidemological evidence for a causal association, The Lancet, 353, 2026, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 47 Wednesday, February 15, 2006 4:33 PM

Past and Future Perspectives on Autism Epidemiology

47

33. Kadesjö, B., Gillberg, C., and Hagberg, B., Autism and Asperger syndrome in sevenyear old children: a total population study, Journal of Autism and Developmental Disorders, 29(4), 327, 1999. 34. Baird, G., Charman, T., Baron-Cohen, S. et al., A screening instrument for autism at 18 months of age: a 6-year follow-up study, Journal of the American Academy of Child and Adolescent Psychiatry, 39, 694–702, 2000. 35. Powell, J., Edwards, A., Edwards, M. et al., Changes in the incidence of childhood autism and other autistic spectrum disorders in preschool children from two areas in the West Midlands, U.K. Developmental Medicine and Child Neurology, 42(9), 624, 2000. 36. Bertrand, J., Mars, A., Boyle, C. et al., Prevalence of autism in a United States population: the Brick Township, New Jersey, Investigation, Pediatrics, 108, 1155–1161, 2001. 37. Fombonne, E., Simmons, H., Ford, T., Meltzer, H., and Goodman, R., Prevalence of pervasive developmental disorders in the British nationwide survey of child mental health, Journal of the American Academy of Child and Adolescent Psychiatry, 40(7), 820, 2001. 38. Magnusson, P. and Saemundsen, E., Prevalence of autism in Iceland, Journal of Autism and Developmental Disorders, 31, 153, 2001. 39. Chakrabarti, S. and Fombonne, E., Pervasive developmental disorders in preschool children, Journal of the American Medical Association, 285(24), 3093, 2001. 40. Davidovitch, M., Holtzman, G., and Tirosh, E., Autism in the Haifa area: an epidemiological perspective, Israeli Medical Association Journal, 3, 188, 2001. 41. Chakrabarti, S. and Fombonne, E., Pervasive developmental disorders in preschool children: confirmation of high prevalence, American Journal of Psychiatry, 162(6), 1133–1141, 2005. 42. Fombonne, E., Zakarian, R., Bennett, A., Meng, L., and McLean-Heywood, D., Pervasive developmental disorders in Montréal, Québec: prevalence and links with immunizations (submitted). 43. Wing, L. and Gould, J., Severe impairments of social interactions and associated abnormalities in children: epidemiology and classification, Journal of Autism and Developmental Disorders, 9(1), 11, 1979. 44. Fombonne, E., Prevalence of childhood disintegrative disorder (CDD), Autism, 6, 2, 147, 2002. 45. Fombonne, E. and Tidmarsh, L., Epidemiological data on Asperger disorder, Child and Adolescent Psychiatric Clinics of North America, 12, 15, 2003. 46. Ehlers, S. and Gillberg, C., The epidemiology of Asperger syndrome: a total population study, Journal of Child Psychology and Psychiatry, 34, 1327, 1993. 47. Webb, E., Morey, J., Thompsen, W., Butler, C., Barber, M., and Fraser, W.I., Prevalence of autistic spectrum disorder in children attending mainstream schools in a Welsh education authority, Developmental Medicine and Child Neurology, 45(6), 377, 2003. 48. Scott, F.J., Baron-Cohen, S., Bolton, P., and Brayne, C., Brief report: prevalence of autism spectrum conditions in children aged 5–11 years in Cambridgeshire, U.K., Autism, 6(3), 231, 2002. 49. Yeargin-Allsopp, M., Rice, C., Karapurkar, T., Doernberg, N., Boyle, C., and Murphy, C., Prevalence of autism in a U.S. metropolitan area, Journal of the American Medical Association, 289(1), 49, 2003. 50. Fombonne, E., The prevalence of autism, Journal of the American Medical Association, 289(1), 1, 2003.

© 2006 by Taylor & Francis Group, LLC

2732_C002.fm Page 48 Wednesday, February 15, 2006 4:33 PM

48

Understanding Autism: From Basic Neuroscience to Treatment 51. Fombonne, E., Is there an epidemic of autism? Pediatrics, 107, 411, 2001. 52. California Department of Developmental Services, Changes in the population of persons with autism and pervasive developmental disorders in California’s Developmental Services System: 1987 through 1998, Report to the Legislature March 1, 1999 19 p, http://www.dds.ca.gov. 53. California Department of Developmental Services, Autism Spectrum Disorders: Changes in the California caseload — An Update 1999 through 2002, April 2003, http://www.dds.ca.gov/Autism/pdf/AutismReport2003.pdf. 54. MIND Institute Report to the Legislature on the Principal Findings from the epidemiology of autism in California, A comprehensive pilot study, University of California, Davis, October 17, 2002. 55. Eagle, R.S., Commentary: further commentary on the debate regarding increase in autism in California, Journal of Autism and Developmental Disorders, 34(1), 87, 2004. 56. Jick, H. and Kaye, J.A., Epidemiology and possible causes of autism, Pharmacotherapy, 23(12), 1524, 2003. 57. Sturmey, P. and Vernon, J., Administrative prevalence of autism in the Texas school system, Journal of the American Academy of Child and Adolescent Psychiatry, 40(6), 621, 2001. 58. Hillman, R., Kanafani, N., Takahashi, T., and Miles, J., Prevalence of autism in Missouri: changing trends and the effect of a comprehensive State autism project, Missouri Medicine, 97, 159, 2000. 59. Gillberg, C., Infantile autism and other childhood psychoses in a Swedish region: epidemiological aspects, Journal of Child Psychology and Psychiatry, 25, 35, 1984. 60. Kaye, J., Melero-Montes, M., and Jick, H., Mumps, measles, and rubella vaccine and the incidence of autism recorded by general practitioners: a time trend analysis, British Medical Journal, 322, 0, 2001. 61. Smeeth, L., Cook, C., Fombonne, E., Rodrigues, L., Smith, P., and Hall, A., Rate of first recorded diagnosis of autism and other pervasive developmental disorders in United Kingdom general practice, 1988 to 2001, BMC Medicine, 2, 39, 2004. 62. Wing, L., Childhood autism and social class: a question of selection? British Journal of Psychiatry, 137, 410, 1980. 63. Gillberg, C., Infantile autism in children of immigrant parents, A population-based study from Göteborg, Sweden, British Journal of Psychiatry, 150, 856, 1987. 64. Gillberg, C., Schaumann, H., and Gillberg, I.C. Autism in immigrants: children born in Sweden to mothers born in Uganda, Journal of Intellectual Disability Research, 39, 141, 1995. 65. Schopler, E., Andrews, C.E., and Strupp, K., Do autistic children come from upper-middle-class parents? Journal of Autism and Developmental Disorders, 9(2), 139, 1979. 66. Baron-Cohen, S., Saunders, K., and Chakrabarti, S., Does autism cluster geographically? A research note, Autism, 3, 39–43, 1999. 67. IOM (Institute of Medicine), Immunization Safety Review: Vaccines and Autism, Washington, D.C.: National Academic Press, 2004. 68. Parker, S.K., Schwartz, B., Todd, J., Pickering, L.K., Thimerosal-containing vaccines and autistic spectrum disorder: a critical review of published original data, Pediatrics, 114(3): 793–804, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 49 Monday, February 13, 2006 5:02 PM

3 Genetic Basis of Autism Elena Bonora, Janine A. Lamb, Gabrielle Barnby, Anthony J. Bailey, and Anthony P. Monaco CONTENTS Autism: General Characteristics..............................................................................49 Evidence for Autism as a Genetic Disorder ...........................................................50 Genetic Models for Autism Susceptibility..............................................................51 Chromosomal Abnormalities in Autism..................................................................52 Identification of Autism Susceptibility Genes ........................................................53 Linkage Analysis in Autism ....................................................................................54 Candidate Gene Studies for Autism........................................................................57 Candidate Genes on Chromosome 7................................................................58 Candidate Genes on Chromosome 15..............................................................60 Functional Candidate Genes: A Role for Other Neurotransmitters in Autism ............................................................................60 Studies of Other Candidate Genes...................................................................61 Large-Scale Association Studies in Autism ............................................................61 The Use of DNA Array Technology in Autism ......................................................62 Considerations for Complex Disease Gene Mapping.............................................62 Conclusions and Future Perspectives ......................................................................65 Acknowledgments....................................................................................................66 Electronic Resources ...............................................................................................66 References................................................................................................................67

AUTISM: GENERAL CHARACTERISTICS Autism (MIM209850) is a neurodevelopmental disorder characterized by impairments in communication and reciprocal social interaction, and accompanied by restricted and stereotyped behaviors and interests. The sex ratio in autism is heavily skewed toward males, the male to female ratio being approximately 4:1.1 Autism has an onset in the first three years of life and persists throughout adulthood,1 but the clinical picture varies in severity and changes over the course of development. The term autism spectrum disorder (ASD) is now commonly used to refer to idiopathic pervasive developmental disorders, which include childhood autism, atypical autism, Asperger syndrome, and pervasive developmental disorder not otherwise specified (PDDNOS). There is a variable range of symptoms, characterized by 49

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 50 Monday, February 13, 2006 5:02 PM

50

Understanding Autism: From Basic Neuroscience to Treatment

different degrees of impairment in the social and communication domains, and different levels of intelligence and language abilities. Epileptic seizures also develop in approximately one third of individuals with autism.2 The general population prevalence of core autism is approximately 10 in 10,000,3 but when other ASDs are also considered, the prevalence rate may be as high as 60 in 10,000 children.4–6 The most notable aspect of recent epidemiological studies is the increase in the identification of autism cases in studies reported between the mid-1960s and the 1990s, from 4.4 to 12.7 cases per 10,000 in the last decade3. This apparent rise has led to the hypothesis that environmental factors such as vaccines, chemical pollutants, and diet have a role in autism susceptibility. This rising prevalence may be due, however, to increased recognition of the variability in disease expression, a broadening of related diagnostic criteria, and improved public and professional awareness of the disorder. Despite the inclusion of more cases as a result of these changes, ASD diagnosis has become increasingly consistent through the use of standardized diagnostic instruments, such as the Autism Diagnostic Interview Revised (ADI-R)7 and the Autism Diagnostic Observation Schedule (ADOS).8,9

EVIDENCE FOR AUTISM AS A GENETIC DISORDER Autism is an etiologically heterogeneous disorder. Nevertheless, it is now widely accepted that most cases arise because of a complex genetic predisposition. Since autism was first described in 1943,10 hypotheses about its etiology have ranged from the psychological to the biological. During the 1970s and 1980s, however, twin and family studies provided unequivocal evidence for a genetic component to autism susceptibility. In the first three epidemiological same-sex twin studies of autism in same-sex twins,11–13 all twins (one or both of whom were affected with autism) who lived in a geographically defined area were sought out. Such a sampling strategy was used to reduce ascertainment bias, because twins recruited through advertisements yield a biased sample. More monozygotic (MZ) twin pairs than expected for the population frequency and twins concordant for a disorder or trait are likely to participate in such studies. The inclusion of only same-sex twin pairs was used to counter skewing toward greater discordance rates in opposite-sex twin pairs that results from the high male to female ratio in autism. The first U.K. study of twin pairs reported a 36% concordance for autism in monozygotic twins compared to 0% concordance in dizygotic (DZ) twins, suggesting a strong heritable effect for autism susceptibility.11 Among the MZ twin pairs who were discordant for autism, most of their nonautistic counterparts had languagebased learning disabilities, and several were also socially reticent. The social reticence was more striking when the twins were reexamined in adulthood.14 Subsequently, a Scandinavian study13 reported a higher concordance of 96% vs. 0% in MZ and DZ twins respectively, confirming the high heritability of the disorder. In the subsequent U.K. study, the MZ concordance rate was 60%, vs. 0% in DZ twins,12 but when the range of social and language impairments typical of a milder phenotype was considered, the concordance rate was as high as 92% in MZ twins, compared to 10% in DZ twins. These differences in the concordance rates reported between

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 51 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

51

the three studies may reflect differences in ascertainment or diagnostic criteria, or variability in the range of phenotypic expression. At that time, estimates from family studies on the prevalence of autism in siblings of affected children was 4%,15,16 giving rise to a recurrence risk (λs) of 100 times that of the population prevalence. These twin and family studies indicated that the heritability estimate (H) for autism (calculated from the recurrence risk and the MZ:DZ concordance rates) is ≥ 90%, suggesting that autism is one of the most heritable neuropsychiatric disorders. More recent studies have confirmed the high heritability of autism, despite the increased population prevalence rates. It is also recognized that sibling recurrence risk is actually higher (~ 6 to 8%), but the prevalence in siblings is lower (~ 3 to 4%), because of “stoppage rules,” or the tendency of parents of an affected child to defer subsequent pregnancies.17–19 In addition, phenotypic analyses have shown that several characteristics such as social reticence or impairment, communication difficulties, and rigidity are seen more frequently in the relatives of children with autism compared to relatives of children with other neuropsychiatric disorders or controls.15,20 These traits are qualitatively similar to autism but are much milder and are not associated with global developmental delay, suggesting the existence of a broader autistic phenotype with a shared genetic etiology. The delayed onset of speech and difficulty with reading are also more common in family members of individuals with autism, and relatives appear to be at increased risk of other psychiatric disorders including affective disease, particularly major and recurrent depressive illness, and elevated rates of anxiety-related personality traits.21,22

GENETIC MODELS FOR AUTISM SUSCEPTIBILITY Several studies have sought to identify a genetic model for autism susceptibility. The first step in disease model identification is to investigate the pedigrees of affected individuals. Such analyses indicate that idiopathic autism does not follow a simple monogenic mode of inheritance, because relative recurrence risks, although greatly elevated, are much less than would be expected under monogenic inheritance. The difference in pairwise concordance between MZ and DZ twins and the rapid decline in recurrence rate with decreasing genetic relatedness indicates a non-Mendelian, complex mode of inheritance. The falloff in monozygotic to dizygotic twin concordance rates is too steep to be explained by an additive hypothesis, regardless of the number of genes involved, and evidence for multiplicative genetic interaction (epistasis) is provided when risk ratios decrease exponentially across different degrees of relationship, as seen in autism.23 Hence, multiplicative inheritance, imprinting, maternal effects, genes affecting expression rather than susceptibility (modifiers), and allelic and locus heterogeneity have been suggested as possible models for the genetic liability of autism.5,24 It is currently proposed that the genetic model for autistic disorder involves several variants interacting in some manner to produce the clinical phenotype, i.e., a multilocus model with epistasis. Latent class analysis of twin and family data has suggested that as few as 3 to 4 predisposing genes may be implicated,18 although as many as 15 loci have been proposed to be involved.25 In the case of a multilocus

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 52 Saturday, March 18, 2006 10:56 AM

52

Understanding Autism: From Basic Neuroscience to Treatment

model, no single variant is necessary or sufficient per se, and multiple predisposing variants in one or more genes have to be inherited to develop the phenotype. Other factors, such as sex and environmental influences may also influence the severity of phenotypic expression. The multilocus-epistatic model of autism is congruent with the aggregation of features of the broader autism phenotype among first-degree relatives of individuals with autism. This may reflect the possession of only a few predisposing variants, rather than the whole complement necessary to develop the full autism phenotype, or may represent variable expression of the entire complement of genes. This multilocus-epistatic model is different from the traditional concept of genetic heterogeneity, in which mutations in different genes lead to an identical or similar phenotype, and each mutation is by itself sufficient to cause disease. A wellestablished example of genetic heterogeneity is tuberous sclerosis, which has an elevated co-occurrence with autism, in which mutations in two different genes lead to the same phenotype.26,27 The defective genes (TSC1 and TSC2) are located on different chromosomes, and their gene products (hamartin and tuberin, respectively) have been shown to physically associate in vivo and may function as part of the same protein complex.28 It has been proposed that some genetic variants, present in the normal population, increase the risk for a particular disability and are found at a higher frequency among affected individuals. The three areas of impairment in autism — social ability, the development and use of language for communication, and restricted interests — may be measured as quantitative traits distributed throughout the population. For example, several quantitative traits related to autism analyzed by the Broader Phenotype Autism Symptom Scale (BPASS) have been shown to capture the continuum of severity of impairments.29 This implies that statistically more powerful quantitative trait locus (QTL) strategies, widely utilized to identify genes for complex disorders such as dyslexia,30 may be applied successfully to autism research.

CHROMOSOMAL ABNORMALITIES IN AUTISM The identification of chromosomal abnormalities in autism has helped reinforce the view that genetic influences are important in the development of this disorder. Numerous reports in the literature have documented chromosomal aberrations associated with autism31,32; abnormalities of chromosome 15, and structural and numerical abnormalities of the sex chromosomes are most frequently reported. The most prevalent chromosome 15 abnormalities are supernumerary isodicentric chromosome 15 and maternally derived interstitial duplications of the 15q11–q13 region, particularly in individuals with mental retardation and seizures.33,34 Interestingly, the duplicated 15q11–q13 region is the critical region for Angelman syndrome, a neurological disorder in which some of the clinical symptoms resemble autistic behaviors.35 The Angelman syndrome gene, UBE3A,36 is maternally imprinted (expressed only from the paternal allele), suggesting that gene dosage of UBE3A, GABRB3, or other genes in the 15q11–q13 region37 might contribute to the autistic phenotype in these cases.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 53 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

53

There have been several reports of autistic features in females showing whole or partial deletion of one X chromosome. Skuse et al. described individuals with Turner’s syndrome among whom those with a single paternally derived X chromosome had superior verbal and social skills compared to others with a single maternally derived X chromosome.38 This led the authors to suggest that there might be an imprinted locus for social cognition, expressed only from the paternal X chromosome. The imprinted-X liability threshold model hypothesizes that this protective locus may explain the higher threshold of phenotypic expression in females than in males and the higher incidence of autism in males.39 This original report has, however, not been replicated. Reports of linkage of autism to chromosome 7 have led to further reports of cytogenetic abnormalities in this chromosomal region. Ashley-Koch et al. described a family with a paracentric inversion of the long arm of chromosome 7 (Inv(7)(q22–q31.2)) transmitted from an unaffected mother to all three of her children.40 Two of the three siblings in the family were affected by autism, and the third presented a severe expressive language disorder. Warburton and colleagues described two unrelated cases, one with autism and a second with severe expressive language impairment, both of whom showed de novo abnormalities involving breakpoints on chromosome 7q31 (inv(7)(p12.2;q31.3) and t(2;7)(p23;q31.3), respectively).41 Vincent and colleagues characterized a translocation (t(7;13)(q31.2;q21)) transmitted from an unaffected mother to an autistic child and mapped the breakpoint within a highly conserved, brain-expressed gene of unknown function (RAY1).42 Deletions ranging in size from 5 to ≥ 200 kb have been identified in a small number of families at markers D7S630 and D7S517.43 A database of chromosome 7 abnormalities associated with autism and reported in the literature is maintained and updated at The Centre for Applied Genomics (TCAG; http://www.tcag.bioinfo.sickkids.on.ca/). Several other chromosomal abnormalities have been reported in individuals with autism. These include deletions in the region of linkage on chromosome 244,45 and in chromosomal regions 2q37 and 22q1333. Deletions ranging from 5.9 kb to 1.2 Mb have also been described on chromosome 8 at marker D8S624.43 However, the functional significance of these abnormalities in relation to autistic disorders remains unclear. The recent use of comparative genomic hybridization (CGH) array analysis, a powerful molecular cytogenetic technique, enables sensitive, high-resolution detection of alterations of DNA copy number, including aneuploidies and submicroscopic imbalances, on a previously unmanageable scale.46

IDENTIFICATION OF AUTISM SUSCEPTIBILITY GENES With no clear clues as to the underlying nature of the genetic liability for development of idiopathic autism, a first step towards identifying susceptibility genes is mapping the disease to a specific subchromosomal location. In the absence of any chromosomal abnormalities or animal models of the disease that may indicate the region of the disrupted gene, this localization is usually carried out by linkage analysis, followed by higher-resolution association mapping. This enables the identification of suitable candidate genes within the chromosomal interval, based on knowledge

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 54 Monday, February 13, 2006 5:02 PM

54

Understanding Autism: From Basic Neuroscience to Treatment

of the disease phenotype and pathology, or a structural or functional relationship with other genes known to be involved in similar disease phenotypes. Variants within these candidate genes can subsequently be tested for association with the disease, under the assumption of the “common disease common variant” hypothesis.

LINKAGE ANALYSIS IN AUTISM Several groups have adopted a whole genome screen approach using affected sibling pairs to identify autism susceptibility loci. Traditionally, this involves genotyping 300 to 400 highly polymorphic microsatellite markers evenly spaced throughout the genome (Figure 3.1). Linkage information is subsequently used to calculate two-point LOD scores at each marker examined and to generate multipoint LOD score profiles along the 23 chromosomes. The peaks in these profiles represent a probability measure for increased sharing between affected sibling pairs (ASP).

Genomic DNA preparation

DNA masterplate

Automatic pre-PCR reagent handling

PCR amplification of polymorphic markers

Automated sampling

Marker pooling

DNA plating

Genetic analyser

268 272

Data collection

Data analysis

FIGURE 3.1 Schematic overview of whole genome screen linkage approach using polymorphic microsatellite markers. Genomic DNA is prepared for each individual from whole blood, cell lines, or buccal swabs and put into a masterplate before plating out. PCR amplification is carried out for each individual DNA sample using 300 to 400 polymorphic microsatellite markers across the genome. The PCR products are entered into a capillary or gel-based genetic analyzer, and the resulting data collected and analyzed.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 55 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

55

If significant or otherwise supported by independent studies, such peaks may identify the location of genetic risk factors.47 An ongoing debate in the study of complex traits is the question of what represents significant linkage in a whole genome screen, as analysis of genotype data involves multiple statistical tests and an unknown number of loci. Traditionally, a LOD score threshold of 3 is accepted as significant evidence supporting linkage for a monogenic disorder. A whole genome screen threshold of 3.6 for ASPs (resulting in a 5% false-positive rate in the whole genome) has been proposed.48 Empirical determination of significance, based on data randomization and permutation testing, is an alternative to using a common threshold. However, the proposed thresholds are often not achievable owing to limitations in sample size and corresponding power. A typical study may involve 100 to 200 sib pairs and has only moderate power when a given disease locus exerts modest genetic effect. The power of a sample is dependent on the contribution each locus makes to the overall genetic variation and to the extent of potential interaction between two or more loci. Adopting a higher threshold reduces the ability to detect linkage. Power limitations and confounding factors such as locus heterogeneity have led to the creation of guidelines for suggestive linkage of a multipoint MLS > 2.2.48 It has also been suggested that it may be more efficient to perform a staged genomic screen to avoid false-negatives in the detection of loci of small effect in a moderate-sized population. By initially applying a lower threshold for linkage, a second stage is performed, typing additional families and genetic markers in those regions showing suggestive linkage in the first stage. This allows for a distinction to be made between convincing evidence for linkage and regions of interest to help ensure that loci with weaker effects are not missed.49 To date, nine genomewide scans for autism susceptibility loci have been carried out, identifying a number of chromosomal regions of interest.25,50–59 In the first genome screen, on a sample of 98 affected relative pairs, the International Molecular Genetic Study of Autism Consortium (IMGSAC) reported a region on chromosome 7q as generating the most significant evidence for linkage, with an MLS of 3.55 in a subset of 56 affected sib-pair families in the U.K., and an MLS of 2.53 in all 86 affected sib-pair families.53 A second-stage study by IMGSAC, including additional families and markers, supported initial linkage results on chromosomes 7q and 16p, and identified two additional regions on chromosomes 2q and 17q with a multipoint MLS greater than 2.54, 60 In total, other linkage studies have identified suggestive linkage on chromosomes 6,56 7, and 1352; 1,25 2,51 5, 19, and X55; X,57 1, 3, 7, and X50; 5, 11, and 1758; and 17 and 19.59 Overlapping results across studies have identified a few common regions of linkage (Figure 3.2). The region on chromosome 2q reported by IMGSAC54 with a multipoint MLS of 3.72 reaches genomewide significance and overlaps that identified by Buxbaum et al.51 with a maximum multipoint, nonparametric linkage score of 2.39. Similarly, for the locus on chromosome 7q identified by IMGSAC,60 regions have also been identified by the Collaborative Linkage Study of Autism with a maximum multipoint heterogeneity LOD of 2.2,52 and by Ashley-Koch et al.40 Furthermore, most studies report some degree of increased sharing in this region of chromosome 7q, designated AUTS1. The region on chromosome 1q identified by the Finnish study with a maximum multipoint parametric LOD score of 2.63 under a dominant model50

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 56 Monday, February 13, 2006 5:02 PM

56

Understanding Autism: From Basic Neuroscience to Treatment

1

2

3

4

5

6

7

8

9

10

11

14

15

16

17

18

19

20

21

22

X

Y

12

13

MLS > 3.6 MLS > 2.2 MLS > 1

FIGURE 3.2 Ideogram showing the results of genome screens for autism susceptibility loci. (Reproduced with permission from Barnby G. and Monaco A.P., Strategies for autism candidate gene analysis, Novartis Found Symp, 2003  John Wiley & Sons.)

overlaps with that previously reported by Risch et al.25 with a multipoint maximum LOD score of 2.15. In addition, the locus on chromosome 17q identified by the IMGSAC54 has also been reported.58,59 Recently, the first genomewide scan for Asperger syndrome loci61 showed several regions of interest on chromosomes 1q, 3p, and 13q, which overlapped with previously reported regions of linkage for autism. Nominal evidence for linkage was also found for several markers on chromosome 7q. Multiple factors, including ascertainment differences, inclusion–exclusion criteria, differing analytical approaches, marker maps and density, and varying sample sizes may complicate comparison of results between different groups. A preliminary meta-analysis of the first four genome scans revealed that the regions on chromosome 7q and distal 13q were suggestive of linkage by genomewide criteria.62 It is worth noting that if there are as many as 15 or more loci, each of small effect, contributing to autism susceptibility,25 a lack of replication does not permit the exclusion of a particular locus. The nature of genomewide linkage screening methods tends to indicate a relatively large genetic interval in the range of tens of centimorgans (cM). Furthermore, computer simulations have shown that for complex traits, linkage peaks resulting from different samples may vary from 10 to 30 cM.63 It is encouraging that there is convergent evidence for linkage to several chromosome regions. However, the regions identified are large (for example, chromosome 7q), and in the absence of replicated association findings, investigators have tried to minimize clinical and genetic heterogeneity by examining phenotypic subsets, for example, the presence of obsessive-compulsive behaviors or language delay, to improve localization of the linkage signals.51,64–66 Similarly, the male preponderance in autism suggests involvement of sex-specific factors in autism susceptibility.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 57 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

57

However, there is little evidence from genetic studies to suggest the involvement of genes on the X chromosome in the majority of individuals with autism. Recent analysis, according to the sex of affected sib pairs, suggests that linkage to previously identified regions shows a specific sex-related effect. Linkage to chromosome 16 in the IMGSAC sample is limited to male affected sibling pairs; the same trend is observed on chromosome 7 but narrowly fails to reach significance.67 Linkage to chromosome 15q in this sample is limited to the nonmale ASP, but it is not yet clear whether the cytogenetic abnormalities of 15q11-q13 occur at an increased frequency in autistic females. A male-specific effect has also been identified for the locus on chromosome 17 using a similar linkage analysis approach.68 Linkage dependent on the sex of affected individuals has been reported previously in studies of other complex diseases including inflammatory bowel disease, hypertension, diabetes, and personality trait neuroticism.69–72 These results support the utility of this approach and suggest that it may be useful for other neurodevelopmental disorders in which a similar sex bias is observed, although, as with all analyses of sample subsets, statistical power is reduced because of the smaller sample size. The continued collection of larger family samples is necessary to increase power for detecting loci of small effect and discriminating true from false positives. The increasing number of publicly available autism family collections, such as the Autism Genetic Resource Exchange (AGRE)73 and the Coriell Autism Research Resource (http://locus.umdnj.edu/autism), may prove useful in this respect. The utility of expanding collections of affected relative pair families has been demonstrated in terms of strengthening putative true linkages, identifying linkages that were not originally detected in the smaller family sample, and reducing “noise” levels across the genome.53–55,58 Several studies have looked for evidence of epistatic interactions between unlinked putative autism susceptibility loci.55,59,74,75 It has been suggested that this may be a more powerful approach than single-locus analysis offering improved localization for the identification of genes for complex disease.76–78

CANDIDATE GENE STUDIES FOR AUTISM The linkage approach undertaken in the last decade has identified several chromosomal regions of interest for autism susceptibility. However, it has not, so far, significantly narrowed the chromosomal intervals for autism susceptibility genes. For this reason many groups have adopted a candidate gene screening and association strategy for genes mapping to the subchromosomal locations defined by linkage analysis. This approach also enables researchers to test the alternative “common disease multiple rare variants” hypothesis and to ask whether allelic heterogeneity may be present at a given locus. By selecting affected individuals contributing to linkage in a given region, it is possible to enrich a smaller subsample for molecular studies that seek to identify susceptibility variants in genes within this region. One may screen coding, promoter, and noncoding conserved sequences at candidate genes to identify variants of potential functional significance relating to disease.79 The availability of the complete human genome sequence80,81 has greatly facilitated the ease and efficiency of screening candidate genes. Scientists have direct

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 58 Monday, February 13, 2006 5:02 PM

58

Understanding Autism: From Basic Neuroscience to Treatment

35 30 25 20 15 10 5 0 1995

1996

1997

1998

1999 2000 Year

2001

2002

2003

2004

(a)

FIGURE 3.3 Candidate gene screening and association studies for autism susceptibility genes over the last decade. (a) Number of publications; (b) position of genes studied across all chromosomes from pter to qter with number of studies of each gene shown on the y-axis. For clarity, gene names for only genes with more than five publications are indicated. Positions are from Build 35 of the Human Reference Sequence. (From PubMed search terms “gene” OR “association” AND “autis*” in Title/Abstract since 1/1/1995. This excludes isolated case studies and reports of karyotypic abnormalities.)

access to databases on the Internet (for example, the Ensembl Genome Browser, UCSC Genome Bioinformatics, NCBI), where the annotated genome sequence is constantly curated and updated. A survey of the existing literature on PubMed (using the search terms gene OR association AND autism) indicates that many such studies have been reported over the last decade (since January 1, 1995) in over 100 publications and with over 100 candidate genes examined (Figure 3.3). An exhaustive report of all the candidate gene studies in autism is beyond the scope of this chapter, and the reader is directed to reviews elsewhere.33,34,82

CANDIDATE GENES

ON

CHROMOSOME 7

The concordance of the linkage findings on chromosome 7 has recently made this region the focus of several candidate gene studies. An investigation of WNT2, a member of the WNT gene family influencing the development of the central nervous system, identified two families carrying nonconservative sequence variants that segregated with autism,83 but other studies have failed to replicate these findings.84,85 Another gene of interest on chromosome 7q encodes Reelin (RELN), which is involved in the migration of cortical and hippocampal neurons and cerebellar Purkinje cells. Postmortem studies of the brains of individuals with autism have shown reduced Purkinje cell numbers,86,87 and, interestingly, a progressive loss of Purkinje cells has been reported in the cerebellum of male heterozygous reeler mice,

© 2006 by Taylor & Francis Group, LLC

SLC6A4 12

Number of studies

10 HOXA1 8

RELN GABRB3 UBE3A

6

HOXB1

MECP2 FMR1

4 2 0

1

2

3

4

5

6

7

8 9 Chromosome

10

11

12

13

14

15 16 17 18 19 20 21 22 23

(b)

FIGURE 3.3 (Continued).

59

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 59 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

14

2732_C003.fm Page 60 Monday, February 13, 2006 5:02 PM

60

Understanding Autism: From Basic Neuroscience to Treatment

whereas the females are spared.88 Three studies have shown preferential transmission of specific alleles of a repeat in the 5 untranslated region of the RELN gene in families with autism,89–91 although negative results have also been reported.84, 92, 93 Screening the coding sequence of this gene for autism susceptibility variants in a large sample did not reveal any putative etiological variants with a high enough frequency to explain the strength of the linkage findings.94 Recently, evidence for association of variants in the laminin beta 1 gene (LAMB1) with autism has been shown.95,96 Association results also have been reported for alleles at the gamma catalytic subunit of the phosphatidyl 3-OH-kinase gene (PIK3CG) on 7q22,97 and three homeobox genes: HOXA1 and HOXB1 (on 7p and 7q, respectively98,99) and engrailed-2 (EN2) on 7q36.100–102 These results are difficult to interpret as they have not been replicated in independent samples.84,103,104

CANDIDATE GENES

ON

CHROMOSOME 15

Reports of cytogenetic abnormalities of chromosome 15q11-q13 have also put this region at the center of investigations to identify susceptibility genes for autism. A cluster of GABAA (γ-aminobutyric acid) receptor genes is located in this region (GABRB3, GABRA5, and GABRG3). GABA is the predominant inhibitory neurotransmitter in the central nervous system (CNS), and a decrease in glutamic acid decarboxylase, which synthesizes GABA from its precursor glutamate, has been observed in the brain tissue from individuals with autism.105 Several studies have reported association of autism with marker alleles in the GABAA receptor gene cluster, although other studies relying predominantly on microsatellite markers have not found evidence for association (see, for example106–110) Similarly, studies of the maternally expressed genes UBE3A111,112 and ATP10C113,114 in the same cytogenetic region have generated conflicting results. The recent detection of the reduced expression of both UBE3A and GABRB3 in autism and Rett syndrome brain tissue samples further supports the possibility of epigenetic dysregulation affecting this region and raises the possibility that alterations in both UBE3A and GABRB3 could be involved in autism.37

FUNCTIONAL CANDIDATE GENES: A ROLE NEUROTRANSMITTERS IN AUTISM

FOR

OTHER

The potential role of the serotonin system in autism has led to several analyses of the serotonin transporter gene (SLC6A4/HTT) on chromosome 17q. Association with alleles at a functional insertion/deletion polymorphism (HTTLPR) in the promoter that affects gene expression has been detected in several studies. There are contrary findings, with association to either short (S) or long (L) alleles, or an absence of association.109,115–119 Additional polymorphisms across the gene have also shown association.120 The 5-HT7121 and 5-HT2A122 serotonin receptor genes have also been studied, but the results are conflicting. Glutamate is an important excitatory neurotransmitter within the CNS. Studies of glutamate receptor genes in autism have been reported, including association of the muscarinic glutamate receptor 6 (GluR6 or GRIK2) gene on chromosome

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 61 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

61

6q21123,124 and that of the metabotrophic glutamate receptor 8 (GRM8) gene on chromosome 7q31.125 Replication of these results has yet to be reported.

STUDIES

OF

OTHER CANDIDATE GENES

The increased co-occurrence of autism, in a minority of cases, with disorders of known genetic etiology, principally fragile X syndrome, tuberous sclerosis, and neurofibromatosis type 1,1 has led to studies of these genes in autism. An association at the phenotypic level may suggest disruption to common neurobiological pathways. Several studies have reported analysis of candidate genes on chromosome 2 in which different groups have reported linkage with autism.51,54 New coding variants, not found in large control groups, have been identified in the GEFII/ATF and TBR1 genes on chromosome 2q34,126 but the low frequency of these coding changes does not offer a clear explanation of the role of these genes in autism susceptibility. Positive association has also been presented for the inositol polyphosphate-1phosphatase (INPP1) gene on 2q3297 and the mitochondrial aspartate/glutamate carrier SLC25A12 on 2q31,127 although replication of these results is awaited. Mutations of the methyl-CpG-binding protein-2 (MECP2) gene on chromosome Xq28 are responsible for the pervasive developmental disorder Rett syndrome.128, 129 Screening the coding sequence of this gene for variation has lead to the identification of mutations in a small minority of patients with autism.130–134 Jamain et al. reported the identification of mutations in the X-linked neuroligin genes, NLGN3 on Xq13 and NLGN4 on Xp22.3. A missense change in a highly conserved position was identified in NLGN3 in two affected brothers, and a frameshift insertion in NLGN4, producing a premature stop codon was found in two other male siblings.135 Recently, a novel mutation in NLGN4 was detected in a large pedigree with individuals affected by mental retardation or autism, showing that this change is not specific to the autistic phenotype.136 Screening these genes in further patient samples has generated conflicting rather than conclusive results.137–139 It is clear from the preceding overview that many genes are being tested for their contribution to autism susceptibility. Until positive findings are reliably replicated or refuted in large, independent samples, the significance of these findings is difficult to determine.

LARGE-SCALE ASSOCIATION STUDIES IN AUTISM A common strategy currently used for mapping disease genes in complex disorders is to follow up results from linkage analysis with association mapping in refined genomic intervals for more accurate localization. Inheritance of chromosomal intervals absent from ancestral recombination results in particular multiallelic combinations, or haplotypes, in the population. Such multiallelic combinations are in linkage disequilibrium (LD).140 Indirect association studies of genetic variants (markers) use the principle that markers in close proximity to a disease-risk variant will be more often co-inherited than expected under independent assortment because they are in linkage disequilibrium. Population-based association studies typically examine specific alleles at a marker locus that are more frequent in affected individuals (cases)

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 62 Monday, February 13, 2006 5:02 PM

62

Understanding Autism: From Basic Neuroscience to Treatment

than in the unaffected population (controls),141 but this method may be prone to false-positives owing to population admixture. An alternative method is based on comparing allele frequencies in cases using nontransmitted alleles from their parents as family-based controls. This is implemented in the transmission disequilibrium test (TDT). The TDT requires that a marker be both in linkage disequilibrium with the disease variant and that it segregate with disease status to be significant.142 As a result, the TDT is unlikely to identify spurious associations due to very recent population admixture or population samples that are stratified with respect to genetically differentiated groups. Several studies have shown the existence of a series of high-LD regions separated by short, discrete segments of very low LD, typically defined as recombination hotspots.143,144 These regions of high LD, known as haplotype blocks, exhibit limited haplotype diversity, so that a small number of distinct haplotypes account for most of the genetic variation in a population within that block. The presence of regions with high LD could benefit association studies because recombination sites would delimit the boundaries for candidate gene analysis and indicate how far to extend the search for functional variants.145 These ideas have prompted the International Haplotype Mapping (HapMap) Consortium, a genomewide catalog of common polymorphisms — principally singlenucleotide polymorphisms (SNPs) — and haplotype blocks in multiple human populations (Figure 3.4). The HapMap Consortium and other SNP discovery resources allow genetic researchers access to approximately 5 million SNPs with varying degrees of validation (dbSNP, Build 124). Traditionally, association studies have been limited to the investigation of a small number of variants or candidate genes. However, these recent advances in technology and bioinformatics facilitate extensive association studies of complex disorders, such as autism, enabling much higher throughput in terms of the numbers of variants and samples tested, although the fruits of these endeavors have yet to be reported.

THE USE OF DNA ARRAY TECHNOLOGY IN AUTISM Advances in DNA microarray-based technology enable scientists to now conduct high-throughput SNP genome screens for complex disorders and also provide an opportunity to measure the parallel expression levels of thousands of candidate genes in cases relative to controls.46,146 Obtaining samples suitable for expression analysis remains problematic for autism because of its neurobiological nature and developmental profile, but it remains a promising means of identifying and investigating susceptibility genes.

CONSIDERATIONS FOR COMPLEX DISEASE GENE MAPPING The lack of reproducibility and difficulty in interpreting candidate gene studies may be due to similar problems as those encountered in linkage analysis: differences in disease ascertainment and inclusion–exclusion criteria, population differences, clinical and genetic heterogeneity and phenocopies, publication or reporting bias, and

© 2006 by Taylor & Francis Group, LLC

Genotyped SNPS

Gene location

© 2006 by Taylor & Francis Group, LLC

63

FIGURE 3.4 The International HapMap project. This genomewide catalog of common polymorphisms (principally singlenucleotide polymorphisms [SNPs]) and haplotype blocks in multiple human populations provides researchers with a powerful resource for candidate gene screening and association studies.

2732_C003.fm Page 63 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

Sequence coverage

2732_C003.fm Page 64 Monday, February 13, 2006 5:02 PM

64

Understanding Autism: From Basic Neuroscience to Treatment

a large number of statistically underpowered studies. The complex genetics underlying autism susceptibility might also make variants difficult to identify. In a disorder of complex genetic architecture, as described previously, no single locus contains alleles that are either necessary or sufficient to cause disease, and it is not known if the genetic variance underlying autism susceptibility is due to a small number of loci where alleles are common (the “common disease common variant” hypothesis) or due to a larger number of loci with multiple rare alleles of independent origin.147 The success of association studies critically depends on the degree of genetic heterogeneity (both allelic and locus) underlying the disease phenotype, as these methods are based on the assumption that a few penetrant or common susceptibility alleles are present in the population at the disease locus. It is likely that in the presence of multiple etiological variants, association studies may have limited power and only a fraction of disease-causing SNPs may be identified by this approach. Marker density and the degree of LD between a causative mutation and the marker loci are also important determinants of power in association studies.148,149 The expected decrease in LD with recombination distance is rapid, suggesting that association studies will not be able to detect a relatively old disease-associated allele that is more than 5 to 10 kb away from the variant tested.150 Allele frequency can also be important in determining whether association will be detected. For common alleles, marker-disease allele frequency and LD relationship become crucial, and even moderate deviations from the optimal model affect the successful detection of association, as shown by the ApoE4 allele and Alzheimer’s disease151. For infrequent variants (frequency < 0.1), the power to detect association is also related to effect size. For rare variants of large effect size, as found in Crohn’s disorder, marker-disease allele frequency discrepancies and moderate LD have relatively less impact on the ability to detect association,152 but if the effect size is small, available sample sizes and association designs may have insufficient power.148 The presence of disease-associated alleles of moderate effect in complex disorders is suggested by the poor reproducibility of the linkage signals. Although linkage signals are a function of both relative risk and allele frequency, even rare mutations are identified in Mendelian disorders because of their high relative risk. Assumptions about the underlying disease allele frequencies depend on their impact on reproductive fitness and the degree of selection that has acted on them in the past. Early-onset, severe diseases may have lower frequency alleles compared with later onset or milder diseases, such as autism, because the etiological variants are less subject to negative selection. However, selecting SNPs with moderate-torare minor allele frequencies affects the power of the TDT to detect association because more families are uninformative. Hence, the variants selected for association studies should be chosen carefully. In summary, markers in strong LD with the disease allele and with similar frequencies have the highest power to detect association. However, a negative association result in a particular genomic region does not exclude a significant gene effect in the region if these conditions are not met.153

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 65 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

65

If the number of independent susceptibility variants with small penetrance is high, the frequency of the disease alleles may be very similar in affected and control individuals. In these circumstances, analysis of haplotypes has more power to detect association than analysis of single variants.154 Affected individuals with at least one affected sibling on average carry more copies of the disease allele than singleton cases, resulting in increased power.155 In addition, selecting families that show evidence of linkage to a region has been suggested,156 and choosing affected siblings who show the most evidence for pairwise allele sharing with the other affected siblings in the family has been shown to increase statistical power.157 If a disease is caused by multiple, rare susceptibility alleles of independent origin, their combined frequency in samples enriched for disease-related alleles is expected to be significantly different from that in controls; screening candidate genes in these samples can be useful for the identification of variants with a role in disease etiology.141 However, despite recent technological and bioinformatic advances, it remains difficult to identify good candidates for a complex disease as current knowledge of the biological pathways involved in the disease is still largely incomplete. Improvements in our knowledge and understanding of the human genome sequence have revealed unexpected levels of complexity in terms of gene transcription, processing, and regulation. The possible involvement of epigenetic variation rather than coding polymorphisms may explain the difficulty in identifying autism-related variants in candidate genes to date.

CONCLUSIONS AND FUTURE PERSPECTIVES A wealth of linkage, and cytogenetic and candidate gene studies have implicated several regions of the genome that may harbor autism susceptibility genes. It is likely that autism may be caused by a combination of common and rare alleles, and a multidisciplinary range of approaches is needed to identify these variants. Recently, large multinational collaborations such as the International Human Genome Sequencing Consortium, the HapMap project, and the ENCODE project to identify all functional elements of the human genome in different populations have been formed on an unprecedented scale. Large collaborative samples of multiplex families with autism, for example, the National Alliance for Autism Research (NAAR)-organized Autism Genome Project, have also been established. Access to these resources will assist in the search for complex disease genes. The ongoing development of novel, more cost-effective, and more accurate high-throughput genotyping and sequencing techniques provides a formidable amount of information that can be used in the detailed investigation of potential etiological variants, and has prompted researchers to move toward genome wide SNP screens and exhaustive association studies across the most significant chromosomal regions. The use of comparative genomics and, in particular, the completed genomic sequences of the mouse and chimpanzee will also prove to be powerful tools in the interpretation of the human genome sequence. Sequence comparisons can be used to reveal regions of high similarity, identifying conserved coding regions and

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 66 Monday, February 13, 2006 5:02 PM

66

Understanding Autism: From Basic Neuroscience to Treatment

regulatory sequences. Identifying and mapping homologous genes in different species will also aid in the prediction of likely gene function. Once the risk haplotypes for autism susceptibility are identified, the mechanisms by which they affect gene or protein function and the regulation of gene expression can be studied. The discovery of functionally relevant SNPs will aid the identification of upstream targets such as transcription factors. In the absence of variants with obvious functional effects, effects on the level of transcriptional regulation can be assessed by quantitative analysis of allele-specific effects.158 Alternatively, genomicbased expression studies in human cell lines can be used to dissect gene regulation and determine whether promoter or other regions are responsible for any expressionlevel differences. Detailed functional analysis of autism susceptibility genes can then be undertaken. This work is essential for understanding the mechanism of pathogenesis in autism and will guide studies of neuropathology and the development of potential animal models. The phenotypic impact of susceptibility genes can be assessed through the behavioral, neuroanatomical, and neurophysiological study of mouse mutants mapped to syntenic regions. Identifying downstream targets of susceptibility genes will yield new genes to be tested in association studies and will lead to a better understanding of the genetic pathways affecting brain development in autism and other related disorders. The parallel development of increasingly sophisticated diagnostic measures and the possible classification of autism endophenotypes will lead to more sophisticated quantitative trait analyses of the disorder. This approach should result in substantially more power to better disentangle the genetic variability underlying typical autism impairments. Over the past years, autism researchers have become increasingly aware of the limitations of candidate gene analysis with previous, somewhat naive, approaches based on unknown numbers and effects of gene variants and levels of LD. Advances in understanding the structure of the human genome and solving technological problems will, in turn, lead to new issues of data handling and statistical analysis, and generate future ethical dilemmas. However, it is hoped that promising new technologies will have a positive impact on elucidating the genetic basis of this devastating disorder, which continues to elude the medical research community.

ACKNOWLEDGMENTS This work is supported by the U.K. Medical Research Council, the Wellcome Trust, the NLM Foundation, and EC Fifth Framework (QLG2-CT-1999-0094). We thank two reviewers for their helpful comments on the manuscript.

ELECTRONIC RESOURCES The Centre for Applied Genomics (TCAG) http://chr7.org Coriell Autism Research Resource http://locus.umdnj.edu/autism Ensembl Genome Browser http://www.ensembl.org/ UCSC Genome Bioinformatics http://genome.ucsc.edu/ HapMap project http://www.hapmap.org/ ENCODE project www.genome.gov/10005107

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 67 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

67

REFERENCES 1. Volkmar, F.R., Lord, C., Bailey, A., Schultz, R.T., and Klin, A., Autism and pervasive developmental disorders, J Child Psychol Psychiatry 45, 135–170, 2004. 2. Rutter, M., Autistic children: infancy to adulthood, Semin Psychiatry 2, 435–350, 1970. 3. Fombonne, E., Epidemiological surveys of autism and other pervasive developmental disorders: an update, J Autism Dev Disord 33, 365–382, 2003. 4. Fombonne, E., The epidemiology of autism: a review, Psychol Med 29, 769–786, 1999. 5. Szatmari, P., Heterogeneity and the genetics of autism, J Psychiatry Neurosci 24, 159–65, 1999. 6. Chakrabarti, S. and Fombonne, E., Pervasive developmental disorders in preschool children, JAMA 285, 3093–3099, 2001. 7. Lord, C., Rutter, M., and Le Couteur, A., Autism diagnostic interview-revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders, J Autism Dev Disord 24, 659–85, 1994. 8. Lord, C. et al., The autism diagnostic observation schedule-generic: a standard measure of social and communication deficits associated with the spectrum of autism, J Autism Dev Disord 30, 205–223, 2000. 9. Lord, C. et al., Autism diagnostic observation schedule: a standardized observation of communicative and social behavior, J Autism Dev Disord 19, 185–212, 1989. 10. Kanner, L., Autistic disturbances of affective content, Nerv Child 2, 217–250, 1943. 11. Folstein, S. and Rutter, M., Infantile autism: a genetic study of 21 twin pairs, J Child Psychol Psychiatry 18, 297–321, 1977. 12. Bailey, A. et al., Autism as a strongly genetic disorder: evidence from a British twin study, Psychol Med 25, 63–77, 1995. 13. Steffenberg, S. et al., A twin study of autism in Denmark, Finland, Iceland, Norway and Sweden, J Child Psychol Psychiatry 30, 405–416, 1989. 14. Le Couteur, A. et al., A broader phenotype of autism: the clinical spectrum in twins, J Child Psychol Psychiatry 37, 785–801, 1996. 15. Bailey, A., Palferman, S., Heavey, L., and Le Couteur, A., Autism: the phenotype in relatives, J Autism Dev Disord 28, 369–392, 1998. 16. Bolton, P. et al., A case-control family history study of autism, J Child Psychol Psychiatry 35, 877–900, 1994. 17. Jones, M.B. and Szatmari, P., A risk-factor model of epistatic interaction, focusing on autism, Am J Med Genet 114, 558–565, 2002. 18. Pickles, A. et al., Latent-class analysis of recurrence risks for complex phenotypes with selection and measurement error: a twin and family history study of autism, Am J Hum Genet 57, 717–726, 1995. 19. Szatmari, P. and Jones, M.B., Effects of misclassification on estimates of relative risk in family history studies, Genet Epidemiol 16, 368–381, 1999. 20. Bailey, A., Phillips, W., and Rutter, M., Autism: towards an integration of clinical, genetic, neuropsychological, and neurobiological perspectives, J Child Psychol Psychiatry 37, 89–126, 1996. 21. Piven, J. and Palmer, P., Psychiatric disorder and the broad autism phenotype: evidence from a family study of multiple-incidence autism families, Am J Psychiatry 156, 557–563, 1999. 22. Lord, C. and Bailey, A., Child and Adolescent Psychiatry, Ed., Rutter, M.T., Blackwell Publishing, 2002, pp. 636–663.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 68 Monday, February 13, 2006 5:02 PM

68

Understanding Autism: From Basic Neuroscience to Treatment 23. Risch, N., Linkage strategies for genetically complex traits. I. Multilocus models, Am J Hum Genet 46, 222–228, 1990. 24. Cook, E.H., Jr. et al., Autism or atypical autism in maternally but not paternally derived proximal 15q duplication, Am J Hum Genet 60, 928–934, 1997. 25. Risch, N. et al., A genomic screen of autism: evidence for a multilocus etiology, Am J Hum Genet 65, 493–507, 1999. 26. van Slegtenhorst, M. et al., Identification of the tuberous sclerosis gene TSC1 on chromosome 9q34, Science 277, 805–808, 1997. 27. The European Chromosome 16 Tuberous Sclerosis Consortium, Identification and characterization of the tuberous sclerosis gene on chromosome 16, Cell 75, 1305–1315, 1993. 28. van Slegtenhorst, M. et al., Interaction between hamartin and tuberin, the TSC1 and TSC2 gene products, Hum Mol Genet 7, 1053–1057, 1998. 29. Sung, Y.J. et al., Genetic investigation of quantitative traits related to autism: use of multivariate polygenic models with ascertainment adjustment, Am J Hum Genet 76, 68–81, 2005. 30. Francks, C. et al., A 77-kilobase region of chromosome 6p22.2 is associated with dyslexia in families from the United Kingdom and from the United States, Am J Hum Genet 75, 1046–1058, 2004. 31. Veenstra-Vanderweele, J., Christian, S.L., and Cook, E.H., Jr., Autism as a paradigmatic complex genetic disorder, Annu Rev Genomics Hum Genet 5, 379–405, 2004. 32. Gillberg, C., Chromosomal disorders and autism, J Autism Dev Disord 28, 415–425, 1998. 33. Cook, E.H., Jr., Genetics of autism, Child Adolesc Psychiatr Clin N Am 10, 333–350, 2001. 34. Lamb, J.A., Moore, J., Bailey, A., and Monaco, A.P., Autism: recent molecular genetic advances, Hum Mol Genet 9, 861–868, 2000. 35. Clayton-Smith, J. and Laan, L., Angelman syndrome: a review of the clinical and genetic aspects, J Med Genet 40, 87–95, 2003. 36. Matsuura, T. et al., De novo truncating mutations in E6-AP ubiquitin-protein ligase gene (UBE3A) in Angelman syndrome, Nat Genet 15, 74–77, 1997. 37. Samaco, R.C., Hogart, A., and LaSalle, J.M., Epigenetic overlap in autism-spectrum neurodevelopmental disorders: MECP2 deficiency causes reduced expression of UBE3A and GABRB3, Hum Mol Genet 14, 483–492, 2005. 38. Skuse, D.H. et al., Evidence from Turner's syndrome of an imprinted x-linked locus affecting cognitive function, Nature 387, 705–708, 1997. 39. Skuse, D.H., Imprinting, the X-chromosome, and the male brain: explaining sex differences in the liability to autism, Pediatr Res 47, 9–16, 2000. 40. Ashley-Koch, A. et al., Genetic studies of autistic disorder and chromosome 7, Genomics 61, 227–236, 1999. 41. Warburton, P. et al., Support for linkage of autism and specific language impairment to 7q3 from two chromosome rearrangements involving band 7q31, Am J Med Genet 96, 228–234, 2000. 42. Vincent, J.B. et al., Identification of a novel gene on chromosome 7q31 that is interrupted by a translocation breakpoint in an autistic individual, Am J Hum Genet 67, 510–514, 2000. 43. Yu, C.E. et al., Presence of large deletions in kindreds with autism, Am J Hum Genet 71, 100–115, 2002. 44. Gallagher, L. et al., Brief report: a case of autism associated with del(2)(q32.1q32.2) or (q32.2q32.3), J Autism Dev Disord 33, 105–108, 2003.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 69 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

69

45. Borg, I. et al., A cryptic deletion of 2q35 including part of the PAX3 gene detected by breakpoint mapping in a child with autism and a de novo 2;8 translocation, J Med Genet 39, 391–399, 2002. 46. Wang, N.J., Liu, D., Parokonny, A.S., and Schanen, N.C., High-resolution molecular characterization of 15q11-q13 rearrangements by array comparative genomic hybridization (array CGH) with detection of gene dosage, Am J Hum Genet 75, 267–281, 2004. 47. Strachan, T. and Read, A.P., Human Molecular Genetics 2, BIOS Scientific Publishers, Oxford, U.K., 2001. 48. Lander, E. and Kruglyak, L., Genetic dissection of complex traits: guidelines for interpreting and reporting linkage results, Nat Genet 11, 241–247, 1995. 49. Weeks, D.E. and Lathrop, G.M., Polygenic disease: methods for mapping complex disease traits, Trends Genet 11, 513–519, 1995. 50. Auranen, M. et al., A genomewide screen for autism-spectrum disorders: evidence for a major susceptibility locus on chromosome 3q25-27, Am J Hum Genet 71, 777–790, 2002. 51. Buxbaum, J.D. et al., Evidence for a susceptibility gene for autism on chromosome 2 and for genetic heterogeneity, Am J Hum Genet 68, 1514–1520, 2001. 52. CLSA, An autosomal genomic screen for autism, Am J Med Genet 88, 609–615, 1999. 53. IMGSAC, A full genome screen for autism with evidence for linkage to a region on chromosome 7q, Hum Mol Genet 7, 571–578, 1998. 54. IMGSAC, A genomewide screen for autism: strong evidence for linkage to chromosomes 2q, 7q, and 16p, Am J Hum Genet 69, 570–581, 2001. 55. Liu, J. et al., A genomewide screen for autism susceptibility loci, Am J Hum Genet 69, 327–340, 2001. 56. Philippe, A. et al., Genome-wide scan for autism susceptibility genes, Paris Autism Research International Sibpair Study [published erratum appears in Hum Mol Genet 8(7): 1353, July 1999], Hum Mol Genet 8, 805–812, 1999. 57. Shao, Y. et al., Genomic screen and follow-up analysis for autistic disorder, Am J Med Genet 114, 99–105, 2002. 58. Yonan, A.L. et al., A genomewide screen of 345 families for autism-susceptibility loci, Am J Hum Genet 73, 886–897, 2003. 59. McCauley, J.L. et al., Genome-wide and Ordered-Subset linkage analyses provide support for autism loci on 17q and 19p with evidence of phenotypic and interlocus genetic correlates, BMC Med Genet 6, 1, 2005. 60. IMGSAC, Further characterization of the autism susceptibility locus AUTS1 on chromosome 7q, Hum Mol Genet 10, 973–982, 2001. 61. Ylisaukko-oja, T. et al., Genome-wide scan for loci of Asperger syndrome, Mol Psychiatry 9, 161–168, 2004. 62. Badner, J.A. and Gershon, E.S., Regional meta-analysis of published data supports linkage of autism with markers on chromosome 7, Mol Psychiatry 7, 56–66, 2002. 63. Roberts, S.B., MacLean, C.J., Neale, M.C., Eaves, L.J., and Kendler, K.S., Replication of linkage studies of complex traits: an examination of variation in location estimates, Am J Hum Genet 65, 876–884, 1999. 64. Shao, Y. et al., Phenotypic homogeneity provides increased support for linkage on chromosome 2 in autistic disorder, Am J Hum Genet 70, 1058–1061, 2002. 65. Buxbaum, J.D. et al., Linkage analysis for autism in a subset families with obsessivecompulsive behaviors: evidence for an autism susceptibility gene on chromosome 1 and further support for susceptibility genes on chromosome 6 and 19, Mol Psychiatry 9, 144–150, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 70 Monday, February 13, 2006 5:02 PM

70

Understanding Autism: From Basic Neuroscience to Treatment 66. Bradford, Y. et al., Incorporating language phenotypes strengthens evidence of linkage to autism, Am J Med Genet 105, 539–547, 2001. 67. Lamb, J.A. et al., Analysis of IMGSAC autism susceptibility loci: evidence for sex limited and parent of origin specific effects, J Med Genet 42, 132–137, 2005. 68. Stone, J.L. et al., Evidence for sex-specific risk alleles in autism spectrum disorder, Am J Hum Genet 75, 1117–1123, 2004. 69. Fullerton, J. et al., Linkage analysis of extremely discordant and concordant sibling pairs identifies quantitative-trait loci that influence variation in the human personality trait neuroticism, Am J Hum Genet 72, 879–890, 2003. 70. Fisher, S.A. et al., Sex stratification of an inflammatory bowel disease genome search shows male-specific linkage to the HLA region of chromosome 6, Eur J Hum Genet 10, 259–265, 2002. 71. Kristiansen, O.P. et al., Association of a functional 17beta-estradiol sensitive IL6174G/C promoter polymorphism with early-onset type 1 diabetes in females, Hum Mol Genet 12, 1101–1110, 2003. 72. O’Donnell, C.J. et al., Evidence for association and genetic linkage of the angiotensinconverting enzyme locus with hypertension and blood pressure in men but not women in the Framingham Heart Study, Circulation 97, 1766–1772, 1998. 73. Geschwind, D.H. et al., The autism genetic resource exchange: a resource for the study of autism and related neuropsychiatric conditions, Am J Hum Genet 69, 463–466, 2001. 74. Dowd, M.F., Manoski, R., Santangelo, S., and Folstein, S., American Society of Human Genetics, San Diego, CA, 2001. 75. Shao, Y. et al., American Society of Human Genetics, San Diego, CA, 2001. 76. Cordell, H.J., Wedig, G.C., Jacobs, K.B., and Elston, R.C., Multilocus linkage tests based on affected relative pairs, Am J Hum Genet 66, 1273–1286, 2000. 77. Cox, N.J. et al., Loci on chromosomes 2 (NIDDM1) and 15 interact to increase susceptibility to diabetes in Mexican Americans, Nat Genet 21, 213–215, 1999. 78. Xu, J. et al., Genomewide screen and identification of gene-gene interactions for asthma-susceptibility loci in three U.S. populations: collaborative study on the genetics of asthma, Am J Hum Genet 68, 1437–1446, 2001. 79. Collins, F.S., Positional cloning moves from perditional to traditional, Nat Genet 9, 347–350, 1995. 80. Lander, E.S. et al., Initial sequencing and analysis of the human genome, Nature 409, 860–921, 2001. 81. Finishing the euchromatic sequence of the human genome, Nature 431, 931–945, 2004. 82. Keller, F. and Persico, A.M., The neurobiological context of autism, Mol Neurobiol 28, 1–22, 2003. 83. Wassink, T.H. et al., Evidence supporting WNT2 as an autism susceptibility gene, Am J Med Genet 105, 406–413, 2001. 84. Li, J. et al., Lack of association between HoxA1 and HoxB1 gene variants and autism in 110 multiplex families, Am J Med Genet 114, 24–30, 2002. 85. McCoy, P.A. et al., No association between the WNT2 gene and autistic disorder, Am J Med Genet 114, 106–109, 2002. 86. Kemper, T.L. and Bauman, M., Neuropathology of infantile autism, J Neuropathol Exp Neurol 57, 645–652, 1998. 87. Bailey, A. et al., A clinicopathological study of autism, Brain 121, 889–905, 1998. 88. Hadj-Sahraoui, N., Frederic, F., Delhaye-Bouchaud, N., and Mariani, J., Gender effect on Purkinje cell loss in the cerebellum of the heterozygous reeler mouse, J Neurogenet 11, 45–58, 1996.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 71 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

71

89. Persico, A.M. et al., Reelin gene alleles and haplotypes as a factor predisposing to autistic disorder, Mol Psychiatry 6, 150–159, 2001. 90. Zhang, H. et al., Reelin gene alleles and susceptibility to autism spectrum disorders, Mol Psychiatry 7, 1012–1017, 2002. 91. Skaar, D.A. et al., Analysis of the RELN gene as a genetic risk factor for autism, Mol Psychiatry 10, 563–571, 2005. 92. Devlin, B. et al., Alleles of a reelin CGG repeat do not convey liability to autism in a sample from the CPEA network, Am J Med Genet B Neuropsychiatr Genet 126, 46–50, 2004. 93. McCoy, P.A. et al., American Society of Human Genetics, San Diego, CA, 2001. 94. Bonora, E. et al., Analysis of reelin as a candidate gene for autism, Mol Psychiatry 8, 885–892, 2003. 95. Bonora, E. et al., Mutation screening and association analysis of six candidate genes for autism on chromosome 7q, Eur J Hum Genet 13, 198–207, 2005. 96. Hutcheson, H.B. et al., Examination of NRCAM, LRRN3, KIAA0716, and LAMB1 as autism candidate genes, BMC Med Genet 5, 12, 2004. 97. Serajee, F.J., Nabi, R., Zhong, H., and Mahbubul Huq, A.H., Association of INPP1, PIK3CG, and TSC2 gene variants with autistic disorder: implications for phosphatidylinositol signalling in autism, J Med Genet 40, e119, 2003. 98. Ingram, J.L. et al., Discovery of allelic variants of HOXA1 and HOXB1: genetic susceptibility to autism spectrum disorders, Teratology 62, 393–405, 2000. 99. Conciatori, M. et al., Association between the HOXA1 A218G polymorphism and increased head circumference in patients with autism, Biol Psychiatry 55, 413–419, 2004. 100. Gharani, N., Benayed, R., Mancuso, V., Brzustowicz, L.M., and Millonig, J.H., Association of the homeobox transcription factor, ENGRAILED 2, 3, with autism spectrum disorder, Mol Psychiatry 9, 474–484, 2004. 101. Petit, E. et al., Association study with two markers of a human homeogene in infantile autism, J Med Genet 32, 269–274, 1995. 102. Zhong, H., Serajee, F.J., Nabi, R., and Huq, A.H., No association between the EN2 gene and autistic disorder, J Med Genet 40, e4, 2003. 103. Devlin, B. et al., No evidence for linkage of liability to autism to HOXA1 in a sample from the CPEA network, Am J Med Genet 114, 667–672, 2002. 104. Gallagher, L., Hawi, Z., Kearney, G., Fitzgerald, M., and Gill, M., No association between allelic variants of HOXA1/HOXB1 and autism, Am J Med Genet B Neuropsychiatr Genet 124, 64–67, 2004. 105. Fatemi, S.H. et al., Glutamic acid decarboxylase 65 and 67 kDa proteins are reduced in autistic parietal and cerebellar cortices, Biol Psychiatry 52, 805–810, 2002. 106. Buxbaum, J.D. et al., Association between a GABRB3 polymorphism and autism, Mol Psychiatry 7, 311–316, 2002. 107. Cook, E.H., Jr. et al., Linkage-disequilibrium mapping of autistic disorder, with 15q11-13 markers, Am J Hum Genet 62, 1077–1083, 1998. 108. Martin, E.R. et al., Analysis of linkage disequilibrium in gamma-aminobutyric acid receptor subunit genes in autistic disorder, Am J Med Genet 96, 43–48, 2000. 109. Maestrini, E. et al., Serotonin transporter (5-HTT) and gamma-aminobutyric acid receptor subunit beta3 (GABRB3) gene polymorphisms are not associated with autism in the IMGSA families, The International Molecular Genetic Study of Autism Consortium, Am J Med Genet 88, 492–496, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 72 Monday, February 13, 2006 5:02 PM

72

Understanding Autism: From Basic Neuroscience to Treatment

110. McCauley, J.L. et al., A linkage disequilibrium map of the 1-Mb 15q12 GABA(A) receptor subunit cluster and association to autism, Am J Med Genet B Neuropsychiatr Genet 131, 51–59, 2004. 111. Nurmi, E.L. et al., Linkage disequilibrium at the Angelman syndrome gene ube3a in autism families, Genomics 77, 105–113, 2001. 112. Veenstra-VanderWeele, J., Gonen, D., Leventhal, B.L., and Cook, E.H., Jr., Mutation screening of the UBE3A/E6-AP gene in autistic disorder, Mol Psychiatry 4, 64–67, 1999. 113. Nurmi, E.L. et al., Dense linkage disequilibrium mapping in the 15q11-q13 maternal expression domain yields evidence for association in autism, Mol Psychiatry 8, 624–634, 2003. 114. Kim, S.-J. et al., Mutation screening and transmission disequilibrium study of ATP10C in autism., 2002. 115. Conroy, J. et al., Serotonin transporter gene and autism: a haplotype analysis in an Irish autistic population, Mol Psychiatry 9, 587–593, 2004. 116. Cook, E.H., Jr. et al., Evidence of linkage between the serotonin transporter and autistic disorder, Mol Psychiatry 2, 247–250, 1997. 117. Betancur, C. et al., Serotonin transporter gene polymorphisms and hyperserotonemia in autistic disorder, Mol Psychiatry 7, 67–71, 2002. 118. Persico, A.M. et al., Lack of association between serotonin transporter gene promoter variants and autistic disorder in two ethnically distinct samples, Am J Med Genet 96, 123–127, 2000. 119. McCauley, J.L. et al., Linkage and association analysis at the serotonin transporter (SLC6A4) locus in a rigid-compulsive subset of autism, Am J Med Genet B Neuropsychiatr Genet 127, 104–112, 2004. 120. Kim, S.J. et al., Transmission disequilibrium mapping at the serotonin transporter gene (SLC6A4) region in autistic disorder, Mol Psychiatry 7, 278–288, 2002. 121. Lassig, J.P. et al., Physical mapping of the serotonin 5-HT(7) receptor gene (HTR7) to chromosome 10 and pseudogene (HTR7P) to chromosome 12, and testing of linkage disequilibrium between HTR7 and autistic disorder, Am J Med Genet 88, 472–475, 1999. 122. Herault, J. et al., Serotonin and autism: biochemical and molecular biology features, Psychiatry Res 65, 33–43, 1996. 123. Jamain, S. et al., Linkage and association of the glutamate receptor 6 gene with autism, Mol Psychiatry 7, 302–310, 2002. 124. Shuang, M. et al., Family-based association study between autism and glutamate receptor 6 gene in Chinese Han trios, Am J Med Genet B Neuropsychiatr Genet 131, 48–50, 2004. 125. Serajee, F.J., Zhong, H., Nabi, R., and Huq, A.H., The metabotropic glutamate receptor 8 gene at 7q31: partial duplication and possible association with autism, J Med Genet 40, e42, 2003. 126. Bacchelli, E. et al., Screening of nine candidate genes for autism on chromosome 2q reveals rare nonsynonymous variants in the cAMP-GEFII gene, Mol Psychiatry 8, 916–924, 2003. 127. Ramoz, N. et al., Linkage and association of the mitochondrial aspartate/glutamate carrier SLC25A12 gene with autism, Am J Psychiatry 161, 662–669, 2004. 128. Amir, R.E. et al., Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl- CpG-binding protein 2, Nat Genet 23, 185–188, 1999. 129. Lam, C.W. et al., Spectrum of mutations in the MECP2 gene in patients with infantile autism and Rett syndrome, J Med Genet 37, E41, 2000.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 73 Monday, February 13, 2006 5:02 PM

Genetic Basis of Autism

73

130. Beyer, K.S. et al., Mutation analysis of the coding sequence of the MECP2 gene in infantile autism, Hum Genet 111, 305–309, 2002. 131. Vourc’h, P. et al., No mutations in the coding region of the Rett syndrome gene MECP2 in 59 autistic patients, Eur J Hum Genet 9, 556–558, 2001. 132. Carney, R.M. et al., Identification of MeCP2 mutations in a series of females with autistic disorder, Pediatr Neurol 28, 205–211, 2003. 133. Lobo-Menendez, F. et al., Absence of MeCP2 mutations in patients from the South Carolina autism project, Am J Med Genet 117B, 97–101, 2003. 134. Shibayama, A. et al., MECP2 structural and 3'-UTR variants in schizophrenia, autism and other psychiatric diseases: a possible association with autism, Am J Med Genet B Neuropsychiatr Genet 128, 50–53, 2004. 135. Jamain, S. et al., Mutations of the X-linked genes encoding neuroligins NLGN3 and NLGN4 are associated with autism, Nat Genet 34, 27–29, 2003. 136. Laumonnier, F. et al., X-linked mental retardation and autism are associated with a mutation in the NLGN4 gene, a member of the neuroligin family, Am J Hum Genet 74, 552–557, 2004. 137. Gauthier, J. et al., NLGN3/NLGN4 gene mutations are not responsible for autism in the Quebec population, Am J Med Genet B Neuropsychiatr Genet 132, 74–75, 2005. 138. Yan, J. et al., Analysis of the neuroligin 3 and 4 genes in autism and other neuropsychiatric patients, Mol Psychiatry 10, 329–332, 2005. 139. Vincent, J.B. et al., Mutation screening of X-chromosomal neuroligin genes: no mutations in 196 autism probands, Am J Med Genet B Neuropsychiatr Genet 129, 82–84, 2004. 140. Ardlie, K.G., Kruglyak, L., and Seielstad, M., Patterns of linkage disequilibrium in the human genome, Nat Rev Genet 3, 299–309, 2002. 141. Tabor, H.K., Risch, N.J., and Myers, R.M., Opinion: candidate-gene approaches for studying complex genetic traits: practical considerations, Nat Rev Genet 3, 391–397, 2002. 142. Spielman, R.S., McGinnis, R.E., and Ewens, W.J., Transmission test for linkage disequilibrium: the insulin gene region and insulin-dependent diabetes mellitus (IDDM), Am J Hum Genet 52, 506–516, 1993. 143. Gabriel, S.B. et al., The structure of haplotype blocks in the human genome, Science 296, 2225–2229, 2002. 144. Patil, N. et al., Blocks of limited haplotype diversity revealed by high-resolution scanning of human chromosome 21, Science 294, 1719–1723, 2001. 145. Cardon, L.R. and Abecasis, G.R., Using haplotype blocks to map human complex trait loci, Trends Genet 19, 135–140, 2003. 146. Purcell, A.E., Jeon, O.H., Zimmerman, A.W., Blue, M.E., and Pevsner, J., Postmortem brain abnormalities of the glutamate neurotransmitter system in autism, Neurology 57, 1618–1628, 2001. 147. Pritchard, J.K., Are rare variants responsible for susceptibility to complex diseases? Am J Hum Genet 69, 124–137, 2001. 148. Zondervan, K.T. and Cardon, L.R., The complex interplay among factors that influence allelic association, Nat Rev Genet 5, 89–100, 2004. 149. Ke, X. et al., The impact of SNP density on fine-scale patterns of linkage disequilibrium, Hum Mol Genet 13, 577–588, 2004. 150. Kruglyak, L., Prospects for whole-genome linkage disequilibrium mapping of common disease genes, Nat Genet 22, 139–144, 1999. 151. Rubinsztein, D.C. and Easton, D.F., Apolipoprotein E genetic variation and Alzheimer’s disease: a meta-analysis, Dement Geriatr Cognit Disord 10, 199–209, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C003.fm Page 74 Monday, February 13, 2006 5:02 PM

74

Understanding Autism: From Basic Neuroscience to Treatment

152. Hugot, J.P. et al., Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn’s disease, Nature 411, 599–603, 2001. 153. Risch, N.J., Searching for genetic determinants in the new millennium, Nature 405, 847–856, 2000. 154. Morris, R.W. and Kaplan, N.L., On the advantage of haplotype analysis in the presence of multiple disease susceptibility alleles, Genet Epidemiol 23, 221–233, 2002. 155. Risch, N. and Teng, J., The relative power of family-based and case-control designs for linkage disequilibrium studies of complex human diseases I. DNA pooling, Genome Res 8, 1273–1288, 1998. 156. Horikawa, Y. et al., Genetic variation in the gene encoding calpain-10 is associated with type 2 diabetes mellitus, Nat Genet 26, 163–175, 2000. 157. Fingerlin, T.E., Boehnke, M., and Abecasis, G.R., Increasing the power and efficiency of disease-marker case-control association studies through use of allele-sharing information, Am J Hum Genet 74, 432–443, 2004. 158. Knight, J.C., Keating, B.J., Rockett, K.A., and Kwiatkowski, D.P., In vivo characterization of regulatory polymorphisms by allele-specific quantification of RNA polymerase loading, Nat Genet 33, 469–475, 2003.

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 75 Friday, March 10, 2006 2:16 PM

4

Finding Genes in Spite of Heterogeneity: Endophenotypes, QTL Mapping, and Expression Profiling in Autism Daniel H. Geschwind and Maricela Alarcón

CONTENTS Introduction..............................................................................................................76 Making Sense of Complexity: The Need to Link Brain, Genes, and Behavior................................................................................................76 Autism: A Complex and Genetically Heterogeneous Trait ....................................77 Current Status of Autism Linkage...........................................................................77 Approach #1 — Endophenotypes in Mental Disorders..........................................78 Endophenotypes also Inform the Use of Animal Models ...............................78 Linkage in Large Families ...............................................................................79 Extending Endophenotype Analysis to Autism: Preliminary Success or Signs of Trouble? ................................................................................................80 Language...........................................................................................................81 Social Behavior and Cognition ........................................................................81 RRBs.................................................................................................................83 Modeling Other Endophenotypes and the Autism Continuum: Structural Equation Modeling .................................................................................83 The Next Step — Structural Equation Modeling (SEM)................................84 Phenotypic Models ...........................................................................................85 Genetic Models.................................................................................................86 Approach #2 — Microarray and Bioinformatic Analysis of Gene Expression: Untested but not Unworthy .....................................................................................87 References................................................................................................................89

75

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 76 Friday, March 10, 2006 2:16 PM

76

Understanding Autism: From Basic Neuroscience to Treatment

INTRODUCTION Although it is among the most genetically based of neuropsychiatric syndromes, with a heritability estimated to be over 90%,1 the identification of the genes underlying autism risk has been challenging. The reasons for this are manifold, including those that plague most genetic studies of complex diseases (see Table 4.1 and Bonora et al. this volume). Another putative reason highlighted in the chapter by Beaudet and Zoghbi (Chapter 5) is the potential role of epigenetic factors, such as changes in chromatin structure. However, beyond genetic complexity, studies of autism must also deal with the enormous challenges posed by the phenotype. Autism is a disorder that is a mixture of a spectrum of behavioral and cognitive disabilities that are rarely observed in an identical manner even among monozygotic (MZ) twins. The reconciliation of the enormous clinical heterogeneity with the equally vast genetic complexity presents considerable challenges. Here we discuss two approaches that we expect will aid in the effort to identify the genetic basis of autism: the use of endophenotypes for quantitative trait locus (QTL) analysis and genomic expression profiling for subject classification and gene discovery.

TABLE 4.1 Challenges of Complex Genetic Disorders 1. Many genes contribute to the phenotype. 2. Each gene contributes a portion of the risk. 3. The mix of genes that contribute varies between patients (genetic heterogeneity). 4. Common variants (present in unaffected subjects) may contribute. 5. Environmental influences may be significant. 6. Disease status is not defined relative to underlying biological causes, so affection status may be a relatively arbitrary quantitative determination or cutoff. 7. Lack of replication of linkage.

MAKING SENSE OF COMPLEXITY: THE NEED TO LINK BRAIN, GENES, AND BEHAVIOR Classical behavior genetics has focused primarily on psychometric measurements of behavior or cognitive performance as the phenotypes of interest rather than the brain structure that clearly must underlie these traits. However, the brain, its development, and its subsequent modifiability by experience need to be considered clearly within the domain of phenotypes most relevant to disease. Any gene, or any behavior, needs to be understood in its structural and systems context within the CNS. In the same vein, our current notion of disease is predicated primarily on clinical utility rather than underlying genetic etiology. In some cases, exemplified by Alzheimer’s disease (AD), frontotemporal dementia (FTD), or Parkinson’s disease, a defining molecular or cellular pathology has been identified that allows rigorous classification of the disease. However, even in molecularly defined dementias, such

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 77 Friday, March 10, 2006 2:16 PM

Finding Genes in Spite of Heterogeneity

77

as AD or FTD, there is not always a correlation between pathology, clinical phenotype, and the mutational basis of disease.2–4

AUTISM: A COMPLEX AND GENETICALLY HETEROGENEOUS TRAIT In a heterogeneous clinical syndrome such as autism, where there is no biomarker or defining pathology, the situation is even more vexing. Patients are classified syndromically by behavioral observation along many dimensions. Thus two patients can differ significantly along many of these individual behavioral or cognitive dimensions, yet share the same composite diagnosis. In this case, the brain systems involved may be distinct, or overlapping systems that are affected may be affected in different ways. Furthermore, two patients may look identical at one time point and yet have very different developmental trajectories. The prognosis of these two patients would be entirely different, suggesting different brain mechanisms have been activated either by the disease itself or by the nervous system’s response to the disease. From a genetic standpoint, this creates enormous complexity, as it clearly involves the role of many genes interacting with each other and environmental factors, some of which we may not be able to measure accurately. Even in cases where two patients have a very similar behavioral and cognitive profile, it is probable that they may have distinct molecular etiologies, as has been observed in Mendelian diseases (SCA, dementia). Genetic researchers in autism share the hope that knowledge of these multiple molecular etiologies will lead to some understanding of the final common pathways, highlighting the key brain systems that are involved as well as the mechanism of their dysfunction. Classical autism has been observed in many single-gene Mendelian disorders, including fragile X, tuberous sclerosis, and Joubert’s syndrome, although within these conditions, a diagnosis of autistic disorder is only observed in a small percentage.1 In addition, a single chromosomal abnormality — a maternally inherited duplication of chromosome 15q11-13 — accounts for between 1–2% of autistic children5 (Christa Lese Martin, unpublished observations). However, autism is typically inherited in a nonmendelian pattern, in which it is likely that multiple genes interacting with the environment underlie autism susceptibility. This and the likely accompanying clinical and genetic heterogeneity pose significant challenges for conventional positional cloning approaches. Genetic heterogeneity in autism is supported by many lines of evidence, not the least of which are the results of independent genome scans, which provide empiric evidence for multiple loci.

CURRENT STATUS OF AUTISM LINKAGE Several genome-wide screens of multiplex families have identified possible susceptibility regions of interest, but few have reached a level of genome-wide significance1 (Bonora et al., 2006, this volume). Moreover, most of the published regions have not attained a level of significance that strongly warrant large-scale fine mapping efforts and few specific gene–disease associations have been confirmed in independent

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 78 Friday, March 10, 2006 2:16 PM

78

Understanding Autism: From Basic Neuroscience to Treatment

samples (Bonora et al., 2006, this volume). Even if the current strategies do identify reproducible underlying genetic variants, the question still remains: What aspect of the autism phenotype are they related to? Thus, strategies involving the definition of more homogeneous subsets of cases within current collections, or more specific endophenotypes that may be more proximal to a given genetic risk, rather than the clinical diagnosis of autistic disorder or pervasive development disorder (PDD) are warranted. These endophenotypes must be understood in the context of cognitive neuroscience and neurobiology. Furthermore, new analytic approaches are needed to increase the power to detect and confirm multiple interacting genetic loci. These strategies are the focus of the rest of this review.

APPROACH #1 — ENDOPHENOTYPES IN MENTAL DISORDERS Even in the context of a single-gene cause of autism, it seems reasonable that rather than trying to find genes for autism per se, we could take the route of defining simpler individual components or features related to autism. Optimally, these components may be seen to a greater extent in the first-degree relatives of autistic probands than in unrelated individuals and yet be on a continuum with such features in the general population. But the hope is that they are less complex and thus may be closer to individual genetic risk factors than autism itself. Such a component is called an endophenotype.15,16 The term endophenotype was originally coined to refer to phenotypes that were not visible but, rather, were microscopic or biochemical.15,17 The meaning of the term has expanded considerably to include virtually any measurable feature — cognitive, behavioral, temporal progression, structural, etc. — that is associated with a disease state, which meets certain criteria to ensure both their underlying genetic simplicity and their relationship with a psychiatric disorder.15,18 The rationale for using an endophenotype in psychiatric genetics was that it represented a more “elementary” heritable trait that could be explained by fewer genes than the complete clinical syndrome itself, making genetic risk factors easier to find. An endophenotype must be present more frequently in first-degree relatives than in the general population, a factor that can significantly improve the efficiency of genetic-mapping studies by allowing extension to larger pedigrees. An example of a relevant endophenotype for a neurologic condition is the use of the electroencephalogram (EEG), an electrophysiologic measure, to define abnormalities underlying various genetic epilepsy syndromes,19–21 or phonological processing to define a core deficit in dyslexia.22,23

ENDOPHENOTYPES

ALSO INFORM THE

USE

OF

ANIMAL MODELS

This highlights another important advantage of an endophenotype for understanding disease pathophysiology. Certain endophenotypes may allow more clear translation between genetic findings in humans and animal models.20 Modeling the complex behavior associated with an epilepsy syndrome is far more challenging than electrophysiological monitoring by EEG. This issue is clearly amplified by several orders of magnitude in autism, where the full-blown syndrome of autism will be impossible

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 79 Friday, March 10, 2006 2:16 PM

Finding Genes in Spite of Heterogeneity

79

to monitor, but a clear synaptic or brain structural abnormality or a behavior such as repetitive behavior or abnormal social interaction can be assessed in many animal model systems.20 It should be made clear, however, that the goal may not be to study social behavior in a number of animal systems, from zebra fish to the mouse, because the brain systems and even molecules involved in complex social behaviors could easily be distinct, especially in evolutionarily distant species. However, similar to the study of learning in model organisms, such approaches will be useful for untangling the synaptic or cellular mechanisms involved by asking questions such as what structural, developmental, or electrophysiological alterations occur when an autism susceptibility allele is expressed in a given model.24,25 A preliminary example of the use of endophenotypes is the recent study of subsonic vocalizations in a mouse model with a disruption in the gene FoxP2, in which a dominant mutation causes a human speech and language disorder.26

LINKAGE

IN

LARGE FAMILIES

Endophenotypes provide another important avenue to significantly increase the power of genetic studies of autism. Because endophenotypes should be measurable in unaffected family members, they may be particularly useful in employing the power of large pedigrees to identify autism-related loci. The concept of the extended phenotypes in families has been called “the broader autism phenotype.”27–32 Using these concepts, the Autism Genetic Resource Exchange (AGRE) group33,34 and others35–40 have analyzed the dimensions of language delay and repetitive restrictive behavior (RRB) to increase the power to detect linkage within their samples, either by sample stratification (e.g., Reference 36) or by using quantitative trait information in the entire sample (e.g., Reference 33) as discussed in the text that follows. Endophenotype scan results are summarized in Table 4.2. Few scan results overlap and, with a few exceptions,34,41 there have been no attempts at independent replication. These studies have focused on nuclear families. None has used the full power of extensively phenotyping unaffected family members, which many believe is the main advantage of this approach.16,41 In cases where extended families are available, linkage methods have considerable power relative to sibling pair strategies, especially when the number of affected sibs is low.16,42–44 Linkage analysis of large extended families defined by the genealogy can also increase the resolution of linkage mapping, an important issue, given the often broad regions identified in sibling studies of complex diseases. Perhaps the best examples of this approach are from Iceland, where loci for a number of genetically complex diseases, including schizophrenia, have been identified and confirmed in other populations.45,46 Efficient use of this approach requires well-validated endophenotypes that can be efficiently measured and access to large families or genealogical records. Unfortunately, the use of endophenotypes in genetic studies of autism is hampered by our current lack of knowledge as to the nature of robust brain structural (imaging or measurement of head circumference), biochemical, gene expression, or cognitivebehavioral, clearly heritable, autism-related endophenotypes and the difficulties associated with collecting such data. Although there have been efforts to reduce genetic heterogeneity by stratifying autism families as a function of the proband’s language

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 80 Friday, March 10, 2006 2:16 PM

80

Understanding Autism: From Basic Neuroscience to Treatment

TABLE 4.2 Endophenotype Scans in Autism Endophenotype Age at first phrase Age at first word Age at first word Age at first word Insistence on sameness Obsessive-compulsive behavior Obsessive-compulsive behavior Obsessive-compulsive behavior Phrase speech delay Phrase speech delay Phrase speech delay Repetitive behaviors Rigid-compulsive behavior Savant skill Word speech delay

Study Size (n)

Best Linkage

Chromosome

Reference

291 AGRE families 291 AGRE families 152 AGRE families 291 AGRE families 23 ASD families 115 Affected sib pairs 115 Affected sib pairs 115 Affected sib pairs 95 Multiplex families 99 Multiplex families 133 AGRE families 291 AGRE families 137 Multiplex families 94 Multiplex families 133 AGRE families

Z = 2.22 Z = 3.10 Z = 2.98 Z = 2.84 MLS = 4.71 NPL = 3.06

17p 3q 7q35 17q 15q11-q12 1q42.2

Alarcón et al., 2005 Alarcón et al., 2005 Alarcón et al., 2002 Alarcón et al., 2005 Shao et al., 2003 Buxbaum et al., 2004

NPL = 2.61

6q14.3

Buxbaum et al., 2004

NPL = 2.31

19p13.12

Buxbaum et al., 2004

NPL = 2.45

02q31-q32

Buxbaum et al., 2001

NPL = 2.86

2q33.1

Shao et al., 2002

NPL = 2.35 Z = 2.31 HLOD = 3.62

15q 17q 17p11.2

Spence et al., in press Alarcón et al., 2005 McCauley et al., 2004

MLS = 2.6

15q11-q12

NPL = 2.41

12q

Nurmi et al., 2003 Spence et al., in press

delay35,37 and by including endophenotypes as covariates in linkage analysis,40,47 relatively little attention has been given to better define the phenotype of autism used for research purposes. The expression of autism varies greatly among those affected with the disorder. For example, one patient may lead an independent life and have above average language ability, whereas another may be unable to speak or to communicate, have below average cognitive ability, and may require assistance for daily living. It is crucial to exploit this variability in autism to facilitate the detection of susceptibility genes for the disorder. Thus, the first step is to find novel measures that capture the variability in the expression of autism. These measures may then be used as endophenotypes in subsequent genetic analyses.

EXTENDING ENDOPHENOTYPE ANALYSIS TO AUTISM: PRELIMINARY SUCCESS OR SIGNS OF TROUBLE? Investigators have begun to use autism characteristics or endophenotypes (such as language delay, social and communication impairments, and repetitive behaviors) as covariates in linkage analyses to increase both phenotypic and genetic homogeneity of

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 81 Friday, March 10, 2006 2:16 PM

Finding Genes in Spite of Heterogeneity

81

the sample. Here we discuss results of genetic studies with endophenotypes related to the core autism deficits: language, social cognition, and repetitive, restrictive behavior.

LANGUAGE The use of stratification by language phenotype34,35,48 and QTL mapping of speechand language-related endophenotypes in multiplex autistic families33 have demonstrated that a significant portion of the signal present in the 7q3 region may be related to a language-related endophenotype in multiplex autism families.38 The use of this endophenotype of word or phrase speech delay has helped to refine linkage signals and shows significant promise not only for the 7q region but potentially for the 2q region37 and another possibly linked region on chromosome 13q.35 A 9.6-Mb region on chromosome 7q3 was defined by quantitative linkage and association results.33 In a preliminary exploration of this region, 130 single-nucleotide polymorphisms (SNPs) in 11 candidate genes were tested. Each marker was tested for association with the diagnosis of autism as well as with the language endophenotypes. Results suggested modest associations of the contactin-associated proteinlike 2 (CNTNAP2) gene with broad autism (p < .05), WORD (p < .05), and PHRASE (p < .05). The CNTNAP2 gene is one of the largest genes in the human genome and encompasses 1.5% of chromosome 7.49 It is comprised of 25 exons distributed over 2 Mb on chromosome 7q35, the region that underlies our reported language QTL. The gene encodes transmembrane proteins that mediate cell-to-cell interactions in the nervous system48 and was recently reported to be disrupted in a family with Gilles de la Tourette’s syndrome and obsessive-compulsive disorder.50 Due to the overlapping characteristics of these disorders with autism,51 and the position of CNTNAP2 relative to our QTL region, this gene is a good candidate for possible involvement with autism. Over 500 additional SNPs were typed to cover the CNTNAP2 gene more thoroughly in 212 affected individuals and their parents from AGRE. Association analyses of these SNPs using the diagnosis of autism as well as age at first word (an item from the Autism Diagnostic Interview – Revised, ADI – R) support the relationship of the CNTNAP2 gene to autism susceptibility in this sample.52,53 Replication of this association result is the necessary next step in identifying a QTL for autism. Although the modest p-values for the association results may not meet the criteria for multiple testing, it is possible that a better measure of the disorder will show a more significant result with the CNTNAP2 gene.

SOCIAL BEHAVIOR

AND

COGNITION

Deficits in social cognition and behavior are at the core of autism spectrum disorders.54–58 Screening instruments for the milder variants of autism spectrum disorders in adolescents and adults are scarce. Two standardized rating scales designed for this purpose, the Autism Spectrum Quotient and the Social Responsiveness Scale (SRS), have been developed and used in family studies.58–62 In addition, Sung et al. have used a measure developed by Dawson and colleagues (unpublished) — the Broader Autism Phenotype Symptom Scale (BPASS) — to assess heritability of

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 82 Friday, March 10, 2006 2:16 PM

82

Understanding Autism: From Basic Neuroscience to Treatment

broader phenotypes, including social behavior.63 According to Sung et al.,63 the BPASS shows high correlation with the Family History Interview,64 and although it appears very promising,63 little has been published on its reliability metrics and validity. For the last decade or so, Constantino and colleagues have been working to develop and validate quantitative measures of autism-related social impairment, which has led to the development of the SRS.58 The SRS is a 65-item quantitative measure of autistic social impairment that has been extensively tested in both clinically-ascertained and population-based samples of subjects.58 Scores on the SRS are highly heritable, unrelated to IQ, and continuously distributed in the general population. Most importantly, the SRS clearly distinguishes patients with autism spectrum disorders from those with other child psychiatric conditions. Although the SRS deals with multiple autism diagnostic issues beyond social cognition, it focuses on a core feature of autism: a subject’s capacity to engage in emotionally appropriate reciprocal social interaction. Analyses of the SRS have consistently indicated the presence of a singular continuously-distributed variable underlying autistic impairment, characterized by general deficiency in reciprocal social behavior and for which a single index score is generated by the SRS.57 A recent study demonstrates the distribution of the SRS scores among autistic children, their siblings from the AGRE cohort, PDD subjects, siblings, and normal controls (Constantino et al., in press). There is a clear separation of SRS distributions in autistic and PDD subjects and normal controls, with siblings of autistic probands falling in between, consistent with a potential major gene effect. This distribution represents a nearly ideal situation from the standpoint of a quantitative endophentype for linkage mapping, because it indicates that genetic liability is manifested by an aggregation of deficiency in reciprocal social behavior that is continuously distributed among the siblings of affected probands. These results also indicate that clinical categories within the “broader autism phenotype” may share common genetic underpinnings.58 These findings are supported by data from the BPASS, in which social behavior appears to be one of the important heritable features identified.63 These findings of the heritability and distribution of social-related variables has important implications for quantitative genetic studies, in that the search for genetic susceptibility loci for autism may be advanced by studying the genetic factors that result in variation in reciprocal social behavior in large population-based samples that include a wide range of autistic social impairment. Based on this exciting supposition, we have performed a preliminary genomewide linkage analysis of the SRS in 125 AGRE pedigrees.65 Nonparametric QTL analysis was performed using the all-sibling pairs, equally weighted option, implemented in GENEHUNTER. Despite the fact that this sample was approximately 1/3 of the Yonan et al. cohort,66 surprisingly strong positive findings were obtained. Seven loci attained a nominal p value of < .01 using the SRS in this small cohort, vs. only four loci reaching this level in the much larger qualitative scan, illustrating the power of quantitative traits vs. categorical diagnoses in linkage analysis. The SRS peaks were also highly overlapping with previous autism peaks, demonstrating the power and relevance of the QTL approach using this specific endophenotype.

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 83 Friday, March 10, 2006 2:16 PM

Finding Genes in Spite of Heterogeneity

83

RRBS A number of studies support the familiality of RRB in families of autistic probands. One of the first clear characterizations of the broader phenotype was by Piven and colleagues; they identified a dimension of “rigid–aloof” personality in families of autistic probands.29,67 Other studies using populations ascertained for genetic investigation of autism have also identified evidence for significant familiality or heritability of this dimension of behavior.33,63,68 Genome scans in the AGRE sample have been performed using a composite measure of RRB (item DD total from the ADI – R), but have not yielded strong evidence for loci related to this trait.33,36 In the study by Buxbaum and colleagues,33,36 composite measures related to OCD (D1 and D2) are used to stratify families, resulting in the identification of only modest statistical support for even the best locus. This is not surprising, given the small sample size used (less than 50 famlies with the phenotype). However, Shao et al. have used an “insistence on sameness” measure, derived from a principal components analysis of selected items from the ADI – R, as a covariate on which to stratify families in a linkage analysis of autism.40 This method, called ordered subset analysis, increased the Lod score by nearly 3-fold to 4.7. This region harbors the GABARB3 gene, which has been implicated in some preliminary reports of association in autism. It remains to be seen whether variation in this gene underlies the 15q11-13 linkage peak. In contrast to other studies that use the phenotype of RRB on which to stratify families into presumably more homogeneous groups based on phenotype, the study by Alarcón et al.33 uses RRB directly as a quantitative trait on which to base linkage. The QTL approach has the power of avoiding categorical, sometimes arbitrary, cutoffs to stratify families and the power of including all families rather than a subset with the categorical phenotype in question.63 As mentioned above, the RRB measure used in the QTL genome scan was an item from the ADI – R called DD total. This composite measure is familial (with a significant sibling correlation of 0.31) and when used in QTL genome scans of AGRE families, it produced modest linkage peaks on chromosomes 16p and 17q.33,34 As mentioned earlier, QTL analyses may be extended to include large families, rather than the nuclear pedigrees so far used in most genetic studies of autism. We have used the same QTL approach to look at other familial behaviors related to autism, such as nonverbal communication in the AGRE sample, which has also shown considerable power when compared with traditional linkage approaches, identifying at least one locus of genome-wide significance.69 These studies on RRB, the SRS, and language all highlight the promise of QTL approaches in autism. More focus on independent replication, which will necessitate collaborations and data sharing between groups, is needed to move this area forward.

MODELING OTHER ENDOPHENOTYPES AND THE AUTISM CONTINUUM: STRUCTURAL EQUATION MODELING Most of the endophenotypes used to date consist of items from the ADI – R.50 Although these have been useful to identify or narrow down chromosomal regions of interest, a more extensive investigation of the covariation among several behavioral

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 84 Friday, March 10, 2006 2:16 PM

84

Understanding Autism: From Basic Neuroscience to Treatment

and cognitive measures in autistic individuals is warranted. By refining the autism phenotype that is used for research purposes, power to detect susceptibility genes for this disorder will be increased. Researchers and clinicians have recognized the value of studying the variability in the expression of the autism phenotype to understand the nature of the disorder. To this end, several investigators have begun to explore the covariation among items not only from the ADI – R itself but also from psychometric instruments such as the Childhood Autism Rating Scale (CARS),70 the Autism Behavior Checklist (ABC), the Vineland Adaptive Behavior Scales,71–73 and the SRS58 by applying traditional multivariate techniques such as principal components analysis (PCA) and factor analysis (FA).74,75 Tadevosyan-Leyfer et al.76 ran a PCA of most of the items from the ADI – R irrespective of their inclusion in its diagnostic algorithm. They identified six factors that summarize the instrument: spoken language, social intent, compulsions, developmental milestones, savant skills, and sensory aversions. These were validated against an independent sample that was extensively characterized with tests of intelligence, executive function, vocabulary, adaptive behavior, and psychiatric diagnosis. The sibling intraclass correlations of the PCA factors were modest, ranging from .15 to .30, but were significant and suggest that they are familial. Thus, these six factors may prove useful in linkage studies. In contrast, FA of 12 ADI – R items related to restricted and repetitive behaviors resulted in 2 factors (repetitive sensory motor actions and resistance to change) that did not have significant sibling intraclass correlations.75 These studies illustrate that although factors may represent two distinct classes of patients and their use may decrease phenotypic heterogeneity in research studies, they may not be useful endophenotypes for genetic analysis.

THE NEXT STEP — STRUCTURAL EQUATION MODELING (SEM) One of our research goals is to expand this factor analytic approach using a more flexible statistical method to include more behavioral and cognitive measures of autism as well as to simultaneously incorporate genetic information in the analysis. SEM provides a powerful method with which to investigate the covariation among behavioral tests of children and adolescents with autism spectrum disorder (ASD). SEM is a statistical tool that describes the linear relationships among observed and latent (i.e., unobserved) variables. Latent variables, such as alcoholism, intelligence, and neuroticism, are constructs that are not directly measured and may require multiple indicators for their assessment. For example, intelligence may be assessed by the Wechsler Intelligence Scale for Children — Revised and the Raven’s Progressive Matrices test. FA, PCA, regression, canonical correlation, discriminant analysis, and multivariate analysis of variance are all special cases of SEM.77 SEM has been applied to many fields including education, economics, sociology, marketing, and behavior genetics,78 and may possibly provide a framework through which we can address several questions regarding the nature of autism. For example, one can investigate whether there is an underlying “autism” factor that determines the co-variation among different measures of the disorder. Such a factor of severity has been proposed by several and supported by data using the ADI – R, Vineland

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 85 Friday, March 10, 2006 2:16 PM

Finding Genes in Spite of Heterogeneity

85

Adaptive Behavior Scales,79 and the SRS.74 The main components of such a factor can be used in a composite or separately for genetic analysis. The goal of the SEM approach is to develop a phenotypic model that includes the strongest indicators of autism and the optimum stratification of the sample with regard to diagnostic subtypes (autism, Asperger’s, PDD), communication level (verbal vs. nonverbal), and sex of the affected individual. A genetic model can be based on this parsimonious phenotypic model and incorporate data from the siblings of the probands. We believe that such a careful investigation of the behavioral and cognitive measures of individuals with autism can lead to an improved measure of the disorder that may be useful in the identification of autism genes.

PHENOTYPIC MODELS Several phenotypic models can be developed and tested to examine the covariation of the behavioral and cognitive measures of children and adolescents with autism across different diagnostic subtypes. The most general model is shown in Figure 4.1. This is the common factor model that can examine the phenotypic overlap in the dependent variables.80 The variance in each measure is partitioned into that which is shared among the measures (represented as a single latent factor) and that which is measure specific or residual. In this case, five components of the Vineland Adaptive Behavior Scales are used to illustrate the model. Communication (COMM), daily living skills (DLS), socialization (SOC), motor skills (MOT), and maladaptive behaviors (MB) are the dependent variables. A latent (i.e., unobserved) factor called AUTISM loads on all of the dependent variables and is solely responsible for their covariation. The factor loadings on each measure can be squared to represent the proportion of variance accounted for by the common factor. For example, the squared coefficient for COMM, b12, is the proportion of variance in communication that is due to the AUTISM factor, and the squared residual for COMM, e12, is the variance that is not shared with the latent factor. If the measures were uncorrelated, we would expect negligible loadings on the common factor and substantial loadings on the residual portion of the model. In contrast, if the measures are highly intercorrelated, we would expect substantial factor loadings on the common factor and negligible loadings on the residuals. If the factor loadings were significant,

AUTISM

b1

b2

b3

COMM

DLS

SOC

MOT

∈3

∈4

∈1

∈2

b4

b5 MB ∈5

FIGURE 4.1 Phenotypic common factor model of the Vineland Adaptive Behavior Scales.

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 86 Friday, March 10, 2006 2:16 PM

86

Understanding Autism: From Basic Neuroscience to Treatment

these results would support the existence of a latent factor, i.e., an underlying vulnerability or risk for autism. Moreover, the common factor model also allows the presence of independent factors that are unique to each measure. This model can be expanded to include the Vineland subtests, nonverbal IQ (IQ), the picture vocabulary (VOC) tests, and any other behavioral or cognitive tests. The overall fit of the factor models to the data can be evaluated using various goodness-of-fit indices such as the χ2 statistic, and χ2 difference tests can be used to assess the fit of various submodels.

GENETIC MODELS The genetic model will be an extension of the most parsimonious phenotypic model and will include a sibling of the original proband. The genetic model can test whether a candidate gene, such as CNTNAP2, influences the AUTISM factor. One example of a possible genetic model is shown in Figure 4.2. In this example, COMM, DLS, and SOC are the measures that best represent the AUTISM factor in the phenotypic model and are now used in the genetic model. In Figure 4.2, Q1 and Q2 are the latent QTL variables for sibs 1 and 2, respectively; R1 and R2 are the latent variables for sibs 1 and 2, respectively, that represent the sibs’ genetic relationship, independent of the QTL; and E1 and E2 represent the error terms for sibs 1 and 2, respectively. The correlation coefficients, π and s, are the estimates of the correlation between the additive effects of the QTL in question and the residual family resemblance not accounted for by the QTL, respectively. The coefficients q, r, and e are the influences of the latent genetic (Q1, Q2, R1, R2) s π E1

R1 e

r

Q1

Q2

q

R2 q

AUTISM

b1 COMM e1

b2

E2

r

e

AUTISM

b3

b1

DLS1

SOC1

e2

e3

COMM e1

b2

b3

DLS2

SOC2

e2

e3

FIGURE 4.2 The common factor QTL model is used to test whether the candidate gene CNTNAP2 influences the “autism” factor. Symbols are described in the text.

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 87 Friday, March 10, 2006 2:16 PM

Finding Genes in Spite of Heterogeneity

87

and environmetal (E1, E2) variables on the common factor of AUTISM. Nonadditive effects can also be incorporated into the genetic model. The QTL model described above is typically employed to test linkage, which can be used for finding candidates as well as gene mapping.81 However, our strategy is to consider models of joint association and linkage applicable to parent–sibling designs.81,82 Association is defined as the direct outcome of allele substitution on the observed trait value whereby any true allelic effects will contribute to both withinand between-family mean trait differences. The test of linkage is accomplished by decomposing the sibling covariance structure into variance accounted for by the QTL and remaining background variance. When association is present, linkage will be reduced81,83 and thereby provides clues indicating that a SNP candidate is the QTL or is close to the QTL. These phenotypic and genetic models illustrate the next steps in our strategy to both define a novel endophenotype for autism that captures the disorder’s variability as well as use to this measure to test the association of previously identified candidate genes.

APPROACH #2 — MICROARRAY AND BIOINFORMATIC ANALYSIS OF GENE EXPRESSION: UNTESTED BUT NOT UNWORTHY After we identify regions of linkage that reach genome-wide significance and are replicated in independent populations using QTL analysis or endophenotypes to stratify subjects (e.g., Cantor et al.84), we are left with the difficult task of identifying the genes and risk alleles underlying the phenotype. High-density SNP typing is now economically plausible to test common variants. But rare variants need to be assessed carefully on a gene-by-gene basis by sequencing, and such resequencing on a large scale is prohibitively expensive. How do we identify candidates under linkage peaks for this kind of exhaustive analysis? Sutcliff et al.85 focused on a gene, the serotonin transporter, based on a carefully founded biological hypothesis and some positional information. However, such well-founded candidate approaches are few and far between. One approach championed by some is to study single-gene disorders in which autism is sometimes or often seen, such as the fragile X or Joubert’s syndrome.1,6–10 The study of rare Mendelian forms of other brain disorders has proven fruitful for improving understanding of disease pathophysiology in many cases, including in Alzheimer’s disease and related dementias.11 It is likely that by studying such disorders with autism and autism-related phenotypes in mind, one can identify the shared underlying biological or genetic features that lead to autism. Because the pathways from gene to brain and subsequently to behavior are likely to be far less complex in these single-gene conditions, they may provide a relatively efficient avenue to the understanding of autism pathophysiology. Some of the issues of shared pathophysiology in autism-related single-gene disorders, Rett syndrome, fragile X, and 15q11-13 duplication syndrome, such as chromosome structure or gene expression regulation have been addressed in the chapter by Beaudet and Zoghbi. Given current work in

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 88 Friday, March 10, 2006 2:16 PM

88

Understanding Autism: From Basic Neuroscience to Treatment

the neurobiology of fragile X, it is also reasonable to postulate that the coupling of a rapid translational response at the dendrite in response to neuronal activity could also link these syndromes.12–14 Using microarrays to identify potentially overlapping changes in gene expression between causes of autism with known monogenetic etiologies provides another avenue, albeit untested, to find final common pathways or biomarkers of the disease (see Approach #2). We think that it is worth considering assessment of lymphoblast gene expression, using microarrays to prioritize a subset of candidates based on positional or other genetic data. There are many pitfalls with this approach, including the notion that lymphoblasts are not the brain and, therefore, gene expression changes observed will not be relevant.85 However, the same is true for human brain postmortem tissue, which is not available from the period of development most relevant to the disorder. Experimental support for the utility of blood gene expression profiles to distinguish between acute and chronic central neurologic conditions in animals and man, such as seizures or stroke, comes from the work of Sharp and colleagues.86–89 Recent studies in familial Alzheimer’s disease further support this approach.90 So, although perhaps a high-risk endeavor, we have taken an initial proof of principal approach, asking several related questions: 1. Can we identify gene expression profiles that distinguish between the known less complex forms of autism such as fragile X, 15q11-13, and controls? 2. Can we identify the genes separating autistic from nonautistic subjects? 3. Can we identify candidate genes for repetitive behaviors or other endophenotypes in autism on the basis of gene expression? We first performed a pilot experiment to see whether we could distinguish autistic probands from AGRE with fragile X or 15q11-13 from each other and from controls. We reasoned that this would provide a proof of principal that cases of autism with distinct etiologies could be identified based on lymphoblastoid gene expression. Preliminary experiments in a small group of patients have identified a set of genes that classify these subgroups (Nishimura and Geschwind, unpublished). Interestingly, several genes within the 15q11-13 duplicated region are upregulated in patients lymphoblasts, including GABARB3. In a second set of experiments we have used MZ twin pairs from AGRE who are discordant for RRB to identify a small class of genes highly correlated with this behavior in four twin pairs. One of the most robustly correlated genes (p = .004) was then tested in an independent sample of 10 nontwin autistic probands and shown to have an r2 of .56 (p < .01) with the ADI – R C total score of repetitive behaviors. This gene has no known function but may be located under some suggestive peaks identified in previous studies. We have taken a similar approach with the reciprocal social interaction domain of the ADI, and found that four of the nine genes identified that showed robust correlation with this domain were under a two-Z score support interval of the seven peaks identified by SRS analysis. This suggests that these genes are good candidates for the probable cause underlying the social impairments related to autism, and they are currently undergoing testing for association. All of these microarray-based approaches must be considered preliminary as they are largely untested in autism.

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 89 Friday, March 10, 2006 2:16 PM

Finding Genes in Spite of Heterogeneity

89

However, we hope that by the time the chapter is published to have at least confirmed some of these findings independently and established more conclusively whether this approach is worth pursuing in more depth. Similarly, just as genes contributing to autism are identified with a variety of methods and confirmed, additional candidates for autism may be identified using bioinformatic approaches. For example, using gene expression QTL databases such as WebQTL,91–93 loci that modify the expression of a given autism candidate gene, either in cis or trans can be identified. Those in trans become potential candidates for novel disease loci that can than be tested in patients to identify variants and in models to assess functional interactions. In this manner, single genes associated with autism can be expanded to define disease-relevant pathways. The same, of course, can be done now with the single-gene causes of autism. Genes identified in many such disorders (i.e., fragile X, 15q11-13, Joubert’s, etc.) can be viewed as having garnered circumstantial support for their candidacy in idiopathic autism — thus generating easily testable hypotheses. The hope is that the confluence of such experimental and bioinformatic approaches will help us to more rapidly clarify our current notions of the biological basis of this complex disease.

REFERENCES 1. Spence, S.J., The genetics of autism, Semin Pediatr Neurol 11, 196–204, 2004. 2. Nasreddine, Z.S. et al., From genotype to phenotype: a clinical pathological, and biochemical investigation of frontotemporal dementia and parkinsonism (FTDP-17) caused by the P301L tau mutation, Ann Neurol 45, 704–715, 1999. 3. Boeve, B.F. et al., Corticobasal degeneration and frontotemporal dementia presentations in a kindred with nonspecific histopathology, Dement Geriatr Cognit Disord 13, 80–90, 2002. 4. Bird, T. et al., Epidemiology and genetics of frontotemporal dementia/Pick’s disease, Ann Neurol 54(Suppl. 5), S29–31, 2003. 5. Wassink, T.H., Piven, J., and Patil, S.R., Chromosomal abnormalities in a clinic sample of individuals with autistic disorder, Psychiatr Genet 11, 57–63, 2001. 6. Brown, W.T. et al., Frequency of fragile X in multiplex autism: testing AGRE families, Am J Hum Genet 71, 347, 2002. 7. Feinstein, C. and Reiss, A.L., Autism: the point of view from fragile X studies, J Autism Dev Disord 28, 393–405, 1998. 8. Gillberg, C. and Wahlstrom, J., Chromosome abnormalities in infantile autism and other childhood psychoses: a population study of 66 cases, Dev Med Child Neurol 27, 293–304, 1985. 9. Hallmayer, J. et al., Molecular analysis and test of linkage between the FMR-1 gene and infantile autism in multiplex families, Am J Hum Genet 55, 951–959, 1994. 10. Piven, J., Gayle, J., Landa, R., Wzorek, M., and Folstein, S., The prevalence of fragile X in a sample of autistic individuals diagnosed using a standardized interview, J Am Acad Child Adolesc Psychiatry 30, 825–830, 1991. 11. Tanzi, R.E. and Bertram, L., New frontiers in Alzheimer’s disease genetics, Neuron 32, 181–184, 2001. 12. Dolen, G. and Bear, M.F. Courting a cure for fragile X, Neuron 45, 642–644, 2005. 13. Bear, M.F., Huber, K.M., and Warren, S.T., The mGluR theory of fragile X mental retardation, Trends Neurosci 27, 370–377, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 90 Friday, March 10, 2006 2:16 PM

90

Understanding Autism: From Basic Neuroscience to Treatment 14. Huber, K.M., Gallagher, S.M., Warren, S.T., and Bear, M.F., Altered synaptic plasticity in a mouse model of fragile X mental retardation, Proc Natl Acad Sci USA 99, 7746–7750, 2002. 15. Gottesman, II and Gould, T.D., The endophenotype concept in psychiatry: etymology and strategic intentions, Am J Psychiatry 160, 636–645, 2003. 16. Almasy, L. and Blangero, J., Endophenotypes as quantitative risk factors for psychiatric disease: rationale and study design, Am J Med Genet 105, 42–44, 2001. 17. John, B. and Lewis, K.R., Chromosome variability and geographical distribution in insects: chromosome rather than gene variation provide the key to differences among populations, Science 152, 711–721, 1966. 18. Freimer, N. and Sabatti, C., The human phenome project, Nat Genet 34, 15–21, 2003. 19. Berkovic, S.F. and Scheffer, I.E., Genetics of the epilepsies, Curr Opin Neurol 12, 177–182, 1999. 20. Noebels, J.L., Modeling human epilepsies in mice, Epilepsia 42(Suppl. 5), 11–15, 2001. 21. Frankel, W.N. et al., Development of a new genetic model for absence epilepsy: spikewave seizures in C3H/He and backcross mice, J Neurosci 25, 3452–3458, 2005. 22. Fisher, S.E. et al., Independent genome-wide scans identify a chromosome 18 quantitative-trait locus influencing dyslexia, Nat Genet 30, 86–91, 2002. 23. Willcutt, E.G. et al., Quantitative trait locus for reading disability on chromosome 6p is pleiotropic for attention-deficit/hyperactivity disorder, Am J Med Genet 114, 260–268, 2002. 24. Brown, R. and Silva, A.J., Molecular and cellular cognition; the unraveling of memory retrieval, Cell 117, 3–4 2004. 25. Zoghbi, H.Y., Postnatal neurodevelopmental disorders: meeting at the synapse? Science 302, 826–830, 2003. 26. Shu, W. et al., Altered ultrasonic vocalization in mice with a disruption in the Foxp2 gene, Proc Natl Acad Sci USA 102, 9643–9648, 2005. 27. Bailey, A. et al., Autism as a strongly genetic disorder: evidence from a British twin study, Psychol Med 25, 63–77, 1995. 28. Bolton, P.F., Pickles, A., Murphy, M., and Rutter, M., Autism, affective and other psychiatric disorders: patterns of familial aggregation, Psychol Med 28, 385–395, 1998. 29. Piven, J., Palmer, P., Jacobi, D., Childress, D., and Arndt, S., Broader autism phenotype: evidence from a family history study of multiple-incidence autism families, Am J Psychiatry 154, 185–190, 1997. 30. Pickles, A. et al., Variable expression of the autism broader phenotype: findings from extended pedigrees, J Child Psychol Psychiatry 41, 491–502, 2000. 31. Le Couteur, A. et al., A broader phenotype of autism: the clinical spectrum in twins, J Child Psychol Psychiatry 37, 785–801, 1996. 32. Cook, E.H., Jr., Genetics of autism, Child Adolesc Psychiatr Clin N Am 10, 333–350, 2001. 33. Alarc´on, M., Cantor, R.M., Liu, J., Gilliam, T.C., and Geschwind, D.H., Evidence for a language quantitative trait locus on chromosome 7q in multiplex autism families, Am J Hum Genet 70, 60–71, 2002. 34. Alarc´on, M., Yonan, A.L., Gilliam, T.C., Cantor, R.M., and Geschwind, D.H., Quantitative genome scan and Ordered-Subsets Analysis of autism endophenotypes support language QTLs, Mol Psychiatry, 10(8): 747–757, 2005. 35. Bradford, Y. et al., Incorporating language phenotypes strengthens evidence of linkage to autism, Am J Med Genet 105, 539–547, 2001.

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 91 Friday, March 10, 2006 2:16 PM

Finding Genes in Spite of Heterogeneity

91

36. Buxbaum, J.D. et al., Linkage analysis for autism in a subset families with obsessivecompulsive behaviors: evidence for an autism susceptibility gene on chromosome 1 and further support for susceptibility genes on chromosome 6 and 19, Mol Psychiatry 9, 144–150, 2004. 37. Buxbaum, J.D. et al., Evidence for a susceptibility gene for autism on chromosome 2 and for genetic heterogeneity, Am J Hum Genet 68, 1514–1520, 2001. 38. Folstein, S.E. and Mankoski, R.E., Chromosome 7q: where autism meets language disorder? Am J Hum Genet 67, 278–281, 2000. 39. Nurmi, E.L. et al., Exploratory subsetting of autism families based on savant skills improves evidence of genetic linkage to 15q11-q13, J Am Acad Child Adolesc Psychiatry 42, 856–863, 2003. 40. Shao, Y. et al., Fine mapping of autistic disorder to chromosome 15q11-q13 by use of phenotypic subtypes, Am J Hum Genet 72, 539–548, 2003. 41. Spence, S.J. et al., Stratification based on language-related endophenotypes in autism: attempt to replicate previous reported linkage, Neuropsychiatric Genetics (in press). 42. Blangero, J., Williams, J.T., and Almasy, L., Variance component methods for detecting complex trait loci, Adv Genet 42, 151–181, 2001. 43. Williams, J.T. and Blangero, J., Power of variance component linkage analysis to detect quantitative trait loci, Ann Hum Genet 63, 545–563, 1999. 44. Blangero, J., Williams, J.T., and Almasy, L., Quantitative trait locus mapping using human pedigrees, Hum Biol 72, 35–62, 2000. 45. Stefansson, H. et al., Association of neuregulin 1 with schizophrenia confirmed in a Scottish population, Am J Hum Genet 72, 83–87, 2003. 46. Stefansson, H. et al., Neuregulin 1 and susceptibility to schizophrenia, Am J Hum Genet 71, 877–892, 2002. 47. Shao, Y. et al., Phenotypic homogeneity provides increased support for linkage on chromosome 2 in autistic disorder, Am J Hum Genet 70, 1058–1061, 2002. 48. Warburton, P. et al., Support for linkage of autism and specific language impairment to 7q3 from two chromosome rearrangements involving band 7q31, Am J Med Genet 96, 228–234, 2000. 49. Nakabayashi, K. and Scherer, S.W. The human contactin-associated protein-like 2 gene (CNTNAP2) spans over 2 Mb of DNA at chromosome 7q35, Genomics 73, 108–112, 2001. 50. Verkerk, A.J. et al., CNTNAP2 is disrupted in a family with Gilles de la Tourette syndrome and obsessive compulsive disorder, Genomics 82, 1–9, 2003. 51. Lord, C., Rutter, M., and Le Couteur, A., Autism diagnostic interview-revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders, J Autism Dev Disord 24, 659–685, 1994. 52. Duvall, J.A. et al., Testing candidate genes on 7q and 16 for autism association, Am J Hum Genet (Suppl. 1975), 2004. 53. Duvall, J. et al., Association analysis of 657 SNPS in a candidate autism gene: CNTNAP2, Annual Meeting, International Meeting for Autism Research, Boston, MA, May 5–7, 2005, IMFAR Abstracts, P4A.1.7, 2005. 54. Lord, C. et al., The autism diagnostic observation schedule-generic: a standard measure of social and communication deficits associated with the spectrum of autism, J Autism Dev Disord 30, 205–223, 2000. 55. Lord, C. et al., Diagnosing autism: analyses of data from the autism diagnostic interview, J Autism Dev Disord 27, 501–517, 1997. 56. Bailey, A., Palferman, S., Heavey, L., and Le Couteur, A., Autism: the phenotype in relatives, J Autism Dev Disord 28, 369–392, 1998.

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 92 Friday, March 10, 2006 2:16 PM

92

Understanding Autism: From Basic Neuroscience to Treatment 57. Baron-Cohen, S. et al., Psychological markers in the detection of autism in infancy in a large population, Br J Psychiatry 168, 158–163, 1996. 58. Constantino, J.N. et al., Validation of a brief quantitative measure of autistic traits: comparison of the social responsiveness scale with the autism diagnostic interviewrevised, J Autism Dev Disord 33, 427–433, 2003. 59. Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., and Clubley, E., The autism-spectrum quotient (AQ): evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians, J Autism Dev Disord 31, 5–17, 2001. 60. Constantino, J.N. and Todd, R.D., Intergenerational transmission of subthreshold autistic traits in the general population, Biol Psychiatry 57, 655–660, 2005. 61. Constantino, J.N. and Todd, R.D., Autistic traits in the general population: a twin study, Arch Gen Psychiatry 60, 524–530, 2003. 62. Constantino, J.N., Hudziak, J.J., and Todd, R.D., Deficits in reciprocal social behavior in male twins: evidence for a genetically independent domain of psychopathology, J Am Acad Child Adolesc Psychiatry 42, 458–467, 2003. 63. Sung, Y.J. et al., Genetic investigation of quantitative traits related to autism: use of multivariate polygenic models with ascertainment adjustment, Am J Hum Genet 76, 68–81, 2005. 64. Bolton, P. et al., A case-control family history study of autism, J Child Psychol Psychiatry 35, 877–900, 1994. 65. Geschwind, D. et al., A genomewide linkage scan using the Social Responsiveness Scale (SRS) as a quantitative trait for autism. (Annual Meeting, International Meeting for Autism Research, Boston, MA, May 5–7, 2005). IMFAR Abstracts No. S5.6. 2005. 66. Yonan, A.L. et al., A genomewide screen of 345 families for autism-susceptibility loci, Am J Hum Genet 73, 886–897, 2003. 67. Piven, J. et al., Personality characteristics of the parents of autistic individuals, Psychol Med 24, 783–795, 1994. 68. Silverman, J.M. et al., Symptom domains in autism and related conditions: evidence for familiality, Am J Med Genet 114, 64–73, 2002. 69. Chen, G., Kono, N., Geschwind, D., and C. R., Linkage of autism spectrum disorder: evidence of a non-verbal communication locus on 8q24. (Annual Meeting, International Meeting for Autism Research, Boston, MA, May 5–7, 2005), IMFAR Abstracts No. P1B.2.17, 2005. 70. Schopler, E., Reichler, R.J., DeVellis, R.F., and Daly, K., Toward objective classification of childhood autism: Childhood Autism Rating Scale (CARS), J Autism Dev Disord 10, 91–103, 1980. 71. Carter, A.S. et al., The Vineland adaptive behavior scales: supplementary norms for individuals with autism, J Autism Dev Disord 28, 287–302, 1998. 72. Freeman, B.J., Ritvo, E.R., Yokota, A., Childs, J., and Pollard, J., WISC-R and Vineland Adaptive Behavior Scale scores in autistic children, J Am Acad Child Adolesc Psychiatry 27, 428–429, 1988. 73. Sparrow, S., Balla, D., and Cicchetti, D., Vineland Adaptive Behavior Scales: Interview Edition Survey Form, American Guidance Service, Circle Pines, MN, 1984. 74. Constantino, J.N. et al., The factor structure of autistic traits, J Child Psychol Psychiatry 45, 719–726, 2004. 75. Cuccaro, M.L. et al., Factor analysis of restricted and repetitive behaviors in autism using the autism diagnostic interview-R, Child Psychiatry Hum Dev 34, 3–17, 2003. 76. Tadevosyan-Leyfer, O. et al., A principal components analysis of the autism diagnostic interview-revised, J Am Acad Child Adolesc Psychiatry 42, 864–872, 2003.

© 2006 by Taylor & Francis Group, LLC

2732_C004.fm Page 93 Friday, March 10, 2006 2:16 PM

Finding Genes in Spite of Heterogeneity

93

77. Neale, M.C. and Cardon, L.R., Methodology for Genetic Studies of Twins and Families, Kluwer Academic, Dordrecht, The Netherlands, 1992. 78. Hayduk, L.A., Structural Equation Modeling with LISREL: Essentials and Advances, Johns Hopkins University Press, Baltimore, 1987. 79. Szatmari, P. et al., Quantifying dimensions in autism: a factor-analytic study, J Am Acad Child Adolescent Psychiatr 41, 467–474, 2002. 80. Young, S.E., Stallings, M.C., Corley, R.P., Krauter, K.S., and Hewitt, J.K., Genetic and environmental influences on behavioral disinhibition, Am J Med Genet 96, 684–695, 2000. 81. Posthuma, D., de Geus, E.J.C., Boomsma, D.I., and Neale, M.C., Combined linkage and association tests in Mx, Behavior Genetics 34, 179–196, 2004. 82. Fulker, D.W., Cherny, S.S., Sham, P.C., and Hewitt, J.K., Combined linkage and association sib-pair analysis for quantitative traits, Am J Hum Genet 64, 259–267, 1999. 83. Sham, P.C., Recent developments in quantitative trait loci analysis. In Behavioral Genetics in the Postgenomic Era, eds. Plomin, R., DeFries, J.C., Craig, I.W., and McGuffin, P., APA, Washington, DC, 2002. 84. Cantor, R.M. et al., Replication of autism linkage: fine-mapping peak at 17q21, Am J Hum Genet 76, 2005. 85. Sutcliffe, J.S. et al., Allelic heterogeneity at the Serotonin Transporter Locus (SLC6A4) confers susceptibility to autism and rigid-compulsive behaviors, Am J Hum Genet 77, 265–279, 2005. 86. Geschwind, D.H., DNA microarrays: translation of the genome from laboratory to clinic, Lancet Neurol 2, 275–282, 2003. 87. Tang, Y., Lu, A., Aronow, B.J., Wagner, K.R., and Sharp, F.R., Genomic responses of the brain to ischemic stroke, intracerebral haemorrhage, kainate seizures, hypoglycemia, and hypoxia, Eur J Neurosci 15, 1937–1952, 2002. 88. Tang, Y., Nee, A.C., Lu, A., Ran, R., and Sharp, F.R., Blood genomic expression profile for neuronal injury, J Cereb Blood Flow Metab 23, 310–319, 2003. 89. Tang, Y. et al., Valproic acid blood genomic expression patterns in children with epilepsy — a pilot study, Acta Neurol Scand 109, 159–168, 2004. 90. Tang, Y. et al., Human blood genomics: distinct profiles for gender, age and neurofibromatosis type 1, Brain Res Mol Brain Res 132, 155–167, 2004. 91. Li, H. et al., Inferring gene transcriptional modulatory relations: a genetical genomics approach, Hum Mol Genet 14, 1119–1125, 2005. 92. Bystrykh, L. et al., Uncovering regulatory pathways that affect hematopoietic stem cell function using genetical genomics, Nat Genet 37, 225–232, 2005. 93. Carlborg, O. et al., Methodological aspects of the genetic dissection of gene expression, Bioinformatics 21, 2383–2393, 2005.

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 95 Friday, March 10, 2006 2:17 PM

5

A Mixed Epigenetic and Genetic and Mixed De Novo and Inherited Model for Autism Arthur L. Beaudet and Huda Y. Zoghbi

CONTENTS Possibility of Epigenetic and de novo Factors in Autism.......................................95 Epigenetics and Disease ..........................................................................................96 Genomic Imprinting and Disease............................................................................97 Genetics vs. Epigenetics in Autism.........................................................................98 Relevance of Fragile X Syndrome, Rett Syndrome, and Chromosome 15q11-q13 ................................................................................100 Rett Syndrome and Autism ............................................................................101 Chromosome 15q11-q13 and Autism ............................................................102 Possible Role of Genes on the X or Y Chromosomes in Autism .................105 References..............................................................................................................107

POSSIBILITY OF EPIGENETIC AND DE NOVO FACTORS IN AUTISM The genetic basis of autism is most often hypothesized to follow a multilocus epistatic model with the involvement of as few as 3 to 4 or as many as 15 loci. Some authors1 have favored a model of “oligogenecity with epistasis.” Chapter 3 in this book provides an excellent overview of the genetic data related to the etiology of autism and describes the multilocus epistatic model. These 3 to 15 autism susceptibility loci are usually hypothesized to be mostly autosomal with some possibility of loci on the X chromosome as seen in Figure 3.2 of Chapter 3. The multilocus epistatic model usually assumes that the susceptibility alleles are inherited from the parents and that they involve nucleotide sequence abnormalities. These assumptions are essential for any success in identifying autism susceptibility loci using a genomewide search for genetic linkage or association. 95

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 96 Friday, March 10, 2006 2:17 PM

96

Understanding Autism: From Basic Neuroscience to Treatment

In this chapter, we will review some alternative hypotheses to the multilocus epistatic model that have dramatically different implications for research strategies. We discuss the possibility that the presumed abnormalities of gene expression may be epigenetic (not involving differences in nucleotide sequence) rather than genetic (involving differences in nucleotide sequence) and the possibility that the abnormalities may be de novo in the symptomatic individual and not present, at least in their full form, in the parents. One of us has proposed a mixed epigenetic and genetic and mixed de novo and inherited (MEGDI) model,2 with emphasis on the possibility that de novo and epigenetic factors might predominate, although the model allows for genetic and inherited contributions that are expected to predominate in the multilocus epistatic model. The high male to female sex ratio must be taken into account in any model of autism, and there are two contrasting alternative mechanisms: sex-limited effects of autosomal loci or genes on the X and/or Y chromosomes. The multilocus epistatic model could be compatible with either mechanism, but most of the focus is on autosomal loci. In this chapter, we will explore multiple mechanisms whereby genes on the X or Y chromosomes might contribute to autism. Epigenetics is often defined as the study of changes in gene function that are stable and heritable (or potentially heritable as in terminally differentiated neurons) and do not entail a change in DNA sequence. An alternative perspective (source Adrian Bird) defines epigenetics as the adaptation of chromosomal regions so as to perpetuate local activity states, whether of long or short duration and whether inherited or not. The first definition focuses on mitotic and meiotic heritability and can include posttranscriptional mechanisms such as prion-based inheritance. The second definition is narrower mechanistically, with the focus on chromatin and the processes for its modification, but broader in terms of duration and heritability of the effects. The chromatin modifications that accompany the cell cycle might be excluded by the first definition and included by the second. Typically, epigenetic regulation involves a signal or stimulus that changes gene expression, and when the signal is no longer present, the change in gene expression persists. There is extensive information regarding the biochemical basis of epigenetic regulation, particularly in the form of chromatin remodeling, and excellent reviews are available.3–10 Epigenetic regulation includes DNA methylation, covalent modification of histones (the so-called “histone code”), and the action of a wide variety of nonhistone proteins involved in chromatin remodeling, including the polycomb group of proteins, methylated DNA binding proteins, and proteins with chromodomains. Because genomic DNA must exist in a particular chromatin configuration, the genotype can only give rise to phenotype through the prism of the epigenotype (Figure 5.1). The role of epigenetics in the etiology of cancer has become widely appreciated,11,12 but its contribution to phenotype, apart from cancer, has only recently begun attracting greater attention,13–15 including the possibility of roles in learning, memory, behavior, and psychiatric illness.16–18

EPIGENETICS AND DISEASE There are genetic disorders that affect chromatin structure and remodeling. These disorders can affect chromatin in trans or in cis. Rett syndrome and fragile X syndrome are two X-linked genetic conditions that include autism spectrum disorders

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 97 Friday, March 10, 2006 2:17 PM

A Mixed Epigenetic and Genetic

97

Environment

Genotype

Phenotype Epigenotype

Stochastic events

FIGURE 5.1 (A color version of this figure follows page 236.) The genotype affects the phenotype only through the prism of the epigenotype. (Taken from Beaudet, A.L., Is medical genetics neglecting epigenetics? Genet Med 4, 399–402, 2002. With permission.)

in their phenotypic variation, and they alter chromatin in trans or in cis, respectively. Rett syndrome is caused by mutations, usually de novo, affecting the locus encoding methyl-CpG-binding protein 2 (MeCP2). In the case of Rett syndrome, it is likely that expression is altered for many loci with cell autonomous deficiency of MeCP2 leading to failure to deacetylate histones and repress gene expression. Fragile X syndrome is caused by expansion of a CGG triplet repeat in the 5-noncoding region of transcript from the fragile X mental retardation gene (FMR1) encoding the fragile X protein (FMRP). The repeat expansion causes a secondary DNA methylation and silencing of expression at the FMR1 locus. Thus, the two single-gene disorders that most often lead to a typical autism phenotype are X-linked and have secondary effects on chromatin structure.

GENOMIC IMPRINTING AND DISEASE Genomic imprinting is a subset of epigenetic regulation in which the activity of a gene is reversibly modified, depending on the sex of the parent who transmits it; see reviews.4,19,20 This leads to unequal expression from the maternal and paternal alleles of a diploid locus. Some of the most obvious evidence for the role of epigenetics in the etiology of human disease comes from disorders of genes subject to genomic imprinting. Mutations and epimutations in imprinted genes can give rise to genetic and epigenetic phenotypes, respectively; uniparental disomy (UPD, inheritance of two copies of a chromosome from one parent and none from the other) and imprinting defects represent epigenetic disease phenotypes, whereas Mendelian disorders typically involve genetic mutations rather than epimutations. The imprinted domain on human chromosome 15q11-q13 is of particular interest in autism because maternal, but not paternal, duplications of this region can cause autism as discussed in the text that follows. Prader–Willi syndrome (PWS) and Angelman syndrome (AS) are caused by mixed epigenetic and genetic effects leading to deficiency of gene expression from the paternal chromosome 15q11-q13 in PWS and from the maternal chromosome in AS. Most cases of PWS and AS are caused by 5 to 6 Mb

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 98 Friday, March 10, 2006 2:17 PM

98

Understanding Autism: From Basic Neuroscience to Treatment

Prader-Willi Deletion Genetic

70%

UPD Epigenetic

30%

Paternal deficiency 15q11-q13

Angelman Deletion Genetic

70%

UPD Epigenetic

Rare

Maternal deficiency 15q11-q13

All deletions and UPD are de novo events

FIGURE 5.2 (A color version of this figure follows page 236.) Depiction of genomic deletion or uniparental disomy (UPD) giving rise to Prader–Willi or Angelman syndromes. (Modified from Jiang, Y.H., Bressler, J., and Beaudet, A.L., Epigenetics and human disease, Annu Rev Genomics Hum Genet 5, 479–510, 2004. With permission.)

deletions of paternal or maternal 15q11-q13, respectively, but maternal UPD for chromosome 15 also causes PWS, and paternal UPD causes AS. At a gene level, AS is caused by maternal deficiency for the E6-AP ubiquitin ligase (gene symbol UBE3A), which is imprinted with brain-specific silencing of the paternal allele. PWS is caused by paternal deficiency for some subset of the multiple imprinted protein coding transcripts and noncoding snoRNAs that are transcriptionally silenced on the maternal chromosome. The data from PWS and AS highlight the fact that the same clinical phenotype can be caused by genetic or epigenetic defects as in the case of deletion or UPD causing PWS or AS, respectively (Figure 5.2). UPD most often is caused by trisomic or monosomic conception for that chromosome followed by cytogenetic “rescue” to yield two copies of the chromosome from one parent.21,22 In deletion cases of PWS or AS, there is a sequence abnormality involving absence of 5 to 6 Mb of DNA. In UPD cases, the genomic sequence of the individual is entirely normal. The genetic and epigenetic changes in disorders such as PWS and AS can occur de novo or be inherited, and these disorders are excellent examples of monogenic or oligogenic phenotypes caused by a MEGDI model (Figure 5.3). The MEGDI model is also applicable to Beckwith–Wiedemann syndrome and to phenotypes associated with altered expression of the guanine nucleotide-binding protein’s -stimulating activity (GNAS) locus/complex, including McCune-Albright syndrome, Albright hereditary osteodystrophy, and various forms of pseudohypoparathyroidism.23–25

GENETICS VS. EPIGENETICS IN AUTISM The genetic contribution to the etiology of autism is thought to be significant largely because the concordance in monozygotic (MZ) twins is quite high, being approximately ~60% if the autism phenotype is narrowly defined and ~90% if the phenotype

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 99 Friday, March 10, 2006 2:17 PM

A Mixed Epigenetic and Genetic

99

De novo

Epigenetic

Monogenic or oligogenic MEGDI model

Inherited

Genetic

MEGDI = mixed epigenetic & genetic and de novo & inherited

FIGURE 5.3 An oligogenic mixed epigenetic and genetic and mixed de novo and inherited (MEGDI) model for Prader–Willi syndrome, Angelman syndrome, Beckwidth–Weidemann syndrome, disorders of the GNAS locus, and possibly autism. (From Jiang, Y.H., Bressler, J., and Beaudet, A.L., Epigenetics and human disease, Annu Rev Genomics Hum Genet 5, 479–510, 2004. With permission.)

is broadly defined; see Chapter 3 for a detailed review and discussion of the twin studies. The concordance in dizygotic (DZ) twins is much lower and is similar to the recurrence risk in siblings. As stated in Chapter 3, “the heritability estimate (H) for autism (calculated from the recurrence risk and the MZ:DZ concordance rates) is ≥ 90%, suggesting that autism is one of the most heritable neuropsychiatric disorders.” Heritability is usually defined as that proportion of the observed variation in a particular phenotype, and in a particular study, that can be attributed to the contribution of genotype (inheritance). Interestingly, the role of the epigenotype is rarely mentioned in the context of heritability. The twin data for autism have been interpreted by Risch et al.26 in the context of the multilocus epistatic model as follows: “The very high (25-fold) MZ:DZ concordance ratio is indicative of at least several interacting loci and, potentially, of many such loci.” In contrast, we believe that it is instructive to compare the heritability of autism to that of trisomy 21. For trisomy 21, the concordance for MZ twins is essentially 100%, whereas that for DZ twins is approximately the same as the recurrence risk in siblings. Thus, the heritability for trisomy 21 is ~100%, but the phenotype is not inherited in the sense that the offspring receives a mutant genotype or nucleotide sequence aberration that is present in a parent. Thus, there is a seeming contradiction in that trisomy 21 has a high heritability but is usually not inherited but, rather, is of de novo origin. We believe that the etiology of autism could be similar to trisomy 21 in the sense that there may be a major component involving de novo mutation or epimutation. A high MZ concordance and low DZ concordance are also to be expected for a disorder such as Rett syndrome, where new mutations cause the majority of cases, demonstrating that a very high MZ:DZ concordance ratio can occur for a single-gene disorder.

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 100 Friday, March 10, 2006 2:17 PM

100

Understanding Autism: From Basic Neuroscience to Treatment

Based on a multilocus model, numerous genome-wide scans for genetic linkage or association have been performed as is reviewed in Chapter 3. The predictions of the multilocus epistatic model and the MEGDI model could hardly be more different in the context of genome-wide scans. The multilocus epistatic model predicts that autism loci, mutations, and haplotypes should be discoverable. Failure to discover these might be caused by heterogeneity; stratification of the patients by subphenotypes as well as the study of genetically isolated populations might increase the chance for success. Genome-wide scans based on the study of nucleotide sequence differences in blood-derived DNA from affected sib pairs or parent–child trios cannot identify autism genes if the events are epigenetic rather than genetic or if the events are de novo rather than inherited. In addition, epigenetic events cannot be detected using blood-derived DNA if the epigenetic changes are brain-specific. Although the MEGDI model would be the same as the multilocus model to the extent that factors are genetic mutations inherited from the parents, it would emphasize the importance of characterizing the epigenotype in brain tissue from affected individuals. This would likely involve studies of DNA methylation and characterization of chromatin, using methods such as chromatin-immunoprecipitation (ChIP). The MEGDI model would allow for the possibility that some of the autism loci as shown in Figure 3.2 of Chapter 3 might represent true genetic factors for susceptibility genotypes in the parents that predispose to de novo events affecting offspring or for premutations or preepimutations in parents progressing to full mutations or full epimutations in the offspring.

RELEVANCE OF FRAGILE X SYNDROME, RETT SYNDROME, AND CHROMOSOME 15q11-q13 There are many examples in medical genetics where findings in rare or unique patients have led to the discovery of genes or principles that are relevant to the majority of individuals with a similar phenotype. Studies of rare cases of homozygous familial hypercholesterolemia led to a cascade of subsequent studies that have provided extensive understanding of cholesterol metabolism in general. Unique individuals with large genomic deletions led to the identification of the genes for polyposis of the colon27 and for CHARGE syndrome,28 although the majority of affected patients have point mutations. It was the recognition of the deletion cause for PWS and AS that allowed for the discovery of the more enigmatic UPD as an epigenetic cause of these syndromes. Multiple families in the Autism Genetic Resource Exchange (AGRE) collection have been found to have fragile X mutations (see http://www.agre.org/), and at least one family has a duplication of chromosome 15q11-q13.2 There are multiple reports of MECP2 mutations in cases correctly diagnosed as having an autism phenotype. Can further studies of fragile X syndrome, Rett syndrome, and the PWS or AS imprinted domain clarify the etiology of the majority of cases of autism? The MeCP2 Rett syndrome protein, the FMRP fragile X protein, and the E6AP AS protein all have downstream effects on other genes and gene products. FMRP is an RNA-binding protein that affects translation of multiple transcripts, whereas the effects are thought to be on transcription for MeCP2 and on protein degradation

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 101 Friday, March 10, 2006 2:17 PM

A Mixed Epigenetic and Genetic

101

for E6-AP. One possibility is that all of these disorders affect some common set of downstream genes and protein products.

RETT SYNDROME

AND

AUTISM

Rett syndrome (RTT) is a dominant X-linked postnatal neurodevelopmental disorder characterized by motor abnormalities, seizures, loss of hand use, and language regression.29 RTT is classified as one of the autistic spectrum disorders (ASD) in DSM-IV.30 Both Rett and autism (to a less consistent extent) are manifested after a period of apparently normal development, both disrupt social and language development, and both are accompanied by unusual stereotyped movements.31 The overlap in clinical features, together with the timing of onset, raises the possibility that some neurobiological mechanisms are shared between these disorders. The discovery that RTT is caused by mutations in the gene encoding the MeCP2 protein32 provided molecular evidence for a relationship between Rett syndrome and autism. MECP2 mutations cause classic RTT as well as a broad spectrum of phenotypes, including isolated mental retardation, ASD, or an Angelman-like syndrome.33 MECP2 mutations usually cause classic Rett syndrome in females if the patterns of X-chromosome inactivation (XCI) are balanced (~50:50) but cause mild mental retardation or autism if XCI patterns are unbalanced, favoring the normal MECP2 allele.34–36 More recently we identified MECP2 mutations in four female patients and two males that fulfill the DSM-IV criteria for autism (Neul, Zoghbi, Roa, and Glaze, unpublished data). These patients have impaired social and language functions but none of the other features of Rett syndrome. The finding of isolated autism phenotypes in females with Rett syndrome-causing MECP2 mutations argues that mosaic loss of function of MeCP2 in cells (due to XCI patterns) is sufficient to produce autism spectrum phenotypes. Mouse studies lend further support to the role of MeCP2 in mediating social behavior. Studies of the Mecp2308/Y mice, which reproduce most phenotypes of Rett syndrome,37 revealed social behavior abnormalities reminiscent of autism-like behaviors.38 Taken together, the human data and mouse behavioral data underscore the importance of MeCP2 in regulating the expression and function of the genes involved in social behavior. In addition, there is evidence that approximately doubling the levels of MeCP2 protein using BAC transgenic lines causes postnatal developmental phenotypes in mice that can be rescued by breeding the transgenic lines to MeCP2-null mice.39 Restoring MeCP2 levels to close to the normal level rescues both the lethality of null males and the postnatal phenotypes seen with overexpression. These data argue that MeCP2 levels are very tightly regulated and that any disturbances of such levels lead to progressive postnatal neurological disorders. Human studies are beginning to corroborate the data observed in mice. A duplication of a genomic region spanning MECP2 was found to cause a Rett-like phenotype in a male,40 and we (Zoghbi, Beaudet, and colleagues) have identified a second unpublished case; in addition, increased MECP2 RNA levels have been observed in autism spectrum patients.41 Duplication of the MECP2 region appears to be a frequent cause of severe neurological abnormalities in males.41a These data together make the prediction that subtle alterations in MECP2 RNA levels

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 102 Friday, March 10, 2006 2:17 PM

102

Understanding Autism: From Basic Neuroscience to Treatment

(either due to epigenetic, genetic, or RNA stability regulation) might explain autismlike phenotypes in males or females. There are two reports suggesting that deficiency of MeCP2 lowers expression of UBE3A/Ube3a in human and mouse brain, and this would provide a link whereby a Rett mutation might cause a delayed onset of deficiency for E6-AP.41,42 One report indicates that male mice null for MeCP2 have biallelic expression of the antisense and biallelic silencing of the sense transcripts for Ube3a.42 There is another report of reduced expression of UBE3A in Mecp2deficient mice and in human Rett, AS, and autism brains compared with controls.43 This report also described reduced expression of GABRB3 in multiple Rett, AS, and autism brain samples.

CHROMOSOME 15q11-q13

AND

AUTISM

Very extensive information is available regarding the genomic architecture, protein coding genes, noncoding RNAs, and the related phenotypic abnormalities for the PWS or AS imprinted domain within human chromosome 15q11-q13 (Figure 5.4). As discussed earlier, PWS and AS provide excellent examples of monogenic or oligogenic disorders caused by a mixture of genetic and epigenetic and of de novo and inherited factors that led to the formulation of the MEGDI model.

E6-AP

bp1

bp2

UBE3A sense UBE3A antisense

IC-DMR

bp3

PWS proteins & RNAs

ER C2 H

P

3A BE U

AT P1 0A G AB G RB AB 3 G RA AB 5 RG 3

N sn oR

N

cen

IP A N 1 IP CY A2 FI G P1 CP S M KR M N3 AG N EL D 2 C1 N 5 SN OR U F2 RF -S N

RP

As

N

5-6 Mb region tel

FIGURE 5.4 (A color version of this figure follows page 236.) Genomic structure of the imprinted domain of human chromosome 15q11-q13. In the top section, black arrows represent products from nonimprinted genes, blue arrows products expressed preferentially from the paternal chromosome, and pink arrows products expressed preferentially from the maternal chromosome. E6-AP is the protein produced from the Angelman ubiquitin ligase gene UBE3A. IC-DMR is the imprinting center differentially methylated region. The identity of gene symbols can be found in any genome browser.

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 103 Friday, March 10, 2006 2:17 PM

A Mixed Epigenetic and Genetic

103

Apart from abnormalities of the sex chromosomes, duplications of chromosome 15 are the most common cytogenetic abnormality in autism. The duplications can be interstitial or take the form of supernumerary isodicentric chromosomes. A report that maternal, but not paternal, interstitial duplication of human chromosome 15q11q13 caused autism greatly increased the focus on chromosome 15 and the potential for involvement of genomic imprinting in autism.44 There are many reports of extra isodicentric 15q chromosomes being found in association with mental retardation and autistic features.45–47 The isodicentric chromosomes are most often of maternal origin, but supernumerary chromosomes of this type are predominantly of maternal origin for other chromosomes as well,48 making the interpretation of the parent of origin of the isodicentric cases less compelling than that for the interstitial duplications. A series of such patients was characterized at a molecular level using array comparative genomic hybridization.49 The extent to which maternal, but not paternal, interstitial duplications of 15q11q13 cause autism is of considerable relevance to the potential for a broader involvement of this region in autism. In one recent report of 16 patients with independent interstitial duplications and 3 interstitial triplications, the rearrangement was maternal in origin for all but 1 proband.50 There was one case with paternal transmission, with an abnormal phenotype of developmental delay and a behavioral disorder, and this exception could be explained by a variety of molecular mechanisms. Another unusual case is that of three sibs with a maternally inherited interstitial duplication of 15q11-q13 and a language disorder; the phenotype of the mother was normal although the duplication was of grandmaternal origin.51 However, other reports find that the phenotype for interstitial duplications is often not typical for autism but intellectual impairment is frequent; they confirm that the interstitial duplications have greater phenotypic effects when of maternal origin.52 The UBE3A gene maps within 15q11-q13 and deserves special attention because of its brain- or neuron-specific imprinting53–56 and known neurobehavioral phenotype, but an adjacent cluster of GABA receptor genes (GABRB3, GABRA5, and GABRG3) also deserves attention because of the relevance to neuronal function. In an attempt to identify epigenetic defects of 15q11-q13 in autism brain samples, particularly focusing on UBE3A, one of the most significant findings was the discovery of an abnormality of DNA methylation involving the CpG island at the 5-end of UBE3A in one of 17 autism brains as shown in Figure 5.5.2 Given the evidence that maternal, but not paternal, duplications of chromosome 15q11-q13 cause autism, this finding of an epigenetic abnormality involving the Angelman gene in this region is highly significant in our view, even though it involves only a single case. One obvious hypothesis related to the role of maternal duplications of 15q11-q13 in autism would be that overexpression of UBE3A might cause autism, but analysis of E6-AP protein in autism brain samples using western blotting was not consistent with this possibility.2 Instead, the data suggested that reduction of E6-AP may be present in some cases of autism, although the potential for postmortem degradation artifacts makes the significance of these findings uncertain. The level of E6-AP was not reduced in the brain with the abnormality of DNA methylation for UBE3A. Given the varied evidence for involvement of 15q11-q13 in autism, the indication of a parent-of-origin effect, an abnormality of DNA methylation at UBE3A in one autism brain, and the parallels with AS, the

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 104 Friday, March 10, 2006 2:17 PM

104

Understanding Autism: From Basic Neuroscience to Treatment

Cx

Cx

Ce

Liv

LB

FB

a EcoRI

EcoRI + NotI

12 Control cerebellum samples b EcoRI

EcoRI + NotI

Ha-3511 Ce Cx

c

Ha-3511 Ce Cx

Co

Co

d

EcoRI + NotI

BssHII + EcoRI

FIGURE 5.5 Southern blot analysis of methylation of the CpG island at the 5of UBE3A. (a) Control samples. Cx, cerebral cortex; Ce, cerebellum; Liv, liver; LB, lymphoblasts; FB, fibroblasts. (b) Control autopsy cerebellum samples. (c)–(d): Brain from autism case Ha-3511 (cerebellum, Ce and cerebral cortex, Cx) and cerebellum for two other autism cases and from a control (Co) using two different methylation-sensitive restriction enzymes NotI and BssHII within the CpG island of UBE3A. (Reproduced from Jiang, Y.H. et al., A mixed epigenetic/ genetic model for oligogenic inheritance of autism with a limited role for UBE3A, Am J Med Genet A 131, 1–10, 2004. With permission.)

possibility of involvement of 15q11-q13 in a MEGDI model for autism is quite attractive (Figure 5.6). The potential involvement of the GABA receptor genes within 15q11-q13 in autism is unclear. The entire cluster of GABA receptor genes is duplicated in the maternal duplications causing autism. Many studies have focused on the GABRB3 gene, which is the member of the cluster that is nearest to UBE3A. Most investigators report that these GABA receptor genes are not subject to genomic imprinting,57 although there are conflicting reports.58,59 There are reports of linkage disequilibrium and association with autism across the GABA(A) receptor subunit cluster.60,61 Another group found only marginal evidence for linkage disequilibrium but emphasized the need for further assessment of whether GABRB3, in particular, might be subject to genomic imprinting.62 Another gene in 15q11-q13 encodes an ATPase (ATP10A) and is imprinted with preferential maternal expression, but studies to date have not found evidence for involvement of this gene in autism.63 One interesting possibility would be that events in the PWS or AS domain have position effects on the GABA receptor cluster. For example, GABRB3 may not be

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 105 Friday, March 10, 2006 2:17 PM

A Mixed Epigenetic and Genetic

105 Autism?

Angelman

De novo De novo De novo & & & inherited inherited inherited

De novo

De novo

Deletion

UPD

Imprint defect

UBE3A null

Genetic

Epigenetic

Mixed

Genetic

De novo

?

?

?

Interstial Isodiduplic centric

Paternal imprint defect?

Maternal imprint defect?

Other gene ?

Genetic Genetic

Mixed?

Mixed?

Mixed?

?

?

∗

Frequent?

FIGURE 5.6 (A color version of this figure follows page 236.) Depiction of the MEGDI model for Angelman syndrome and autism. The various molecular forms of Angelman syndrome are depicted on the left. The interstitial duplication and isodicentric examples for autism are well documented. The DNA methylation abnormality in one of 17 autism brains likely represents a paternal or maternal imprinting defect.

imprinted normally, but spreading of silencing from UBE3A to GABRB3 might cause autism. There are numerous examples in epigenetic regulation where extra copies of a gene lead to trans-acting silencing of all copies of the gene.64

POSSIBLE ROLE

OF

GENES

ON THE

X

OR

Y CHROMOSOMES

IN

AUTISM

The sex bias in autism is a very strong effect, with a male to female ratio of 4:1 in most studies. The sex ratio rises to 7.5:1 for cases without malformations or dysmorphic features65 and to 8:1 for autistic children being educated in mainstream schools66; even higher ratios have been reported occasionally for Asperger syndrome. There are two contrasting mechanisms for the major effect of sex on the risk of autism: either a sex-limited effect on autosomal loci or a major gene on the X or Y chromosome. Because most of the focus in the multilocus epistatic model is on autosomal regions, it would appear that most investigators favor the sex-limited hypothesis. This would be similar to the risk of breast cancer, where major autosomal loci such as BRCA1 and BRCA2 can confer increased risk but the level of risk is very dependent on the sex of the individual.

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 106 Friday, March 10, 2006 2:17 PM

106

Understanding Autism: From Basic Neuroscience to Treatment

There are numerous possible mechanisms whereby genes on the X or Y chromosome might account for the biased sex ratio in autism; these include the following: 1. Mutations or epimutations on Y, with loss or gain of function 2. Mutations or epimutations on X, with loss or gain of function and with recessiveness in females We are not aware of much evidence as to whether epimutations might be dominant or recessive, although they presumably could be either and might involve imprinted loci. For known cases of imprinting defects, the involved gene is subject to genomic imprinting such that the epimutation might be considered to be dominant in one sex and recessive in the other. Similarly, there is limited information as to whether epimutations are heritable, semiheritable, or nonheritable, although they could be any of these three, depending on whether the epimutation affects germ cells or not and whether there is complete, partial, or no erasure and resetting of epigenetic marks affecting germ cells. Turning to genetic factors first and then to epigenetics, based on available data, certain types of mutations on the X or Y chromosome are unlikely, or at least less likely, to be major contributors. Inherited mutations on the X chromosome are unlikely because they would have been detected by linkage studies such as analysis of sib pairs. Although there may be a weak locus on the X chromosome (see Figure 3.2 of Chapter 3), the effect is not large enough to explain the sex bias. Rarely, mutations in the genes for neuroligins 3 and 4 may cause X-linked mental retardation with autistic features67 or typical autism,68 although such mutations were rare when large series of patients were studied.69,70 Postsynaptic neuroligins 3 and 4 are thought to play an important role in synapse formation. Interestingly, neuroligin 2 is exclusively expressed in inhibitory synapses,71–73 and knocking down the function of Nlgn 1, 2, and 3 in rodents causes a selective decrease in inhibitory synaptic function.73 These data suggest that functional inhibitory synapses are more dependent on neuroligins and that decreased inhibitory synapse function might contribute to autism spectrum phenotypes. The role of the fragile X syndrome in autism was discussed earlier. De novo mutations on the X chromosome cannot be ruled out. In the case of the Y chromosome, it is very difficult to distinguish sex-limited effects from effects of mutations in genes on the Y chromosome because maleness and almost all the genes on Y are inherited as a block of genome, without recombination. The fact that a father transmits a Y chromosome rather than an X is by far the largest genetic risk factor in autism, but this fact does not help to distinguish sex-limited effects on autosomal genes from the effects of genes on the X or Y chromosomes. Ancient mutations on the Y chromosome with an association to autism are unlikely, based on at least one study of Y chromosome haplotypes in autism compared to control subjects.74 Inherited mutations on the Y chromosome are unlikely because father-to-son transmission of a severe autism phenotype is rare. This interpretation is subject to some doubt because severely affected individuals probably have a marked reduction in reproductive fitness, although there is little published data on this issue. The possibility that milder (e.g., Asperger) phenotypes might be transmitted from father to son is less certain, with little published

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 107 Friday, March 10, 2006 2:17 PM

A Mixed Epigenetic and Genetic

107

in the medical literature on the reproductive fitness in Asperger syndrome, to our knowledge. Turning to epigenetic defects, the vast majority of genetic data on autism are based on analysis of nucleotide sequence variations in leukocyte or cultured cell DNA, and they do not address the possibility of epigenetic defects directly. Therefore, virtually any form of epigenetic defect involving a gene on the X and/or Y chromosomes is possible and could easily explain the male predominance in autism. Epigenetic effects are variable in their heritability and are often described as being semiheritable. As an example, one could hypothesize that recessive epigenetic defects increasing or decreasing expression of MECP2 are a common cause of autism, and this could explain the sex bias. Epigenetic defects of chromatin may or may not include abnormalities of DNA methylation, so that ChIP studies might be required in addition to studies of DNA methylation, and the abnormalities might be brain-specific. Loss-of-function genetic or epigenetic defects on the Y chromosome may be relatively unlikely because patients with a 45X karyotype are not autistic in most cases. However, there are reports of autism in 45X females, and there is a suggestion that the autistic patients more often have a maternal X and lack a paternal X.75,76 It has been suggested that there is a maternally silenced, imprinted gene near the X centromere that accounts for the more significant cognitive differences in 45X patients lacking a maternal X compared to those lacking a paternal X.76 Gain-offunction mutations or epimutations on the Y chromosome are feasible. As an example, the gene for synaptobrevin-like 1 (SYBL1) is located in the small pseudoautosomal region on the long arm of the X and Y chromosomes. Functional homology implicates this gene in vesicle trafficking. The gene is subject to random X-inactivation in females, but the copy on the Y chromosome is silenced in males. We have hypothesized that an epigenetic failure to silence the copy on the Y chromosome could cause autism. However, we have been unable to detect any abnormalities of DNA methylation or any sequence abnormalities in autism brain or lymphoblasts (Ben-Shachar, Shinawi, Jiang, and Beaudet, unpublished). Candidate genes on the X and Y chromosome that might be affected by de novo epigenetic abnormalities to cause autism could include MECP2, FMR1, neuroligin 4 on X or Y (NLGN4 or NLGN4Y), neuroligin 3 on X (NLGN3), protocadherin on X or Y (PCDHX or PCDHY). A thorough testing of these candidate genes for epigenetic abnormalities would require, at a minimum, studies of DNA methylation and ChIP on autism brain tissue.

REFERENCES 1. Folstein, S.E. and Rosen-Sheidley, B., Genetics of autism: complex aetiology for a heterogeneous disorder, Nat Rev Genet 2, 943–955, 2001. 2. Jiang, Y.H. et al., A mixed epigenetic/genetic model for oligogenic inheritance of autism with a limited role for UBE3A, Am J Med Genet A 131, 1–10, 2004. 3. Jenuwein, T. and Allis, C.D., Translating the histone code, Science 293, 1074–1080, 2001. 4. Reik, W., Dean, W., and Walter, J., Epigenetic reprogramming in mammalian development, Science 293, 1089–1093, 2001.

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 108 Friday, March 10, 2006 2:17 PM

108

Understanding Autism: From Basic Neuroscience to Treatment

5. Li, E., Chromatin modification and epigenetic reprogramming in mammalian development, Nat Rev Genet 3, 662–673, 2002. 6. Felsenfeld, G. and Groudine, M., Controlling the double helix, Nature 421, 448–453, 2003. 7. Jaenisch, R. and Bird, A., Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals, Nat Genet 33(Suppl.), 245–254, 2003. 8. Lachner, M., O’Sullivan, R.J., and Jenuwein, T., An epigenetic road map for histone lysine methylation, J Cell Sci 116, 2117–2124, 2003. 9. Studitsky, V.M., Walter, W., Kireeva, M., Kashlev, M., and Felsenfeld, G., Chromatin remodeling by RNA polymerases, Trends Biochem Sci 29, 127–135, 2004. 10. Ringrose, L. and Paro, R., Epigenetic regulation of cellular memory by the Polycomb and Trithorax group proteins, Annu Rev Genet 38, 413–443, 2004. 11. Jones, P.A. and Laird, P.W., Cancer epigenetics comes of age, Nat Genet 21, 163–167, 1999. 12. Feinberg, A.P., The epigenetics of cancer etiology, Semin Cancer Biol 14, 427–432, 2004. 13. Egger, G., Liang, G., Aparicio, A., and Jones, P.A., Epigenetics in human disease and prospects for epigenetic therapy, Nature 429, 457–463, 2004. 14. Bjornsson, H.T., Fallin, M.D., and Feinberg, A.P., An integrated epigenetic and genetic approach to common human disease, Trends Genet 20, 350–358, 2004. 15. Jiang, Y.H., Bressler, J., and Beaudet, A.L., Epigenetics and human disease, Annu Rev Genomics Hum Genet 5, 479–510, 2004. 16. Petronis, A. et al., Psychiatric epigenetics: a new focus for the new century, Mol Psychiatry 5, 342–346, 2000. 17. Petronis, A., The origin of schizophrenia: genetic thesis, epigenetic antithesis, and resolving synthesis, Biol Psychiatry 55, 965–970, 2004. 18. Levenson, J.M. and Sweatt, J.D., Epigenetic mechanisms in memory formation, Nat Rev Neurosci 6, 108–118, 2005. 19. Reik, W. and Walter, J., Genomic imprinting: parental influence on the genome, Nat Rev Genet 2, 21–32, 2001. 20. Arney, K.L., Erhardt, S., Drewell, R.A., and Surani, M.A., Epigenetic reprogramming of the genome — from the germ line to the embryo and back again, Int J Dev Biol 45, 533–540, 2001. 21. Ledbetter, D.H. and Engel, E., Uniparental disomy in humans: development of an imprinting map and its implications for prenatal diagnosis, Hum Mol Genet 4(Spec. No.), 1757–1764, 1995. 22. Engel, E. and Antonarakis, S.E., Genomic Imprinting and Uniparental Disomy in Medicine: Clinical and Molecular Aspects, Wiley-Liss, New York, 2002. 23. Nicholls, R.D. and Knepper, J.L., Genome organization, function, and imprinting in Prader–Willi and Angelman syndromes, Annu Rev Genomics Hum Genet 2, 153–175, 2001. 24. Weinstein, L.S., Chen, M., and Liu, J., Gs(alpha) mutations and imprinting defects in human disease, Ann N Y Acad Sci 968, 173–197, 2002. 25. Ruggieri, V.L. and Arberas, C.L. Behavioral phenotypes. Biologically determined neuropsychological patterns, Rev Neurol 37, 239–253, 2003. 26. Risch, N. et al., A genomic screen of autism: evidence for a multilocus etiology, Am J Hum Genet 65, 493–507, 1999. 27. Bodmer, W.F. et al., Localization of the gene for familial adenomatous polyposis on chromosome 5, Nature 328, 614–616, 1987.

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 109 Friday, March 10, 2006 2:17 PM

A Mixed Epigenetic and Genetic

109

28. Vissers, L.E. et al., Mutations in a new member of the chromodomain gene family cause CHARGE syndrome, Nat Genet 36, 955–957, 2004. 29. Hagberg, B., Aicardi, J., Dias, K., and Ramos, O., A progressive syndrome of autism, dementia, ataxia, and loss of purposeful hand use in girls: Rett’s syndrome: report of 35 cases, Ann Neurol 14, 471–479, 1983. 30. Association, A.P., Diagnostic and Statistical Manual of Mental Disorders 4th ed. (text revision), Arlington, VA, 2000. 31. Zoghbi, H.Y., Postnatal neurodevelopmental disorders: meeting at the synapse? Science 302, 826–830, 2003. 32. Amir, R.E. et al., Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2, Nat Genet 23, 185–188, 1999. 33. Neul, J.L. and Zoghbi, H.Y., Rett syndrome: a prototypical neurodevelopmental disorder, Neuroscientist 10, 118–128, 2004. 34. Shahbazian, M.D., Sun, Y., and Zoghbi, H.Y., Balanced X chromosome inactivation patterns in the Rett syndrome brain, Am J Med Genet 111, 164–168, 2002. 35. Wan, M. et al., Rett syndrome and beyond: recurrent spontaneous and familial MECP2 mutations at CpG hotspots, Am J Hum Genet 65, 1520–1529, 1999. 36. Carney, R.M. et al., Identification of MeCP2 mutations in a series of females with autistic disorder, Pediatr Neurol 28, 205–211, 2003. 37. Shahbazian, M. et al., Mice with truncated MeCP2 recapitulate many Rett syndrome features and display hyperacetylation of histone H3, Neuron 35, 243–254, 2002. 38. Moretti, P., Bouwknecht, J.A., Teague, R., Paylor, R., and Zoghbi, H.Y., Abnormalities of social interactions and home-cage behavior in a mouse model of Rett syndrome, Hum Mol Genet 14, 205–220, 2005. 39. Collins, A.L. et al., Mild overexpression of MeCP2 causes a progressive neurological disorder in mice, Hum Mol Genet 13, 2676–2689, 2004. 40. Meins, M. et al., Submicroscopic duplication in Xq28 causes increased expression of the MECP2 gene in a boy with severe mental retardation and features of Rett syndrome, J Med Genet 42, e12, 2005. 41. Samaco, R.C., Nagarajan, R.P., Braunschweig, D., and LaSalle, J.M., Multiple pathways regulate MeCP2 expression in normal brain development and exhibit defects in autism-spectrum disorders, Hum Mol Genet 13, 629–639, 2004. 41a. Van Esch, H. et al., Duplication of the MECP2 region is a frequent cause of severe mental retardation and progressive neurological symptoms in males. Am J Hum Genet 77, 442–453, 2005. 42. Makedonski, K., Abuhatzira, L., Kaufman, Y., Razin, A., and Shemer, R., MeCP2 deficiency in Rett syndrome causes epigenetic aberrations at the PWS/AS imprinting center that affect UBE3A expression, Hum Mol Genet 14, 1049–1058, 2005. 43. Samaco, R.C., Hogart, A., and LaSalle, J.M., Epigenetic overlap in autism-spectrum neurodevelopmental disorders: MECP2 deficiency causes reduced expression of UBE3A and GABRB3, Hum Mol Genet 14, 483–492, 2005. 44. Cook, E.H., Jr. et al., Autism or atypical autism in maternally but not paternally derived proximal 15q duplication, Am J Hum Genet 60, 928–934, 1997. 45. Crolla, J.A., Harvey, J.F., Sitch, F.L., and Dennis, N.R., Supernumerary marker 15 chromosomes: a clinical, molecular and FISH approach to diagnosis and prognosis, Hum Genet 95, 161–170, 1995. 46. Browne, C.E. et al., Inherited interstitial duplications of proximal 15q: genotypephenotype correlations, Am J Hum Genet 61, 1342–1352, 1997.

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 110 Friday, March 10, 2006 2:17 PM

110

Understanding Autism: From Basic Neuroscience to Treatment

47. Ungaro, P. et al., Molecular characterization of four cases of intrachromosomal triplication of chromosome 15q11-q14, J Med Genet 38, 26–34, 2001. 48. Crolla, J.A., Youings, S.A., Ennis, S., and Jacobs, P.A., Supernumerary marker chromosomes in man: parental origin, mosaicism and maternal age revisited, Eur J Hum Genet 13, 154–160, 2005. 49. Wang, N.J., Liu, D., Parokonny, A.S., and Schanen, N.C., High-resolution molecular characterization of 15q11-q13 rearrangements by array comparative genomic hybridization (array CGH) with detection of gene dosage, Am J Hum Genet 75, 267–281, 2004. 50. Roberts, S.E. et al., Characterization of interstitial duplications and triplications of chromosome 15q11-q13, Hum Genet 110, 227–234, 2002. 51. Boyar, F.Z. et al., A family with a grand-maternally derived interstitial duplication of proximal 15q, Clin Genet 60, 421–430, 2001. 52. Bolton, P.F. et al., The phenotypic manifestations of interstitial duplications of proximal 15q with special reference to the autistic spectrum disorders, Am J Med Genet 105, 675–685, 2001. 53. Vu, T.H. and Hoffman, A.R., Imprinting of the Angelman syndrome gene, UBE3A, is restricted to brain, Nat Genet 17, 12–13, 1997. 54. Rougeulle, C., Glatt, H., and Lalande, M., The Angelman syndrome candidate gene, UBE3A/E6-AP, is imprinted in brain, Nat Genet 17, 14–15, 1997. 55. Albrecht, U. et al., Imprinted expression of the murine Angelman syndrome gene, Ube3a, in hippocampal and Purkinje neurons, Nat Genet 17, 75–78, 1997. 56. Yamasaki, K. et al., Neurons but not glial cells show reciprocal imprinting of sense and antisense transcripts of Ube3a, Hum Mol Genet 12, 837–847, 2003. 57. Buettner, V.L., Longmate, J.A., Barish, M.E., Mann, J.R., and Singer-Sam, J., Analysis of imprinting in mice with uniparental duplication of proximal chromosomes 7 and 15 by use of a custom oligonucleotide microarray, Mamm Genome 15, 199–209, 2004. 58. Meguro, M. et al., Evidence for uniparental, paternal expression of the human GABAA receptor subunit genes, using microcell-mediated chromosome transfer, Hum Mol Genet 6, 2127–2133, 1997. 59. Liljelund, P., Handforth, A., Homanics, G.E., and Olsen, R.W. GABA(A) receptor beta3 subunit gene-deficient heterozygous mice show parent-of-origin and genderrelated differences in beta3 subunit levels, EEG, and behavior, Brain Res Dev Brain Res 157, 150–161, 2005. 60. Menold, M.M. et al., Association analysis of chromosome 15 GABAA receptor subunit genes in autistic disorder, J Neurogenet 15, 245–259, 2001. 61. McCauley, J.L. et al., A linkage disequilibrium map of the 1-Mb 15q12 GABA(A) receptor subunit cluster and association to autism, Am J Med Genet B Neuropsychiatr Genet 131, 51–59, 2004. 62. Curran, S. et al., An association analysis of microsatellite markers across the PraderWilli/Angelman critical region on chromosome 15 (q11-13) and autism spectrum disorder,. Am J Med Genet B Neuropsychiatr Genet 2005. 63. Kim, S.J. et al., Mutation screening and transmission disequilibrium study of ATP10C in autism, Am J Med Genet 114, 137–143, 2002. 64. Cogoni, C. and Macino, G., Post-transcriptional gene silencing across kingdoms, Curr Opin Genet Dev 10, 638–643, 2000. 65. Miles, J.H. and Hillman, R.E., Value of a clinical morphology examination in autism, Am J Med Genet 91, 245–253, 2000. 66. Scott, F.J., Baron-Cohen, S., Bolton, P., and Brayne, C., Brief report: prevalence of autism spectrum conditions in children aged 5–11 years in Cambridgeshire, U.K., Autism 6, 231–237, 2002.

© 2006 by Taylor & Francis Group, LLC

2732_C005.fm Page 111 Friday, March 10, 2006 2:17 PM

A Mixed Epigenetic and Genetic

111

67. Laumonnier, F. et al., X-linked mental retardation and autism are associated with a mutation in the NLGN4 gene, a member of the neuroligin family, Am J Hum Genet 74, 552–557, 2004. 68. Jamain, S. et al., Mutations of the X-linked genes encoding neuroligins NLGN3 and NLGN4 are associated with autism, Nat Genet 34, 27–29, 2003. 69. Vincent, J.B. et al., Mutation screening of X-chromosomal neuroligin genes: no mutations in 196 autism probands, Am J Med Genet B Neuropsychiatr Genet 129, 82–84, 2004. 70. Gauthier, J. et al., NLGN3/NLGN4 gene mutations are not responsible for autism in the Quebec population, Am J Med Genet B Neuropsychiatr Genet 132, 74–75, 2005. 71. Varoqueaux, F., Jamain, S., and Brose, N., Neuroligin 2 is exclusively localized to inhibitory synapses, Eur J Cell Biol 83, 449–456, 2004. 72. Graf, E.R., Zhang, X., Jin, S.X., Linhoff, M.W., and Craig, A.M., Neurexins induce differentiation of GABA and glutamate postsynaptic specializations via neuroligins, Cell 119, 1013–1026, 2004. 73. Chih, B., Engelman, H., and Scheiffele, P., Control of excitatory and inhibitory synapse formation by neuroligins, Science 307, 1324–1328, 2005. 74. Jamain, S. et al., Y chromosome haplogroups in autistic subjects, Mol Psychiatry 7, 217–219, 2002. 75. Donnelly, S.L. et al., Female with autistic disorder and monosomy X (Turner syndrome): parent-of-origin effect of the X chromosome, Am J Med Genet 96, 312–316, 2000. 76. Skuse, D.H. et al., Evidence from Turner’s syndrome of an imprinted x-linked locus affecting cognitive function, Nature 387, 705–708, 1997. 77. Beaudet, A.L., Is medical genetics neglecting epigenetics? Genet Med 4, 399–402, 2002.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 113 Friday, March 10, 2006 2:19 PM

6

Neurobiology of Related Disorders: Fragile X Syndrome Usha Narayanan and Stephen T. Warren

CONTENTS Introduction............................................................................................................113 The Gene Causing Fragile X: FMR1 ....................................................................114 Triplet Repeat Expansion in Fragile X .................................................................115 FMRP: The Protein Product of the Disease Gene in Fragile X Syndrome .........116 FMRP and the Neuronal Phenotype in Fragile X Syndrome...............................117 Fragile X Therapeutics ..........................................................................................120 Fragile X and Autism ............................................................................................121 Etiology of Autism ................................................................................................121 Finding the Disease-Causing Gene: Insight from Fragile X................................122 References..............................................................................................................123

INTRODUCTION Cognition is a complex set of active intellectual processes through which information is obtained, transformed, retrieved, and used by the brain. The essence of human cognition is memory formation. Mechanistic insight into this process can often be gleaned by the study of genetic aberrations in individuals that result in a disruption in the capacity for memory formation and pronounced cognitive deficits. Early studies by Lehrke indicated that genes influencing human cognition lie on the X chromosome and estimated that one fourth of all mental retardation can be traced to X-linked factors [1]. Although many X-linked loci can result in mental retardation, a substantial fraction is due to fragile X syndrome. Fragile X is the most commonly inherited form of mild to moderate mental retardation affecting 1/4000 males and 1/8000 females. It was initially identified by Lubs in 1969 and reported as a novel X-chromosome anomaly cosegregating with mental retardation over three generations [2]. Cytological studies ascertained a constriction or fragile site in the distal long arm of the X chromosome in metaphase preparations from four mentally retarded males and one normal female. Subsequently, Harrison et al. localized this break to X27q.3 and called it the fragile X chromosome [3]. 113

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 114 Friday, March 10, 2006 2:19 PM

114

Understanding Autism: From Basic Neuroscience to Treatment

The identification of a cytogenetic diagnostic marker expedited a more complete evaluation of what is now appreciated as fragile X syndrome. The clinical presentation of a male fragile X patient includes moderate-to-severe mental retardation, with most patients exhibiting IQs of ~50 [4,5]. Other classical symptoms include macroorchidism and subtle connective tissue signs along with the characteristic appearance of a long, narrow face and large ears [6–9]. Certain features that presumably form a byproduct of the connective tissue disorder are: velvet-like skin, finger joint hyperextensibility, recurrent otitis media, aortic root dilatation, and mitral valve prolapse [9–11]. Importantly, patients with fragile X very often display autistic features including shyness, poor eye contact and social anxiety, as well as hyperactivity, hand flapping, hand biting, and perseverative speech [11]. Females may also be affected but usually display a milder phenotype than do male patients. This is due to the X-linked nature of fragile X syndrome and is substantiated with the finding that phenotypic severity of fragile X in females is directly correlated with the degree of X-inactivation on the fragile X chromosome [12,13]. Fragile X syndrome was appreciated as having a complex inheritance pattern in the early 1980s. The X-linked recessive model proposed initially failed to explain the existence of affected females who were heterozygous and pedigrees in which unaffected males transmitted the fragile X chromosome to their daughters [14–16]. These observations were reiterated by Sherman, who performed an exhaustive segregation analysis on 206 fragile X syndrome pedigrees, establishing that disease penetrance increases with succeeding generations of the pedigree, a phenomenon that became known as the Sherman paradox [17,18]. Thus, fragile X syndrome displays a form of anticipation with increasing penetrance in succeeding generations [19–22]. This atypical inheritance pattern was understood better with the discovery of the disease-causing gene in 1991.

THE GENE CAUSING FRAGILE X: FMR1 Identifying the gene responsible for fragile X was the result of extensive physical and genetic mapping. Pedigree analyses using the fragile site, which cosegregated with the syndrome, revealed the causal locus on a 22 cM region of the Xq chromosome [23]. Further mapping reduced the interval to 1 to 2 Mb, supporting the localization of the disease locus to the fragile site [24–29]. Using these probes and methylation-sensitive restriction fragment digests, a large methylated region in fragile X patients was identified that typically was unmethylated in normal males or unaffected carriers [30,31]. Warren et al. used somatic cell hybrids to precisely mark the fragile X site by selecting for events of X chromosome breakage and translocation generated by inducing fragility [32,33]. These hybrids greatly accelerated the pace of molecular discoveries regarding fragile X syndrome, being critical in (1) showing that the abnormally methylated fragments from affected individuals were unstable and increased in size with successive generations of a pedigree [34,35] and (2) identifying a yeast artificial chromosome (YAC) containing markers near the fragile site that revealed an aberrantly methylated CpG island in fragile X patients [36]. Using cosmids probes derived from a YAC that crossed the presumptive fragile site in the somatic cell hybrids, the gene responsible for fragile X syndrome, fragile X mental retardation 1 (FMR1), was identified [37].

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 115 Friday, March 10, 2006 2:19 PM

Neurobiology of Related Disorders: Fragile X Syndrome

115

The lack of FMR1 expression in patients confirmed FMR1 as the causative gene in fragile X syndrome [38–41]. FMR1 is located at Xq27.3 precisely at a unique fragile site on the X chromosome of patients [42].

TRIPLET REPEAT EXPANSION IN FRAGILE X An abnormal expansion of repeated trinucleotide sequences within FMR1 causes fragile X and resembles other inherited neurodegenerative disorders (i.e., myotonic dystrophy) that, subsequent to the FMR1 discovery, were also found to be due to unstable repeats within the responsible genes. These repeat expansions were termed dynamic mutations because they can change in each generation, accounting for certain unusual inheritance patterns such as anticipation. In fragile X syndrome, the deleterious CGG-repeat responsible for the loss of gene expression was found at the 5UTR of FMR1 [43–45]. The repeat was shown to be polymorphic with a range of 6 to 60 and a mode of 30 [46,47]. Moreover, it was ascertained that normal protein is produced, and no substantial dysfunction is noticed if the repeat number was between ~50 and 200. However, if the expansion exceeded 200 CGG repeats, the disease became clinically discernible. Fragile X pedigree analyses explained this observation with the discovery of two classes of alleles: nonpenetrant premutations of 60 to 200 repeats and completely penetrant full mutations with >200 repeats, often near 1000 repeats. Premutation alleles were found to be unstable with a tendency to expand through successive generations. The expansion to a full mutation occurred when the disease allele was transmitted by a female, as male spermatogenesis is unable to maintain the unstable repeat [48]. Thus, a large-scale CGG expansion was found to be the causative mutation in more than 95% of patients with fragile X syndrome. Characterizing the CGG repeats in the 5UTR and the upstream CpG island of the FMR1 promoter revealed that they were hypermethylated, following the expansion ultimately leading to transcriptional silencing of FMR1 [49,50]. Moreover, treating full-mutation cell lines with methylation inhibitors reactivated a low level of FMR1 expression, indicating that methylation causes gene silencing [51]. To dissect the precise mechanism of silencing, footprinting analyses were performed in vivo, revealing binding sites for transcription factors in the FMR1 promoter region protected in normal cells but not in fragile X patient cells [52,53]. These findings were corroborated by the discovery of transcription factors such as USF1, USF2, and α-Pal/Nrf-1, binding the FMR1 promoter and required for promoter-driven gene expression. Alongside, it was found that methylating reporter constructs failed to completely abolish gene expression. Thus, both histone-dependent and histoneindependent mechanisms could be involved to produce the methylation-dependent gene silencing seen in FMR1. Given the link between repeat expansion, methylation, and disease manifestation, identifying the timing of expansion is pivotal to disease diagnosis and treatment. Initial studies suggested that expansion occurred postzygotically, following germ line differentiation [54]. However, the degree of mosaicism observed was not inversely proportional to the length of the maternal permutation as would be predicted by the postzygotic model, suggesting another disease mechanism. Malter et al.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 116 Friday, March 10, 2006 2:19 PM

116

Understanding Autism: From Basic Neuroscience to Treatment

[41] found only full-mutation alleles in the ovaries of a 16-week-old fetus, suggesting that the timing of expansion must precede germ line differentiation. In addition, testis tissue from a 13-week-old full-mutation fetus showed only full mutations in the germ cells while that from an older fetus showed evidence of both premutation and full-mutation alleles in the germ cells [41]. In effect, the timing of CGG repeat expansion, ultimately leading to fragile X, is thought to be prezygotic although early embryogenesis cannot be ruled out [55,56].

FMRP: THE PROTEIN PRODUCT OF THE DISEASE GENE IN FRAGILE X SYNDROME Fragile X syndrome etiology has revealed that the lack of functional FMRP is both necessary and sufficient for the development of the disease. Preliminary evidence for a specific role for FMRP in cognition came from examining brains of three fragile X patients that showed increased dendritic spine density in the CNS [57]. Thus, elucidation of the physiological function of FMRP in vivo has been a major goal of fragile X research over the recent years. FMRP has about 20 speculated splice variants as the FMR1 mRNA undergoes extensive alternative splicing. However, only four of the five protein isoforms are readily detectable. Almost all FMRP consists of FMR1 isoform 7, which lacks exon 12. FMRP shares functional domains with proteins known to form large ribonucleoprotein complexes in vivo. Indeed, it contains three RNA-binding motifs: two KH domains that show homology to hnRNP K and an RGG (Arg-Gly-Gly) box similar to hnRNP U [58,59]. These RNA-binding domains were found to be functional as both recombinant- and purified FMRP bound RNA homopolymers and 4% of fetal brain messages in vitro [60]. Recent studies have elucidated the nature of the RNA motif bound by FMRP and its physiological mRNA ligands. The former was found to be an intramolecular G-quartet structure bound by the RGG box of FMRP [61,62]. This RNA structure was further validated in FMRP binding as it was found in ~70% of the mRNA transcripts immunoprecipitated with FMRP [54]. Apart from biochemical analyses, FMRP was also the subject of cellular localization studies, which established it as a primarily cytoplasmic protein in neurons [63–66]. An analysis of FMRP truncations has led to the delineation of both a nuclear localization signal (NLS) and a nuclear export signal (NES), suggesting that FMRP shuttles between the nucleus and the cytoplasm. FMRP has been thought to have a cytoplasmic function as it associates with translating polyribosomes in a large mRNP complex in the cytoplasm [67,68]. This mRNP is greater than 660 kD in size and contains a number of proteins, including FMRP autosomal homologs, FXR1 and FXR2 [69,70]. The functional importance of this polyribosome association became clear with the discovery of a severely affected fragile X patient with a missense mutation in the KH2 RNA-binding domain of FMRP [71]. This isoleucine to asparagine change, I304N, is known to compromise the RNA-binding ability of the KH2 domain although the absolute RNA-binding capacity of FMRP remains unaffected [72,73]. However, the mutant fails to associate with translating polyribosomes but rather is part of a smaller abnormal mRNP. Thus, the mutant FMRP protein causes improper mRNP formation and polyribosome association, leading to

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 117 Friday, March 10, 2006 2:19 PM

Neurobiology of Related Disorders: Fragile X Syndrome

117

a severe fragile X phenotype in the patient, establishing the crucial role of cytoplasmic FMRP. Another role for cytoplasmic FMRP in translation was revealed by determining the identities of the mRNA ligands precipitated by FMRP in the mouse brain. A substantial number of mRNAs exhibited a shift in translation status as seen from the polyribosomal profiles of full-mutation lymphoblastoid cells compared to normal cells. Previous in vitro studies had ascertained that purified FMRP suppresses translation, suggesting that cells lacking FMRP would show increased polysome loading of select messages [74,75]. However, Brown et al. [53] observed both increased and decreased polysome loading of messages in fragile X syndrome cells. The decreased loading of certain messages could be explained by it being now outcompeted for ribosome binding by the messages normally bound by FMRP and now increased in polysomes. Posttranslational modifications, especially phosphorylation, have been known to modify RNA-binding and protein–protein interactions in proteins structurally similar to FMRP [76–80]. This became important with the findings that both the drosophila and mouse orthologs of FMRP, dFxr and Fmr1, respectively, are highly phosphorylated on a conserved serine residue [81,82]. The phosphorylated serine was observed both in the cultured mouse cells and, more importantly, in the mouse brain. In drosophila, in vitro phosphorylation of FMRP was known to increase its binding affinity for RNA homopolymers but no such effect was detected in murine FMRP. Rather, phosphorylated FMRP in the mouse was found to associate with stalled ribosomes, suggesting that dephosphorylation may initiate the release of the FMRP mRNA ligands from translational suppression. Given these data, it seems likely that phosphorylation may play a regulatory role in FMRP suppression of translation. As proposed in Figure 6.1, the FMRP-specific kinases and phosphatases may work together to maintain a balance between the phosphorylated and nonphosphorylated forms of FMRP. The downstream effects of this balance could include ribosome stalling and translational suppression, supported by the association between phosphorylated FMRP and stalled ribosomes. These studies supported a cytoplasmic role of FMRP in gene expression. However, the most compelling evidence for FMRP’s function in translational suppression came from its association with microRNAs [83,84]. Immunoprecipitation analyses with human cell lines revealed that FMRP interacted with microRNA pathway components, i.e., Dicer and the eIF2C2 (the mammalian ortholog of Argonaute1, Ago1). Furthermore, studies in drosophila indicated that Ago1 is vital for FMRP’s role in neural development [85]. Thus, the emerging picture reveals FMRP protein function in translational suppression and synaptogenesis to be central to the etiology of fragile X syndrome.

FMRP AND THE NEURONAL PHENOTYPE IN FRAGILE X SYNDROME Apart from the gross clinical phenotype, fragile X has also been reported to have a neuronal phenotype since the first autopsies performed by Rudelli et al. [86,87]. These studies showed normal neuron counts but abnormal dendritic spines — the postsynaptic protrusions from dendrites where the majority of excitatory synapses

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 118 Friday, March 10, 2006 2:19 PM

118

Understanding Autism: From Basic Neuroscience to Treatment

AAAA

FMRP

Kinase

Phosphatase

FMRP

P

AAAA

FIGURE 6.1 (A color version of this figure follows page 236.) Model for a regulatory role of phosphorylation in FMRP suppression of translation. Phosphorylated FMRP is known to preferentially associate with stalled ribosomes as compared to nonphosphorylated FMRP. Thus, it is possible that the FMRP-specific kinase and phosphatase maintain a balance between phosphorylated and nonphosphorylated FMRP, ultimately regulating translational suppression of the FMRP mRNA ligands.

occur [78]. Dendritic spines are known to change density and morphology in response to a variety of environmental stimuli [88]. Spine abnormalities have been known to be associated with mental retardation; however, it is unknown if they are a cause or an effect of the disease [89]. In fragile X, the neuronal phenotype includes an increase in dendritic spine density in the CNS. The dendritic spines have a long, thin morphology, closely resembling the immature spines in developing neurons. This suggested a role for FMRP in neuronal maturation or development, but a putative mechanism of FMRP action emerged upon investigation of the mouse model for fragile X, the Fmr1 knockout mouse. Although the animal model differs from the human form of gene disruption, absence of FMRP results in a phenotype resembling that in fragile X syndrome in humans. This includes macroorchidism and deficits in spatial learning and associative fear-conditioned learning [90]. Interestingly, there is also an overall increase in the number of dendritic spines in the Fmr1 knockout mice, and the spine © 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 119 Friday, March 10, 2006 2:19 PM

Neurobiology of Related Disorders: Fragile X Syndrome

mGluR 1/5

119

mGluR 1/5

+

+ AAAA

AAAA

− FMRP

Fragile X syndrome FMRP

FIGURE 6.2 (A color version of this figure follows page 236.) The mGluR theory at fragile X mental retardation. The activation of the group I metabotropic receptors (mGluR) is known to result in increased translation of messages (including FMRP ligands) in the dendrite. In the presence of FMRP, translation of these mRNA ligands is suppressed. However, in the case of fragile X, the absence of FMRP leads to increased ligand translation. Ultimately, this disrupts the translational regulation in the dendrites, contributing to the synaptic deficits in fragile X.

morphology mimics the human condition in being long and thin. This overall increase in spine density seen in the Fmr1 knockout mice and the fragile X patients suggests that the lack of mature synapses may be the basis of the cognitive deficits. The morphology of spines may be due to a developmental derangement attributable to changes in protein expression controlled by FMRP. Evidence for such a role of FMRP in synaptic plasticity comes from studies comparing the wild-type and knockout mice using two synaptic models for information storage: long-term depression (LTD) and long-term potentiation (LTP). LTD is a decrease in the strength of the same synapses after prolonged, low-frequency stimulation [91]. One form of LTD is metabotropic glutamate receptors (mGLuR) dependent and requires protein synthesis [92]. This form of LTD is enhanced in Fmr1 knockout mice, indicating that the absence of FMRP alters synaptic plasticity. This becomes more significant as FMRP synthesis is induced in synaptoneuronsomes by the activation of Group1 mGluRs (mGluR1 and 5) [92a]. Taking into account the role of FMRP in translational repression along with the presence of FMRP mRNA in dendrites, Bear et al. proposed a model (Figure 6.2) theorizing its role in mGluR LTD [93]. According to this model, mGluR activation typically induces the synthesis of proteins involved in stabilizing LTD and, in addition, FMRP. It is conceivable that the FMRP synthesized inhibits further synthesis of mRNA ligands (by virtue of the role of FMRP as a translational repressor) resulting in the suppression of LTD. As there are mRNA ligands expressed in positive correlation with FMRP, it has been proposed that both the pools of mRNA ligands (negatively and positively regulated by FMRP) compete for a limited number of © 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 120 Friday, March 10, 2006 2:19 PM

120

Understanding Autism: From Basic Neuroscience to Treatment

polysomes, providing an internal check on protein expression. In fragile X, the absence of FMRP would result in an imbalance in synthesis of the mRNA ligands leading to enhanced LTD. This model is in agreement with the current knowledge of FMRP and fragile X biology. Apart from LTD, the other mode of information storage, LTP, could also be affected in the absence of FMRP. LTP is a long-term increase in synaptic strength in response to high-frequency stimulation and is thought to occur in three phases: the first requiring neither transcription nor protein synthesis, a second intermediate mGluR-dependent phase, and a final phase requiring both transcription and protein synthesis [94,95]. Although the first and the last phases are unaffected in Fmr1 knockout mice, it is unknown if the intermediate phase is affected [96,97]. The neuronal phenotype seen in fragile X could thus be explained as changes in spine morphology in response to local changes in expression of the mRNA ligands controlled by FMRP. This theory is supported by two pieces of evidence: (1) The synaptic immaturity seen in the Fmr1 knockout mice and human patients resembles that which can be caused by sensory deprivation, and (2) FMRP translocates to the areas surrounding synapses following stimulation of mGluR.

FRAGILE X THERAPEUTICS Treating fragile X syndrome successfully demands a more complete understanding of the etiology. Despite progress made in understanding the molecular basis of the disease, the therapeutic options remain palliative, a combination of behavioral and cognitive therapy with symptom-specific therapy for medical problems [98]. Given the role of FMRP in translation suppression, mRNP trafficking, and synaptic plasticity, it appears that the therapeutic approach needs to target events downstream of FMRP function. Previous studies have rescued the full-mutation FMR1 from transcriptional silencing by applying histone deactelyase and methylase inhibitors, but the reactivation was marginal and accompanied by considerable cytotoxicity. In addition, the transcripts with large repeats are not translated well [99–102]. These data suggest that reactivating transcription of FMR1 alone may not prove sufficient for restoring wild-type FMRP levels. FMRP replacement therapy takes a step closer to treating the problem at the protein level but needs to overcome the undesirable overexpression phenotypes similar to that seen with expressing human FMRP in the mouse [103]. In this regard, characterizing the in vivo FMRP ligands by Brown et al. may prove to be a viable launching pad for treatment options targeted downstream of FMRP. Changes in the translational expression profiles of FMRP mRNA ligands, along with changes in the activation of mGluR expression, provide attractive therapeutic options. The action of Group 1 mGluR antagonists is one such case in point. The mGluR5 antagonists are being pursued more intensively because mGluR1 blockers cause ataxia by disrupting cerebellar function. The prototypical mGluR5 antagonist is 2-methyl-6- (phenylethynyl)-pyridine MPEP. In animal models, systemically administered MPEP has broad and potent anticonvulsant and antioxylytic effects without any overt effects on locomotor activity. MPEP can reverse inflammation-induced mechanical hyperalgesia by inhibiting mGluR5 receptors in the C-fibers of the skin. And by inhibiting mGluR receptors in the gut, MPEP can reduce bowel motility.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 121 Friday, March 10, 2006 2:19 PM

Neurobiology of Related Disorders: Fragile X Syndrome

121

Antagonists of mGluR5 might correct the mild cognitive defects seen in the Fmr1 knockout mice by bringing synaptic plasticity back into its proper range. The mGluR5 antagonists have great promise as a potential treatment for the neurological and psychiatric symptoms of fragile X expressed in adults. However, if the syndrome is a lasting consequence of brain development with exaggerated Group 1 mGluR signaling, it is possible that early intervention with receptor antagonists could prevent some symptoms from occurring altogether. Gene therapy is an exciting avenue that has been researched for treating fragile X syndrome, although it currently lacks a reliable assay for studying the effect of introducing a normal FMR1 gene to restore FMRP expression. Also, the percentage of mouse brain neurons transduced by this approach needs to be recognized to ascertain the numbers that significantly alter spatial learning deficits. Finally, any effects observed at the cellular level would need to be translated into correction of the neurobehavioral defect in adult animals. Given the association between mGluR and FMRP, successful gene therapy could start with assessing if the same glutamate receptor expression defects are seen in human fragile X as those seen in the knockout mouse [104,105]. This may offer a more immediate neuropharmacological intervention for fragile X syndrome. The studies on the etiology and therapeutics of fragile X may provide valuable insight for other neurobehavioral disorders with cognitive deficits, e.g., autism.

FRAGILE X AND AUTISM Autism is a syndrome with multiple etiologies, which is made obvious by the range of disorders that present autistic behaviors. Fragile X mental retardation syndrome has been well documented to show autistic behavior as one of its classical symptoms. Clinically, autism is defined as a pervasive developmental disorder, which can be defined behaviorally as a syndrome consisting of abnormal development of social skills (withdrawal, lack of interest in peers), limitations in the use of interactive language (speech as well as nonverbal communication), and sensorimotor deficits (inconsistent responses to environmental stimuli) with considerable variation in severity [106,107]. Other impairments include that of language, often exhibiting echolalia (the involuntary repetition of a word or sentence just spoken by another person), deficiencies in symbolic thinking, stereotypic behaviors (e.g., repetitive nonproductive movements of hands and fingers, rocking, meaningless vocalization), self-stimulation, self-injury behavior, and seizures. Mental retardation, although not a diagnostic feature, is present in moderate to severe range. Autism is a common disorder with a median rate of incidence that is 7/ 10,000 [108]. The occurrence in siblings is thought to be from 3 to 7% representing a 50- to 100-fold increase in risk [109].

ETIOLOGY OF AUTISM No single cause has been identified for the development of autism as yet. A genetic origin for autism is suggested by twin studies and the higher incidence rates seen among siblings [110]. Contributing factors include infections, errors in metabolism, lead poisoning, and fetal alcohol syndrome [111].

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 122 Friday, March 10, 2006 2:19 PM

122

Understanding Autism: From Basic Neuroscience to Treatment

An association between fragile X and autism has long been established. There is known to be an increased frequency of occurrence in patients with genetic conditions such as fragile X syndrome [112]. Approximately 30% of the individuals with fragile X are on the autism spectrum [113–115]. There is a disagreement, however, over the degree of fragile X prevalence in patients with autism, with some studies demonstrating little or no association between fragile X and autism, and others finding a high association [116,117]. Using molecular markers, recent epidemiological studies have documented incidence rates between 7 to 8% of fragile X in populations with autism [118–122].

FINDING THE DISEASE-CAUSING GENE: INSIGHT FROM FRAGILE X To dissect the etiology of autism, the association between specific genes and the disease is being studied. Toward this end, several approaches have been used including cytogenetic analyses of cosegregating markers, whole genome screens, and clinical studies of pathogenetic developmental models. Autism candidate gene search may also benefit from research of other neurogenetic disorders such as fragile X syndrome. Given the overlap in the phenotypic characteristics of autism and fragile X, it is conceivable that the molecular bases show similarities. A molecular dysfunction in fragile X syndrome is caused by the lack of FMRP protein, leading to misregulated expression of certain mRNA ligands. Analyses of the FMRP mRNPs in the mammalian brain demonstrated that the FMRP mRNA ligands contain an intramolecular G quartet structure, a planar conformation of 4 guanine residues [95]. Probing microarrays with mRNAs co-immunoprecipitated with murine FMRP RNPs revealed 432 mRNAs enriched in the FMRP mRNPs. Importantly, probing microarrays with mRNAs isolated from the polysome fraction of a human fragile X cell line yielded 251 human mRNAs with an altered polyribosome profile in the absence of FMRP. The murine orthologs of 50% of the human mRNAs assessed were found to be associated with wild-type FMRP RNPs [95]. Further, Darnell et al. demonstrated that 70% of these mRNAs contained a G quartet directly bound by FMRP. Unigene database searches for G-quartetcontaining sequences identified several potential FMRP target mRNAs in the genome. After assessing these sequences for their ability to form G quartets, 31 RNAs were selected. Twelve were tested for FMRP binding, six were found to bind FMRP with affinities of 75 to 467 nM kd,and the others failed to interact with FMRP. The mRNAs from the Brown et al. [95] study showing altered polysome distribution for G quartets were also tested for FMRP binding [55]. These data yielded a list of eight mRNA targets — RNAs with G quartets directly binding FMRP and showing aberrant polysome profiles in fragile X cells. Of these mRNA ligands, some may serve as good candidates in autism; MAP1B is a case in point. MAP1B is a member of the cytoskeletal proteins called the microtubule associated proteins and is known to be the first of the neural MAPs to be expressed in situ [123,124]. It gets downregulated during normal brain development [125,126] and is known to play an important role in neurite and synapse development [127]. Fragile X

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 123 Friday, March 10, 2006 2:19 PM

Neurobiology of Related Disorders: Fragile X Syndrome

123

syndrome shows delayed dendritic spine maturation and a MAP1B expression deregulation. A drosophila model for fragile X revealed that futsch, the MAP1B fly ortholog, is regulated by dFXR. Intriguingly, deficiency of drosophila Fmr1 results in increased expression of futsch and synaptic overgrowth at the neuromuscular junction [128]. These data were confirmed in the mammalian system by co-immunoprecipitation and microarray analyses [55,95]. Crucially, MAP1B expression was found to increase in fragile X patient cell lines as compared to the wild-type cell cultures. In vivo evidence in mice confirmed that developmentally programmed FMRP expression represses MAP1B translation and facilitates the decline of MAP1B during active synaptogenesis in neonatal brain development. In the absence of FMRP, MAP1B is misregulated, showing a delayed MAP1B decline in the Fmr1 KO brain. This inappropriate MAP1B expression results in increased microtubule stability in Fmr1 KO neurons. Thus, cytoskeleton organization during neuronal development and the abnormal microtubule dynamics is a conceivable underlying factor for the pathogenesis of fragile X mental retardation. Intriguingly, MAP1B has recently been linked to another neurogenetic disorder with a mental retardation phenotype, i.e., autism [124]. MAP1B has been found to associate with Reelin (RELN), a neuronal signaling protein that plays a pivotal role in the migration of several neuronal cell types and a genetic risk factor for autism [117]. RELN has been mapped to 7q22 and is known to lie in the autism susceptibility locus within the chromosome long arm, 7q [129]. Preliminary data indicate a role for MAP1B in Reelin-dependent neuronal migration. Interestingly, MAP1B-deficient mice display an abnormal structuring of the nervous system, especially in the brainlaminated areas indicating a failure in neuronal migration [124]. In summary, the FMRP mRNA ligand MAP1B seems to emerge as an interesting candidate gene for autism. Other FMRP mRNA ligands could hold potential clues to the molecular pathogenesis in autism.

REFERENCES 1. Lehrke, R.G. and Bergsma, D., Minnesota Association for Retarded Children, and National Foundation, X-linked mental retardation and verbal disability Symposia Specialists; distributed by International Medical Book Corp., Miami, FL, 1974. 2. Lubs, H.A., A marker X-chromosome, Am J Hum Genet 21, 231, 1969. 3. Harrison, C.J., Jack, E.M., Allen, T.D., and Harris, R., The fragile X: a scanning electron microscope study, J Med Genet 20(4), 280, 1983. 4. Bennetto, L. and Pennington, B.F., The neuropsychology of fragile X syndrome, in Fragile X Syndrome: Diagnosis, Treatment and Research, Hagerman, R.J. and Silverman, A.C., Eds., The John Hopkins University Press, Baltimore, MD, 1996, p. 210. 5. Weeber, E.J., Levenson, J.M., and Sweatt, J.D., Molecular genetics of human cognition, Mol Interventions 2(6), 376, 2002. 6. Turner, G., Daniel, A., and Frost, M., X-linked mental retardation, macro-orchidism, and the Xq27 fragile site, J Pediatr 96, 837, 1980. 7. Opitz, J.M., Westphal, J.M., and Daniel, A., Discovery of a connective tissue dysplasia in the Martin-Bell syndrome, Am J Med Genet 17(1), 101, 1984. 8. Hagerman, R.J., Van Housen, K., Smith, A.C., and McGavran, L., Consideration of connective tissue dysfunction in the fragile X syndrome, Am J Med Genet 17(1), 111, 1984.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 124 Saturday, March 18, 2006 11:02 AM

124

Understanding Autism: From Basic Neuroscience to Treatment

9. Loehr, J.P., Synhorst, D.P., Wolfe, R.R., and Hagerman, R.J., Aortic root dilatation and mitral valve prolapse in the fragile X syndrome, Am J Med Genet 23(1–2), 189, 1986. 10. Hagerman, R., Physical and behavioural phenotype, in Fragile-X Syndrome: Diagnosis, Treatment and Research, Hagerman, R.J. and Cronister, A., Eds., The Johns Hopkins University Press, Baltimore, MD, 1996, p. 3. 11. Merenstein, S.A., Sobesky, W.E., Taylor, A.K., Riddle, J.E., Tran, H.X., and Hagerman, R.J., Molecular-clinical correlations in males with an expanded FMR1 mutation, Am J Med Genet 64, 388, 1996. 12. Abrams, M.T., Reiss, A.L., Freund, L.S., Baumgardner, T.L., Chase, G.A., and Denckla, M.B., Molecular-neurobehavioral associations in females with the fragile X full mutation, Am J Med Genet 51, 317, 1994. 13. Sobesky, W.E., Taylor, A.K., Pennington, B.F., Bennetto, L., Porter, D., Riddle, J., and Hagerman, R.J., Molecular/clinical correlations in females with fragile X, Am J Med Genet 64(2), 340, 1996. 14. Martin, J.P. and Bell, J., A pedigree of mental defect showing sex-linkage, J Neurol Psychiatry 6, 154, 1943. 15. Losowsky, M.S., Hereditary mental defect showing the pattern of sex influence, J Ment Defic Res 5, 60, 1961. 16. Loesch, D.Z., Hay, D.A., and Mulley, J., Transmitting males and carrier females in fragile X: revisited, Am J Med Genet 51(4), 392, 1994. 17. Sherman, S.L., Maddalena, A., Howard-Peebles, P.N., Brown, W.T., Nolin, S., Jenkins, E., Schwartz, C., Tarrelton, J., Shapiro, L.R., and Smits, A.P., Characteristics of the transmission of the FMR1 gene from carrier females in a prospective sample of conceptuses, Am J Med Genet 51(4), 503, 1994. 18. Sherman, S.L., Jacobs, P.A., Morton, N.E., Froster~Iskenius, U., Howard-Peebles, P.N., Nielsen, K.B., Partington, M.W., Sutherland, G.R., Turner, G., and Watson, M., Further segregation analysis of the fragile X syndrome with special reference to transmitting males, Hum Genet 69(4), 289, 1985. 19. Szot, M., The fragile X syndrome, Folia Histochem Cytobiol 35(3), 135, 1997. 20. Carpenter, N.J., Genetic anticipation: expanding tandem repeats, Neurol Clin 12(4), 683, 1994. 21. Richards, R.I., Holman, K., Friend, K., Kremer, E., Hillen, D., Staples, A., Brown, W.T., Goonewardena, P., Tarleton, J., Schwartz, C., and Sutherland, G.R., Evidence of founder chromosomes in fragile X syndrome, Nat Genet 1(4), 257, 1992. 22. Sutherland, G.R. and Richards, R.I., Unusual inheritance patterns due to dynamic mutation in fragile X syndrome, Ciba Found Symp 197, 119, 1996. 23. Oberle, I., Heilig, R., Moisan, J.P., Kloepfer, C., Mattei, G.M., Mattei, J.F., Boue, J., Froster-Iskenius, U., Jacobs, P.A., and Lathrop, G.M., Genetic analysis of the fragileX mental retardation syndrome with two flanking polymorphic DNA markers, Proc Natl Acad Sci U S A 83(4), 1016, 1986. 24. Suthers, G.K., Hyland, V.J., Callen, D.F., Oberle, I., Rocchi, M., Thomas, N.S., Morris, C.P., Schwartz, C.E., Schmidt, M., and Ropers, H.H., Physical mapping of new DNA probes near the fragile X mutation (FRAXA) by using a panel of cell lines, Am J Hum Genet 47(2), 187, 1990. 25. Suthers, G.K., Hyland, V.J., Callen, D.F., Baker, E., Eyre, H., Roberts, S.H., Hors-Cayla, M.C., Davies, K.E., Bell, M.V., and Sutherland, G.R., Physical and genetic mapping of new DNA probes at Xq26-qter, Cytogenet Cell Genet 51, 1087, 1989.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 125 Friday, March 10, 2006 2:19 PM

Neurobiology of Related Disorders: Fragile X Syndrome

125

26. Hirst, M.C., Rack, K., Nakahori, Y., Roche, A., Bell, M.V., Flynn, G., Christadoulou, Z., MacKinnon, R.N., Francis, M., and Littler, A.J., A YAC contig across the fragile X site defines the region of fragility, Nucleic Acids Res 19(12), 3283, 1991. 27. Hirst, M.C., Roche, A., Flint, T.J., MacKinnon, R.N., Bassett, J.H., Nakahori, Y., Watson, J.E., Bell, M.V., Patterson, M.N., and Boyd, Y., Linear order of new and established DNA markers around the fragile site at Xq27.3, Genomics 10(1), 243, 1991. 28. Rousseau, F., Heitz, D., Biancalana, V., Blumenfeld, S., Kretz, C., Boue, J., Tommerup, N., Van Der Hagen, C., DeLozier-Blanchet, C., Croquette, M.F., Gilgenkranz, S., Jalbert, P., Voelckel, M.A., Oberlé, I., and Mandel, J.L., Direct diagnosis by DNA analysis of the fragile X syndrome of mental retardation, N Engl J Med 325(24), 1673, 1991. 29. Rousseau, F., Vincent, A., Rivella, S., Heitz, D., Triboli, C., Maestrini, E., Warren, S.T., Suthers, G.K., Goodfellow, P., Mandel, J.L., Toniolo, D., and Oberlé, I., Four chromosomal breakpoints and four new probes mark out a 10-cM region encompassing the fragile-X locus (FRAXA), Am J Hum Genet 48(1), 108, 1991. 30. Vincent, A., Heitz, D., Petit, C., Kretz, C., Oberle, I., and Mandel, J.L., Abnormal pattern detected in fragile-X patients by pulsed-field gel electrophoresis, Nature 349 (6310), 624, 1991. 31. Bell, M.V., Hirst, M.C., Nakahori, Y., MacKinnon, R.N., Roche, A., Flint, T.J., Jacobs, P.A., Tommerup, N., Tranebjaerg, L., and Froster-Iskenius, U., Physical mapping across the fragile X: hypermethylation and clinical expression of the fragile X syndrome, Cell 64(4), 861, 1991. 32. Warren, S.T., Knight, S.J., Peters, J.F., Stayton, C.L., Consalez, G.G., and Zhang, F.P., Isolation of the human chromosomal band Xq28 within somatic cell hybrids by fragile X site breakage, Proc Natl Acad Sci U S A 87(10), 3856, 1990. 33. Yu, S., Pritchard, M., Kremer, E., Lynch, M., Nancarrow, J., Baker, E., Holman, K., Mulley, J.C., Warren, S.T., and Schlessinger, D., Fragile X genotype characterized by an unstable region of DNA, Science 252(5010), 1179, 1991. 34. Oberle, I., Rousseau, F., Heitz, D., Kretz, C., Devys, D., Hanauer, A., Boue, J., Bertheas, M.F., and Mandel, J.L., Instability of a 550-base pair DNA segment and abnormal methylation in fragile X syndrome, Science 252(5010), 1097, 1991. 35. Yu, S., Pritchard, M., Kremer, E., Lynch, M., Nancarrow, J., Baker, E., Holman, K., Mulley, J.C., Warren, S.T., Schlessinger, D., Sutherland, G.R., and Richards, R.I., Fragile X genotype characterized by an unstable region of DNA, Science 252(5010), 1179, 1991. 36. Heitz, D., Rousseau, F., Devys, D., Saccone, S., Abderrahim, H., Le Paslier, D., Cohen, D., Vincent, A., Toniolo, D., and Della Valle, G., Isolation of sequences that span the fragile X and identification of a fragile X-related CpG island.[erratum appears in Science 252(5005), 494, April 26, 1991], Science 251(4998), 1236, 1991. 37. Verkerk, A.J., Pieretti, M., Sutcliffe, J.S., Fu, Y.H., Kuhl, D.P., Pizzuti, A., Reiner, O., Richards, S., Victoria, M.F., and Zhang, F.P., Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in fragile X syndrome, Cell 65(5), 905, 1991. 38. Pieretti, M., Zhang, F.P., Fu, Y.H., Warren, S.T., Oostra, B.A., Caskey, C.T., and Nelson, D.L., Absence of expression of the FMR-1 gene in fragile X syndrome, Cell 66(4), 817, 1991. 39. Gedeon, A.K., Baker, E., Robinson, H., Partington, M.W., Gross, B., Manca, A., Korn, B., Poustka, A., Yu, S., Sutherland, G.R., and Mulley, J.C., Fragile X syndrome without CCG amplification has an FMR1 deletion, Nat Genet 1(5), 341, 1992.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 126 Friday, March 10, 2006 2:19 PM

126

Understanding Autism: From Basic Neuroscience to Treatment

40. Wöhrle, D., Hirst, M.C., Kennerknecht, I., Davies, K.E., and Steinbach, P., Genotype mosaicism in fragile X fetal tissues, Hum Genet 89(1), 114, 1992. 41. Wohrle, D., Kotzot, D., Hirst, M.C., Manca, A., Korn, B., Schmidt, A., Barbi, G., Rott, H.D., Poustka, A., and Davies, K.E., A microdeletion of less than 250 kb, including the proximal part of the FMR-I gene and the fragile-X site, in a male with the clinical phenotype of fragile-X syndrome, Am J Hum Genet 51(2), 299, 1992. 42. Ashley, C.T., Sutcliffe, J.S., Kunst, C.B., Leiner, H.A., Eichler, E.E., Nelson, D.L., and Warren, S.T., Human and murine FMR-1: alternative splicing and translational initiation downstream of the CGG-repeat, Nat Genet 4(3), 244, 1993. 43. Kremer, E.J., Pritchard, M., Lynch, M., Yu, S., Holman, K., Baker, E., Warren, S.T., Schlessinger, D., Sutherland, G.R., and Richards, R.I., Mapping of DNA instability at the fragile X to a trinucleotide repeat sequence p(CCG)n, Science 252(5013), 1711, 1991. 44. Kremer, E.J., Yu, S., Pritchard, M., Nagaraja, R., Heitz, D., Lynch, M., Baker, E., Hyland, V.J., Little, R.D., and Wada, M., Isolation of a human DNA sequence which spans the fragile X, Am J Hum Genet 49(3), 656, 1991. 45. Verkerk, A.J., Pieretti, M., Sutcliffe, J.S., Fu, Y.H., Kuhl, D.P., Pizzuti, A., Reiner, O., Richards, S., Victoria, M.F., Zhang, F.P., Eussen, B.E., Van Ommen, G.J.B., Blonden, L.A.J., Riggins, G.J., Chastain, J.L., Kunst, C.B., Galjaard, H., Caskey, C.T., Nelson, D.L., Oostra, B.A., and Warren, S.T., Identification of a gene (FMR-1) containing a CGG repeat coincident with a breakpoint cluster region exhibiting length variation in fragile X syndrome, Cell 65(5), 905, 1991. 46. Fu, Y.H., Kuhl, D.P., Pizzuti, A., Pieretti, M., Sutcliffe, J.S., Richards, S., Verkerk, A.J., Holden, J.J., Fenwick, R., Jr., Warren, S.T., Oostra, B.A., Nelson, D.L., and Caskey, C.T., Variation of the CGG repeat at the fragile X site results in genetic instability: resolution of the Sherman paradox, Cell 67(6), 1047, 1991. 47. Snow, K., Doud, L.K., Hagerman, R., Pergolizzi, R.G., Erster, S.H., and Thibodeau, S.N., Analysis of a CGG sequence at the FMR-1 locus in fragile X families and in the general population, Am J Hum Genet 53(6), 1217, 1993. 48. Malter, H.E., Iber, J.C., Willemsen, R., De Graaff, E., Tarleton, J.C., Leisti, J., Warren, S.T., and Oostra, B.A., Characterization of the full fragile X syndrome mutation in fetal gametes, Nat Genet 15, 165, 1997. 49. Sutcliffe, J.S., Zhang, F., Caskey, C.T., Nelson, D.L., and Warren, S.T., PCR amplification and analysis of yeast artificial chromosomes, Genomics 13(4), 1303, 1992. 50. Hornstra, I.K., Nelson, D.L., Warren, S.T., and Yang, T.P., High resolution methylation analysis of the FMR1 gene trinucleotide repeat region in fragile X syndrome, Hum Mol Genet 2(10), 1659, 1993. 51. Chiurazzi, P., Pomponi, M.G., Willemsen, R., Oostra, B.A., and Neri, G., In vitro reactivation of the FMR1 gene involved in fragile X syndrome, Hum Mol Genet 7(1), 109, 1998. 52. Schwemmle, S., de Graaff, E., Deissler, H., Glaser, D., Wohrle, D., Kennerknecht, I., Just, W., Oostra, B.A., Doerfler, W., Vogel, W., and Steinbach, P., Characterization of FMR1 promoter elements by in vivo-footprinting analysis [erratum appears in Am J Hum Genet 74(3), 598, March 2004. Note: Dorfler W [corrected to Doerfler W]], Am J Hum Genet 60(6), 1354, 1997. 53. Drouin, R., Angers, M., Dallaire, N., Rose, T.M., Khandjian, E.W., and Rousseau, F., Structural and functional characterization of the human FMR1 promoter reveals similarities with the hnRNP-A2 promoter region, Hum Mol Genet 6(12), 2051, 1997.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 127 Friday, March 10, 2006 2:19 PM

Neurobiology of Related Disorders: Fragile X Syndrome

127

54. Reyniers, E., Vits, L., De Boulle, K., Van Roy, B., Van Velzen, D., de Graaff, E., Verkerk, A.J., Jorens, H.Z., Darby, J.K., and Oostra, B., The full mutation in the FMR-1 gene of male fragile X patients is absent in their sperm, Nat Genet 4(2), 143, 1993. 55. Moutou, C., Vincent, M.C., Biancalana, V., and Mandel, J.L., Transition from premutation to full mutation in fragile X syndrome is likely to be prezygotic, Hum Mol Genet 6(7), 971, 1997. 56. Jansen, G., Willems, P., Coerwinkel, M., Nillesen, W., Smeets, H., Vits, L., Howeler, C., Brunner, H., and Wieringa, B., Gonosomal mosaicism in myotonic dystrophy patients: involvement of mitotic events in (CTG)n repeat variation and selection against extreme expansion in sperm, Am J Hum Genet 54, 575, 1993. 57. Hinton, V.J., Brown, W.T., Wisniewski, K., and Rudelli, R.D., Analysis of neocortex in three males with the fragile X syndrome, Am J Med Genet 41(3), 289, 1991. 58. Siomi, H., Matunis, M.J., Michael, W.M., and Dreyfuss, G., The pre-mRNA binding K protein contains a novel evolutionarily conserved motif, Nucleic Acids Res 21(5), 1193, 1993. 59. Kiledjian, M. and Dreyfuss, G., Primary structure and binding activity of the hnRNP U protein: binding RNA through RGG box, Embo J 11(7), 2655, 1992. 60. Brown, V., Small, K., Lakkis, L., Feng, Y., Gunter, C., Wilkinson, K.D., and Warren, S.T., Purified recombinant Fmrp exhibits selective RNA binding as an intrinsic property of the fragile X mental retardation protein, J Biol Chem 273(25), 15521, 1998. 61. Schaeffer, C., Bardoni, B., Mandel, J.L., Ehresmann, B., Ehresmann, C., and Moine, H., The fragile X mental retardation protein binds specifically to its mRNA via a purine quartet motif, Embo J 20(17), 4803, 2001. 62. Darnell, J.C., Jensen, K.B., Jin, P., Brown, V., Warren, S.T., and Darnell, R.B., Fragile X mental retardation protein targets G quartet mRNAs important for neuronal function, Cell 107(4), 489, 2001. 63. Devys, D., Lutz, Y., Rouyer, N., Bellocq, J.P., and Mandel, J.L., The FMR-1 protein is cytoplasmic, most abundant in neurons and appears normal in carriers of a fragile X premutation, Nat Genet 4(4), 335, 1993. 64. Verheij, C., Bakker, C.E., de Graaff, E., Keulemans, J., Willemsen, R., Verkerk, A.J., Galjaard, H., Reuser, A. J., Hoogeveen, A.T., and Oostra, B.A., Characterization and localization of the FMR-1 gene product associated with fragile X syndrome, Nature 363(6431), 722, 1993. 65. Eberhart, D.E., Malter, H.E., Feng, Y., and Warren, S.T., The fragile X mental retardation protein is a ribosonucleoprotein containing both nuclear localization and nuclear export signals, Hum Mol Genet 5, 1083, 1996. 66. Feng, Y., Gutekunst, C.A., Eberhart, D.E., Yi, H., Warren, S.T., and Hersch, S.M., Fragile X mental retardation protein: nucleocytoplasmic shuttling and association with somatodendritic ribosomes, J Neurosci 17(5), 1539, 1997. 67. Khandjian, E.W., Corbin, F., Woerly, S., and Rousseau, F., The fragile X mental retardation protein is associated with ribosomes, Nat Genet 12(1), 91, 1996. 68. Corbin, F., Bouillon, M., Fortin, A., Morin, S., Rousseau, F., and Khandjian, E.W., The fragile X mental retardation protein is associated with poly(A)(+) mRNA in actively translating polyribosomes, Hum Mol Genet 6(9), 1465, 1997. 69. Ceman, S., Brown, V., and Warren, S.T., Isolation of an FMRP-associated messenger ribonucleoprotein particle and identification of nucleolin and the fragile x-related proteins as components of the complex, Mol Cell Biol 19(12), 7925, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 128 Friday, March 10, 2006 2:19 PM

128

Understanding Autism: From Basic Neuroscience to Treatment

70. Ceman, S., Nelson, R., and Warren, S.T., Identification of Mouse YB1/p50 as a Component of the FMRP-Associated mRNP Particle, Biochem Biophys Res Commun 279, 904, 2000. 71. De Boulle, K., Verkerk, A.J., Reyniers, E., Vits, L., Hendrickx, J., Van Roy, B., Van den Bos, F., de Graaff, E., Oostra, B.A., and Willems, P.J., A point mutation in the FMR-1 gene associated with fragile X mental retardation, Nat Genet 3(1), 31, 1993. 72. Musco, G., Stier, G., Joseph, C., Castiglione Morelli, M.A., Nilges, M., Gibson, T.J., and Pastore, A., Three-dimensional structure and stability of the KH domain: molecular insights into the fragile X syndrome, Cell 85(2), 237, 1996. 73. Lewis, H.A., Musunuru, K., Jensen, K.B., Edo, C., Chen, H., Darnell, R.B., and Burley, S.K., Sequence-specific RNA binding by a Nova KH domain: implications for paraneoplastic disease and the fragile X syndrome, Cell 100(3), 323, 2000. 74. Laggerbauer, B., Ostareck, D., Keidel, E.M., Ostareck-Lederer, A., and Fischer, U., Evidence that fragile X mental retardation protein is a negative regulator of translation, Hum Mol Genet 10(4), 329, 2001. 75. Li, Z., Zhang, Y., Ku, L., Wilkinson, K.D., Warren, S.T., and Feng, Y., The fragile X mental retardation protein inhibits translation via interacting with mRNA, Nucleic Acids Res 29(11), 2276, 2001. 76. Wang, J., Pegoraro, E., Menegazzo, E., Gennarelli, M., Hoop, R.C., Angelini, C., and Hoffman, E.P., Myotonic dystrophy: evidence for a possible dominant-negative RNA mutation, Hum Mol Genet 4(4), 599, 1995. 77. Wang, L.L., Richard, S., and Shaw, A.S., P62 association with RNA is regulated by tyrosine phosphorylation, J Biol Chem 270(5), 2010, 1995. 78. Idriss, H., Kumar, A., Casas-Finet, J.R., Guo, H., Damuni, Z., and Wilson, S.H., Regulation of in vitro nucleic acid strand annealing activity of heterogeneous nuclear ribonucleoprotein protein A1 by reversible phosphorylation, Biochemistry 33(37), 11382, 1994. 79. Municio, M.M., Lozano, J., Sanchez, P., Moscat, J., and Diaz-Meco, M.T., Identification of heterogeneous ribonucleoprotein A1 as a novel substrate for protein kinase C zeta, J Biol Chem 270(26), 15884, 1995. 80. Jans, D.A. and Hubner, S., Regulation of protein transport to the nucleus: central role of phosphorylation, Physiol Rev 76(3), 651, 1996. 81. Siomi, M.C., Higashijima, K., Ishizuka, A., and Siomi, H., Casein kinase II phosphorylates the fragile X mental retardation protein and modulates its biological properties, Mol Cell Biol 22(24), 8438, 2002. 82. Ceman, S., O’Donnell, W.T., Reed, M., Patton, S., Pohl, J., and Warren, S.T., Phosphorylation influences the translation state of FMRP-associated polyribosomes, Hum Mol Genet 12(24), 3295, 2003. 83. Caudy, A.A., Myers, M., Hannon, G.J., and Hammond, S.M., Fragile X-related protein and VIG associate with the RNA interference machinery, Genes Dev 16(19), 2491, 2002. 84. Ishizuka, A., Siomi, M.C., and Siomi, H., A Drosophila fragile X protein interacts with components of RNAi and ribosomal proteins, Genes Dev 16(19), 2497, 2002. 85. Jin, P., Zarnescu, D.C., Ceman, S., Nakamoto, M., Mowrey, J., Jongens, T.A., Nelson, D.L., Moses, K., and Warren, S.T., Biochemical and genetic interaction between the fragile X mental retardation protein and the microRNA pathway, Nat Neurosci 7(2), 113, 2004. 86. Rudelli, R.D., Brown, W.T., Wisniewski, K., Jenkins, E.C., Laure-Kamionowska, M., Connell, F., and Wisniewski, H.M., Adult fragile X syndrome, Clinico-neuropathologic findings, Acta Neuropathol 67, 289, 1985.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 129 Friday, March 10, 2006 2:19 PM

Neurobiology of Related Disorders: Fragile X Syndrome

129

87. Harris, K.M. and Kater, S.B., Dendritic spines: cellular specializations imparting both stability and flexibility to synaptic function, Annu Rev Neurosci 17, 341, 1994. 88. Yuste, R. and Majewska, A., On the function of dendritic spines, Neuroscientist 7(5), 387, 2001. 89. Purpura, R.P., Dendritic spine dysgenesis and mental retardation, Science 186, 1126, 1974. 90. D’Hooge, R., Nagels, G., Franck, F., Bakker, C.E., Reyniers, E., Storm, K., Kooy, R.F., Oostra, B.A., Willems, P.J., and De Deyn, P.P., Mildly impaired water maze performance in male Fmr1 knockout mice, Neuroscience 76(2), 367, 1997. 91. Bear, M.F. and Abraham, W.C., Long-term depression in hippocampus, Annu Rev Neurosci 19, 437, 1996. 92. Huber, K.M., Kayser, M.S., and Bear, M.F., Role for rapid dendritic protein synthesis in hippocampal mGluR-dependent long-term depression, Science 288(5469), 1254, 2000. 92a. Weiler, I.J., Irwin, S.A., Klintsova, A.Y., Spencer, C.M., Brazelton, A.D., Miyashiro, K., Comery, T.A., Patel, B., Eberwine, J., Greenough, W.T., Proc Natl Acad Sci U.S.A 94, 5395, 1997. 93. Bear, M.F., Huber, K.M., and Warren, S.T., The mGluR theory of fragile X mental retardation, Trends Neurosci 27(7), 370, 2004. 94. Bliss, T.V. and Collingridge, G.L., A synaptic model of memory: long-term potentiation in the hippocampus, Nature 361(6407), 31, 1993. 95. Raymond, C.R., Thompson, V.L., Tate, W.P., and Abraham, W.C., Metabotropic glutamate receptors trigger homosynaptic protein synthesis to prolong long-term potentiation, J Neurosci 20(3), 969, 2000. 96. Godfraind, J.M., Reyniers, E., Deboulle, K., Dhooge, R., Dedeyn, P.P., Bakker, C.E., Oostra, B.A., Kooy, R.F., and Willems, P.J., Long-term potentiation in the hippocampus of fragile X knockout mice, Am J Med Genet 64(2), 246, 1996. 97. Paradee, W., Melikian, H.E., Rasmussen, D.L., Kenneson, A., Conn, P.J., and Warren, S.T., Fragile X mouse: strain effects of knockout phenotype and evidence suggesting deficient amygdala function, Neuroscience 94(1), 185, 1999. 98. Hagerman, R.J., Medical follow-up and pharmacotherapy, in Fragile X Syndrome: Diagnosis, Treatment, and Research, Hagerman, R.J. and Cronister, A., Eds., 1996, The Johns Hopkins University Press, Baltimore, p. 283. 99. Feng, Y., Zhang, F., Lokey, L.K., Chastain, J.L., Lakkis, L., Eberhart, D., and Warren, S.T., Translational suppression by trinucleotide repeat expansion at FMR1, Science 268(5211), 731, 1995. 100. Chiurazzi, P., Pomponi, M.G., Pietrobono, R., Bakker, C.E., Neri, G., and Oostra, B.A., Synergistic effect of histone hyperacetylation and DNA demethylation in the reactivation of the FMR1 gene, Hum Mol Genet 8(12), 2317, 1999. 101. Coffee, B., Zhang, F., Warren, S. T., and Reines, D., Acetylated histones are associated with FMR1 in normal but not fragile X-syndrome cells [erratum appears in Nat Genet 22(2), 209, June 1999], Nat Genet 22(1), 98, 1999. 102. Kenneson, A., Zhang, F., Hagedorn, C.H., and Warren, S.T., Reduced FMRP and increased FMR1 transcription is proportionally associated with CGG repeat number in intermediate-length and premutation carriers, Hum Mol Genet 10(14), 1449, 2001. 103. Peier, A.M., McIlwain, K.L., Kenneson, A., Warren, S.T., Paylor, R., and Nelson, D.L., (Over)correction of FMR1 deficiency with YAC transgenics: behavioral and physical features, Hum Mol Genet 9(8), 1145, 2000.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 130 Friday, March 10, 2006 2:19 PM

130

Understanding Autism: From Basic Neuroscience to Treatment

104. Huber, K.M., Gallagher, S.M., Warren, S.T., and Bear, M.F., Altered synaptic plasticity in a mouse model of fragile X mental retardation, Proc Natl Acad Sci U S A 99, 7746, 2002. 105. Brown, V., Jin, P., Ceman, S., Darnell, J.C., O’Donnell, W.T., Tenenbaum, S.A., Jin, X., Feng, Y., Wilkinson, K.D., Keene, J.D., Darnell, R.B., and Warren, S.T., Microarray identification of FMRP-associated brain mRNAs and altered mRNA translational profiles in fragile X syndrome, Cell 107(4), 477, 2001. 106. Micali, N., Chakrabarti, S., and Fombonne, E., The broad autism phenotype: findings from an epidemiological survey, Autism 8(1), 21, 2004. 107. Chakrabarti, S. and Fombonne, E., Pervasive developmental disorders in preschool children, JAMA 285(24), 3093, 2001. 108. Fombonne, E., The epidemiology of autism: a review, Psychol Med 29(4), 769, 1999. 109. Rutter, M., Genetic studies of autism: from the 1970s into the millennium, J Abnormal Child Psychol 28(1), 3, 2000. 110. Szatmari, P., Jones, M.B., Zwaigenbaum, L., and MacLean, J.E., Genetics of autism: overview and new directions, J Autism Dev Disord 28(5), 351, 1998. 111. Farber, J.M., Therapy for acute otitis media, Arch Pediatr Adolesc Med 150(12), 1315, 1996. 112. Baker, P., Piven, J., and Sato, Y., Autism and tuberous sclerosis complex: prevalence and clinical features, J Autism Dev Disord 28(4), 279, 1998. 113. Rogers, S.J., Wehner, D.E., and Hagerman, R., The behavioral phenotype in fragile X: symptoms of autism in very young children with fragile X syndrome, idiopathic autism, and other developmental disorders, J Dev Behav Pediatr 22(6), 409, 2001. 114. Bailey, D.B., Jr., Mesibov, G.B., Hatton, D.D., Clark, R.D., Roberts, J.E., and Mayhew, L., Autistic behavior in young boys with fragile X syndrome, J Autism Dev Disord 28(6), 499, 1998. 115. Bailey, D.B., Jr., Hatton, D.D., and Skinner, M., Early developmental trajectories of males with fragile X syndrome, Am J Ment Retard 103(1), 29, 1998. 116. Ritvo, E.R., Jorde, L.B., Mason-Brothers, A., Freeman, B.J., Pingree, C., Jones, M.B., McMahon, W.M., Petersen, P.B., Jenson, W.R., and Mo, A., The UCLA-University of Utah epidemiologic survey of autism: recurrence risk estimates and genetic counseling, Am J Psychiatry 146(8), 1032, 1989. 117. Fisch, G.S., Cohen, I.L., Wolf, E.G., Brown, W.T., Jenkins, E.C., and Gross, A., Autism and the fragile X syndrome, Am J Psychiatry 143(1), 71, 1986. 118. Fombonne, E., Du Mazaubrun, C., Cans, C., and Grandjean, H., Autism and associated medical disorders in a French epidemiological survey, J Am Acad Child Adolesc Psychiatry 36(11), 1561, 1997. 119. Fombonne, E., Bolton, P., Prior, J., Jordan, H., and Rutter, M., A family study of autism: cognitive patterns and levels in parents and siblings, J Child Psychol Psychiatry 38(6), 667, 1997. 120. Estecio, M., Fett-Conte, A.C., Varella-Garcia, M., Fridman, C., and Silva, A.E., Molecular and cytogenetic analyses on Brazilian youths with pervasive developmental disorders, J Autism Dev Disord 32(1), 35, 2002. 121. Watson, M.S., Leckman, J.F., Annex, B., Breg, W.R., Boles, D., Volkmar, F.R., Cohen, D.J., and Carter, C., Fragile X in a survey of 75 autistic males, N Engl J Med 310(22), 1462, 1984. 122. Li, S.Y., Chen, Y.C., Lai, T.J., Hsu, C.Y., and Wang, Y.C., Molecular and cytogenetic analyses of autism in Taiwan, Hum Genet 92(5), 441, 1993.

© 2006 by Taylor & Francis Group, LLC

2732_C006.fm Page 131 Friday, March 10, 2006 2:19 PM

Neurobiology of Related Disorders: Fragile X Syndrome

131

123. Crino, P.B., Trojanowski, J.Q., and Eberwine, J., Internexin, MAP1B, and nestin in cortical dysplasia as markers of developmental maturity, Acta Neuropathol (Berl) 93(6), 619, 1997. 124. Cheng, A. and Krueger, B.K., and Bambrick, L.L., MAP5 expression in proliferating neuroblasts, Brain Res Dev Brain Res 113(1–2), 107, 1999. 125. Diaz-Nido, J. and Avila, J., Quantitation of microtubule-associated protein MAP-1B in brain and other tissues, Int J Biochem 21(7), 723, 1989. 126. Schoenfeld, T.A., McKerracher, L., Obar, R., and Vallee, R. B., MAP 1A and MAP 1B are structurally related microtubule associated proteins with distinct developmental patterns in the CNS, J Neurosci 9(5), 1712, 1989. 127. Gonzalez-Billault, C., Jimenez-Mateos, E.M., Caceres, A., Diaz-Nido, J., Wandosell, F., and Avila, J., Microtubule-associated protein 1B function during normal development, regeneration, and pathological conditions in the nervous system, J Neurobiol 58(1), 48, 2004. 128. Zhang, Y.Q., Bailey, A.M., Matthies, H.J., Renden, R.B., Smith, M.A., Speese, S.D., Rubin, G.M., and Broadie, K., Drosophila fragile X-related gene regulates the MAP1B homolog Futsch to control synaptic structure and function, Cell 107(5), 591, 2001. 129. Skaar, D.A., Shao, Y., Haines, J.L., Stenger, J.E., Jaworski, J., Martin, E.R., Delong, G.R., Moore, J.H., McCauley, J.L., Sutcliffe, J.S., Ashley-Koch, A.E., Cuccaro, M.L., Folstein, S.E., Gilbert, J.R., and Pericak-Vance, M.A., Analysis of the RELN gene as a genetic risk factor for autism, Mol Psychiatry, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 133 Friday, March 10, 2006 2:56 PM

7

Fear and Anxiety Pathways Kevin S. LaBar and Joseph E. LeDoux

CONTENTS Introduction............................................................................................................133 The Functions of Fear and Anxiety ......................................................................134 Fear Conditioning as a Model System for Studying Emotional Learning and Memory ...........................................................................................134 The Neural Circuitry for Fear Learning................................................................136 Acquiring Fears to Danger Signals: Role of the Amygdala .........................136 Prefrontal–Amygdala Interactions during Fear Extinction............................139 The Hippocampus and Fear of Environmental Contexts ..............................141 Fear Conditioning in Humans ...............................................................................143 Psychophysiological Studies ..........................................................................143 Neuropsychological Investigations in Brain-Lesioned Patients ....................143 Functional Neuroimaging of Conditioned Fear Pathways ............................144 Fear Conditioning in Anxiety Disorders........................................................145 Fear and Anxiety Pathways: Implications for Autism Research ..........................147 Conclusions............................................................................................................147 References..............................................................................................................148

INTRODUCTION One of the behavioral hallmarks of autistic spectrum disorders is a deficit in socioemotional processing. Historically, such functions have been associated with activity in limbic forebrain structures, which are anatomically positioned to control internal body reactions to salient environmental stimuli. In recent years, the brain–behavior relationships that mediate emotional functions have been elucidated in greater detail. This knowledge has been made possible by scientific advances that allow more accurate probing of limbic brain regions in humans and nonhuman animals, as well as renewed academic interest in studying motivated and social behaviors at the neural level. As scientists approach a new era for understanding the genetic and neurobiological underpinnings of autism, it is important to consider methodological lines of inquiry that have shaped current views of emotional information processing in the brain. In this regard, much progress has been made by focusing research on specific emotions using experimental paradigms that can be

133

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 134 Friday, March 10, 2006 2:56 PM

134

Understanding Autism: From Basic Neuroscience to Treatment

adapted from animal models to clinical populations. Here, we review the brain systems that regulate the emotion fear, with an emphasis on the neural mechanisms that subserve learning and memory of fear associations established through classical conditioning procedures.

THE FUNCTIONS OF FEAR AND ANXIETY Emotions can be construed as action dispositions that enable the body to respond appropriately when confronted with salient environmental stimuli. The adaptive value of fear as an emotion can be understood within an evolutionary context whereby detecting the presence of threat in the environment is critical for survival. Situations of danger generate a complex stress response with motoric, autonomic, endocrine, and immunological sequelae that mobilize the body’s energy resources for defense against potential imminent harm. Vigilance systems must act quickly and exert a powerful influence over sensory, cognitive, and visceromotor domains to prepare the body for attack, execute coping strategies, and remember the properties and locations of threatening stimuli so they can be avoided in the future. When the organism is in a prolonged state of readiness to confront potential situations of danger, anxiety ensues. In humans, the range of contexts capable of eliciting anxiety has expanded physically, psychologically, and socially, and fear-processing circuitry has adapted to respond to man-made environmental threats. A breakdown of defensive mechanisms makes the organism vulnerable to environmental stressors. The importance of fear in behavioral regulation is accordingly mirrored by the prevalence of anxiety disorders in humans, which are estimated to have a lifetime occurrence of about 19.2% in men and 30.5% in women.1

FEAR CONDITIONING AS A MODEL SYSTEM FOR STUDYING EMOTIONAL LEARNING AND MEMORY Although many stressors in the environment have innate signal value (e.g., the sight of a snake), others are learned through experience, either directly or by observation. A fundamental way in which emotional associations are formed was discovered by Pavlov a century ago. Pavlov showed that neutral stimuli (e.g., ringing of a bell) presented in anticipation of a biologically significant reinforcer (e.g., food) come to elicit physiological responses (e.g., salivation) that indicate a change in motivational salience. When such a conditioned stimulus (CS) reliably predicts an aversive reinforcer (unconditioned stimulus, or US), a conditioned response develops that is indicative of a state of fear (Figure 7.1). Multiple bodily responses are triggered in parallel, including alterations in the animal’s mobility (freezing), heart rate, sweat gland activity, reflex potentiation, and stress hormone release, until the reinforcement contingency changes and the CS provides new information that the reinforcer is no longer being delivered. At this point, the fear response is reduced in frequency or magnitude and is said to undergo extinction. Subsequent studies have shown that the predictive relationship between the CS and US is more important than their

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 135 Friday, March 10, 2006 2:56 PM

Fear and Anxiety Pathways

off

135

Skin conductance response

on

CS (tone)

Habituation

US (shock) CS

Acquisition

US

4 3 Trial # 2 1

CS

Extinction Time

US CS

US

CR UR

FIGURE 7.1 Design of a fear conditioning experiment. Left panel: The procedure typically involves three phases — habituation, acquisition, and extinction. In the habituation phase, the conditioned stimulus or CS (typically an auditory or visual stimulus) is presented alone for a few trials to diminish any novelty or orienting responses prior to learning. In the acquisition phase, the CS predicts the occurrence of a noxious unconditioned stimulus or US (typically a shock or loud noise). Conditioned fear responses develop to the presentation of the CS as well as to the environmental context in which the CS–US associations are learned. In the extinction phase, the CS is again presented alone to extinguish the learned fear response. Only three trials of each phase are depicted. Responses to a predictive control stimulus (CS+) are typically compared to an unpredictive control stimulus (CS−) (not shown). Right panel: Example of skin conductance response (SCR) as a physiological index of conditioned and unconditioned fear in a healthy human participant. Four acquisition trials are illustrated. Skin conductance is measured from the palmar surface of the second and third digits of the participant’s hand. Time of CS onset is indicated by dashed vertical line. Increases in SCR amplitude elicited by the CS (CR, open arrow) and US (UR, closed arrow) are measured on a trial-by-trial basis and averaged across blocks of trials during each experimental phase.

temporal contiguity, that some conditioned associations are more readily learned than others, and that conditioned responses can be acquired in a single learning trial if the reinforcer is of sufficient intensity.2 By arranging more intricate spatiotemporal relationships among stimuli and reinforcers, it is possible to model complex and dynamic aspects of fear-related behaviors. Because this form of emotional learning is nonverbal, observed throughout the phyla, elicited by a variety of sensory stimuli and reinforcers, and easy to measure, quantify, and manipulate experimentally, it has provided a tractable solution to the neuroscientific study of fear and anxiety. Fear conditioning is thought to play a broad role in extracting information about warning cues and their relationship to environmental dangers, in initiating the engagement of coping skills, and in the establishment and maintenance of human fears in a variety of social contexts.3 Importantly, it has provided a window into brain mechanisms that underlie fear across species, including human clinical populations. As discussed in the following text, identifying pathophysiological changes in the relevant fear pathways is key to understanding the neural bases of anxiety disorders and their normalization following treatment.

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 136 Friday, March 10, 2006 2:56 PM

136

Understanding Autism: From Basic Neuroscience to Treatment

THE NEURAL CIRCUITRY FOR FEAR LEARNING ACQUIRING FEARS

TO

DANGER SIGNALS: ROLE

OF THE

AMYGDALA

Studies in nonhuman animals have been instrumental in defining the relevant neuroanatomical pathways, neurotransmitter systems, and intracellular events that contribute to the acquisition, extinction, and postextinction recovery of fear learning. The amygdala, an almond-shaped structure in the medial temporal lobe, has emerged as a key brain region in which important neural processing takes place to form the CS–US associations that underlie conditioned fear learning.4 Initial studies had shown that lesions to this structure impair fear conditioning.5 Subsequent research has focused on identifying input pathways that provide sensory information about fear-relevant stimuli to the amygdala, intra-amygdalar processing of fear signals, and output pathways that lead to motoric and visceral expressions of defensive behavior (Figure 7.2). The amygdala receives sensory input via direct subcortical (thalamo-amygdala) and indirect cortical (thalamo-cortico-amygdala) pathways. Each of these routes of information processing is capable of mediating conditioned fear associations to discrete cues.6 The direct, subcortical route is adaptive to the organism because it provides a rapid means of identifying sources of threat in the environment so that

Neocortex Ventromedial PFC Central

ITC

Lateral

Thalamus

Sensory input

Hippocampus

Basal Striatum Hypothalamus

Fear output

Amygdala

Midbrain/brainstem

FIGURE 7.2 A simplified model of the neural pathways for acquiring and extinguishing conditioned fear. Input pathways (top) are indicated by solid arrows; output pathways (bottom) are indicated by dashed arrows. Subcortical structures are illustrated as ellipses, and cortical structures are illustrated as rectangles. Many input projections to the amygdala are reciprocal (not shown). Subdivisions of the lateral, central, and basal nuclei, as well as additional subnuclei of the amygdala are not depicted. ITC = inhibitory intercalated cells, PFC = prefrontal cortex.

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 137 Friday, March 10, 2006 2:56 PM

Fear and Anxiety Pathways

137

immediate threats can be responded to quickly. This pathway, however, can only support crude information processing of sensory signals. It is thought that processing along the subcortical route serves to prime the amygdala to receive additional information from the cortical pathway, which is capable of extracting sensory and perceptual details to identify the exact nature of the potential threat. Both of these processing streams converge onto neurons in the lateral nucleus of the amygdala, which serves as its sensory interface.7 Neurons in the lateral nucleus are sensitive to multiple sensory modalities and can thus integrate auditory or visual CS information with somatosensory information relayed by the noxious US.8 On reaching the lateral nucleus, fear signals are processed forward within the amygdala via intrinsic connections to both the basal and central nuclei.9 The basal nucleus receives additional inputs, including information about spatial contexts that is transmitted via the hippocampus. The basal nucleus, in turn, sends projections out to the striatum and association cortex in the frontal and temporal lobes, which may be important for initiating coping strategies once fearful stimuli have been detected.10,11 Information from both the lateral and basal nuclei converge on the central nucleus, which serves as the amygdala’s primary interface to brain stem and hypothalamic centers that control the expression of defensive behaviors. Importantly, lesions to these efferent structures only impair the expression of fear in specific output channels. For example, lesions of the central gray region of the midbrain impair the expression of conditioned immobility (freezing behavior) but not conditioned heart rate changes, whereas lesions to the hypothalamus produce the opposite effect.12 The central nucleus is therefore the last way station along the conditioned fear pathway that plays an integrative role in CS processing by coordinating diverse species-typical defensive reactions to generate a complex fear and stress response. Recent lesion studies have shown that damage to just the lateral and central nuclei (but not other subnuclei) are sufficient to yield deficits in fear conditioning to simple cues. 13 Therefore, these two structures constitute the minimal neural circuitry necessary to form conditioned fear associations within the amygdala (Figure 7.3). Electrophysiological studies have confirmed the anatomic and lesion findings to support a key role for the amygdala in fear conditioning. Recordings of neurons in the lateral nucleus of awake, behaving rats during fear conditioning show that the most prominent increases in firing rates occur within 15 msec of CS onset.14–16 This latency implicates CS-evoked changes in processing along the rapid thalamoamygdala pathway. There is also evidence for synaptic plasticity in the central nucleus of the amygdala, cortex, and thalamus, which may contribute to fear conditioning in important ways, although these changes in neuronal signaling of the CS occur later in time and only after additional training trials.15,17 Moreover, lesions of the amygdala reduce synaptic plasticity observed at cortical and thalamic sites, which suggests that the amygdala may be a driving force behind fear-induced changes in neuronal excitation occurring elsewhere in the brain, although additional anatomic pathways may be involved.18,19 Parallel investigations of long-term potentiation (LTP) have provided important links between synaptic changes taking place in amygdala neurons and conditioned fear behavior. LTP is induced by high-frequency stimulation of afferent pathways

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 138 Friday, March 10, 2006 2:56 PM

138

Understanding Autism: From Basic Neuroscience to Treatment

Pre-CS 60

CS

Freezing (sec)

50 40

All amy

30 20 10 M

B + AB

AB

B

LA

CE

All amy

Sham

Unoperated

LA Unpaired

0

CE

FIGURE 7.3 Lesions of the amygdala impair the acquisition of conditioned fear in the rat. Lesions to the lateral nucleus (LA), central nucleus (CE), or the entire amygdala (all Amy) yield similar deficits in acquisition of conditioned freezing responses to a tone CS that predicts the occurrence of a footshock US. Lesions to other subregions of the amygdala, including the basal (B), accessory basal (AB), and medial (M) nuclei have no effect in comparison to sham-lesioned and unoperated animals. These data illustrate that the lateral and central subnuclei form the minimal circuitry necessary for acquiring conditioned fear associations within the amygdala. (Adapted from Nader, K. et al., Damage to the lateral and central, but not other amygdaloid nuclei prevents the acquisition of auditory fear conditioning, Learn. Mem., 8, 156, 2001. With permission.)

to a given structure and is measured as a long-lasting increase in the amplitude of excitatory postsynaptic potentials. Because LTP yields an enduring, activity-dependent increase in synaptic transmission, it is a currently prevailing model of how learning and memory processes are instantiated neurophysiologically.20 LTP has been induced in the thalamo-amygdala, cortico-amygdala, and hippocampal-amygdala pathways.21–23 The classic form of LTP is dependent on activation of glutamatergic NMDA and AMPA receptors, and the amygdala is a rich source of glutamatergic transmission. Intra-amygdala blockade of NMDA receptors reduces both LTP and conditioned fear behavior, although at some synapses a non-NMDA-dependent form of LTP is involved.24–27 Importantly, LTP enhances auditory-evoked potentials along the thalamo-amygdala pathway, and fear conditioning induces LTP along the thalamo-amygdala and cortico-amygdala pathways in vitro and in vivo.28–31 This line of research has yielded a tight link between the behavioral and molecular levels of analysis. Following the induction of synaptic plasticity, intracellular second messengers, gene transcription, and protein synthesis consequently generate cytoskeletal and adhesion remodeling, which stabilize the functional alterations in synaptic transmission over time. Morphological changes in dendritic spines include modifications of existing spine volume and number, spine morphogenesis, glutamate receptor trafficking into spines, and adhesion between the pre- and postsynaptic

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 139 Friday, March 10, 2006 2:56 PM

Fear and Anxiety Pathways

139

elements.32 Such changes are hypothesized to form the substrate of long-term retention of fear memories at the cellular level.

PREFRONTAL–AMYGDALA INTERACTIONS

DURING

FEAR EXTINCTION

The ability to respond flexibly to changes in the salience of stimuli is a hallmark of emotional well-being. Fear reactions may no longer be appropriate if the stimulus loses its predictive relationship to a dangerous outcome. Persistent and maladaptive fear responses to environmental elicitors are hallmarks of anxiety disorders, including phobias, panic disorder, and posttraumatic stress disorder. Exposure therapy (behavioral desensitization or extinction training) is predicated on the notion that presenting oncefeared cues in the absence of reinforcement converts them into safety signals, which suppresses the fear response and alleviates the patient’s distress. It has been long recognized that extinction training does not eradicate the original fear memory. Rather, the fear memory is superseded by a memory for extinction training that should retroactively inhibit the fear association and dominate behavior.33 Because the treatment of anxiety disorders rests on the ability to control fear responses after they have already been acquired, delineating the brain mechanisms of fear extinction has immense clinical value. Given the anatomy of fear pathways reviewed in the preceding text, it is useful to approach the neurobiology of extinction learning by considering how conditioned-fear associations established through amygdalar circuitry come under the influence of executive control mechanisms. In this regard, the prefrontal cortex is a prime candidate, because this brain region is involved in a variety of executive functions across sensoryperceptual, cognitive-mnemonic, and affective-social domains.34 In the rat, lesions of the ventromedial prefrontal cortex (vmPFC) or its dopaminergic inputs selectively prolong extinction learning but do not impact the initial acquisition of conditioned fear.35–37 Recent studies have shown that vmPFC lesions particularly affect the retention of extinction learning 24 h later rather than the short-term suppression of fear responses across training trials.38 Electrical stimulation of the vmPFC paired with presentation of a CS suppresses freezing behavior in rats, and recordings of vmPFC neurons show selective increases in spike firing in response to an extinguished CS 24 h after the initial extinction training session (Figure 7.4).39,40 LTP along the thalamo-vmPFC pathway is associated with maintenance but not initial acquisition of extinction memories, and retrieval of prior extinction training enhances metabolic activity in the vmPFC.41,42 These results suggest that the vmPFC contributes to the consolidation and delayed retrieval of extinction memories. The vmPFC exerts its influence over the expression of fear behavior via inhibitory interactions with the amygdala. A network of GABAergic interneurons is present within the intra-amygdala circuitry relevant for conditioned fear learning.43,44 These neurons are located within the lateral, basal, and central nuclei as well as in an intercalated cell mass that is positioned between these structures (Figure 7.2). Electrical stimulation of the vmPFC yields a reduction of amygdalar signaling in response to electrical pulses or presentations of a feared CS.45–47 Blockade of NMDA receptors or second messenger signaling in the basolateral amygdala impairs extinction.48,49 Neurons in the lateral amygdala typically reduce

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 140 Friday, March 10, 2006 2:56 PM

140

Understanding Autism: From Basic Neuroscience to Treatment

Infralimbic cortex (IL)

(a)

Freezing to tone (%)

100

Day 1 Cond.

Day 2

Extinction

Extinction

75 50 25 0 Trial blocks (b)

Spikes

Habit. + cond.

Extinction

Extinction

20

20

20

10

10

10

0 −1 0

1

2

0 −1 0 1 2 Time (s) (c)

0 −1 0

1

2

FIGURE 7.4 Electrophysiological correlates of fear extinction in the ventromedial prefrontal cortex (vmPFC) of the rat. (a) Recordings were obtained from the infralimbic region of the vmPFC. (b) Rats exhibited increases and decreases in conditioned freezing behavior in response to a tone CS commensurate with the training protocol on day 1. On day 2, a second extinction training session was conducted, and rats exhibited little freezing, which shows good retention of extinction learning from day 1. (c) Although neurons did not respond to the CS during initial acquisition or extinction training on day 1, they did show enhanced signaling of the CS during the extinction training session conducted on day 2. These results indicate that the vmPFC plays a selective role in the consolidation and retrieval of memories for fear extinction over time. (From Milad M.R. and Quirk, G.J., Neurons in medial prefrontal cortex signal memory for fear extinction, Nature, 420, 70, 2002.)

their CS-evoked firing rates during extinction training, although some cells remain responsive to the CS after fear responses are suppressed behaviorally.50 Moreover, cells within the lateral amygdala retain their conditioning-induced functional coupling as evidenced by synchronous firing patterns after extinction training, even

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 141 Friday, March 10, 2006 2:56 PM

Fear and Anxiety Pathways

141

when the CS is no longer actively signaled.14 These latter observations provide evidence that traces of the original fear memory remain present in amygdala neurons even after fear is no longer expressed. In the absence of cortical influences, conditioned-fear associations in the amygdala remain indelible, and dysfunctional prefrontal–amygdala interactions could contribute to emotional perseveration as exemplified in anxiety disorders.51,52

THE HIPPOCAMPUS

AND

FEAR

OF

ENVIRONMENTAL CONTEXTS

Fears may be elicited not only by specific sights or sounds but also by certain locations or settings (“contexts”) in the environment. For example, one may fear the context of dark alleys even in the absence of explicit danger signs being present. It is adaptive for fear responses to be gated by appropriate contextual cues, as fear responses elicited in safe environments needlessly expend the body’s energy resources. Generalization of fears to irrelevant contexts is characteristic of certain anxiety disorders, including posttraumatic stress disorder and phobias. Recovery of fears following exposure therapy complicates the treatment of anxiety disorders, and contextual factors are thought to play an important role in relapse.53 Animal models have shown that extinguishing of fear is particularly sensitive to the setting in which extinction training takes place.33 Thus, the neural mechanisms that contribute to contextual fear and context-dependent recovery of fear are as important to understand as those that contribute to acquisition and extinction of cued fear. In conditioning paradigms, contextual fear is assessed by presenting the CS in a distinctive testing chamber and then measuring fear responses to the chamber itself in the absence of CS presentation. Lesions of the basolateral amygdala impair the acquisition of both cued and contextual fear.54,55 As mentioned in the preceding text, information about spatial contexts is communicated to the amygdala via hippocampal projections that terminate primarily in the basal nucleus. Pretraining lesions of the dorsal hippocampus impair contextual fear acquisition but leave cued fear conditioning intact.54,56 This effect is especially prominent when the environment is processed configurally (in a conjunctive, unified representation) rather than elementally (as separate features), as determined by context discrimination and context preexposure tests.57,58 Posttraining lesions of the dorsal hippocampus lead to a temporally limited retrograde amnesia of contextual fear (Figure 7.5). Lesioned rats exhibit impaired memory for contexts that were fear-conditioned 1 to 28 d prior to surgery but no memory impairment for contexts that were fear-conditioned 28 to 100 d prior to surgery. This temporal gradation in retrograde amnesia following hippocampal damage parallels findings from spatial memory in rats and declarative (explicit) memory in humans.59 After a period of memory consolidation, it is thought that contextual representations are maintained directly in the neocortex. In contrast to the hippocampus, posttraining lesions of the basolateral amygdala lead to a more permanent retrograde loss of cued and contextual fear.60,61 Thus, the integrity of the amygdala appears to be important for the retention and expression of fear memories regardless of its source of input regarding the sensory features of the environment (i.e., hippocampus or cortex).

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 142 Friday, March 10, 2006 2:56 PM

142

Understanding Autism: From Basic Neuroscience to Treatment

Freezing (% time)

Remote contextual fear

Recent contextual fear

Context summary

100

100

100

75

75

75

50

50

50

25

25

25

Sham Hipp

0

1 2 3 4 5 6 7 8 Minutes (a)

0

1 2 3 4 5 6 7 8 Minutes (b)

0

Remote Recent (first six min) (c)

FIGURE 7.5 Posttraining hippocampal lesions produce a temporally graded retrograde amnesia of contextual fear. Rats underwent two fear conditioning procedures, each conducted in distinct testing chambers using different CSs. The first training session took place 50 d prior to surgery, and the second training session took place 1 d prior to surgery. Following recovery from surgery, rats were tested for their memory of the two contexts as measured by the amount of freezing behavior expressed in each testing chamber. Lesions to the dorsal hippocampus impaired recent memory for contextual fear associations learned 1 d prior to surgery (b and c) but did not affect remote memory for contextual fear associations learned 50 d prior to surgery (a and c). These findings provide evidence that the hippocampus plays a time-limited role in consolidating and retrieving memories of fearful environments. (Adapted from Anagnostaras, S.G., Maren, S. and Fanselow, M.S., J. Neurosci., 119, 1106, 1999. With permission.)

Two experimental procedures have led to key insights into the context dependency of fear recovery following extinction — reinstatement and renewal.62 Both tasks involve context manipulations after fear to a specific CS has already been acquired and extinguished. Following extinction, the animal has two associative memories that may compete for behavior — the CS–US memory from acquisition training and the CS–no US memory from extinction training. Context manipulations following extinction training may influence which of these memories is recalled. During fear reinstatement, the animal is reexposed to the noxious US in a particular context, and then the extinguished CS is presented either in the reexposure context or an alternate one. Fear responses to the CS are recovered only when the CS is presented in the same context as the reexposed US. During fear renewal, fear responses to an extinguished CS are recovered if the CS is presented in a novel context (one that differed from the extinction context). Pretraining lesions or postextinction inactivation of the dorsal hippocampus impairs fear recovery on these paradigms, particularly when there is ambiguity between the recovery test context and the contexts used for acquisition and extinction training.63–66 Neuronal firing patterns in the lateral amygdala also exhibit renewal in that responses to an extinguished CS are higher in a novel context than in the extinction context.67 Contextual influences over fear extinction are hypothesized to involve interactions among the hippocampus, vmPFC, and amygdala, but the details of the regional interactions have not been established yet.52

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 143 Friday, March 10, 2006 2:56 PM

Fear and Anxiety Pathways

143

FEAR CONDITIONING IN HUMANS PSYCHOPHYSIOLOGICAL STUDIES In the presence of conditioned fear stimuli, humans exhibit a complex physiological response, including increases in skin conductance responses (SCR), potentiated startle reflexes, and alterations in heart rate. Although conditioned fear responses are relatively stable within individuals, they vary widely across them.68 Intersubject variability is associated with several factors, including personality, genetics, and interactions between biological sex and stress hormones.69–73 In humans, the conditioning process is also accompanied by declarative knowledge (conscious awareness) of the reinforcement parameters. Although there is debate concerning the necessary and sufficient conditions of its influence, declarative knowledge can modulate the expression of fear behavior, especially on complex tasks, such as discrimination tasks, in which one CS is reinforced (CS+) and other CSs are not (CS−), and trace conditioning tasks, in which a retention interval is interposed between the offset of the CS and onset of the US.74–78 Consistent with biological preparedness theory, conditioning to fear-relevant (phobic) stimuli, including snakes, spiders, and fearful or angry facial expressions, tends to be less resistant to extinction and can be acquired unconsciously using visual masking techniques.79–81 Subliminal fear conditioning has been influential as a model for understanding how fear and anxiety can arise in the absence of awareness on the part of the individual.82

NEUROPSYCHOLOGICAL INVESTIGATIONS

IN

BRAIN-LESIONED PATIENTS

The anatomical pathways of fear conditioning are highly conserved, as shown by studies of neurologic patients with damage to the brain regions implicated in animal research. Although organic syndromes that affect the human brain are not as spatially precise as those created by lesion techniques in nonhuman animals, the functions of the amygdala, hippocampus, and vmPFC in emotional learning have been confirmed in human patients. Patients with damage to the medial temporal lobe that includes the amygdala are impaired at acquiring conditioned fear associations (Figure 7.6).83–87 Conditioning deficits occur even when patients can verbalize the reinforcement contingency and demonstrate intact fear responses to the noxious US itself. These observations implicate a deficit in an implicit emotional learning system rather than a deficit in declarative knowledge or fear expression following amygdala damage. In contrast, patients with selective damage to the hippocampus but intact amygdalae show the opposite dissociation — they cannot verbalize the reinforcement contingency but nonetheless acquire fear normally on simple conditioning tasks.83,88 Recent evidence also suggests that the human hippocampus is important for aspects of contextual fear conditioning. Despite showing normal fear acquisition, amnesics with selective hippocampal damage do not exhibit contextdependent reinstatement of conditioned fear after extinction (Figure 7.7).88 The role of the vmPFC in extinction learning has been somewhat difficult to specify, because patients with damage to this region tend to have reduced autonomic expression.89 Nonetheless, patients with vmPFC damage have difficulty extinguishing and reversing learned visual associations and exhibit other evidence of emotional perseveration

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 144 Friday, March 10, 2006 2:56 PM

144

Understanding Autism: From Basic Neuroscience to Treatment

1.5

0.2

∗

SCR (√uS)

SCR (√uS)

0.15 1

0.5

Control SP

0.1 0.05 0

−0.05

0 Control

SP

Habituation

Acquisition

Extinction

FIGURE 7.6 Impaired fear conditioning in a patient with bilateral amygdala damage consequent to temporal lobe epilepsy. Each data point represents an average of four trials. Right panel: In contrast to healthy controls, patient S.P. does not exhibit acquisition of fear, as measured by skin conductance response (SCR) to a visual CS paired with a mild wrist shock (*p < .05). Left panel: Patient S.P., however, does show intact SCRs to the shock itself, indicating a normal ability to process noxious signals and to express unconditioned fear. She also has intact declarative knowledge regarding the CS–US relationship. These observations implicate a deficit in an implicit emotional learning system following amygdala damage in humans. uS = microSiemens. (Adapted from Phelps, E.A. et al., Specifying the contributions of the human amygdala to emotional memory: a case study, Neurocase, 4, 527, 1998. With permission.)

similar to that found during extinction training in rats with infralimbic and orbitofrontal damage.90–92

FUNCTIONAL NEUROIMAGING

OF

CONDITIONED FEAR PATHWAYS

The functional anatomy of conditioned fear has been confirmed by neuroimaging studies of the healthy brain. CS-evoked changes in amygdala activation occur during both acquisition and extinction training (Figure 7.8).93–98 Moreover, the degree of amygdala activation is correlated with physiologic indices of fear learning in individual participants.94,96,99 The amygdala’s role is greatest during the first few training trials when the conditioned associations are initially formed, similar to the electrophysiological response profiles of some lateral amygdala neurons in the rat.93,94 Amygdala activation is also elicited by angry facial expressions that had undergone subliminal fear conditioning.100 The amygdala has greater interactions with subcortical regions, including the superior colliculus and thalamus, during subliminal conditioning compared to supraliminal conditioning.101 This latter result provides evidence for preferential engagement of the subcortical fear pathway during unconscious emotional learning in the human brain. Other structures that participate in conditioned fear acquisition include the anterior cingulate gyrus, sensory neocortex, thalamus, and regions of the prefrontal cortex.93,94,97,102,103 Trace conditioning procedures engage additional structures, including the hippocampus and dorsal frontoparietal cortices, which are hypothesized to mediate declarative and working memory processes that contribute to the processing of fear associations on this more complex task.95,97 Regions of the prefrontal cortex, including the vmPFC, contribute selectively to fear extinction, and reciprocal functional interactions between the vmPFC

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 145 Friday, March 10, 2006 2:56 PM

Fear and Anxiety Pathways

145 ∗

Normalized SCR (sqr uS)

0.6 0.5 0.4 0.3 0.2 0.1 0

Habituation Acquisition Extinction Recovery test Re-extinction Early Late Early Late Early Late Pre Post Early Late Different context - ctrl

Same context - ctrl

Amnesic

FIGURE 7.7 Reinstatement of extinguished fears is context- and hippocampal-dependent in humans. Mean (±S.E.M.) conditioned responses by group and experimental phase are depicted. All participants underwent habituation, acquisition, and extinction of a visual CS–noise US association in the same environmental context. Following extinction, participants were given several presentations of the US alone, and recovery of fear responses to the CS was then assessed. US-alone presentations took place either in the same context or in a different context at the time point indicated by the vertical dashed line. Recovery of extinguished fear (between vertical solid lines) occurred only for control participants who underwent reinstatement in the same environmental context (black bars, *p < .05). Mean data from two amnesic patients with hippocampal lesions are indicated by the gray bars. Amnesics acquired and extinguished fear normally during the initial training session, but they did not show fear recovery following reinstatement, despite being tested in the same environmental context. These results indicate a selective role for the hippocampus in the contextual recovery of latent fear associations. Ctrl = control, Post = first postreinstatement trial, Pre = last reinstatement trial, SCR = skin conductance response, sqr = square-root transformed, uS = microSiemens. (Adapted from LaBar, K.S. and Phelps, E.A., Reinstatement of conditioned fear in humans is context-dependent and impaired in amnesia, Behav. Neurosci., 119, 677, 2005. With permission.)

and amygdala have been identified during extinction training and during presentations of previously fear-conditioned stimuli.98,101 In sum, the neuroimaging data have led to novel insights regarding regional specificity, regional interactions, and timing of responding in the relevant components of the conditioned fear network in the human brain.

FEAR CONDITIONING

IN

ANXIETY DISORDERS

Studies of fear conditioning provide an important foundation for characterizing fear dysregulation in anxiety disorders, including panic disorder, posttraumatic stress disorder, generalized anxiety disorder, social phobia, and specific phobias. Although fear conditioning models do not account for all aspects of psychopathology, they are especially relevant for understanding unconscious aspects of fear learning as well as the influence of executive control processes over acquired fear representations, including their generalization, extinction, context specificity, and recovery

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 146 Friday, March 10, 2006 2:56 PM

146

Understanding Autism: From Basic Neuroscience to Treatment

30

Extinction

% Active voxels

Acquisition

20

10

0 0

2 4 Normalized difference GSR

6

FIGURE 7.8 (A color version of this figure follows page 236.) Amygdala activation and skin conductance responses in healthy adults during acquisition and extinction of conditioned fear. Participants acquired and extinguished a visual CS–shock US association while undergoing functional magnetic resonance imaging. Left panel: Group-averaged hemodynamic responses elicited by the CS are displayed in a coronal image through the amygdala (green box). Data are averaged from the first half of trials for each phase of training (acquisition, extinction). Right panel: The extent of amygdala activation elicited during acquisition was correlated with the magnitude of skin conductance responses in a subset of participants. GSR = galvanic skin response. (Adapted from LaBar, K.S. et al., Human amygdala activation during conditioned fear acquisition and extinction: a mixed trial fMRI study, Neuron, 20, 937, 1998. With permission.)

over time.82,104–107 Even in circumstances for which fear conditioning models do not account for the genesis or maintenance of the disorder, fear conditioning tasks are useful as experimental probes during brain imaging and psychophysiological measurements to establish the integrity of the relevant frontolimbic pathways and their response to treatment. Anxiety disorder patients typically have exaggerated conditioned fear responses and generalize their fear to safety signals or safe contexts.108–114 Exposure therapy is an efficacious behavioral treatment of anxiety disorders, especially specific phobias, and the mechanisms underlying fear extinction via behavioral desensitization are being understood in terms of reciprocal vmPFC–amygdala interactions. Nonetheless, fears can recover following exposure therapy, especially when patients are exposed to contextual cues outside the therapeutic setting, and the context specificity of fear relapse is being understood in terms of the interactions among the hippocampus, vmPFC, and amygdala.53 Pharmacological and genetic studies in nonhuman animals are pointing to cellular and molecular processes that can provide novel insights into the efficacy of drug therapies.115 For example, Davis and colleagues have recently shown that acute administration of an NMDA agonist in conjunction with exposure therapy facilitates extinction learning and reduces fear

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 147 Friday, March 10, 2006 2:56 PM

Fear and Anxiety Pathways

147

symptomatology in acrophobics for up to 3 months after treatment.116 In this way, fear conditioning research has provided an exemplary model of a translational approach to neuroscience that bridges basic knowledge of emotion-processing pathways from nonhuman animals to investigations of the human brain and its psychopathology in affective disorders.

FEAR AND ANXIETY PATHWAYS: IMPLICATIONS FOR AUTISM RESEARCH The amygdala and related structures implicated in fear and anxiety functions also contribute to socioemotional disturbances associated with autistic spectrum disorders.117 Although we are unaware of any fear conditioning studies in autism, structural MRI studies have revealed abnormalities in amygdala and hippocampal volume. Autistic children have enlarged amygdalae and hippocampi compared to normally developing children, even when adjusting for presence of mental retardation and total cerebral volume, although the group differences tend to diminish by adolescence or adulthood.118–121 Anxiety symptoms are often comorbid with autistic spectrum disorders, and enlarged amygdala volume is also observed in children with generalized anxiety disorder.122,123 Amygdala lesions in monkeys and humans produce socioemotional changes in common with autism, albeit without full-blown autistic symptomatology.117,124 For instance, infant monkeys with neonatal lesions of the medial temporal lobe, including the amygdala, exhibit impaired social interactions in dyadic relationships, including reduced social contact, averted eye gaze, and avoidant behavior.125 In addition, autistics perform similarly to amygdala-lesioned patients on tasks involving visual scanning of faces and social and emotional judgments of facial features.126–129 Of particular interest is that the processing of fear in facial expressions depends heavily on analysis of eye gaze, and both amygdala-lesioned patients and autistics do not spontaneously scan the eye region for facial expressions of fear or other emotions.130,131 Functional MRI studies have shown reduced activation of the amygdala, fusiform gyrus, superior temporal sulcus, and other frontolimbic structures during the perception of faces and facial expressions in autism.132,133 Davidson and colleagues recently reported a correlation between the degree of amygdala activation in autistic individuals and the amount of eye gaze they directed at emotionally expressive and neutral faces.134 These researchers postulated that averted gaze in autism reflects a coping strategy to reduce arousal and anxiety associated with social contact, which leads to decreased expertise in evaluating socioemotional signals from faces and a concomitant reduction in brain activation of the relevant faceprocessing regions of the brain.

CONCLUSIONS In conclusion, characterizing the circuitry for fear and anxiety has relevance for understanding psychological and neural mechanisms underlying emotional perception and social communication deficits in autism. Furthermore, lessons learned from fear

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 148 Friday, March 10, 2006 2:56 PM

148

Understanding Autism: From Basic Neuroscience to Treatment

conditioning research over the past 30+ years can inform studies of autism in ways that go beyond characterization of frontolimbic dysfunction. Much progress has been made in the neurobiology of fear through the application of complementary methodologies and common behavioral paradigms across species. To have maximal impact, translational research into higher mental functions depends on analysis at multiple levels, from genetic and molecular markers to descriptions of pharmacological and neural systems and, ultimately, to the level of behavior in healthy humans and that exhibited in clinical disorders. The analysis of fear through classical conditioning procedures illustrates how complex psychological concepts can be broken down into component parts for which the neurobiology is specified with greater precision. In this regard, the functional anatomy of fear has been fostered by the systematic investigation of processes involved in the acquisition, extinction, and contextual modulation of conditioned fears. As scientists enter the postgenomic era, more intricate questions can be answered regarding the relationship among genes, brain, and behavior. New technologies, however, are only as useful as the experimental hypotheses that are posed. Studies of socioemotional processing in autism would benefit by applying the same scientific rigor as that exemplified by the fear conditioning research reviewed herein. By adopting a structured and comprehensive interdisciplinary program, researchers can begin to unravel the biological mysteries surrounding the most interesting and perplexing aspects of neurodevelopmental disorders.

REFERENCES 1. Kessler, R.C. et al., Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States: results from the National Comorbidity survey, Arch. Gen. Psychiatry, 51, 8, 1994. 2. Rescorla, R.A., Pavlovian conditioning: it’s not what you think it is, Am. Psychol., 43, 151, 1988. 3. Turkkan, J.S., Classical conditioning: the new hegemony, Behav. Brain Sci., 12, 121, 1989. 4. Fanselow, M.S. and LeDoux, J.E., Why we think plasticity underlying Pavlovian fear conditioning occurs in the basolateral amygdala, Neuron, 23, 229, 1999. 5. Kapp, B.S. et al., Amygdala central nucleus lesions: effect on heart rate conditioning in the rabbit, Physiol. Behav., 23, 1109, 1979. 6. Romanski, L.M. and LeDoux, J.E., Equipotentiality of thalamo-amygdala and thalamo-cortico-amygdala circuits in auditory fear conditioning, J. Neurosci., 12, 4501, 1992. 7. Li, X.F. et al., Convergent but temporally separated inputs to lateral amygdala neurons from the auditory thalamus and auditory cortex use different postsynaptic receptors: in vivo intracellular and extracellular recordings in fear conditioning pathways, Learn. Mem., 3, 229, 1996. 8. Romanski, L.M. et al., Somatosensory and auditory convergence in the lateral nucleus of the amygdala, Behav. Neurosci., 107, 444, 1993. 9. Pitkänen, A. et al., Intrinsic connections of the rat amygdaloid complex: projections originating in the lateral nucleus, J. Compar. Neurol., 356, 288, 1995. 10. Killcross, S. et al., Different types of fear-conditioned behaviour mediated by separate nuclei within amygdala, Nature, 388, 377, 1997.

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 149 Friday, March 10, 2006 2:56 PM

Fear and Anxiety Pathways

149

11. Amorapanth, P. et al., Different lateral amygdala outputs mediate reactions and actions elicited by a fear-arousing stimulus, Nat. Neurosci., 3, 74, 2000. 12. LeDoux, J.E. et al., Different projections of the central amygdaloid nucleus mediate autonomic and behavioural correlates of conditioned fear, J. Neurosci., 8, 2517, 1988. 13. Nader, K. et al., Damage to the lateral and central, but not other amygdaloid nuclei prevents the acquisition of auditory fear conditioning, Learn. Mem., 8, 156, 2001. 14. Quirk, G.J. et al., Fear conditioning enhances short-latency auditory responses of lateral amygdala neurons: parallel recordings in the freely behaving rat, Neuron, 15, 1029, 1995. 15. Quirk, G.J. et al., Fear conditioning enhances different temporal components of toneevoked spike trains in auditory cortex and lateral amygdala, Neuron, 19, 613, 1997. 16. Maren, S., Auditory fear conditioning increases CS-elicited spike firing in lateral amygdala neurons even after extensive overtraining, Eur. J. Neurosci., 12, 4047, 2000. 17. Pascoe, J.P. and Kapp, B.S., Electrophysiological characteristics of amygdaloid central nucleus neurons during Pavlovian fear conditioning in the rabbit, Behav. Brain Res., 16, 117, 1985. 18. Armony, J.L. et al., Differential effects of amygdala lesions on early and late plastic components of auditory cortex spike trains during fear conditioning, J. Neurosci., 18, 2592, 1998. 19. Maren, S., Yap, S.A., and Goosens, K.A., The amygdala is essential for the development of neuronal plasticity in the medial geniculate nucleus during auditory fear conditioning in rats, J. Neurosci., 21, RC135, 2001. 20. Malenka, R.C. and Nicoll, R.A., Long-term potentiation-a decade of progress?, Science, 285, 1870, 1999. 21. Chapman, P.F. et al., Long-term synaptic potentiation in the amygdala, Synapse, 6, 271, 1990. 22. Gean, P. et al., Long-term enhancement of EPSP and NMDA receptor-mediated synaptic transmission in the amygdala, Brain Res. Bull., 31, 7, 1993. 23. Maren, S. and Fanselow, M.S., Synaptic plasticity in the basolateral amygdala induced by hippocampal formation stimulation in vivo, J. Neurosci., 15, 7548, 1995. 24. Miserendino, M.J. et al., Blocking of acquisition but not expression of conditioned fear-potentiated startle by NMDA antagonists in the amygdala, Nature, 345, 716, 1990. 25. Chapman, P.F. and Bellavance, L.L., Induction of long-term potentiation in the basolateral amygdala does not depend on NMDA receptor activation, Synapse, 11, 310, 1992. 26. Huang, Y.-Y. and Kandel, E.R., Postsynaptic induction and PKA-dependent expression of LTP in the lateral amygdala, Neuron, 21, 169, 1998. 27. Weisskopf, M.G. et al., L-type voltage-gated calcium channels mediate NMDAindependent associative long-term potentiation at thalamic input synapses to the amygdala, J. Neurosci., 19, 10512, 1999. 28. Rogan, M.T. and LeDoux, J.E., LTP is accompanied by commensurate enhancement of auditory-evoked responses in a fear conditioning circuit, Neuron, 15, 127, 1995. 29. Rogan, M.T. et al., Fear conditioning induces associative long-term potentiation in the amygdala, Nature, 390, 604, 1997. 30. McKernan, M.G. and Shinnick-Gallagher, P., Fear conditioning induces a lasting potentiation of synaptic currents in vitro, Nature, 390, 607, 1997. 31. Tsvetkov, E. et al., Fear conditioning occludes LTP-induced presynaptic enhancement of synaptic transmission in the cortical pathway to the lateral amygdala, Neuron, 34, 289, 2002.

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 150 Friday, March 10, 2006 2:56 PM

150

Understanding Autism: From Basic Neuroscience to Treatment

32. Lamprecht, R. and LeDoux, J.E., Structural plasticity and memory, Nat. Rev. Neurosci., 5, 45, 2004. 33. Bouton, M.E., Context and behavioral processes in extinction, Learn. Mem., 11, 485, 2004. 34. Goldman-Rakic, P.S., The prefrontal landscape: implications of functional architecture for understanding human mentation and the central executive, Proc. Royal Soc. London, 1445, 1996. 35. Morgan, M.A. et al., Extinction of emotional learning: contribution of medial prefrontal cortex, Neurosci. Lett., 163, 109, 1993. 36. Morgan, M.A. and LeDoux, J.E., Differential contribution of dorsal and ventral medial prefrontal cortex to the acquisition and extinction of conditioned fear in rats, Behav. Neurosci., 109, 681, 1995. 37. Morrow, B.A. et al., The role of mesoprefrontal dopamine neurons in the acquisition and expression of conditioned fear in the rat, Neuroscience, 92, 553, 1999. 38. Lebron, K. et al., Delayed recall of fear extinction in rats with lesions of ventral medial prefrontal cortex, Learn. Mem., 11, 544, 2004. 39. Milad M.R. and Quirk, G.J., Neurons in medial prefrontal cortex signal memory for fear extinction, Nature, 420, 70, 2002. 40. Milad, M.R. et al., Electrical stimulation of medial prefrontal cortex reduces conditioned fear in a temporally specific manner, Behav. Neurosci., 118, 389, 2004. 41. Herry, C. and Garcia, R., Prefrontal cortex long-term potentiation, but not long-term depression, is associated with the maintenance of extinction of learned fear in mice, J. Neurosci., 22, 577, 2002. 42. Barrett, D. et al., Metabolic mapping of mouse brain activity after extinction of a conditioned emotional response, J. Neurosci., 23, 5740, 2003. 43. McDonald, A.J. and Augustine, J.R., Localization of GABA-like immunoreactivity in the monkey amygdala, Neuroscience, 52, 281, 1993. 44. Paré, D. and Smith, Y., The intercalated cell masses project to the central and medial nuclei of the amygdala in cats, Neuroscience, 57, 1077, 1993. 45. Rosenkranz, J.A. and Grace, A.A., Cellular mechanisms of infralimbic and prelimbic prefrontal cortical inhibition and dopaminergic modulation of basolateral amygdala neurons in vivo, J. Neurosci., 22, 324, 2002. 46. Quirk, G.J. et al., Stimulation of medial prefrontal cortex decreases the responsiveness of central amygdala output neurons, J. Neurosci., 23, 8800, 2003. 47. Rosenkranz, J.A. et al., The prefrontal cortex regulates lateral amygdala neuronal plasticity and responses to previously conditioned stimuli, J. Neurosci., 23, 11054, 2003. 48. Falls, W.A. et al., Extinction of fear-potentiated startle: blockade by infusion of an NMDA antagonist into the amygdala, J. Neurosci., 12, 854, 1992. 49. Lu, K.T. et al., Mitogen-activated protein kinase cascade in the basolateral nucleus of amygdala is involved in extinction of fear-potentiated startle, J. Neurosci., 21, RC162, 2001. 50. Repa, J.C. et al., Two different lateral amygdala cell populations contribute to the initiation and storage of memory, Nat. Neurosci., 4, 724, 2001. 51. LeDoux, J.E. et al., Indelibility of subcortical emotional memories, J. Cognit. Neurosci., 8, 2517, 1989. 52. Sotres-Boyen, F. et al., Emotional perseveration: an update on prefrontal-amygdala interactions in fear extinction, Learn. Mem., 11, 525, 2004. 53. Rodriguez, B.I. et al., Context-specificity of relapse: effects of therapist and environmental context on return of fear, Behav. Res. Ther., 37, 845, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 151 Friday, March 10, 2006 2:56 PM

Fear and Anxiety Pathways

151

54. Phillips, R.G. and LeDoux, J.E., Differential contribution of amygdala and hippocampus to cued and contextual fear conditioning, Behav. Neurosci., 106, 274, 1992. 55. LaBar, K.S. and LeDoux, J.E., Partial disruption of fear conditioning in rats with unilateral amygdala damage: correspondence with unilateral temporal lobectomy in humans, Behav. Neurosci., 110, 1, 1996. 56. Selden, N.R.W. et al., Complementary roles for the amygdala and hippocampus in aversive conditioning to explicit and contextual cues, Neuroscience, 42, 335, 1991. 57. Anagnostaras, S.G. et al., Hippocampus and contextual fear conditioning: recent controversies and advances, Hippocampus, 11, 8, 2001. 58. Rudy, J.W., Huff, N.C., and Matus-Amat, P., Understanding contextual fear conditioning: insights from a two-process model, Neurosci. Biobehav. Rev., 28, 675, 2004. 59. Squire, L.R., Memory and the hippocampus: a synthesis from findings with rats, monkeys, and humans, Psychol. Rev., 99, 195, 1992. 60. Lee, Y. et al., Lack of temporal gradient of retrograde amnesia following NMDAinduced lesions of the basolateral amygdala assessed with the fear-potentiated startle paradigm, Behav. Neurosci., 110, 836, 1996. 61. Maren, S. et al., Retrograde abolition of conditional fear after excitotoxic lesions in the basolateral amygdala of rats: absence of a temporal gradient, Behav. Neurosci., 110, 718, 1996. 62. Bouton, M.E. and Bolles, R.C., Role of conditioned contextual stimuli in reinstatement of extinguished fear, J. Exp. Psychol., 5, 368, 1979. 63. Wilson, A. et al., The role of the rat hippocampal system in several effects of context in extinction, Behav. Neurosci., 109, 828, 1995. 64. Frohardt, R.J. et al., The effects of neurotoxic hippocampal lesions on two effects of context after fear extinction, Behav. Neurosci., 114, 227, 2000. 65. Corcoran, K. and Maren, S., Hippocampal inactivation disrupts contextual retrieval of fear memory after extinction, J. Neurosci., 21, 1720, 2001. 66. Corcoran, K.A. and Maren, S., Factors regulating the effects of hippocampal inactivation on renewal of conditional fear after extinction, Learn. Mem., 11, 598, 2004. 67. Hobin, J.A. et al., Context-dependent neuronal activity in the lateral amygdala represents fear memories after extinction, J. Neurosci., 23, 8410, 2003. 68. Fredrikson, M. et al., Internal consistency and temporal stability of classically conditioned skin conductance responses, Biol. Psychol., 35, 153, 1993. 69. Franks, C.M., Conditioning and personality: a study of normal neurotic subjects, J. Abnorm. Soc. Psychol., 52, 143, 1956. 70. Guimaraes, F.S. et al., Characterization of a psychophysiological model of classical fear conditioning in healthy volunteers: influence of gender, instruction, personality and placebo, Psychopharmacology, 104, 231, 1991. 71. Garpenstrand, H. et al., Human fear conditioning is related to dopaminergic and serotonergic biological markers, Behav. Neurosci., 115, 358, 2001. 72. Hettema, J.M. et al., A twin study of the genetics of fear conditioning, Arch. Gen. Psychiatry, 60, 702, 2003. 73. Zorawski, M. et al., Sex, stress, and fear: individual differences in conditioned learning, Cognit. Affect. Behav. Neurosci., 5, 191, 2005. 74. LaBar, K.S. and Disterhoft, J.F., Conditioning, awareness, and the hippocampus, Hippocampus, 8, 620, 1998. 75. Clark, R.E. et al., Classical conditioning, awareness, and brain systems, Trends Cognit. Sci., 6, 524, 2002.

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 152 Friday, March 10, 2006 2:56 PM

152

Understanding Autism: From Basic Neuroscience to Treatment

76. Lovibond, P.F. and Shanks, D.R., The role of awareness in Pavlovian conditioning: empirical evidence and theoretical implications, J. Exp. Psychol. Anim. Behav. Process., 28, 3, 2002. 77. Wiens, S. and Öhman, A., Unawareness is more than a chance event: comment on Lovibond and Shanks, J. Exp. Psychol. Anim. Behav. Process., 28, 27, 2002. 78. LaBar, K.S. et al., Impact of healthy aging on awareness and fear conditioning, Behav. Neurosci., 118, 905, 2004. 79. Marks, I. and Tobena, A., Learning and unlearning fear: a clinical and evolutionary perspective, Neurosci. Biobehav. Rev., 14, 365, 1990. 80. Öhman, A. et al., Preparedness and preattentive associative learning: electrodermal conditioning to masked stimuli, J. Psychophysiol., 9, 99, 1995. 81. Öhman, A. and Soares, J.J., On the automatic nature of phobic fear: conditioned electrodermal responses to masked fear-relevant stimuli, J. Abnorm. Psychol., 102, 121, 1993. 82. Öhman, A. and Mineka, S., Fears, phobias, and preparedness: toward an evolved module of fear and fear learning, Psychol. Rev., 108, 483, 2001. 83. Bechara, A. et al., Double dissociation of conditioning and declarative knowledge relative to the amygdala and hippocampus in humans, Science, 269, 1115, 1995. 84. LaBar, K.S. et al., Impaired fear conditioning following unilateral temporal lobectomy in humans, J. Neurosci., 15, 6846, 1995. 85. Phelps, E.A. et al., Specifying the contributions of the human amygdala to emotional memory: a case study, Neurocase, 4, 527, 1998. 86. Peper, M. et al., Aversive learning in patients with unilateral lesions of the amygdala and hippocampus, Biol. Psychol., 58, 1, 2000. 87. Hamann, S., Monarch, E.S., and Goldstein, F.C., Impaired fear conditioning in Alzheimer’s disease, Neuropsychologia, 40, 1187, 2002. 88. LaBar, K.S. and Phelps, E.A., Reinstatement of conditioned fear in humans is contextdependent and impaired in amnesia, Behav. Neurosci., 119, 677, 2005. 89. Zahn, T.P. et al., Frontal lobe lesions and electrodermal activity: effects of significance, Neuropsychologia, 37, 1227, 1999. 90. Rolls, E.T. et al., Emotion-related learning in patients with social and emotional changes associated with frontal lobe damage, J. Neurol. Neurosurg. Psychiatry, 57, 1518, 1994. 91. Bechara, A. et al., Insensitivity to future consequences following damage to human prefrontal cortex, Cognition, 50, 7, 1994. 92. Davidson, R.J., Putnam, K.M., and Larson, C.L., Dysfunction in the neural circuitry of emotion regulation — a possible prelude to violence, Science, 289, 591, 2000. 93. Buchel, C. et al., Brain systems mediating aversive condtioning: an event-related fMRI study, Neuron, 20, 1998. 94. LaBar, K.S. et al., Human amygdala activation during conditioned fear acquisition and extinction: a mixed trial fMRI study, Neuron, 20, 937, 1998. 95. Büchel, C. et al., Amygdala-hippocampal involvement in human aversive trace conditioning revealed through event-related functional magnetic resonance imaging, J. Neurosci., 19, 10869, 1999. 96. Cheng, D.T. et al., Functional MRI of human amygdala activity during Pavlovian fear conditioning: stimulus processing versus response expression, Behav. Neurosci., 117, 3, 2003. 97. Knight, D.C. et al., Neural substrates mediating human delay and trace fear conditioning, J. Neurosci., 24, 218, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 153 Friday, March 10, 2006 2:56 PM

Fear and Anxiety Pathways

153

98. Phelps, E.A. et al., Extinction learning in humans: role of the amygdala and vmPFC, Neuron, 43, 897, 2004. 99. Furmark, T. et al., The amygdala and individual differences in human fear conditioning, NeuroReport, 8, 3957, 1997. 100. Morris, J.S., Öhman, A., and Dolan, R.J., Conscious and unconscious emotional learning in the human amygdala, Nature, 393, 467, 1998. 101. Morris, J.S., Öhman, A., and Dolan, R.J., A subcortical pathway to the right amygdala mediating “unseen” fear, Proc. Natl. Acad. Sci. USA, 96, 1680, 1999. 102. Morris, J.S., Büchel, C., and Dolan, R.J., Parallel neural responses in amygdala subregions and sensory cortex during implicit fear conditioning, NeuroImage, 13, 1044, 2001. 103. Fischer, H. et al., Right-sided human prefrontal brain activation during acquisition of conditioned fear, Emotion, 2, 233, 2002. 104. Bouton, M.E., Context and ambiguity in the extinction of emotional learning: implications for exposure therapy, Behav. Res. Ther., 26, 137, 1988. 105. Wolpe, J. and Rowan, V.C., Panic disorder: a product of classical conditioning, Behav. Res. Ther., 26, 441, 1988. 106. Charney, D.S. et al., Psychobiologic mechanisms of posttraumatic stress disorder, Arch. Gen. Psychiatry, 50, 294, 1993. 107. Rauch, S.L., Shin, L.M., and Wright, C.I., Neuroimaging studies of amygdala function in anxiety disorders, Ann. N.Y. Acad. Sci., 985, 389, 2003. 108. Welch, L. and Kubis, J., Conditioned PGR (psychogalvanic response) in states of pathological anxiety, J. Nerv. Ment. Dis., 105, 372, 1947. 109. Bitterman, M.E. and Holtzman, W.H., Conditioning and extinction of the galvanic skin response as a function of anxiety, J. Abnorm. Soc. Psychol., 47, 615, 1952. 110. Pitman, R.K. and Orr, S.P., Test of the conditioning model of neurosis: differential aversive conditioning of angry and neutral facial expressions in anxiety disorder patients, J. Abnorm. Psychol., 95, 208, 1986. 111. Ashcroft, K.R. et al., Evaluation of a psychophysiological model of classical fear conditioning in anxious patients, Psychopharmacology, 104, 215, 1991. 112. Grillon, C. et al., Baseline and fear-potentiated startle in panic disorder patients, Biol. Psychiatry, 35, 431, 1994. 113. Grillon, C. and Morgan, C.A., III, Fear-potentiated startle conditioning to explicit and contextual cues in Gulf war veterans with posttraumatic stress disorder, J. Abnorm. Psychol., 108, 134, 1999. 114. Orr, S.P. et al., De Novo conditioning in trauma-exposed individuals with and without posttraumatic stress disorder, J. Abnorm. Psychol., 109, 290, 2000. 115. Davis, M., Role of NMDA receptors and MAP kinase in the amygdala in extinction of fear: implications for exposure therapy, Eur. J. Neurosci., 16, 395, 2002. 116. Ressler, K.J. et al., Cognitive enhancers as adjunct to psychotherapy: use of DCycloserine in phobic individuals to facilitate extinction of fear, Arch. Gen. Psychiatry, 61, 1136, 2002. 117. Baron-Cohen, S. et al., The amygdala theory of autism, Neurosci. Biobehav. Rev., 24, 355, 2000. 118. Aylward, E.H. et al., MRI volumes of amygdala and hippocampus in non-mentally retarded autistic adolescents and adults, Neurology, 53, 2145, 1999. 119. Haznedar, M.M. et al., Limbic circuitry in patients with autism spectrum disorders studied with positron emission tomography and magnetic resonance imaging, Amer. J. Psychiatry, 157, 1994, 2000.

© 2006 by Taylor & Francis Group, LLC

2732_C007.fm Page 154 Friday, March 10, 2006 2:56 PM

154

Understanding Autism: From Basic Neuroscience to Treatment

120. Sparks, B.F. et al., Brain structural abnormalities in young children with autism spectrum disorder, Neurology, 59, 184, 2002. 121. Schumann, C.M. et al., The amygdala is enlarged in children but not adolescents with autism; the hippocampus is enlarged at all ages, J. Neurosci., 24, 6392, 2004. 122. Muris, P. et al., Comorbid anxiety symptoms in children with pervasive developmental disorders, J. Anxiety Disord., 12, 387, 1998. 123. De Bellis, M.D. et al., A pilot study of amygdala volumes in pediatric generalized anxiety disorder, Biol. Psychiatry, 48, 51, 2000. 124. Amaral, D.G. et al., The amygdala and autism: implications from non-human primate studies, Genes Brain Behav., 2, 295, 2003. 125. Bachevalier, J., Malkova, L., and Mishkin, M., Effects of selective neonatal temporal lobe lesions on socioemotional behavior in infant rhesus monkeys (Macaca mulatta), Behav. Neurosci., 115, 545, 2001. 126. Adolphs, R. et al., Impaired recognition of emotion in facial expressions following bilateral damage to the human amygdala, Nature, 372, 669, 1994. 127. Adolphs, R. et al., The human amygdala in social judgement, Nature, 393, 470, 1998. 128. Adolphs, R., Sears, L., and Piven, J., Abnormal processing of social information from faces in autism, J. Cognit. Neurosci., 13, 232, 2001. 129. Klin, A. et al., Visual fixation patterns during viewing of naturalistic social situations as predictors of social competence in individuals with autism, Arch. Gen. Psychiatry, 59, 809, 2002. 130. Pelphrey, K.A. et al., Visual scanning of faces in autism, J. Autism Dev. Disord., 32, 249, 2002. 131. Adolphs, R. et al., A mechanism for impaired fear recognition after amygdala damage, Nature, 433, 68, 2005. 132. Pierce, K. et al., Face processing occurs outside the fusiform “face area” in autism: evidence from functional MRI, Brain, 124, 2059, 2001. 133. Pelphrey, K.A. et al., Perception of dynamic changes in facial expressions of emotion in autism, J. Cognit. Neurosci., Suppl. S., 63, 2005. 134. Dalton, K.M. et al., Gaze fixation and the neural circuitry of face processing in autism, Nat. Neurosci., 8, 519, 2005.

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 155 Wednesday, February 15, 2006 3:19 PM

Networks and 8 Cerebellar Autism: An Anatomical Hypothesis Richard P. Dum and Peter L. Strick CONTENTS Introduction............................................................................................................155 Cerebellar Structural Alterations in Autism..........................................................156 Circuit Dysfunction Hypotheses ...........................................................................158 Cerebello-Limbic Circuit................................................................................158 Cerebello-Thalamo-Cortical Circuit...............................................................159 Macroarchitecture of Cerebro-Cerebellar Loops ..................................................163 What Is the Full Extent of the Cerebellar Influence over the Cerebral Cortex?......................................................................................166 Summary and Conclusions....................................................................................167 References..............................................................................................................168

INTRODUCTION Autism is characterized by a spectrum of behavioral disorders, including motor and sensory abnormalities, speech and language difficulties, deficits in emotional and social development, attention problems, and disordered cognitive processing. In this chapter we briefly review the data that suggest that there are abnormalities in cerebellar structure associated with autism. This evidence has led to suggestions that many of the deficits seen in autism are a consequence of abnormal activity in cerebellar circuits. Cerebellar connections with the limbic system and with the cerebral cortex have been posited as mediating the abnormal activity, which is then expressed as a constellation of behavioral disorders. We present our recent evidence that the output of the cerebellum projects via the thalamus to multiple motor and nonmotor areas of the cerebral cortex, including regions of the prefrontal and posterior parietal cortex. Thus, the neural substrate exists for the output of the cerebellum to influence a wide range of motor, cognitive, and visuospatial operations. As a corollary, we argue that some of the motor and behavioral deficits seen in autism may be due to the abnormal operation of cerebellar circuits with the cerebral cortex. 155

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 156 Wednesday, February 15, 2006 3:19 PM

156

Understanding Autism: From Basic Neuroscience to Treatment

Autism is a neurological syndrome that manifests itself in the first several years after birth. It is characterized by a wide variety of behavioral disorders, including motor and sensory abnormalities, speech and language difficulties, deficits in emotional and social development, attention problems, and disordered cognitive processing. The widespread nature of the neurological deficits suggests that autism is caused by biological mechanisms that influence multiple neural systems. An alternative is that autism results from abnormalities in a neural system with widespread connections. Of course, these are not mutually exclusive alternatives. In this chapter we will briefly review the data indicating that there are abnormalities in cerebellar structure that are associated with autism. This evidence has led to suggestions that many of the deficits seen in autism are a consequence of abnormal activity in cerebellar circuits. Cerebellar connections with the limbic system and with the cerebral cortex have been posited as mediating the abnormal activity, which is then expressed as a constellation of behavioral disorders. We will present our recent evidence that cerebellar output (via the thalamus) to the cerebral cortex targets not only the cortical motor areas, but also multiple regions of the prefrontal and posterior parietal cortex. Thus, the anatomical substrate exists for the output of the cerebellum to influence a wide range of motor, cognitive, and visuospatial operations. As a corollary, abnormalities in cerebellar structure and function have the potential to produce multiple motor and nonmotor deficits such as those seen in autism.

CEREBELLAR STRUCTURAL ALTERATIONS IN AUTISM One of the most consistent neuropathological observations associated with autism is a significant reduction (typically 35–50%) in the number of Purkinje cells within the cerebellar cortex (for reviews see Palmen et al., 2004; Bauman and Kemper, 2005). This decrease is widespread, being most evident in the paraflocculus, flocculus, and posterior portions of the hemispheres below the horizontal fissure (Bauman and Kemper, 1985 and1994; Bauman, 1991; Bauman and Kemper, 2005). A reduction in the number of Purkinje cells in the vermis was noted in some but not all studies (Ritvo et al., 1986; Bailey et al., 1998). Overall, reductions in the number of Purkinje cells were found in 21 of 29 brains of autistic subjects examined by eight independent laboratories (Palmen et al., 2004). However, Palmen et al. also emphasized that the frequent comorbidity of mental retardation, epilepsy, and anticonvulsant drug therapy, as well as the small number of cases examined by each laboratory, may confound these observations. Indeed, Fatemi and colleagues (2002) found no decrease in Purkinje cell numbers in a sample of five autistic individuals, although they did find a reduction in the size of Purkinje cells (24%). The alterations in cerebellar structure in autism involve the deep cerebellar nuclei as well as the cerebellar cortex (e.g., Bauman and Kemper, 1994; Bailey et al., 1998). The changes in the deep nuclei may depend on the age of the subject. Cells in the cerebellar nuclei were enlarged in brains from autistic children (9 to 13 years old), whereas they were reduced in size and number in the brains from autistic adults (>22 years old) (Bauman and Kemper, 1994 and 2005). Thus, the significance and generality of the alterations in the deep nuclei remain unclear.

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 157 Wednesday, February 15, 2006 3:19 PM

Cerebellar Networks and Autism: An Anatomical Hypothesis

157

A number of additional, but by no means universal, findings support an association between autism and structural abnormalities in the cerebellum. Using magnetic resonance (MR) imaging, Courchesne and his colleagues found a reduction in the size of the posterior vermis (lobules VI and VII) in patients with autism (Courchesne et al., 1988, Courchesne, Saitoh et al., 1994, Courchesne, Townsend et al., 1994; for review see Courchesne 1997). Hypoplasia of the posterior vermis was seen in autistic children and adults, and was present in some infants under a year old (Hashimoto et al., 1995). Hypoplasia of the posterior vermis has not been observed in all studies (e.g., Garber and Ritvo, 1992; Holttum et al., 1992; Kleiman et al., 1992; Piven et al., 1992). However, Courchesne and his colleagues (Courchesne, Saitoh et al., 1994, Courchesne, Townsend et al., 1994, 1997) have argued that a subset (~15%) of autistic subjects display hyperplasia of vermal lobules VI and VII. Thus, they reasoned that averaging subject populations with hyper- and hypoplasia masked the vermal changes in some studies. Most imaging studies have concentrated on the vermis in autistic subjects because it can be captured in a single sagittal slice in MR imaging. However, it is important to note that structural and neurochemical changes in the cerebellar cortex of autistic subjects are not limited to the vermis. Indeed, the most striking Purkinje cell loss has been observed in the cerebellar hemispheres (e.g., Palmen et al., 2004; Bauman and Kemper, 2005). Only one study has reported a correlation between the size of vermal lobules VI and VII and the size of the hemispheres in autistic subjects (Murakami et al., 1989). This issue and its relationship to changes in Purkinje cell number and size needs to be examined in a larger sample, with more complete reconstructions of the cerebellum. A number of “neurochemical markers” also are altered in the cerebellar cortex of autistic patients. Lee et al. (2002) found changes in specific nicotinic receptors in the cerebellum but no alterations in the amount of either M1 or M2 receptors or in choline acetyltransferase in the same region. Fatemi et al. (2001) observed reductions in the expression of reelin (>40%) and Bcl-2 (34–51%) in the cerebellar cortex. Reelin is an important secretory glycoprotein responsible for normal layering of the brain and Bcl-2 is a regulatory protein responsible for control of programmed cell death. The authors argued that dysregulation of reelin and Bcl-2 may cause abnormal development of cerebellar circuits that, in turn, result in some of the behavioral abnormalities observed in autism. Regardless of the variations noted in the preceding text, it should be clear that alterations in the cerebellum are present in significant numbers of autistic patients. Even so, there is considerable variation among patients in the extent of cerebellar changes. Only a subset of patients displayed a striking loss of Purkinje cells. Similarly, only a subset of subjects showed hypoplasia or hyperplasia of vermal lobules VI and VII. At present, there is no explanation for the origin of these extreme variations. Many other issues about the nature of the cerebellar pathology present in autism remain to be explored. For example, the precise distribution and age-related nature of the changes in the cerebellar cortex and in the deep cerebellar nuclei are unclear. This issue needs to be examined using a variety of neurochemical markers along with modern stereological methods. In addition, major insights could come from correlating motor, cognitive, and behavioral deficits with cerebellar alterations in

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 158 Wednesday, February 15, 2006 3:19 PM

158

Understanding Autism: From Basic Neuroscience to Treatment

individual subjects. For example, eye movement abnormalities are present in some autistic patients (Takarae, Minshew, Luna, and Sweeney, 2004; Takarae, Minshew, Luna, Krisky et al., 2004). Perhaps these subjects display alterations in vermal lobules VI and VII. These lobules have come to be termed the “oculomotor vermis” because of their involvement in the control of saccadic and smooth-pursuit eye movements (Noda and Fujikado, 1987; Barash et al., 1999; Suh et al., 2000; see Ito, 2005). Overall, clear associations between specific abnormal behaviors and sites of cerebellar abnormality remain to be identified.

CIRCUIT DYSFUNCTION HYPOTHESES The evidence we have just reviewed has led to suggestions that many of the deficits seen in autism are a consequence of abnormal activity in cerebellar circuits. In general, two models have been proposed to explain how cerebellar dysfunction could lead to the range of behavioral abnormalities seen in autism. One of these models involves dysfunction in cerebellar connections with the limbic system and the other involves dysfunction in cerebellar connections with the cerebral cortex. Basic aspects of these models will be reviewed in separate sections in the text that follows.

CEREBELLO-LIMBIC CIRCUIT There is a longstanding notion that the cerebellum interacts with the limbic system. Lee et al. (2003) recently reviewed the complex literature on this subject and proposed that autism is “the final common manifestation of altered functioning” in a cerebellolimbic circuit. They argued that lesions of the limbic system, including the amygdala, lead to abnormal behaviors in nonhuman primates that are similar to those observed in autism. They also noted that cerebellar damage in human subjects can result in cognitive and affective impairments similar to those seen in autism (Schmahmann and Sherman, 1998). In addition, some autistic subjects display classic cerebellar deficits (Takarae, Minshew, Luna, and Sweeney, 2004; Takarae, Minshew, Luna, Krisky et al., 2004). It is known that cerebellar stimulation can alter limbic function and elicit behaviors like sham rage, predatory attack, grooming, and eating (e.g., Zanchetti and Zoccolini, 1954; Berntson et al., 1973; Reis et al., 1973). Furthermore, electrophysiological evidence suggests that cerebellar stimulation, especially in portions of the fastigial nucleus and associated regions of vermal cortex, can evoke responses at limbic sites, including the cingulate cortex and amygdala (e.g., Anand et al., 1959; Snider and Maiti, 1976). Taken together, these observations, along with the structural changes in the cerebellum, are the basis of the proposal that abnormal cerebellar output to the limbic system is a cause of autistic behavior. The major weakness in this proposal is the absence of a clear anatomical substrate that links the output of the cerebellum and, especially, the fastigial nucleus with limbic sites like the amygdala. Although neuroanatomical evidence indicates that the deep cerebellar nuclei are interconnected with the hypothalamus (Haines et al., 1990), these connections do not appear to be sufficient to mediate all of the behavioral effects evoked by cerebellar stimulation. Thus, the circuits that link the output of the deep cerebellar nuclei with regions of the limbic system need to be explored using modern neuroanatomical methods. © 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 159 Wednesday, February 15, 2006 3:19 PM

Cerebellar Networks and Autism: An Anatomical Hypothesis

159

CEREBELLO-THALAMO-CORTICAL CIRCUIT It has long been known that the cerebellum is massively interconnected with the cerebral cortex. The “classical” view of these interconnections was that the cerebellum gathers information from widespread cortical areas including portions of the frontal, parietal, temporal, and occipital lobes. This information was then thought to be “funneled” through cerebellar circuits, where it ultimately converged on the ventrolateral nucleus of the thalamus (e.g., Evarts and Thach, 1969; Allen and Tsukahara, 1974). The ventrolateral nucleus was believed to project to a single cortical area, the primary motor cortex (M1). Thus, cerebellar connections with the cerebral cortex were viewed as a means of collecting information from widespread regions of the cerebral cortex and conveying it to M1 for the generation and control of movement. According to this view, abnormal activity in this circuit would lead to purely motor deficits. Recent findings have caused us to challenge this view (e.g., Schell and Strick, 1984; Hoover and Strick, 1999; Middleton and Strick, 1994, 1996, and 2001; Clower et al., 2001 and 2005; Kelly and Strick, 2003). It is now clear that efferents from the cerebellar nuclei project to multiple subdivisions of the ventrolateral thalamus (e.g., Percheron, 1977; Stanton, 1980; Asanuma et al., 1983; Ilinsky and Kultas-Ilinsky, 1987; Percheron et al., 1996) which, in turn, project to a myriad of cortical areas (e.g., Schell and Strick, 1984; Wiesendanger and Wiesendanger, 1985(a) and (b); Orioli and Strick, 1989; Matelli et al., 1989; Hoover and Strick, 1999; Rouiller et al., 1994 and 1999; Middleton and Strick, 1994 and 1996, 2001; Matelli and Luppino, 1996; Sakai et al., 1999 and 2002; Clower et al., 2001 and 2005; Dum and Strick, 2003). Thus, the outputs from the cerebellum influence more widespread regions of the cerebral cortex than previously recognized. This change in perspective is important, because it provides the anatomical substrate for the output of the cerebellum to influence motor and nonmotor areas of the cerebral cortex. As a consequence, abnormal activity in these circuits could lead not only to motor deficits but also to some of the cognitive, attentional, and affective impairments seen in autism. The next section presents some of the new observations that form the basis for this change in perspective. These results depended on the development of a new technique — the use of neurotropic viruses as transneuronal tracers. In the past, attempts to determine the cortical targets of cerebellar output were faced with a number of technical and conceptual limitations. The key difficulty was the disynaptic nature of the link between the cerebellum and the cerebral cortex. Conventional tracing methods are unable to reveal disynaptic connections or do so with great difficulty. To overcome this problem, we developed the use of herpes simplex virus type 1 [HSV1] and rabies virus as transneuronal tracers in nonhuman primates (e.g., Zemanick et al., 1991; Strick and Card, 1992; Kelly and Strick, 2000 and 2003). Virus tracing is unique as an anatomical method in its ability to define a chain of synaptically linked neurons. Selected strains of virus move transneuronally in either the retrograde or anterograde direction (Zemanick et al., 1991; Kelly and Strick, 2003). Thus, one can examine either the inputs to or the outputs from a site. The viruses we use as tracers move from neuron to neuron exclusively at synapses and the transneuronal transport occurs in a time-dependent fashion. Thus, by careful adjustment of the survival time after a virus injection, we can study neural circuits

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 160 Wednesday, February 15, 2006 3:19 PM

160

Understanding Autism: From Basic Neuroscience to Treatment

composed of two or even three synaptically connected neurons.We have used virus tracing to examine cerebello-thalamo-cortical pathways to a variety of cortical areas (Lynch et al., 1994; Middleton and Strick, 1994 and 1996; 2001; Hoover and Strick, 1999; Clower et al., 2001, 2005; Kelly and Strick, 2003). In our initial studies, we used retrograde transneuronal transport of HSV1 to examine the organization of cerebellar projections to M1 (Hoover and Strick, 1999). We found that the projections to M1 originate largely from the dentate. Therefore, we created unfolded maps of the dentate to provide a consistent frame of reference for comparing the distribution of labeled neurons that project from this nucleus to the different cortical areas (Dum and Strick, 2003). Virus injections into the arm representation of M1 labeled a compact cluster of neurons in the dorsal portion of the dentate at mid-rostrocaudal levels (Figure 8.1, middle panel). Virus injections into the leg or the face representation of M1 also labeled dense clusters of neurons in a dorsal portion of the dentate. In each case, the

M1 leg (Jo17)

M1 arm (Jo19)

M1 face (Jo18) Dorsal

7 6 5 4

2 1 0

Ventral

Distance (mm)

3

−1 −2 −4 −5

13–16 10–12 8–9 4–7

50 70 90 110 130 150 Section number

11–13 10 7–9 3–6

100 120 140 160 180 Section number

10–11 9 7–8 3–6

Caudal

Dorsal

−3

200 220 240 260 280 300 Section number

FIGURE 8.1 Organization of dentate output to M1. Unfolded maps of the dentate illustrate the distribution and density of “second-order” neurons (shaded squares) labeled after retrograde transneuronal transport of HSV1 from the leg (left panel), arm (middle panel), and face (right panel) representations of M1. These maps of the dentate were created by unfolding serial coronal sections through the nucleus (see Dum and Strick, 2003 for details of unfolding and determination of cell density). The rostro-caudal midpoint of the nucleus is indicated by a vertical dashed line. These maps were reconstructed from every other coronal section. The density key is in the lower left of each panel. (Adapted from Dum, R.P. and Strick, P.L., An unfolded map of the cerebellar dentate nucleus and its projections to the cerebral cortex, J. Neurophysiol. 89: 634–639, 2003 [first published November 6, 2002; 10.1152/jn.00626. 2002].)

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 161 Wednesday, February 15, 2006 3:19 PM

Cerebellar Networks and Autism: An Anatomical Hypothesis

161

location of labeled neurons was spatially separate from the location of neurons that projected to the arm representation of M1. Virus injections into the leg representation labeled neurons in the rostral pole of the dorsal portion of the dentate (Figure 8.1, left panel), whereas virus injections into the face representation labeled neurons at caudal levels of the dorsal dentate (Fig. 1, right panel). We have used the term “output channels” to describe the topographically distinct origin of dentate projections to different cortical areas (for example, see Middleton and Strick, 1998). The rostralto-caudal sequence of dentate output channels to the leg, arm, and face representations in M1 corresponds well with the somatotopic organization of the dentate previously proposed on the basis of physiological studies (e.g., Allen et al., 1978; Stanton, 1980; Rispal-Padel et al., 1982; Asanuma et al., 1983; Thach et al., 1993). The region of the dentate that contains neurons which project to M1 occupies only 30% of the nucleus (Hoover and Strick, 1999; Dum and Strick, 2003). This implies that a substantial portion of the dentate projects to cortical targets other than M1. To test this proposal, we injected virus into selected prefrontal and posterior parietal areas of the cortex. In every case, these injections resulted in transneuronal labeling of neurons that were most concentrated in ventral regions of the dentate (Middleton and Strick, 2001; Clower et al., 2001 and 2005). For example, virus injections into a portion of area 46 in the prefrontal cortex labeled neurons in a region of dentate that was ventral to the arm representation of M1 (Figure 8.2, center left panel). Virus injections into a portion of area 91 in the prefrontal cortex labeled neurons at even more caudal and ventral levels of the dentate (Figure 8.2, center right panel). Virus injections into a portion of area 7b in the posterior parietal cortex labeled neurons in the caudal pole of the dentate (Figure 8.2, far right panel). These and other results demonstrated that the dentate has substantial projections to nonmotor areas of the frontal and posterior parietal cortex. Furthermore, these projections originate from output channels that are located in the ventral dentate (e.g., Middleton and Strick, 1994 and 2001; Clower et al., 2001). We have summarized our findings by plotting the observations of individual experiments onto a single unfolded map of the dentate nucleus (Figure 8.3, left). On this map, we marked the average location of the output channel to each cortical area examined. The map includes the location of the output channel that projects to the arm representation in the ventral premotor area (PMv)(Middleton and Strick, 1997). This summary diagram emphasizes that a sizeable portion of the dentate projects to parts of prefrontal and posterior parietal cortex, in addition to its classical target, M1. Furthermore, the dentate map reveals an important feature of the topographic organization of the dentate output. The output channels to prefrontal and posterior parietal areas of cortex are spatially segregated from those that target motor areas of the cortex. Thus, the dentate appears to be anatomically divided into separate domains that target motor and nonmotor areas of the cerebral cortex. The division of the dentate into separate motor and nonmotor domains is reinforced by an underlying molecular gradient. We recently discovered a monoclonal antibody, “8B3,” that differentially stains regions of the dentate in monkeys (Figure 8.3, right) (Pimenta et al., 2001; Dum et al., 2002). Immunoreactivity for 8B3 is most intense in ventral regions of the dentate which project to the prefrontal and posterior parietal areas of cortex. Antibody staining is lowest in the dorsal two thirds of the

© 2006 by Taylor & Francis Group, LLC

Area 9L (F11L)

Area 7b (W21) Dorsal

7 6 5 4

1 Ventral

Distance (mm)

2

0 −1 −2

−4 −5

14–17 11–13 9–10 2–8

100 120 140 160 180 Section number

7–12 6 4–5 2–3

13–16 10–12 7–9 2–6

100 120 140 160 180 Section number

100 120 140 160 180 Section number

4–7 3 2 1

Caudal

Dorsal

−3

90 110 130 150 170 190 Section number

FIGURE 8.2 Dentate output to the prefrontal and posterior parietal cortex. Unfolded maps of the dentate illustrate the distribution and density of neurons labeled after HSV1 injections into the arm representation of M1, area 46, area 9l, and area 7b. The M1 arm case shown is different from the one illustrated in Figure 8.1. Conventions and abbreviations are as in Figure 8.1. (Adapted from Dum, R.P. and Strick, P.L., An unfolded map of the cerebellar dentate nucleus and its projections to the cerebral cortex, J. Neurophysiol. 89: 634–639, 2003 [first published November 6, 2002; 10.1152/jn.00626.2002].) © 2006 by Taylor & Francis Group, LLC

Understanding Autism: From Basic Neuroscience to Treatment

3

2732_C008.fm Page 162 Wednesday, February 15, 2006 3:19 PM

Area 46 (F11R)

162

M1 arm (Z10)

2732_C008.fm Page 163 Wednesday, February 15, 2006 3:19 PM

Cerebellar Networks and Autism: An Anatomical Hypothesis 8B3 staining

Dorsal

Dentate

163

M1 face M1 arm “Motor” PMv M1 leg

Ventral

7b 9L 46d

D

1 mm

C

>1.5 1–1.5 0.5–1 <0.5

Dorsal

“Nonmotor”

FIGURE 8.3 Functional topography within the dentate. Left panel: Dentate output channels. Labels mark the location of dentate neurons that project to each cortical area studied. The fine dashed line marks the border between the motor and non-motor domains in the dentate (see text and right panel for further explanation). (Dum, R.P. and Strick, P.L., An unfolded map of the cerebellar dentate nucleus and its projections to the cerebral cortex. J. Neurophysiol. 89: 634–639, 2003 [first published November 6, 2002; 10.1152].) Right panel: Distribution of immunostaining for antibody 8B3. The intensity of immunostaining was contoured into four grayscale levels based on standard deviations from the mean intensity of labeling (see key). PMv, ventral premotor area. (Dum, R.P., Li, C., and Strick, P.L., Motor and nonmotor domains in the monkey dentate. Ann N Y Acad Sci. 978: 289–301, 2002. Copyright 2002 by New York Academy of Sciences.)

dentate which project to the cortical motor areas. These observations suggest that 8B3 recognizes the nonmotor domain within the dentate (Figure 8.3). Similarly, immunostaining for two calcium-binding proteins, calretinin and parvalbumin, is reported to be greatest in the ventral regions of the squirrel monkey dentate (Fortin et al., 1998). These molecular markers raise the possibility that it will be possible to identify anatomically homologous domains in the human dentate.

MACROARCHITECTURE OF CEREBRO-CEREBELLAR LOOPS By prolonging the survival time, we were able to map the location of third-order neurons in the cerebellar cortex (Purkinje cells) that were labeled by retrograde transneuronal transport of rabies virus from injection sites in the cerebral cortex

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 164 Wednesday, February 15, 2006 3:19 PM

164

Understanding Autism: From Basic Neuroscience to Treatment

M1

Area 46 I

I

II II

III

III IV IV

V

V

p

F.pr. VI

Crus I

a

F.in.cr.

VI

Crus II

F.ps.

VII

a

VII VIII

HVIIB

S.int.cr.2

F.ppd.

p VIII

VIIB

F.apm.

IX HVIII

IX

X X

FIGURE 8.4 Regions of the cerebellar cortex that target M1 and area 46. Left: The distribution of Purkinje cells (black dots) that were labeled after retrograde transneuronal transport of rabies from injections into the arm area of M1. Right: The distribution of Purkinje cells that were labeled after retrograde transneuronal transport of rabies from injections into dorsal area 46. The maps represent surface reconstructions of the cerebellar cortex (see Kelly and Strick, 2003, for details). Nomenclature and abbreviations are according to Larsell (1970). (Adapted from Kelly, R.M. and Strick, P.L., Cerebellar loops with motor cortex and prefrontal cortex of a nonhuman primate, J. Neurosci. 23: 8832–8444, 2003. © The Society for Neuroscience.)

(Kelly and Strick, 2003). Using this approach, we compared the location of Purkinje cells that project to the arm area of M1 with those that project to area 46 in the prefrontal cortex. Injections of rabies into the arm representation of M1 labeled Purkinje cells mainly in the hemispheric portions of lobules IV–VI (Figure 8.4, left panel). These injections also labeled a second smaller group of Purkinje cells more caudally in lobules HVIIB–HVIII. In contrast, injections of rabies into area 46 labeled Purkinje cells mainly in crus II of the ansiform lobule (Figure 8.4, right panel). A few labeled Purkinje cells also were found in vermal lobules VII and X. Thus, both M1 and area 46 are the targets of output from the cerebellar cortex.

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 165 Wednesday, February 15, 2006 3:19 PM

Cerebellar Networks and Autism: An Anatomical Hypothesis

M1

165

46

THM1

PNM1

TH46

PN46

DNM1

CBM1

DN46

CB46

FIGURE 8.5 Separate closed-loop circuits link the cerebellum with M1 and area 46. Note that each area of cerebral cortex provides the major input to a separate region of the cerebellar cortex and dentate. This region of the cerebellar cortex and dentate then project back upon the same area of the cerebral cortex. CB, cerebellar cortex; DN, dentate nucleus; PN, pontine nuclei; TH, subdivisions of the thalamus. (Adapted from Kelly, R.M. and Strick, P.L., Cerebellar loops with motor cortex and prefrontal cortex of a nonhuman primate, J. Neurosci. 23: 8832–8444, 2003. © The Society for Neuroscience.)

However, the output to each area of the cerebral cortex originates from Purkinje cells in very different regions of the cerebellar cortex. We saw no evidence of overlap between the two systems. In separate experiments we used anterograde transneuronal transport of the H129 strain of HSV1 to determine the distribution of granule cells in the cerebellar cortex that receive input from M1 or from area 46 (Zemanick et al., 1991; Kelly and Strick, 2003). We found that neurons in the arm area of M1 project via the pons to granule cells located mainly in lobules IV–VI, whereas neurons in area 46 project to granule cells mainly in crus II. These findings indicate that the regions of the cerebellar cortex which receive input from M1 are the same as those that project to M1 (Figure 8.5, left). Similarly, the regions of the cerebellar cortex which receive input from area 46 are the same as those that project to area 46 (Figure 8.5, right). Thus, our observations suggest that multiple closed-loop circuits represent a fundamental architectural feature of cerebro-cerebellar interactions. Furthermore, the separation of motor and nonmotor functions seen in the dentate extends to the level of the cerebellar cortex. These results have a number of important functional implications. Even though the internal wiring diagram of the cerebellar cortex does not vary from lobule to lobule, the cerebellar cortex is not functionally homogeneous. Instead, it contains localized regions that are interconnected with specific motor and nonmotor areas of the cerebral cortex. In fact, we have argued that the map of function in the cerebellar cortex is likely to be as rich and complex as that in the cerebral cortex (Kelly and Strick, 2003). As a consequence, global dysfunction of the cerebellar cortex can cause wide-ranging effects on behavior (e.g., Schmahmann, 2004). On the other hand, localized dysfunction of a portion of the cerebellar

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 166 Wednesday, February 15, 2006 3:19 PM

166

Understanding Autism: From Basic Neuroscience to Treatment

cortex can lead to more limited deficits and have the potential to be motor or nonmotor, depending on the specific site of the abnormality (e.g., Fiez et al., 1992; Schmahmann and Sherman, 1998; Allen and Courchesne, 2003; Gottwald et al., 2004). Thus, precisely defining the location of a lesion, site of activation, or recording site is as important for studies of the cerebellum as it is for studies of the cerebral cortex.

WHAT IS THE FULL EXTENT OF THE CEREBELLAR INFLUENCE OVER THE CEREBRAL CORTEX? Our results clearly demonstrate that a number of motor and nonmotor areas are the targets of output from the dentate. However, substantial portions of the dentate did not contain labeled neurons following virus injections into any of the cortical areas we examined. Thus, we have yet to identify all of the cortical targets of dentate output. In addition, fastigial and interpositus nuclei send efferents to the thalamus (Batton et al., 1977; Stanton, 1980; Asanuma et al., 1983; Kalil et al., 1981), but our studies have not identified the cortical targets for most of these efferents. As a consequence, the full extent of the cerebellar influence over the cerebral cortex remains to be determined. We have argued elsewhere that our observations allow us to make some predictions about the additional cortical targets of cerebellar output (Middleton and Strick, 1998; Kelly and Strick, 2003; Dum and Strick, 2003). The dentate output channels that we have identified appear to be part of closed-loop, cerebro-cerebellar circuits. In essence, cortical areas that are the target of dentate output also appear to project back upon the cerebellar cortex via the cortico-pontine system (Figure. 8.6)(e.g., Brodal, 1978; Wiesendanger et al., 1979; Vilensky and Van Hoesen, 1981; Leichnetz et al., 1984; Glickstein et al., 1985; Schmahmann and Pandya, 1991, 1993, and 1997). We have shown this explicitly for M1 and for dorsal portions of area 46 (see preceding text and Kelly and Strick, 2003). In contrast, cortical areas that do not receive projections from the dentate (e.g., the ventral portion of area 46, the lateral portion of area 12, and area TE in the inferotemporal cortex)(Middleton and Strick, 1998 and 2001) do not appear to project back upon the cerebellar cortex. If these results reflect a general rule, then all of the cortical areas that project to the cerebellum are the targets of cerebellar output. In addition to the areas we have already studied, the cerebellum receives input from a wide variety of higher-order nonmotor areas. These cortical regions include regions of extrastriate cortex, large portions of the posterior parietal cortex on the lateral surface of the hemisphere (Figure 8.6, bottom-shaded region), and large portions of the cingulate cortex and the parahippocampal gyrus on the medial surface of the hemisphere (Figure 8.6, top-shaded region)(Brodal, 1978; Wiesendanger et al., 1979; Vilensky and Van Hoesen, 1981; Leichnetz et al., 1984; Glickstein et al., 1985; Schmahmann and Pandya, 1991, 1993, and 1997). In future studies we will examine whether each of these cortical areas is the target of cerebellar output. If some or all of these areas turn out to be cerebellar targets, then the full extent of cerebellar influence over the cerebral cortex is much larger than previously suspected.

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 167 Wednesday, February 15, 2006 3:19 PM

Cerebellar Networks and Autism: An Anatomical Hypothesis

preSMA

M1

SMA

9m

167

CgS

M1 leg prePMd

12l

IP

PMv

AIP

M1 face

7b

Ar S

PS

S

M1 arm

FEF 46d 46v

CS

PMd

9l

LS

STS TE 1 cm

FIGURE 8.6 Cortical targets of cerebello-thalamocortical circuits. Cortical areas that are the target of cerebellar output all fall within the region that projects to the cerebellum via the pons (shading). In contrast, cortical regions that are not the target of cerebellar output (46v, 121, TE) all fall outside the region that projects to the cerebellum (see text for references and reviews). AIP, anterior intraparietal area; ArS, arcuate sulcus; CgS, cingulate sulcus; CS, central sulcus; FEF, frontal eye fields; IPS, intraparietal sulcus; LS, lateral sulcus; PMd, dorsal premotor area; PMv, ventral premotor area; prePMd, pre-dorsal premotor area; preSMA, presupplementary motor area; PS, principal sulcus; SMA, supplementary motor area; STS, superior temporal sulcus. (Adapted from Dum, R.P. and Strick, P.L., An unfolded map of the cerebellar dentate nucleus and its projections to the cerebral cortex. J. Neurophysiol. 89: 634–639, 2003 [first published November 6, 2002; 10.1152].)

SUMMARY AND CONCLUSIONS The dominant view of cerebellar function over the last century has been that it is concerned with the coordination and control of motor activity (Brooks and Thach, 1981). Consistent with that view, we found that multiple traditional motor areas in

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 168 Wednesday, February 15, 2006 3:19 PM

168

Understanding Autism: From Basic Neuroscience to Treatment

the cerebral cortex are the targets of dentate output. Nevertheless, we also found that a significant portion of the output of the dentate projects to nonmotor areas of the cerebral cortex, including regions of the prefrontal and posterior parietal cortex (Middleton and Strick, 1998; Hoover and Strick, 1999; Middleton and Strick, 2001; Clower et al., 2001 and 2005). Thus, the anatomical substrate exists for cerebellar output to influence the cognitive and visuospatial computations performed in the prefrontal and posterior parietal cortex (Middleton and Strick, 2001; Clower et al., 2001 and 2005). As a corollary, abnormalities in cerebellar structure and function have the potential to produce multiple motor and nonmotor deficits, such as those seen in autism. The output to nonmotor areas of the cerebral cortex originates specifically from a ventral portion of the dentate. This nonmotor region of the dentate is “recognized” by several molecular markers. There is evidence from human and monkey studies that ventral portions of the dentate and regions of the cerebellar cortex are activated during tasks that involve short-term working memory, rulebased learning, and other higher executive functions (Mushiake and Strick, 1993; Kim et al., 1994; Jueptner et al., 1997; Liu et al., 2000; Chen and Desmond, 2005). Several authors have argued that ventral dentate and related regions of the cerebellar hemispheres are selectively enlarged in great apes and humans (Leiner et al., 1991; Matano, 2001). Indeed, the enlargement of the ventral dentate in humans is thought to parallel the enlargement of the prefrontal the cortex. These observations have led to the proposal that the dentate’s participation in nonmotor functions may be especially prominent in humans (e.g., Leiner et al., 1991; Schmahmann and Sherman, 1998). In summary, we have reviewed the concept that autistic behaviors could result from pathology in a single neural system with widespread connections. The cerebellum, and especially the cerebellar cortex, is the most consistent site of pathology in autism. The anatomical connections of the cerebellum that we have just reviewed fit the description of a system with widespread connections. Thus, cerebellar pathology could lead to abnormal activity in cerebello-limbic and cerebello-thalamocortical pathways, which could then be expressed as autistic behavior. Future tests of this anatomical hypothesis will require further exploration of the link between the cerebellar nuclei and regions of the limbic system, along with additional studies that define the full extent of cerebellar projections to nonmotor areas of the cerebral cortex. Once this anatomical substrate has been fully defined, a test of this cerebellar hypothesis will be to determine whether abnormal activity in these circuits results in autistic-like behavior.

REFERENCES Akkal, D., Dum, R.P. and Strick, P.L., Cerebellar and pallidal inputs to the supplementary motor area (SMA), Soc. Neurosci. Abstr. 27: 825.4, 2001. Akkal, D., Dum, R.P., and Strick, P.L., Cerebellar and basal ganglia inputs to the presupplementary motor area (PreSMA), Soc. Neurosci. Abstr. 28: 462.14, 2002.

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 169 Wednesday, February 15, 2006 3:19 PM

Cerebellar Networks and Autism: An Anatomical Hypothesis

169

Allen, G.I., Gilbert, P.F., and Yin, T.C., Convergence of cerebral inputs onto dentate neurons in monkey, Exp. Brain Res. 32: 151–170, 1978. Allen, G. and Courchesne, E., Differential effects of developmental cerebellar abnormality on cognitive and motor functions in the cerebellum: an fMRI study of autism, Am. J. Psychiatry 160: 262–273, 2003. Allen, G.I. and Tsukahara, N., Cerebrocerebellar communication systems, Physiol. Rev. 54: 957–1006, 1974. Anand, B.K., Malhotra, C.L., Singh, B., and Dua, S., Cerebellar projections to limbic system, J. Neurophysiol. 22: 451–457, 1959. Asanuma, C., Thach, W.T., and Jones, E.G., Distribution of cerebellar terminations and their relation to other afferent terminations in the ventral lateral thalamic region of the monkey, Brain Res. 286: 237–265, 1983. Bailey, A., Luthert, P., Dean, A., Harding, B., Janota, I., Montgomery, M., Rutter, M., and Lantos, P., A clinicopathological study of autism, Brain 121: 889–905, 1998. Barash, S., Melikyan, A., Sivakov, A., Zhang, M., Glickstein, M., and Thier, P., Saccadic dysmetria and adaptation after lesions of the cerebellar cortex, J. Neurosci. 19: 10931–10939, 1999. Batton, R.R., Jayaraman, A.I., Ruggiero, D., and Carpenter, M.B., Fastigial efferent projections in the monkey: an autoradiographic study, J. Comp. Neurol. 174: 281–306, 1977. Bauman, M.L., Microscopic neuroanatomical abnormalities in autism, Pediatrics 87: 791–796, 1991. Bauman, M. and Kemper, T.L., Histoanatomic observations of the brain in early infantile autism, Neurology 35: 866–874, 1985. Bauman, M.L. and Kemper, T.L., Neuroanatomic observation of the brain in autism, in The Neurobiology of Autism, Baltimore, MD: Johns Hopkins University Press, 1994, pp. 119–145, Bauman, M.L. and Kemper, T.L., Neuroanatomic observations of the brain in autism: a review and future directions, Int. J. Dev. Neurosci. 23: 183–187, 2005. Berntson, G.G., Potolicchio, S.J. Jr., and Miller, N.E., Evidence for higher functions of the cerebellum: eating and grooming elicited by cerebellar stimulation in cats, Proc. Natl. Acad. Sci. U S A. 70: 2497–2499, 1973. Brodal, P., The corticopontine projection in the rhesus monkey. Origin and principles of organization, Brain 101: 251–283, 1978. Brooks, V.B. and Thach, W.T., Cerebellar control of posture and movement, in Handbook of Physiology. Section I: The Nervous System, Vol. II: Motor Control, Part I, Ed., V. B. Brooks. Bethesda, MD: American Physiological Society, 1981, pp. 877–946. Chen, S.H. and Desmond, J.E., Temporal dynamics of cerebro-cerebellar network recruitment during a cognitive task, Neuropsychologia 43(9): 1227–1237, 2005. (Epub February 25, 2005.) Clower, D.M., West, R.A., Lynch, J.C., and Strick, P.L., The inferior parietal lobule is the target of output from the superior colliculus, hippocampus and cerebellum, J. Neurosci. 21: 6283–6291, 2001. Clower, D.M., Dum, R.P., and Strick, P.L., Basal ganglia and cerebellar Inputs to ‘AIP,’ Cerebral Cortex, 15: 913–920, 2005. doi:10.1093/cercor/bhh190, Advance Access published on September 30, 2004. Courchesne, E., Brainstem, cerebellar and limbic neuroanatomical abnormalities in autism, Curr. Opin. Neurobiol. 7: 269–278, 1997. Courchesne, E, Yeung-Courchesne, R., Press, G.A., Hesselink, J.R., Jernigan, T.L., Hypoplasia of cerebellar vermal lobules VI and VII in autism, N. Engl. J. Med. 318: 1349–1354, 1988.

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 170 Wednesday, February 15, 2006 3:19 PM

170

Understanding Autism: From Basic Neuroscience to Treatment

Courchesne, E., Saitoh, O., Yeung-Courchesne, R., Press, G.A., Lincoln, A.J., Haas, R.H., and Schreibman, L., Abnormality of cerebellar vermian lobules VI and VII in patients with infantile autism: identification of hypoplastic and hyperplastic subgroups with MR imaging, Am. J. Roentgenol. 162: 123–130, 1994. Courchesne, E., Townsend, J., and Saitoh, O., The brain in infantile autism: posterior fossa structures are abnormal, Neurology 44: 214–223, 1994. Dum, R.P. and Strick, P.L., Pallidal and cerebellar inputs to the digit representations of the dorsal and ventral premotor areas (PMd and PMv), Soc. Neurosci. Abstr. 25: 1925, 1999. Dum, R.P. and Strick, P.L., An unfolded map of the cerebellar dentate nucleus and its projections to the cerebral cortex, J. Neurophysiol. 89: 634–639, 2003 [first published November 6, 2002; 10.1152/jn.00626.2002]. Dum, R.P., Li, C., and Strick, P.L., Motor and nonmotor domains in the monkey dentate, Ann. N.Y. Acad. Sci. 978: 289–301, 2002. Dum, R.P., Akkal, D. and Strick, P.L., The dorsal premotor area (PMd) and the Pre-PMd area are the target of basal ganglia and cerebellar output in the cebus monkey, Program No. 535.12. 2004. Abstract Viewer/Itinerary Planner, Washington, D.C.: Society for Neuroscience. Evarts, E.V. and Thach, W.T., Motor mechanisms of the CNS: cerebrocerebellar interrelations, Annu. Rev. Physiol. 31: 451–498, 1969. Fatemi, S.H., Stary, J.M., Halt, A.R., and Realmuto, G.R., Dysregulation of Reelin and Bcl2 proteins in autistic cerebellum, J. Autism. Dev. Disord. 31: 529–535, 2001. Fatemi, S.H., Halt, A.R., Realmuto, G., Earle, J., Kist, D.A., Thuras, P., and Merz, A., Purkinje cell size is reduced in cerebellum of patients with autism, Cell Mol. Neurobiol. 22: 171–175, 2002. Fiez,, J.A., Petersen, S.E., Cheney, M.K., and Raichle, M.E., Impaired non-motor learning and error detection associated with cerebellar damage, A single case study, Brain 115: 155–178, 1992. Fortin, M., Marchand, R., and Parent, A., Calcium-binding proteins in primate cerebellum, Neurosci. Res. 30: 155–168, 1998. Garber, H.J. and Ritvo, E.R., Magnetic resonance imaging of the posterior fossa in autistic adults, Am. J. Psychiatry 149: 245–247, 1992. Glickstein, M., May, J.G. III, and Mercier, B.E., Corticopontine projection in the macaque: the distribution of labeled cortical cells after large injections of horseradish peroxidase in the pontine nuclei, J Comp. Neurol. 235: 343–359, 1985. Gottwald, B., Wilde, B., Mihajlovic, Z., and Mehdorn, H.M., Evidence for distinct cognitive deficits after focal cerebellar lesions. J. Neurol. Neurosurg. Psychiatry 75(11): 1524–1531, 2004. Haines, D.E., May, P.J., and Dietrichs, E., Neuronal connections between the cerebellar nuclei and hypothalamus in Macaca fascicularis: cerebello-visceral circuits, J. Comp. Neurol. 299: 106–122, 1990. Hashimoto, T., Tayama, M., Muarkawa, K., Yoshimoto, T., Miyazali, M., Harada, M., and Kuroda, Y., Development of the brainstem and cerebellum in autistic patients, J. Autism Dev. Disord. 25: 1–18, 1995. Holttum, J.R., Minshew, N.J., Sanders, R.S., and Phillips, N.E., Magnetic resonance imaging of the posterior fossa in autism, Biol. Psychiatry 32: 1091–101, 1992. Hoover, J.E. and Strick, P.L., The organization of cerebellar and basal ganglia outputs to primary motor cortex as revealed by retrograde transneuronal transport of herpes simplex virus type 1, J Neurosci. 19: 1446–1463, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 171 Wednesday, February 15, 2006 3:19 PM

Cerebellar Networks and Autism: An Anatomical Hypothesis

171

Ilinsky, I.A. and Kultas-Ilinsky, K., Sagittal cytoarchitectonic maps of the Macaca mulatta thalamus with a revised nomenclature of the motor-related nuclei validated by observations on their connectivity, J. Comp. Neurol. 262: 331–364, 1987. Ito, M., Bases and implication of learning in the cerebellum-adaptive control and internal model mechanism, Prog. Brain Res. 148: 95–109, 2005. Jueptner, M., Frith, C.D., Brooks, D.J., Frackowiak, R.S.J., and Passingham, R.E., Anatomy of motor learning. II. Subcortical structures and learning by trial and error, J. Neurophysiol. 77: 1325–1337, 1997. Kalil, K., Projections of the cerebellar and dorsal column nuclei upon the thalamus of the rhesus monkey, J. Comp. Neurol. 195: 25–50, 1981. Kelly, R.M. and Strick, P.L., Rabies as a transneuronal tracer of circuits in the central nervous system, J. Neurosci. Methods 103: 63–71, 2000. Kelly, R.M. and Strick, P.L., Cerebellar loops with motor cortex and prefrontal cortex of a nonhuman primate, J. Neurosci. 23: 8832–8444, 2003. Kim, S.-G., Ugurbil, K., and Strick, P.L., Activation of cerebellar output nucleus during cognitive processing, Science 265: 949–951, 1994. Kleiman, M.D., Neff, S., and Rosman, N.P., The brain in infantile autism: are posterior fossa structures abnormal? Neurology 42: 753–760, 1992. Larsell, O., The Comparative Anatomy and Histology of the Cerebellum from Monotremes through Apes, Minneapolis, MN: The University of Minnesota Press, 1970. Lee, D.A., Lopez-Alberola, R., and Bhattacharjee, M., Childhood autism: a circuit syndrome? Neurologist 9: 99–109, 2003. Lee, M., Martin-Ruiz, C., Graham, A., Court, J., Jaros, E., Perry, R., Iversen, P., Bauman, M., and Perry, E., Nicotinic receptor abnormalities in the cerebellar cortex in autism, Brain 125: 1483–1495, 2002. Leichnetz, G.R., Smith, D.J., and Spencer, R.F., Cortical projections to the paramedian tegmental and basilar pons in the monkey, J. Comp. Neurol. 228(3): 388–408, 1984. Leiner, H.C., Leiner, A.L., and Dow, R.S., The human cerebro-cerebellar system: its computing, cognitive, and language skills, Behav. Brain Res. 44: 113–128, 1991. Liu, Y., Pu, Y., Gao, J.H., Parsons, L.M., Xiong, J., Liotti, M., Bower, J.M., and Fox, P.T., The human red nucleus and lateral cerebellum in supporting roles for sensory information processing, Hum Brain Mapp. 10(4): 147–159, 2000. Lynch, J.C., Hoover, J.E., and Strick, P.L., Input to the primate frontal eye field from the substantia nigra, superior colliculus, and dentate nucleus demonstrated by transneuronal transport, Exp. Brain Res. 100: 181–186, 1994. Matano, S., Brief communication: proportions of the ventral half of the cerebellar dentate nucleus in humans and great apes, Am. J. Phys. Anthro. 114: 163–165, 2001. Matelli, M., Luppino, G., Fogassi, L., and Rizzolatti, G., Thalamic input to inferior area 6 and area 4 in the macaque monkey, J. Comp. Neurol. 280: 468–488, 1989. Matelli, M. and Luppino, G., Thalamic input to mesial and superior area 6 in the macaque monkey, J. Comp. Neurol. 372: 59–87, 1996. Middleton, F.A. and Strick, P.L., Anatomical evidence for cerebellar and basal ganglia involvement in higher cognitive function, Science 266: 458–461, 1994. Middleton, F.A. and Strick, P.L., New concepts regarding the organization of basal ganglia and cerebellar output, Excerpta Med. 1116: 253–269, 1996. Middleton, F.A. and Strick, P.L., Dentate output channels: motor and cognitive components, in Progress in Brain Research. 114: The Cerebellum from Structure to Control, Eds., De Zeeuw, C.I., Strata, P., and Voogd, J., Amsterdam, VA: Elsevier, 1997, pp. 553–566.

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 172 Wednesday, February 15, 2006 3:19 PM

172

Understanding Autism: From Basic Neuroscience to Treatment

Middleton, F.A. and Strick, P.L., Cerebellar output: motor and cognitive channels, Trends Cognit. Sci. 2: 348–354, 1998. Middleton, F.A. and Strick, P.L., Cerebellar “projections” to the prefrontal cortex of the primate, J. Neurosci. 21: 700–712, 2001. Murakami, J.W., Courchesne, E., Press, G.A., Yeung-Courchesne, R., and Hesselink, J.R., Reduced cerebellar hemisphere size and its relationship to vermal hypoplasia in autism, Arch. Neurol. 46: 689–694, 1989. Mushiake, H. and Strick, P.L., Preferential activity of dentate neurons during limb movements guided by vision, J. Neurophysiol. 70: 2660–2664, 1993. Noda, H. and Fujikado, T., Topography of the oculomotor area of the cerebellar vermis in macaques as determined by microstimulation, J. Neurophysiol. 58: 359–378, 1987. Orioli, P. and Strick, P.L., Cerebellar connections with the motor cortex and the arcuate premotor area: an analysis employing retrograde transneuronal transport of WGAHRP, J. Comp. Neurol. 28: 612–626, 1989. Palmen, S.J., van Engeland, H., Hof, P.R., and Schmitz, C., Neuropathological findings in autism, Brain 127: 2572–2583, 2004. (Epub August 25, 2004.) Percheron, G., The thalamic territory of cerebellar afferents and the lateral region of the thalamus of the macaque in stereotaxic ventricular coordinates, J. Hirnforsch. 18: 375–400, 1977. Percheron, G., Francois, C., Talbi, B., Yelnik, J., and Fenelon, G., The primate motor thalamus, Brain Res. Brain Res. Rev. 22: 93–181, 1996. Pimenta, A.F., Strick, P.L., and Levitt, P.R., A novel proteoglycan epitope is expressed in functionally discrete patterns in primate cortical and subcortical regions, J. Comp. Neurol. 430: 369–388, 2001. Piven, J., Nehme, E., Simon, J., Barta, P., Pearlson, G., and Folstein, S.E., Magnetic resonance imaging in autism: measurement of the cerebellum, pons, and fourth ventricle, Biol. Psychiatry 31: 491–504, 1992. Reis, D.J., Doba, N., and Nathan, M.A., Predatory attack, grooming, and consummatory behaviors evoked by electrical stimulation of cat cerebellar nuclei, Science 182(114): 845–847, 1973. Rispal-Padel, L., Cicirata, F., and Pons, C., Cerebellar nuclear topography of simple and synergistic movements in the alert baboon (Papio papio), Exp. Brain Res. 47: 365–380, 1982. Ritvo, E.R., Freeman, B.J., Scheibel, A.B., Duong, T., Robinson, H., Guthrie, D., and Ritvo, A., Lower Purkinje cell counts in the cerebella of four autistic subjects: initial findings of the UCLA-NSAC Autopsy Research Report, Am. J. Psychiatry 143: 862–866, 1986. Rouiller, E.M., Liang, F., Babalian, A., Moret, V., and Wiesendanger, M., Cerebellothalamocortical and pallidothalamocortical projections to the primary and supplementary motor cortical areas: a multiple tracing study in macaque monkeys, J. Comp. Neurol. 345: 185–213, 1994. Rouiller, E.M., Tanne, J., Moret, V., and Boussaoud, D., Origin of thalamic inputs to the primary, premotor, and supplementary motor cortical areas and to area 46 in macaque monkeys: a multiple retrograde tracing study. J. Comp. Neurol. 409: 131–152, 1999. Sakai, S.T., Inase, M., and Tanji, J., Pallidal and cerebellar inputs to thalamocortical neurons projecting to the supplementary motor area in Macaca fuscata: a triple-labeling light microscopic study, Anat. Embryol. (Berl).199: 9–19, 1999. Sakai, S.T., Inase, M., and Tanji, J., The relationship between MI and SMA afferents and cerebellar and pallidal efferents in the macaque monkey, Somatosens. Mot. Res. 19: 139–148, 2002.

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 173 Wednesday, February 15, 2006 3:19 PM

Cerebellar Networks and Autism: An Anatomical Hypothesis

173

Schell, G.R. and Strick, P.L., The origin of thalamic inputs to the arcuate premotor and supplementary motor areas, J. Neurosci. 4: 539–560, 1984. Schmahmann, J.D., Dysmetria of thought: clinical consequences of cerebellar dysfunction on cognition and affect, Trends Cognit. Sci. 2: 362–371, 1998. Schmahmann, J.D. and Pandya. D.N., Projections to the basis pontis from the superior temporal sulcus and superior temporal region in the rhesus monkey, J. Comp. Neurol. 308(2): 224–248, 1991. Schmahmann, J.D. and Pandya, D.N., Prelunate, occipitotemporal, and parahippocampal projections to the basis pontis in rhesus monkey, J. Comp. Neurol. 337(1): 94–112. 1993. Schmahmann, J.D. and Pandya, D.N., Anatomic organization of the basilar pontine projections from prefrontal cortices in rhesus monkey. J. Neurosci. 17: 438–458, 1997. Schmahmann, J.D. and Sherman, J.C., The cerebellar cognitive affective syndrome, Brain 121: 561–579, 1998. Schmahmann, J.D., Disorders of the cerebellum: ataxia, dysmetria of thought, and the cerebellar cognitive affective syndrome. J. Neuropsychiatry Clin. Neurosci. 16: 367–378, 2004. Snider, R.S. and Maiti, A., Cerebellar contributions to the Papez circuit, J. Neurosci. Res. 2: 133–146, 1976. Stanton, G.B., Topographical organization of ascending cerebellar projections from the dentate and interposed nuclei in Macaca mulatta: an anterograde degeneration study. J. Comp. Neurol. 190: 699–731, 1980. Strick, P.L. and Card, J.P., Transneuronal mapping of neural circuits with alpha herpesviruses, in Experimental Neuroanatomy: A Practical Approach, Eds., Bolam, J.P., Oxford: Oxford University Press, 1992, pp. 81–101. Strick, P.L., Hoover, J.E., and Mushiake, H., Evidence for “output channels” in the basal ganglia and cerebellum, in Role of the Cerebellum and Basal Ganglia in Voluntary Movement, Eds., Mano, N., Hamada, I., and DeLong, M.R., Elsevier, Amsterdam, VA, 1993, pp. 171–180. Suh, M., Leung, H.C., and Kettner, R.E., Cerebellar flocculus and ventral paraflocculus Purkinje cell activity during predictive and visually driven pursuit in monkey, J. Neurophysiol. 84: 1835–1850, 2000. Takarae, Y., Minshew, N.J., Luna, B., and Sweeney, J.A., Oculomotor abnormalities parallel cerebellar histopathology in autism, J. Neurol. Neurosurg. Psychiatry 75: 1359–1361, 2004. Takarae, Y., Minshew, N.J., Luna, B., Krisky, C.M., and Sweeney, J.A., Pursuit eye movement deficits in autism, Brain 127: 2584–2594, 2004b. (Epub October 27, 2004.) Thach, W.T., Perry, J.G., Kane, S.A., and Goodkin, H.P., Cerebellar nuclei: rapid alternating movement, motor somatotopy, and a mechanism for the control of muscle synergy, Rev. Neurol. (Paris) 149: 607–628, 1993. Vilensky, J.A. and van Hoesen, G.W., Corticopontine projections from the cingulate cortex in the rhesus monkey, Brain Res. 205(2): 391–395, 1981. Wiesendanger, R. and Wiesendanger, M., The thalamic connections with medial area 6 (supplementary motor cortex) in the monkey (Macaca fascicularis), Exp. Brain Res. 59: 91–104, 1985a. Wiesendanger, R. and Wiesendanger, M., Cerebella-cortical linkage in the monkey as revealed by transcellular labeling with the lectin wheat germ agglutinin conjugated to the marker horseradish peroxidase. Exp. Brain Res. 59: 105–117, 1985b. Wiesendanger, R., Wiesendanger, M., and Ruegg, D.G., An anatomical investigation of the corticopontaine projection in the primate (Macaca fascicularis and Saimiri sciureus) — II, The projection from frontal and parental association areas, Neuroscience 4: 747–765, 1979.

© 2006 by Taylor & Francis Group, LLC

2732_C008.fm Page 174 Wednesday, February 15, 2006 3:19 PM

174

Understanding Autism: From Basic Neuroscience to Treatment

Zanchetti, A. and Zoccolini, A., Autonomic hypothalamic outbursts elicited by cerebellar stimulation, J. Neurophysiol. 17: 475–483, 1954. Zemanick, M.C., Strick, P.L., and Dix, R.D., Direction of transneuronal transport of herpes simplex virus 1 in the primate motor system is strain-dependent, Proc. Natl. Acad. Sci. U.S.A. 88: 8048–8051, 1991.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 175 Wednesday, February 15, 2006 3:19 PM

9 Language in Autism Matthew Walenski, Helen Tager-Flusberg, and Michael T. Ullman CONTENTS Introduction............................................................................................................175 Language and Communication in ASD: The Evidence........................................176 Pragmatic Deficits ..........................................................................................176 Nonverbal Communicative Gesture...........................................................177 Speech Acts................................................................................................177 Conversational Discourse ..........................................................................177 Pragmatic Functions of Prosody ...............................................................178 Interpreting Nonliteral Language ..............................................................178 Grammar and Lexicon....................................................................................178 Grammatical Abilities................................................................................179 Lexical Abilities.........................................................................................181 Neuroimaging Studies ...............................................................................181 Formulaic Speech ...........................................................................................183 Integrative Theories of Language in ASD ............................................................183 Pragmatics and Theory of Mind ....................................................................183 Grammar, Lexicon, and the PDH ..................................................................185 ASD Profile of Procedural System Functions ..........................................188 ASD Profile of the Declarative Memory System .....................................191 The Neurobiology of Procedural and Declarative Memory Brain Structures in ASD.............................................................193 Summary and Conclusion .....................................................................................193 Acknowledgments..................................................................................................194 References..............................................................................................................194

INTRODUCTION Autism, often referred to as autism spectrum disorder (ASD), is a developmental disorder of the brain that is strongly associated with deficits in language and communication as well as a variety of other impairments, including abnormal social interaction and motor function.1 Roughly 20% of children with autism are essentially nonverbal, using fewer than five words per day.2 Others acquire functional language to varying degrees, although the exact profile of language and communicative abilities appears to be somewhat heterogeneous. 175

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 176 Wednesday, February 15, 2006 3:19 PM

176

Understanding Autism: From Basic Neuroscience to Treatment

In this chapter, we first summarize the evidence regarding language abilities in ASD. (Note that we use the terms autism and ASD in the broadest sense, including diagnoses of autism and Asperger’s syndrome.) The data suggest that ASD is associated with a particular pattern of relatively spared and impaired language functions. We then examine in depth two explanatory theories of language in ASD. The two theories are complementary in that they focus on different sets of language functions. However, both theories take the same broad approach in that they address relations between language and nonlanguage domains with a view to exploring whether similar behavioral profiles across apparently distinct cognitive domains can be explained by common neurocognitive substrates. According to one theory, deficits of “theory of mind” in ASD can explain pragmatic impairments of language and communication in terms of social deficits and their neurocognitive underpinnings.3–5 In contrast, the procedural deficit hypothesis (PDH) posits that grammatical impairments in the disorder — including syntax, morphology, and phonology — can be largely explained by neurocognitive abnormalities of the procedural memory system, whereas lexical knowledge, which depends on the declarative memory system, remains relatively spared.6,7 Thus, rather than investigating language deficits in isolation, we examine integrative explanatory theories that attempt to account for these deficits in the broader context of brain and behavior in ASD.

LANGUAGE AND COMMUNICATION IN ASD: THE EVIDENCE Here we first present data related to pragmatic linguistic functions and then turn to evidence pertaining to grammatical and lexical aspects of language. Formulaic speech in ASD, which may be related to more than one of these domains, is addressed at the end of this section. Our discussion will concentrate on cross-sectional studies and, by necessity, on those children who have acquired at least some functional language.

PRAGMATIC DEFICITS Pragmatics concerns the practical knowledge that is necessary to use and interpret language appropriately for the social and real-world contexts in which utterances are made. Social aspects of pragmatics crucially include knowledge of the social rules that govern speaker–hearer interactions (which often involves interpreting a speaker’s intended meaning across different social contexts), whereas real-world aspects of pragmatics include knowledge of people and objects and how they are likely to interact (e.g., knowing that girls are more likely than boys to play with dolls). Pragmatics encompasses both verbal and nonverbal aspects of communication, including gestures, prosodic cues (i.e., of intonation and rhythm), and facial expressions, all of which combine to enhance effective communication in face-toface social contexts. A range of impairments consistent with pragmatic deficits are inherent to ASD. These impairments are widespread and are found in both children and adults, with diagnoses ranging from classic autism to Asperger’s syndrome.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 177 Wednesday, February 15, 2006 3:19 PM

Language in Autism

177

Nonverbal Communicative Gesture Young ASD children present with well-documented problems in nonverbal communication.8–10 Data based on parental report indicate significant delays in the use of early gestures.11 Studies of nonverbal intentional gestural communication in young ASD children consistently show that pointing gestures that are used to communicate requests are generally produced and understood, whereas pointing gestures that are used to share interest in an object or to direct attention to an event are virtually absent (even though these emerge at the same time as requesting gestures in typically developing children).12 Speech Acts Speech acts are utterances that serve a communicative function, such as requests, comments, or commands. Importantly, they require knowledge of how language is used within a culture.13 Evidence suggests that ASD children are missing speech acts that emphasize social engagement rather than speech acts that regulate others’ behavior.14 Wetherby and Prutting15 examined the range of speech acts that were expressed by ASD children in both gestural and spoken language at early stages of development in comparison to language-matched typically developing children. They found that the ASD children were not significantly different from the controls in their use of language for requesting objects or actions, for protests, and for selfregulation (e.g., “Don’t do that.”). However, speech acts with social functions, such as comments, showing off, acknowledging the listener, and requesting information, were completely absent from ASD discourse. In another study, ASD children used declarative sentences that were direct responses to questions but did not make other types of declarative statements or comments, which are thought to involve more significant social awareness.16 Compared to children with specific language impairment (SLI), ASD children have been found to use fewer affirming or agreeing utterances.17 Finally, in a study comparing ASD children to Down syndrome children, the ASD children rarely communicated about objects that were the focus of their mothers’ attention.18 Conversational Discourse Deficits in conversational ability are found throughout childhood and adolescence in ASD.19 Older, higher-functioning ASD adolescents are likely to speak too much and in a monologue style during interviews.20,21 They have problems responding adequately to questions, especially when discussing an unusual event or personal narrative,22 and tend to have difficulty making clear reference in their conversations to people or places.23 These latter findings are consistent with evidence of impaired performance on referential communication tasks.24 They also have difficulty judging the amount of information that needs to be included for effective communication.25 Other studies have shown that people with autism seem less able to shift their discourse when there are failures in communication — for example, if a listener has not heard or understood an utterance. Paul and Cohen26 found that although adults

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 178 Wednesday, February 15, 2006 3:19 PM

178

Understanding Autism: From Basic Neuroscience to Treatment

with autism were just as likely to respond to requests for clarification as mentally retarded adults matched on nonverbal IQ, their answers were less specific than those of the nonautistic participants. They were also less likely to add information that might be of help to the listener, suggesting problems in judging the relevance of a piece of information. This finding was recently replicated in high-functioning schoolage children with autism.27 Finally, we note that several studies have demonstrated abnormalities in autism in other kinds of discourse skills, especially in telling stories or recounting personal narratives.28–30 Pragmatic Functions of Prosody Prosody describes the timing, rhythm, and intonation of speech. It has numerous pragmatic functions, including nongrammatical pauses and the use of stress to direct attention to words or other elements. In a recent survey of investigations of prosody in ASD, all studies (ten of ten) in which pragmatic functions of prosody could be isolated reported deficits in those functions.31 Interpreting Nonliteral Language Spoken and written discourse often includes different forms of nonliteral language, including idioms, metaphors, lies, jokes, and so forth. To understand nonliteral forms of language, one must infer the speaker’s intended meaning (as in jokes or metaphors, for example) or understand cultural–linguistic expressions (as in idioms). Both of these entail pragmatic knowledge of how language is used in different social contexts. ASD individuals, even those who are older and high functioning, have great difficulty interpreting nonliteral or figurative speech.32–35 These difficulties include problems with idioms,36 with metaphor and irony,32,37 and with the ability to explain nonliteral utterances embedded in stories (e.g., lies, jokes, pretence, irony, sarcasm, or double bluff).33,38,39 Using a more structured task, it has also been found that children with autism have difficulty interpreting a speaker’s intended meaning in a conversational context and, unlike matched controls, they interpret utterances in a literal way instead of in relation to the speaker’s stated desire.40

GRAMMAR

AND

LEXICON

Language depends upon two mental abilities.41,42 Idiosyncratic information must be memorized in some sort of mental dictionary, which is often referred to as the mental lexicon. The lexicon necessarily includes all words with arbitrary sound–meaning pairings, such as the noncompositional (“simple”) word cat. But language also consists of regularities, which can be captured by rules of grammar. The rules constrain how lexical forms and other basic units in language combine to make complex representations, including phrases and sentences (e.g., the cat took the train to work; syntax), complex words (e.g., walk + -ed -> walked; morphology), and the structured sound patterns of words (e.g., t + u + p -> tup; phonology). Importantly, although complex representations (such as the phrase the cat) could be computed anew each time (the + cat), they could in principle also be stored in the mental lexicon (e.g., the cat could be stored as a single unit). As we will see, the evidence

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 179 Wednesday, February 15, 2006 3:19 PM

Language in Autism

179

suggests that grammatical composition is impaired in ASD, whereas lexical knowledge remains relatively spared. Grammatical Abilities Syntax underlies the rule-governed combination of words into the sequential and hierarchical structures of phrases and sentences. Impairments of sentence comprehension have been widely reported in ASD in both auditory and visual (i.e., reading) domains.43–50 However, not all subjects show these deficits.44,46 Intriguingly, one study of ASD adults reports normal accuracy but abnormally fast reaction times in a sentence comprehension task.51 In expressive language (i.e., in speaking), studies have found that the spontaneous speech of ASD children shows reduced syntactic complexity compared to children with Down syndrome or to developmentally delayed or typically developing control children.52–54 Similarly, ASD children have been found to omit required closed-class items (e.g., definite articles such as the), which play important grammatical roles.55,56 Children with ASD have also been shown to have lower rates of novel, nonimitative utterances compared to typically developing and Down syndrome children, instead relying more on “formulaic” utterances (see following text for further discussion on formulaic speech).57 Finally, ASD children have shown impairments on the portion of the Clinical Evaluation of Language Fundamentals (CELF) that tests immediate sentence repetition and reflects (at least in part) expressive syntax.46,58 Morphology, which refers to the structure of words with respect to their meaningful parts, comes in two flavors. Derivational morphology (the creation of new words; e.g., solemnity and toughness are derived from solemn and tough) has, to our knowledge, not been investigated in ASD. Several studies of ASD have examined inflectional morphology, which concerns the modification of a word to fit its grammatical role (e.g., sang and walked are past-tense inflected). It is important to note that inflectional morphology involves both morphosyntax (the choice of inflection based on the syntax of a sentence — for example, choosing present tense or past tense, depending on the syntactic context) and morphophonology (phonological changes to a word that reflect morphological processes, such as the vowel change in the irregular past tense formation of dug from dig). Although morphosyntax strongly depends upon combinatorial rule-governed (i.e., grammatical) processes, this is not necessarily the case for morphophonology. Whereas regular morphophonological transformations, as in English regular past tenses, follow rule-governed compositional patterns (walk + -ed), irregular morphophonology is at least partly unpredictable (e.g., sing–sang, bring–brought), and therefore must rely on stored lexical knowledge. In two studies that examined samples of spontaneous speech, ASD children omitted inflectional morphemes (e.g., they produced play instead of playing) more often than unimpaired or mentally retarded control subjects.55,56 This pattern appeared to hold particularly for certain regularly inflected forms (especially -ingsuffixed forms, such as playing, but also for regular past tenses in the Bartolucci et al. study), whereas irregular past-tenses were relatively spared in both studies.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 180 Wednesday, February 15, 2006 3:19 PM

180

Understanding Autism: From Basic Neuroscience to Treatment

Similarly, a more recent study of spontaneous speech reported impairments on thirdperson singular present tense (e.g., in the production of washes), which is almost completely regular.59 A recent examination of elicited verbal morphology (subjects were asked to produce inflected forms in past tense picture sentence contexts) found impaired production of past tense forms in ASD (regulars and irregulars were not distinguished).58 Similarly, high rates of omissions and incorrect inflections in present tense and (both regular and irregular) past tense production (e.g., wash or washing for washes; catch or catching for caught) were found in an elicitation study of language-impaired ASD children compared to ASD children without evident language deficits.60 A more recent study of elicited past tense production in highfunctioning ASD and age-matched typically developing children found normal accuracy scores for regular, irregular, and novel (e.g., plag–plagged) verbs. However, response times revealed important differences between the groups: the production of regularized (stem + -ed, e.g., in walked, plagged, digged) but not irregularized (e.g., dug, splim–splam) forms was abnormally fast in ASD compared to controls.61 In summary, all six studies of spontaneous or elicited speech reported abnormalities in the production of inflectional morphology. All but one reported actual impairments. The remaining study, in which ASD children were abnormally fast at producing regularized forms, differed from the others in having the highest-functioning ASD subjects (the highest IQs) and in having little or no requirement for social interaction during production, as the items were presented visually on a computer screen (for more discussion of this last point see Reference 62). Additionally, accuracy or response time differences between ASD and control subjects were found in all six studies for regulars, but were not observed for irregulars in three of the four studies that distinguished regular, and irregular morphological forms, including the one study in which regular and irregular forms were explicitly well-matched on frequency and other factors.61 Thus, ASD seems to be associated with abnormalities of inflectional morphology, particularly for regular forms, although it is not yet clear to what extent this is due to problems of regular morphophonology, morphosyntax, or both. Phonology refers to the sound patterns of language. In ASD, certain aspects of phonology may be relatively preserved, whereas others are somewhat impaired. Phonology is concerned both with individual speech sounds (i.e., phonemes, such as the /a/ sound in father) and how they are combined sequentially and hierarchically into syllables and words. Several studies have reported no particular ASD deficits with individual phonemes in either receptive or expressive language46 (but see Reference 31; for a review of earlier studies see Reference 63). Impairments in ASD have more often been reported in the combination of sounds into complex structures. Thus, deficits have been found in the repetition of auditorily presented nonsense words (e.g., barrazon),46,64 though the presence of such impairments is less clear in some studies.53,58 Because this task is posited to involve both disassembling the input into smaller units (e.g., phonemes or syllables) and then reassembling these units in production, it is expected to involve compositional processes. The evidence therefore suggests that at least compositional aspects of phonology may be somewhat impaired.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 181 Wednesday, February 15, 2006 3:19 PM

Language in Autism

181

Lexical Abilities Behavioral studies suggest that lexical knowledge remains largely normal in ASD. First of all, word-learning abilities seem to be essentially intact.53 Second, performance has been found to be unimpaired on a range of receptive lexical tasks, such as word–picture correspondence (“Is this a … ?”), word–picture matching (“Pick the correct picture to match …”), picture selection (“Show me all the …”), and word definition (“What is a …?”).44,46,65 Third, expressive lexical abilities seem to remain largely spared in single-word production tasks, such as picture naming, synonym and antonym generation, and reading single letters or words out loud.45,46,48,66 When ASD subjects are asked to name pictures as rapidly as possible (so-called “rapid automatic naming”), a mixed profile can be seen, with some subjects showing normal performance, whereas others are impaired.47 Similarly, performance on verbal fluency tasks, in which subjects are asked to name as many words as possible in a given period of time, seems to be generally, but not always, spared. Normal performance has been found in both letter fluency (e.g., “Name as many words as you can that begin with the letter F.”) and category fluency (e.g., “Name as many animals as you can.”).44,47,48,67 Other studies, in contrast, have reported deficits in both types of verbal fluency tasks49,66 as well as in unconstrained (“miscellaneous”) verbal fluency (e.g., “Say as many words as you can think of, any words at all.”).67 Finally, as we have seen earlier, the production of irregular past-tense forms (which depend on memorized lexical knowledge) is generally, but not always, spared. Thus, the evidence suggests that although lexical knowledge itself may remain spared in ASD, there seem to be some deficits in retrieving or searching for this knowledge, perhaps particularly under speeded conditions. Neuroimaging Studies There have been few neuroimaging studies of lexicon or grammar in ASD. We are aware of only two functional neuroimaging studies of either domain, both of which examined syntactic processing. (A study of conceptual processing in ASD is discussed below.) In addition, one structural magnetic resonance imaging (MRI) experiment examined the relation between brain abnormalities and language impairments in ASD. In presenting these three studies, we focus specifically on structures previously implicated in language functions (i.e., we do not discuss visual areas, the pons, etc.). In the section on the PDH, we discuss the potential significance of these structures, specifically the relation between frontal, basal ganglia, and cerebellar structures and procedural memory and grammar on the one hand and between temporal/ temporo-parietal structures and declarative memory and lexical memory on the other. A functional MRI (fMRI) study of visual sentence comprehension (as compared to visual fixation) found that ASD adults showed greater activation than unimpaired age- and IQ-matched controls in the left posterior superior temporal sulcus (between the superior and middle temporal gyri, i.e., between Brodmann’s areas — BA – 22 and 21), in right temporo-parietal/inferior-parietal cortex (BA 39), and in the parahippocampal gyrus bilaterally; but less activation than controls in the left inferior frontal gyrus (“Broca’s region,” i.e., BA 44, 45, and 47), right inferior frontal gyrus,

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 182 Wednesday, February 15, 2006 3:19 PM

182

Understanding Autism: From Basic Neuroscience to Treatment

and left lateral and medial premotor cortex and nearby frontal cortex, including the SMA/pre-SMA (SMA refers to the supplementary motor area in medial BA 6; preSMA lies just anterior to the SMA).51 Similarly, a positron emission tomography (PET) study of auditory sentence comprehension (in comparison to silence) reported greater mean regional activation (we do not discuss deactivations, which are difficult to interpret) in adult ASD subjects (who were impaired at an off-line measure of auditory sentence comprehension) than age-matched healthy controls in the right superior temporal gyrus (BA 22, 41/42) and right inferior parietal cortex (BA 40), but less activation than controls in left premotor cortex (middle/inferior BA 6) as well as in the mid portion (i.e., not posterior) of the left middle temporal gyrus (BA 21).47 Additionally, within left temporal cortex, the ASD subjects showed a peak activation in a posterior portion of this region, whereas the control subjects showed peak activations in mid and anterior portions. Similar to the fMRI study, there was also less activation in ASD than control subjects in left inferior frontal gyrus (within Broca’s region, i.e., in BA 44, 45, and posterior 47), as well as in the left caudate nucleus and the left lentiform nucleus (both of which are part of the basal ganglia), although these differences did not reach significance. In a follow-up reanalysis of a subset of this data,68 the authors reported reduced activation in the ASD subjects, compared to the controls, in left dorsolateral prefrontal cortex (BA 46) and in the right dentate nucleus of the cerebellum, suggesting abnormalities of frontal–cerebellar circuitry, in particular between left frontal cortex and the right cerebellum (note that the right cerebellum is connected primarily to the left cerebrum). These functional imaging findings are paralleled by findings from a structural MRI study that related neuroanatomical volumes to performance on the CELF (which probes multiple domains of language) and a nonword repetition task.69 ASD children were first divided into those who performed within the normal range on these two language tests and those who did not. These two ASD groups were compared against two age-matched comparison groups, one with specific language impairment (SLI) and one composed of typically developing control children. Both language-impaired groups (ASD with impaired language and SLI) had significant right hemisphere asymmetry (i.e., right larger than left), as compared to the two comparison groups, in classical Broca’s area — that is, the pars triangularis (BA 45) and pars opercularis (BA 44) of the inferior frontal gyrus. Reported volumes from this region in each hemisphere suggest that the increased asymmetry stemmed from both decreased left hemisphere volume and increased right hemisphere volume in the two language-impaired groups. In contrast, both language-impaired groups showed left hemisphere asymmetry (left larger than right) in the planum temporale (the posterior portion of the upper bank of the temporal lobe), whereas the two comparison groups did not. Analogously to the Broca’s area asymmetry, this increased leftward asymmetry in the planum temporale seems to be explained by a combination of decreased right hemisphere volume and increased left hemisphere volume in the two language-impaired groups. In sum, in both of the functional neuroimaging studies of receptive syntax, ASD subjects showed greater activation in posterior superior-temporal/inferior-parietal cortex — including the posterior superior temporal gyrus and sulcus and temporoparietal/inferior-parietal cortex — but less in certain left frontal regions, i.e., left

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 183 Wednesday, February 15, 2006 3:19 PM

Language in Autism

183

premotor cortex, dorsolateral prefrontal cortex, and Broca’s region, as well as in left basal ganglia structures and the right cerebellum (note that only one of the studies addressed subcortical or cerebellar structures). The structural MRI study found that the language-impaired ASD (and SLI) subjects showed a volumetric decrease in left frontal regions, especially in Broca’s area, but an increase in a left temporal lobe region — an intriguingly similar pattern to the patterns of increased and decreased activation found in the two functional neuroimaging studies. Interpretations of these findings are discussed below.

FORMULAIC SPEECH ASD speech is marked by several striking features: repetitive and stereotyped utterances (e.g., overuse of routine utterances such as “thank you” or “you’re welcome”); idiosyncratic sound–meaning associations (i.e., “metaphorical language” such as “I want to go blue” to express a desire to go outside); excessively literal language (e.g., responding “No, it’s raining water” to the statement “It’s raining cats and dogs”); difficulty with pronouns (e.g., a child with autism may say “Would you like a cookie?” to request one) and with other deictic terms (i.e., terms whose reference depends on contextual factors, such as this, that, here, there, etc.); and immediate or delayed echolalia (lexically, prosodically, and syntactically faithful repetitions).57,63,70 These various aspects of speech are highly characteristic of ASD and feature among the diagnostic criteria in DSM-IV for qualitative communication impairments in the disorder.1 They are observed at much higher rates in ASD than in typically developing children (for a review, see Reference 63) or children with other developmental disorders, including SLI71 and Down’s syndrome.57 These aspects of autistic speech are often described together as formulaic speech. 57,72 Formulas are defined as prefabricated sequences of words that are stored and retrieved whole from memory (in any population, impaired or unimpaired).73 It has been suggested that at least certain aspects of formulaic speech in ASD (e.g., echolalia and an overreliance on a restricted range of formulaic forms) may reflect pragmatic or social deficits in language use.63,72 It has also been suggested that formulas offer a shortcut (via memorization) by which to bypass grammatical processing, particularly when grammatical processing is difficult.73 Thus the overreliance in ASD on these forms may reflect more than one underlying deficit.

INTEGRATIVE THEORIES OF LANGUAGE IN ASD Here we present two explanatory neurocognitive theories of language in ASD. Both theories integrate findings from language and nonlanguage domains. One focuses on pragmatics and the other on lexicon and grammar.

PRAGMATICS

AND

THEORY

OF

MIND

As we have seen earlier, ASD involves pervasive impairments in the pragmatic aspects of language use. The disorder is also strongly associated with nonverbal social deficits.74,75 The theory of mind hypothesis of autism has been proposed to

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 184 Wednesday, February 15, 2006 3:19 PM

184

Understanding Autism: From Basic Neuroscience to Treatment

integrate deficits across these domains.76,77 According to this hypothesis, people with ASD are fundamentally impaired at causally linking their own and other people’s behavior to mental states. These deficits of theory of mind are posited to underlie impairments both of pragmatics and of nonverbal social abilities.33 The evidence suggests that individuals with ASD have theory of mind impairments, as indicated by deficits in tasks that tap their understanding of minds and mental states such as belief (e.g., false belief), knowledge, and emotion.78,79 Crucially, these impairments appear to be linked not only to social deficits in ASD75 but also to pragmatic deficits.33 Several lines of evidence support an association between impairments of theory of mind and of pragmatics. First of all, whereas ASD is associated with impairments of pragmatics that involve viewing people as mental beings, certain other aspects of language use that do not require knowledge of people’s mental states tend to remain relatively spared — such as turn-taking skills, requesting behavior, or regulatory speech acts (see preceding text). A second line of evidence comes from studies of joint attention in young children. Joint attention involves the triadic engagement between a child, another person, and an object or event of interest. It has been viewed as an early theory of mind ability in that it demonstrates the child’s capacity to monitor and manipulate the attentional focus of another person. Deficits in joint attention are evident in looking behavior, communicative pointing, and sharing interest with another person.8,10 Impairments in joint attention may explain why language is delayed in ASD, as joint attention skills are important developmental predictors of the rate of language development in the disorder.80 Third, deficits in conversational discourse in ASD seem to reflect problems in understanding that communication is both about exchanging information with others and about the expression and interpretation of the speaker’s intended meaning.32,81 For example, Capps et al.19 found a positive correlation between performance on theory of mind tasks and the ability to maintain an ongoing topic of conversation among children with autism. Similarly, Hale and Tager-Flusberg82 found an inverse relationship between performance at theory of mind tasks and the frequency of non-topic related utterances, a relationship that was independent of overall language ability. Difficulties with nonliteral language also seem to stem from impairments with theory of mind problems, in particular with understanding intentional aspects of communication. For example, Happé’s studies have demonstrated a direct relationship between theory of mind performance and the ability to interpret nonliteral meaning in people with autism.32,33 Finally, it is interesting to note that functional neuroimaging studies in ASD suggest abnormalities in the neural structures that underlie the processing of theory of mind. Several brain structures have been implicated in typically developing individuals in theory of mind processing, in particular medial frontal cortex (BA 8/9, bordering on the cingulate gyrus), lateral inferior frontal cortex (primarily Broca’s area, namely BA 44/45), and posterior superior temporal/temporo-parietal cortex.83,84 Intriguingly, the small number of neuroimaging studies of theory of mind in ASD suggest a tendency for decreased activation (relative to controls) in both of these frontal regions and increased activation (relative to controls) in the temporal

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 185 Wednesday, February 15, 2006 3:19 PM

Language in Autism

185

region.84–87 This pattern of decreased activation in Broca’s area and increased activation in posterior superior temporal/temporo-parietal cortex is strikingly similar to that found in the functional neuroimaging studies of language discussed earlier. It remains to be seen whether similar patterns are found in the processing of pragmatic aspects of language.

GRAMMAR, LEXICON,

AND THE

PDH

As we have seen in the preceding text, ASD is associated with impairments of combinatorial aspects of grammar, across syntax, morphology, and phonology. According to the procedural deficit hypothesis (PDH) of autism, these language impairments are explained by abnormalities of the brain structures that underlie the procedural memory system, resulting in impairments of the particular language and nonlanguage functions that depend on these structures. For a discussion of the PDH with respect to a number of different disorders, including autism, SLI, attention deficit hyperactivity disorder (ADHD), and dyslexia, see Reference 6. For an indepth examination of the hypothesis as it applies to SLI, see Reference 7. For a related hypothesis of procedural learning deficits in ASD, see Reference 88. The procedural memory system (for the sake of simplicity, we also refer to the system as the “procedural system”) is implicated in the learning of new, and the control of long-established, motor and cognitive “skills,” “habits,” and other procedures, such as typing, riding a bicycle, or skilled game playing.89–91 The system is particularly important for acquiring and performing skills involving sequences91,92 and has been shown to underlie nonlinguistic rule learning.93,94 Evidence also suggests that the system subserves aspects of the learning and use of grammatical rulegoverned combination, across syntax, morphology, and phonology, in both receptive and expressive language.6,95 The procedural system is composed of a network of several interconnected brain structures. It depends especially on structures in the left hemisphere of the cerebrum.96,97 It is rooted in neural circuits that encompass the frontal lobes and the basal ganglia (a set of subcortical structures that include the caudate nucleus and the putamen, which together constitute the neostriatum). Within frontal cortex, two areas play particularly important roles: premotor areas (especially the region of the supplementary motor area [SMA] and pre-SMA) and Broca’s area. Other brain structures also form part of the procedural system network, including portions of both inferior parietal cortex and the cerebellum (including the dentate nucleus).6,89,91,98,99 Finally, the procedural system appears to be closely related to the so-called “dorsal” stream pathway.6,100 The PDH of autism posits that ASD is associated with neural abnormalities of the procedural memory system, resulting in impairments of the various functions that depend on this system. Thus, one should expect motor and cognitive deficits, especially those involving sequencing. In the linguistic domain, one should observe deficits of grammar, across syntax, morphology and phonology. One should also find impairments of any other functions that depend on the brain structures that underlie the procedural system, such as aspects of temporal processing and working memory (whether or not these functions are related to procedural memory).7

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 186 Wednesday, February 15, 2006 3:19 PM

186

Understanding Autism: From Basic Neuroscience to Treatment

Thus the PDH is a brain-based hypothesis. That is, unlike many other accounts of language deficits in autism or other developmental disorders, the PDH is based on the premise that that certain brain structures are abnormal, and that the functions that depend on these structures are therefore likely to be impaired. The nature of the observed impairments is expected to vary with the precise nature of the neural dysfunction, potentially leading to substantial heterogeneity in the linguistic and nonlinguistic impairments in ASD. First of all, because each brain structure in the procedural system plays a somewhat different functional role, the dysfunction of different structures should lead to different types of impairment. For example, abnormalities of structures thought to underlie the acquisition of procedural knowledge, such as the basal ganglia, should yield different behavioral phenotypes than abnormalities of structures that may subserve the execution of procedural skills, which might be the case for certain frontal regions.101 We expect that a portion of the heterogeneity of language (and other cognitive functions) in ASD may be explained by variation across individuals as to which structures are affected and to what degree. Just as abnormalities of different structures within the procedural memory system should lead to behavioral heterogeneity, so should the dysfunction of different portions of structures — especially of those structures that constitute frontal/basal ganglia and frontal/cerebellar circuits. These circuits are composed of parallel and largely functionally segregated “channels” (also referred to as “loops”).102 For example, in the basal ganglia, each channel receives projections at the neostriatum — some channels primarily at the caudate and others at the putamen — from a particular set of cortical and subcortical structures. Each channel then follows the same set of internal connections within the basal ganglia and then projects outward via the thalamus to a particular cortical region (from which it also receives input), primarily in frontal cortex. A somewhat analogous architecture can be found in the circuitry projecting from the cerebellum via the thalamus to frontal cortex. Each of the frontal/ basal ganglia and frontal/ cerebellar channels underlies functions associated with the cortical region to which it projects. For example, the frontal/ basal ganglia channel projecting to the primary motor cortex subserves motor functions. Dysfunction in a given structure in this circuitry (e.g., in the neostriatum) is unlikely to be limited to a single channel. Rather, evidence from neurodegenerative and neurodevelopmental disorders suggests a tendency for dysfunction to co-occur across multiple channels.103,104 However, it is unlikely that exactly the same channels will be affected in all individuals with a particular disorder. Therefore abnormalities in these structures in ASD should result in variability across individuals with respect to the combination of channels that are affected and the severity of their dysfunction. Nevertheless, evidence from other disorders suggests that within a given disorder, certain channels should be more likely than others to be dysfunctional.103 So, although variability with regard to which channels are affected should lead to some functional heterogeneity across subjects, an important degree of similarity among individuals with the disorder is to be expected. Additionally, different types of deficits should be associated with the dysfunction of different pathways within these circuits, in particular with differential dysfunction of the “direct” and “indirect” pathways within the basal ganglia. These two pathways

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 187 Wednesday, February 15, 2006 3:19 PM

Language in Autism

187

have opposing effects on the basal ganglia’s influence on frontal cortex. Via a series of inhibitory and excitatory projections, the direct pathway ultimately disinhibits frontal cortical activity, whereas the indirect pathway ultimately inhibits it. Imbalances between the two pathways can lead to the excessive inhibition or disinhibition of functions that depend on the frontal cortical regions to which the basal ganglia project.105,106 This is thought to explain the inhibited/suppressed (“hypo”) and disinhibited/unsuppressed (“hyper”) motor, grammar, and other behaviors found in various neurodevelopmental as well as neurodegenerative disorders affecting the basal ganglia, including Parkinson’s disease, Huntington’s disease, Tourette syndrome, obsessive-compulsive disorder (OCD) and ADHD.104,105,107,108 For example, Huntington’s disease patients show unsuppressible grammatical rule use (e.g., in affixation, saying walkeded and dugged) as well as unsuppressible movements, whereas Parkinson’s disease patients show the suppression of both grammatical rule use and of movement.108 Given the highly plastic nature of the developing brain, compensation is likely to occur. It has been shown that the functions of abnormal neural tissue can be taken over by similar or proximate intact tissue.109 Thus, abnormalities of specific portions of the striatum or frontal cortex may be compensated for by other portions of these structures in the same, or perhaps even the contralateral, hemisphere. In addition, if a function can be performed by more than one computational mechanism, it could be taken over by a brain structure whose type of computation is distinct from that of the abnormal tissue. Indeed, we posit that the declarative memory system can and will take over certain grammatical functions from the abnormal procedural memory system. The declarative memory system normally subserves the long-term learning, representation, and use of knowledge about facts (conceptual–semantic knowledge, i.e., “semantic memory”) and personally experienced events (“episodic memory”).110,111 The knowledge learned in this system is at least partly (but not completely) explicit — that is, available to conscious awareness.111,112 Medial temporal structures (including the hippocampus and nearby regions such as the parahippocampal gyrus) consolidate new memories, which eventually depend largely on neocortical regions, particularly in the temporal lobes.110,111,113,114 Other brain structures also play a role in declarative memory, including portions of Broca’s region (BA 44, 45, and 47) and frontal polar cortex (BA 10), which seem to underlie aspects of the selection or retrieval of declarative memories,115–117 and the right cerebellum, which may underlie searching for declarative knowledge.118 Importantly, some of these other structures also constitute part of the procedural memory system, suggesting that procedural system abnormalities may be expected to result in impairments of lexical search or retrieval — though not necessarily of the acquisition and organization of lexical knowledge.7 Finally, evidence suggests that the declarative memory system subserves the learning and use not only of fact and event knowledge but also of lexical knowledge. Middle and inferior aspects of the temporal lobe may be particularly important for storing word meanings, whereas superior temporal and temporo-parietal regions may be more important for storing phonological word forms, and possibly also for stored morphological and syntactic structures (e.g., in formulaic speech).6,119

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 188 Wednesday, February 15, 2006 3:19 PM

188

Understanding Autism: From Basic Neuroscience to Treatment

The declarative and procedural memory systems do not function completely independently from each other. Rather, evidence suggests that they interact, yielding both cooperative and competitive learning and processing.6,120–122 First, the two systems can complement each other in acquiring the same or analogous knowledge, including knowledge of sequences. Second, animal and human studies suggest that the two systems also interact competitively. This leads to a “seesaw effect,”6 such that a dysfunction of one system results in an enhancement of the other, or that learning in one system depresses the function of the other. We posit that in ASD the declarative memory system will tend to take over certain grammatical functions from the dysfunctional procedural memory system. In particular, complex structures that can be composed by the grammatical/procedural system (e.g., walk + -ed) in normally developing individuals may simply be stored as chunks (e.g., walked) in lexical/declarative memory in individuals with ASD. Structures that are easier to memorize should be more likely to be stored. Thus, forms that are of higher frequency, shorter, and perhaps less complex should be particularly likely to be memorized. Moreover, ASD individuals should be able to compensate for their grammatical deficits by learning explicit rules in declarative memory, such as “add -ed to the end of the verb when the event has already happened.” Such increases in reliance on the lexical/declarative memory system in ASD should also be reflected in measures at the level of the brain, such as activation changes (compared to controls) in lexical/declarative memory brain structures in functional neuroimaging studies and possibly even in changes in the neuroanatomy of these structures (e.g., in their volumes or areas). Indeed, both behavioral and brain evidence suggests that such declarative-memory compensation takes place in other populations who appear to be afflicted with a grammatical/procedural dysfunction, including children with SLI 7 and agrammatic aphasics.123 Finally, such compensation should depend on the extent to which declarative memory abilities remain intact and should vary with respect to the functionality of this system. Thus, where declarative memory is dysfunctional, such compensation should be less available. In the extreme, a lack of such compensation in the context of a highly dysfunctional procedural memory system would be expected to result in a virtual absence of language. However, we suggest that declarative memory is often (though not necessarily) largely spared in ASD, resulting in a relative sparing of lexical knowledge, though the retrieval or search of this knowledge may tend to be problematic (see preceding text). In the following two sections we provide brief overviews of the status in ASD of language and other functions that depend on the brain structures that underlie the procedural and declarative memory systems. ASD Profile of Procedural System Functions Grammar As we have seen, rule-governed compositional aspects of grammar seem to be largely abnormal in ASD. This pattern holds across language domains, from syntax to morphology to phonology. In contrast, aspects of knowledge or processing in these domains that do not as clearly involve composition, such as irregularly

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 189 Wednesday, February 15, 2006 3:19 PM

Language in Autism

189

inflected forms or individual phonemes, seem to remain relatively spared. These contrasting patterns support the prediction of the PDH that compositional aspects of grammar, which are posited to depend on the procedural system, should be impaired in ASD. Evidence also suggests that ASD individuals compensate for these impairments by relying on lexical/declarative memory. The widespread dependence on formulaic speech indicates a dependence on the use of memorized complex representations, even of structures that would normally be composed in syntax.7,73 It is important to point out that even ASD language that appears to be normal may actually depend more heavily on lexical/declarative memory than it does in typically developing individuals. Although such a dependence may be difficult to detect in spontaneous language samples or in many standardized tests, it can be revealed by a variety of behavioral measures or analyses (e.g., “frequency effect” correlations between the forms’ frequency of use and production or recognition times, suggesting storage of these forms) as well as by neuroimaging techniques.7 Indeed, we have seen that neuroimaging studies of grammatical processing find not only reduced activation and volumes in procedural system brain structures but also increased activation and volumes in certain declarative memory structures, which may reflect an increased dependence on these structures due to compensation and the seesaw effect. The two recent investigations that reported both normal accuracy and faster than normal grammatical processing in syntax and regular morphology are highly intriguing.51,61 It is not yet clear what to make of these findings. One possibility is that these subjects are highly successful at compensating for compositional deficits with the lexical/declarative memory system. The increased temporal lobe activation in the Just et al. study is consistent with this view. Alternatively, these ASD subjects may possess a disinhibited/unsuppressed (hyper) profile, leading to particularly rapid responses, whereas subjects in other studies may instead display an inhibited/suppressed (hypo) profile, leading to impaired accuracy. Indeed, as we will see, both profiles seem to be found in ASD in certain nonlinguistic domains. A hyper profile is consistent with a different interpretation of the underactivation observed in procedural system brain structures during grammatical processing, namely that this reflects particularly efficient rather than deficient processing. Interestingly, the ASD subjects were high functioning in both of the investigations that reported fast processing, suggesting the possibility that such fast processing might be at least somewhat associated with high-functioning ASD.124 Whether or not such an association turns out to hold, it is important to note that very few language studies of ASD report response times (we are not aware of any apart from the two discussed here), suggesting the possibility that such fast performance is actually not uncommon. Nonlinguistic Procedural System Functions Autism is strongly associated with a number of motor and nonmotor deficits that suggest abnormalities of the brain structures of the procedural system. First of all, the processing, as well as the acquisition of both verbal and nonverbal nonmotor sequences, has been reported to be abnormal in ASD, particularly for complex

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 190 Wednesday, February 15, 2006 3:19 PM

190

Understanding Autism: From Basic Neuroscience to Treatment

sequences.88,125,126 For example, abnormalities have been reported in the processing (production or recall) of hierarchically structured sequences, with relative sparing of linear, repetitive sequences.125,126 This finding is particularly relevant to claims of grammatical impairments because grammar (and especially complex syntax) depends on the hierarchical rather than simply linear arrangement of smaller units into larger ones (e.g., the hierarchical combination of words to form phrases and sentences in syntax).127,128 In addition to these findings concerning the processing of sequences, the one published study we are aware of that examined nonmotor sequence learning in a serial reaction time task found no evidence of learning in high-functioning ASD children and adolescents, whereas age- and IQ-matched typically developing control subjects indeed learned the sequence over the course of the task.88 A large number of studies have shown the existence of motor impairments in ASD (for reviews, see Reference 129 and Reference 130). Indeed, the prevalence of such findings has led to the suggestion that motor dysfunction in autism should be considered a core feature of the disorder.131,132 Consistent with procedural memory abnormalities, deficits seem to be much worse for performing complex sequential motor skills than simple motor acts such as finger tapping.44 A large number of studies have reported impairments in the pantomime and imitation of complex actions (for a review, see Reference 129). One particularly comprehensive study found that pantomiming actions (e.g., “Show how you would use a toothbrush.”) was highly impaired, especially for sequential as compared to single actions; imitating actions was somewhat less impaired (but again, more so for sequential than single actions); and actual object use did not show deficits at all.133 This is exactly the pattern seen in ideomotor apraxia, which is linked to left SMA, basal ganglia, and inferior parietal structures.96 Finally, it has also been observed that individuals with ASD have difficulty learning complex sequential motor skills such as dancing or skipping.129 Other motor deficits in ASD are suggestive of abnormalities of specific brain structures of the procedural system. Impairments of balance and other motor-related functions linked to the cerebellum have often been reported in ASD.124,134,135 Still other motor impairments suggest basal ganglia abnormalities and seem to indicate the existence of both hypo and hyper deficits in the disorder. On the one hand, a number of studies have reported hypokinetic (i.e., bradykinetic, or slow) movements similar to those of Parkinson’s patients.124,136 On the other hand, ASD is also strongly associated with unsuppressed motor activity, such as motor (and vocal) tics and stereotypies (repetitive movements or behaviors),137 which have been linked to basal ganglia abnormalities not only in developmental disorders such as Tourette syndrome138,139 but also in ASD itself140,141 (but see Reference 142). Hyperkinetic (choreiform) movements such as isolated jerking of the extremities, which are characteristic of Huntington’s disease, have also been observed in ASD.135 Interestingly, in this last study, all ten subjects showed hyperkinetic movements, whereas none showed movement impairments typical of Parkinson’s disease. Thus, there appears to be at least some separation between ASD groups that show hypo and hyper behaviors. Consistent with such a separation, Mari et al.124 found that lowfunctioning ASD individuals showed bradykinetic movements similar to those found

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 191 Wednesday, February 15, 2006 3:19 PM

Language in Autism

191

in Parkinson’s disease, whereas high-functioning individuals instead showed abnormally fast movements relative to normal controls. This lends credence to the hypothesis discussed earlier that hypo and hyper profiles may also be found in ASD in the domain of language. Finally, other functions associated with the brain structures of the procedural system have also been found to be impaired. (For a discussion of the relationship of the following functions to procedural memory, see Reference 7.) Thus, investigations of ASD have reported deficits of rapid temporal processing143 as well as impairments in the estimation of stimulus duration144 (but see Reference 88). Additionally, working memory has been found to be impaired in ASD in some studies145,146 (but see Reference 62). There have also been reports of deficits in ASD of motion processing, a dorsal stream function.147,148 Indeed, one study147 found deficits in motion processing, but not in form processing, which depends on the ventral stream (which is closely related to declarative memory6). Further examination in ASD of these and other functions that may depend on procedural system brain structures seems warranted. ASD Profile of the Declarative Memory System Lexical Memory As we have seen, evidence suggests that word learning and lexical knowledge remain largely normal in ASD. For example, individuals with the disorder show intact performance at receptive lexical processing tasks. Similarly, expressive lexical abilities seem to be spared in single-word production tasks, though impairments have been found in both rapid naming and verbal fluency tasks, suggesting deficits in lexical retrieval or search, especially under speeded conditions. This profile of lexical abilities is consistent with a relatively normal lexical memory, accompanied by abnormalities to brain structures such as frontal and cerebellar regions that underlie lexical search and retrieval as well as aspects of rapid processing.6,7 Conceptual Knowledge Conceptual knowledge appears to be largely spared in ASD. This seems to hold for both individual word meanings and their categorical organization.65,149–152 For example, ASD children have been found to show a normal pattern of prototypicality ratings of members of numerous categories at both basic and superordinate levels, such as chairs, furniture, dogs, and animals.149 Additionally, semantic priming (e.g., nurse will be responded to faster or more accurately following doctor than following table due to semantic associations between doctor and nurse) has been found to be normal,153 suggesting intact conceptual–semantic representations in ASD. Another study reported normal interference in a Stroop task with a range of word categories, including not only color names but also concrete and abstract words.150 Similarly, in a word–picture matching task, ASD subjects did not differ from controls on either concrete or abstract words.151 However, not all conceptual categories seem to be spared in ASD. Specifically, ASD individuals have been found to show impairments at processing words that are

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 192 Wednesday, February 15, 2006 3:19 PM

192

Understanding Autism: From Basic Neuroscience to Treatment

related to mental or emotional states5,151 (but see Reference 154). Importantly, these impairments may be explained in terms of deficits of emotion or theory of mind.5,155 Indeed, ASD deficits in the conceptual knowledge of verbs that are related to mental states (“cognition” verbs such as know, think) were found to be related to measures of theory of mind (but not to grammatical performance).5 We are aware of only one functional neuroimaging study of conceptual processing in ASD.156 This fMRI study examined conceptual–semantic processing (“Judge if a word is positive or negative.”) as compared to perceptual processing (“Judge if a word is upper or lower case.”). In this comparison, control subjects robustly activated regions in the left BA 45, left BA 47 (and, borderline significantly, right BA 47), and left medial frontal cortex (two clusters of activation, one on the border of BA 6 and BA 8 — i.e., pre-SMA — and one more anterior in BA 8), as well as in the right cerebellum. In contrast, ASD subjects showed only minimal frontal activation, with a small cluster in left BA 47. They showed no activation in the cerebellum or in any other frontal region, including left BA 45, medial frontal cortex, or right BA 47. Moreover, and unlike the controls, they activated the posterior portion of the superior temporal sulcus/middle temporal gyrus. This is strikingly similar to the neuroimaging patterns discussed earlier and reinforces the view that the temporal lobe structures of lexical/declarative memory may not only be spared, but in fact might be relied on to a greater degree than in typically developing subjects, even for conceptual processing. Learning in Declarative Memory An increasing body of evidence suggests that the learning of verbal and nonverbal knowledge in declarative memory (i.e., the acquisition of knowledge in semantic memory) is essentially spared in ASD.125,145,157,158 (Tasks probing immediate recall or recognition will not be discussed here as they may reflect processing in shortterm or working memory rather than learning in declarative memory, which can perhaps be most clearly ascertained with recall or recognition after a delay; we only discuss such post-delay performance here.) First of all, numerous studies have suggested that learning in “rote memory” (i.e., memorizing individual items such as telephone numbers or addresses) is a strength in ASD.145,159,160 Second, ASD subjects show normal performance on tasks of paired-associate learning (e.g., presenting a pair of words that are studied together, then later prompting with the first word of the pair and asking the subject to recall the second word).44,49,157,161 Third, normal performance is also observed on cued recall measures after a delay (e.g., after the presentation of a list of words, subjects can be successfully cued with the initial sounds of the word, such as fr to cue the recall of fruit).161–163 However, evidence also suggests the existence of episodic memory impairments in ASD.125,145,158,164 Thus, memory for recent, personally experienced events is consistently reported to be impaired in ASD relative to controls.165–168 The basis of this episodic memory impairment in the context of spared semantic memory is not yet clear, although it has been suggested that these impairments may reflect the particular dependence of episodic memory on frontal structures.158 Note that in any case such impairments would not be expected to impact the acquisition of lexical knowledge, which should depend on semantic rather than episodic memory.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 193 Wednesday, February 15, 2006 3:19 PM

Language in Autism

193

The Neurobiology of Procedural and Declarative Memory Brain Structures in ASD Although studies of the neurobiology of procedural and declarative memory brain structures have produced quite a few inconsistent results, some patterns are beginning to emerge (for overviews of these and other brain structures in ASD, see References 169, 170, and 171). Of particular interest here are abnormalities of left frontal cortex, and especially Broca’s area, which have consistently been found in studies that have examined this region.172–176 Interestingly, neuroanatomical abnormalities of frontal–cerebellar circuitry have also been implicated in several recent reports.175–177 Moreover, postmortem studies reliably report reduced numbers of Purkinje cells in the cerebellum.170 Nevertheless, results from experiments probing the cerebellum with other methods, such as structural MRI, have been more variable (for a recent review, see Reference 171). Studies of other brain structures of the procedural and declarative memory systems, including the hippocampus and the basal ganglia, have also produced somewhat inconsistent results, with some (but not other) studies suggesting abnormalities (see reviews mentioned earlier and discussion in Reference 142). (Note that the hippocampus was not examined in any of the language-related imaging studies discussed earlier, and the basal ganglia were only examined in one of these studies.) Further investigations are clearly needed to clarify the inconsistencies in these (and other) structures and to examine the functional consequences for language of any observed structural abnormalities.

SUMMARY AND CONCLUSION In this chapter we have argued that language and communicative impairments in autism can be better understood in light of integrative explanatory frameworks that examine these deficits in the broader context of brain and behavior in ASD. First, impairments in pragmatic language abilities are argued to be related to theory of mind impairments. Second, we show that ASD is associated with impairments of compositional aspects of grammar, across linguistic domains, and argue that these impairments are related to abnormalities of brain structures of the procedural memory system, including at least Broca’s area. In contrast, lexical and declarative (especially semantic) memory appear to be a relative strength, and evidence suggests that these capacities may be used to compensate for deficient grammatical processing in ASD. The two classes of language and communication impairments (i.e., of pragmatics and grammar) have been presented as independent deficits. However, possible relations between the two have yet to be explored. Of particular interest is evidence suggesting that frontal cortex, including Broca’s area, is implicated in both theory of mind and grammatical/procedural functions. Intriguingly, it has been suggested that theory of mind depends on the dorsal stream,83 suggesting another potential neuroanatomical link between the two domains. Thus, it may be worthwhile to explore the possibility that pragmatic and grammatical deficits in ASD result from distinct functional deficits, which nevertheless both ultimately depend on the same or related underlying brain structures.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 194 Wednesday, February 15, 2006 3:19 PM

194

Understanding Autism: From Basic Neuroscience to Treatment

In conclusion, autism is associated with a particular profile of impaired (pragmatic, grammatical) and spared (lexical) language abilities. The status of these abilities in ASD can be at least partly explained in terms of their dependence on neurocognitive substrates that also subserve nonlanguage functions, specifically theory of mind (pragmatics), procedural memory (grammar), and declarative memory (lexicon). Whether or how the neurocognitive abnormalities underlying theory of mind and procedural memory are related to each other or to other abnormalities in ASD (e.g., underconnectivity, 51,178 weak central coherence,179 or impaired executive function180) remains to be examined. Overall, the explanatory theories presented here not only provide wide-ranging accounts of linguistic and nonlinguistic data but also generate new questions and new directions for research.

ACKNOWLEDGMENTS Support was provided to MTU by NSF SBR-9905273, NIH R01 MH58189, and NIH R01 HD049347, and to MTU and MW by research grants from the National Alliance for Autism Research and the Mabel Flory Trust. Support was provided to HTF by U19 DC 03610, which is part of the NICHD/NIDCD funded CPEA, and U54 MH 66398, which is part of the NIH funded STAART Centers, and a grant from the Nancy Lurie Marks Family Foundation. Please address correspondence to Michael T. Ullman ([email protected]).

REFERENCES 1. APA, DSM-IV, 4th ed., American Psychiatric Association, Washington, D.C., 1994. 2. Lord, C., Risi, S., and Pickles, A., Trajectory of language development, in Developmental Language Disorders, Eds., Rice, M.L. and Warren, S.F., Lawrence Erlbaum Associates, Mahwah, NJ, 2004, pp. 7–29. 3. Baron-Cohen, S., Social and pragmatic deficits in autism: cognitive or affective?, Journal of Autism and Developmental Disorders 18, 379–402, 1988. 4. Tager-Flusberg, H., What language reveals about the understanding of minds in children with autism, in Understanding Other Minds: Perspectives from Autism, Eds., Baron-Cohen, S., Tager-Flusberg, H., and Cohen, D.J., Oxford University Press, Oxford, 1993, pp. 138–157. 5. Tager-Flusberg, H., Language and understanding minds: connections in autism, in Understanding Other Minds, 2nd ed., Eds., Baron-Cohen, S., Tager-Flusberg, H., and Cohen, D.J., Oxford University Press, New York, 2000, pp. 124–149. 6. Ullman, M.T., Contributions of memory circuits to language: the declarative/procedural model, Cognition 92(1–2), 231–270, 2004. 7. Ullman, M.T. and Pierpont, E.I., Specific language impairment is not specific to language: the procedural deficit hypothesis, Cortex 41(3), 399–433, 2005. 8. Mundy, P., Sigman, M., and Kasari, C., A longitudinal study of joint attention and language development in autistic children, Journal of Autism and Developmental Disorders 20, 115–123, 1990. 9. Mundy, P., Sigman, M., Ungerer, J., and Sherman, T., Defining the social deficits in autism: The contribution of nonverbal communication measures, Journal of Child Psychology and Psychiatry 27, 657–669, 1986.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 195 Wednesday, February 15, 2006 3:19 PM

Language in Autism

195

10. Mundy, P., Sigman, M., and Kasari, C., The theory of mind and joint-attention deficits in autism, in Understanding Other Minds: Perspectives from Autism, Eds., BaronCohen, S., Tager-Flusberg, H., and Cohen, D.J., Oxford University Press, Oxford, 1993, pp. 181–203. 11. Charman, T., Drew, A., Baird, C., and Baird, G., Measuring early language development in preschool children with autism spectrum disorder using the MacArthur communicative development inventory, Journal of Child Language 30, 213–236, 2003. 12. Mundy, P., Sigman, M., and Kasari, C., Nonverbal communication, developmental level and symptom presentation in autism, Development and Psychopathology 6, 389–401, 1994. 13. Grice, H.P., Logic and conversation, in Syntax and Semantics, Eds., Cole, P. and Morgan, J.L., Academic Press, New York, 1975. 14. Wetherby, A., Ontogeny of communication functions in autism, Journal of Autism and Developmental Disorders 16, 295–316, 1986. 15. Wetherby, A. and Prutting, C., Profiles of communicative and cognitive-social abilities in autistic children, Journal of Speech and Hearing Research 27, 364–377, 1984. 16. Ball, J., A Pragmatic Analysis of Autistic Children’s Language with respect to Aphasic and Normal Language Development, Doctoral dissertation, Melbourne University, 1978. 17. Loveland, K., Landry, S., Hughes, S., Hall, S., and McEvoy, R., Speech acts and the pragmatic deficits of autism, Journal of Speech and Hearing Research 31, 593–604, 1988. 18. Rollins, P., A Case Study of the Development of Communicative Skills for Six Autistic Children, Doctoral dissertation, Harvard University, 1994. 19. Capps, L., Kehres, J., and Sigman, M., Conversational abilities among children with autism and children with developmental delays, Autism 2, 325–344, 1998. 20. Ghaziuddin, M. and Gerstein, L., Pedantic speaking style differentiates Asperger syndrome from high-functioning autism, Journal of Autism and Developmental Disorders 26, 585–595, 1996. 21. Ramberg, C., Ehlers, S., Nyden, A., Johansson, M., and Gillberg, C., Language and pragmatic functions in school-age children on the autism spectrum, European Journal of Disorders of Communication 31, 387–413, 1996. 22. Adams, C., Green, J., Gilchrist, A., and Cox, A., Conversational behaviour of children with Asperger syndrome and conduct disorder, Journal of Child Psychology and Psychiatry 43, 679–690, 2002. 23. Fine, J., Bartolucci, G., Szatmari, P., and Ginsberg, G., Cohesive discourse in pervasive developmental disorders, Journal of Autism and Developmental Disorders 24, 315–330, 1994. 24. Loveland, K., Tunali, B., Kelley, M.L., and McEvoy, R.E., Referential communication and response adequacy in autism and Down’s syndrome, Applied Psycholinguistics 10, 301–313, 1989. 25. Surian, L., Baron-Cohen, S., and Van der Lely, H.K.J., Are children with autism deaf to Grician Maxims?, Cognitive Neuropsychiatry 1(1), 55–71, 1996. 26. Paul, R. and Cohen, D.J., Responses to contingent queries in adults with mental retardation and pervasive developmental disorders, Applied Psycholinguistics 5, 349–357, 1984. 27. Volden, J., Conversational repair in speakers with autism spectrum disorder, International Journal of Language and Communication Disorders 39, 171–189, 2004. 28. Loveland, K. and Tunali, B., Narrative language in autism and the theory of mind hypothesis: a wider perspective, in Understanding Other Minds: Perspectives from Autism, Eds., Baron-Cohen, S., Tager-Flusberg, H., and Cohen, D.J., Oxford University Press, Oxford, 1993.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 196 Wednesday, February 15, 2006 3:19 PM

196

Understanding Autism: From Basic Neuroscience to Treatment

29. Tager-Flusberg, H., “Once upon a ribbit”: stories narrated by autistic children, British Journal of Developmental Psychology 13, 45–59, 1995. 30. Tager-Flusberg, H. and Sullivan, K., Attributing mental states to story characters: a comparison of narratives produced by autistic and mentally retarded individuals, Applied Psycholinguistics 16, 241–256, 1995. 31. Shriberg, L., Paul, R., McSweeney, J.L., Klin, A., Cohen, D., and Volkmar, F.R., Speech and prosody characteristics of adolescents and adults with high-functioning autism and asperger syndrome, Journal of Speech Language and Hearing Research 44(5), 1097–1115, 2001. 32. Happé, F., Communicative competence and theory of mind in autism: a test of relevance theory, Cognition 48, 101–119, 1993. 33. Happé, F., An advanced test of theory of mind: Understanding of story characters’ thoughts and feelings by able autistic, mentally handicapped and normal children and adults, Journal of Autism and Developmental Disorders 24, 129–154, 1994. 34. MacKay, G. and Shaw, A., A comparative study of figurative language in children with autistic spectrum disorders, Child Language Teaching and Therapy 20, 13–32, 2004. 35. Minshew, N.J., Sweeney, J.A., and Furman, J.M., Evidence for a primary neocortical systems abnormality in autism, Society for Neuroscience Abstracts 21, 735, 1995. 36. Kerbel, D. and Grunwell, P., A study of idiom comprehension in children with semantic-pragmatic difficulties. Part II: Between-groups results and discussion, International Journal of Language and Communication Disorders 33(1), 23–44, 1998. 37. Martin, I. and McDonald, S., An exploration of causes of non-literal language problems in individuals with Asperger syndrome, Journal of Autism and Developmental Disorders 34, 311–328, 2004. 38. Baron-Cohen, S., Hey! It was just a joke! Understanding propositions and propositional attitudes by normally developing children and children with autism, Israel Journal of Psychiatry 34, 174–178, 1997. 39. Jolliffe, T. and Baron-Cohen, S., The strange stories test: a replication with highfunctioning adults with autism or Asperger syndrome, Journal of Autism and Developmental Disorders 29, 395–406, 1999. 40. Mitchell, P., Saltmarsh, R., and Russell, H., Overly literal interpretations of speech in autism: understanding that messages arise from minds, Journal of Child Psychology and Psychiatry 38, 685–691, 1997. 41. Chomsky, N., Aspects of the Theory of Syntax, The MIT Press, Cambridge, MA, 1965. 42. Pinker, S., The Language Instinct, William Morrow, New York, 1994. 43. Tager-Flusberg, H., Sentence comprehension in autistic children, Applied Psycholinguistics 2, 5–24, 1981. 44. Minshew, N.J., Goldstein, G., and Siegel, D., Neuropsychologic functioning in autism: profile of a complex information processing disorder, Journal of the International Neuropsychological Society 3(4), 303–316, 1997. 45. Minshew, N.J., Goldstein, G., Taylor, G.H., and Siegel, D.J., Academic achievement in high functioning autistic individuals, Journal of Clinical and Experimental Neuropsychology 16(2), 261–270, 1994. 46. Kjelgaard, M.M. and Tager-Flusberg, H., An investigation of language impairment in autism: implications for genetic subgroups, Language and Cognitive Processes 16(2), 287–308, 2001. 47. Müller, R.A., Behen, M.E., Rothermel, R.D., Chugani, D.C., Muzik, O., Mangner, T.J., and Chugani, H.T., Brain mapping of language and auditory perception in highfunctioning autistic adults: a PET study, Journal of Autism and Developmental Disorders 29(1), 19–31, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 197 Wednesday, February 15, 2006 3:19 PM

Language in Autism

197

48. Minshew, N.J., Goldstein, G., and Siegel, D., Speech and language in high-functioning autistic individuals, Neuropsychology 9, 255–261, 1995. 49. Minshew, N.J., Goldstein, G., Muenz, L.R., and Payton, J.B., Neuropsychological functioning in nonmentally retarded autistic individuals, Journal of Clinical and Experimental Neuropsychology 14(5), 749–761, 1992. 50. Prior, M.R. and Hall, L.C., Comprehension of transitive and intransitive phrases by autistic, retarded, and normal children, Journal of Communication Disorders 12(2), 103–111, 1979. 51. Just, M., Cherkassky, V., Keller, T., and Minshew, N., Cortical activation and synchronization during sentence comprehension in high-functioning autism: evidence of underconnectivity, Brain 127(Pt. 8), 1811–1821, 2004. 52. Scarborough, H.S., Rescorla, L., Tager-Flusberg, H., Fowler, A.E., and Sudhalter, V., The relation of utterance length to grammatical complexity in normal and languagedisordered groups, Applied Psycholinguistics: Psycholigcal Studies of Language Processes 12(1), 23–45, 1991. 53. Eigsti, I.-M. and Bennetto, L., Syntactic abilities and memory functions in young children with autism, under review. 54. Tager-Flusberg, H., Calkins, S., Nolin, T., Baumberger, T., Anderson, M., and Chadwick-Dias, A., A longitudinal study of language acquisition in autistic and Down syndrome children, Journal of Autism and Developmental Disorders 20(1), 1–21, 1990. 55. Bartolucci, G., Pierce, S.J., and Streiner, D., Cross-sectional studies of grammatical morphemes in autistic and mentally retarded children, Journal of Autism and Developmental Disorders 10(1), 39–50, 1980. 56. Howlin, P., The acquisition of grammatical morphemes in autistic children: a critique and replication of the findings of Bartolucci, Pierce, and Streiner, 1980, Journal of Autism and Developmental Disorders 14(2), 127–136, 1984. 57. Tager-Flusberg, H. and Calkins, S., Does imitation facilitate the acquisition of grammar? Evidence from a study of autistic, Down's syndrome and normal children, Journal of Child Language 17(3), 591–606, 1990. 58. Botting, N. and Conti-Ramsden, G., Autism, primary pragmatic difficulties, and specific language impariment: can we distinguish them using psycholinguistic markers?, Developmental Medicine and Child Neurology 45, 515–524, 2003. 59. Condouris, K., Evancie, L., and Tager-Flusberg, H., Tense error patterns in children with autism and children with SLI, International Meeting for Autism Research, Orlando, FL, 2002. 60. Roberts, J.A., Rice, M.L., and Tager-Flusberg, H., Tense marking in children with autism, Applied Psycholinguistics 25, 429–444, 2004. 61. Walenski, M., Mostofsky, S.H., Gidley Larson, J., and Ullman, M.T., Fast Grammatical Processing in autism, in Twelfth Annual Meeting of the Cognitive Neuroscience Society, New York, 2005. 62. Ozonoff, S. and Strayer, D.L., Further evidence of intact working memory in autism, Journal of Autism Developmental Disorders 31(3), 257–263, 2001. 63. Lord, C. and Paul, R., Language and communication in autism, in Handbook of Autism and Pervasive Developmental Disorders, 1997, pp. 195–225. 64. Tager-Flusberg, H., Do autism and specific language impairment represent overlapping language disorders, in Developmental Language Disorders, Eds., Rice, M.L. and Warren, S.F., 2004, pp. 31–52. 65. Tager-Flusberg, H., The conceptual basis for referential word meaning in children with autism, Child Development 56(5), 1167–1178, 1985.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 198 Wednesday, February 15, 2006 3:19 PM

198

Understanding Autism: From Basic Neuroscience to Treatment

66. Rumsey, J.M., Neuropsychological divergence of high-level autism and severe dyslexia, Journal of Autism and Childhood Disorders 20(2), 155–168, 1990. 67. Boucher, J., Word Fluency in High-Functioning Autistic Children, Journal of Autism and Childhood Schizophrenia, 18(4), 637–645, 1988. 68. Müller, R.-A., Chugani, D.C., Behen, M.E., Rothermel, R.D., Muzik, O., Chakraborty, P.K., and Chugani, H.T., Impairment of dento-thalamo-cortical pathway in autistic men: language activation data from positron emission tomography, Neuroscience Letters 245(1), 1–4, 1998. 69. De Fossé, L., Hodge, S., Makris, N., Kennedy, D., Caviness, V., McGrath, L., Steele, S., Ziegler, D., Herbert, M., Frazier, J., Tager-Flusberg, H., and Harris, G., Language association cortex asymmetry in autism and specific language impairment, Annals of Neurology 56, 757–766, 2004. 70. Howlin, P. and Rutter, M., Treatment of Autistic Children, John Wiley & Sons, Chichester, U.K., 1987. 71. Bartak, L., Rutter, M., and Cox, A., A comparative study of infantile autism and specific developmental receptive language disorder, British Journal of Psychiatry 126, 127–145, 1975. 72. Dobbinson, S., Perkins, M., and Boucher, J., The interactional significance of formulas in autistic language, Clinical Linguistics and Phonetics 17(4–5), 299–307, 2003. 73. Wray, A. and Perkins, M.R., The functions of formulaic language: an integrated model, Language and Communication 20, 1–28, 2000. 74. Tager-Flusberg, H., A psychological approach to understanding the social and language impairments in autism, International Review of Psychiatry 11, 325–334, 1999. 75. Tager-Flusberg, H., Exploring the relationships between theory of mind and socialcommunicative functioning in children with autism, in Individual Differences in Theory of Mind: Implications for Typical and Atypical Development, Eds., Repacholi, B. and Slaughter, V., Psychology Press, Hove, UK, 2003. 76. Baron-Cohen, S., Leslie, A.M., and Frith, U., Does the autistic child have a theory of mind?, Cognition 21, 37–46, 1985. 77. Baron-Cohen, S., Tager-Flusberg, H., and Cohen, D.J., Understanding Other Minds: Perspectives from Autism, Oxford University Press, Oxford, 1993. 78. Baron-Cohen, S., From attention-goal psychology to belief-desire psychology: the development of a theory of mind, and its dysfunction, in Understanding Other Minds: Perspectives from Autism, Baron-Cohen, S., Tager-Flusberg, H., and Cohen, D.J., Oxford University Press, Oxford, 1993, pp. 60–82. 79. Hobson, R.P., Understanding persons: the role of affect, in Understanding Other Minds: Perspectives from Autism, Baron-Cohen, S., Tager-Flusberg, H., and Cohen, D.J., Oxford University Press, Oxford, 1993, pp. 204–227. 80. Sigman, M. and Ruskin, E., Continuity and Change in the Social Competence of Children with Autism, Down Syndrome and Developmental Delays, Blackwell Publishers, Malden, MA, 1999. 81. Sperber, D. and Wilson, D., Relevance: Communication and Cognition, Harvard University Press, Cambridge, MA, 1986. 82. Hale, C.M. and Tager-Flusberg, H., Social communication in children with autism: the relationship between theory of mind in discourse development, Autism, 9(2), 157–178, 2005. 83. Frith, C.D. and Frith, U., Interacting minds — a biological basis, Science 286, 1692–1695, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 199 Wednesday, February 15, 2006 3:19 PM

Language in Autism

199

84. Happé, F., Theory of mind and the self, Annals of the New York Academy of Sciences 1001, 134–144, 2003. 85. Baron-Cohen, S., Ring, H.A., Wheelwright, S., Bullmore, E.T., Brammer, M.J., Simmons, A., and Williams, S.C., Social intelligence in the normal and autistic brain: an fMRI study, European Journal of Neuroscience 11, 1891–1898, 1999. 86. Happé, F., Ehlers, S., Fletcher, P., Frith, U., Johansson, M., Gillberg, C., Dolan, R., Frackowiak, R., and Frith, C., “Theory of mind” in the brain: evidence from a PET scan study of Asperger syndrome, Neuroreport 8, 197–201, 1996. 87. Frith, C., What do imaging studies tell us about the neural basis of autism?, Novartis Foundation Symposium 251, 149–166, 2003. 88. Mostofsky, S.H., Goldberg, M.C., Landa, R.J., and Denckla, M.B., Evidence for a deficit in procedural learning in children and adolescents with autism: implications for cerebellar contribution, Journal of the International Neuropsychological Society 6(7), 752–759, 2000. 89. Schacter, D.L. and Tulving, E., Memory Systems 1994, The MIT Press, Cambridge, MA, 1994. 90. Squire, L.R. and Knowlton, B.J., The medial temporal lobe, the hippocampus, and the memory systems of the brain, in The New Cognitive Neurosciences, Ed., Gazzaniga, M.S., MIT Press, Cambridge, MA, 2000, pp. 765–780. 91. Willingham, D.B., A neuropsychological theory of motor skill learning, Psychological Review 105(3), 558–584, 1998. 92. Graybiel, A.M., Building action repertoires: Memory and learning functions of the basal ganglia, Current Opinion in Neurobiology 5, 733–741, 1995. 93. Knowlton, B.J. and Squire, L.R., Artificial grammar learning depends on implicit acquisition of both abstract and exemplar-specific information, Journal of Experimental Psychology: Learning Memory and Cognition 22(1), 169–181, 1996. 94. Poldrack, R.A., Prabhakaran, V., Seger, C.A., and Gabrieli, J.D., Striatal activation during acquisition of a cognitive skill, Neuropsychology 13(4), 564–574, 1999. 95. Ullman, M.T., A neurocognitive perspective on language: the declarative/procedural model, Nature Reviews Neuroscience 2, 717–726, 2001. 96. Heilman, K.M., Watson, R.T., and Rothi, L.G., Disorders of skilled movements: limb apraxia, in Behavioral Neurology and Neuropsychology, Eds., Feinberg, T.E. and Farah, M.J., McGraw-Hill, New York, 1997, pp. 227–235. 97. Schluter, N.D., Krams, M., Rushworth, M.F.S., and Passingham, R.E., Cerebral dominance for action in the human brain: the selection of actions, Neuropsychologia 39(2), 105–113, 2001. 98. Squire, L.R. and Zola, S.M., Structure and function of declarative and nondeclarative memory systems, Proceedings of the National Academy of Science USA 93, 13515–13522, 1996. 99. Hikosaka, O., Sakai, K., Nakahara, H., Lu, X., Miyachi, S., Nakamura, K., and Rand, M.K., Neural mechanisms for learning of sequential procedures, in The New Cognitive Neurosciences, Ed., Gazzaniga, M.S., MIT Press, Cambridge, MA, 2000, pp. 553–572. 100. Goodale, M.A., Perception and action in the human visual system, in The New Cognitive Neurosciences, Ed., Gazzaniga, M.S., MIT Press, Cambridge, MA, 2000, pp. 365–378. 101. Ullman, M.T., Is Broca’s area part of a basal ganglia thalamocortical circuit?, Cortex, in press. 102. Middleton, F.A. and Strick, P.L., Basal ganglia and cerebellar loops: motor and cognitive circuits, Brain Research 31(2–3), 236–250, 2000.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 200 Wednesday, February 15, 2006 3:19 PM

200

Understanding Autism: From Basic Neuroscience to Treatment

103. Jankovic, J. and Tolosa, E.E., Parkinson’s Disease and Movement Disorders, Williams and Wilkins, Baltimore, MD, 1993. 104. Bradshaw, J.L., Developmental Disorders of the Frontostriatal System, Psychology Press, Hove, East Sussex, Great Britain, 2001. 105. Young, A.B. and Penney, J.B., Biochemical and functional organization of the basal ganglia, in Parkinson’s Disease and Movement Disorders, 2nd ed., Eds., Jankovic, J. and Tolosa, E., Williams and Wilkins, Baltimore, MD, 1993, pp. 1–11. 106. Albin, R.L., Young, A.B., and Penney, J.B., The functional anatomy of basal ganglia disorders, Trends in Neurosciences 12(10), 366–375, 1989. 107. Middleton, F.A. and Strick, P.L., Basal ganglia output and cognition: evidence from anatomical, behavioral, and clinical studies, Brain and Cognition 42(2), 183–200, 2000. 108. Ullman, M.T., Corkin, S., Coppola, M., Hickok, G., Growdon, J.H., Koroshetz, W.J., and Pinker, S., A neural dissociation within language: evidence that the mental dictionary is part of declarative memory, and that grammatical rules are processed by the procedural system, Journal of Cognitive Neuroscience 9(2), 266–276, 1997. 109. Merzenich, M.M., Recanzone, G., Jenkins, W.M., Allard, T.T., and Nudo, R.J., Cortical representational plasticity, in Neurobiology of Neocortex, Eds., Rakic, P. and Singer, W., John Wiley & Sons, Chichester, U.K., 1988, pp. 41–67. 110. Squire, L.R., Clark, R.E., and Bayley, P.J., Medial temporal lobe function and memory, in The Cognitive Neurosciences, 3rd ed., Ed., Gazzaniga, M.S., MIT Press, Cambridge, MA, 2004, pp. 691–708. 111. Eichenbaum, H. and Cohen, N.J., From Conditioning to Conscious Recollection: Memory Systems of the Brain, Oxford University Press, New York, 2001. 112. Chun, M.M., Contextual cueing of visual attention, Trends in Cognitive Sciences 4(5), 170–178, 2000. 113. Hodges, J.R. and Patterson, K., Semantic memory disorders, Trends in Cognitive Sciences 1(2), 68–72, 1997. 114. Martin, A., Ungerleider, L.G., and Haxby, J.V., Category specificity and the brain: the sensory/motor model of semantic representations of objects, in The Cognitive Neurosciences, Ed., Gazzaniga, M.S., The MIT Press, Cambridge, MA, 2000, pp. 1023–1036. 115. Binder, J.R., Frost, J.A., Hammeke, T.A., Rao, S., and Cox, R.W., Function of the left planum temporale in auditory and linguisitic production, Brain 119, 1239–1247, 1996. 116. Thompson-Schill, S.L., D’Esposito, M., Aguirre, G.K., and Farah, M.J., Role of left inferior prefrontal cortex in retrieval of semantic knowledge: a reevaluation, Proceedings of the National Academy of Science USA 94(26), 14792–14797, 1997. 117. Buckner, R.L. and Wheeler, M.E., The cognitive neuroscience of remembering, Nature Review Neuroscience 2(9), 624–634, 2001. 118. Desmond, J.E. and Fiez, J.A., Neuroimaging studies of the cerebellum: language, learning, and memory, Trends in Cognitive Sciences 2(9), 355–362, 1998. 119. Ullman, M.T., A cognitive neuroscience perspective on second language acquisition: the declarative/procedural model, in Mind and Context in Adult Second Language Acquisition, Ed., Sanz, C., Georgetown University Press, Washington, D.C., 2005, pp. 141–178. 120. Poldrack, R.A. and Packard, M.G., Competition among multiple memory systems: converging evidence from animal and human brain studies, Neuropsychologia 41(3), 245–251, 2003. 121. Packard, M.G. and Knowlton, B.J., Learning and memory functions of the basal ganglia, Annual Review of Neuroscience 25, 563–593, 2002.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 201 Wednesday, February 15, 2006 3:19 PM

Language in Autism

201

122. Fletcher, P.C., Zafiris, O., Frith, C.D., Honey, R.A.E., and Corlett, P.R., On the benefits of not trying: brain activity and connectivity reflecting the interactions of explicit and implicit sequence learning, Cerebral Cortex, 15(7), 1002–1015, 2004. 123. Drury, J.E. and Ullman, M. T., The memorization of complex forms in aphasia: implications for recovery, Brain and Language 83, 139–141, 2002. 124. Mari, M., Castiello, U., Marks, D., Marraffa, C., and Prior, M., The reach-to-grasp movement in children with autism spectrum disorder, Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 358(1430), 393–403, 2003. 125. Boucher, J., “Lost in a sea of time”: timing-parsing and autism, in Time and Memory: Issues in Philosophy and Psychology, Eds., Perkins, M. and Howard, S., Clarendom Press, Oxford, 2001, pp. 111–135. 126. Hermelin, B. and Frith, U., Psychological studies of childhood autism: can autistic children make sense of what they see and hear?, Focus on Autistic Behavior 6(1), 6–13, 1991. 127. Chomsky, N., Syntactic Structures, The Hague, Mouton, 1957. 128. Chomsky, N., The Minimalist Program, The MIT Press, Cambridge, MA, 1995. 129. Smith, I.M. and Bryson, S.E., Imitation and action in autism: a critical review, Psychological Bulletin 116(2), 259–273, 1994. 130. Leary, M.R. and Hill, D.A., Moving on: autism and movement disturbance, Mental Retardation 34(1), 39–53, 1996. 131. Rogers, S. and Pennington, B.F., A theoretical approach to the deficits in infantile autism, Development and Psychopathology 3, 137–163, 1991. 132. Müller, R.A., Pierce, K., Ambrose, J.B., Allen, G., and Courchesne, E., Atypical patterns of cerebral motor activation in autism: a functional magnetic resonance study, Biological Psychiatry 49(8), 665–676, 2001. 133. Rogers, S.J., Bennetto, L., McEvoy, R., and Pennington, B.F., Imitation and pantomime in high-functioning adolescents with autism spectrum disorders, Child Development 67(5), 2060–2073, 1996. 134. Pierce, K. and Courchesne, E., Evidence for a cerebellar role in reduced exploration and stereotyped behavior in autism, Biological Psychiatry 49(8), 655–664, 2001. 135. Jones, V. and Prior, M., Motor imitation abilities and neurological signs in autistic children, Journal of Autism Developmental Disorders 15(1), 37–46, 1985. 136. Vilensky, J.A., Damasio, A.R., and Maurer, R.G., Gait disturbances in patients with autistic behavior: a preliminary study, Archives of Neurology 38(10), 646–649, 1981. 137. Ringman, J. and Jankovic, J., Occurrence of tics in Asperger's syndrome and autistic disorder, Journal of Child Neurology 15(6), 394–400, 2000. 138. Swerdlow, N.R. and Young, A.B., Neuropathology in Tourette syndrome: an update, Advances in Neurology 85, 151–161, 2001. 139. Mink, J., Neurobiology of basal ganglia circuits in Tourette syndrome: faulty inhibition of unwanted motor patterns?, Advances in Neurology 85, 113–122, 2001. 140. Sears, L.L., Vest, C., Mohamed, S., Bailey, J., Ranson, B.J., and Piven, J., An MRI study of the basal ganglia in autism, Neuropsychopharmacology and Biological Psychiatry 23(4), 613–624, 1999. 141. Hollander, E., Anagnostou, E., Chaplin, W., Esposito, K., Haznedar, M., Licalzi, E., Wasserman, S., Soorya, L., and Buchsbaum, M., Striatal volume on magnetic resonance imaging and repetitive behaviors in autism, Biological Psychiatry, 2005. 142. Hardan, A.Y., Kilpatrick, M., Keshavan, M.S., and Minshew, N.J., Motor performance and anatomic magnetic resonance imaging (MRI) of the basal ganglia in autism, Journal of Child Neurology 18(5), 317–324, 2003.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 202 Wednesday, February 15, 2006 3:19 PM

202

Understanding Autism: From Basic Neuroscience to Treatment

143. Oram Cardy, J.E., Flagg, E.J., Roberts, W., Brian, J., and Roberts, T.P., Magnetoencephalography identifies rapid temporal processing deficit in autism and language impairment, Neuroreport 16(4), 329–332, 2005. 144. Szelag, E., Kowalska, J., Galkowski, T., and Poppel, E., Temporal processing deficits in high-functioning children with autism, Br J Psychol 95(Pt. 3), 269–282, 2004. 145. Bennetto, L., Pennington, B.F., and Rogers, S.J., Intact and impaired memory functions in autism, Child Development 67, 1816–1835, 1996. 146. Luna, B., Minshew, N.J., Garver, K.E., Lazar, N.A., Thulborn, K.R., Eddy, W.F., and Sweeney, J.A., Neocortical system abnormalities in autism: an fMRI study of spatial working memory, Neurology 59(6), 834–840, 2002. 147. Spencer, J., O’Brien, J., Riggs, K., Braddick, O., Atkinson, J., and Wattam-Bell, J., Motion processing in autism: evidence for a dorsal stream deficiency, Neuroreport 11(12), 2765–2767, 2000. 148. Pellicano, E., Gibson, L., Maybery, M., Durkin, K., and Badcock, D.R., Abnormal global processing along the dorsal visual pathway in autism: a possible mechanism for weak visuospatial coherence?, Neuropsychologia 43(7), 1044–1053, 2005. 149. Tager-Flusberg, H., Basic level and superordinate level categorization by autistic, mentally retarded, and normal children, Journal of Experimental Child Psychology 40(3), 450–469, 1985. 150. Eskes, G.A., Bryson, S.E., and McCormick, T.A., Comprehension of concrete and abstract words in autistic children, Journal of Autism and Childhood Schizophrenia 20(1), 61–73, 1990. 151. Hobson, R.P. and Lee, A., Emotion-related and abstract concepts in autistic people: evidence from the British picture vocablulary scale, Journal of Autism and Developmental Disorders 19(4), 601–623, 1989. 152. Bryson, S.E., Interference effects in autistic children: evidence for the comprehension of single stimuli, Journal of Abnormal Psychology 92(2), 250–254, 1983. 153. Toichi, M. and Kamio, Y., Verbal association for simple common words in highfunctioning autism, Journal of Autism and Developmental Disorders 31, 483–490, 2001. 154. Van Lancker, D., Cornelius, C., and Needleman, R., Comprehension of verbal terms for emotions in normal, autistic, and schizophrenic children, Developmental Neuropsychology 7(1), 1–18, 1991. 155. Beversdorf, D.Q., Anderson, J.M., Manning, S.E., Anderson, S.L., Nordgren, R.E., Felopulos, G.J., Nadeau, S.E., Heilman, K.M., and Bauman, M.L., The effect of semantic and emotional context on written recall for verbal language in high functioning adults with autism spectrum disorder, Journal of Neurology, Neurosurgery, and Psychiatry 65(5), 8734–8737, 1998. 156. Harris, G.J., Chabris, C.F., Clark, J., Urban, T., Aharon, I., Steele, S., McGrath, L., Condouris, K., and Tager-Flusberg, H., Brain activation during semantic processing in autism spectrum disorders via functional magnetic resonance imaging, Brain and Cognition, in press. 157. Minshew, N.J. and Goldstein, G., The pattern of intact and impaired memory functions in autism, Journal of Child Psychology and Psychiatry 42, 1095–1101, 2001. 158. Ben Shalom, D., Memory in autism: review and synthesis, Cortex 39(4–5), 1129–1138, 2003. 159. VanMeter, L., Fein, D., Morris, R., Waterhouse, L., and Allen, D., Delay versus deviance in autistic social behavior, Journal of Autism and Developmental Disorders 27(5), 557–569, 1997. 160. Minshew, N.J. and Goldstein, G., Is autism an amnesic disorder? Evidence from the California verbal learning test, Neuropsychology 7, 209–216, 1993.

© 2006 by Taylor & Francis Group, LLC

2732_C009.fm Page 203 Wednesday, February 15, 2006 3:19 PM

Language in Autism

203

161. Boucher, J. and Warrington, E.K., Memory deficits in early infantile autism: some similarities to the amnesic syndrome, British Journal of Psychology 67(1), 73–87, 1976. 162. Bowler, D.M., Matthews, N.J., and Gardiner, J.M., Asperger’s syndrome and memory: similarity to autism but not amnesia, Neuropsychologia 35, 65–70, 1997. 163. Tager-Flusberg, H., Semantic processing in the free recall of autistic children: further evidence for a cognitive deficit, British Journal of Psychology 9, 417–430, 1991. 164. Toichi, M. and Kamio, Y., Long-term memory in high-functioning autism: controversy on episodic memory in autism reconsidered, Journal of Autism and Developmental Disorders 33(2), 151–161, 2003. 165. Boucher, J., Memory for recent event in autistic children, Journal of Autism and Childhood Schizophrenia 11(3), 293–301, 1981. 166. Klien, S.B., Chan, R.L., and Loftus, J., Independence of episodic and semantic selfknowledge: the case from autism, Social Cognition 17(4), 413–436, 1999. 167. Millward, C., Powell, S., Messer, D., and Jordan, R., Recall for self and other in autism: children’s memory for events experienced by themselves and their peers, Journal of Autism and Developmental Disorders 30(1), 15–28, 2000. 168. Bowler, D.M., Gardiner, J.M., Grice, S., and Saavalainen, P., Memory illusions: false recall and recognition in adults with Asperger’s syndrome, Journal of Abnormal Psychology 109(4), 663–672, 2000. 169. Rumsey, J. and Ernst, M., Functional Neuroimaging of autistic disorders, Mental Retardation and Developmental Disabilities Research Reviews 6(3), 171–179, 2000. 170. Bauman, M.L. and Kemper, T.L., The Neurobiology of Autism, The Johns Hopkins University Press, Baltimore, MD, 2005, p. 404. 171. Brambilla, P., Hardan, A., di Nemi, S.U., Perez, J., Soares, J.C., and Barale, F., Brain anatomy and development in autism: review of structural MRI studies, Brain Research Bulletin 61(6), 557–569, 2003. 172. Abell, F., Krams, M., Ashburner, J., Passingham, R., Friston, K., Frackowiak, R., Happe, F., Frith, C., and Frith, U., The neuroanatomy of autism: a voxel-based whole brain analysis of structural scans, Neuroreport 10(8), 1647–1651, 1999. 173. Herbert, M.R., Harris, G.J., Adrien, K.T., Ziegler, D.A., Makris, N., Kennedy, D.N., Lange, N.T., Chabris, C.F., Bakardjiev, A., Hodgson, J., Takeoka, M., Tager-Flusberg, H., and Caviness, V.S., Abnormal asymmetry in language association cortex in autism, Annals of Neurology 52(5), 588–596, 2002. 174. Herbert, M.R., Ziegler, D.A., Deutsch, C.K., O’Brien, L.M., Kennedy, D.N., Filipek, P.A., Bakardjiev, A.I., Hodgson, J., Takeoka, M., Makris, N., and Caviness, V.S., Jr., Brain asymmetries in autism and developmental language disorder: a nested wholebrain analysis, Brain 128(Pt. 1), 213–226, 2005. 175. Courchesne, E. and Pierce, K., Brain overgrowth in autism during a critical time in development: implications for frontal pyramidal neuron and interneuron development and connectivity, International Journal of Developmental Neuroscience 23(2–3), 153–170, 2005. 176. Carper, R. and Courchesne, E., Localized enlargement of the frontal cortex in early autism, Biological Psychiatry 57(2), 126–133, 2005. 177. Carper, R. and Courchesne, E., Inverse correlation between frontal lobe and cerebellum sizes in children with autism, Brain 123(4), 836–844, 2000. 178. Belmonte, M.K., Allen, G., Beckel-Mitchener, A., Boulanger, L.M., Carper, R.A., and Webb, S.J., Autism and abnormal development of brain connectivity, Journal of Neuroscience 24(42), 9228–9231, 2004. 179. Frith, U. and Happé, F., Autism: beyond “theory of mind,” Cognition 50, 115–132, 1994. 180. Hill, E.L., Executive dysfunction in autism, Trends in Cognitive Sciences 8(1), 26–32, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 205 Friday, March 10, 2006 2:21 PM

10

Prefrontal Cortex Joseph L. Price

CONTENTS Introduction............................................................................................................205 Structure.................................................................................................................206 Dorsolateral Prefrontal Cortex: Caudal, Dorsal, and Ventral ........................207 Ventromedial Prefrontal Cortex: Orbital and Medial ....................................209 Connections............................................................................................................210 Intrinsic Connections......................................................................................211 Sensory Inputs to Ventrolateral Prefrontal Region ........................................213 Parietal Connections of Caudolateral Prefrontal Region...............................213 Nonsensory Associations of Dorsomedial and Medial Prefrontal Regions ..........................................................................................214 Motor and Visceromotor Outputs...................................................................215 Functions................................................................................................................215 Orbital Cortex: Food and Reward? ................................................................215 Ventrolateral Convexity: Object-Related Selection and Judgment, Working Memory and Language?..................................................................216 Caudolateral Prefrontal Cortex: Visuo-Spatial and Auditory-Spatial Coordination? .................................................................................................218 Dorsomedial Prefrontal Cortex: Monitoring of Self-Referential Information? ...................................................................................................218 Medial Prefrontal Network: Visceral Modulation, Emotion, and Monitoring of Internal State?..................................................................219 References..............................................................................................................221

INTRODUCTION Autism is a multisystem disorder that disrupts most forms of cognition, but particularly affects social interaction, emotional behavior, and language. Because these functions all depend on integrative mechanisms within the prefrontal cortex, it is appropriate to consider the structure and function of this cortical region in a volume devoted to autism. At the same time, specific neuropathological defects related to autism have not been identified in the prefrontal cortex, although it is likely that any defects are subtle and difficult to identify morphologically. This chapter will focus on the structure, connections, and attributed functions of the prefrontal cortex. Even if the primary defects in autism are in some other part of the brain, it is to be expected 205

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 206 Friday, March 10, 2006 2:21 PM

206

Understanding Autism: From Basic Neuroscience to Treatment

that the expression of the disorder will depend on interaction with neural systems in the prefrontal cortex. The prefrontal cortex is somewhat loosely defined as the region of the frontal lobe rostral to the motor (and premotor) cortex. It represents a large fraction of the cerebral cortex in monkeys and an even larger fraction in humans. In a relatively short chapter it is impossible to present more than an overview of the structure, organization, and functions of such a large and complex part of the brain. This chapter will begin with the architectonic subdivisions of the prefrontal cortex, because these provide the basic map that is necessary for any further discussion. This will be followed by an overview of the connections of the many parts of the prefrontal cortex, with emphasis on cortico-cortical relations. Finally, a brief outline of the functions that have been attributed to the several domains of the prefrontal cortex will be attempted. The discussion will be focused on observations in nonhuman primates, but the architectonic maps will be correlated with equivalent maps in humans. Functional imaging observations in humans will also be discussed where appropriate.

STRUCTURE Both functionally and anatomically, the prefrontal cortex is usually divided into dorsolateral and ventromedial regions. (As discussed in the following text, however, the pattern of connections with other parts of the cerebral cortex suggests an organization that is almost orthogonal to this subdivision.) The dorsolateral region has been associated with several functions related to memory and executive control, and extends from the premotor areas and frontal eye fields to area 10 at the frontal pole. The ventromedial region, in contrast, has been associated with reward and emotion. It includes the anterior cingulate cortex and other medial prefrontal areas, as well as the cortex on the orbital surface of the frontal lobe. The cortex in both the dorsolateral and ventromedial regions extends from agranular cortical regions caudally, to the very granular and even eulaminate (having all layers well developed) cortex of the frontal pole. However, the nature of this progression is different in the two regions, proceeding from premotor areas dorsolaterally and from paralimbic areas ventromedially. Several architectonic maps of the prefrontal cortex have been proposed over the past century, but most current maps have been based on the maps by Brodmann (1909; Garey, 1994). Working almost 100 years ago, Brodmann provided what are still the most widely referenced maps for the cerebral cortex as a whole, in a variety of animals. The fact that his maps are still useful is a remarkable tribute to Brodmann’s observations, because the maps were drawn at a time when understanding of cortical function and connections was still relatively rudimentary. At the same time, there are limitations to Brodmann’s maps. He did not illustrate the orbital surface of the frontal lobe, so current maps of that region are based on Walker’s (1940) map of the monkey. There are also discrepancies between maps of different animals; these are especially apparent between the human and monkey maps, which cause problems when experimental work done in monkeys is compared with functional imaging data from human subjects. And finally, increased

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 207 Friday, March 10, 2006 2:21 PM

Prefrontal Cortex

207

understanding of the functional role of specific cortical subdivisions has made it necessary to analyze the cortex in greater detail, such that many or even most of Brodmann’s areas have been subdivided in current maps. The description in this chapter will be based primarily on two recent maps of the frontal lobe, one by Petrides and Pandya (1994, 1999, 2001) (Figure 10.1), and the other by Carmichael and Price (1994) and Öngür et al. (2003) (Figure 10.2). Both of these have attempted to correlate architectonic areas in monkeys and humans. Although each has provided additional subdivisions of Brodmann’s areas, both are consistent with the basic plan proposed by Brodmann (and by Walker for the orbital cortex).

DORSOLATERAL PREFRONTAL CORTEX: CAUDAL, DORSAL,

AND

VENTRAL

The caudal part of the dorsolateral prefrontal cortex contains agranular areas 6 (premotor cortex) and 8 (frontal eye fields), both of which are subdivided into several subareas. Area 6 is usually divided into dorsal and ventral components (6d and 6v), and the ventral region is further divided into areas 6va and 6vb (Preuss and GoldmanRakic, 1991). In addition, the dorsomedial part is recognized as area 6m or the supplementary motor area (SMA). In both monkeys and humans, area 8 is divided into a ventral area 8a and dorsal area 8b. In monkeys area 8a is situated within and just rostral to the arcuate sulcus, and is further subdivided into dorsal and ventral regions. Area 8b is located dorsal to the arcuate sulcus and extends dorsomedially onto the medial surface (Preuss and Goldman-Rakic, 1991; Petrides and Pandya, 1999). The cortex rostral to area 8 is occupied by two well-developed areas, area 9 dorsally and area 46 more ventrally. Of these, area 46 is fully granular, whereas 9 is less granular, or even dysgranular. Petrides and Pandya (1999) have pointed out that there is a discrepancy between the usual delineations of these areas in monkeys and humans. In most human maps (Brodmann, 1909; Economo and Koskinas, 1925; Sarkissov et al., 1955), area 9 extends ventrally onto the middle frontal gyrus, between area 8 and area 46. Brodmann did not recognize an area 46 in his monkey map, and most investigators have therefore used Walker’s (1940) map, which places area 46 on both banks of the full length of the principal sulcus, between area 8 caudally and area 10 rostrally; in this map area 9 is situated dorsal to area 46 and does not intervene between it and area 8. To deal with this discrepancy, and also on the basis of their own architectonic observations, Petrides and Pandya recognized an area 9/46 between area 8 and area 46 in both monkeys and humans. Area 46, then, is restricted to the rostral part of the principal sulcus (in monkeys) or middle frontal gyrus (in humans). A progression of three areas occupies the ventrolateral convexity of the frontal cortex (in monkeys) or inferior frontal gyrus (in humans). Caudally, the dysgranular area 44 adjoins area 6v; this area, which is very small but apparently present in monkeys, is usually equated with the premotor speech area. Rostral to this is area 45, which is characterized by large pyramidal cells in the deep part of layer III, and a moderately well-developed granular layer IV. Area 45 was divided into rostral and caudal subdivisions (areas 45a and 45b) by Petrides and Pandya (1999). Finally, ventral and rostral to area 45 is a region that Brodmann (1909) labeled area 12 in monkeys and area 47 in humans. To deal with this difference, Petrides and Pandya

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 208 Friday, March 10, 2006 2:21 PM

208

Understanding Autism: From Basic Neuroscience to Treatment

Dorsomedial prefrontal cortex

8b 8ad

9

8av

9/46d

9/46v 46

44

10

45a

45b

47/12

Caudolateral prefrontal cortex

8b 9 46

9/46d

8ad

9/46v

8av

10

45 47/12

Ventrolateral prefrontal cortex

FIGURE 10.1 (A color version of this figure follows page 236.) Architectonic maps of the lateral prefrontal cortex in humans (above) and monkeys (below), as delineated by Petrides and Pandya (1994, 1999, 2001). The dorsomedial, ventrolateral, and caudolateral domains discussed in the text are marked by red, yellow, and green overlays, respectively.

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 209 Friday, March 10, 2006 2:21 PM

Prefrontal Cortex

209

10p 11m

11l

47/12r 14 13 47/ r b 131 47/ 121 13 12m 45 14c m 13a

10 mm

Iam Iapm

Orbital prefrontal network

9

24 25 32 10m 10r pI 14c 14r 11m

47/12s Iai

Medial prefrontal network

24c 11m 14r

24b 12l

12m

24a

13b

32

13m 14c

9

12r 11l

13a Iam AON

10p

Dorsomedial prefrontal cortex

OB 10o

32ac

cc

13l

12o G

Iapm Ial Iai

10m

25

PrCo

14c

14r

5 mm

FIGURE 10.2 (A color version of this figure follows page 236.) Architectonic maps of the ventral (orbital) and medial prefrontal cortex in humans (above) and monkeys (below), as delineated by Carmichael and Price (1994) and Öngür et al. (2003). The dorsomedial domain and the medial and orbital prefrontal networks discussed in the text are marked by red, blue, and beige overlays, respectively.

(1994, 2001) proposed that it be labeled area 47/12. It is a relatively large heterogeneous area that varies from dysgranular to granular and shows substantial staining differences for myelinated fibers. Based on these differences, and differences in connections, Carmichael and Price (1994) divided it into four areas in monkeys: 12o (for orbital), 12l (for lateral), 12m (for medial), and 12r (for rostral). In humans, Öngür et al. (2003) recognized the same areas, but used the designation 47/12, following Petrides and Pandya (1994, 2001). Because the equivalent of area 12o is located within the horizontal ramus of the lateral sulcus in humans, it was designated area 47/12s (for sulcal).

VENTROMEDIAL PREFRONTAL CORTEX: ORBITAL

AND

MEDIAL

The caudal part of the ventral or orbital surface of the frontal lobe is continuous with the insula. As marked by the presence of the claustrum deep to the insular

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 210 Friday, March 10, 2006 2:21 PM

210

Understanding Autism: From Basic Neuroscience to Treatment

cortex, the most caudal part of the orbital surface can be distinguished as a rostral extension of the insula. Carmichael and Price (1994) recognized five subdivisions: the medial, posteromedial, intermediate, lateral, and posterolateral agranular insular areas (Iam, Iapm, Iai, Ial, and Iapl, respectively). These areas are all agranular, although they vary in structure from relatively thin, poorly laminated areas (e.g., Iam, Iapm) to thicker, more fully laminated areas (Iai, Ial, and Iapl). Rostral to the agranular insular areas, the map by Walker (1940) has been used as the model for most subsequent descriptions. It recognizes areas 13 and 14 in the central and medial orbital cortex, area 11 rostral to these, and area 10 at the frontal pole, as well as area 12 at the ventrolateral corner of the hemisphere. Carmichael and Price (1994) subdivided each of these, for a total of 13 areas (12l, 12m, 12o, 12r, 13a, 13b, 13l, 13m, 14c, 14r, 11l, 11m, and 10o). In humans, Öngür et al. (2003) were able to delineate all these areas. On the medial wall of the frontal lobe, Brodmann (1909) recognized three agranular areas: area 25 at the caudal edge, ventral to the genu of the corpus callosum, area 32 rostral to this, and area 24 in the anterior cingulate gyrus dorsal to the corpus callosum. As in the orbital cortex, the granular cortex at the frontal pole of the medial cortex is area 10. All of these areas have been described in recent maps (Preuss and Goldman-Rakic, 1991; Carmichael and Price, 1994; Petrides and Pandya, 1999). The part of area 10 on the medial wall was termed area 10m by Carmichael and Price (1994), who also described a caudal extension of this area ventral to area 32. Area 24 is usually divided into 24a, 24b, and 24c, representing strips parallel to the corpus callosum. These areas are also present in humans, but inconsistencies in the description by Brodmann (1909) have produced confusion about area 32 (Öngür et al., 2003). In monkeys, Brodmann called area 32 the prelimbic area, and located it rostral and ventral to the genu of the corpus callosum. In humans, his area 32, which he called the dorsal anterior cingulate area, extends dorsally and caudally, around area 24 and the corpus callosum. Brodmann (1909) himself wrote that his anterior cingulate area 32 in humans is not comparable to the prelimbic area 32 in monkeys (see Garey, 1994, p. 138). Öngür et al. (2003) therefore recognized area 32pl (for prelimbic) in the human subgenual region and area 32ac (for anterior cingulate) in the supracallosal area. The major difference in the prefrontal cortex between monkeys and humans is the great expansion and increased granularity of area 10 in humans (Öngür et al., 2003). Because of this, the two subdivisions of area 10 recognized in monkeys (areas 10m and 10o) were expanded to three subdivisions in humans (areas 10m, 10r, and 10p).

CONNECTIONS The dorsolateral and ventromedial parts of the prefrontal cortex (PFC) have usually been considered to have different connections with other parts of the cortex and subcortical structures (Figure 10.3). Both are interconnected with the mediodorsal thalamic nucleus, but the ventromedial PFC is related to the medial part of the nucleus, whereas dorsolateral PFC is related to the more lateral part. In addition, the dorsolateral PFC has been characterized by connections with the posterior parietal cortex, whereas

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 211 Friday, March 10, 2006 2:21 PM

Prefrontal Cortex

211

Rostral Superior Temporal Cortex 24c

Posterior Cingulate/ Retrosplenial Cortex

9 9

24b

9/46d 9/46v

46

24a 32

Hypothalamus PAG

Posterior parietal cortex

8b

10

10m

47/12

8ad 8av

45

25 14c

Posterior auditory cortex Ventral premotor cortex

14r OB

5 mm

10o 12r 11m

Amygdala Hippocampus

14r

Parahippocampal cortex

Somatosensory (1, 2, SII, 7b)

13m

14c

Visual (inf temporal)

12m

13b

Entorhinal cortex

Auditory (rostral)

12l

11l

13a

3l

12o G

Iam AON

Gustatory PrCo

Iapm

Ial

Visceral Olfactory

Iai

FIGURE 10.3 (A color version of this figure follows page 236.) Summary of major connections of the prefrontal domains discussed in the text.

the ventromedial PFC is related to limbic structures such as the amygdala, hippocampus, and related medial temporal areas. In spite of such differences, however, there are also substantial similarities between connections of the dorsolateral and ventromedial prefrontal cortex, especially in connections to other cortical regions. Strikingly, these suggest an orthogonal organization into dorsomedial and ventrolateral prefrontal regions.

INTRINSIC CONNECTIONS There are substantial interconnections between areas within the prefrontal cortex, which have a significant degree of specificity that define distinct systems or networks. These have been studied in most detail in the orbital and medial regions of the cortex, in which two interconnected networks have been defined. An equivalent, comprehensive scheme for the cortico-cortical connections within the dorsolateral prefrontal cortex has not been done, but some analysis is possible from published illustrations of experiments in the dorsolateral prefrontal cortex. It should be noted, however, that the description of intrinsic connections for the dorsolateral cortex given in the following text is not yet definitive.

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 212 Friday, March 10, 2006 2:21 PM

212

Understanding Autism: From Basic Neuroscience to Treatment

In the ventromedial part of the prefrontal cortex (the orbital and medial prefrontal cortex), experiments on cortico-cortical connections formed the basis for two distinct systems, or networks (Carmichael and Price, 1996). One of these, the orbital prefrontal network, includes most of the areas in the orbital cortex, including areas Iam, Iapm, Ial, 13m, 13l, 12m, 12l, 12r, and 11l. These areas have extensive connections with each other, but they have relatively few connections to areas in the medial prefrontal cortex. Most of the areas within the orbital network also have interconnections with areas on the ventrolateral prefrontal cortex, including areas 6v, 45, and the rostroventral part of area 46, on the ventral bank of the principal sulcus (Barbas and Pandya, 1989; Carmichael and Price, 1995b, 1996; Petrides and Pandya, 2001). The ventrolateral prefrontal areas are also interconnected with each other, although they have relatively few connections with the dorsomedial prefrontal cortex. The areas on the medial wall and frontal pole (areas 24, 25, 32, 10m, and 10o), as well as areas on the ventromedial corner of the frontal lobe (areas 14c, 14r, and 11m), have been included in the medial prefrontal network (Carmichael and Price, 1996). As with the orbital network, these areas have extensive connections with each other, but not to the orbital network areas. Two other medial orbital areas (13a and 13b) have connections to both the medial and orbital networks. In addition, two areas in the caudolateral orbital cortex (areas Iai and 12o) are preferentially connected to the medial network, although area 12o also has some connections with the orbital network. Strikingly, many of the medial network areas, including areas 12o and Iai, are connected to dorsomedial area 9, the rostrodorsal part of area 46 in the dorsal bank and fundus of the principal sulcus, and area 8b (Barbas and Pandya, 1989; Carmichael and Price, 1996; Petrides and Pandya, 1999). These dorsomedial areas are also interconnected with each other, but they have relatively few connections to the ventrolateral prefrontal cortex ventral to the principal sulcus. Where a relatively small projection has been identified in ventrolateral areas, they are often illustrated in the approximate position of areas 12o and Iai, the lateral parts of the medial network, although the published analyses have not been sufficiently detailed to confirm this (e.g., Petrides and Pandya, 1999). These results suggest that there are two overall systems of interconnected areas within the prefrontal cortex. One involves the medial prefrontal network, together with most of the areas in the dorsomedial part of the frontal lobe, dorsal to the principal sulcus. Interestingly, this system also involves a small region in the caudolateral orbital cortex (including at least areas Iai and 12o). The other system includes the orbital prefrontal network and the ventrolateral convexity, up to and including the rostroventral bank of the principal sulcus. Many of the details of these systems, especially the relations of the medial and orbital networks to the areas on the dorsolateral surface and extent to which the systems can be subdivided into functional subunits, remain to be worked out. In addition, the caudal part of the dorsolateral prefrontal cortex, including areas 8a and several parts of area 6, may form a third interconnected system, which has outputs to the motor areas through the ventral premotor area (area 6v) (Dum and Strick, 2005). These areas also have substantial interconnections. They are also

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 213 Friday, March 10, 2006 2:21 PM

Prefrontal Cortex

213

connected with the caudal part of area 46 (area 9/46 of Petrides and Pandya, 1999), but have relatively few connections with more rostral prefrontal areas.

SENSORY INPUTS

TO

VENTROLATERAL PREFRONTAL REGION

The ventrolateral prefrontal region, including the ventrolateral convexity and much of the orbital cortex, is characterized by its sensory inputs, mainly from cortical sensory associational areas. Projections from the inferior temporal cortex carry visual object information to several architectonic areas, including areas 45, 12l, 13m, 13l, and 11l (Webster et al., 1994; Carmichael and Price, 1995b). Projections from the anterior auditory belt or parabelt areas have been reported to area 45, as well as to the lateral frontal pole (Romanski et al., 1999). It has been suggested that these carry auditory object information (in contrast to projections from the caudal auditory belt/parabelt cortex to the areas 8a and 9/46 that may carry auditory-spatial information; Romanski et al., 1999). Inputs from several somatic sensory cortical areas including parts of areas 1 and 2, SII, and 7b also reach the ventrolateral prefrontal region, including at least the ventral part of area 46 and orbital areas 12m and 13l (Preuss and Goldman-Rakic, 1989; Carmichael and Price, 1995b). Recordings of unit activity in the orbital cortex have identified neurons that respond to all sensory modalities except audition (Rolls, 2005). Auditory responses have been recorded in the cortex on the ventrolateral convexity (Romanski and Goldman-Rakic, 2002). Further, there are taste, visceral, and olfactory inputs to the orbital cortex. The primary gustatory cortex, in the rostrodorsal insula, is connected with parts of the agranular insula and areas 13l and 13m. In addition, there are connections with parts of the agranular insular cortex (areas Ial and Iapl) from the ventral lamina of the taste/visceral thalamic relay nucleus (Carmichael and Price, 1995b); these may relay visceral afferent information. The olfactory inputs arise in the primary olfactory cortex at the caudal edge of the orbital surface and project into the agranular insular areas. From there, olfactory sensory activity can be distributed through the orbital network (Carmichael et al., 1994; Carmichael and Price, 1995b). Together with the visual and somatic sensory inputs, these suggest that the orbital network is particularly involved in assessment of food. Indeed, neurons in this region respond to multimodal food or food-related stimuli (Rolls, 2005). The responses are not limited to the sensory aspects of the stimuli, however, but also reflect affective properties, i.e., whether the stimulus is rewarding or aversive (Rolls, 2000; Schultz, 2000; Hikosaka and Watanabe, 2000).

PARIETAL CONNECTIONS

OF

CAUDOLATERAL PREFRONTAL REGION

The caudal part of the dorsomedial region, including primarily areas 46/9, and the more caudal frontal eye field and premotor areas (areas 8a and 6), are substantially connected with the posterior parietal cortex, areas 7a and 7ip (LIP) in the lateral bank of the intraparietal sulcus and the adjacent gyrus, and the medial parietal area 7m (Cavada and Goldman-Rakic, 1989; Petrides and Pandya, 1984, 1999; Lewis and Van Essen, 2000). This parietal region is often considered as a visuospatial region and, as part of the dorsal visual stream conveying information about

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 214 Friday, March 10, 2006 2:21 PM

214

Understanding Autism: From Basic Neuroscience to Treatment

location and movement of visual stimuli, it may be involved in monitoring extrapersonal space. Inputs to this region have been reported from the caudal part of the auditory cortex, which has been suggested to process auditory-spatial activity (Romanski et al., 1999).

NONSENSORY ASSOCIATIONS PREFRONTAL REGIONS

OF

DORSOMEDIAL

AND

MEDIAL

In contrast to the ventrolateral part of the prefrontal cortex, the dorsomedial and medial prefrontal areas receive little direct sensory inputs from sensory association cortical areas. As described in the following text, they are connected to a relatively consistent set of cortical areas, which include cortex on the rostral superior temporal gyrus and dorsal bank of the superior temporal sulcus, and the posterior cingulate and retrosplenial cortex. In addition, the medial prefrontal areas are strongly connected to limbic structures, including the amygdala, hippocampus, entorhinal cortex, and parahippocampal cortex. Several studies have consistently observed that areas in the medial and dorsomedial prefrontal cortex are connected with the cortex in the rostral superior temporal gyrus and dorsal bank of the superior temporal sulcus (Carmichael and Price, 1995b; Barbas et al., 1999; Petrides and Pandya, 1999; Kondo et al., 2003). This region includes the dorsal part of the temporal pole and a triangular region extending caudally from the temporal pole into the rostral part of the dorsal bank of the superior temporal sulcus. Although the region is very near the auditory cortex and has been suggested to have auditory function, it is in fact just outside the auditory belt and parabelt areas. The belt and parabelt areas are located within the lateral sulcus rostrally, whereas the region that is connected to the medial network is situated more laterally and ventrally, extending into the dorsal bank of the superior temporal sulcus. There appears to be little direct evidence that this region is directly (particularly unimodally) involved in audition. In addition, the medial prefrontal network and the dorsomedial prefrontal areas are connected to the posterior cingulate and retrosplenial cortex, including parts of area 23 and areas 29 and 30 (Carmichael and Price, 1995a; Petrides and Pandya, 1999; Morris et al., 1999; Kobayashi and Amaral, 2003). These connections are not uniform, and there are considerable differences in the specific connections of individual areas within both the posterior cingulate or retrosplenial region and the medial and dorsomedial prefrontal region. Further analysis is needed to understand the organization of this system, but it is clear that the two regions are linked. The medial prefrontal network also has substantial limbic connections (Amaral and Price, 1984; Barbas and Blatt, 1995; Carmichael and Price, 1995a; Kondo et al., 2005). The amygdala provides the densest and most extensive projection, to the entire medial wall and to the lateral areas Iai and 12o that are associated with the medial network. The projections also extend to a lesser extent into area 9 and even into the rostral part of area 46 (Amaral and Price, 1984). The hippocampal formation provides a more restricted projection from the subiculum to the medial edge of the orbital cortex and the caudal part of the medial wall around the genu

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 215 Friday, March 10, 2006 2:21 PM

Prefrontal Cortex

215

of the corpus callosum. Other limbic areas such as the entorhinal and parahippocampal cortex have comparable projections to the medial network (Carmichael and Price, 1995a; Kondo et al., 2005).

MOTOR

AND

VISCEROMOTOR OUTPUTS

Virtually all of the connections of the prefrontal cortex discussed in the preceding text as inputs are reciprocal, so that much of the output of the prefrontal cortex can be thought of as feedback onto other areas of the cerebral cortex. In addition, there are outputs to motor and premotor cortical areas, and to visceromotor control structures in the hypothalamus and brain stem. Although it is generally presumed that the dorsolateral prefrontal cortex influences behavior through outputs to the motor and premotor cortical areas, these connections are remarkably restricted. The major pathway appears to be through the ventral premotor area (6v) (Dum and Strick, 2005). This area is connected to both the dorsal premotor area as well as to the primary motor cortex. The only substantial prefrontal connection of the ventral premotor area is with the ventral part of area 46, in the midportion of the ventral bank of the principal sulcus. The other premotor areas do not appear to receive substantial prefrontal projections, although prefrontal connections have been reported with cingulate motor areas (Dum and Strick Cingulate book). The visceromotor outputs arise in the medial prefrontal network and project to the medial and lateral hypothalamus, the ventral midbrain, and the periaqueductal gray (An et al., 1998; Öngür et al., 1998; Rempel-Clower and Barbas, 1998). The hypothalamus and periaqueductal gray are both areas that coordinate several aspects of visceral function, and these projections provide a mechanism for cortical modulation of both autonomic (sympathetic and parasympathetic) and endocrine function. The most substantial projections originate from areas 25 and 32 on the caudal part of the medial wall, but projections also arise from all of the other medial network areas, including more rostral medial areas and the lateral areas Iai and 12o. There are also projections to the hypothalamus and periaqueductal gray from area 9 in the dorsomedial prefrontal cortex and from the dorsal part of the temporal pole cortex.

FUNCTIONS The structure and connections described in the preceding text indicate that the prefrontal cortex is not homogeneous, but is made up of many different areas that have distinct structures, connections, and presumably functions. Although it is not yet possible to understand the function of each area, several domains within the prefrontal cortex can be associated with distinct functional roles.

ORBITAL CORTEX: FOOD

AND

REWARD?

As discussed earlier, the ventrolateral half of the prefrontal cortex, including the orbital cortex on the ventral surface and the ventral half of the lateral surface, receives a number of inputs from all sensory modalities. Not surprisingly, functional analysis

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 216 Friday, March 10, 2006 2:21 PM

216

Understanding Autism: From Basic Neuroscience to Treatment

of these areas suggests that they are primarily involved in integration and analysis of sensory information. Both the orbital and ventrolateral convexity, however, appear to process more than just the sensory properties of stimuli. The orbital cortex receives olfactory, gustatory, and visceral afferents, in addition to visual and somatic sensory inputs. These are integrated through the cortico-cortical interconnections within the orbital prefrontal network. This constellation of afferents suggests that the orbital network is involved in the analysis and assessment of food and related stimuli. Physiological recordings, mostly done by Rolls and his colleagues (Rolls, 2005), indicate that neurons in the orbital cortex respond to multimodal aspects of food stimuli. Individual neurons may respond to the appearance, flavor (i.e., taste plus olfaction), and texture of foods. In humans, several functional imaging studies have shown that the orbital cortex is activated by olfactory, taste, somatic sensory, and conjoined stimuli (e.g., Small et al., 1999; Zald and Pardo, 2000; Hagen et al., 2002; Rolls, 2005; Small and Prescott, 2005). In addition to the sensory aspects of these stimuli, however, the orbital cortex neurons also respond to the affective quality of stimuli. That is, the response of orbital neurons changes if the status of a stimulus is changed from rewarding to aversive. This is especially striking when an experimental animal is fed to satiety with a particular food; the response to that food stimulus, but not to other foods, changes in relation to the change in the animal’s acceptance of the food (Rolls, 1997; 2005). Similar effects have been observed in humans with fMRI (O’Doherty et al., 2000; Kringelbach et al., 2003). Further, neurons in the orbital cortex are responsive to more abstract factors such as the expectation of reward and stimuli that predict reward (Schultz et al., 2000; Hikosaka and Watanabe, 2000, 2004; Schultz, 2004). Although the system may have initially evolved to assess sensory properties of food, it is a primary reward and it would appear likely that the system has become elaborated in primates to process more generalized aspects of reward.

VENTROLATERAL CONVEXITY: OBJECT-RELATED SELECTION AND JUDGMENT, WORKING MEMORY AND LANGUAGE? As described earlier, the ventrolateral convexity of the prefrontal cortex, ventral to the principal sulcus, receives inputs from visual, auditory, and somatic sensory areas, particularly unimodal sensory association areas in the temporal and rostral parietal cortex. The visual and auditory inputs arise from the inferior temporal cortex and the anterior belt/parabelt areas that have been associated with object recognition (Romanski et al., 1999). Goldman-Rakic and her colleagues recorded responses to complex sensory stimuli in the ventrolateral prefrontal cortex that included firing during the delay phase of working memory tasks. On the basis of these observations they proposed that the ventrolateral convexity is involved in object-related working memory, in parallel to a spatial working memory system in the dorsally adjacent cortex around the caudal principal sulcus (Wilson et al., 1993; Romanski, 2004).

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 217 Friday, March 10, 2006 2:21 PM

Prefrontal Cortex

217

Human functional imaging studies have also provided evidence for activation of the ventrolateral prefrontal cortex during tasks that enhance memorization or encoding of information. Such tasks, even when the memory formation is not intentional, activate an extended region in the lateral prefrontal cortex that appears to involve parts of areas 45, 47, and ventral 46 (Buckner et al., 1999). Strikingly, when single-event fMRI methods are used on a trial-by-trial basis, activation of this region predicts whether the stimulus presented in that trial will be subsequently remembered or forgotten (Wagner et al., 1998). On the other hand, lesions of the lateral prefrontal cortex do not produce notable amnesia comparable to that seen with medial temporal lobe lesions. Although there are several possible explanations for the lack of memory deficits with frontal lesions, they suggest that the prefrontal cortex is not involved in memory function in a simple way, but may instead have a more basic role that is related to several functions as well as memory. Other groups have argued, in fact, that the function of the ventrolateral prefrontal region is not working memory as such, but that it is involved in the closely related processes of object selection, comparison, and judgment (Passingham et al., 2000; Petrides, 1996, 2005). It is worth noting that such a role would be similar to the role of the orbital cortex in assessing and evaluating food and other stimuli in relation to reward. A meta-analysis by Duncan and Owen (2000) has shown that the region of the ventrolateral prefrontal cortex that is activated by working memory tasks is also activated by tasks with a broad range of cognitive demands, including aspects of perception, response selection, executive control, and problem solving, as well as episodic memory. They remark that these results suggest a specific frontal system that is consistently used for diverse cognitive problems involving higher-level sensory processing. In addition, the ventral part of the lateral prefrontal surface is associated in humans with various language-related functions. These may be related to the auditory and visual processing seen in the region in nonhuman primates, but they also appear to reflect the cognitive and associative roles suggested earlier. Area 44, at the caudal edge of this region, is generally equated with Broca’s language area; functional imaging studies suggest that this area and a more rostral region that includes parts of areas 45 and/or 47/12 are activated in relation to distinct aspects of language. A number of studies have reported differences in these areas in relation to semantic vs. phonological language tasks, with the caudal region being particularly important for phonological tasks, and the more rostral region being more involved in semantic tasks (see Buckner et al., 2005 for a discussion). For example, a PET study by Buckner et al. (1995) found activation in these regions with two speech generation tasks, stem completion and verb generation. Area 44 was activated by both tasks, whereas the more rostral region was activated only by verb generation, which is a more complex task that involves semantic as well as phonological processing. More recently, however, Gold et al. (2005) have suggested that the situation is more complex. In an fMRI study, they indicate that both phonological and semantic aspects of language are distributed to rostral and caudal parts of the ventrolateral prefrontal cortex. There still appear to be rostrocaudal differences, but they are more subtle than was previously thought.

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 218 Friday, March 10, 2006 2:21 PM

218

Understanding Autism: From Basic Neuroscience to Treatment

CAUDOLATERAL PREFRONTAL CORTEX: VISUO-SPATIAL AND AUDITORY-SPATIAL COORDINATION? The connections from the posterior parietal areas with areas 46/9 and 8a within the arcuate sulcus and the caudal part of the principal sulcus are generally thought to be the continuation of the dorsal visual stream that processes visuospatial activity. These areas in the caudolateral prefrontal cortex also are presumed to be involved in visuospatial-related functions, including working memory for spatial position (Romanski, 2004). This analysis depends principally on work by Goldman-Rakic and her colleagues, who showed that units within the caudal part of the principal sulcus fire during the delay period of an oculomotor delayed response task, in which animals were trained to make a saccade to a remembered location. They argued that the delay period firing could hold information about the location that could be used to guide the subsequent saccade. Individual neurons were found to be tuned to a specific peripheral location, with other neurons tuned to other locations (Funahashi et al., 1989). Small lesions in this region of the cortex produced memory deficits for specific visual field locations in this task (Funahashi et al., 1993). Other sensory modalities may also be involved in the spatial function of the caudolateral prefrontal cortex. Although there are some indications of somatic sensory inputs to this region, the best indication is from the auditory system (Romanski, 2004). As mentioned in the preceding text, there are projections from the caudal belt areas around the auditory cortex to the periarcuate/posterior principal sulcus region (Romanski et al., 1999), and there are several studies documenting auditory function in this region (see Romanski, 2004 for references). The caudolateral prefrontal cortex may not be solely involved in spatial working memory, however. Lesions of area 8a have also been shown to produce severe deficits in a nonspatial visual–visual conditional associative task (Petrides, 2005). In this task the animal was required to make a choice between lighted and nonlighted boxes, depending on which of two other visual stimuli were presented. The position of the lighted and nonlighted boxes varies between trials, so no spatial information was involved. It is not clear whether this represents a more fine-grained separation of functions, in which, for example, area 46/9 is involved in spatial working memory whereas area 8a functions in conditional selection of competing stimuli. Alternatively, it is possible that both tasks are affected by a more basic function of this region of the cortex related to processing and maintaining sensory information for subsequent motor response.

DORSOMEDIAL PREFRONTAL CORTEX: MONITORING OF SELF-REFERENTIAL INFORMATION? The connections of the dorsomedial prefrontal cortex, including primarily the dorsal part of area 46 and areas 9 and 8b, do not indicate a clear functional role for this domain. This is because the function of the related cortical regions, in the posterior cingulate/retrosplenial cortex and the rostral superior temporal sulcus, are themselves relatively poorly defined. Based on studies with lesions in this region, however, Petrides and his colleagues have proposed that the dorsomedial prefrontal region is

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 219 Friday, March 10, 2006 2:21 PM

Prefrontal Cortex

219

specifically involved in tracking or monitoring information about previous choices (Petrides, 2005). Human subjects show fMRI activation of the lateral part of area 9, rostral area 9/46, and area 46 in a task that requires them to determine whether they had seen a pair of stimuli before and, if so, which one they selected in the previous exposure. Similarly, monkeys with lesions of the dorsomedial prefrontal cortex showed deficits in a task that required the animal to choose between two equally familiar stimuli based on which one was selected in a previous exposure. In both cases, the key information was what the individual did previously; there were no spatial or object-related cues. A similar result is provided by an fMRI study by Gusnard et al. (2001), which related activity to a task that compared externally vs. internally orientated attention. In both cases the subjects were presented with a series of images. In the externally cued condition, they made a judgment of whether the scene was indoors or outdoors, whereas in the internally cued condition they rated the picture as pleasant or unpleasant. When comparing the two, the internally cued condition evoked more activity in the dorsomedial prefrontal cortex, especially in the medial part of areas 9 and 8b. If this result is compared with the data from Petrides (2005) described earlier, it would appear that the medial part of the dorsomedial region is active when the individual is monitoring internal thoughts or feelings, whereas the more lateral part of this region may monitor actions or choices that the individual has done or made.

MEDIAL PREFRONTAL NETWORK: VISCERAL MODULATION, EMOTION, AND MONITORING OF INTERNAL STATE? The areas of the medial prefrontal network (both those on the medial wall of the frontal lobe and the related areas in the caudolateral orbital cortex) share many of the connections of the dorsomedial prefrontal cortex, particularly with the rostral superior temporal gyrus and dorsal bank of the superior temporal sulcus, and with the posterior cingulate/retrosplenial region. It may be expected, therefore, that this prefrontal domain would play a similar functional role as the dorsomedial domain, and indeed it can be suggested that the medial network is involved in monitoring of self-referential information. In addition, however, the medial network is also implicated in visceral modulation in relation to emotional stimuli, and in emotion itself. As described briefly in the preceding text, the medial network has substantial outputs to visceral coordination centers in the hypothalamus and brain stem. This correlates well with older observations by Kaada et al. (1949) that stimulation of the medial prefrontal cortex can elicit responses in visceral functions such as respiration, heart rate, and blood pressure. More recently, several functional imaging studies have shown that activity in areas within the medial network is correlated with visceral activation (Critchley, Elliott et al., 2000; Critchley, Corfield et al., 2000; Critchley et al., 2005; Patterson et al., 2002; Nagai et al., 2004; Teves et al., 2004). In most of these cases, the visceral reaction was induced by emotional stimuli, but activity in the medial prefrontal cortex has also been shown to be correlated with visceral reactions evoked by hypoglycemia, a nonemotional stimulus (Teves et al., 2004). It appears that the medial network and the closely related amygdala

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 220 Friday, March 10, 2006 2:21 PM

220

Understanding Autism: From Basic Neuroscience to Treatment

are the principal structures through which the cerebral hemisphere can control and modulate visceral activity. In humans, lesions of the ventromedial prefrontal cortex abolish the automatic visceral response to emotive stimuli, blunt emotional reactions, and produce a severe deficit in effective decision making (Bechara et al., 2000; Bechara, 2004). Most strikingly, subjects with these lesions appear to have difficulty in understanding the long-term consequences of their actions and instead base their decisions on prospects for immediate, short-term gain. Damasio (1994) and his colleagues have argued that that visceral reactions to emotive stimuli provide a critical somatic marker or warning signal that is used to guide behavior, especially against making inappropriate or disadvantageous decisions. The loss of the visceral reaction following ventromedial prefrontal lesions is proposed to be the causative factor in the behavioral deficit in decision making. In some cases the visceral reaction itself may be monitored, but a parallel “as if ” circuit may also provide a signal without necessarily requiring sensation of the visceral effect. In addition, recent observations suggest that the syndrome produced by ventromedial cortical leseions may also reflect deficits in the ability to suppress previously learned reactions or associations (Fellows and Farah, 2005); this probably reflects damage to the cortico-striatal-pallido-thalamic loop from the medial prefrontal network through the ventromedial striatum (including the nucleus accumbens), ventral pallidum, and mediodorsal thalamic nucleus (Öngür and Price, 2000). In addition to visceral reactions in relation to emotive stimuli, the medial prefrontal network areas also appear to be involved in the emotion itself. This has been best shown by functional imaging studies of mood disorders, which indicate that the amygdala, medial prefrontal cortex, and a lateral region in the anterior insula are all abnormally active in depression (Drevets et al., 1992; Drevets, 2000; Mayberg et al., 2003) and induced sadness (Mayberg et al., 1999). The anterior insular region appears to correspond to the lateral component of the medial network (Öngür et al., 2003). In a recent paper, Mayberg and her colleagues (2005) report that deep brain stimulation in the caudal, medial prefrontal cortex (near area 25) can produce near complete remission of major depressive disorder in severe, treatment-resistant cases. The stimulation produced decreases activity in areas of the medial prefrontal network (both on the medial wall of the hemisphere and in anterior insula) and in the hypothalamus, similar to the effect of antidepressive drug treatment seen in a previous study of less severe depression (Mayberg et al., 2000). Raichle and his colleagues (Raichle et al., 2001; Raichle and Gusnard, 2005) have recently drawn attention to the observation from functional imaging that there is a consistent, default pattern of brain activity at rest in which some areas are more active than others. The areas that are most active at rest often show an apparent decrease in activity during externally directed tasks. The most striking of these areas are in the ventral part of the medial prefrontal cortex, and in the posterior cingulate/ retrosplenial region, with an additional area in the rostromedial temporal lobe. In a further investigation of this phenomenon, Greicius et al. (2003) correlated activity in different parts of the brain to define functional connectivity during distinct brain states. They found that the ventromedial prefrontal cortex is correlated with the posterior cingulate/retrosplenial region, both at rest, when activity is high, and during

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 221 Friday, March 10, 2006 2:21 PM

Prefrontal Cortex

221

engagement in a visual task, when activity is low. Taken together, these results indicate that these areas form an organized network of areas that are engaged during a baseline mode of brain function but are reduced in activity during specific goaldirected behaviors. It is tempting to suggest that this system is involved in monitoring the internal state of the body during rest, providing an awareness of self that is switched off when attention is directed to an external task.

REFERENCES Amaral, D.G. and Price, J.L., Amygdalo-cortical projections in the monkey (Macaca fascicularis), J Comp Neurol., 230: 465–496, 1984. An, X., Bandler, R., Ongur, D., and Price, J.L., Prefrontal cortical projections to longitudinal columns in the midbrain periaqueductal gray in macaque monkeys, J Comp Neurol., 401: 455–479, 1998. Barbas, H. and Pandya, D.N., Architecture and intrinsic connections of the prefrontal cortex in the rhesus monkey, J Comp Neurol., 286: 353–375, 1989. Barbas, H. and Blatt, G.J., Topographically specific hippocampal projections target functionally distinct prefrontal areas in the rhesus monkey, Hippocampus, 5(6): 511–533, 1995. Barbas, H., Ghashghaei, H., Dombrowski, S.M., and Rempel-Clower, N.L., Medial prefrontal cortices are unified by common connections with superior temporal cortices and distinguished by input from memory-related areas in the rhesus monkey, J Comp Neurol., 410: 343–367, 1999. Bechara, A., Damasio, H., and Damasio, A.R., Emotion, decision making and the orbitofrontal cortex, Cereb Cortex, 10: 295–307, 2000 Bechara, A., Disturbances of emotion regulation after focal brain lesions, Int Rev Neurobiol., 62: 159–193, 2004. Brodmann, K., Vergleichende Lokalisationslehre der Grosshirnrinde in ihren Prinzipien dargestellt auf Grund des Zellenbaues, Leipzig, J.A. Barth, 1909. Buckner, R.L., Raichle, M.E., and Petersen, S.E., Dissociation of human prefrontal cortical areas across different speech production tasks and gender groups, J Neurophysiol., 74: 2163–2173, 1995. Buckner, R.L., Kelley, W.M., and Petersen, S.E., Frontal cortex contributes to human memory formation, Nat Neurosci., 2: 311–314, 1999 Carmichael, S.T. and Price, J.L., Architectonic subdivision of the orbital and medial prefrontal cortex in the macaque monkey, J Comp Neurol., 346: 366–402, 1994. Carmichael, S.T. and Price, J.L., Limbic connections of the orbital and medial prefrontal cortex in macaque monkeys, J Comp Neurol., 363: 615–641, 1995a. Carmichael, S.T. and Price, J.L., Sensory and premotor connections of the orbital and medial prefrontal cortex, J Comp Neurol., 363: 642–664, 1995b. Carmichael, S.T. and Price, J.L., Connectional networks within the orbital and medial prefrontal cortex of macaque monkeys, J. Comp. Neurol., 371: 179–207, 1996. Carmichael, S.T., Clugnet, M.-C., and Price, J.L., Central olfactory connections in the macaque monkey, J Comp Neurol., 346: 403–434, 1994. Cavada, C. and Goldman-Rakic, P.S., Posterior parietal cortex in rhesus monkey: II. Evidence for segregated corticocortical networks linking sensory and limbic areas with the frontal lobe, J Comp Neurol., 287: 422–445, 1989. Critchley, H.D., Elliott, R., Mathias, C.J., and Dolan, R.J., Neural activity relating to generation and representation of galvanic skin conductance responses: a functional magnetic resonance imaging study, J Neurosci., 20: 3033–3040, 2000.

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 222 Friday, March 10, 2006 2:21 PM

222

Understanding Autism: From Basic Neuroscience to Treatment

Critchley, H.D., Corfield, D.R., Chandler, M.P., Mathias, C.J., and Dolan, R.J., Cerebral correlates of autonomic cardiovascular arousal: a functional neuroimaging investigation in humans, J Physiol., 523(Pt. 1): 259–270, 2000. Critchley, H.D., Rotshtein, P., Nagai, Y., O’Doherty, J., Mathias, C.J., and Dolan, R.J., Activity in the human brain predicting differential heart rate responses to emotional facial expressions, Neuroimage, 24: 751–762, 2005. Damasio, A.R., Descartes’ Error: Emotion, Reason, and the Human Brain, Grosset/Putnam, New York, 1994. Drevets, W.C., Videen, T.O., Price, J.L., Preskorn, S.H., Carmichael, S.T., and Raichle, M.E., A functional anatomical study of unipolar depression, J Neurosci., 12: 3628–3641, 1992. Drevets, W.C., Functional anatomical abnormalities in limbic and prefrontal cortical structures in major depression, Prog Brain Res., 126: 413–431, 2000. Dum, R.P. and Strick, P.L., Frontal lobe inputs to the digit representations of the motor areas on the lateral surface of the hemisphere, J Neurosci., 25: 1375–1386, 2005. Duncan, J. and Owen, A.M., Common regions of the human frontal lobe recruited by diverse cognitive demands, Trends Neurosci., 23: 475–483, 2000 Economo C. von and Koskinas, G.N., Die Cytoarchitektonik der Hirnrinde des Erwachsenen Menschen, Springer-Verlag, Berlin, 1925. Fellows, L.K. and Farah, M.J., Different underlying impairments in decision-making following ventromedial and dorsolateral frontal lobe damage in humans, Cereb Cortex, 15: 58–63, 2005. Funahashi, S., Bruce, C.J., and Goldman-Rakic, P.S., Mnemonic coding of visual space in the monkey’s dorsolateral prefrontal cortex, J Neurophysiol., 61: 331–334, 1989. Funahashi, S., Bruce, C.J., and Goldman-Rakic, P.S., Dorsolateral prefrontal lesions and oculomotor delayed-response performance: evidence for mnemonic “scotomas,” J Neurosci., 13: 1479–1497, 1993. Garey, L.J., Brodmann’s Localization in the Cerebral Cortex, Smith-Gordon, London, 1994. Gold, B.T., Balota, D.A., Kirchhoff, B.A., and Buckner, R.L., Common and dissociable activation patterns associated with controlled semantic and phonological processing: evidence from fMRI adaptation, Cereb Cortex., 2005. [Epub ahead of print.] Greicius, M.D., Krasnow, B., Reiss, A.L., and Menon, V., Functional connectivity in the resting brain: a network analysis of the default mode hypothesis, Proc Natl Acad Sci USA, 100: 253–258, 2003. Gusnard, D.A., Akbudak, E., Shulman, G.L., and Raichle, M.E., Medial prefrontal cortex and self-referential mental activity: relation to a default mode of brain function, Proc Natl Acad Sci U S A, 98: 4259–4264, 2001. Hagen, M.C., Zald, D.H., Thornton, T.A., and Pardo, J.V., Somatosensory processing in the human inferior prefrontal cortex, J Neurophysiol., 88: 1400–1406, 2002. Hikosaka, K. and Watanabe, M., Delay activity of orbital and lateral prefrontal neurons of the monkey varying with different rewards, Cereb Cortex, 10: 263–271, 2000. Hikosaka, K. and Watanabe, M., Long- and short-range reward expectancy in the primate orbitofrontal cortex, Eur J Neurosci., 19: 1046–1054, 2004 Kaada B.R., Pribram, K.H., and Epstein, J.A., Respiratory and vascular responses in monkeys from temporal pole, insula, orbital surface and cingulate gyrus: a preliminary report, J Neurophysiol., 12: 347–356, 1949. Kobayashi, Y. and Amaral, D.G., Macaque monkey retrosplenial cortex: II. Cortical afferents, J Comp Neurol., 466: 48–79, 2003. Kondo, H., Saleem, K.S., and Price, J.L., Differential connections of the temporal pole with the orbital and medial prefrontal networks in macaque monkeys, J Comp Neurol., 465: 499–523, 2003.

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 223 Friday, March 10, 2006 2:21 PM

Prefrontal Cortex

223

Kondo, H., Saleem, K.S., and Price, J.L., Differential connections of the perirhinal and parahippocampal cortex with the orbital and medial prefrontal networks in macaque monkeys, J Comp Neurol., 2005, in press. Kringelbach, M.L., O’Doherty, J., Rolls, E.T., and Andrews, C., Activation of the human orbitofrontal cortex to a liquid food stimulus is correlated with its subjective pleasantness, Cereb Cortex, 13(10): 1064–1071, October 2003. Lewis, J.W. and Van Essen, D.C., Corticocortical connections of visual, sensorimotor, and multimodal processing areas in the parietal lobe of the macaque monkey, J Comp Neurol., 428: 112–137, 2000 Mayberg, H.S., Liotti, M., Brannan, S.K., McGinnis, S., Mahurin, R.K., Jerabek, P.A., Silva, J.A., Tekell, J.L., Martin, C.C., Lancaster, J.L., and Fox, P.T., Reciprocal limbiccortical function and negative mood: converging PET findings in depression and normal sadness, Am J Psychiatry, 156: 675–682, 1999. Mayberg, H.S., Brannan, S.K., Tekell, J.L., Silva, J.A., Mahurin, R.K., McGinnis, S., and Jerabek, P.A., Regional metabolic effects of fluoxetine in major depression: serial changes and relationship to clinical response, Biol Psychiatry, 48: 830–843, 2000. Mayberg, H.S., Positron emission tomography imaging in depression: a neural systems perspective, Neuroimaging Clin N Am., 13: 805–815, 2003. Mayberg, H.S., Lozano, A.M., Voon, V., McNeely, H.E., Seminowicz, D., Hamani, C., Schwalb, J.M., and Kennedy, S.H., Deep brain stimulation for treatment-resistant depression, Neuron, 45: 651–660, 2005. McGaugh, J.L., 2004 The amygdala modulates the consolidation of memories of Morris, R., Petrides, M., and Pandya, D.N., Architecture and connections of retrosplenial area 30 in the rhesus monkey (Macaca mulatta), Eur J Neurosci., 11: 2506–2518, 1999. Nagai, Y., Critchley, H.D., Featherstone, E., Trimble, M.R., and Dolan, R.J., Activity in ventromedial prefrontal cortex covaries with sympathetic skin conductance level: a physiological account of a “default mode” of brain function, Neuroimage, 22: 243–251, 2004. O’Doherty, J., Rolls, E.T., Francis, S., Bowtell, R., McGlone, F., Kobal, G., Renner, B., and Ahne, G., Sensory-specific satiety-related olfactory activation of the human orbitofrontal cortex, Neuroreport, 11(4): 893–897, March 20, 2000. Öngür, D., An, X., and Price, J.L., Prefrontal cortical projections to the hypothalamus in macaque monkeys, J Comp Neurol., 401: 480–505, 1998. Öngür, D. and Price, J.L., The organization of networks within the orbital and medial prefrontal cortex of rats, monkeys and humans, Cereb Cortex, 10: 206–219, 2000. Öngür, D., Ferry, A.T., and Price, J.L., Architectonic subdivision of the human orbital and medial prefrontal cortex, J Comp Neurol., 460: 425–449, 2003. Passingham, R.E., Toni, I., and Rushworth, M.F., Specialization within the prefrontal cortex: the ventral prefrontal cortex and associative learning, Exp Brain Res., 133: 103–113, 2000. Patterson, Ungerleider, and Bandettini, Task-independent functional brain activity correlation with skin conductance changes: an fMRI study, Neuroimage, 17: 1797–1806, 2002. Petrides, M., Specialized systems for the processing of mnemonic information within the primate frontal cortex, Philos Trans R Soc Lond B Biol Sci., 351: 1455–1461, 1996. Petrides, M., Lateral prefrontal cortex: architectonic and functional organization, Philos Trans R Soc Lond B Biol Sci., 360: 781–795, 2005. Petrides, M. and Pandya, D.N., Projections to the frontal cortex from the posterior parietal region in the rhesus monkey, J Comp Neurol., 228: 105–116, 1984. Petrides, M. and Pandya, D.N., Comparative cytoarchitectonic analysis of the human and the macaque frontal cortex, in Boller, F. and Grafman, J., Eds., Handbook of Neuropsychology, Vol. 9, Elsevier Science B.V., Amsterdam,1994, pp. 17–58.

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 224 Friday, March 10, 2006 2:21 PM

224

Understanding Autism: From Basic Neuroscience to Treatment

Petrides, M. and Pandya, D.N., Dorsolateral prefrontal cortex: comparative cytoarchitectonic analysis in the human and the macaque brain and corticocortical connection patterns, Eur J Neurosci., 11: 1011–1036, 1999. Petrides, M. and Pandya, D.N., Comparative cytoarchitectonic analysis of the human and the macaque ventrolateral prefrontal cortex and corticocortical connection patterns in the monkey, Eur J Neurosci., 16: 291–310, 2001. Preuss, T.M. and Goldman-Rakic, P.S., Connections of the ventral granular frontal cortex of macaques with perisylvian premotor and somatosensory areas: anatomical evidence for somatic representation in primate frontal association cortex, J Comp Neurol., 282: 293–316, 1989. Preuss, T.M. and Goldman-Rakic, P.S., Myelo- and cytoarchitecture of the granular frontal cortex and surrounding regions in the strepsirhine primate Galago and the anthropoid primate Macaca, J Comp Neurol., 310: 429–474, 1991. Raichle, M.E. and Gusnard, D.A., Intrinsic brain activity sets the stage for expression of motivated behavior, J Comp Neurol., 2005, in press. Raichle, M.E., MacLeod, A.M., Snyder, A.Z., Powers, W.J., Gusnard, D.A., and Shulman, G.L., A default mode of brain function, Proc Natl Acad Sci U S A, 98(2): 676–682, January 16, 2001. Rempel-Clower, N.L. and Barbas, H., Topographic organization of connections between the hypothalamus and prefrontal cortex in the rhesus monkey, J Comp Neurol., 398(3): 393–419, August 31, 1998. Rolls, E.T., Taste and olfactory processing in the brain and its relation to the control of eating, Crit Rev Neurobiol., 11: 263–287, 1997 Rolls, E.T., The orbitofrontal cortex and reward, Cereb Cortex, 10: 284–294, 2000. Rolls, E.T., Taste, olfactory, and food texture processing in the brain, and the control of food intake, Physiol Behav., 85: 45–56, 2005. Romanski, L.M., Domain specificity in the primate prefrontal cortex, Cognit Affect Behav Neurosci., 4(4): 421–429, December 2004. Romanski, L.M., Tian, B, Fritz, J., Mishkin, M., Goldman-Rakic, P.S., and Rauschecker, J.P., Dual streams of auditory afferents target multiple domains in the primate prefrontal cortex, Nat Neurosci., 2: 1131–1136, 1999. Romanski, L.M. and Goldman-Rakic, P.S., An auditory domain in primate prefrontal cortex, Nat Neurosci., 5: 15–16, 2002. Sarkissov, S.A., Filimonoff, I.N., Kononova, E.P., Preobraschenskaya, I.S., and Kukuev, L.A., Atlas of the Cytoarchitectonics of the Human Cerebral Cortex, Medgiz, Moscow, 1955. Schultz, W., Tremblay, L., and Hollerman, J.R., Reward processing in primate orbitofrontal cortex and basal ganglia, Cereb Cortex, 10: 272–284, 2000. Schultz, W., Neural coding of basic reward terms of animal learning theory, game theory, microeconomics and behavioural ecology, Curr Opin Neurobiol., 14: 139–147, 2004. [Review.] Small, D.M., Zald, D.H., Jones-Gotman, M., Zatorre, R.J., Pardo, J.V., Frey, S., and Petrides, M., Human cortical gustatory areas: a review of functional neuroimaging data, Neuroreport, 10(1): 7–14, January 18, 1999. Small, D.M. and Prescott, J., Odor/taste integration and the perception of flavor, Exp Brain Res., 2005. [Epub ahead of print] Teves, D., Videen, T.O., Cryer, P.E., and Powers, W.J., Activation of human medial prefrontal cortex during autonomic responses to hypoglycemia, Proc Natl Acad Sci U S A, 101: 6217–6221, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C010.fm Page 225 Friday, March 10, 2006 2:21 PM

Prefrontal Cortex

225

Wagner, A.D., Schacter, D.L., Rotte, M., Koutstaal, W., Maril, A., Dale, A.M., Rosen, B.R., and Buckner, R.L., Building memories: remembering and forgetting of verbal experiences as predicted by brain activity, Science, 281: 1188–1191, 1998. Walker, A.E., A cytoarchitectural study of the prefrontal area of the macaque monkey, J Comp Neurol., 73: 59–86, 1940. Webster, M.J., Bachevalier, J., Ungerleider, L.G., Connections of inferior temporal areas TEO and TE with parietal and frontal cortex in macaque monkeys, Cereb Cortex, 4: 470–483, 1994. Wilson, F.A., Scalaidhe, S.P., Goldman-Rakic, P.S., Dissociation of object and spatial processing domains in primate prefrontal cortex, Science, 260(5116): 1955–1958, June 25, 1993. Zald, D.H. and Pardo, J.V., 2000 Functional neuroimaging of the olfactory system in humans, Int J Psychophysiol., 36: 165–181.

© 2006 by Taylor & Francis Group, LLC

11

The Social Brain, Amygdala, and Autism Cynthia M. Schumann, Melissa D. Bauman, Christopher J. Machado, and David G. Amaral

CONTENTS Introduction............................................................................................................228 Social Behavior: What and Where ........................................................................228 Role of the Amygdala in Social Behavior.....................................................229 Electrophysiological Recording and Functional Neuroimaging Studies ...............................................................................229 Lesion Studies with Nonhuman Primates .................................................231 Current Theories regarding the Amygdala and Social Behavior ..................231 The Amygdala Assigns an Emotional Significance to Social Stimuli ............................................................................................231 The Amygdala Detects Threat or Danger in the Environment...............................................................................................232 Behavioral Changes Resulting from Damage to the Amygdala in Adult Macaque Monkeys ..........................................232 Behavioral Changes Resulting from Damage to the Amygdala in Neonatal Macaque Monkeys.....................................233 Human Patients with Amygdala Lesions ..................................................236 Summary of Amygdala and Social Behavior............................................237 The Amygdala and Autism.............................................................................237 Functional MRI Studies of the Amygdala in Autism ...............................237 Structural MRI Studies of the Amygdala in Autism ................................238 Postmortem Studies of the Amygdala in Autism......................................240 Amygdala Dysfunction May Contribute to Symptoms of Autism.............................................................................243 Abnormal Amygdala Function May Impact Social Processing .......................................................................................244 Conclusions............................................................................................................245 References..............................................................................................................246

227

© 2006 by Taylor & Francis Group, LLC

228

Understanding Autism: From Basic Neuroscience to Treatment

INTRODUCTION The mechanistic definition of the social brain is fast becoming one of the most exciting and active areas of modern cognitive neuroscience (Blakemore et al., 2004; Insel and Fernald, 2004; Ochsner, 2004; Lieberman, 2005). Knowledge in this area not only addresses weighty issues related to the evolution of the brain and cognitive faculties that differentiate humans from other species, but also has implications for the behavioral pathology that underlies disorders as diverse as autism, schizophrenia, and Alzheimer’s disease (Insel and Fernald, 2004). Much has been written in recent years about brain regions that are putatively involved in one or more facets of social behavior. Ralph Adolphs, an articulate spokesperson for the field, has aptly stated that “social cognition is a domain with fuzzy boundaries and vaguely specified components” (Adolphs, 2001). Adolphs has provided a number of scholarly reviews on the state of research on the social brain (Adolphs, 2001, 2003a, 2003b), and the reader is referred to these papers for a comprehensive discussion of the behavioral components and putative brain regions involved in social cognition. However, much work remains to be done to adequately define the component processes of social behavior. Moreover, there is currently little clarity concerning whether certain brain regions have evolved specifically to mediate component processes of social behavior or whether the more general functions of these brain regions have been recruited in the service of social cognition. The general approach that we have taken in our nonhuman primate studies is to explore the dependency of component processes of social behavior on putative brain regions. This is done experimentally by making selective lesions of structures such as the amygdala and then carrying out detailed behavioral observations to define how the behavioral repertoire of the subjects has been altered. We have also explored the relationship of putative components of the social brain in the pathophysiology of autism. These studies are carried out by using either structural magnetic resonance imaging or postmortem stereological analyses. In this chapter, we will summarize findings from each of these research programs. However, we will begin with a short overview that defines the components of social behavior and the social brain.

SOCIAL BEHAVIOR: WHAT AND WHERE For humans and other group-living primates, the efficient production of speciestypical social behavior is of paramount importance. It contributes to the formation and maintenance of relationships with others that are critical for acquiring resources necessary to sustain life and ensuring the propagation of one’s genetic material. It has been speculated that the pressures of a complex social environment have selected for a sophisticated and flexible neural network in the primate brain that is charged with mediating social behavior (Dunbar, 2002). What processes are involved in social behavior? If one imagines the behavior resulting from a social stimulus such as a facial expression, a number of processes must take place. These include: (1) perception of the stimulus and (2) evaluation of its social significance. Whether a response will be generated will depend on (1) motivation to respond, (2) whether emotions such as fear are generated that might

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

229

Context

Evaluation

Motivation

Perception

Emotion

Social stimulus

Social behavior

FIGURE 11.1 Essential component process of social behavior.

modulate a response, and (3) whether the context is conducive to a social response. Finally, given adequate perception, evaluation, motivation, and context, the brain must execute an appropriate behavioral and physiological response (Figure 11.1). What brain regions are involved in social behavior? Obviously, some of the components of social behavior are mediated by brain regions that are common to many types of behavior. Thus, the visual perception of facial expression capitalizes on processing throughout much of the ventral stream visual pathway. However, both single-unit studies in the monkey (Perrett et al., 1988, 1992) as well as functional imaging studies in human subjects indicates that certain temporal lobe areas such as the fusiform gyrus appear to be particularly relevant to the perception of invariant aspects of facial structure (Kanwisher et al., 1997), whereas other brain regions such as the superior temporal gyrus evaluate the social significance of changing facial attributes such as eye gaze (Haxby et al., 2000). The brain regions that appear to be consistently implicated in aspects of social function are illustrated in Figure 11.2 (adapted from Adolphs, 2001). These include the amygdala (blue), the orbitofrontal cortex (red), the cingulate cortex (yellow), and the somatosensory cortex (green). The fusiform cortex and superior temporal gyrus are not indicated in this illustration. Of these regions, we have focused most of our attention thus far on the amygdala. As outlined in the next section, this brain region has been implicated in several component processes of social behavior including social motivation, evaluation of social significance, and generation of emotion in social situations.

ROLE

OF THE

AMYGDALA

IN

SOCIAL BEHAVIOR

The strong link between the amygdala, normal primate social behavior, and its potential dysfunction in neurodevelopmental disorders such as autism has been forged through convergent data generated across several different, yet complementary, experimental approaches. This evidence, along with current theories of amygdala function, will be described in the following sections. Electrophysiological Recording and Functional Neuroimaging Studies Electrophysiological recordings from the nonhuman primate amygdala have indicated that this structure is responsive to many features of the social environment, including particular individuals and genders (Brothers et al., 1990; Nakamura et al., © 2006 by Taylor & Francis Group, LLC

230

Understanding Autism: From Basic Neuroscience to Treatment

Somatosensory cortex

Amygdala

Cingulate cortex

Orbitofrontal cortex

FIGURE 11.2 (A color version of this figure follows page 236.) Brain regions implicated in the mediation of social behavior. (Adapted with permission from Adolphs, R. [2001], The neurobiology of social cognition, Curr Opin Neurobiol 11: 231–239.)

1992; Brothers and Ring, 1993), gross body or discrete limb movements (Brothers and Ring, 1993), as well as faces, specific facial expressions, and direct eye contact (Rolls, 1984; Leonard et al., 1985; Brothers et al., 1990; Brothers and Ring, 1993). Recordings from the human amygdala have also indicated heightened activity when patients view human face stimuli relative to objects (Fried et al., 2002). Kling and colleagues (1979) found that the nonhuman primate amygdala may also be sensitive to the level of potential danger indicated by social signals, because amygdala activity © 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

231

seems to be highest when the outcome of an interaction could have either positive or negative repercussions (i.e., receiving a threat, being chased or approached, etc), whereas activity is lower during less tense interactions (i.e., grooming, huddling, sitting alone, etc.). Similarly, noninvasive functional neuroimaging experiments have indicated that the amygdala can be activated by a variety of social stimuli, such as facial expressions depicting fear (Breiter et al., 1996; Morris et al., 1996; Morris, Friston et al., 1998; Canli et al., 2002), disgust (Phillips et al., 1997), anger (Morris, Ohman et al., 1998), and sadness (Blair et al., 1999), as well as several positive emotions (Breiter et al., 1996; Canli et al., 2002). In addition, the amygdala is also activated when subjects make complex social judgments based on faces, such as determining whether an individual is trustworthy (Winston et al., 2002) or discerning what someone may be thinking (Baron-Cohen et al., 1999). Furthermore, during a social attribution task, subjects were shown a video of three simple geometric shapes moving in a way that mimicked a social interaction. The amygdala was activated when subjects viewed the socially moving shapes, but not when the shapes moved randomly (Schultz et al., 2003). These results, along with those from electrophysiological recording studies, suggest that the amygdala and perhaps several other brain regions such as the fusiform gyrus, orbitofrontal cortex, and temporal pole, may be preferentially involved in detecting and/or decoding the meaning of social signals. Lesion Studies with Nonhuman Primates The earliest studies using amygdala lesions in nonhuman primates linked this structure to several aspects of social behavior, including affiliation (Dicks et al., 1968; Kling et al., 1970), aggression (Kling and Cornell, 1971), dominance (Rosvold et al., 1954), maternal care (Steklis and Kling, 1985), and sexual behavior (Kling, 1968). Typically, bilateral amygdala damage resulted in decreased affiliative behavior and subsequent social isolation when tested in seminaturalistic environments (Dicks et al., 1968; Kling et al., 1970; Kling and Cornell, 1971; Steklis and Kling, 1985). These early results led to the proposal that the amygdala is essential for interpreting and producing species-typical social behaviors (Brothers, 1990). However, variability in social testing conditions, social group size, species, gender, and lesion technique across these early studies produced results that were highly variable (Kling and Brothers, 1992).

CURRENT THEORIES

REGARDING THE

AMYGDALA

AND

SOCIAL BEHAVIOR

The experimental results described in the preceding text have generated several specific theories regarding how the amygdala interacts with other regions of the primate brain to mediate appropriate social interactions across multiple contexts. The Amygdala Assigns an Emotional Significance to Social Stimuli The extensive research efforts of Antonio Damasio (see Damasio, 1994, and Bechara et al., 2000, for reviews), Edmund Rolls (see Rolls, 1999, 2002, for review), Wolfram Schultz (see Schultz, 2002, for review), and their collaborators

© 2006 by Taylor & Francis Group, LLC

232

Understanding Autism: From Basic Neuroscience to Treatment

have contributed to the view that the amygdala may attach a learned emotional state or reward/punishment valence to social stimuli. The amygdala, in conjunction with the orbitofrontal cortex, may also facilitate appropriate behavioral responses through interactions with the basal ganglia, basal forebrain cholinergic system, and noradrenergic system. The Amygdala Detects Threat or Danger in the Environment Another emerging theory suggests that one function of the amygdala is to evaluate environmental stimuli for potential danger and to coordinate an appropriate behavioral and physiological response (Davis, 1992; LeDoux, 2000; Pare et al., 2004). This idea is based on studies which demonstrated that lesions or temporary inactivation of the rodent amygdala interfere with the acquisition and expression of conditioned fear (LeDoux et al., 1990; Killcross et al., 1997; Wilensky et al., 1999; Amorapanth et al., 2000) and alter species-typical fear responses in nonhuman primates (Emery et al., 2001; Kalin et al., 2001; Prather et al., 2001). Similarly, the human amygdala is activated in response to a variety of fear-inducing or aversive stimuli (Whalen et al., 1998; Phelps et al., 2001; Dilger et al., 2003; Fredrikson and Furmark, 2003; Hariri et al., 2003). Thus, the amygdala may play a modulatory role in social behavior by regulating one’s perception and response to environmental danger, particularly social threats (Amaral et al., 2003). This theory is supported by recent amygdala lesion studies in adult and infant nonhuman primates indicating that a functional amygdala is not needed for the production of social behavior. Rather, the amygdala may play a modulatory role by regulating fear behaviors during social interactions. Behavioral Changes Resulting from Damage to the Amygdala in Adult Macaque Monkeys Although the amygdala has been cast as a major component of the social brain for some time now, specifically testing and defining its function has been complicated by a number of methodological shortcomings (see Bachevalier and Meunier, 2005, for review). In the past decade, several laboratories have reexamined the role of the amygdala in social behavior and emotional processing. In one particularly relevant study, we investigated the impact of selective bilateral amygdala lesions on adult primate social behavior (Emery et al., 2001). Guided by presurgical magnetic resonance images (MRI), a neurotoxin (ibotenic acid) was stereotaxically injected bilaterally into the amygdala of six adult rhesus monkeys of middle social dominance rank. To precisely identify and quantify any differences in social behavior between operated monkeys and unoperated control animals, a specifically defined behavioral ethogram was developed that classified 40 discrete macaque social and nonsocial behaviors. Given the previous literature, we expected that these selective bilateral amygdala lesions would result in operated animals being less affiliative and more socially withdrawn during social interactions with age-matched conspecifics. However, during dyadic interaction, animals with amygdala lesions were not socially withdrawn.

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

233

Instead, operated monkeys displayed more social signals and social behaviors toward the stimulus animals than control animals, especially during initial encounters. Furthermore, stimulus animals sought out animals with amygdala lesions more often for social interactions (social contact, grooming, or sitting in close proximity) than unoperated controls. This suggests that the operated animals were perceived as less threatening and more attractive as social partners. The result from this study that was consistent with previous findings was that operated animals showed increased oral and tactile exploration of the testing environment. These results suggest that the amygdala is not necessarily involved in the generation of species-typical social behavior. However, the amygdala damage did appear to decrease the normal amount of social inhibition displayed by macaques when they encounter an unfamiliar social partner. Therefore, it is possible that the amygdala is a critical neural structure for restraining social behavior until an adequate assessment of the intentions and disposition of a novel social partner can be made. This conclusion has recently been substantiated by Málková and colleagues (2003), who found that intra-amygdala infusion of the GABAA antagonist bicuculline, which effectively disinhibits the amygdala by blocking the inhibitory affect of GABA, resulted in decreased social contact, a complete loss of social play, and an increase in active withdrawal in monkeys. Further support for the notion that the amygdala restrains inappropriate behavior in times of potential danger comes from recent studies of monkeys in nonsocial settings. These studies have demonstrated a consistent lack of inhibition and avoidance of dangerous stimuli (such as a rubber snake) following bilateral neurotoxic lesions of the entire amygdala (Meunier et al., 1999; Kalin et al., 2001), neurotoxic lesions of the central amygdaloid nucleus alone (Kalin et al., 2004), or surgical disconnection of the amygdala and orbitofrontal cortex in monkeys (Izquierdo and Murray, 2004). Furthermore, when the same animals from the study by Emery and colleagues (2001) were examined in tests that specifically probed their behavioral reactions to stimuli that normal macaques find aversive or potentially dangerous (rubber snake, unfamiliar staring human, etc.), a generalized lack of behavioral inhibition was found (Mason et al., in press). For example, animals with amygdala lesions, relative to control animals, were faster to contact novel movable metal objects, spent more time near a human who was standing near the test cage, were quicker to take a food reward positioned near complex, animal-like stimuli, spent more time touching the objects, and showed a longer duration of contact with objects as the complexity of their features increased. In fact, this study also found a significant positive correlation between the disinhibited behavior of animals with amygdala lesions in this nonsocial context and with the magnitude of their disinhibited behavior displayed in a social context (Emery et al., 2001). Behavioral Changes Resulting from Damage to the Amygdala in Neonatal Macaque Monkeys The results of the Emery et al. (2001) study suggest that the amygdala is not a critical component of the neural circuitry required for fundamental aspects of primate social behavior. One caveat to this interpretation is that these subjects were adult

© 2006 by Taylor & Francis Group, LLC

234

Understanding Autism: From Basic Neuroscience to Treatment

monkeys, which acquired the fundamental tools of social behavior much earlier in development, prior to the lesion. Although the amygdala is not needed to produce social behavior in adulthood, it remains possible that the amygdala may be needed to gain social knowledge at earlier time points. Initial attempts to evaluate the role of the amygdala in early development were carried out by Kling and colleagues (1967). They created bilateral amygdala lesions in infant macaques within the first postnatal week. Following the surgery, the amygdala-lesioned animals were successfully returned to their mothers, developed normal reflexes, and did not display obvious behavioral abnormalities, leading the authors to suggest that early damage to the amygdala did not alter behavioral development. However, an equally plausible explanation suggests that the lack of reported behavioral abnormalities may also be attributed to the relatively limited assessment of behavioral development. A more extensive behavioral assessment following early amygdalectomy suggested that early lesions do alter social development (Kling and Green, 1967; Thompson et al., 1969; Thompson and Towfighi, 1976; Thompson et al., 1977; Thompson, 1981). Six individually reared female rhesus macaques received bilateral amygdalectomies at approximately two months of age and were then observed in social interactions with conspecifics. During these social interactions, the amygdalalesioned infants produced far more fear behaviors than control subjects. The social behavior of these amygdala-lesioned subjects became increasingly abnormal over time as they initiated fewer social interactions and became exceedingly submissive to conspecifics. More recently Bachevalier and colleagues (1994) reevaluated the effects of early amygdala damage. In this study, six peer-reared monkeys received aspiration lesions of the amygdala within the first postnatal month. When placed in social dyads at two months of age, the neonatal amygdala-lesioned infants showed less overall activity, exploration, and social behavior initiation than age-matched controls. At six months of age, activity levels were relatively normal, but social interactions were reduced in amygdalectomized infants compared to the controls. The authors also reported that more extensive lesions of the medial temporal lobe, including the amygdala, hippocampus, and ventromedial temporal cortex produced a more profound effect on social interactions, including “lack of social skills,” flat affect, and increased stereotypic behaviors. Given that impaired social communication and a lack of social interest is the hallmark of autism, the authors proposed that lesions of the medial temporal lobe, specifically the amygdala, might provide an animal model of autism (Bachevalier, 1994, 2000). The disparate results of previous neonatal amygdala lesion studies (i.e., no behavioral changes, changes in fear behavior, changes in social interest, etc.) may be explained by methodological differences. As in the adult lesion studies, the lesion technique and relatively limited behavioral observations may complicate the interpretations of prior neonatal lesion studies. In addition, the rearing conditions of the infant monkeys introduce a methodological complication that is specific to developmental studies (Capitanio, 1986). Rearing nonhuman primates in isolate conditions, or with only their mothers, or with only peers has a profound effect on social and neurobiological development (Mason, 1960; Harlow et al.,

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

235

1965; Griffin and Harlow, 1966; Haxby et al., 2002; Winslow et al., 2003). It is possible that these restricted rearing conditions may exacerbate behavioral changes associated with brain lesions. Previous infant amygdala lesion research has utilized isolation-reared infants (Thompson et al., 1969), mother-only-reared infants (Kling and Green, 1967), or peer-only-reared infants (Bachevalier, 1994), but no study had examined amygdala-lesioned infants raised in a social setting. Thus, it was unknown if the behavioral deficits observed in previous neonatal amygdala lesion studies were due to the brain lesion, the rearing environment, or a combination of the two factors. To address this question, our laboratory designed a series of experiments to evaluate the effects of early amygdala damage in socially reared rhesus monkeys (Prather et al., 2001; Bauman et al., 2004a, 2004b). We predicted that if the amygdala is a core component of the social brain, then removal of the amygdala early in development would profoundly alter fundamental features of social behavior, including social attention and the ability to produce and interpret social signals. However, if these core features of social behavior remain intact following amygdala damage, then it is reasonable to conclude that the amygdala is not essential for the development of fundamental aspects of social behavior. Twenty-four rhesus monkeys were given selective ibotenic-acid-induced lesions of the amygdala or hippocampus, or a sham lesion procedure at two weeks of age. These infants were reared by their mothers and given daily access to other monkeys to simulate features of the social organization of free-ranging macaques (Berman, 1980), which appear necessary to facilitate species-typical social and hormonal development (Mason, 1960; Mason and Green, 1962; Mason and Sponholz, 1963; Shannon et al., 1998; Bastian et al., 2003; Winslow et al., 2003). We then observed the social development of these animals during the first year of life, by systematically quantifying social interactions with their mothers, familiar peers, and novel peers. When observed at three months of age, the amygdala-lesioned monkeys spent the same amount of time nursing and in contact or proximity with their mothers as did control or hippocampus-lesioned subjects (Bauman et al., 2004b). There were no differences among the experimental groups in the frequency of mother–infant behaviors, including maternal rejection, maternal restraint, maternal retrieval, aggression, threats, or grooming. Furthermore, none of the infants engaged in maladaptive behaviors, such as self-clasping, crouching, rocking, or motor stereotypies, which are indicative of abnormal social behavior development (Capitanio, 1986). After the amygdala-lesioned subjects were weaned from their mothers at six months of age, they continued to develop a species-typical repertoire of social behavior (Bauman et al., 2004a). They were observed successfully interacting with members of their social rearing group, with familiar conspecifics at six and nine months of age, and with novel conspecifics at 1 yr of age, indicating that a functional amygdala is not needed to develop fundamental aspects of social behavior. In fact, our research demonstrated that infant macaques with neonatal amygdala lesions are able to (1) form filial bonds, (2) develop a species-typical repertoire of social behavior, (3) display interest in conspecifics, and (4) interact with conspecifics in various social contexts.

© 2006 by Taylor & Francis Group, LLC

236

Understanding Autism: From Basic Neuroscience to Treatment

Amygdala Control Hippocampus 3.0

∗

∗

0.5 0.4 ∗

2.0

∗

∗

∗

0.3

∗

∗

0.2

1.0

0.1 0

0 Fear grimace

Flee

Scream

Freeze

FIGURE 11.3 Neonatal amygdala-lesioned monkeys produced more fear behaviors during novel dyadic interactions than did control or hippocampus-lesioned monkeys. Each bar represents the average number of fear behaviors ±SEM per 5-min observation period across all 72 observation periods. Asterisks denote significant post hoc Fisher PLSD tests (p < .05).

Although the amygdala-lesioned subjects developed fundamental aspects of social behavior, they differed from control and hippocampus-lesioned subjects in some significant ways. The most striking abnormality of the amygdala-lesioned subjects was their inappropriate fear behaviors in various contexts (Figure 11.3). We observed that amygdala-lesioned subjects produced fear behaviors more frequently than control or hippocampus-lesioned subjects during social interactions with both novel and familiar social partners (Bauman et al., 2004a). We have proposed that through its mediation of fear, the amygdala plays a modulatory, rather than essential, role in social processing. Human Patients with Amygdala Lesions One of the most extensively studied individuals with a selective and complete bilateral amygdala lesion, patient S.M., developed her lesion during adolescence from Urbach-Wiethe disease (Adolphs et al., 1994; Bechara et al., 1995). Patient S.M.’s social behavior remains relatively intact, despite her lack of amygdala function. Human patients with damage to the amygdala, including S.M., do display deficits in fear conditioning (Bechara et al., 1995; LaBar et al., 1995) and recognizing emotions in facial expressions, primarily fear (Adolphs et al., 1994; Adolphs et al., 1995; Bechara et al., 1995). This deficit in processing fear and potential threat may also affect more complex social judgments. Patients with bilateral amygdala damage, including S.M., were impaired in judging the trustworthiness of another person from viewing a photo of their face (Adolphs et al., 1998). The patients judged the people in the photos as more trustworthy and more approachable than did normal viewers. Recently, Adolphs et al. (2005) found that patient S.M. is impaired in her ability to make normal use of information from the eye region of the face when judging

© 2006 by Taylor & Francis Group, LLC

Environment

Genotype

Phenotype Epigenotype

Stochastic events

COLOR FIGURE 5.1 The genotype affects the phenotype only through the prism of the epigenotype. (Taken from Beaudet, A.L., Is medical genetics neglecting epigenetics? Genet Med 4, 399–402, 2002. With permission.)

Prader-Willi Deletion Genetic

Angelman UPD

Deletion

Epigenetic

Genetic

70%

30%

Paternal deficiency 15q11-q13

70%

UPD Epigenetic

Rare

Maternal deficiency 15q11-q13

All deletions and UPD are de novo events

COLOR FIGURE 5.2 Depiction of genomic deletion or uniparental disomy (UPD) giving rise to Prader–Willi or Angelman syndromes. (Modified from Jiang, Y.H., Bressler, J., and Beaudet, A.L., Epigenetics and human disease, Annu Rev Genomics Hum Genet 5, 479–510, 2004. With permission.)

© 2006 by Taylor & Francis Group, LLC

E6-AP

bp1

bp2

UBE3A sense UBE3A antisense

IC-DMR

bp3

PWS proteins & RNAs

ER C2

P

H

3A BE U

AT P1 0A G AB G RB AB 3 G RA AB 5 RG 3

As N oR

sn

N

cen

IP A N 1 IP C Y A2 FI G P1 CP S M KR M N3 AG N EL D 2 N C1 5 SN OR U F2 RF -S N

RP N

5-6 Mb region tel

COLOR FIGURE 5.4 Genomic structure of the imprinted domain of human chromosome 15q11-q13. In the top section, black arrows represent products from nonimprinted genes, blue arrows products expressed preferentially from the paternal chromosome, and pink arrows products expressed preferentially from the maternal chromosome. E6-AP is the protein produced from the Angelman ubiquitin ligase gene UBE3A. IC-DMR is the imprinting center differentially methylated region. The identity of gene symbols can be found in any genome browser. Autism?

Angelman

De novo De novo De novo & & & inherited inherited inherited

De novo

De novo

Deletion

UPD

Imprint defect

UBE3A null

Genetic

Epigenetic

Mixed

Genetic

De novo

?

?

?

Interstial Isodiduplic centric

Paternal imprint defect?

Maternal imprint defect?

Other gene ?

Genetic Genetic

Mixed?

Mixed?

Mixed?

?

?

∗

Frequent?

COLOR FIGURE 5.6 Depiction of the MEGDI model for Angelman syndrome and autism. The various molecular forms of Angelman syndrome are depicted on the left. The interstitial duplication and isodicentric examples for autism are well documented. The DNA methylation abnormality in one of 17 autism brains likely represents a paternal or maternal imprinting defect.

© 2006 by Taylor & Francis Group, LLC

AAAA

FMRP

Kinase

Phosphatase

FMRP

P

AAAA

COLOR FIGURE 6.1 Model for a regulatory role of phosphorylation in FMRP suppression of translation. Phosphorylated FMRP is known to preferentially associate with stalled ribosomes as compared to nonphosphorylated FMRP. Thus, it is possible that the FMRP-specific kinase and phosphatase maintain a balance between phosphorylated and nonphosphorylated FMRP, ultimately regulating translational suppression of the FMRP mRNA ligands. mGluR 1/5

mGluR 1/5

+

+ AAAA

AAAA

− FMRP

Fragile X syndrome FMRP

COLOR FIGURE 6.2 The mGluR theory. The activation of the group I metabotropic receptors (mGluR) is known to result in increased translation of messages (including FMRP ligands) in the dendrite. In the presence of FMRP, translation of these mRNA ligands is suppressed. However, in the case of Fragile X, the absence of FMRP leads to increased ligand translation. Ultimately, this disrupts the translational regulation in the dendrites, contributing to the synaptic deficits in Fragile X.

© 2006 by Taylor & Francis Group, LLC

30

Extinction

% Active voxels

Acquisition

20

10

0 0

2 4 Normalized difference GSR

6

COLOR FIGURE 7.8 Amygdala activation and skin conductance responses in healthy adults during acquisition and extinction of conditioned fear. Participants acquired and extinguished a visual CS–shock US association while undergoing functional magnetic resonance imaging. Left panel: Group-averaged hemodynamic responses elicited by the CS are displayed in a coronal image through the amygdala (green box). Data are averaged from the first half of trials for each phase of training (acquisition, extinction). Right panel: The extent of amygdala activation elicited during acquisition was correlated with the magnitude of skin conductance responses in a subset of participants. GSR = galvanic skin response. (Adapted from LaBar, K.S. et al., Human amygdala activation during conditioned fear acquisition and extinction: a mixed trial fMRI study, Neuron, 20, 937, 1998. With permission.) © 2006 by Taylor & Francis Group, LLC

COLOR FIGURE 10.1 Architectonic maps of the lateral prefrontal cortex in humans (above) and monkeys (below), as delineated by Petrides and Pandya (1994, 1999, 2001). The dorsomedial, ventrolateral, and caudolateral domains discussed in the text are marked by red, yellow, and green overlays, respectively.

© 2006 by Taylor & Francis Group, LLC

COLOR FIGURE 10.2 Architectonic maps of the ventral (orbital) and medial prefrontal cortex in humans (above) and monkeys (below), as delineated by Carmichael and Price (1994) and Öngür et al. (2003). The dorsomedial domain and the medial and orbital prefrontal networks discussed in the text are marked by red, blue, and beige overlays, respectively. © 2006 by Taylor & Francis Group, LLC

COLOR FIGURE 10.3 Summary of major connections of the prefrontal domains discussed in the text. © 2006 by Taylor & Francis Group, LLC

COLOR FIGURE 11.2 Brain regions implicated in the mediation of social behavior. (Adapted with permission from Adolphs, R. [2001], The Neurobiology of Social Cognition, Curr Opin Neurobiol 11: 231–239.)

COLOR FIGURE 11.4 MRI showing human amygdala (red) and adjacent hippocampus (blue).

© 2006 by Taylor & Francis Group, LLC

(a)

(b)

COLOR FIGURE 11.5 Linear regression scatterplot for absolute amygdala volume (cm3) by age. Typically developing subjects show a positive correlation of age with amygdala volume for both the (a) left and (b) right amygdala (∗p < .05). Amygdala volume in participants with autism did not correlate with age. Abbreviations: LFA, participants with low-functioning autism; HFA, participants with high-functioning autism; ASP, participants with Asperger syndrome; CON, typically developing control participants. (From Schumann, C.M. et al., 2004, J Neurosci, 24(28), 6392–6401. With permission.)

COLOR FIGURE 16.2 Cerebral enlargement in 2- and 3-yr-olds with autism was greatest in the frontal lobe (green) followed by the parietal (yellow) and temporal (blue) lobes.9 Striping in frontal areas indicates primary and secondary motor cortex. Striping in other areas indicates primary sensory (somatosensory, auditory, and visual), and assorted unimodal association areas. The ranking of overgrowth in autism also corresponds with the relative proportions of higher-order multimodal association cortex (unpatterned areas). © 2006 by Taylor & Francis Group, LLC

A

B

COLOR FIGURE 16.4 A typical anatomical MRI, with resolution around 1 mm, is unable to show the detail actually present in the cerebellum. Where the convolutions of the cerebellum are smaller than the voxel size, the MRI voxels must necessarily contain part white and part gray matter, resulting in some intermediate intensity level.

© 2006 by Taylor & Francis Group, LLC

Deform calc

T1 MRI

Reference MRI Deform Deformation field

Deformed “spatially normalized”

Segment

CSF

Grey

Other subjects, two or more groups

White

Blur

“Grey concentration”

Voxelwise statistics (SPM)

Significant differences map

COLOR FIGURE 16.5 A simplified conceptual view of VBM, incorporating deformation (“warp”), tissue segmentation, and blur, followed by statistics across multiple subjects and groups. More recent versions of VBM integrate the warp and segment processing so that the registration process benefits from segmentation. Recent versions are also more sophisticated about how voxel contributions are copied from original volume to warped volume, weighting for the degree to which each voxel is stretched or shrunk.

© 2006 by Taylor & Francis Group, LLC

A

B

COLOR FIGURE 16.6 Surface reconstruction software, such as Freesurfer (http://surfer. nmr.mgh.harvard.edu), calculates a surface that is represented as a mesh of triangles. (A) A view of the surface where white matter meets gray matter. Green indicates convex surfaces, and red indicates concave surfaces. (B) A magnified view of the same reconstruction with the mesh visible on the surface.

COLOR FIGURE 17.1 Activation during a spatial working memory task. (Courtesy of Dr. John Sweeney and Dr. Beatriz Luna.)

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

237

emotions and that the eyes may be the most notable feature for identifying emotions such as fear. Patient S.M.’s deficits extend beyond the processing of faces. As mentioned previously, typically developing individuals will attribute social meaning to ambiguous, moving geometric shapes that appear to have goal-directed movement and intentions, i.e., the social attribution task (Heider and Simmel, 1944), whereas patient S.M.’s description is limited to strictly physical, asocial, and geometric terms (Heberlein and Adolphs, 2004). Summary of Amygdala and Social Behavior There have been relatively few experimental studies of the role of the amygdala in social behavior. And there are an extremely small number of patients with selective and complete bilateral lesions of the amygdala. We are impressed by the fact that human subjects such as S.M. can function in a complex social environment with only very subtle impairments. This individual has been married, is raising children, and has been normally employed. Our monkeys that received bilateral amygdala lesions as adults or infants demonstrate an essentially normal repertoire of social interactions. Moreover, they appear to be able to perceive and generate speciestypical social communications. They are motivated to interact because they are involved in more, rather than less, affiliative behaviors. What is perturbed in these animals is the amount of social behavior and their ability to produce an appropriate fear response to a stimulus. Thus, in relation to Figure 11.1, we would propose that the amygdala is involved in the generation of emotion to a social stimulus that, depending on the context, may either inhibit or facilitate social interaction. The amygdala is not essential for social behavior but is a modulatory influence on its expression.

THE AMYGDALA

AND

AUTISM

Given the variation and complexity of symptoms seen across the autistic spectrum, it is probable that several areas of the brain may develop abnormally. The amygdala, given its involvement in the production and recognition of emotions, is a logical candidate to study in autism. In addition, patients with amygdala lesions and individuals with autism share some common deficits. Although patients with amygdala lesions are clearly not autistic, Adolphs and colleagues (2001) found that individuals with autism perform similar to patients with amygdala lesions in judging people to be more trustworthy and approachable than normal individuals. Klin (2000) also found that individuals with autism provide narrations similar to patient S.M. that are strictly limited to physical, asocial, and geometric terms (Heberlein and Adolphs, 2004). Functional MRI Studies of the Amygdala in Autism Previous functional neuroimaging studies have indicated that individuals with autism spectrum disorders show abnormal patterns of amygdala activation in response to social stimuli. High-functioning adults with autism or Asperger

© 2006 by Taylor & Francis Group, LLC

238

Understanding Autism: From Basic Neuroscience to Treatment

syndrome demonstrate deficits in the ability to judge from images of another person’s eyes what that person might be thinking (Baron-Cohen et al., 1997). When combined with functional imaging, this task revealed that control subjects activate the amygdala and superior temporal gyri when inferring the mental/emotional state of another person. In contrast, individuals with autism or Asperger syndrome activated the frontotemporal regions but not the amygdala when making social inferences from the eyes (Baron-Cohen et al., 1999). Another functional imaging study revealed that adults with autism did not activate the left amygdala and left cerebellum region when implicitly processing emotional facial expressions (i.e., attending to gender of the stimuli), but demonstrated relatively preserved amygdala function when explicitly evaluating emotion (Critchley et al., 2000). Pierce et al. (2001) found that the amygdala was activated when typically developing individuals viewed unfamiliar faces, but that it was not activated in individuals with autism during this task. Children and adolescents with autism spectrum disorders have also shown abnormal amygdala activation while matching faces by emotion and assigning a label to facial expressions (Wang et al., 2004). Although children in the control group showed more amygdala activation when matching faces by emotion than assigning a verbal label, the children with autism spectrum disorder did not demonstrate this pattern of task-dependent amygdala modulation. However, one caveat to interpreting findings from face-processing studies is that subjects with autism are reluctant to make eye contact, and there is some controversy as to whether they are actually examining the face in a similar manner as controls (Davidson and Slagter, 2000). Structural MRI Studies of the Amygdala in Autism Until recently, structural MRI studies painted an unclear and inconsistent picture of the amygdala in autism. Some studies report decreased volume (Aylward et al., 1999; Pierce et al., 2001), others report increased volume (Howard et al., 2000; Sparks et al., 2002; Schumann et al., 2004; Mosconi et al., 2005), and still others find no difference in volume (Haznedar et al., 2000; Schumann et al., 2004) in individuals with autism. These studies vary in the age groups studied, diagnostic and exclusionary criteria, and neuroanatomical methods for defining the amygdala in MRI scans. Aylward et al. (1999) manually traced the amygdala in 14 cases of high-functioning autism with ages ranging from 11 to 37 years and reported decreased amygdala volume in autism subjects compared to age-matched control cases. Pierce et al. (2001) analyzed six cases of high-functioning autism with ages ranging from 23 to 41 years also reporting amygdala volumes to be significantly smaller by approximately 15% relative to controls. Howard et al. (2000) employed both manual tracing and stereological point counting in ten males with high-functioning autism and Asperger syndrome ranging from 15 to 40 years of age. They reported the volume of the amygdala to be larger in subjects with autistic spectrum disorder relative to controls. Sparks et al. (2002) were the only group to measure the volume of the amygdala in young children (36 to 56 months of age). They found that in males with autism, the amygdala is larger by 16% on the right and 13% on the left, relative

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

239

FIGURE 11.4 (A color version of this figure follows page 236.) MRI showing human amygdala (red) and adjacent hippocampus (blue).

to controls. Studies from other groups are now emerging that confirm this finding of a larger amygdala in very young children (Mosconi et al., 2005). We recently carried out a study to (1) compare volume measurements of the amygdala in children across the autistic spectrum and (2) attempt to reconcile contradictory results in previously published MRI studies (Schumann et al., 2004). The volume of the amygdala was measured in 85 male children 7 to 18 years of age in four diagnostic groups: low-functioning autism; high-functioning autism; Asperger syndrome; and age-matched, typically developing controls (Figure 11.4). One striking finding is that the amygdala in typically developing male children increases in size by approximately 40% from 7 to 18 years of age. This finding is consistent with studies from Giedd and colleagues (1996) and Giedd (1997) who report a 50% increase in volume from 4 to 18 years of age in males, but not females. However, we found that children with autism do not undergo this same pattern of development (Figure 11.5). The amygdala in young children 7 to 12 years of age with autism is initially larger than controls by approximately 15%. We found a significant difference in amygdala volume in both low- and high-functioning children with autism, indicating that the difference is related to autism rather than to mental retardation. This enlargement was not paralleled by an overall enlarged brain, because there was no difference in total cerebral volume in this age range. There was no difference in the volume of the amygdala between children with autism and typically developing controls aged 13 to 18 years. Thus, the amygdala is initially larger than normal in the children with autism, but does not undergo the

© 2006 by Taylor & Francis Group, LLC

240

Understanding Autism: From Basic Neuroscience to Treatment

2.75 Right amygdala volume (cm3)

Left amygdala volume (cm3)

2.75 2.50 2.25 2.00 1.75 1.50 8.0

10.0

12.0 14.0 Age (a)

16.0

2.50 2.25 2.00 1.75 1.50

18.0

8.0

10.0

12.0 14.0 Age (b)

16.0

18.0

LFA HFA ASP CON

FIGURE 11.5 (A color version of this figure follows page 236.) Linear regression scatterplot for absolute amygdala volume (cm3) by age. Typically developing subjects show a positive correlation of age with amygdala volume for both the (a) left and (b) right amygdala (∗p < .05). Amygdala volume in participants with autism did not correlate with age. Abbreviations: LFA, participants with low-functioning autism; HFA, participants with highfunctioning autism; ASP, participants with Asperger syndrome; CON, typically developing control participants. (From Schumann, C.M. et al., 2004, J Neurosci, 24(28), 6392–6401. With permission.)

same age-related increase in volume that takes place in typically developing children. These findings help to explain the variability in reports from previous MRI studies of individuals with autism. In younger children, the amygdala is larger in those with autism relative to age-matched, typically developing controls (Sparks et al., 2002; Schumann et al., 2004; Mosconi et al., 2005). However, studies focused primarily on adults or a wide age range of subjects have found no difference in (Haznedar et al., 2000) or have found potentially smaller (Aylward et al., 1999; Pierce et al., 2001) amygdala volumes in individuals with autism. Although we found that the amygdala of older children with autism is approximately the same size, we would predict that there are fundamental abnormalities in the neuroanatomical and functional organization of the amygdala in individuals with autism that results from the abnormal developmental time course. These differences would likely persist into adulthood. Postmortem Studies of the Amygdala in Autism Bauman and Kemper (1985, 1994) were the first to report abnormalities in the microscopic organization of the amygdala in postmortem autism cases. Their initial report was of a 29-year-old male with autism, seizure disorder, and mental retardation compared to a 25-year-old typically developing control male (Bauman and Kemper, 1985). Nissl-stained whole-brain serial sections from the autism and control

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

241

case were viewed side-by-side under a microscope at the same magnification. Density measures were made in the central part of each cytoarchitectonic region. Bauman and Kemper observed increased cell packing density in the central, medial, and cortical nuclei (40%, 28%, and 35%, respectively) in the autism case. They also noted that cell size in these areas was reduced. The basal, lateral, and accessory basal nuclei showed little difference. Kemper and Bauman (1993) followed up their initial case study with five additional cases of autism (four males and one female) aged 9, 10, 12, 22, and 28. Nisslstained sections of brain tissue from autism cases were compared to age-matched controls in which corresponding areas were again viewed side-by-side under a microscope. Qualitative observations indicated that neurons in the amygdala of autism cases appeared unusually small and more densely packed than in age-matched controls. This was most pronounced in the cortical, medial, and central nuclei, whereas the lateral nucleus generally appeared to be comparable to controls. The basal nucleus of the amygdala also showed an intermediate degree of involvement. Kemper and Bauman (1993) suggested that densely packed amygdala neurons may manifest during an early stage of maturation, a time at which the neuronal size and complexity of neuropil have not reached adult proportions. These changes could result from a curtailment of normal maturation. The results of Bauman and Kemper are complicated by the fact that four of the six cases had a seizure disorder. Studies focusing on cases of epilepsy without autism indicate a reduction in amygdala volume of 10 to 30%, with neuronal cell loss reported in the lateral and basal nuclei of the amygdala (Pitkanen et al., 1998). In addition, recent studies have raised methodological concerns about the interpretation of density measurements as an indication of neuropathology. Tissue undergoes variable shrinkage during processing, and the only way to unambiguously interpret pathological changes in cell number or density is to estimate actual neuron number in the entire amygdala. As described earlier in detail, MRI studies have found the amygdala to be larger in young children with autism compared to age-matched controls (Sparks et al., 2002; Schumann et al., 2004; Mosconi et al., 2005). One possibility to account for this greater volume is that there is a greater number of neurons in the autistic amygdala. This would be consistent with the findings of Bauman and Kemper (1985, 1994). We recently carried out a study using unbiased stereological methods to estimate the number of neurons in the autistic amygdala (Schumann and Amaral, submitted). The goal of our study was to measure neuron number, regional volume, neuronal density, and mean neuronal cross-sectional area in the entire amygdaloid complex and in individual nuclei in male postmortem cases of autism, without seizure disorder, compared to typically developing, age-matched male controls. The intact amygdala was collected from one hemisphere of 9 autism and 10 age-matched control brains 10 to 44 years of age at death. A principle of design-based stereological techniques is that the entire area of interest must be reliably sampled. Prior to initiating our study of autism cases, we extensively defined the borders of the amygdaloid complex in 10 control cases (Schumann and Amaral, 2005). We then outlined the amygdaloid complex on every 100-µm Nissl section in which it was present (approximately 20 to 25 sections per case) on the blinded control and autism cases. The amygdaloid

© 2006 by Taylor & Francis Group, LLC

242

Understanding Autism: From Basic Neuroscience to Treatment

VC

M

C

OT

I AB

PAC

B

SAS

L

PL

EC

FIGURE 11.6 Brightfield photomicrograph of Nissl-stained coronal section through midrostrocaudal amygdala. Abbreviations: (AB) accessory basal nucleus, (B) basal nucleus, (C) central nucleus, (EC) entorhinal cortex, (I) intercalated nuclei, (L) lateral nucleus, (M) medial nucleus, (OT) optic tract, (PAC) periamygdaloid cortex, (PL) paralaminar nucleus, (SAS) semiannular sulcus, and (VC) ventral claustrum.

complex was further partitioned into five subdivisions: (1) lateral nucleus, (2) basal nucleus, (3) accessory basal nucleus, (4) central nucleus, and (5) remaining nuclei (Figure 11.6). Neurons were counted in the total amygdala and each of the five subdivisions using the optical fractionator technique (West et al., 1991). The major finding was that the total autistic amygdala and the lateral nucleus have significantly fewer neurons than controls. This finding was surprising in that it appears to be in conflict with what one might expect from previous MRI and postmortem studies of autism. We did not find increased neuronal density as Bauman and Kemper (1985, 1994) had previously reported. What might account for the lower number of neurons in the autistic amygdala? Two possible hypotheses have emerged: (1) fewer neurons were generated during early development or (2) a normal or even excessive number of neurons was generated initially, but some of these have subsequently degenerated during adulthood. Unfortunately, there is currently no evidence to support or reject either of these possibilities. The early increased size of the autistic amygdala reported in MRI studies (Sparks et al., 2002; Schumann et al., 2004) cannot be used as evidence of increased neuronal proliferation because the size difference could be accounted for by other pathological changes, such as an increase in the number of fibers, rather than differences in neuron number. The resolution of this issue would require similar

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

243

postmortem stereological studies of the amygdala in younger autistic subjects. It is interesting to note, however, that studies carried out with young autistic children indicate that the amygdala is larger than normal (Sparks et al., 2002; Schumann et al., 2004; Mosconi et al., 2005) whereas studies carried out in adults suggest a smaller amygdala (Aylward et al., 1999; Pierce et al., 2001). This raises the prospect that the amygdala has a normal or perhaps even an increased number of neurons in early postnatal life and that a regressive process takes place at some later time. Amygdala Dysfunction May Contribute to Symptoms of Autism Baron-Cohen et al. (2000) proposed that pathology of the amygdala is responsible for behavioral impairments seen in individuals with autism. This theory was based on the assumption that the amygdala was responsible for mediating social behavior (Brothers, 1990), the primary impairment in autism. However, as discussed earlier, recent evidence indicates that we may need to rethink the role of the amygdala in social behavior. We have suggested that the amygdala plays a modulatory role in social behavior, but is not essential as a substrate for producing a normal repertoire of social behavior. Converging neurobiological evidence from both human and animal models indicates that the amygdala plays an essential role in regulating fear behaviors, which may in turn alter social processing (Adolphs, 2003a). The well-established role of the amygdala in regulating fear responses suggests that dysfunction of the amygdala may contribute to abnormal fear processing in humans (Amaral et al., 2003). The amygdala may be dysregulated in a number of emotional disorders such as anxiety (Davidson and Slagter, 2000; Fredrikson and Furmark, 2003). Using functional imaging, Thomas et al. (2001) reported that anxious children show heightened amygdala activity in response to fearful faces compared with typically developing children. Structural abnormalities of the amygdala are also linked with anxiety in children. De Bellis et al. (2000) found that the right amygdala of children with generalized anxiety disorder was larger than age-matched controls, whereas Milham et al. (2005) recently reported a reduction in left amygdala gray matter volume for children with anxiety disorders relative to comparison subjects. In Leo Kanner’s (1943) original description of autism, he noted unusual fear or anxiety in several of his young patients. One child, Herbert, was “tremendously frightened by running water, gas burners, and many other things.” He became upset by any change of an accustomed pattern. “If he notices change, he is very fussy and cries.” Another child did a “good deal of worrying.” He was upset because the moon did not always appear in the sky at night. He preferred to play alone and would get down from a play apparatus as soon as another child approached. Insistence on sameness leads children with autism to become greatly distressed when anything is broken or incomplete. A great part of the day is spent demanding consistency in the sequence of events. Kanner notes that although many individuals with autism learn to tolerate changes in routine and interactions with other people in their environment as adults, these interruptions cause a great deal of anxiety in young children with autism. Social interactions with other people are an unwelcome intrusion to the child with autism. When social interaction is forced upon the child, Kanner observed that the child, with a great deal of anxiety, will either ignore the person attempting to interact

© 2006 by Taylor & Francis Group, LLC

244

Understanding Autism: From Basic Neuroscience to Treatment

or quickly answer to end the intrusion. This aspect of autism, although consistently described by parents (Wing, 1976) and included as a feature in the DSM-IV (APA, 1994), has not been extensively studied (Muris et al., 1998; Gillott et al., 2001). Though the behavioral manifestation of amygdala pathology in autism is not yet understood, it is plausible that pathology of the amygdala may alter the ability to correctly regulate fear and anxiety responses. Indeed, anxiety appears to be a common, though understated, feature of autism spectrum disorders. Muris and colleagues (1998) examined the presence of co-occurring anxiety symptoms in 44 children diagnosed with autism or pervasive developmental disorder (PDD). Using parental report, they found that 84% of the children met the criteria for at least one anxiety disorder. Gillott et al. (2001) compared high-functioning children with autism to two control groups including children with specific language impairment and normally developing children on measures of anxiety and social worry. Children with autism were found to be significantly more anxious on both indices. Similarly, Kim et al. (2000) evaluated the prevalence of anxiety and mood problems in 59 children with high-functioning autism or Asperger syndrome, reporting that children with autism and Asperger syndrome were at a greater risk for mood and anxiety problems than the general population. Moreover, a recent study reported that 62% of children participating in a mood and anxiety disorders research clinic screened positive for a possible diagnosis of an autism spectrum disorder (Towbin et al., 2005). In addition to abnormal anxiety levels, individuals with autism show a disordered pattern of neural responses to emotional stimuli in faces depicting fear (Dawson et al., 2004). Typically developing children exhibit a larger early negative component (N300) and later negative slow wave of the event-related potential (ERP) to a face depicting fear than to the neutral face. In contrast, children with autism do not show the difference in amplitude. In children with autism, faster speed of early processing (i.e., N300 latency) of the fear face is associated with better performance on tasks assessing social attention (social orienting, joint attention, and attention to distress). In addition, persons with autism who have an enlarged amygdala relative to controls also have selective impairment in the recognition of facial expressions of fear and perception of eye gaze (Howard et al., 2000), similar to the impairments found in individuals with an amygdala lesion (Adolphs et al., 1994; 1995). Thus, based on our current knowledge of amygdala function and pathology, it is plausible that abnormal amygdala development contributes to abnormal fear and anxiety processing in children with autism, which may in turn contribute or exacerbate the hallmark feature of social avoidance. Abnormal Amygdala Function May Impact Social Processing There is substantial evidence that individuals with autism are impaired in their ability to process social information from faces (Grelotti et al., 2002). Early inattention to faces could have profound implications for later social development. Autism subjects show abnormal visual scan paths during eye-tracking studies when viewing faces, typically spending little time on core social features such as the eyes (Klin et al., 2002; Pelphrey et al., 2002). It is unclear as to whether these findings represent active avoidance of the eye region, potentially involving the amygdala, or that failure

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

245

to look at the eyes represents a more global lack of social interest or motivation. An emerging hypothesis is that the amygdala may play a role in mediating or directing visual attention to the eyes (Grelotti et al., 2002; Adolphs et al., 2005; Schultz, 2005). Pierce et al. (2004) found that when autistic subjects viewed familiar faces, they were able to activate the amygdala appropriately in response to familiar and unfamiliar faces, suggesting that the familiar faces may have enhanced motivation or attention to all of the stimuli. Although inattention to faces, particularly the eye region, is one of the earliest and most consistent symptoms of autism, little is known regarding the underlying causes of this abnormal pattern of social attention. One possibility is that children with autism simply lack social motivation and thus lack interest in attending to the face. An alternative view is that individuals with autism perceive social interactions as threatening and therefore avoid the interaction as a means of alleviating anxiety. Indeed, research from typically developing children indicates that children who are physiologically aroused by a distressing film were more likely to avert their gaze from the stimulus (Fabes et al., 1993). It is plausible that children with autism utilize a similar strategy of gaze aversion in response to arousing social stimuli. Given the amygdala’s role in fear and anxiety, one would predict heightened amygdala activation during eye contact in persons with autism if they found the eye contact aversive. Dalton and colleagues (2005) recently carried out a series of studies utilizing functional imaging and eye-tracking technology simultaneously while showing subjects familiar and unfamiliar faces. They found the amount of time persons with autism spent looking at the eye region of the face was strongly positively correlated with amygdala activation, but not in typically developing control subjects. The autistic subjects also showed greater left amygdala activation relative to controls in response to unfamiliar faces and greater right amygdala activation in response to both familiar and unfamiliar faces. This suggests a heightened emotional, or even fearful, response when autistic individuals look at another person’s eyes, regardless of whether they are familiar or a stranger. Additional studies would benefit from measuring the physiological responses associated with arousal and anxiety (i.e., increased heart rate, skin response, etc.) during face processing in individuals with autism.

CONCLUSIONS Understanding the organization of neural systems that underlie normal social behavior provides important clues as to which brain regions may be pathological in autism. Conversely, understanding the pathology commonly associated with autism may provide insights into brain regions putatively involved in normal social processing. We have outlined a series translational neuroscience studies in which we have focused on the role of the amygdala in autism. Although we have obtained both MRI and postmortem histological evidence that the amygdala is pathological in autism, our primate studies indicate that damage to the amygdala may not be responsible for the core deficit of social impairment. Rather, we propose that altered function of the amygdala in autism may lead to abnormal fear or anxiety, which is a common comorbid characteristic of autism. Increased anxiety, particularly,

© 2006 by Taylor & Francis Group, LLC

246

Understanding Autism: From Basic Neuroscience to Treatment

increased social anxiety, may exacerbate social withdrawal. These conclusions are, of course, preliminary and there are alternative hypotheses that could accommodate the neuropathological observations. For example, lesion studies remove a brain region’s influence. But, if the same brain region were hyperactive, this might have a quite different behavioral outcome. It is not clear whether a hyperactive amygdala might have a more specific impact on social function. However, the combination of detailed human studies with appropriate animal models is a powerful approach not only to understanding the normal functioning of the brain but to unravel the processes that underlie disorders such as autism.

REFERENCES Adolphs, R. (2001), The neurobiology of social cognition, Curr Opin Neurobiol 11: 231–239. Adolphs, R. (2003a), Cognitive neuroscience of human social behavior, Nat Rev Neurosci 4: 165–178. Adolphs, R. (2003b), Investigating the cognitive neuroscience of social behavior, Neuropsychologia 41: 119–126. Adolphs, R., Tranel, D., and Damasio, A.R. (1998), The human amygdala in social judgment, Nature 393: 470–474. Adolphs, R., Sears, L., and Piven, J. (2001), Abnormal processing of social information from faces in autism, J Cognit Neurosci 13: 232–240. Adolphs, R., Tranel, D., Damasio, H., and Damasio, A. (1994), Impaired recognition of emotion in facial expressions following bilateral damage to the human amygdala, Nature 372: 669–672. Adolphs, R., Tranel, D., Damasio, H., and Damasio, A.R. (1995), Fear and the human amygdala, J Neurosci 15: 5879–5891. Adolphs, R., Gosselin, F., Buchanan, T.W., Tranel, D., Schyns, P., and Damasio, A.R. (2005), A mechanism for impaired fear recognition after amygdala damage, Nature 433: 68–72. Amaral, D.G., Bauman, M.D., Capitanio, J.P., Lavenex, P., Mason, W.A., Mauldin-Jourdain, M.L., and Mendoza, S.P. (2003), The amygdala: is it an essential component of the neural network for social cognition? Neuropsychologia 41: 517–522. Amorapanth, P., LeDoux, J.E., and Nader, K. (2000), Different lateral amygdala outputs mediate reactions and actions elicited by a fear-arousing stimulus, Nat Neurosci 3: 74–79. APA (1994), Diagnostic and Statistical Manual of Mental Disorders, 4th ed., Washington, D.C.: American Psychiatric Association. Aylward, E.H., Minshew, N.J., Goldstein, G., Honeycutt, N.A., Augustine, A.M., Yates, K.O., Barta, P.E., and Pearlson, G.D. (1999), MRI volumes of amygdala and hippocampus in non-mentally retarded autistic adolescents and adults, Neurology 53: 2145–2150. Bachevalier, J. (1994), Medial temporal lobe structures and autism — a review of clinical and experimental findings, Neuropsychologia 32: 627–648. Bachevalier, J. (2000), The amygdala, social cognition, and autism, in The Amygdala: A Functional Analysis, Aggleton, J.P., Ed., New York: Oxford University Press, pp. 509–543. Bachevalier, J. and Meunier, M. (2005), The neurobiology of social-emotional cognition in nonhuman primates, in The Cognitive Neuroscience of Social Behaviour, Easton, A. and Emery, N.J., Eds., London: Psychology Press, pp. 19–57.

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

247

Baron-Cohen, S., Jolliffe, T., Mortimore, C., and Robertson, M. (1997), Another advanced test of theory of mind: evidence from very high functioning adults with autism or Asperger syndrome, J Child Psychol Psychiatry 38: 813–822. Baron-Cohen, S., Ring, H.A., Bullmore, E.T., Wheelwright, S., Ashwin, C., and Williams, S.C. (2000), The amygdala theory of autism, Neurosci Biobehav Rev 24: 355–364. Baron-Cohen, S., Ring, H.A., Wheelwright, S., Bullmore, E.T., Brammer, M.J., Simmons, A., and Williams, S.C. (1999), Social intelligence in the normal and autistic brain: an fMRI study, Eur J Neurosci 11: 1891–1898. Bastian, M.L., Sponberg, A.C., Suomi, S.J., and Higley, J.D. (2003), Long-term effects of infant rearing condition on the acquisition of dominance rank in juvenile and adult rhesus macaques (Macaca mulatta), Dev Psychobiol 42: 44–51. Bauman, M. and Kemper, T.L. (1985), Histoanatomic observations of the brain in early infantile autism, Neurology 35: 866–874. Bauman, M. and Kemper, T.L. (1994), Neuroanatomic observations of the brain in autism, in The Neurobiology of Autism, Bauman, M. and Kemper, T.L., Eds., Baltimore, MD: Johns Hopkins University Press, pp. 119–145. Bauman, M.D., Lavenex, P., Mason, W.A., Capitanio, J.P., and Amaral, D.G. (2004a), The development of social behavior following neonatal amygdala lesions in rhesus monkeys, J Cognit Neurosci 16: 1388–1411. Bauman, M.D., Lavenex, P., Mason, W.A., Capitanio, J.P., and Amaral, D.G. (2004b), The development of mother-infant interactions after neonatal amygdala lesions in rhesus monkeys, J Neurosci 24: 711–721. Bechara, A., Damasio, H., and Damasio, A.R. (2000), Emotion, decision making and the orbitofrontal cortex, Cereb Cortex 10: 295–307. Bechara, A., Tranel, D., Damasio, H., Adolphs, R., Rockland, C., and Damasio, A.R. (1995), Double dissociation of conditioning and declarative knowledge relative to the amygdala and hippocampus in humans, Science 269: 1115–1118. Berman, C.I. (1980), Mother-infant relationships among free-ranging rhesus monkeys on Cayo Santiago: a comparison with captive pairs, Animal Behav 28: 860–873. Blair, R.J., Morris, J.S., Frith, C.D., Perrett, D.I., and Dolan, R.J. (1999), Dissociable neural responses to facial expressions of sadness and anger, Brain 122 (Pt. 5): 883–893. Blakemore, S.J., Winston, J., and Frith, U. (2004), Social cognitive neuroscience: where are we heading? Trends Cognit Sci 8: 216–222. Breiter, H.C., Etcoff, N.L., Whalen, P.J., Kennedy, W.A., Rauch, S.L., Buckner, R.L., Strauss, M.M., Hyman, S.E., Rosen, B.R. (1996), Response and habituation of the human amygdala during visual processing of facial expression, Neuron 17: 875–887. Brothers, L. (1990), The social brain: a project for integrating primate behavior and neurophysiology in a new domain, Concepts Neurosci 1: 27–51. Brothers, L. and Ring, B. (1993), Mesial temporal neurons in the macaque monkey with responses selective for aspects of social stimuli, Behav Brain Res 57: 53–61. Brothers, L., Ring, B., and Kling, A. (1990), Response of neurons in the macaque amygdala to complex social stimuli, Behav Brain Res 41: 199–213. Canli, T., Sivers, H., Whitfield, S.L., Gotlib, I.H., and Gabrieli, J.D. (2002), Amygdala response to happy faces as a function of extraversion, Science 296: 2191. Capitanio, J. (1986), Behavioral pathology, in Comparative Primate Biology: Behavior, Conservation, and Ecology, Mitchell, G. and Erwin, J., Eds., New York: Alan R. Liss, pp. 411–454. Critchley, H.D., Daly, E.M., Bullmore, E.T., Williams, S.C., Van Amelsvoort, T., Robertson, D.M., Rowe, A., Phillips, M., McAlonan, G., Howlin, P., and Murphy, D.G. (2000), The functional neuroanatomy of social behavior: changes in cerebral blood flow when

© 2006 by Taylor & Francis Group, LLC

248

Understanding Autism: From Basic Neuroscience to Treatment

people with autistic disorder process facial expressions, Brain 123 (Pt. 11): 2203– 2212. Dalton, K.M., Nacewicz, B.M., Johnstone, T., Schaefer, H.S., Gernsbacher, M.A., Goldsmith, H.H., Alexander, A.L., and Davidson, R.J. (2005), Gaze fixation and the neural circuitry of face processing in autism, Nat Neurosci 8: 519–526. Damasio, A.R. (1994), Decartes’ Error, New York: Avon Books. Davidson, R.J. and Slagter, H.A. (2000), Probing emotion in the developing brain: functional neuroimaging in the assessment of the neural substrates of emotion in normal and disordered children and adolescents, Ment Retard Dev Disabil Res Rev 6: 166–170. Davis, M. (1992), The role of the amygdala in fear and anxiety, Annu Rev Neurosci 15: 353–375. Dawson, G., Webb, S.J., Carver, L., Panagiotides, H., and McPartland, J. (2004), Young children with autism show atypical brain responses to fearful versus neutral facial expressions of emotion, Dev Sci 7: 340–359. De Bellis, M.D., Casey, B.J., Dahl, R.E., Birmaher, B., Williamson, D.E., Thomas, K.M., Axelson, D.A., Frustaci, K., Boring, A.M., Hall, J., and Ryan, N.D. (2000), A pilot study of amygdala volumes in pediatric generalized anxiety disorder, Biol Psychiatry 48: 51–57. Dicks, D., Myers, R.E., and Kling, A. (1968), Uncus and amygdala lesions: effects on social behavior in the free-ranging rhesus monkey, Science 165: 69–71. Dilger, S., Straube, T., Mentzel, H.J., Fitzek, C., Reichenbach, J.R., Hecht, H., Krieschel, S., Gutberlet, I., and Miltner, W.H. (2003), Brain activation to phobia-related pictures in spider phobic humans: an event-related functional magnetic resonance imaging study, Neurosci Lett 348: 29–32. Dunbar, R.I.M. (2002), The social brain hypothesis, in Foundations in Social Neuroscience, Cacioppo, J.T., Bernston, G.G., Adolphs, R., Carter, C.S., Davidson, R.J., McClintock, M.K., McEwen, B.S., Meaney, M.J., Schacter, D.L., Sternberg, E.M., Suomi, S.J., and Taylor, S.E., Eds., Cambridge, MA: The MIT Press, pp. 69–87. Emery, N.J., Capitanio, J.P., Mason, W.A., Machado, C.J., Mendoza, S.P., and Amaral, D.G. (2001), The effects of bilateral lesions of the amygdala on dyadic social interactions in rhesus monkeys (Macaca mulatta), Behav Neurosci 115: 515–544. Fabes, R.A., Eisenberg, N., and Eisenbud, L. (1993), Behavioral and physiological correlates of children's reactions to others in distress, Dev Psychol 29: 655–663. Fredrikson, M. and Furmark, T. (2003), Amygdaloid regional cerebral blood flow and subjective fear during symptom provocation in anxiety disorders, Ann N Y Acad Sci 985: 341–347. Fried, I., Cameron, K.A., Yashar, S., Fong, R., and Morrow, J.W. (2002), Inhibitory and excitatory responses of single neurons in the human medial temporal lobe during recognition of faces and objects, Cereb Cortex 12: 575–584. Giedd, J.N. (1997), Normal development, in Child and Adolescent Psychiatric Clinics of North America Neuroimaging, Peterson, B.S., Ed., Philadelphia, PA: W.B. Saunders Company, pp. 265–282. Giedd, J.N., Vaituzis, A.C., Hamburger, S.D., Lange, N., Rajapakse, J.C., Kaysen, D., Vauss, Y.C., and Rapoport, J.L. (1996), Quantitative MRI of the temporal lobe, amygdala, and hippocampus in normal human development: ages 4–18 years, J Comp Neurol 366: 223–230. Gillott, A., Furniss, F., and Walter, A. (2001), Anxiety in high-functioning children with autism, Autism 5: 277–286.

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

249

Grelotti, D.J., Gauthier, I., and Schultz, R.T. (2002), Social interest and the development of cortical face specialization: what autism teaches us about face processing, Dev Psychobiol 40: 213–225. Griffin, G.A. and Harlow, H.F. (1966), Effects of three months of total social deprivation on social adjustment and learning in the rhesus monkey, Child Dev 37: 533–547. Hariri, A.R., Mattay, V.S., Tessitore, A., Fera, F., and Weinberger, D.R. (2003), Neocortical modulation of the amygdala response to fearful stimuli, Biol Psychiatry 53: 494–501. Harlow, H.F., Dodsworth, R.O., and Harlow, M.K. (1965), Total social isolation in monkeys, Proc Natl Acad Sci U S A 54: 90–97. Haxby, J.V., Hoffman, E.A., and Gobbini, M.I. (2000), The distributed human neural system for face perception, Trends Cognit Sci 4: 223–233. Haxby, J.V., Hoffman, E.A., and Gobbini, M.I. (2002), Human neural systems for face recognition and social communication, Biol Psychiatry 51: 59–67. Haznedar, M.M., Buchsbaum, M.S., Wei, T.C., Hof, P.R., Cartwright, C., Bienstock, C.A., and Hollander, E. (2000), Limbic circuitry in patients with autism spectrum disorders studied with positron emission tomography and magnetic resonance imaging, Am J Psychiatry 157: 1994–2001. Heberlein, A.S. and Adolphs, R. (2004) Impaired spontaneous anthropomorphizing despite intact perception and social knowledge, Proc Natl Acad Sci U S A 101: 7487–7491. Heider, F. and Simmel, M. (1944), An experimental study of apparent behavior, Am J Psychol 57: 243–259. Howard, M.A., Cowell, P.E., Boucher, J., Broks, P., Mayes, A., Farrant, A., and Roberts, N. (2000), Convergent neuroanatomical and behavioral evidence of an amygdala hypothesis of autism, Neuroreport 11: 2931–2935. Insel, T.R. and Fernald, R.D. (2004), How the brain processes social information: searching for the social brain, Annu Rev Neurosci 27: 697–722. Izquierdo, A. and Murray, E.A. (2004), Combined unilateral lesions of the amygdala and orbital prefrontal cortex impair affective processing in rhesus monkeys, J Neurophysiol 91: 2023–2039. Kalin, N.H., Shelton, S.E., and Davidson, R.J. (2004), The role of the central nucleus of the amygdala in mediating fear and anxiety in the primate, J Neurosci 24: 5506– 5515. Kalin, N.H., Shelton, S.E., Davidson, R.J., and Kelley, A.E. (2001), The primate amygdala mediates acute fear but not the behavioral and physiological components of anxious temperament, J Neurosci 21: 2067–2074. Kanner, L. (1943), Autistic disturbances of affective contact, in Nervous Child, pp. 217–250. Kanwisher, N., McDermott, J., and Chun, M.M. (1997), The fusiform face area: a module in human extrastriate cortex specialized for face perception, J Neurosci 17: 4302–4311. Kemper, T.L. and Bauman, M.L. (1993), The contribution of neuropathologic studies to the understanding of autism, Neurol Clin 11: 175–187. Killcross, S., Robbins, T.W., and Everitt, B.J. (1997), Different types of fear-conditioned behavior mediated by separate nuclei within amygdala, Nature 388: 377–380. Kim, J.A., Szatmari, P., Bryson, S.E., Streiner, D.L., and Wilson, F.J. (2000), The prevalence of anxiety and mood problems among children with autism and Asperger syndrome, Autism 4: 117–132. Klin, A. (2000), Attributing social meaning to ambiguous visual stimuli in higher-functioning autism and Asperger syndrome: the social attribution task, J Child Psychol Psychiatry 41: 831–846.

© 2006 by Taylor & Francis Group, LLC

250

Understanding Autism: From Basic Neuroscience to Treatment

Klin, A., Jones, W., Schultz, R., Volkmar, F., and Cohen, D. (2002), Visual fixation patterns during viewing of naturalistic social situations as predictors of social competence in individuals with autism, Arch Gen Psychiatry 59: 809–816. Kling, A. (1968), Effects of amygdalectomy and testosterone on sexual behavior of male juvenile macaques, J Comp Physiol Psychol 65: 466–471. Kling, A. and Cornell, R. (1971), Amygdalectomy and social behavior in the caged stumptailed macaque (Macaca speciosa), Folia Primatol (Basel) 14: 190–208. Kling, A., Lancaster, J., and Benitone, J. (1970), Amygdalectomy in the free-ranging vervet (Cercopithecus aethiops), J Psychiatr Res 7: 191–199. Kling, A., Steklis, H.D., and Deutsch, S. (1979), Radiotelemetered activity from the amygdala during social interactions in the monkey, Exp Neurol 66: 88–96. Kling, A.S. and Green, P.C. (1967), Effects of neonatal amygdalectomy in the maternally reared and maternally deprived macaque, Nature 213: 742–743. Kling, A.S. and Brothers, L. (1992), The amygdala and social behavior, in The Amygdala: Neurobiological aspects of Emotion, Memory, and Dysfunction, Aggleton, J.P., Ed., New York: Wiley-Liss. LaBar, K.S., LeDoux, J.E., Spencer, D.D., and Phelps, E.A. (1995), Impaired fear conditioning following unilateral temporal lobectomy in humans, J Neurosci 15: 6846–6855. LeDoux, J.E. (2000), Emotion circuits in the brain, Annu Rev Neurosci 23: 155–184. LeDoux, J.E., Cicchetti, P., Xagoraris, A., and Romanski, L.M. (1990), The lateral amygdaloid nucleus: sensory interface of the amygdala in fear conditioning, J Neurosci 10: 1062–1069. Leonard, C.M., Rolls, E.T., Wilson, F.A., and Baylis, G.C. (1985), Neurons in the amygdala of the monkey with responses selective for faces, Behav Brain Res 15: 159–176. Lieberman, M.D. (2005), Principles, processes, and puzzles of social cognition: an introduction for the special issue on social cognitive neuroscience, Neuroimage 28(4): 745–756. Malkova, L., Barrow, K.V., Lower, L.L., and Gales, K. (2003), Decreased social interactions in monkeys after unilateral blockade of GABAA receptors in the basolateral amygdala, Ann N Y Acad Sci 985: 540–541. Mason, W.A. (1960), The effects of social restriction on the behavior of rhesus monkeys: I. Free social behavior, J Comp Physiol Psychol 53: 582–589. Mason, W.A. and Green, P.C. (1962), The effects of social restriction on the behavior of rhesus monkeys: IV. Responses to a novel environment and to an alien species, J Comp Physiol Psychol 55: 363–368. Mason, W.A. and Sponholz, R.R. (1963), Behavior of rhesus monkeys raised in isolation, J Psychiatric Res 1: 299–306. Mason, W.A., Capitanio, J.P., Machado, C.J., Mendoza, S.P., and Amaral, D.G. (in press), Amygdalaectomy and responsiveness to novelty in rhesus monkeys: generality and individual consistency of effects, Emotion. Meunier, M., Bachevalier, J., Murray, E.A., Malkova, L., and Mishkin, M. (1999), Effects of aspiration versus neurotoxic lesions of the amygdala on emotional responses in monkeys, Eur J Neurosci 11: 4403–4418. Milham, M.P., Nugent, A.C., Drevets, W.C., Dickstein, D.P., Leibenluft, E., Ernst, M., Charney, D., and Pine, D.S. (2005), Selective reduction in amygdala volume in pediatric anxiety disorders: a voxel-based morphometry investigation, Biol Psychiatry 57: 961–966. Morris, J.S., Friston, K.J., Buchel, C., Frith, C.D., Young, A.W., Calder, A.J., and Dolan, R.J. (1998), A neuromodulatory role for the human amygdala in processing emotional facial expressions, Brain 121 (Pt. 1): 47–57.

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

251

Morris, J.S., Frith, C.D., Perrett, D.I., Rowland, D., Young, A.W., Calder, A.J., and Dolan, R.J. (1996), A differential neural response in the human amygdala to fearful and happy facial expressions, Nature 383: 812–815. Morris, J.S., Ohman, A., and Dolan, R.J. (1998), Conscious and unconscious emotional learning in the human amygdala, Nature 393: 467–470. Mosconi, M., Cody, H., Poe, M., Joshi, S., Peterson, S., and Piven, J. (2005), Amygdala and hippocampal enlargement in young children with autism, in IMFAR, Boston, MA. Muris, P., Steerneman, P., Merckelbach, H., Holdrinet, I., and Meesters, C. (1998), Comorbid anxiety symptoms in children with pervasive developmental disorders, J Anxiety Disord 12: 387–393. Nakamura, K., Mikami, A., and Kubota K (1992), Activity of single neurons in the monkey amygdala during performance of a visual discrimination task, J Neurophysiol 67: 1447–1463. Ochsner, K.N. (2004), Current directions in social cognitive neuroscience, Curr Opin Neurobiol 14: 254–258. Pare, D., Quirk, G.J., and Ledoux, J.E. (2004), New vistas on amygdala networks in conditioned fear, J Neurophysiol 92: 1–9. Pelphrey, K.A., Sasson, N.J., Reznick, J.S., Paul, G., Goldman, B.D., and Piven, J. (2002), Visual scanning of faces in autism, J Autism Dev Disord 32: 249–261. Perrett, D.I., Hietanen, J.K., Oram, M.W., and Benson, P.J. (1992), Organization and functions of cells responsive to faces in the temporal cortex, Philos Trans R Soc Lond B Biol Sci 335: 23–30. Perrett, D.I., Mistlin, A.J., Chitty, A.J., Smith, P.A., Potter, D.D., Broennimann, R., and Harries, M. (1988), Specialized face processing and hemispheric asymmetry in man and monkey: evidence from single unit and reaction time studies, Behav Brain Res 29: 245–258. Phelps, E.A., O’Connor, K.J., Gatenby, J.C., Gore, J.C., Grillon, C., and Davis, M. (2001), Activation of the left amygdala to a cognitive representation of fear, Nat Neurosci 4: 437–441. Phillips, M.L., Young, A.W., Senior, C., Brammer, M., Andrew, C., Calder, A.J., Bullmore, E.T., Perrett, D.I., Rowland, D., Williams, S.C., Gray, J.A., and David, A.S. (1997), A specific neural substrate for perceiving facial expressions of disgust, Nature 389: 495–498. Pierce, K., Haist, F., Sedaghat, F., and Courchesne, E. (2004), The brain response to personally familiar faces in autism: findings of fusiform activity and beyond, Brain 127: 2703–2716. Pierce, K., Muller, R.A., Ambrose, J., Allen, G., and Courchesne, E. (2001), Face processing occurs outside the fusiform face area in autism: evidence from functional MRI, Brain 124: 2059–2073. Pitkanen, A., Tuunanen, J., Kalviainen, R., Partanen, K., and Salmenpera, T. (1998), Amygdala damage in experimental and human temporal lobe epilepsy, Epilepsy Res 32: 233–253. Prather, M.D., Lavenex, P., Mauldin-Jourdain, M.L., Mason, W.A., Capitanio, J.P., Mendoza, S.P., and Amaral, D.G. (2001), Increased social fear and decreased fear of objects in monkeys with neonatal amygdala lesions, Neuroscience 106: 653–658. Rolls, E.T. (1984), Neurons in the cortex of the temporal lobe and in the amygdala of the monkey with responses selective for faces, Hum Neurobiol 3: 209–222. Rolls, E.T. (1999), The Brain and Emotion, Oxford: Oxford University Press.

© 2006 by Taylor & Francis Group, LLC

252

Understanding Autism: From Basic Neuroscience to Treatment

Rolls, E.T. (2002), The functions of the orbitofrontal cortex, in Principles of Frontal Lobe Function, Stuss, D.T. and Knight, R.T., Eds., Oxford: Oxford University Press, pp. 354–374. Rosvold, H.E., Mirsky, A.F., and Pribram, K.H. (1954), Influence of amygdalectomy on social behavior in monkeys, J Comp Physiol Psychol 47: 173–178. Schultz, R.T. (2005), Developmental deficits in social perception in autism: the role of the amygdala and fusiform face area, Int J Dev Neurosci 23: 125–141. Schultz, R.T., Grelotti, D.J., Klin, A., Kleinman, J., Van der Gaag, C., Marois, R., and Skudlarski, P. (2003), The role of the fusiform face area in social cognition: implications for the pathobiology of autism, Philos Trans R Soc Lond B Biol Sci 358: 415–427. Schultz, W. (2002), Getting formal with dopamine and reward, Neuron 36: 241–263. Schumann, C.M. and Amaral, D.G., Stereological evidence for fewer neurons in the autistic amygdala, submitted. Schumann, C.M. and Amaral, D.G. (2005), Stereological estimation of the number of neurons in the human amygdaloid complex, J Comp Neurol 491(4): 320–329. Schumann, C.M., Hamstra, J., Goodlin-Jones, B.L., Lotspeich, L.J., Kwon, H., Buonocore, M.H., Lammers, C.R., Reiss, A.L., and Amaral, D.G. (2004), The amygdala is enlarged in children but not adolescents with autism; the hippocampus is enlarged at all ages, J Neurosci 24: 6392–6401. Shannon, C., Champoux, M., and Suomi, S.J. (1998), Rearing condition and plasma cortisol in rhesus monkey infants, Am J Primatology 46: 311–321. Sparks, B.F., Friedman, S.D., Shaw, D.W., Aylward, E.H., Echelard, D., Artru, A.A., Maravilla, K.R., Giedd, J.N., Munson, J., Dawson, G., and Dager, S.R. (2002), Brain structural abnormalities in young children with autism spectrum disorder, Neurology 59: 184–192. Steklis, H. and Kling, A.S. (1985), Neurobiology of affiliative behavior in nonhuman primates, in The Psychobiology of Attachment and Separation, Reite, M. and Field, T., Eds., Orlando, FL: Academic Press, pp. 93–134. Thomas, K.M., Drevets, W.C., Dahl, R.E., Ryan, N.D., Birmaher, B., Eccard, C.H., Axelson, D., Whalen, P.J., and Casey, B.J. (2001), Amygdala response to fearful faces in anxious and depressed children, Arch Gen Psychiatry 58: 1057–1063. Thompson, C.I. (1981), Long-term behavioral development of rhesus monkeys after amygdalectomy in infancy, in The Amygdaloid Complex, Ben-Ari, Y., Ed., Amsterdam: Elsevier, pp. 259–270. Thompson, C.I. and Towfighi, J.T. (1976), Social behavior of juvenile rhesus monkeys after amygdalectomy in infancy, Physiol Behav 17: 831–836. Thompson, C.I., Schwartzbaum, J.S., and Harlow, H.F. (1969), Development of social fear after amygdalectomy in infant rhesus monkeys, Physiol Behav 4: 249–254. Thompson, C.I., Bergland, R.M., and Towfighi, J.T. (1977), Social and nonsocial behaviors of adult rhesus monkeys after amygdalectomy in infancy or adulthood, J Comp Physiol Psychol 91: 533–548. Towbin, K.E., Pradella, A., Gorrindo, T., Pine, D.S., and Leibenluft, E. (2005), Autism spectrum traits in children with mood and anxiety disorders, J Child Adolesc Psychopharmacol 15: 452–464. Wang, A.T., Dapretto, M., Hariri, A.R., Sigman, M., and Bookheimer, S.Y. (2004), Neural correlates of facial affect processing in children and adolescents with autism spectrum disorder, J Am Acad Child Adolesc Psychiatry 43: 481–490.

© 2006 by Taylor & Francis Group, LLC

The Social Brain, Amygdala, and Autism

253

West, M.J., Slomianka, L., and Gundersen, H.J. (1991), Unbiased stereological estimation of the total number of neurons in the subdivisions of the rat hippocampus using the optical fractionator, Anat Rec 231: 482–497. Whalen, P.J., Rauch, S.L., Etcoff, N.L., McInerney, S.C., Lee, M.B., and Jenike, M.A. (1998), Masked presentations of emotional facial expressions modulate amygdala activity without explicit knowledge, J Neurosci 18: 411–418. Wilensky, A.E., Schafe, G.E., and LeDoux, J.E. (1999), Functional inactivation of the amygdala before but not after auditory fear conditioning prevents memory formation, J Neurosci 19: RC48. Wing, L. (1976), Diagnosis, clinical description and prognosis, in Early Childhood Autism, Oxford: Pergamon Press, pp. 15–48. Winslow, J.T., Noble, P.L., Lyons, C.K., Sterk, S.M., and Insel, T.R. (2003), Rearing effects on cerebrospinal fluid oxytocin concentration and social buffering in rhesus monkeys, Neuropsychopharmacology 28: 910–918. Winston, J.S., Strange, B.A., O’Doherty, J., and Dolan, R.J. (2002), Automatic and intentional brain responses during evaluation of trustworthiness of faces, Nat Neurosci 5: 277–283.

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 255 Wednesday, February 15, 2006 3:24 PM

Thalamus and 12 The Neuromodulatory Systems Mircea Steriade CONTENTS Autism and the Thalamus......................................................................................255 Intrathalamic and Corticothalamic Neuronal Circuitry ........................................256 Thalamic Projections of Brain Stem and Forebrain Neuromodulatory Systems ....................................................................................259 Thalamic Projections of Glutamatergic and Cholinergic Brain Stem Reticular Neurons .......................................................................259 Thalamic Projections of Brain Stem and Hypothalamic Monoaminergic Systems ................................................................................262 Basal Forebrain Projections to the Thalamus ................................................262 Brain Stem–Thalamic Neurons during Tonic and Phasic Activation Processes...................................................................................263 Modulatory Actions on Thalamocortical and Thalamic Inhibitory Neurons ................................................................................269 References..............................................................................................................272

AUTISM AND THE THALAMUS The brain structures and neuronal mechanisms that underlie the expression of autistic disorders are not yet elucidated. Although some studies suggested that the cerebral defect in autism is only functional without major morphological substrate,1 and others found evidence of abnormalities localized to the cerebellum or cerebellofugal systems,2–3 there is also evidence of gross anatomical changes in the thalamus and functional dysfunction implicating attention and sensory gating through the thalamic anteroom to the cerebral cortex. Thus, the thalamic volume was found to be significantly different (mainly reduced) in autistic groups relative to normal control subjects,4–6 and emotion processing results in lower blood flow in autism not only in some cortical areas but also in the thalamus.7 These recent studies provide a possible substrate for an earlier hypothesis8 implicating the anterior and medial nuclear groups of the thalamus, together with their projection cortical areas that are direct targets of dopaminergic midbrain neurons, in the autism dysfunction, related

255

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 256 Wednesday, February 15, 2006 3:24 PM

256

Understanding Autism: From Basic Neuroscience to Treatment

to imbalance in these structures. Another factor that may be important in the pathogenesis of autism is serotonin, as decreased serotonin synthesis was found in the cerebellothalamic pathway.9 In this chapter, I will discuss the neuronal circuitry of the thalamus and different neuromodulatory systems arising in the brain stem core, basal forebrain, and posterior hypothalamus,10 which act on three major thalamic neuronal classes: glutamatergic thalamocortical (TC) neurons, GABAergic thalamic reticular (RE) neurons, and local-circuit GABAergic interneurons.11 As some previous data pointed to decreased thalamic volume and blood flow in the thalamus of autistic subjects,4,6,7 I shall mainly focus on systems that normally activate various thalamic neuronal types.

INTRATHALAMIC AND CORTICOTHALAMIC NEURONAL CIRCUITRY Thalamic nuclei can be systematized into sensory-motor (or relay), association, intralaminar, and RE neuronal aggregates. The term relay indicates that those nuclei (among them visual lateral geniculate, auditory medial geniculate, and somatosensory ventroposterior) transfer to the cerebral cortex specific sensory signals arising in ascending afferent pathways. This does not imply that such nuclei operate as merely relays, i.e., as if nothing would change between activities in afferent fibers and in axons of TC neurons. In fact, the presence of local-circuit inhibitory neurons in various nuclei and the relations that TC neurons have with thalamic RE inhibitory neurons account for the integrative processes in thalamic relay nuclei, mainly consisting of higher response selectivity than that recorded at prethalamic levels.11 Basically, the pathways are as follows: (1) ascending fibers from specific systems contact both TC and local-circuit neurons, (2) the axons of the local interneurons contact TC neurons and their presynaptic dendrites contact other interneurons, and (3) RE neurons project to most TC neurons as well as to local interneurons. Three of the four major actors in the intrathalamic and corticothalamic neuronal operations are depicted in Figure 12.1a. For the sake of simplicity, local-circuit interneurons are not illustrated in this panel, although gating processes in the thalamus largely depend on them (Figure 12.8 and Figure 12.9; see also the following text). Corticothalamic glutamatergic neurons are shown (Figure 12.1a) because this pathway outnumbers the TC one by almost one order of magnitude. TC neurons are bushy and their variations are linked to soma size: large neurons project to middle and deep neocortical layers, whereas small neurons project preferentially to superficial layers. The TC-cortico-TC projection (Figure 12.1a) is a typical example of a feedback excitatory neuronal loop. The axons of most TC neurons distribute with a trilaminar pattern, mainly to midlayers but also to layers 6 and 1.11 Some nuclei, such as the ventromedial and rostral intralaminar centrolateral, preferentially project to layer 1. Conjoined stimulation of one specific thalamic nucleus, such as the somatosensory ventroposterior nucleus, and intralaminar centrolateral nucleus

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 257 Wednesday, February 15, 2006 3:24 PM

The Thalamus and Neuromodulatory Systems

257

results in supralinear summation of the two inputs at cortical output layer 5, demonstrating coincidence detection along the apical dendrites of deep-lying pyramidal neurons.12 The inhibition exerted by RE neurons on TC neurons (Figure 12.1a) was demonstrated by: (1) abolition of the prolonged, biphasic GABAA-B inhibitory postsynaptic potentials (IPSPs) in TC neurons and their replacement by short-lasting IPSPs produced by local-circuit neurons after disconnection of TC neurons from the RE nucleus;13 (2) dual intracellular recordings in vivo, showing that spike bursts in RE neurons are followed by IPSPs in target TC neurons;14 and (3) GABA-mediated IPSPs in TC cells by stimulating RE neurons in thalamic slices maintained in vitro.15 However, the RE-to-TC inhibition is a conventional notion that does not take into consideration the connection from RE neurons to local-circuit GABAergic neurons,16 which may produce significant effects at the ultimate targets, the TC neurons. Indeed, a greatly increased incidence of IPSPs in TC neurons was observed following the excitotoxic lesion of RE perikarya, reflecting the release of local inhibitory interneurons from the inhibition exerted by RE neurons.13 There are no connections between TC neurons located in various dorsal thalamic nuclei, and these neurons may only communicate through intercalated relays in the cerebral cortex or RE nucleus. In contrast, RE neurons are linked through both chemical synapses and electrical coupling. The inhibition within the RE nuclear complex is because of dendrodendritic and axodendritic contacts among RE neurons,17,18 which produce IPSPs mainly through GABAA receptor activation.19 These synapses are effectively recruited by cortico-RE inputs, which explains the potency of corticothalamic stimulation in eliciting sleep spindles20 by acting on the intranuclear inhibitory network that generates spindle oscillations in pacemaking RE neurons.21 The electrical coupling of RE neurons is also implicated in intra-RE synchronization of oscillatory spindle activities.22,23 Corticothalamic axons from every neocortical area originate in neurons located in layers 6 or 5. Most corticothalamic axons are thin and slow conducting. Such corticothalamic axons originate in layer 6, whereas thick axons arise from layer 5, and some of these large fibers are collaterals of projections to the striatum, brain stem, and spinal cord. Although all corticothalamic axons release glutamate and are thus excitatory, the effect of electrical stimuli to neocortex or naturally synchronous cortical volleys are different on RE and TC neurons during the burstfiring mode, which occurs in slow-wave sleep or paroxysmal activities. RE neurons respond to cortical inputs with excitatory patterns consisting of rhythmic spike bursts, whereas TC neurons display prolonged GABAA-B-mediated IPSPs owing to the prior excitation of RE neurons (Figure 12.1). This opposite pattern is because of the prevalent excitation of RE neurons by corticothalamic axons as compared to the action on TC neurons, a conclusion based on the greater density of some subunits of glutamate receptors and greater amplitude of excitatory postsynaptic currents in RE neurons.24 The RE-induced inhibition of TC neurons during cortically generated seizures of the absence (petit mal) type (Figure 12.1b) may explain the obliteration of messages from the outside world and unconsciousness during these seizures.25

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 258 Wednesday, February 15, 2006 3:24 PM

258

Understanding Autism: From Basic Neuroscience to Treatment

+ Cortical

+ + Reticular + LTS − Thalamocortical (a)

50 mV

Depth - EEG area 4

0.2 s

Intra - cell area 4 −70 mV

Intra - cell RE −68 mV Intra - cell TC −60 mV (b)

FIGURE 12.1 Relations between corticothalamic, thalamic reticular (RE), and thalamocortical (TC) neurons: (A) Three neurons (cortical, RE, and TC) were intracellularly recorded and stained in cats. Signs of excitation and inhibition are indicated by plus and minus. For the sake of simplicity, local-circuit inhibitory neurons in the cortex and thalamus are not illustrated. Insets represent the response of RE and TC neurons to cortical stimulation (arrowheads point to stimulus artifacts). The GABAergic RE neuron responded to cortical stimulation with a high-frequency spike burst, followed by a sequence of spindle waves on a depolarizing envelope (membrane potential, −68 mV). The TC neuron responded to cortical stimulation (arrowhead) with a biphasic, GABAA-B-mediated IPSP, leading to a low-threshold spike (LTS)

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 259 Wednesday, February 15, 2006 3:24 PM

The Thalamus and Neuromodulatory Systems

259

THALAMIC PROJECTIONS OF BRAIN STEM AND FOREBRAIN NEUROMODULATORY SYSTEMS The pioneering study on the ascending brain stem reticular formation26 regarded this system as a monolith with widespread energizing actions upon the forebrain, mainly relayed in the thalamus, and introduced the notion of nonspecificity in the activating process. This notion and that of the diffuse pathways betrayed the state of primitive knowledge of the brain stem reticular core of that time. Modern tracing techniques combined with immunohistochemistry helped to define the multiple transmitter agents used by brain stem core neurons. This progress, coupled with electrophysiological recordings of identified neurons (see Section “Brain Stem–Thalamic Neurons during Tonic and Phasic Activation Processes”), challenged the idea of hypothetically ubiquitous projections of activating neurons and replaced the notion of a monolithic brain stem reticular core exerting global and undifferentiated energizing actions upon the forebrain by the concept of a differentiated structure. Besides different glutamatergic, cholinergic, and monoaminergic systems located in the brain stem reticular core, other neuromodulatory systems arise in the posterior hypothalamus and basal forebrain, and they also act on the thalamus.10 Most neuromodulatory systems have an activating effect on target neurons. Activation is defined27 as a state of readiness in cerebral networks, a state of membrane polarization that brings neurons closer to the firing threshold, thus ensuring safe synaptic transmission and quick responses without losing the sculpting inhibitory processes of short duration that are necessary during the adaptive state of waking.

THALAMIC PROJECTIONS OF GLUTAMATERGIC BRAIN STEM RETICULAR NEURONS

AND

CHOLINERGIC

Since the early 1980s, the natural temptation was to specify, within a structure previously viewed as nonspecific, the chemical code of activating brain stem core neurons with differential projections to various thalamic nuclei. Two main cholinergic nuclei (pedunculopontine tegmental and laterodorsal tegmental — PPT and LDT — also termed Ch5 and Ch6 groups, respectively) were described at the midbrain–pontine junction.28 Later, glutamate was found to be colocalized with acetylcholine (ACh) in brain stem cholinergic neurons.29 The action of glutamate on TC neurons is, in many respects, similar to that of ACh, namely depolarization by blockage of a “leak” K+ current.11 The number of brain stem reticular neurons found outside the territories of cholinergic neurons (most of them use glutamate as transmitter) exceeds that of cholinergic neurons within PPT/LDT nuclei. I shall

FIGURE 12.1 (CONTINUED) and a sequence of hyperpolarizing spindle waves (membrane potential, −70 mV); (B) relations between cortical (area 4), RE, and TC neurons of cat during spontaneously occurring, cortically generated paroxysmal activity with polyspike-wave complexes at 2 Hz. Note IPSPs in TC neuron (filled circles) in close time relation with spike bursts fired by RE neuron, driven from cortex. Modified from Steriade et al.,74 Contreras and Steriade,20 Lytton et al.,75 and Steriade.76

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 260 Wednesday, February 15, 2006 3:24 PM

260

Understanding Autism: From Basic Neuroscience to Treatment

30 25 20 15 10 5 0 VB

30 25 20 15 10 5 0 VA-VL

MG VM

LG A4 A3 A2 A1 0 P1 P2 P3 P4 Rostrocaudal level

A4 A3 A2 A1 0 P1 P2 P3 P4 Rostrocaudal level

30 25 20 15 10 5 0 MD LP A4 A3 A2 A1 0 P1 P2 P3 P4 Rostrocaudal level

LP

VA-VL

MD

SC IC

VM AC

VB

PAG RFB

FTC RN OC

PB MM

V4 LDT

A4 A3 A2 A1 0 P1 P2 P3 P4

FIGURE 12.2 Brain stem reticular projections to relay nuclei and associational thalamic nuclei in the cat. The parasagittal section shows some of the thalamic nuclei (ventroanteriorventrolateral, VA-VL; ventromedial, VM; ventrobasal, VB; lateroposterior, LP; mediodorsal, MD) where wheat germ agglutinin–horseradish peroxidase (WGA–HRP) was injected and

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 261 Wednesday, February 15, 2006 3:24 PM

The Thalamus and Neuromodulatory Systems

261

describe together the thalamic projections of cholinergic and glutamatergic brain stem core neurons. The ascending axons of cholinergic and glutamatergic reticular neurons at the midbrain-pontine junction are overwhelmingly relayed in the thalamus, with a less massive contingent directed ventrally to the basal forebrain. This stands in contrast to the direct cortical projections of monoaminergic brain stem nuclei (see section titled “Thalamic Projections of Brain Stem and Hypothalamic Monoaminergic Systems”). The only species that seems to have a significant direct brain stem reticular projection to the visual cortex is the chimpanzee.30 Retrograde tracing combined with immunohistochemistry of choline acetyltransferase (ChAT) in rats,31 cats and macaque monkeys32,33 showed that cholinergic nuclei at the midbrain–pontine junction (Ch5–Ch6 groups) project to virtually all thalamic nuclei, i.e., relay, association, intralaminar, and RE. In addition to Ch5–Ch6 projections, thalamic association and intralaminar nuclei receive a massive projection from the noncholinergic (presumably glutamatergic) neurons located in the upper midbrain and the rostral pontine reticular formation (Figure 12.2). Whereas brain stem reticular projections to relay thalamic nuclei modulate the synaptic transmission of impulses from various sensory and motor modalities, the projections to rostral intralaminar and RE nuclei are involved in the generalized processes of activation and oscillation in TC systems. The caudal intralaminar (centromedian-parafascicular) nuclei are essentially related to the striatum. In addition to their efferent connections to the caudate nucleus, the rostral intralaminar (centrolateral-paracentral) neurons project over widespread cortical territories where they exert depolarizing actions and thus represent an important link in the activating circuit from the brain stem reticular formation to the cerebral cortex.34,35 The axon of the same mesopontine cholinergic neuron may innervate more than one thalamic target. Thus, PPT or LDT neurons can be antidromically activated from two different dorsal thalamic nuclei,36 some mesopontine neurons innervate both thalamic RE and related TC neurons,37 and some brain stem cholinergic neurons have dual projections to the thalamus and the basal forebrain.38 The projections from the upper brain stem core to cholinergic neurons of the nucleus basalis (NB) in the basal forebrain mainly arise from neurons other than the mesopontine cholinergic ones, as anterogradely labeled PPT/LDT axons do not generally contact NB cholinergic cells with cortical projections and cholinergic NB cells receive most inputs from noncholinergic terminals.39 Even if direct brain stem–NB, cholinergic-to-cholinergic FIGURE 12.2 (CONTINUED) the main brain stem reticular territories where retrogradely labeled neurons were found (central tegmental field, FTC; peribrachial, PB, coextensive with PPT and LDT nuclei). Frontal stereotaxic planes are indicated (A4–P4). The three computergenerated graphs show the percentage (ordinate) of HRP-positive cells at various rostrocaudal levels (abscissa) from the total number of labeled neurons in the upper brain stem reticular core. For abbreviations of thalamic nuclei, see text. (Other abbreviations: AC, anterior commissure; IC, inferior colliculus; MM, medial mammillary nucleus; OC, optic chiasm; PAG, periaqueductal gray; RFB, retroflex bundle; RN, red nucleus; SC, superior colliculus.) (Modified from Steriade, M. et al., Projections of cholinergic and non-cholinergic neurons of the brain stem core to relay and associational thalamic nuclei in the cat and macaque monkey, Neuroscience 25, 47–67, 1988.)

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 262 Wednesday, February 15, 2006 3:24 PM

262

Understanding Autism: From Basic Neuroscience to Treatment

projections were to be eventually demonstrated, ACh elicits a muscarinic-mediated hyperpolarization of NB cells.40 Thus, the most likely candidates for the brain stem drive reaching NB neurons during activated states are brain stem glutamatergic neurons, within the limits of or outside the PPT/LDT nuclei, or even colocalized with ACh in the same cholinergic mesopontine cells. This possibility was supported by an experimental study.41 The presence of two parallel activating pathways (from the brain stem reticular formation to cortex via synaptic relays within the thalamus or NB) is supported by experiments showing that brain-stem-induced depolarization of cortical neurons, their enhanced excitability, and replacement of slow oscillations by fast rhythms can be achieved after extensive ipsilateral lesions of either thalamus or NB.42 Thus, at variance with some assumptions placing exclusive emphasis on any one cholinergic system, either PPT/LDT or NB cholinergic nuclei are sufficient to activate the cerebral cortex.

THALAMIC PROJECTIONS OF BRAIN STEM MONOAMINERGIC SYSTEMS

AND

HYPOTHALAMIC

In contrast with the congruent results on the brain stem cholinergic and glutamatergic innervation of different thalamic nuclei in various species, the density of norepinephrine (NE) projections from locus coeruleus (LC) and serotonin (5-HT) projections from the dorsal raphe (DR) nucleus to the thalamus greatly varies from nucleus to nucleus and from species to species. Besides the thalamic RE nucleus that is very densely innervated by both LC and DR throughout its extent,43 the highest density of the NE and 5-HT innervation is found in the dorsal and ventral part of the lateral geniculate nucleus, whereas other thalamic nuclei display low- to moderate-density fibers. Immunocytochemical mapping studies showed that posterior hypothalamic neurons releasing histamine (HA) mainly project to the ventral part of the lateral geniculate complex, whereas A and C laminae are sparsely labeled and the perigeniculate (RE) nucleus display intermediate amounts of label.44 Very few, if any, conventional synapses are found in other structures where numerous HA profiles were examined. In view of these data, HA probably achieves its action in a nonsynaptic (volume transmission) fashion, similar to other monoamines and ACh in the cerebral cortex.45 The thalamus is largely bypassed by dopaminergic axons.

BASAL FOREBRAIN PROJECTIONS

TO THE

THALAMUS

Basal forebrain nuclei contribute to the cholinergic innervation of a limited number of nuclei in the medial and rostral thalamus. The rostral pole of thalamic RE nucleus as well as mediodorsal and anteromedial nuclei receive projections from cholinergic and noncholinergic neurons of different basal forebrain cell groups.46,47 The basal forebrain cholinergic projection to the RE thalamic nucleus is crucial in forebrain activation processes because, in addition to the brain stem–RE cholinergic projection, it is a source for the disruption of synchronized spindle oscillations (which accompany early sleep stages) at their very site of genesis (the RE nucleus) because ACh hyperpolarizes RE neurons.11 The basal forebrain projection to the MD nucleus can be involved in the modulation of memory and learning processes.

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 263 Wednesday, February 15, 2006 3:24 PM

The Thalamus and Neuromodulatory Systems

263

The majority of noncholinergic basal forebrain neurons with thalamic projections are GABAergic cells intermingled with the cholinergic ones.48 The inhibitory effect of GABA on thalamic RE neurons is associated with a marked increase in Cl− conductance. These data, combined with the ACh-induced hyperpolarization and increased K+ conductance of thalamic RE neurons, support the idea of a permissive role played by NB (in conjunction with brain stem cholinergic) cells in deactivation and sleep onset by promoting spindle oscillations immediately after the reduction in firing rates of NB neurons.49

BRAIN STEM–THALAMIC NEURONS DURING TONIC AND PHASIC ACTIVATION PROCESSES Here, I focus on changes in firing rates and patterns of glutamatergic and cholinergic neurons recorded from the upper brain stem reticular core and mesopontine cellular aggregates, which lead to tonic and phasic activation processes in the thalamus and cerebral cortex during wakefulness and rapid-eye-movement (REM), or dreaming, sleep. Rostral midbrain50 and bulbar51 reticular neurons with antidromically identified projections to thalamic intralaminar and ventromedial nuclei double their firing rates in advance of the most precocious signs of forebrain activation during transition from non-REM sleep to either waking or REM sleep. Those two thalamic targets were chosen because they project over widespread cortical territories11 and may thus account for the diffuse cortical excitatory processes associated with tonic activation in these two brain-aroused states. Such data were used to evaluate the hypothesis that an increase in discharge rates of brain stem reticular neurons precedes overt signs of activation in rostral targets of neuromodulatory systems. Pooled analysis of midbrain and bulbar reticular neuronal groups revealed that a statistically significant increase in firing rates occurred 15 to 30 seconds before the end of non-REM-sleep epochs that developed into brain-activated states of waking or REM sleep.50,51 There are virtually no cholinergic cells in the rostral mesencephalon, but the established direct brain stem–thalamic excitatory action34 suggests that those midbrain neurons probably use excitatory amino acids as neurotransmitters. Unit recordings were performed in cat PPT and LDT cholinergic nuclei at the mesopontine junction to investigate the relation between the activity of their neurons and the tonic process of forebrain activation during waking and REM sleep.36 At the rostral PPT level, cholinergic neurons represent about 85 to 90% of the neuronal population,52 whereas significant numbers of catecholaminergic neurons within the PPT nucleus appear only more caudal. Therefore, the probability that neurons located in the rostral part of the PPT nucleus are cholinergic is very high. These neurons were antidromically identified as projecting to the different thalamic nuclei.36 The majority of thalamically projecting neurons of the cholinergic PPT/LDT nuclei displayed tonic discharge patterns during waking, decreased firing rates during transition from waking to non-REM sleep, and increased firing rates by about 1 minute before the earliest change from electrical synchronized activity during non-REM sleep to activation patterns during REM sleep (Figure 12.3). These data from unit

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 264 Wednesday, February 15, 2006 3:24 PM

264

Understanding Autism: From Basic Neuroscience to Treatment

b LP

LP

a 2 ms

Hertz

40 W 12.7 30

S 8.8

REM 23.8

20 10 0 03000

03200 Time (hmmss)

03400

FIGURE 12.3 Antidromic identification of neuron recorded from pedunculopontine tegmental (PPT) cholinergic nucleus and sequential firing rate (SFR) of PPT neuron during the wakingsleep cycle. Chronically implanted cat. Top panel: two simultaneously recorded PPT neurons (small action potential a and large action potential b); left: antidromic activation of a cell and synaptic excitation of cell b by stimulating thalamic lateroposterior (LP) nucleus; right: changing the polarity of stimulation led to antidromic invasion of cell b. Bottom panel: SFR of thalamically projecting neuron across the waking-sleep cycle. Abscissa indicates real time. Mean firing rates during waking sleep (W), slow-wave sleep (S) and REM sleep are indicated (Hz) for each state. Transitional WS and pre-REM epochs are indicated by vertical interrupted lines (at 0:30:53 and 0:34:34, respectively). Note cyclic activity and decreased firing rate toward the end of W state, preceding S, and increased firing rate preceding REM sleep. (Modified from Steriade, M. et al., Neuronal activities in brain stem cholinergic nuclei related to tonic activation processes in thalamocortical systems, J. Neurosci. 10, 2541–2559, 1990.)

recordings match the increased levels of ACh released in the thalamus during waking and REM sleep when compared to those of non-REM (slow-wave) sleep.53 Thus, PPT/LDT neurons may be considered the best candidates for inducing the cholinergic processes associated with forebrain activation, namely, direct excitation of TC neurons and blockage of synchronized sleep spindle oscillations by inhibiting RE thalamic neurons. In other studies, presumptive cholinergic neurons from the PPT nucleus were antidromically identified from the posterior hypothalamus and also found to display tonic discharges during waking and REM sleep or highly specifically during REM sleep.54 Brain stem cholinergic neurons are also at the origin of phasic ponto-geniculooccipital (PGO) waves (stigmatic events of REM sleep when dreaming episodes occur). These potentials are generated in different neuronal groups of the brain stem reticular core and are transferred to many TC systems, in addition to the visual one

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 265 Wednesday, February 15, 2006 3:24 PM

The Thalamus and Neuromodulatory Systems

265

where they were originally thought to be confined. The interest for the PGO waves stemmed from the discovery that eye movement direction is related to gaze direction in dream imagery, coupled with data from experiments showing that saccadic REMs are coincident with PGO events.55,56 These observations led to the consensus that PGO waves are physiological correlates of brain activation during dreaming sleep, “the stuff that dreams are made of.” Neurons that transfer the brain stem–generated PGO waves to the thalamus are located in and around the PPT cholinergic nucleus. That these thalamically projecting brain stem neurons are cholinergic results from the demonstration that systemic administration and iontophoretic application of nicotinic antagonists into the lateral geniculate nucleus abolish the thalamic PGO waves, and at chronic stages after chemical lesions of PPT cholinergic neurons, PGO waves are largely suppressed during REM sleep (see Reference 10).10 There are several classes of PGO-on PPT/LDT neurons (antidromically activated from various thalamic nuclei) that fire single action potentials, trains of single spikes, or spike bursts with different patterns, all of which precede the negative peak of PGO field potential recorded from thalamic nuclei:57 1. Some PPT/LDT neurons fire single spikes preceding by 15 to 25 msec the negative peak of the thalamic PGO field potential. Intracellular recordings of such neurons revealed that these action potentials rose from large composite excitatory postsynaptic potentials (EPSPs).58 2. Another cell class discharges trains of single spikes whose onset precede by 100 to 200 msec the thalamic PGO wave.57 3. Two types of PGO-on bursting neurons were recorded in the PPT/LDT nuclei. One of them fire sluggish spike bursts (low-frequency, less than 150 Hz) following a period of neuronal silence,59,60 and they probably represent low-threshold bursts following the inhibition acting upon PPT cells by GABAergic substantia nigra pars reticulata neurons.61 The other type of bursting neurons discharge high-frequency spike bursts (more than 300 Hz) that, distinctly from the previous type, occur on a background of tonically increased discharge rates during REM sleep (Figure 12.4).57 These data raise the possibility that such high-frequency bursts may be generated at a depolarized level, at variance to what is expected for a common low-threshold Ca2+ spike. 4. Still another type of PPT/LDT neurons fire tonically at high rates (more than 30 Hz) during epochs of REM sleep without PGO events and stop firing prior to and during thalamic PGO waves.57 The behavior of these PGO-off cells is unexpected for cells located within the limits of the brain stem cholinergic nuclei and is the functional counterpart of the heterogeneity of PPT/LDT nuclei, probably representing GABAergic neurons within the brain stem cholinergic nuclei. Their silenced firing prior and during PGO waves could disinhibit adjacent neurons with tonically increased discharges during PGO waves. These investigations in behaving animals57 revealed the variety of PPT/LDT neurons related to the genesis of PGO waves as well as the organizational complexity

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 266 Wednesday, February 15, 2006 3:24 PM

266

Understanding Autism: From Basic Neuroscience to Treatment

PB-Unit

1

2

3

LG-PGO

EOG 1

2

1s

3

0.2 s

3

20 ms (a) Hz

Hz

300

500

200

300

100

100

38.0

0 −300

0 mS

0 −300

300

38.0 300

0 mS

(b) Prob × 100 5

Prob × 100 1.5 1

3

0.5

1 0

0

0

5

10

15

20

25

mS

0

5

10

15

20

25

mS (c)

FIGURE 12.4 Activity of PGO-on burst neuron in the PPT nucleus during REM sleep of chronically implanted cat: (A) polygraphic ink-written record (unit discharges, deflections exceeding the common level represent high-frequency spike bursts; lateral geniculate, LG-PGO waves; and electrooculogram, EOG) and original spikes with two different speeds, showing three (1 to 3) PGO-related bursts, as indicated on the first trace of the ink-written records. (B) peri-PGO histograms (10-msec bins) of PPT cell’s discharges (time 0 is the negative peak of the LG-PGO wave) for single and clustered PGO waves (left and right panels, respectively). The level of overall spontaneous discharges in REM-sleep (38 Hz) is also indicated; (C) interspike interval histograms (0.5-msec bins) of PPT cell’s activity during the period of

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 267 Wednesday, February 15, 2006 3:24 PM

The Thalamus and Neuromodulatory Systems

MG

267

SC

RN SNr

CG

PP

FTC BC

PB

BP BP PPRF

TH SB T

−

+ PB/LDT

− Off

+ −

HB

FIGURE 12.5 Tentative schemes of cellular interactions underlying the genesis of PGO waves and their transfer from PB(PPT)/LDT nuclei to the thalamus (TH). Top: three frontal sections (from rostral to caudal) depicting the inputs to PB area from substantia nigra pars reticulata (SNr), central tegmental field (FTC), and the paramedian pontine reticular formation (PPRF). (Other abbreviations: BC, brachium conjunctivum; BP, brachium pontis; CG, central gray; MG, medial geniculate nucleus; PP, pes pedunculi; RN, red nucleus; SC, superior colliculus); bottom: hypothesized excitatory (+) and inhibitory (−) interactions between four cellular types (PGO-on and PGO-off) in PB(PPT)/LDT nuclei. (Symbols: SB, sluggish burst; HB, high-frequency burst; Off, PGO-off cell; T, tonic cell.) (Modified from Steriade, M. et al., Different cellular types in mesopontine cholinergic nuclei related to ponto-geniculo-occipital waves, J. Neurosci. 10, 2560–2579, 1990.)

of PPT/LDT circuits and related structures, such as the midbrain central tegmental field, substantia nigra pars reticulata, and the paramedian pontine reticular formation, with final projections to the thalamus (Figure 12.5). FIGURE 12.4 (CONTINUED) increased firing rate around the PGO events for single and clustered PGP wave (left and right panels, respectively). Note the very short intervals (less than 3 msec) reflecting high-frequency bursts. (Modified from Steriade, M. et al., Different cellular types in mesopontine cholinergic nuclei related to ponto-geniculo-occipital waves, J. Neurosci. 10, 2560–2579, 1990.)

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 268 Wednesday, February 15, 2006 3:24 PM

268

Understanding Autism: From Basic Neuroscience to Treatment

Cortical area 5

Cortical area 7

Reticular thalamic

2s

EEG area 5 (a)

800

200

area 5 - area 7

0

−4

−2

0 0

3000

0

2

4

s

−2

0 0

2

4

−4

−2

200

area 7 - reticular

−4

area 5 - area 7

s

0

2

4

s

4

s

area 7 - reticular

−4

−2

0

2

(b)

FIGURE 12.6 Transition from slowly synchronized (sleep-like) to activated brain states in cat: (A) simultaneous recordings of unit discharges from cortical neurons in areas 5 and 7, neuron from the lateral part of the thalamic RE nucleus, together with EEG from area 5. Note the synchronous discharges in both cortical and thalamic RE neurons (with the frequency of the slow sleep oscillation, 0.4 Hz) and disruption of slow oscillation with appearance of tonic

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 269 Wednesday, February 15, 2006 3:24 PM

The Thalamus and Neuromodulatory Systems

269

MODULATORY ACTIONS ON THALAMOCORTICAL AND THALAMIC INHIBITORY NEURONS Setting into action brain stem cholinergic nuclei produces a prolonged depolarization of TC neurons in different dorsal thalamic nuclei, which is associated with increase in the apparent input resistance (Rin) and is due to muscarinic receptors of TC neurons.62,63 The transition from sleep patterns (represented by long-range synchronized bursting activities within the low-frequency rhythms) to activated patterns (indicated by tonic firing of single action potentials) can be induced by stimulation of brain stem core modulatory systems or occurs spontaneously during the shift from slow-wave sleep activity to brain-activated patterns of waking or REM sleep.64 This change is not only seen in TC neurons, with the consequence of inducing cortical activation, but also in thalamic RE neurons (Figure 12.6). In the latter neuronal type, the depolarization and tonic firing upon brain arousal are produced by glutamatergic brain stem–thalamic neurons (or by activating actions exerted by TC as well as corticoRE glutamatergic neurons), because cholinergic neurons induce hyperpolarization of RE neurons with increased membrane conductance.11,65 Actually, the spontaneous firing of RE neurons is characterized by spike bursts during natural slow-wave sleep and tonic firing of single spikes during waking and REM sleep.66 These changes in spontaneous firing are associated with enhancement of evoked responses during activated states. Antidromic responses of TC neurons increase by about 60% 10 to 25 seconds in advance of the most precocious sign of brain activation with transition from slow-wave sleep patterns to waking or REM sleep, and blockage of antidromic responses occur in advance of the occurrence of sleep spindles.64,67 Intracellular-recorded orthodromic responses of TC neurons are also enhanced by brain stem cholinergic stimulation.68 In the visual thalamus, photically evoked potentials are effected by brain stem reticular stimulation despite there being no alterations in simultaneously recorded responses from the optic tract.69 Inhibitory processes in TC neurons during activated and deactivated states were first investigated by measuring the duration of the period of suppressed firing following an afferent volley. However efficiently the mesencephalic reticular stimulation blocked cyclic periods of suppressed firing, it did not eliminate the early inhibitory phase during which TC neurons remained unresponsive to incoming signals. The idea that the early inhibitory phase in TC neurons is preserved during brain stem reticular-induced arousal, which allows these neurons with feature detection properties that assist in discriminatory functions, was later supported by experiments using intracellular recordings. Midbrain reticular stimulation blocks the long-lasting, cyclic hyperpolarizations of a TC neuron, consisting of an abolition of the prolonged tail of these hyperpolarizations and their rhythmic repetition, but preserves the early, short-lasting IPSP (Figure 12.7).70 The discovery of triphasic FIGURE 12.6 (CONTINUED) firing, associated with an activated EEG pattern; (B) crosscorrelograms between area 5 and area 7 neurons (top) and between area 7 and thalamic RE neurons (bottom) during sleep-like state (left) and during activation period (right). (Modified from Steriade, M., Contreras, D., and Amzica, F., Synchronized sleep oscillations and their paroxysmal developments, Trends Neurosci. 17, 199–208, 1994.)

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 270 Wednesday, February 15, 2006 3:24 PM

270

Understanding Autism: From Basic Neuroscience to Treatment

A

50 mV

1

50 ms

2

MRF 3

a

b 10 mV

B1

20 mV

20 ms

2 a

b

20 ms

FIGURE 12.7 Effects of midbrain reticular formation (MRF) stimulation on inhibitory processes in thalamocortical (TC) neurons. Intracellular recordings in cats under (A) barbiturate anesthesia and (B) urethane anesthesia in reserpine-treated animal. (A) thalamic ventrolateral (VL) neuron. Stimulation of motor cortex (arrowheads) evokes an (1) antidromic spike followed by rhythmic inhibitory-rebound sequences, (2) a conditioning pulse-train (320 Hz) to the MRF leaves intact the cortically elicited early IPSP but abolishes the late phase of hyperpolarization and the following inhibition-rebound sequences, and (3) the effect of the MRF stimulation alone is shown; (B) two (1-2) TC neurons from the lateral geniculate (LG) nucleus. Optic chiasm stimulation was adjusted to enhance the separation of the two (GABAA and GABAB) IPSPs. When preceded by a conditioning pulse train to the cholinergic pedunculopontine tegmental (PPT) nucleus, only the first (GABAA) IPSP persisted. Note that the blockage of the second inhibitory component was obtained with PPT conditioning stimuli subthreshold for inducing an early depolarizing response in these LG neurons. (Modified from Steriade, M., The excitatory-inhibitory sequence in thalamic and neocortical cells: state-related changes and regulatory systems, in Dynamic Aspects of Neocortical Function, Edelman, G.M., Gall, W.E., and Cowan, W.M., Eds., John Wiley & Sons, New York, 1984, pp. 107–157; Hu, B., Steriade, M., and Deschênes, M., The effects of brain stem peribrachial stimulation on neurons of the lateral geniculate nucleus, Neuroscience 31, 13–24, 1989.) © 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 271 Wednesday, February 15, 2006 3:24 PM

−58 mV

271

10 mV

The Thalamus and Neuromodulatory Systems

a A

B

40 ms

A

FIGURE 12.8 Three GABA-mediated IPSPs in anterior thalamic neurons of decorticated cat in vivo. For origin of “miniature” GABAa-mediated IPSP in presynaptic dendrites of localcircuit interneurons, see main text. An isolated GABAa-IPSP could be evoked by a single stimulus, with the lowest intensity, applied to the mammillary body. Two stimuli evoked both GABAa- and GABAA-IPSPs. And, only by increasing the stimulation strength of the two stimuli was the full sequence (GABAa-A-B) evoked. (Modified from Paré, D., Curró Dossi, R., and Steriade, M., Three types of inhibitory postsynaptic potentials generated by interneurons in the anterior thalamic complex of cat, J. Neurophysiol. 66, 1190–1204, 1991; Steriade, M., Local gating of information processing through the thalamus, Neuron 41, 493–494, 2004.)

IPSPs in anterior thalamic neurons (Figure 12.8) shed further light on the selective preservation of the earliest IPSP, with simultaneous abolition of later IPSPs. As cat anterior thalamic nuclei are devoid of connections from the thalamic RE nucleus,71 the earliest IPSP (called GABAa, to distinguish it from the subsequent sequence of

AVG

MN LDT

20 mV

10 mV

−55 mV

40 ms

FIGURE 12.9 GABAa IPSP elicited by local interneurons in anterior thalamic neuron of cat and its enhancement following stimulation of mesopontine cholinergic laterodorsal tegmental (LDT) nucleus; mammillary nucleus (MN)-evoked isolated GABAa IPSP. Left: five traces depicting (from top to bottom) MN-evoked control a-IPSP response; LDT + MN stimulation at three different time-intervals; and LDT stimulation alone; right: averages (AVG) of five responses to stimulation of MN and LDT + MN, as indicated by curved arrows; dotted lines tentatively indicate the baseline. Note the depolarization elicited by LDT stimulation alone (left bottom trace). (Modified from Curró Dossi, R., Paré, D., and Steriade, M., Various types of inhibitory postsynaptic potentials in anterior thalamic cells are differentially altered by stimulation of laterodorsal tegmental cholinergic nucleus, Neuroscience 47, 279–289, 1992.) © 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 272 Wednesday, February 15, 2006 3:24 PM

272

Understanding Autism: From Basic Neuroscience to Treatment

GABAA-B IPSPs) is presumably generated by GABA release from the intraglomerular presynaptic dendrites of local-circuit interneurons.72 During brain activation elicited by stimulation of mesopontine cholinergic nuclei, GABAa IPSP is preserved or even enhanced (Figure 12.9), whereas GABAB IPSP and occasionally GABAA IPSP are abolished.73 To conclude, brain stem modulatory actions on thalamic neurons basically consist of enhanced excitability of TC neurons, consequently leading to activation processes in neocortex, as well as sculpting inhibition resulting from the blockade of prolonged and cyclic hyperpolarizations but with preservation of the earliest phase of inhibition that may assist in the enhancement of center-surround mechanism and lateral inhibition during attentive states.

REFERENCES 1. Ceasey, H. et al., Brain morphometry in autistic men as measured by volumetric computed tomography, Arch. Neurol., 43, 669–672, 1986. 2. Courchesne, E., Neuroanatomic imaging in autism, Pediatrics, 87, 781–790, 1991. 3. Müller, R.A. et al., Impairment of dentato-thalamo-cortical pathway in autistic men: language activation data from positron emission tomography, Neurosci. Lett., 245, 1–4, 1998. 4. Tsatsanis, K.D. et al., Reduced thalamic volume in high-functioning individuals with autism, Biol. Psychiatry, 53, 121–129, 2003. 5. Herbert, M.R. et al., Dissociations of cerebral cortex, subcortical and cerebral white matter volumes in autistic boys, Brain, 126, 1182–1192, 2003. 6. Waiter, G.D. et al., A voxel-based investigation of brain structure in male adolescents with autistic spectrum disorder, NeuroImage, 22, 619–625, 2004. 7. Hall, G.B.C., Szechtman, H., and Nahmias, C., Enhanced salience and emotion recognition in autism: a PET study, Am. J. Psychiatry, 160, 1439–1441, 2003. 8. Damasio, A.R. and Maurer, R.G., A neurological model for childhood autism, Arch. Neurol., 35, 777–786, 1978. 9. Chugani, D.C. et al., Altered serotonin synthesis in the dentatothalamocortical pathway in autistic boys, Ann. Neurol., 42, 666–669, 1997. 10. Steriade, M. and McCarley, R.W., Brain Control of Wakefulness and Sleeping, Kluwer-Plenum-Springer, New York, 2005, 702 pp. 11. Steriade, M., Jones, E.G., and McCormick, D.A., Thalamus (Vol. 1, Organisation and Function), Elsevier, Amsterdam, 1997, 959 pp. 12. Llinás, R.R., Leznik, E., and Urbano, F.J., Temporal binding via cortical coincidence detection of specific and nonspecific thalamocortical inputs: a voltage-dependent dyeimaging study in mouse brain slices, Proc. Natl. Acad. Sci. USA 99, 449–454, 2002. 13. Steriade, M. et al., Abolition of spindle oscillations in thalamic neurons disconnected from nucleus reticularis thalami, J. Neurophysiol. 54, 1473–1497, 1985. 14. Timofeev, I. and Steriade, M., Low-frequency rhythms in the thalamus of intactcortex and decorticated cats, J. Neurophysiol. 76, 4152–4168, 1996. 15. Thomson, A.M., Inhibitory postsynaptic potentials evoked in thalamic neurons by stimulation of the reticularis nucleus evoke slow spikes in isolated rat brain slices, Neuroscience 25, 491–502, 1988. 16. Liu, X.B., Warren, R.A., and Jones, E.G., Synaptic distribution of afferents from reticular nucleus in ventroposterior nucleus of cat thalamus, J. Comp. Neurol. 352, 187–202, 1995.

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 273 Wednesday, February 15, 2006 3:24 PM

The Thalamus and Neuromodulatory Systems

273

17. Deschênes, M., Madariaga-Domich, A., and Steriade, M., Dendrodendritic synapses in cat reticularis thalami nucleus, a structural basis for thalamic spindle synchronization, Brain Res. 334, 169–171, 1985. 18. Yen, C.T. et al., The morphology of physiologically identified GABAergic neurons in the somatic sensory part of the thalamic reticular nucleus in the cat, J. Neurosci. 5, 2254–2268, 1985. 19. Zhang, S.J., Huguenard. J.R., and Prince, D.A., GABAA receptor-mediated Cl currents in rat thalamic reticular and relay neurons, J. Neurophysiol. 78, 2280–2286, 1997. 20. Contreras, D. and Steriade, M., Spindle oscillation: the role of corticothalamic feedback in a thalamically generated rhythm, J. Physiol. (Lond.) 490, 159–179, 1996. 21. Steriade, M. et al., The deafferented reticular thalamic nucleus generates spindle rhythmicity, J. Neurophysiol. 57, 260–273, 1987. 22. Landisman, C.E. et al., Electrical synapses in the thalamic reticular nucleus, J. Neurosci. 22, 1002–1009, 2002. 23. Fuentealba, P. et al., Experimental evidence and modeling studies support a synchronizing role for electrical coupling in the cat thalamic reticular neurons in vivo, Eur. J. Neurosci. 20, 111–119, 2004. 24. Golshani, P., Liu, X.B., and Jones, E.G., Differences in quantal amplitude reflect GluR4-subunit number at corticothalamic synapses on two populations of thalamic neurons, Proc. Natl. Acad. Sci. USA 98, 4172–4177, 2001. 25. Steriade, M., Neuronal Substrates of Sleep and Epilepsy, Cambridge University Press, Cambridge, 2003, 517 pp. 26. Moruzzi, G. and Magoun, H.W., Brain stem reticular formation and activation of the EEG, Electroencephalogr. Clin. Neurophysiol. 1, 455–473, 1949. 27. Steriade, M., Alertness, quiet sleep, dreaming, in Cerebral Cortex, Vol. 9 (Normal and Altered States of Function), Peters, A. and Jones, E.G., Eds., Plenum, New York, 1991, pp. 279–357. 28. Mesulam, M.M. et al., Central cholinergic pathways in the rat: an overview based on an alternative nomenclature (Ch1-Ch6), Neuroscience 10, 1185–1201, 1983. 29. Lavoie, B. and Parent, A., Pedunculopontine nucleus in the squirrel monkey: distribution of cholinergic and monoaminergic neurons in the mesopontine tegmentum with evidence for the presence of glutamate in cholinergic neurons, J. Comp. Neurol. 344, 190–209, 1994. 30. Tigges, J. et al., Subcortical structures projecting to visual cortical areas in squirrel monkey, J. Comp. Neurol. 209, 29–40, 1982. 31. Hallanger, A.E. et al., The origins of cholinergic and other subcortical afferents to the thalamus in the rat, J. Comp. Neurol. 262, 105–124, 1987. 32. Steriade, M. et al., Projections of cholinergic and non-cholinergic neurons of the brainstem core to relay and associational thalamic nuclei in the cat and macaque monkey, Neuroscience 25, 47–67, 1988. 33. Paré, D. et al., Projections of upper brainstem cholinergic and non-cholinergic neurons of cat to intralaminar and reticular thalamic nuclei, Neuroscience 25, 69–88, 1988. 34. Steriade, M. and Glenn, L.L., Neocortical and caudate projections of intralaminar thalamic neurons and their synaptic excitation from the midbrain reticular core, J. Neurophysiol. 48, 352–371, 1982. 35. Kinomura, S. et al., Activation by attention of the human reticular formation and thalamic intralaminar nuclei, Science 271, 512–515, 1996. 36. Steriade, M. et al., Neuronal activities in brainstem cholinergic nuclei related to tonic activation processes in thalamocortical systems, J. Neurosci. 10, 2541–2559, 1990.

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 274 Wednesday, February 15, 2006 3:24 PM

274

Understanding Autism: From Basic Neuroscience to Treatment

37. Spreafico, R. et al., Branching projections from mesopontine nuclei to the nucleus reticularis and related thalamic nuclei: a double labeling study in the rat, J. Comp. Neurol. 336, 481–492, 1993. 38. Losier, B.J. and Semba, K., Dual projections of single cholinergic and aminergic brainstem neurons to the thalamus and basal forebrain in the rat, Brain Res. 604, 41–52, 1993. 39. Hallanger, A.E. and Wainer, B.H., Ascending projections from the pedunculopontine tegmental nucleus and the adjacent mesopontine tegmentum in the rat, J. Comp. Neurol. 274, 483–515, 1988. 40. Khateb, A. et al., Modulation of cholinergic nucleus basalis neurons by acetylcholine and N-methyl-D-aspartate, Neuroscience 81, 47–55, 1997. 41. Rasmusson, D.D., Clow, K., and Szerb, J.C., Modification of neocortical acetylcholine release and electroencephalogram desynchronization due to brainstem stimulation by drugs applied to the basal forebrain, Neuroscience 60, 665–677, 1994. 42. Steriade, M., Amzica, F., and Nuñez, A., Cholinergic and noradrenergic modulation of the slow (about 0.3 Hz) oscillation in neocortical cells, J. Neurophysiol. 70, 1384–1400, 1993. 43. Morrison, J.H. and Foote, S.L., Noradrenergic and serotonergic innervation of cortical, thalamic, and tectal visual structures in old and new world monkeys, J. Comp. Neurol. 243, 117–138, 1986. 44. Uhlrich, D.J., Manning, K.A., and Pienkowski, T.P., The histaminergic innervation of the lateral geniculate complex in the cat, Vis. Neurosci. 10: 225–235, 1993. 45. Descarries, L., Gisiger, V., and Steriade, M., Diffuse transmission by acetylcholine in the CNS, Prog. Neurobiol. 53, 603–625, 1997. 46. Parent, A. et al., Basal forebrain cholinergic and non-cholinergic projections to the thalamus and brainstem in cats and monkeys, J. Comp. Neurol. 277, 281–301, 1988. 47. Asanuma, C., Basal forebrain projections to the reticular thalamic nucleus in rat, Proc. Natl. Acad. Sci. U.S.A. 86, 4746–4750, 1989. 48. Asanuma, C. and Porter, L.L., Light and electron microscopic evidence for a GABAergic projection from the caudal basal forebrain to the thalamic reticular nucleus in rats, J. Comp. Neurol. 302, 159–172, 1990. 49. Steriade, M. and Buzsáki, G., Parallel activation of thalamic and cortical neurons by brainstem and basal forebrain cholinergic systems, in Brain Cholinergic Systems, Steriade, M. and Biesold, D., Eds., Oxford University Press, Oxford, 1990, pp. 3–63. 50. Steriade, M., Oakson, G., and Ropert, N., Firing rates and patterns of midbrain reticular neurons during steady and transitional states of the sleep-waking cycle, Exp. Brain Res. 46, 37–51, 1982. 51. Steriade, M., Sakai, K., and Jouvet, M., Bulbothalamic neurons related to thalamocortical activation processes during paradoxical sleep, Exp. Brain Res. 54, 463–475, 1984. 52. Webster, H.H. and Jones, B.E., Neurotoxic lesions of the dorsolateral pontomesencephalic tegmentum cholinergic area in the cat. II. Effects upon sleep-waking states, Brain Res. 458, 285–302, 1988. 53. Williams, J.A. et al., State-dependent release of acetylcholine in rat thalamus measured by in vivo microdyalisis, J. Neurosci, 14, 5236–5242, 1994. 54. El Mansari, M., Sakai, K., and Jouvet, M., Unitary characteristics of presumptive cholinergic tegmental neurons during the sleep-waking cycle in freely moving cats, Exp. Brain Res. 76, 519–529, 1989.

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 275 Wednesday, February 15, 2006 3:24 PM

The Thalamus and Neuromodulatory Systems

275

55. Dement, W.C. and Kleitman, N., The relation of eye movements during sleep to dream activity: an objective method for the study of dreaming, J. Exp. Psychol. 53, 339–346, 1957. 56. Jouvet, M., The role of monoamine and acetylcholine-containing neurons in the regulation of the sleep-waking cycle, Ergeb. Physiol. 64, 166–307, 1972. 57. Steriade, M. et al., Different cellular types in mesopontine cholinergic nuclei related to ponto-geniculo-occipital waves, J. Neurosci. 10, 2560–2579, 1990. 58. Paré, D. et al., Brainstem genesis of reserpine-induced ponto-geniculo-occipital waves: an electrophysiological and morphological investigation, Exp. Brain Res. 81, 533–544, 1990. 59. McCarley, R.W., Nelson, J.P., and Hobson, J.A., Ponto-geniculo-occipital (PGO) burst neurons: correlative evidence for neuronal generators of PGO waves, Science 201, 269–272, 1978. 60. Sakai, K. and Jouvet, M., Brainstem PGO-on cells projecting directly to the cat lateral geniculate nucleus, Brain Res. 194, 500–505, 1980. 61. Datta, S. et al., Substantia nigra reticulata neurons during sleep-waking states: relation with ponto-geniculo-occipital waves, Brain Res. 566, 344–347, 1991. 62. Hu, B., Steriade, M., and Deschênes, M., The effects of brainstem peribrachial stimulation on neurons of the lateral geniculate nucleus, Neuroscience 31, 13–24, 1989. 63. Curró Dossi, R., Paré, D., and Steriade, M., Short-lasting nicotinic and long-lasting muscarinic depolarizing responses of thalamocortical neurons to stimulation of mesopontine cholinergic nuclei, J. Neurophysiol. 65, 393–406, 1991. 64. Glenn, L.L. and Steriade, M., Discharge rate and excitability of cortically projecting intralaminar thalamic neurons during waking and sleep states, J. Neurosci. 2, 1287–1404, 1982. 65. Hu, B., Steriade, M., and Deschênes, M., The effects of peribrachial stimulation on reticular thalamic neurons: the blockage of spindle waves, Neuroscience 31, 1–12, 1989. 66. Steriade, M., Domich, L., and Oakson, G., Reticularis thalami neurons revisited: activity changes during shifts in states of vigilance, J. Neurosci. 6, 68–81, 1986. 67. Steriade, M., Apostol, V., and Oakson, G., Control of unitary activities in cerebellothalamic pathway during wakefulness and synchronized sleep, J. Neurophysiol. 34, 389–413, 1971. 68. Timofeev, I., Contreras, D., and Steriade, M., Synaptic responsiveness of cortical and thalamic neurons during various phases of slow oscillation in cat, J. Physiol. (Lond.) 494, 265–278, 1996. 69. Steriade, M. and Demetrescu, M., Unspecific systems of inhibition and facilitation of potentials evoked by intermittent light, J. Neurophysiol. 23, 602–617, 1960. 70. Steriade, M., The excitatory-inhibitory sequence in thalamic and neocortical cells: state-related changes and regulatory systems, in Dynamic Aspects of Neocortical Function, Edelman, G.M., Gall, W.E., and Cowan, W.M., Eds., John Wiley & Sons, New York, 1984, pp. 107–157. 71. Steriade, M., Parent, A., and Hada, J., Thalamic projections of nucleus reticularis thalami: a study using retrograde transport of horseradish peroxidase and double fluorescent tracers, J. Comp. Neurol. 229, 531–547, 1984. 72. Paré, D., Curró Dossi, R., and Steriade, M., Three types of inhibitory postsynaptic potentials generated by interneurons in the anterior thalamic complex of cat, J. Neurophysiol. 66, 1190–1204, 1991.

© 2006 by Taylor & Francis Group, LLC

2732_C012.fm Page 276 Wednesday, February 15, 2006 3:24 PM

276

Understanding Autism: From Basic Neuroscience to Treatment

73. Curró Dossi, R., Paré, D., and Steriade, M., Various types of inhibitory postsynaptic potentials in anterior thalamic cells are differentially altered by stimulation of laterodorsal tegmental cholinergic nucleus, Neuroscience 47, 279–289, 1992. 74. Steriade, M., McCormick, D.A., and Sejnowski, T.J., Thalamocortical oscillation in the sleeping and aroused brain, Science 262, 679–685, 1993. 75. Lytton, W.W. et al., Dynamic interactions determine partial thalamic quiescence in a computer network model of spike-and-wave seizures, J. Neurophysiol. 77, 1679– 1696, 1997. 76. Steriade, M., Corticothalamic resonance, states of vigilance, and mentation. Neuroscience 101, 243–276, 2000. 77. Steriade, M., Contreras, D., and Amzica, F., Synchronized sleep oscillations and their paroxysmal developments, Trends Neurosci. 17, 199–208, 1994. 78. Steriade, M., Local gating of information processing through the thalamus, Neuron 41, 493–494, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 277 Wednesday, February 15, 2006 3:25 PM

Features 13 Modeling of Autism in Animals Paul H. Patterson CONTENTS Introduction............................................................................................................277 Genetic Manipulation ............................................................................................279 X Chromosome Loci ......................................................................................279 15q11-q13 Locus ............................................................................................280 Oxytocin and Vasopressin ..............................................................................280 Serotonin.........................................................................................................281 DLX ................................................................................................................283 Engrailed.........................................................................................................284 Acetylcholine Receptor ..................................................................................284 Dishevelled .....................................................................................................285 µ–Opioid Receptor .........................................................................................285 Deer Mouse ....................................................................................................285 Reelin..............................................................................................................286 Environmental Manipulation .................................................................................286 Thalidomide and Valproic Acid .....................................................................286 Maternal Infection ..........................................................................................288 Postnatal Viral Infection .................................................................................290 Postnatal Vaccination......................................................................................290 Lesion.....................................................................................................................291 Amygdala........................................................................................................291 Cerebellum......................................................................................................292 Perspectives............................................................................................................293 Acknowledgments..................................................................................................293 References..............................................................................................................293

INTRODUCTION Animal models of many neurological diseases (Alzheimer’s, Parkinson’s, Huntington’s, multiple sclerosis) have proven enormously useful for determining the roles of genes and environment, understanding pathogenesis, and testing potential therapeutic approaches. There is some skepticism, however, concerning models of psychiatric or mental illnesses (e.g., schizophrenia, depression, autism). After all, can 277

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 278 Wednesday, February 15, 2006 3:25 PM

278

Understanding Autism: From Basic Neuroscience to Treatment

cognitive abnormalities or language deficits be detected in animals? To give up this approach, however, would deny the application of powerful genetic and molecular tools to critical illnesses. Moreover, animal models need not be perfect mimics of human diseases in order to be valuable. For instance, the most commonly used genetic mouse models of Huntington’s disease (HD) do not exhibit the severe loss of striatal neurons that is the signature pathology of the disorder. Nonetheless, the HD models are essential for probing how, for instance, mutant huntingtin protein influences HD-related behavior and how huntingtin aggregates in vivo. In fact, despite their lack of characteristic striatal pathology, these mice are currently being used to screen small molecules for use in HD patients. The power of animal models is in the examination of particularly interesting features of a disease, without having to replicate all aspects of the pathophysiology. Thus, it is important not to hold potential animal models of autism, for instance, to a higher standard than models of other human disorders. The relevance of an animal model to a disease should be judged by how well it reflects one or more features of that disease, which may include genetics, pathology, behavior, etiology, electrophysiology, or molecular changes. Skeptics can argue that autism is a particularly difficult case for animal studies because it has a heterogeneous behavioral phenotype, the susceptibility genes have not been firmly identified, and it does not have a pathognomonic histology that allows definitive diagnosis. Nonetheless, autism does have generally agreed upon neuropathological features that are distinctive [1], such as loss of Purkinje cells (PCs) in a specific region of the cerebellum. In addition, a number of molecular changes have been described for autism, including decreased hippocampal γ−aminobutyric acid (GABAA) receptors, and elevated levels of brain derived neurotrophic factor (BDNF) and platelet serotonin (5-HT) [2]. There is also evidence for immune system abnormalities in autism [3,4]. Moreover, some of the striking behavioral features of autism [2] can be assayed in animals, such as neophobia, abnormal social interactions, stereotyped and repetitive motor behaviors, enhanced anxiety, abnormal pain sensitivity, disturbed sleep patterns, deficient maternal bonding and affiliation, and a deficit in sensorimotor gating (prepulse inhibition, PPI). Although autism has a strong genetic basis, it is not a monogenic disorder and thus, it is not possible to establish an immediately relevant genetic mouse model as was done with HD. It is estimated that at least 10 genes may be associated with increased susceptibility to autism, each contributing a small amount to the overall risk, and it has been difficult to firmly identify such genes [5,6]. Nonetheless, as discussed later, there are several genetic changes that do entail an elevated risk for autism, and mouse models of these changes share some features with the human disorder. There are also mouse mutants of various types that display behavior or neuropathology relevant to autism. In addition, models based on autism etiology are valuable and there are several known environmental risk factors that are being successfully modeled in rodents. Finally, there are brain lesion models of interest. Therefore, even at this early stage of analysis, it is clear that various models can be used to study how a particular gene influences certain autism endophenotypes (PPI, maternal affiliation, cerebellum development, expression of neurotrophic factors, etc.) and how known environmental risk factors influence such endophenotypes.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 279 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

279

It will also be interesting to determine how a particular genotype influences the response to an environmental risk factor and vice versa. This review discusses proposed genetic, environmental risk factor, and lesion models. Several other authors have reviewed various aspects of animal models related to autism [7–10].

GENETIC MANIPULATION X CHROMOSOME LOCI As autism is much more frequent in boys than girls (a male to female ratio of 4:1), the X chromosome is a logical place to start with genetic models, and 4 loci have been implicated in autism thus far. The corresponding genes of interest are Fragile X mental retardation (Fmr1), which is silenced in Fragile X syndrome (FXS), MECP2, mutations which cause Rett syndrome, and neuroligin (NLGN) 3 and 4. FXS involves mental retardation, attention deficit, anxiety, seizures, stereotypic behaviors, communication deficits, and macroorchidism, and it is estimated that 25 to 40% of FXS cases meet the diagnostic criteria for autism. Conversely, about 2% of autism cases involve FXS [11]. The FMR1 protein binds as many as 80 mRNAs, and when Fmr1 is deleted, the expression of many of these genes is altered [12]. Approximately 15 of these RNAs map to suspected autism susceptibility loci. Some of these RNAs may be related to the protein-synthesis-dependent functions of metabotropic glutamate receptors [13]. The Fmr1 knockout (KO) mouse displays several FXS features, including macroorchidism and a subtle learning deficit [14–17]. In terms of neuropathology, this mouse has an increased dendritic spine density, with a greater number of spines with an immature appearance in the visual and somatosensory cortices [18,19]. Human FXS studies have similarly suggested dendritic spines with an immature morphology, as well as increased spine density in several cortical areas [20]. The situation in the human hippocampus in autism may be different, however, where there appears to be a decrease in hippocampal dendritic branching [21]. Clearly, more neuropathology as well as much more behavioral analysis is required to relate this mouse model to autism. The NLGN1, 3, and 4 genes map to three loci associated with predisposition to autism (3q26, Xp22.3, and Xq13, respectively). Moreover, mutations in NLGN3 and 4 are associated with autism and, in some cases, mental retardation [22,23]. These mutations lead to loss of protein processing and loss of the capacity for stimulation of synapse formation [24,25]. Given the likely role of the NLGNs in the development of connectivity, it will be very interesting to examine the behavioral and histological phenotypes of mouse models expressing these human mutations. It will also be important to determine the full behavioral phenotypes of the neurexin KOs, as the neurexins are NLGN binding partners [26]. Although NLGN mutations account for a very small fraction of autism cases, study of these proteins may open the door to an important pathway. This has been the case for study of the similarly rare mutations that cause Parkinson’s or Alzheimer’s diseases. Mutations in the Rett syndrome gene, MECP2, are also found among autism subjects, and Rett patients can display symptoms of autism [27]. Similar to FMR1, MECP2 regulates gene expression, including BDNF, and MECP2 may be a link

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 280 Wednesday, February 15, 2006 3:25 PM

280

Understanding Autism: From Basic Neuroscience to Treatment

between synaptic activity and transcription [28]. In one mouse model, a truncated form of MECP2 leads to a progressive, Rett-like phenotype starting at about 6 weeks, which includes increased anxiety-like behavior, seizures, forelimb stereotypies, tremors, and kyphosis [29]. The first three of these symptoms are similar to those in autism, but the abnormalities in this mouse are clearly much broader than that found in autism. Detailed neuropathology remains to be reported.

15Q11-Q13 LOCUS This is a relatively small locus that has been linked to a subset of autism cases. Moreover, Angelman syndrome (AS), which shares some clinical features with autism, including absence of speech, seizures, and disturbed sleep patterns, primarily arises from deletions of this region on the maternal chromosome. One gene within this region, Ube3a, has been identified as a genetic locus for AS in a small minority of cases [30]. This gene is expressed only in the brain and codes for an E6-AP ubiquitin ligase that targets four known proteins for degradation. A mouse model with a maternal null mutation displays the characteristic localized lack of Ube3a expression in PCs and in the hippocampus [31], sites of particular interest in autism pathology. These mice share with AS patients seizures, motor dysfunction, and deficiencies in contextual learning [31, 32]. It is not clear, however, how their specific physical and motor abnormalities relate to autism, although early motor abnormalities are common in autism [2]. The deficits in learning and long-term potentiation [31] are relevant to autism, as mental retardation is observed in the majority of autism cases [33]. It will be important to test these mice in a variety of other behavioral assays with relevance to autism. The 15q11-q13 locus also contains the genes for a number of subunits of the GABA receptor. The levels of GABAA receptor binding are low in autism [34], and mice lacking the β3 subunit of this receptor have several features of AS, including dysmorphic facial features, learning deficits, motor abnormalities, seizures, hyperactivity, and disturbed sleep patterns [36]. As with the Ube3a mouse, more behavioral analysis as well as neuropathology is required to establish the relationship of the GABAA receptor KO mouse to autism. It will also be of interest to examine mice with duplications or KOs in other loci within 15q11-q13 as other genes of interest for autism may be found there.

OXYTOCIN

AND

VASOPRESSIN

These neuropeptides influence a number of social behaviors relevant to autism, including social recognition, affiliation, and aggression. Animal studies show that these neuropeptides also influence the frequency of repetitive behaviors, which are common in autism. Thus, these peptides have strong potential for roles in autism [37,38]. It is therefore important that plasma oxytocin (OT) levels were reported to be reduced in autism, whereas a precursor form is elevated [39,40]. Consistent with these observations is the finding that peripheral infusion of OT in adult patients with autism reduced repetitive behaviors, although a relatively small number of subjects were studied [41]. In terms of potential animal models, the male OT KO mouse fails to recognize familiar conspecifics even after repeated encounters. This is not a

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 281 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

281

general sensory or learning deficit, because these mice are normal in the hidden food location test, the Morris water maze spatial memory test, and they habituate normally in the acoustic startle test [42]. The relevance of social amnesia to autism is not entirely clear, although it is sometimes said that autistic children do not demonstrate normal affiliative behavior towards their parents and siblings (however, this is strongly disputed; see Reference 43). Is this putative lack of affiliative behavior due to a lack of recognition? There is evidence that facial recognition pathways in the brain are abnormal in autism. It is therefore of interest that the OT KO mice also display altered neural pathways in processing social information. Following a 90-second encounter, WT and OT KO mice display divergent patterns of Fos expression in the brain [44]. It will be important to assess other relevant behaviors in this intriguing model, including stereotypy, PPI, and novel object interaction and to assess relevant neuropathology. It should be kept in mind, however, that autism does not involve a KO of the OT gene; thus study OT heterozygous mice is also merited. There are also important correlations with OT levels in primate models of abnormal, autism-relevant behaviors. Monkeys raised from birth in isolation or in nurseries with human caretakers display severe deficits in social interaction and communication and also engage in repetitive stereotypies and self-stimulation [38]. Interestingly, the nursery-raised monkeys have decreased CSF OT, and the levels are positively correlated with affiliative behavior. There are also differences in brain corticotrophin-releasing hormone and arginine vasopressin (AVP) receptor levels (see following text) [38]. The peptide AVP may also be relevant for autism, because a polymorphism in the AVP 1a receptor (V1aR) was linked to autism using the multiallelic transmission and disequilibrium test (but this was not significant after Bonferroni correction) [45]. Acute, central administration of AVP in mice stimulates stereotyped scratching and autogrooming. In terms of potential animal models, the V1aR KO male mouse exhibits reduced anxiety behavior in the open field, in light and dark boxes, and in elevated plus maze tests, which is not consistent with autism. Moreover, no deficit in PPI was detected [46]. On the other hand, its reduced social recognition in the olfactory habituation test is consistent with autism, as is the finding that infant OT KO mice are less vocal than WTs when separated from their mothers [38]. Another model is the Brattleboro rat, which has a mutation that impairs AVP synthesis. This strain has deficits in memory, emotional reactivity, motivation, attention, and social recognition compared to Long Evans rat controls. A similar deficit in social recognition was observed after V1R antagonist treatment, whereas administration of AVP improved recognition in both Brattelboro and Long Evans rats [47]. In addition to these behaviors with relevance to autism, Brattelboro rats have a PPI deficit [48]. Further pathological studies of the brains of these rats is merited.

SEROTONIN It is generally agreed that there are 5-HT abnormalities in autism. Platelet 5-HT levels are elevated 25 to 50% [49,50], although there is no apparent correlation between platelet 5-HT levels and any of a variety of behavioral tests done on autistic subjects [51]. Unfortunately, the mechanism underlying this elevation is

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 282 Wednesday, February 15, 2006 3:25 PM

282

Understanding Autism: From Basic Neuroscience to Treatment

not understood, thus the relevance of this peripheral change for the brain is presently unclear. There are, however, 5-HT abnormalities in the autistic brain, including asymmetrical changes in 5-HT synthesis that correlate with functional impairment [52–54]. Children with autism also do not show the expected changes in brain 5-HT synthesis with age [53]. These abnormalities could have strong effects on mood, aggression, sleep, and social behavior, because changes in this transmitter are associated with altered behaviors in each of these domains in animals. Pharmacological treatment of autism by perturbation of the 5-HT system is also informative; lowering 5-HT levels by tryptophan depletion can exacerbate symptoms, whereas elevating 5-HT levels by treatment with 5-HT uptake inhibitors can alleviate some symptoms of autism [c.f., 55]. There is a continuing search for autism susceptibility genes associated with 5-HT neurotransmission [2,5]. It is clear that 5-HT is important during fetal brain development, as perturbation of its levels by pharmacological methods causes changes in neuron numbers, maturation of dendrites, organization of the cortex, and myelination [56,57]. Therefore, altering 5-HT levels during development could lead to very relevant autism models. Pharmacological depletion of 5-HT in neonatal rat pups results in decreased dendritic length and spine density in the hippocampus [58], and a decrease in hippocampal dendritic arbors has been observed in autism [59]. Pharmacological depletion of 5-HT in pregnant rats alters the period of mitosis in fetal brain regions with high 5-HT innervation [60]. In the converse experiment, treating pregnant rats with the nonspecific 5-HT agonist 5-methoxytryptamine from embryonic day 12 (E12) through postnatal day 20 (P20) results in a number of behavioral abnormalities. These include overreaction to auditory or tactile stimuli, lack of separation-induced vocalizations in pups removed from the mother, and decreased alternation in the spontaneous alternation task [61]. Such behavioral changes are quite relevant to autism. As assayed by proton magnetic resonance spectroscopy, these offspring also display brain metabolic abnormalities that are consistent with changes seen in autism. In addition, such offspring have alterations in reelin expression [62] and in cortical column development, each of which has been demonstrated in autism brains [63,64]. There is, however, no genetic link between reelin and autism at this point. Neonatal disruption of 5-HT tracts also causes lasting changes in behavior, as well as sexspecific alterations in cortical morphogenesis [65]. One concern with these drug paradigms is the considerable stress induced by repeated injections. Although control animals were injected with saline for comparison, the interaction of 5-HT and behavioral stress may be a confounding variable. To circumvent these issues, manipulations of various genes in the 5-HT neurotransmission system can be utilized, altering expression in region- and stage-specific ways. In fact, many such manipulations have been done and have confirmed the critical importance of 5-HT in both early embryonic and postnatal development [57,66]. For instance, KO experiments have shown that the 5-HT2B receptor regulates neurogenesis, cell specification, and cell survival in early CNS development. At later developmental stages, such control can be mediated by the 5-HT1A and 5-HT2A/2C receptors, depending on the specific brain region. Moreover, 5-HT1A receptors regulate dendritic growth whereas 5-HT1B receptors regulate axonal growth. Many striking behavioral phenotypes also result from KOs of various genes in 5-HT transmission. Knocking

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 283 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

283

out MAOA, which is a prime 5-HT breakdown enzyme (although it also degrades other monoamines), causes increased aggression and reduced exploration — at least some of which is due to influences on early postnatal development. In addition, various 5-HT receptor KO mice exhibit increased frequency of seizures, eating disorders, altered response to stress, reduced sensitivity to nociceptive stimuli, increased anxiety, and altered sleep patterns, all of which are very relevant to autism [57,67]. What is missing in this fascinating and growing literature of genetic manipulation is a detailed analysis of neuropathology, particularly with reference to known changes in the autistic brain. There is also an intriguing connection between autism, 5-HT, and cholesterol. The Smith–Lemli–Opitz syndrome (SLOS) results from mutations in a key enzyme in cholesterol synthesis, DHCR7. In addition to many structural abnormalities and physiological deficits, approximately 50% of these patients meet the criteria for autism spectrum disorders [68]. Moreover, there is a partial to complete agenesis in the cerebellar vermis and corpus callosum [69]. Curiously, neonatal mice lacking DHCR7 activity display a large increase in the area of 5-HT immunoreactivity in the hindbrain as well as an expansion of the regional distribution of this staining [70]. Unfortunately, these homozygote mice die very early, so it will be necessary to examine the heterozygotes for behavior. As discussed in the following text, several drugs known to increase the incidence of autism also modulate the extent of 5-HT neuron distribution.

DLX The homeobox-containing DLX gene complexes are of interest because they regulate the development of subsets of cortical and striatal neurons, and two of the linkage loci for autism, 2q31.1 and 7q21.3, contain the DLX1/2 and DLX5/6 complexes, respectively. Although previous studies failed to identify DLX mutations in autism probands, recent sequencing of exons, exon/intron boundaries, and known enhancers of DLX1, 2, 5, and 6 identified three nonsynonymous mutations in DLX2 and two in DLX5 in ~3% of autistic probands [J.L.R. Rubenstein, personal communication]. Such mutations were either very rare or not present in the control population, suggesting that DLX mutations could contribute to autism susceptibility. It is also of interest that DLX genes regulate the expression of the X-linked ARX gene in basal ganglia progenitor cell populations [I. Cobos and J.L.R. Rubenstein, personal communication], and a subset of patients with a mutation in ARX have autistic features [71,72]. Although it is not yet known how the DLX mutations linked to autism affect transcription, the predicted nonconservative amino acid changes lie in highly conserved regions of the proteins. Thus, it is of interest to consider the phenotypes of the relevant DLX KO mice. Mice lacking DLX1 display seizures [I. Cobos and J.L.R. Rubenstein, personal communication], and mutations in DLX2, 5, or ARX alter the development of forebrain GABA neurons [73]. As mentioned earlier, the GABA system is altered in autism, although not to the extent seen in DLX mutants thus far. All of the autistic individuals identified as having DLX mutations are heterozygous for these mutations, thus it will be important to study mice heterozygous for these SNPs.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 284 Wednesday, February 15, 2006 3:25 PM

284

Understanding Autism: From Basic Neuroscience to Treatment

ENGRAILED The gene Engrailed 2 (En2) maps to a location on chromosome 7 that has been linked to autism, and an association has been reported between two intronic SNPs of En2 and autism [74,75]. This association has been recently extended in two additional sets of families [5]. Interestingly, both transgenic misexpression and KOs of En2 result in cerebellar phenotypes that are reminiscent of abnormalities found in autism. As in autism [76], both types of mutants display reduced numbers of PCs [77]. The loss of PCs is thought to be generally distributed throughout the cerebellum in the En2 KO, whereas the loss appears in stripes or patches in the transgenic mouse, despite the fact that En2 overexpression is driven by a general PC promoter. The latter distribution of PC loss has been observed in some autism cases [78]. Thus, it is interesting that endogenous En2 is expressed by all primary cerebellar cell types in spatially restricted stripes from E17.5 to P4. Other features shared between the En2 KO and at least some human autism cases are deficiencies in the number of deep nuclear, granule, and inferior olive neurons. The relationship to autism of the cerebellar foliation abnormalities in the mouse mutant is less clear. A recent report links changes in other En2 KO brain regions to autism; increased neuronal packing and ectopic location of neuronal subgroups in the amygdala, as well as a smaller hippocampus are seen [80]. One paper on the behavior of En2 KO mice shows that these mice display a motor and balance deficit on the Rotarod and are normal in the open field test [180]. This model is definitely worth further examination, and mice expressing the En2 polymorphisms associated with autism should be examined.

ACETYLCHOLINE RECEPTOR There are several connections between acetylcholine (ACh) and autism. This transmitter is important for sustained attention, which may be important in the PPI deficit in autism. Importantly, the mRNA and protein levels of the α4β2 nicotinic ACh receptor as well as ligand binding to this receptor are reduced in the autistic cerebral cortex, and there are posttranscriptional abnormalities in this receptor in the cerebellum [81–83]. In addition, there is one paper indicating efficacy of an ACh esterase inhibitor in treating some symptoms of autism [84]. These findings raise the question of whether the α4 KO mouse is relevant to autism. This mouse exhibits reduced sensitivity to nicotine in several nociception assays and enhanced anxiety in several tests [85, 86]. It is still a question, however, whether the receptor changes observed in autism are due to cell loss or to changes in the transmitter system. Moreover, it is not clear whether the changes in the ACh receptor observed in the adult autistic brain are present early on in the disorder, or are manifested later as a response to other abnormalities. This caveat applies to many pathological findings, of course. At present, therefore, the ACh receptor KO mouse serves as a possible model for behavioral changes due to loss of this type of nicotinic receptor throughout the brain, but its direct relevance to autism is uncertain.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 285 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

285

DISHEVELLED This gene family (Dvl) is a component of the Wnt pathway, which is involved in cell fate determination. The interest in Dvl1 in the context of autism results from the behavioral phenotype of the KO mice. Although the brains of these mice are structurally normal at a gross level, they exhibit deficits in social interaction (whisker trimming, nest building, and huddling). In contrast, these mice are normal in three memory tasks, two of which require a social component [87]. Unfortunately, an earlier report of a deficit in PPI was not reproducible in the Dvl1 KO [87]. There are no reports of linkage of this gene to autism, nor are detailed descriptions of the neuropathology of the KO available as yet. It is of interest that mutations in Wnt2, which could be in the same signaling pathway as Dvl1, have been found in two families with autism [11], although this finding was not replicated in a larger study [88].

–OPIOID RECEPTOR Attachment or affiliative behavior connotes a social bond that can be measured by the selective approach to, and interaction with, particular individuals, and by the display of distress during acute separation from those individuals. The latter is quantified in rodents by measuring ultrasonic vocalizations, typically in pups separated from their mothers. Because such calls are strongly reduced in pups given µ−opioid agonists, endogenous opioids have been implicated in infant attachment to their mothers, and malfunctioning of the opioid system is suggested to be involved in the social indifference displayed by autistic infants [89,90; but see 43]. In fact, pups lacking a functional µ–opioid receptor gene display a striking deficit in ultrasonic vocalizations when separated from their mothers but do not differ from WT in their vocalizations in response to cold or male cues, suggesting a specific lack of affiliative behavior [91]. Moreover, the KO mice do not respond as specifically to their mother’s olfactory cues as do WT mice. Thus, these mice provide an attractive model in which to study the social reinforcement pathways that may be deficient in autism. It is also of interest that opioids are important for pain sensitivity, although abnormalities in this domain in autism is controversial [92].

DEER MOUSE Another striking feature of autism that can be mimicked in animals is the increased frequency of stereotyped behaviors. These are apparently purposeless movements that are repetitive and topologically invariant. Stereotypies have been well studied in pharmacological models where dopamine, glutamate, and opioid peptides are implicated. Housing conditions can also influence these behaviors [93]. Mouse models of Down syndrome (Ts65Dn) [94] and Rett syndrome [29] also exhibit stereotypies. Spontaneous stereotypies are found in the deer mouse, Peromyscus maniculatus, and these can be pharmacologically distinguished from the druginduced stereotypies [95]. These behaviors can be ameliorated by caging the deer mice in an enriched environment, which is also correlated with increased levels of

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 286 Wednesday, February 15, 2006 3:25 PM

286

Understanding Autism: From Basic Neuroscience to Treatment

BDNF in the striatum [96]. Further behavioral and neuropathological analysis of this model is needed.

REELIN Reelin is a critical signal in neuroblast migration. The protein is secreted by Cajal–Retzius cells and binds to the lipoprotein receptors VLSLR and ApoER2, as well as to the α3β1 integrin. The reelin gene is in the chromosome 7 linkage group for autism, but there are conflicting reports about the association of its polymorphisms with autism [reviewed in 5]. The group with a positive finding on reelin has also recently reported suggestive evidence of an association of APOE2 with autism [97]. This is potentially relevant because APOE can competitively inhibit reelin binding to its lipoprotein receptors. Another link to autism is the finding that reelin protein levels are reported to be lower in autistic cerebella compared to controls [62]. A decrease in blood levels of the 410 kD form of reelin was also reported in autistic twins and their first degree relatives [98]. However, no significant change was found in autistic cerebral spinal fluid samples [99]. Reelin is normally expressed by GABAergic neurons in the adult brain, but the relevance of its levels in the circulation is unclear. In the animal model, the reelin heterozygote was reported to display decreased GAD67 expression and reduced dendritic spine density [100]. Most interesting were the reported deficit in PPI and increased fearfulness in the elevated plus maze, with normal spontaneous motor activity [101,102]. However, further testing of the same strain of reelin heterozygote mice by other laboratories failed to confirm a PPI deficit [103,104]. Moreover, subjecting these mice to extensive batteries of behavioral tests relevant to autism and schizophrenia revealed no differences from WT littermates [105]. Another observation of interest in light of the male predominance of autism is the progressive loss of PCs in the male but not female reelin heterozygote [106]. Surprisingly, this progressive loss occurs in adulthood. Given the uncertain genetic linkage with autism, the mixed results with reelin levels in autism, and the lack of solid, relevant behavioral results in the mouse, it is clear that much more data will be required to prove the relevance of reelin in autism.

ENVIRONMENTAL MANIPULATION THALIDOMIDE

AND

VALPROIC ACID

The use of these very different drugs in humans has provided important insights into autism and has led to useful animal models as well. Before it was recognized as teratogenic, thalidomide was used to treat morning sickness. In addition to causing specific abnormalities in craniofacial and limb structures in the offspring, thalidomide led to a marked increase in the incidence of autism [107]. As the particular dysmorphologies induced by thalidomide depend on the precise time of drug ingestion, it is possible to specify the window of vulnerability for autism by studying the external features of thalidomide-exposed offspring with autism. Postconception days 20 to 24 were identified as the critical time period. During this time the neural tube

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 287 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

287

is closing and very few neurons have been born. Those that have been born include the cranial motor nuclei, which innervate the eyes and face. Importantly, craniofacial structural dysmorphologies as well as abnormalities in facial expression consistent with thalidomide exposure have been observed in idiopathic autism [108]. Moreover, other insults of very early embryonic development, such as maternal ingestion of valproic acid (VPA) (see the following text) or misoprostol, also lead to greatly increased risk for autism in the offspring [107]. Unfortunately, thalidomide is not teratogenic in rodents, so the external features in the offspring of drug-injected rodent mothers are not informative. Nonetheless, several brief papers on this model are provocative. Daily maternal injection of rats on E7-18 yields adult offspring displaying increased errors and latency in the multiple-T Cincinnati water maze; interestingly, this deficit is apparent only in males [109]. An issue with this protocol is that the timing of thalidomide administration does not closely mimic the human situation. An impairment in the radial maze but not the Morris water maze test was also reported, although a sexual dimorphism was not mentioned [110]. Exposure to thalidomide on E9, just prior to neural tube closure in rats, yields adult offspring with increased 5-HT in plasma, hippocampus, and frontal cortex, as well as an altered distribution of 5-HT neurons in the raphe nuclei [111,112]. The hyperserotonemia is particularly intriguing in light of the human findings cited earlier. It has been proposed that the thalidomide model should be followed up in primates, in which the drug would be teratogenic and it is possible to assay more human-like autism behaviors [113]. VPA is used widely in humans as a mood stabilizer and anticonvulsant. The offspring of women taking the drug during early pregnancy have a greatly increased risk for autism and for neural tube defects such as spina bifida, and similar teratogenic effects have been observed in certain strains of mice [108]. Moreover, as in autism, PC number is reduced by prenatal VPA exposure at E12.5 in rats, as is the number of neurons in the cranial nerve motor nuclei [108,114,115]. In a preliminary report, the neurons in the inferior olive that innervate PCs are also reduced by VPA exposure, as are the deep nuclei targets of PCs in the nucleus interpositus [116]. These deficits are relevant to human autism autopsy findings [1]. As with thalidomide, VPA exposure on E9 causes increased levels of 5-HT in the rat hippocampus, frontal cortex, and cerebellum [111]. In behavioral studies, a single prenatal VPA exposure on E9 leads to a deficit in radial maze learning [110], and exposure on E12.5 leads to lower sensitivity to painful stimuli and higher sensitivity to nonpainful stimuli, diminished PPI, locomotor and repetitive/stereotypic-like hyperactivity combined with lower exploratory activity, and delayed nest-seeking responses [117]. There is also evidence for a deficit in social memory and novel object exploration [118]. All of these behavioral changes are reminiscent of autism. Moreover, in an eye blink conditioning paradigm, VPA-exposed offspring display more rapid acquisition, and stronger and faster blinks, as is also seen in human autism [108]. It is striking that performance in this task is better than in controls, given that it is mediated through a brain stem–cerebellar circuit, and there is pathology in these structures in autism and in the animal model. Thus, there are many remarkable parallels between the VPA model and the human disorder.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 288 Wednesday, February 15, 2006 3:25 PM

288

Understanding Autism: From Basic Neuroscience to Treatment

How might VPA be causing these changes? A most interesting finding for the developing brain is the observation that chronic VPA treatment enhances neurogenesis in the hippocampus [119]. The concentration of GABA is also increased by VPA, and expression of neurotrophic factors and cytokines is altered as well [120,121]. Gene expression is altered by VPA through inhibition of histone deacetylase [119]. It is intriguing that VPA can alter HOX gene expression [122,123], as the HOXa1 KO mouse lacks most of the facial nucleus and all of the superior olive [124,125] — a finding similar to the abnormalities observed in thalidomide-exposed autism cases. HOXa1 is expressed exclusively during the period of neural tube closure, the time of vulnerability to thalidomide. These findings led to the question: Are HOX gene SNPs associated with human autism? Although positive linkages were reported for alleles of HOXA1 and HOXB1, this finding was not replicated in other populations [reviewed in 10]. The importance of the nonreplication is not clear, however, as different populations may be expected to display diverse sets of susceptibility genotypes.

MATERNAL INFECTION Another environmental factor known to increase the risk of autism is maternal infection. In fact, one review concluded, “the principal non-genetic cause of autism is prenatal viral infection” [126]. The most dramatic evidence for this conclusion came from a rubella epidemic in which the incidence of autism diagnosis in prenatally exposed offspring was more than 200-fold higher than normal [127]. Although there is no rodent model of rubella infection, respiratory infection with influenza virus has been extensively studied in mice, and considerable evidence links maternal respiratory infection during the second trimester (and associated induction of maternal antiviral antibodies and cytokines) with increased risk for schizophrenia in the offspring [128–130]. Influenza infection of pregnant mice at E9.5 yields adult offspring with histological and behavioral abnormalities reminiscent of autism and schizophrenia. For instance, the offspring display deficits in PPI, novel object, and open field exploration, as well as social interaction [131]. Brain pathology includes a selective loss of PCs in lobule VII of the cerebellum [132], as well as thinning of the neocortex and hippocampus, pyramidal cell atrophy, reduced levels of reelin immunoreactivity, changes in the expression of neuronal nitric oxide synthase and synaptosome associated protein of 25 kD, and macrocephaly [133–135]. Some of these findings closely mimic those in autism pathology. Current evidence suggests that the maternal immune response to the virus, rather than direct viral infection of the fetal brain, may be causative. Influenza is normally confined to the respiratory system, and no evidence of viral RNA was found in the fetus following intranasal infection of pregnant mice on E9.5 [136]. However, a conflicting report did provide evidence of the virus in fetal tissues [137]. Perhaps more important are results using surrogates for microbial infection such as poly(I:C) and lipopolysaccharide (LPS). Poly(I:C) is a synthetic double-stranded RNA that evokes an antiviral type of immune reaction in the absence of virus. Maternal injection of poly(I:C) on E9.5 in the mouse is sufficient to cause a PPI deficit in the adult offspring [131]. Injection on E9 in the rat causes reduced exploratory behavior,

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 289 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

289

deficient PPI, and reduced working memory performance in the Morris water maze [138]. Moreover, extensive study of offspring of rats exposed to poly(I:C) on E15 demonstrated a deficit in latent inhibition and an increased sensitivity to the locomotor effects of amphetamine, tests often used in the context of schizophrenia [139]. These offspring also display striking neuron loss and pyknosis in area CA1 of the hippocampus. It will be interesting to directly compare offspring exposed to poly(I:C) at different stages of embryonic development. A further test of the role of the maternal immune response in altering fetal brain development is the use of LPS, an extract of bacterial cell walls that can evoke an antibacterial type of immune response. In one such experiment, when rats were injected subcutaneously on alternate days with LPS throughout pregnancy, the offspring display PPI deficits as adults [140]. Moreover, the offspring have elevated serum levels of the cytokines IL-2 and IL-6, as well as histological evidence of ongoing brain inflammation as evidenced by astrocyte and microglial activation and MHC II induction. It is particularly striking that this inflammation continued in adulthood, given that the animals were never exposed to living microbes. Similarly striking evidence of ongoing brain inflammation has recently been shown in adult autism as well [141], suggesting some sort of subclinical, autocatalytic inflammatory cascade. However, the prolonged course of LPS administration used in the rat experiment does not mimic the timing of a normal uterine bacterial infection in humans. Injection of LPS only on E18 and 19 yields adult rat offspring with increased amphetamine-induced locomotion but no alteration in PPI [142]. There are many mechanisms through which the maternal immune response to infection could influence fetal brain development, such as altered body temperature, reduced food intake, elevated corticosteroid levels, etc. These and other features of the sickness response are known to be driven by peripheral cytokines as well as by cytokines within the brain itself [143]. Thus, it is possible that the maternal cytokine response to infection may mediate at least some of the changes in fetal brain development and behavior seen in the offspring [144]. Several proinflammatory cytokines are elevated in maternal serum and placenta following maternal injection of LPS, poly(I:C), or influenza infection [145,146; W. Xu, S. Smith, B. Deverman, P.H. Patterson, unpublished data]. Moreover, in human studies, elevated maternal TNFα and IL-8 are correlated with increased risk of psychosis in adult offspring [130,147]. Such cytokines could act on the brain directly or work through other mediators in the placenta, for instance. Another mechanism that has been suggested for mediating the effects of infection on fetal and postnatal brain development is production of antibodies by the mother. It is well known that autoantibodies are associated with several CNS diseases [148], and a subgroup of childhood onset obsessive-compulsive disorders and tic disorders have autoimmune-related etiology following streptococcal infection. Moreover, intravenous immunoglobulin and plasma exchange can reduce symptoms of these disorders [149]. In the case of autism, there are a number of reports of anti-brain antibodies in the serum of mothers as well as autistic children themselves [3,4]. The most provocative paper in this area described experiments using serum from a mother of multiple autism spectrum children [150]. Unlike control sera, this serum immunoglobulin fraction binds strongly to cerebellar PCs and certain brain stem neurons.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 290 Wednesday, February 15, 2006 3:25 PM

290

Understanding Autism: From Basic Neuroscience to Treatment

Moreover, when relatively large amounts of this serum were injected into pregnant mice, the adult offspring display significant deficits in open field exploration and a multiple static rod test [150]. Although this preliminary report requires confirmation, there are numerous reports, albeit often conflicting, of abnormalities in the immune system of autistic subjects [3,4,8,151]. Moreover, the frequency of autoimmune disorders and/or allergies is increased in families with autism, particularly in the mothers [152,153,181]. These various tantalizing connections between autism and the immune system need to be further explored in rigorous experiments.

POSTNATAL VIRAL INFECTION Considerable animal work has been done with postnatal viral infections. The rationale here is that neonatal development of parts of the rodent brain correspond to second or third trimester of human embryonic development, and about a third of autism cases are said to occur by regression in the second or third year. That is, for the latter cases, perhaps a postnatal insult is the “second hit” or precipitating factor. The most extensively studied model involves intracerebral injection of Borna disease virus (BDV) within 12 hours of birth. Although the significance of BDV infection for human psychiatric disorders is very controversial, several aspects of this model are relevant for autism [8,154,155]. Although a high virus load is maintained in the brain throughout life, the injected animals display a mild behavioral phenotype and restricted neuropathology. These mice have altered circadian rhythms, abnormal early locomotor development, spatial and aversive learning deficits, increased motor activity, abnormal anxiety, stereotypies, and reduced play interactions [154]. Alterations in 5-HT in various brain regions have been observed, and there is variable loss of PCs in the cerebellum and granule cells in the dentate gyrus [156]. A transient inflammatory response, which includes elevation of proinflammatory cytokines and chemokines for several weeks, follows infection, and proliferation of microglia is widespread. Whether the inflammation is related to the neuronal loss is not clear [154]. The gradual loss of dentate granule cells with BDV resembles that which is seen with neonatal infection with lymphocytic choriomeningitis virus (LCMV), although LCMV is cleared from the system before the cell loss occurs [8]. A fascinating aspect of this delayed loss of neurons is that it appears to involve the destruction or impairment of hippocampal neural stem cells [8]. Little behavioral data are available for the LCMV model thus far. Infection of 2-day-old rat pups with herpes simplex virus leads to long-term behavioral alterations, and cytomegalovirus (CMV) infection causes a PPI deficit when tested with apomorphine [8]. An association between CMV and autism has been suggested [157], although the histopathology in the CMV, LCMV, and BDV models does not closely resemble that seen in autism.

POSTNATAL VACCINATION Although many epidemiological studies have failed to substantiate a link between early childhood vaccination and autism [158,159], public interest in this issue continues. Although there is presently no reliable animal model of the effects of the

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 291 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

291

measles–mumps–rubella vaccination, there have been reports on the effects on rodent development of the mercury-containing preservative (thimerosal) included in some vaccines. Perhaps the most extensive study to date involved postnatal thimerosal injections in amounts and a time course designed to mimic standard childhood immunizations [160]. Striking effects were seen on behaviors involving anxiety and exploration and on hippocampal histology in the autoimmune-prone SJL/J but not the relatively autoimmune-resistant C57BL/6J and BALB/cJ mouse strains. The hippocampal pathology was said to resemble that found in Rett syndrome. Interestingly, the results observed in the thimerosal-injected mice were apparently identical to those obtained with mice injected with thimerosal-containing vaccines (diphtheria, tetanus, acellular pertussis (DTaP), and haemophilus influenza B (HiB) vaccines) [160]. The mouse strain specificity of this model is attractive and could lead to a better understanding of toxin effects on the immune system and the brain, but the epidemiological relevance to autism is not clear. Moreover, untreated SJL mice have striking behavioral abnormalities, and this strain displays a severe immunodeficiency syndrome, which raises questions about its relevance to studies on autism. It should also be emphasized that recent epidemiological studies have failed to support a link between thimerosal and autism [161,162].

LESION AMYGDALA The amygdala has been implicated in autism in several ways. Functional MRI has revealed that autistic subjects display less amygdala activation than controls in tasks involving judgments of emotional content of faces or eyes, as well as during social attribution tasks in which subjects were assessed for perception of human-like interactions among geometric shapes [163,164]. Autism also involves impairment in the recognition of facial expressions of fear and other emotions, as well as in the memory of faces, which are also characteristic of people with amygdala lesions [165]. These tests are, however, complicated by the reluctance of autistic subjects to make eye contact, and these subjects do not examine faces in the same way as controls [166]. Moreover, structural MRI and neuropathological studies of the amygdala in autism have yielded conflicting results [167,168]. Considerable work has been done with amygdala lesions on primates because these animals offer the great advantage of nearly direct extrapolation of behavioral results to humans. Early work with macaques reported many behavioral abnormalities relevant to autism following neonatal amygdala lesions, such as motor stereotypies, inactivity, lack of exploration, and less initiation of social contact [169,170]. The question has been raised, however, as to how much of this effect was due to raising the lesioned animals in nurseries in the absence of their mothers [167]. In fact, when infants were returned to their mothers directly after the lesion surgery, they developed a complete repertoire of species-typical social signals [171]. The few behavioral abnormalities that were observed in this case were interpreted as deficits in fear processing. On the other hand, neonatally lesioned monkeys returned

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 292 Wednesday, February 15, 2006 3:25 PM

292

Understanding Autism: From Basic Neuroscience to Treatment

to their mothers show social phobia and do not initiate play [172]. Importantly, adult humans with amygdala lesions display normal, reciprocal social interactions and do not show symptoms of autism [167]. The results of ibotenic acid amygdala lesions in rats provide interesting support for at least some of the findings in primates. Early lesion (P7) results in a significant deficit in juvenile play and adult social behavior. These effects are more pronounced than those seen following later (P21) lesions [173]. Recent work from the same group described less pronounced effects of P7 lesions on adult social behavior, but also an interesting synergy of amygdala lesion with juvenile isolation. That is, moving lesioned rats to isolated housing (one rat per cage) after weaning resulted in significant deficits in social behaviors that were not apparent with either experimental perturbation alone [174]. Isolation rearing alters dopamine and serotonin levels in the amygdala and other brain areas [175]. These results raise the possibility that deficits in early brain development that lead to abnormal social interactions could predispose individuals towards social isolation, thus further exacerbating the problem. It will be important to assay a variety of other behaviors in this model and determine the changes in brain pathology and biochemistry. It will also be useful to examine how the rodent lesions may differ from those in primates.

CEREBELLUM Genetic, surgical, or toxin lesions of the cerebellum are of particular interest because of the consistent cerebellar pathology in autism. A comprehensive review has described numerous mutations and toxic insults that lead to diverse patterns of PC loss in rodents [176]. Some of these insults highlight the distinctive nature of the PCs that are selectively vulnerable in lobules VI and VII in autism. For instance, these PCs are less vulnerable to death in the shaker rat, possibly because they selectively express (along with PCs in lobules IX and X) the neuroprotective protein HSP27/25. A similar selective preservation of PCs is seen the mouse models of Niemann–Pick disease type A/B and C [176]. It is clear from the variety of patterns of PC loss seen in various mutants and toxic insults that the selective cell death is not simply due to a PC subset being more or less consistently vulnerable. Rather, the patterned cell loss in each case is specific to the insult. Thus, the selective loss of PCs in lobule VII in the influenza model [132] is an important parallel with autism. A rat model of early (P10) midline cerebellar lesion that bilaterally removed the vermis and fastigial nuclei was behaviorally characterized [177]. As adults, these animals display elevated spontaneous motor activity and exhibit evidence of perseveration and lack of attention to environmental distractors. Unlike the autism phenotype, however, the lesioned animals are neophilic and appear to be less anxious. A lack of attention to environmental distractors is also found in a spontaneous mutant line of Peruvian guinea pigs, GS, which had an almost complete absence of lobules VI and VII. These animals showed a total lack of exploratory behavior in a novel environment and displayed less social interaction than a control strain of guinea pigs [178,179]. It is not clear, however, what the appropriate control strain should be and, unfortunately, the GS line has been lost.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 293 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

293

PERSPECTIVES Despite the fact that we are still very early in the study of features of autism in animals, striking parallels with the human disorder have already been reported. Examples include the adult brain inflammation and selective loss of PCs in the maternal infection models, and the numerous autism-like behavioral changes, neuropathology, and 5-HT abnormalities in the VPA model. The genetic models of OT and µ–opioid receptor deficiencies also involve intriguing behavioral parallels with autism, with the OT model receiving additional support from studies on nonhuman primates raised under deprived conditions. The recent linkages of NLGN, DLX, and En2 to autism also offer attractive possibilities for animal models, particularly if introducing the relevant, specific mutations (as opposed to simple KOs) can cause interesting pathology and behavior. As these are susceptibility genes, it will also be of interest to determine how these genes may affect results obtained with environmental manipulations such as VPA treatment and infection. The potential for this type of experiment is illustrated by the striking specificity of the thimerosal effect for the autoimmune mouse strain.

ACKNOWLEDGMENTS Research quoted from the author’s laboratory was supported by Ginger and Ted Jenkins, the Mettler Autism Fund, the Stanley Medical Research Institute, a McKnight Foundation Neuroscience of Brain Disorder Award, the National Institute of Mental Health, and the Cure Autism Now foundation. Benjamin Deverman, Natalia Malkova, Stephen Smith, and an anonymous reviewer provided useful comments on a draft of the manuscript, and Kathleen Hamilton helped with editing.

REFERENCES 1. Bauman, M.L. and Kemper, T.L., Neuroanatomic observations of the brain in autism: a review and future directions, Int. J. Dev. Neurosci., 23, 183, 2005. 2. Belmonte, M.K. et al., Autism as a disorder of neural information processing: directions for research and targets for therapy, Mol. Psychiatry, 9, 1, 2004. 3. Hornig, M. and Lipkin, W.I., Infectious and immune factors in the pathogenesis of neurodevelopmental disorders: epidemiology, hypotheses, and animal models, Ment. Retard. Dev. Disabil. Res. Rev., 7, 200, 2001. 4. Ashwood, P. and Van de Water, J., A review of autism and the immune response, Clin. Dev. Immunol., 11, 165, 2004. 5. Bartlett, C.W., Gharani, N., Millonig, J.H., and Brzustowicz, L.M., Three autism candidate genes: a synthesis of human genetic analysis with other disciplines, Int. J. Dev. Neurosci., 23, 221, 2005. 6. Muhle, R., Trentacoste, S.V., and Rapin, I., The genetics of autism, Pediatrics, 113, 472, 2004. 7. Andres, C., Molecular genetics and animal models in autistic disorder, Brain Res. Bull., 57, 109, 2002.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 294 Wednesday, February 15, 2006 3:25 PM

294

Understanding Autism: From Basic Neuroscience to Treatment

8. Pearce, B.D., Modeling the role of infections in the etiology of mental illness, Clin. Neurosci. Res., 3, 271, 2003. 9. Keller F. and Perisco, A.M., The neurobiological context of autism, Mol. Neurobiol., 28, 1, 2003. 10. Murcia, C.L., Gulden, F., and Herrup, K., A question of balance: a proposal for new mouse models of autism, Int. J. Dev. Neurosci., 23, 265, 2005. 11. Wassink, T.H., Pivan, J., and Patil, S.R., Chromosomal abnormalities in a clinic sample of individuals with autistic disorder, Psychiatr. Genet., 11, 57, 2001. 12. Miyashiro, K.Y. et al., RNA cargoes associating with FMRP reveal deficits in cellular functioning in FMR1 null mice, Neuron, 37, 417, 2003. 13. Bear, M.F., Huber, K.M., and Warren, S.T., The mGluR theory of fragile X mental retardation, Trends Neurosci., 27, 369, 2004. 14. Kooy, R.F. et al., Transgenic mouse model for the fragile X syndrome, Am. J. Med. Genet., 64, 241, 1996. 15. D’ Hooge, R. et al., Mildly impaired water maze performance in male Fmr1 knockout mice, Neuroscience, 76, 367, 1997. 16. Oostra, B.A. and Hoogeveen, A.T., Animal model for fragile X syndrome, Ann. Med., 29, 563, 1997. 17. Peier, A.M. et al., (Over)correction of FMR1 deficiency with YAC transgenics: behavioral and physical features, Hum. Mol. Genet., 9, 1145, 2000. 18. Comery, T.A. et al., Abnormal dendritic spines in fragile X knockout mice: maturation and pruning deficits, Proc. Natl. Acad. Sci. USA, 94, 5401, 1997. 19. Galvez, R., Gopal, A.R., and Greenough, W.T., Somatosensory cortical barrel dendritic abnormalities in a mouse model of the fragile X mental retardation syndrome, Brain Res., 1, 83, 2003. 20. Irwin, S.A. et al., Abnormal dendritic spine characteristics in the temporal and visual cortices of patients with fragile-X syndrome: a quantitative examination, Am. J. Med. Genet., 98, 161, 2001. 21. Raymond, G.V., Bauman, M.L., and Kemper T.L., Hippocampus in autism: a Golgi analysis, Acta. Neuropathol., 91, 117, 1996. 22. Jamain, S. et al., Paris Autism Research International Sibpair Study, Mutations of the x-linked genes encoding neuroligins NLGN3 and NLGN4 are associated with autism, Nat. Genet., 34, 27, 2003. 23. Laumonnier, R. et al., X-linked mental retardation and autism are associated with a mutation in the NLGN4 gene, a member of the neuroligin family, Am. J. Hum. Genet., 74, 552, 2004. 24. Chih, B., Shehla, K.A., Clark, L., and Scheiffele, P., Disorder-associated mutations lead to functional inactivation of neuroligins, Hum. Mol. Genet., 13, 1474, 2004. 25. Comoletti, D. et al., The Arg451Cys-Neuroligin-3 mutation associated with autism reveals a defect in protein processing, J. Neurosci., 24, 4889, 2004. 26. Kattenstroth, G. et al., Postsynaptic N-methyl-D-aspartate receptor function requires α-neurexins, Proc. Natl. Acad. Sci. USA, 101, 2607, 2003. 27. Glaze, D.G., Rett syndrome: of girls and mice — lessons for regression in autism, Ment. Retard. Dev. Disabil. Res. Rev., 10, 154, 2004. 28. Neul, J.L. and Zoghbi, H.Y., Rett syndrome: a prototypical neurodevelopmental disorder, Neuroscientist, 10, 118, 2004. 29. Shahbazian, M. et al., Mice with truncated MeCP2 recapitulate many Rett syndrome features and display hyperacetylation of histone H3, Neuron, 35, 243, 2002. 30. Jiang, Y.H. and Beaudet, A.L., Human disorders of ubiquitination and proteasomal degradation, Curr. Opin. Pediatr., 16, 419, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 295 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

295

31. Jiang, Y.H. et al., Mutation of the Angelman ubiquitin ligase in mice causes increased cytoplasmic p53 and deficits of contextual learning and long-term potentiation, Neuron, 21, 799, 1998. 32. Weeber, E.J. and Sweatt, J.D., Molecular neurobiology of human cognition, Neuron, 33, 845, 2002. 33. Fombonne, E., Prevalence of childhood disintegrative disorder, Autism, 6, 149, 2002. 34. Blatt, G.J. et al., Density and distribution of hippocampal neurotransmitter receptors in autism: an autoradiographic study, J. Autism Dev. Disord., 31, 537, 2001. 35. Homanics, G.E. et al., Mice devoid of γ-aminobutyrate type A receptor β3 subunit have epilepsy, cleft palate, and hypersensitive behavior, Proc. Natl. Acad. Sci. USA, 94, 4143, 1997. 36. DeLorey, T.M. et al., Mice lacking the β3 subunit of the GABAA receptor have the epilepsy phenotype and many of the behavioral characteristics of Angelman syndrome, J. Neurosci., 18, 8505, 1996. 37. Lim, M.M., Bielsky, I.F., and Young, L.J., Neuropeptides and the social brain: potential rodent models of autism, Int. J. Dev. Neurosci., 23, 235, 2005. 38. Winslow, J.T., Neuropeptides and non-human primate social deficits associated with pathogenic rearing experience, Int. J. Dev. Neurosci., 23, 245, 2005. 39. Modahl, C. et al., Plasma oxytocin levels in autistic children, Biol. Psychiatry, 43, 270, 1998. 40. Green, L.A. et al., Oxytocin and autistic disorder: alterations in peptide forms, Biol. Psychiatry, 50, 609, 2001. 41. Hollander, E. et al., Oxytocin infusion reduces repetitive behaviors in adults with autistic and Asperger’s disorders, Neuropsychopharmacology, 28, 193, 2003. 42. Ferguson, J.N. et al., Social amnesia in mice lacking the oxytocin gene, Nat. Genet., 25, 284, 2000. 43. Gernsbacher, M.A. et al., Autism and deficits in attachment behavior, Science, 307, 1201, 2005. 44. Ferguson, J.N., Aldag, J.M., Insei, T.R., and Young, L.J., Oxytocin in the medial amygdala is essential for social recognition in the mouse, J. Neurosci., 21, 8278, 2001. 45. Kim, S.J. et al., Transmission disequilibrium testing of arginine vasopressin receptor 1A (AVPR1A) polymorphisms in autism, Mol. Psychiatry, 7, 503, 2002. 46. Bielsky, I.F. et al., Profound impairment in social recognition and reduction in anxietylike behavior in vasopressin VIa receptor knockout mice, Neuropharmacology, 29, 483, 2004. 47. Engelmann, M. and Landgraf, R., Microdialysis administration of vasopressin into the septum improves social recognition in Brattelboro rats, Physiol. Behav., 55, 145, 1994. 48. Feifel, D., Melendez, G., and Shilling, P.D., Reversal of sensimotor gating deficits in Brattelboro rats by acute administration of clozapine and a neurotensin agonist, but not haloperidol: a potential predictive model for novel antipsychotic effects, Neuropsychopharmacology, 29: 731, 2004. 49. Cook, E.H., Jr. and Leventhal, B.L., The serotonin system in autism, Curr. Opin. Pediatr., 4, 348, 1996. 50. Anderson, G.M. et al., Serotonin transporter promoter variants in autism: functional effects and relationship to platelet hyperserotonemia, Mol. Psychiatry, 7, 831, 2002. 51. Mulder., E.J. et al., Platelet serotonin levels in pervasive developmental disorders and mental retardation: diagnostic group differences, within-group distribution, and behavioral correlates, J. Am. Acad. Child Adolesc. Psychiatry, 43, 491, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 296 Wednesday, February 15, 2006 3:25 PM

296

Understanding Autism: From Basic Neuroscience to Treatment

52. Chugani, D.C. et al., Altered serotonin synthesis in the dentatothalamocortical pathway in autistic boys, Ann. Neurol., 42, 666, 1997. 53. Chugani, D.C. et al., Developmental changes in brain serotonin synthesis capacity in autistic and nonautistic children, Ann. Neurol., 45, 287, 1999. 54. Chandana, S.R. et al., Significance of abnormalities in developmental trajectory and asymmetry of cortical serotonin synthesis in autism, Int. J. Dev. Neurosci., 23, 171, 2005. 55. Hollander, E., Phillips, A.T., Chaplin, W., Zaqursky, K., Novotny, S., and Iyengar, R., A placebo controlled crossover trial of liquid fluoxetine on repetitive behaviors in childhood and adolescent autism, Neuropsychopharmacology, 30, 582, 2005. 56. Whitaker-Azmitia, P.W., Behavioral and cellular consequences of increasing serotonergic activity during brain development: a role in autism?, Int. J. Dev. Neurosci., 23, 75, 2005. 57. Gaspar, P., Cases, O., and Maroteaux, L., The developmental role of serotonin: news from mouse molecular genetics, Nat. Rev. Neurosci., 4, 1002, 2003. 58. Yan, W., Wilson, C.C., and Haring, J.H., Effects of neonatal serotonin depletion on the development of rat dentate granule cells, Dev. Brain Res., 98, 177, 1997. 59. Raymond, G.V., Bauman, M.L., and Kemper, T.L., Hippocampus in autism: a Golgi analysis, Acta. Neuropathol. (Berl), 91, 117, 1996. 60. Lauder, J.M., and Krebs, H., Serotonin as a differentiation signal in early embryogenesis, Dev. Neurosci., 1, 15, 1978. 61. Kahne, D. et al., Behavioral and magnetic resonance spectroscopic studies in the rat hyperserotonemic model of autism, Physiol. Behav., 75, 403, 2002. 62. Fatemi, S.H. et al., Dysregulation of Reelin and Bcl-2 proteins in autistic cerebellum, J. Autism Dev. Disord., 31, 529, 2001. 63. Casanova, M.F. et al., Minicolumnar pathology in autism, Neurology, 58, 428, 2002. 64. Casanova, M.F., White matter volume increase and minicolumns in autism, Ann. Neurol., 56, 453, 2004. 65. Connell, S., Karikari, C. and Hohmann, D.F., Sex-specific development of cortical monamine levels in mouse, Brain Res. Dev. Brain Res., 151, 187, 2004. 66. Gingrich, J.A. et al., New lessons from knockout mice: the role of serotonin during development and its possible contribution to the origins of neuropsychiatric disorders, CNS Spectr., 8, 572, 2003. 67. Luo, X.P.A.L.J., Serotonergic regulation of somatosensory cortical development: lessons from genetic mouse models, Dev. Neurosci., 25, 173, 2003. 68. Tierney, E. et al., Behavior phenotype in the RSH/Smith-Lemli-Opitz syndrome, Am. J. Med. Genet., 98, 191, 2001. 69. Kelley, R.I. and Hennekam, R.C., The Smith-Lemli-Opitz syndrome, J. Med. Genet., 37, 321, 2000. 70. Waage-Baudet, H. et al., Abnormal serotonergic development in a mouse model for the Smith-Lemli-Opitz syndrome: implications for autism, Int. J. Dev. Neurosci., 21, 451, 2003. 71. Turner, G. et al., Variable expression of mental retardation, autism, seizures, and dystonic hand movements in two families with an identical ARX gene mutation, Am. J. Med. Genet., 112, 405, 2002. 72. Stromme, P. et al., Infantile spasms, dystonia, and other X-linked phenotypes caused by mutations in Aristaless related homeobox gene, ARX, Brain Dev., 24, 266, 2002. 73. Stuhmer, T. et al., Ectopic expression of the Dlx genes induces glutamic acid decarboxylase and Dlx expression, Development, 129, 245, 2002.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 297 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

297

74. Petit, E. et al., Association study with two markers of a human homeogene in infantile autism, J. Med. Genet., 32, 269, 1995. 75. Gharani, N. et al., Association of the homeobox transcription factor, ENGRAILED 2, 3, with autism spectrum disorder, Mol. Psychiatry, 9, 474, 2004. 76. Courchesne, E., Redcay, E., and Kennedy, D.P., The autistic brain: birth through adulthood, Curr. Opin. Neurol., 17, 489, 2004. 77. Liu, A.M. and Joyner, A.L., Early anterior/posterior patterning of the midbrain and cerebellum, Annu. Rev. Neurosci., 24, 869, 2001. 78. Bailey, A. et al., A clinicopathological study of autism, Brain, 121, 889, 1998. 79. Kuemerle, B. et al., Pattern deformities and cell loss in Engrailed-2 mutant mice suggest two separate patterning events during cerebellar development, J. Neurosci., 17, 7881, 1997. 80. Kuemerle, B., Williams, E.A., and Herrup, K., The engrailed-2 mutant as a model of the neuropathology of autism, Program No. 116.9, Abstract Viewer/Itinerary Planner, Washington, D.C.: Society for Neuroscience, 2004. 81. Lee, M. et al., Nicotinic receptor abnormalities in the cerebellar cortex in autism, Brain, 125, 1483 2002. 82. Perry, E.K. et al., Cholinergic activity in autism: abnormalities in the cerebral cortex and basal forebrain, Am. J. Psychiatry, 158, 1058, 2001. 83. Martin-Ruiz, C.M. et al., Molecular analysis of nicotinic receptor expression in autism, Brain Res. Mol. Brain Res., 123, 81, 2004. 84. Hardan, A.Y. and Handen, B.L., A retrospective open trial of adjunctive donepezil in children and adolescents with autistic disorder, J. Child Adolesc. Psychopharmacol., 12, 237, 2002. 85. Marubio, L.M. et al., Reduced antinociception in mice lacking neuronal nicotinic receptor subunits, Nature, 398, 805, 1999. 86. Picciotto, M.R., Brunzell, D.H., and Caldarone, B.J., Effect of nicotine and nicotinic receptors on anxiety and depression, Neuroreport, 13, 1097, 2002. 87. Long, J.M et al., Expanded characterization of the social interaction abnormalities in mice lacking Dvl1, Genes Brain Behav., 3, 51, 2004. 88. Li. J. et al., Lack of evidence for an association between WNT2 and RELN polymorphisms and autism, Am. J. Med. Genet. B Neuropsychiatr. Genet., 126, 51, 2004. 89. Panksepp, J., A neurochemical theory of autism, Trends Neurosci., 2, 174, 1989. 90. Nelson, E.E. and Panksepp, J., Brain substrates of infant-mother attachment: contributions of opioids, oxytocin, and norepinephrine, Neurosci. Biobehav. Rev., 22, 437, 1998. 91. Moles, A., Kieffer, B.L., and D’Amato, F.R., Deficit in attachment behavior in mice lacking the µ-opioid receptor gene, Science, 304, 1983, 2004. 92. Nader, R. et al., Expression of pain in children with autism, Clin. J. Pain, 20, 88, 2004. 93. Mason, G., Stereotypies: a critical review, Anim. Behav., 41, 1015, 1991. 94. Turner, C.A. et al., Spontaneous stereotypy in an animal model of Down syndrome: Ts65Dn mice, Behav. Genet., 31, 393, 2001. 95. Presti, M.F., Gibney, B.C., and Lewis, M.H., Effects of intrastriatal administration of selective dopaminergic ligands on spontaneous stereotypy in mice, Physiol. Behav., 80, 433, 2004. 96. Turner, C.A. and Lewis, M.H., Environmental enrichment: effects on stereotyped behavior and neurotrophin levels, Physiol. Behav., 80, 259, 2003. 97. Persico, A.M. et al., Reelin gene alleles and haplotypes as a factor predisposing to autistic disorder, Mol. Psychiatry, 6, 150, 2001.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 298 Wednesday, February 15, 2006 3:25 PM

298

Understanding Autism: From Basic Neuroscience to Treatment

98. Fatemi, S.H., Stary, J.M., and Egan, E.A., Reduced blood levels of reelin as a vulnerability factor in pathophysiology of autistic disorder, Cell. Mol., Neurobiol., 22, 139, 2002. 99. Ignatova, N., Sindic, C.J.M., and Goffinet, A.M., Characterization of the various forms of the Reelin protein in the cerebrospinal fluid of normal subjects and in neurological diseases, Neurobiol. Dis., 15, 326, 2004. 100. Liu, W.S. et al., Down-regulation of dendritic spine and glutamic acid decarboxylase 67 expressions in the reelin haploinsufficient heterozygous reeler mouse, Proc. Natl. Acad. Sci. USA, 98, 3477, 2001. 101. Tueting, P. et al., The phenotypic characteristics of heterozygous reeler mouse, Neuroreport, 10, 1329, 1999. 102. Tremolizzo, L. et al., An epigenetic mouse model for molecular and behavioral neuropathologies related to schizophrenia vulnerability, Proc. Natl. Acad. Sci. USA, 99, 17095, 2002. 103. Groves, J.O. et al., Evaluation of the heterozygous reeler mouse as a potential model for schizophrenia, J. Psychopharmacol., 17, A43, 2003. 104. Salinger, W.L., Ladrow, P., and Wheeler, C., Behavioral phenotype of the reeler mutant mouse: effects of RELN gene dosage and social isolation, Behav. Neurosci.,117, 1257, 2003. 105. Podhorna, J. and Didriksen, M., The heterozygous reeler mouse: behavioral phenotype, Behav. Brain Res., 153, 43, 2004. 106. Hadj-Sahraoui, N. et al., Gender effect on Purkinje cell loss in the cerebellum of the heterozygous reeler mouse, J. Neurogenet., 11, 45, 1996. 107. Miller, M.T. et al., Autism associated with conditions characterized by developmental errors in early embryogenesis: a mini review, Int. J. Dev. Neurosci., 23, 201, 2005. 108. Arndt, T.L., Stodgell, C.J., and Rodier, P.M., The teratology of autism, Int. J. Dev. Neurosci., 23, 189, 2005. 109. Vorhees, C.V., Weisenburger, W.P., and Minck, D.R., Neurobehavioral teratogenic effects of thalidomide in rats, Neurotoxicol. Teratol., 23, 255, 2001. 110. Miyahara, T. et al., Autism model rat: marked behavioral changes in the embryonal thalidomide/valproic acid exposed rats, Program No. 318.8, Abstract Viewer/Itinerary Planner, Washington, D.C., Society for Neuroscience, 2003. 111. Narita, N. et al., Increased monoamine concentration in the brain and blood of fetal thalidomide- and valproic acid-exposed rat: putative animal models for autism, Pediatr. Res., 52, 576, 2002. 112. Miyazaki, K., Narita, N., and Narita, M., Maternal administration of thalidomide or valproic acid causes abnormal serotonergic neurons in the offspring: implication for pathogenesis of autism, Int. J. Dev. Neurosci., 23, 287, 2005. 113. Teitelbaum, P., A proposed primate animal model of autism, Eur. Child Adolesc. Psychiatry, 12, 48, 2003. 114. Ingram, J.L. et al., Prenatal exposure of rats to valproic acid reproduces the cerebellar anomalies associated with autism, Neurotoxicol. Teratol., 22, 319, 2000. 115. Rodier, P.M. et al., Embryological origin for autism: developmental anomalies of the cranial nerve motor nuclei, J. Comp. Neurol., 370, 247, 1996. 116. Rodier, P.M., Converging evidence for brain stem injury in autism, Dev. Psychopathol., 14, 537, 2002. 117. Schneider, T. and Przewtocki, R., Behavioral alterations in rats prenatally exposed to valproic acid: animal model of autism, Neuropsychopharmacology, 30, 80, 2005.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 299 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

299

118. Brost, D.R., Cross, J.D., and Hollerman, J.R., Social recognition memory following prenatal valproic acid: an animal model of autism, Abstract Viewer/Itinerary Planner, Washington, D.C.: Society for Neuroscience, 318, 10, 2003. 119. Hao, Y. et al., Mood stabilizer valproate promotes ERK pathway-dependent cortical neuronal growth and neurogenesis, J. Neurosci., 24, 6590, 2004. 120. Costa, L.G. et al., Developmental neuropathology of environmental agents, Annu. Rev. Pharmacol. Toxicol., 44, 87, 2004. 121. Bennett, G.D. et al., Valproic acid-induced alterations in growth and neurotrophic factor gene expression in murine embryos, Reprod. Toxicol., 14, 1, 2000. 122. Faiella, A. et al., A mouse model for valproate teratogenicity: parental effects, homeotic transformation, and altered HOX expression, Hum. Mol. Genet., 9, 227, 2000. 123. Finnell, R.H. et al., Molecular basis of environmentally induced birth defects, Annu. Rev. Pharmacol. Toxicol., 42, 181, 2002. 124. Carpenter, E.M. et al., Loss of HOX-A1 (HOX-1.6) function results in the reorganization of the murine hindbrain, Development, 118, 1063, 1993. 125. Mark, M. et al., Two rhombomeres are altered in Hoxa-1 mutant mice, Development, 119, 319, 1993. 126. Ciaranello, A.L. and Ciaranello, R.D., The neurobiology of infantile autism, Annu. Rev. Neurosci., 18, 101, 1995. 127. Chess, S., Follow-up report on autism in congenital-rubella, J. Autism Child. Schizophr., 7, 69, 1977. 128. Brown, A.S. and Susser, E.S., In utero infection and adult schizophrenia, Ment. Retard. Dev. Disabil. Res. Rev., 8, 51, 2002. 129. Brown, A.S. et al., Elevated maternal interleukin-8 levels and risk of schizophrenia in adult offspring, Am. J. Psychiatry, 161, 889, 2004. 130. Brown, A.S. et al., Serologic evidence of prenatal influenza in the etiology of schizophrenia, Arch. Gen. Psychiatry, 61, 774, 2004. 131. Shi, L., Fatemi, S.H., Sidwell, R.W., and Patterson, P.H., Maternal influenza infection causes marked behavioral and pharmacological changes in the offspring, J. Neurosci., 23, 297, 2003. 132. Patterson, P.H. and Shi, L., Maternal influenza virus infection causes a cerebellar defect resembling that in autism, Int. Meeting. Autism Res., S3.3.4, 2002 133. Fatemi, S.H. et al., Prenatal viral infection leads to pyramidal cell atrophy and macrocephaly in adulthood: implications for genesis of autism and schizophrenia, Cell. Mol. Neurobiol., 22, 25. 2002. 134. Fatemi, S.H. et al., Differential expression of synaptosome-associated protein 25 kDa [SNAP-25] in hippocampi of neonatal mice following exposure to human influenza virus in utero, Brain Res., 800, 1, 1998. 135. Fatemi, S.H. et al., Prenatal viral infection causes alterations in nNOS expression in developing mouse brains, Neuroreport., 11, 1493, 2000. 136. Shi, L., Tu, N., and Patterson, P.H., Maternal influenza infection is likely to alter fetal brain development indirectly: the virus is not detected in the fetus, Int. J. Dev. Neurosci., 23, 299, 2005. 137. Aronsson, F. et al., Persistence of viral RNA in the brain of offspring to mice infected with influenza A/WSN/33 virus during pregnancy, J. Neurovirol., 4, 353, 2003. 138. Meyer, U., Feldon, J., Schedlowski, M., and Yee, B.K., Towards an immunoprecipitated neurodevelopmental animal model of schizophrenia, Neurosci. Biobehav. Rev., 29, 913, 2005.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 300 Wednesday, February 15, 2006 3:25 PM

300

Understanding Autism: From Basic Neuroscience to Treatment

139. Zuckerman, L., Rehavi, M., Nachman, R., and Weiner, I., Immune activation during pregnancy in rats leads to a postpubertal emergence of disrupted latent inhibition, dopaminergic hyperfunction, and altered limbic morphology in the offspring: a novel neurodevelopmental model of schizophrenia, Neuropsychopharmacology, 28, 1778, 2003. 140. Borrell, J. et al., Prenatal immune challenge disrupts sensorimotor gating in adult rats: implications for the etiopathogenesis of schizophrenia, Neuropsychopharmacology, 36, 204, 2002. 141. Vargas, D.L. et al., Neuroglial activation and neuroinflammation in the brain of patients with autism, Ann. Neurol., 57, 67, 2005. 142. Fortier, M.E., Joober, R., Luheshi, G.N., and Boksa, P., Maternal exposure to bacterial endotoxin during pregnancy enhances amphetamine-induced locomotion and startle responses in adult rat offspring, J. Psychiatr. Res., 38, 335, 2004. 143. Dantzer, R., Cytokine-induced sickness behavior: a neuroimmune response to activation of innate immunity, Eur. J. Pharmacol., 500, 399, 2004. 144. Patterson, P.H., Maternal infection: window on neuroimmune interactions in fetal brain development and mental illness, Curr. Opin. Neurobiol., 12, 115, 2002. 145. Fidel, P.L. et al., Systemic and local cytokine profiles in endotoxin-induced preterm parturition in mice, Am. J. Obstet. Gynecol., 170, 1467, 1994. 146. Urakubo, A., Jarskog, L.F., Lieberman, J.A., and Gilmore, J.H., Prenatal exposure to maternal infection alters cytokine expression in the placenta, amniotic fluid, and fetal brain, Schizophr. Res., 47, 27, 2001. 147. Buka, S.L. et al., Maternal cytokine levels during pregnancy and adult psychosis, Brain Behav. Immun., 4, 411, 2001. 148. Lang, B., Dale, R.C., and Vincent, A., New autoantibody disorders of the central nervous system, Curr. Opin. Neurol., 16, 351, 2003. 149. Snider, L.A., and Swedo, S.E., PANDAS: current status and directions for research, Mol. Psychiatry, 9, 900, 2004. 150. Dalton, P. et al., Maternal neuronal antibodies associated with autism and a language disorder, Ann. Neurol., 53, 533, 2003. 151. Krause, I. et al., Brief report: immune factors in autism: a critical review, J. Autism. Dev. Disord., 32, 337, 2002. 152. Comi, A.M. et al., Familial clustering of autoimmune disorders and evaluation of medical risk factors in autism, J. Child. Neurol., 14, 388, 1999. 153. Sweeten, T.L. et al., Increased prevalence of familial autoimmunity in probands with pervasive developmental disorders, Pediatrics, 112, 420, 2003. 154. Hornig, M., Briese, T., and Lipkin, W.I., Borna disease virus, J. Neurovirol., 9, 259, 2003. 155. Pletnikov, M.V. et al., Effects of genetic background on neonatal Borna disease virus infection-induced neurodevelopmental damage. I. Brain pathology and behavioral deficits, Brain Res., 944, 97, 2002. 156. Dietz, D.M. et al., Developmental alterations in serotoninergic neurotransmission in Borna disease virus (BDV)-infected rats: a multidisciplinary analysis, J. Neurovirol., 5, 267, 2004. 157. Yamashita, Y. et al., Possible association between congenital cytomegalovirus infection and autistic disorder, J. Autism Dev. Disord., 33, 455, 2003. 158. Madsen, K.M. and Vestergaard, M., MMR vaccination and autism: what is the evidence for a causal association? Drug Saf., 27, 831, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 301 Wednesday, February 15, 2006 3:25 PM

Modeling Features of Autism in Animals

301

159. Smeeth, L. et al., MMR vaccination and pervasive developmental disorders: a case control study, Lancet, 11, 963, 2004. 160. Hornig, M., Chian, D., Lipkin, W.I., Neurotoxic effects of postnatal thimerosal are mouse strain dependent, Mol. Psychiatry, 9, 833, 2004. 161. Parker, S.K., Schwartz, B., Todd, J., and Pickering, L.K., Thimerosal-containing vaccines and autistic spectrum disorder: a critical review of published original data, Pediatrics, 114, 793, 2004. 162. Andrews, N. et al., Thimerosal exposure in infants and developmental disorders: a retrospective cohort study in the United Kingdom does not support a causal association, Pediatrics, 114, 584, 2004. 163. Critchley, H.D. et al., The functional neuroanatomy of social behavior: changes in cerebral blood flow when people with autistic disorder process facial expressions, Brain, 123, 2203, 2000. 164. Schultz, R.T. et al., The role of the fusiform face area in social cognition: implications for the pathobiology of autism, Philos. Trans. R. Soc. Lond. B Biol. Sci., 358, 415, 2003. 165. Howard, H.A. et al., Convergent neuroanatomical and behavioral evidence of an amygdala hypothesis of autism, Neuroreport, 11, 2931, 2000. 166. Davidson, R.J. and Slagter, H.A., Probing emotion in the developing brain: functional neuroimaging in the assessment of the neural substrates of emotion in normal and disordered children and adolescents, Ment. Retard. Dev. Disabil. Res. Rev., 6, 166, 2000. 167. Amaral, D.G., Bauman, M.D., and Mills Schumann, C., The amygdala and autism: implications from non-human primate studies, Genes Brain Behav., 2, 295, 2003. 168. Schumann, C.M. et al., The amygdala is enlarged in children but not adolescents with autism; the hippocampus is enlarged at all ages, J. Neurosci., 24, 6392, 2004. 169. Bachevalier, J., The amygdala, social cognition, and autism, in Aggleton, J.P. (Ed.), The Amygdala: a Functional Analysis, Oxford University Press, New York, 2000, p. 509. 170. Machado, C.J. and Bachevalier, J., Non-human primate models of childhood psychopathology: the promise and the limitations, J. Child Psychol. Psychiatry, 44, 64, 2003. 171. Amaral, D.G. and Corberr, B.A., The amygdala, autism and anxiety, Novartis Found. Symposium, Wiley, Chichester, U.K., 251, 177, 2003. 172. Prather, M.D. et al., Increased social fear and decreased fear of objects in monkeys with neonatal lesions, Neuroscience, 106, 653, 2001. 173. Wolterink, G. et al., Early amygdala damage in the rat as a model for neurodevelopmental psychopathological disorders, Eur. Neuropsychopharmacol., 11, 51, 2001. 174. Diergaarde, L. et al., Neonatal amygdala lesions and juvenile isolation in the rat: differential effects on locomotor and social behavior later in life, Behav. Neurosci., 118, 298, 2004. 175. Hall, F.S., Social deprivation of neonatal, adolescent, and adult rats has distinct neurochemical and behavioral consequences, Crit. Rev. Neurobiol., 12, 129, 1998. 176. Sarna, J.R. and Hawkes, R., Patterned Purkinje cell death in the cerebellum, Prog. Neurobiol., 70, 473, 2003. 177. Bobee, S., Mariette, E., Tremblay-Leveau, H., and Caston, J., Effects of early midline cerebellar lesion on cognitive and emotional functions in the rat, Behav. Brain Res., 112, 107, 2000.

© 2006 by Taylor & Francis Group, LLC

2732_C013.fm Page 302 Wednesday, February 15, 2006 3:25 PM

302

Understanding Autism: From Basic Neuroscience to Treatment

178. Lev-Ram, V. et al., A novel non-ataxic guinea pig strain with cerebrocortical and cerebellar abnormalities, Brain Res., 606, 325, 1993. 179. Caston, J. et al., An animal model of autism: behavioral studies in the GS guineapig, Eur. J. Neurosci., 10, 2677, 1998. 180. Gerlai, R., Millen, K.J., Herrup, K., Fabien, K., Joyner, A.L., and Roder, J., Impaired motor learning performance in cerebellar EN-2 mutant mice, Behav. Neurosci., 110, 126, 1996. 181. Croen, L.A., Grether, J.K., Yoshida, C.K., Odouli, R., and Van de Water, J., Maternal autoimmune diseases, asthma and allergies, and childhood autism spectrum disorders, Arch. Pediatr. Adolesc. Med., 159, 151, 2005.

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 303 Wednesday, February 15, 2006 3:25 PM

and 14 Neuroanatomical Neurochemical Studies of the Autistic Brain: Current Thought and Future Directions Margaret L. Bauman, George Anderson, Elaine Perry, and Melissa Ray CONTENTS Introduction............................................................................................................303 Neuroanatomical Observations..............................................................................304 The Limbic System ...............................................................................................305 The Cerebellum and Brain Stem...........................................................................306 Neurotransmitter Systems......................................................................................308 Norepinephrine ...............................................................................................308 Dopamine........................................................................................................309 Serotonin.........................................................................................................309 Acetylcholine..................................................................................................311 Summary and Future Directions ...........................................................................315 Acknowledgments..................................................................................................316 References..............................................................................................................317

INTRODUCTION Early infantile autism is a behaviorally defined disorder, first described by Kanner in 1943 (Kanner, 1943). The syndrome is characterized by the presence of delayed and disordered language development, impaired social interaction and cognitive skills, lack of imaginary play, poor eye contact, isolated areas of interest, and an obsessive insistence on sameness. Repetitive and stereotypic behavior and disordered modulation of sensory input may be present in some cases. Many affected individuals appear to have unusual islands of rote memory, and some show exceptional talents in the presence of otherwise general functional disability. Although the cause of 303

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 304 Wednesday, February 15, 2006 3:25 PM

304

Understanding Autism: From Basic Neuroscience to Treatment

autism remains largely unknown, twin and family studies strongly suggest evidence for a genetic liability, the mechanism of which has yet to be determined. Over the past 60 years, our understanding of some of the underlying biological mechanisms related to the autism spectrum disorders (ASD) has expanded through a combination of basic science and clinical approaches, aided by advances in technology and more precise phenotypic definition of clinical symptomatology. Increasingly, there has been the attempt to correlate what is known about the neurobiology of the brain with the behavioral, cognitive, language, and social characteristics associated with this complex syndrome. This chapter will review what is known about the neuroanatomy and some of the neurotransmitter systems of the brain in autism. The potential relationship of these findings to the clinical features of ASD will be discussed and directions for future research suggested.

NEUROANATOMICAL OBSERVATIONS In recent years, in vivo analysis of brain structure and function has become more available, and morphometric study of specific brain regions has become an increased focus of neuroimaging research. Spurred by the observation that head circumference in autistic children appears to atypically accelerate in size between birth and the later preschool years (Lainhart et al., 1997), a number of imaging studies have been devoted to the analysis of brain volume in both children and adults with the disorder. In 2001, Courchesne et al. (2001) made the observation that brain volume in a series of autistic children, as studied by magnetic resonance imaging (MRI), increased most markedly between the ages of 2.0 and 4.5 yr, followed by a deceleration of brain growth thereafter. Although both gray and white matter volumes were found to be increased, the major changes involved both the cerebellar and cerebral white matter. Subsequent studies of the same nature have noted that the greatest increase in volume appears to be localized to the frontal lobes with relative sparing of the occipital lobe (Carper et al. 2002). More recently, Herbert et al. (2004) observed that, in a series of autistic children and those with developmental language disorders (DLD), the increase in brain volume appeared to primarily involve the radiate white matter, regions that myelinate later than the deep white matter. This observation would be consistent with the unusual postnatal head growth reported in autism. Supporting these reports of increased brain volume is the observation that fresh brain weights obtained at autopsy are also increased in autistic children in relation to that of adults. In a series of 19 postmortem cases, comprising ages 5 to 13 years, brain weight was found to be greater by 100 to 200 g than the expected weight for that age and sex, and this difference was statistically significant when compared to controls. In contrast, the brain weight in a series of adult autistic cases, of ages 18 to 54 years, were found to be lighter than that expected for that age and sex (Bauman and Kemper, 2005). Although the pathogenesis of this brain enlargement is as yet unknown, a number of mechanisms have been suggested, including a possible increase in the number of neurons and/or glia, premature and accelerated proliferation of synapses, axonal and dendritic arbors, and/or the presence of increased myelination. To date, however, postmortem studies in the autistic brain have failed to show consistent abnormalities

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 305 Wednesday, February 15, 2006 3:25 PM

Neuroanatomical and Neurochemical Studies of the Autistic Brain

305

in the cerebral cortex, and myelination has been found both microscopically and by MRI, to be comparable to controls (Bauman and Kemper, 1994). In a survey in which the whole-brain serial section of nine postmortem cases were studied, the only cortical abnormality found was a minor malformation of the orbitofrontal cortex in one hemisphere obtained from a 9-year-old autistic child with a history of severe seizures (Kemper, 2004). Bailey et al. (1998), however, has reported neocortical malformations in four out of the six brains examined. These abnormalities included atypically oriented pyramidal cells, areas of increased neuronal and cortical density, and irregular laminar patterns. More recently, Casanova et al. (2002) observed that, in comparison with controls, neocortical minicolumns were smaller, less compact, and more numerous in the three cortical areas studied. The authors suggested that, because inhibitory GABAergic double-bouquet cells define the minicolumnar organization, these cells might be the primary sites of these structural differences. Thus, although cortical findings have been reported in some autistic brains, they appear to vary from case to case, and those that have been observed do not appear to be sufficient to explain the presence of increased brain weight and volume. Very few studies of myelin structure or composition have been reported in the autistic brain. Preliminary analysis of the corpus callosum obtained from deceased adult autistic subjects has suggested that alterations in glycolipids and phospholipids may be present in the autism, at least in some cases (Koul, 2005). Thus, although data are still very limited, given the presence of atypical information processing described in autism (Minshew et al., 1997) and the dramatic myelination of brain circuitry during the first year of life (Yakovlev and Lecours, 1967), it is reasonable to consider the possibility that abnormalities of the composition and/or structure of the myelin sheath may be a contributing factor.

THE LIMBIC SYSTEM In 1984, the brain obtained from a well-documented autistic man was studied by means of whole-brain serial section in comparison with an identically processed age- and sex-matched control (Bauman and Kemper, 1984, 1985). Since that initial report, eight additional cases, similarly studied, have since been described (Bauman and Kemper, 1994). All nine brains showed abnormalities of the limbic system and in the cerebellum and related inferior olive. Reduced neuronal cell size and increased cell packing density (number of cells per unit tissue volume) was noted throughout the hippocampus, the medially placed nuclei of the amygdala, the medial mammillary body, the anterior cingulate gyrus, and in the medial septal nucleus (MSN). Selected hippocampal pyramidal neurons, obtained from two autistic childhood brains and studied with the rapid Golgi method, have shown a reduced complexity and extent of dendritic arbors (Raymond et al., 1995). This pattern of reduced neuronal cell size, increased cell packing density, and decreased complexity and extent of dendritic arbors are associated with early stages of brain maturation, and may therefore represent a selective curtailment of normal brain development within these structures. All of these regions are known to be related to each other by interconnecting circuits and make up a major portion of the limbic system of the brain.

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 306 Wednesday, February 15, 2006 3:25 PM

306

Understanding Autism: From Basic Neuroscience to Treatment

More recently, Schumann and Amaral (2005) studied the amygdala of from 11 autistic brains in comparison with age- and sex-matched controls by using stereological techniques. They found no difference in neuronal cell number, size, or density in these autistic brains when compared with controls. However, the amygdala in the brains of autistic children has been reported to be enlarged in comparison with that of typically developing children when studied using MRI (Sparks et al., 2002; Schumann et al., 2004; Mosconi et al., 2005). These investigators suggest that these volumetric differences are unlikely to be due to an increased number of neurons, but may be related to the complexity of dendritic or axonal fiber systems. Given that the results of this microscopic study run counter to previous reports, further studies of the histoanatomy of the amygdala in the autistic brain is warranted with detailed clinical characterization of the subjects during life. Quantitative analysis of the parvalbumin (PV)-positive subpopulation of GABAergic interneurons throughout the hippocampal complex has been recently reported in postmortem brain tissue obtained from five adult male autistic subjects and five age- and sex-matched controls (Lawrence, 2005). All sampled hippocampal subfields demonstrated an increase in PV-positive interneuron cell density with area CA1 reaching statistical significance. The investigators concluded that the increased packing density of PV-positive interneurons appeared to be a feature shared with the total neuronal population in the hippocampal complex, suggesting that both respond to the same unknown pathological process. Further studies investigating additional GABAergic subpopulations in the hippocampus are presently underway. Although small cell size and increased cell packing density have been reported throughout the limbic system in some early studies (Bauman and Kemper, 1994), a different pattern of abnormality has been observed in the nucleus of the diagonal band of Broca (NDB) located in the septum. In this nucleus, enlarged cells that otherwise appear normal were noted in the NDB in all of the autistic subjects of age less than 13 years. In contrast, these same neurons were noted to be smaller in size and reduced in number in all of the autistic adult brains over the age of 22 years (Bauman and Kemper, 1994). It is difficult to interpret the differences in cell size and cell packing density that vary with age in this nucleus. It is possible that these findings may represent an unstable circuit involving the NDB. In the adult monkey, this nucleus provides a highly focused cholinergic projection to the amygdala and hippocampus (Rosene and van Hoesen, 1987). The extent of its nuclear projection during fetal life is unknown. It is possible that the small neurons observed in the hippocampal complex and amygdala in autistic brains may fall within the fetal distribution of this septal projection.

THE CEREBELLUM AND BRAIN STEM Outside of the forebrain, abnormalities have also been reported in the cerebellum and related inferior olive. One of the most consistently reported observations has been the reduction in the number of Purkinje cells throughout the cerebellar cortex, most prominently in the posterolateral neocerebellar cortex and adjacent archicerebellar cortex (Ritvo et al., 1986; Arin et al., 1991). In contrast to these findings in the hemispheres, detailed quantitative analysis of the Purkinje cell number throughout

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 307 Wednesday, February 15, 2006 3:25 PM

Neuroanatomical and Neurochemical Studies of the Autistic Brain

307

all lobules of the vermis has shown no statistically significant differences when compared with controls (Bauman and Kemper, 1996). Within the deep cerebellar nuclei, and similar to the findings observed in the NDB of the septum, abnormalities have been found to vary with age. As in the NDB, the cells of the fastigial, globose, and emboliform nuclei located in the roof of the cerebellum are small, pale, and reduced in number in all of the adult brains. In all of the childhood brains less than 13 years of age, the same neurons as well as those of the dentate nucleus are enlarged and present in adequate numbers (Bauman and Kemper, 1994). We have hypothesized that the enlarged cells seen in the childhood brains may represent the persistence of a fetal circuit between the neurons of the inferior olive and those of the deep cerebellar nuclei. Because this fetal circuit may not be “designed” to be retained for the long term as the dominant postnatal circuit, it may not be able to sustain itself over time. This may then account for the reduced size and eventual loss of neurons in the adult autistic (Bauman and Kemper, 1994). In the brain stem, no retrograde cell loss or atrophy has been noted in the principal olivary nucleus in any of the autistic brains, areas known to be related to the abnormal regions of the cerebellar cortex (Holmes and Stewart, 1908). In humans, neuropathology, atrophy and neuronal cell loss in the inferior olive has been invariably observed following perinatal and postnatal Purkinje cell loss (Norman, 1940; Greenfield, 1954). This cell loss is believed to be due to the close relationship of the olivary climbing fiber axons to the Purkinje cell dendrites (Eccles et al., 1967). It has been shown in fetal monkey that the olivary climbing fibers synapse with the Purkinje cell dendrites in a transitory zone beneath the Purkinje cells (called the lamina dessicans) prior to establishing their definitive relationship with the Purkinje cells (Rakic, 1971). Since this zone is no longer present in the human fetus after 30 weeks gestation (Rakic and Sidman, 1970), it is likely that the cerebellar lesions observed in the autistic brain occurred at or before that time. Although the olivary neurons are present in adequate numbers in all of the adult autistic brains studied to date, they appear to be small in size. In contrast and similar to the findings in the NDB and in the deep cerebellar nuclei, all of the childhood autistic brains show enlarged but otherwise appearing normal olivary neurons (Kemper and Bauman, 1998). Rodier et al. (1996) have noted additional abnormalities in the brain stem. These investigators described decreased numbers of neurons in the facial nucleus and superior olive and shortening of the distance between the trapezoid body and the inferior olive in a patient with autism and Mobius syndrome. In another study, Bailey et al. (1998) have observed a dysplastic configuration of the lamella in the inferior olive and the presence of ectopic neurons lateral to the inferior olive in some cases. In both studies, the findings suggest an abnormality in early brain development occurring during the prenatal period. Thus, microscopic abnormalities in the autistic brain suggest a prenatal onset of the disorder, provide evidence for the involvement of limbic system and cerebellar structures and related circuitry, and indicate some findings that appear to alter with age. Variable abnormalities in the cerebral cortex have been described. However, it is not yet known whether these findings are related to the presence or absence of mental retardation, seizures, or other as-yet-undefined clinical variables in the

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 308 Wednesday, February 15, 2006 3:25 PM

308

Understanding Autism: From Basic Neuroscience to Treatment

subjects studied. Likewise, there is as yet no clear explanation for the apparent brain overgrowth early in life, nor its clinical significance. The possible presence of increased numbers of neurons, interneurons, and/or glial cells, increased dendritic arbors, altered synaptic structures, and, possibly, abnormal myelin layering or composition have all been suggested as potential contributors. Future research will need to address these and many other questions in light of advancing technology and the complex and heterogeneous clinical characteristics of associated with ASD.

NEUROTRANSMITTER SYSTEMS Although defining neuroanatomical abnormalities in the autistic brain will continue to be an important area of investigation, it will be equally important to consider the underlying neurochemical features, if any, that may be associated with the brain regions identified as being morphologically abnormal. Until recently, neurochemical research in autism has focused on the three major monoamine transmitters norepinephrine (NE), dopamine (DA), and serotonin (5-HT) (Anderson, 1987; Cook et al., 1990; Lotspeich and Ciaranello, 1993). The strategies employed in neurochemical research have included (1) the measurement of neurotransmitters and their metabolites in postmortem brain tissue, cerebrospinal fluid (CSF), plasma or serum and blood elements, and urine; (2) the measurement of neurotransmitter-related enzymes in brain and blood; and (3) the measurement of neurotransmitter transporters and receptors in brain and blood specimens. Most of the resulting data has been confined to the more narrowly defined autism phenotype. Some limited neurochemical data are available regarding Rett syndrome, but little or no neurochemical research has been carried out on individuals with PDDNOS (pervasive developmental disorder not otherwise specified) or Asperger’s syndrome.

NOREPINEPHRINE The research on norepinephrine (NE), and the related catecholamine — epinephrine (EPI), is most appropriately discussed in the context of stress response systems. The two major components of the stress response system (Chrousos and Gold, 1992), the noradrenergic sympathetic-adrenomedullary system and the hypothalamicpituitary-adrenal (HPA) axis, have been of interest because of their hyperarousal, hyperactivity, and overreactivity to novel situations often in autism. The functioning of the sympathetic-adrenomedullary system has been assessed through measurements of NE and EPI in plasma and urine. In addition, plasma and urine levels of the major NE metabolites, 3-methoxy-4-hydroxyphenylethylglycol (MHPG) and vanillylmandelic acid (VMA), have also been determined. Serum levels of dopamineβ-hydroxylase (DBH), the synthetic enzyme secreted along with NE from sympathetic neurons, have likewise been analyzed. These studies have been thoroughly reviewed in the literature (Martineau et al., 1994; Minderaa et al., 1994; Schultz and Anderson, 2004). To interpret the varied results obtained across the studies of stress response, it is necessary to appreciate that certain of the measures reflect basal or baseline function (urine measures in general and serum DBH), while others index the acute

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 309 Wednesday, February 15, 2006 3:25 PM

Neuroanatomical and Neurochemical Studies of the Autistic Brain

309

stress response (plasma NE, blood pressure, and heart rate). Indices reflecting basal functioning of the sympathetic-adrenomedullary system consistently have been found to be normal in patients with autism. On the other hand, most of the studies measuring indices of acute stress response have found elevations in patients with autism. Taken together, the data support the idea that stress response systems are hyperresponsive when individuals with autism are stressed but that autistic patients are not in a chronic state of hyperarousal. Findings from studies of HPA axis function are consistent with the sympathetic-adrenomedullary results, and support the same conclusions regarding basal functioning and acute responsivity (Tordjman et al., 1997). The apparent increased response to stressors could be due to a difference in the level of perceived stress, an overelicitation of the physiologic response, or an abnormality in the stress response and arousal systems themselves.

DOPAMINE Altered dopamine (DA) functioning in autism can be postulated based on DA’s clear role in mediating repetitive behaviors and stereotypies, the similarities between the negative symptoms of schizophrenia and autism, and the observation that DA blocking agents (antipsychotics) are effective in treating some aspects of autism. Most studies of DA have examined levels of the major DA metabolite, homovanillic acid (HVA). Unfortunately, results have been inconsistent, with the concentration of HVA in CSF being reported as slightly decreased, apparently normal, or significantly increased (approximately 50%) in autism (reviewed in Narayan et al., 1993). Measurements of HVA in urine have also been discrepant, and the only study of plasma HVA reported similar levels in autistic and control subjects (Minderaa et al., 1989). Other relevant measures include urinary DA and plasma prolactin, which have also been found to be normal in autistic subjects (McBride et al., 1989; Minderaa et al., 1989). Thus, taken together, the available neurochemical and neuroendocrine evidence suggests that central dopaminergic functioning is normal in autism. However, the existence of localized alterations in DA metabolism (Ernst et al., 1997) or changes in specific aspects of DA neurotransmission cannot be ruled out yet.

SEROTONIN One of the best documented biological alterations associated with autism is the group-mean elevation in blood levels of 5-HT (Anderson, 2002; Anderson, 1990), first reported by Schain and Freedman in 1961 (Schain and Freedman, 1961). The basic finding of elevated platelet 5-HT has been robust and well replicated, with group-mean increases of 25 to 50% usually observed (Anderson et al., 2002). More recent research has better characterized this finding (McBride et al., 1998; Mulder et al., 2004), and it now appears that the distribution of platelet 5-HT concentration in autism may be bimodal (Mulder et al., 2004). Although hyperserotonemia in autism is not specific enough to allow the use of blood 5-HT levels as a diagnostic screening tool, its role in the etiology of autism and as a biological marker of autism remain important areas for research. Further, interest in the role of 5-HT in autism

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 310 Wednesday, February 15, 2006 3:25 PM

310

Understanding Autism: From Basic Neuroscience to Treatment

has been strengthened by the positive effect of 5-HT reuptake inhibitors on some autism symptoms, the exacerbation of autistic symptoms observed after tryptophan (a serotonin precursor) depletion, and a consideration of 5-HT’s role in a range of autism-relevant behaviors. Moreover, the growing appreciation of the role of 5-HT in early neurodevelopment points to additional pathways by which 5-HT may contribute to the etiology of autism (Janusonis et al., 2004). There had been a continuing effort to understand the significance of elevated platelet 5-HT, frequently observed in autism, and its underlying mechanisms (Anderson et al., 2002). The normal excretion rates observed for 5-HT and its major metabolite, 5-hydroxyindoleacteic acid (5-HIAA), in autistic subjects indicate that 5-HT synthesis and catabolism are not altered (Anderson et al., 1989). The circulating platelet’s exposure to 5-HT appears normal as measured levels of free 5-HT in plasma are similar to those seen in controls (Cook et al., 1988). Uptake studies have not been definitive, and group differences have not typically been found when 5-HT transporter binding sites have been studied in autistic and normal individuals. Moreover, it now appears that an insertion or deletion promoter polymorphism in the 5-HT transporter gene, although functional, probably does not contribute substantially to the hyperserotonemia of autism, given the allele frequencies and functional effects observed (Anderson et al., 2002). Although most of the 5-HT-related research has focused on the hyperserotonemia of autism, a number of studies of CSF 5-HIAA and several neuroendocrinological studies of central 5-HT functioning have been reported. The CSF studies have found few or no differences between autistic and control groups (Narayan et al., 1993). Neuroendocrine challenge studies have measured plasma levels of growth hormones and prolactin after administration of either fenfluramine (McBride et al., 1989), the 5-HT 1B/D receptor agonist sumatriptan (Novotny et al., 2000), or the 5-HT precursor 5-hydroxytryptophan (5-HTP). The fenfluramine and 5-HTP studies both observed a lowering or blunting of the prolactin response while sumatriptan produced an elevated response in autism. Although the neuroendocrine results are consistent with decreased central presynaptic 5-HT functioning in autism, the fenfluramine challenge data have also been interpreted as indicating reduced central postsynaptic serotonin type 2A receptor (5-HT2A) functioning. Accumulating data regarding 5-HT2A receptors may also be relevant to the presence of platelet hyperserotonemia in autism. Initial work by McBride and colleagues found that both central and peripheral 5-HT2A responses were blunted in autism, with both fenfluramine-induced release of prolactin and 5-HT augmentation of ADP-induced platelet aggregation being reduced (McBride et al., 1989). In addition, it was observed that the mean density of platelet 5-HT2A receptors was reduced, and that central and peripheral blunting was greatest in those individuals with the highest levels of platelet 5-HT, findings later supported by Cook and colleagues in 1993. More recently, additional evidence has come from a recent postmortem study (Antzoulatos and Blatt, 2004) and from two recent neuroimaging studies, both showing findings consistent with reduced central expression of 5-HT2A receptor in autism (Goldberg, unpublished, 2005; Murphy, personal communication, 2004). The causes and consequences of this apparent reduction in central and peripheral expression and functioning of the 5-HT2A receptor are unknown and warrant further

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 311 Wednesday, February 15, 2006 3:25 PM

Neuroanatomical and Neurochemical Studies of the Autistic Brain

311

investigation as does the possible relationship between altered 5-AHT2Aa receptor function and the platelet hyperserotonemia. The recent increased availability of postmortem brain specimens from individuals with autism provides a tremendous opportunity to study the neurobiology of autism. Preliminary studies of 5-HT-related neurochemical measures, including cortical and cerebellar levels of 5-HT, 5-HIAA, tryptophan, and transporter binding, have not identified differences between autistic and control groups (Anderson, unpublished data, 2005). However, there are initial neurochemical studies reporting altered GABAA receptor binding in the hippocampus (Blatt et al., 2000), glutamaterelated abnormalities (Purcell et al., 2001), reduced reelin protein in the cerebellum (Fatemi et al., 2001), and reduced nicotinic receptors in cortical regions (Perry et al., 2001), as well as the report of reduced cortical 5-HT2A receptor binding mentioned earlier (Antzoulatos and Blatt, 2004). These studies require confirmation, but are an indication that postmortem brain research in autism is beginning to catch up with that of other neuropsychiatric disorders.

ACETYLCHOLINE Compared to 5-HT, acetylcholine has only been recently implicated in autism. The cholinergic system consists of (1) neurons in the basal forebrain projecting to the hippocampus, amygdala, entire neocortex and thalamus, (2) striatal interneurons, (3) pontine nuclear groups, and (4) other brain stem nuclei projecting to the cerebellum and elsewhere. Cholinergic afferents innervate the cerebral cortex during the most dynamic periods of neuronal differentiation and synapse formation, suggesting that they play a regulatory role in these events (Holemann and Berger-Sweeney, 1998). Disruption of cholinergic innervation during early postnatal development can result in delayed cortical neuronal development, permanent changes in cortical cytoarchitecture, and cognitive function (Hohmann and Berger-Sweeney, 1998). In the cerebellum, the predominant cholinergic innervation originates from the vestibular nuclei (Barmack et al., 1992). Other projections include those originating in the midline medullar periventicular gray, the C3 adrenergic area raphe obscurus nucleus, and a variety of reticular formation nuclei, together with various sensory nuclei (Lan et al., 1995). Based on autoradiographic analysis in human cerebellum, nicotine binding has been detected in the molecular and granule cell layers but not in the Purkinje cell layer (Court et al., 1995). Paralleling developmental changes in ChAT (choline acetyltransferase), cerebellar cholinergic receptors decline postnatally. Muscarinic M2 and nicotinic receptors are higher in prenatal than in adult human cerebellum (Hellstrom-Lindahl et al., 1998; Hellstrom-Lindahl and Court, 2000), and in rat brain, M1, M3, and M4 muscarinic receptor subtypes decrease from juvenile to adult (Tice et al., 1996). No presynaptic abnormalities have been found in ChAT activity located in the frontal and parietal cortex or the basal forebrain of adult autistic subjects (Perry et al., 2001). Given the observation of reduced numbers of neurons in the basal forebrain cholinergic system reported in postmortem brain obtained from adult autistics (Bauman and Kemper, 1994), the finding of normal ChAT both in frontal and parietal cortex and in the basal forebrain suggests that the presynaptic cholinergic

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 312 Wednesday, February 15, 2006 3:25 PM

312

Understanding Autism: From Basic Neuroscience to Treatment

innervation of the cortex is structurally intact in autism. This indicates that either cholinergic neurons are not depleted in the cases examined or that compensatory axonal sprouting has occurred in conjunction with cell loss. In the Rett syndrome, in which ChAT and the vesicular acetylcholine transporter are reduced in various areas including cortex (Wenk et al., 1991; Wenk and Mobley, 1996), disruption of cholingeric innervation is likely due to developmental or degenerative neuronal abnormalities occurring before or shortly after birth in the absence of compensation. The contrast with autism is striking and elevated BDNF (brain-derived neurotrophic factor) in autism (see the following text) may play a role in maintaining cortical cholinergic projections. Of the two neurotrophins that control cholinergic function so far investigated in the basal forebrain, NGF (nerve growth factor) levels have shown no significant changes, but BDNF levels were significantly elevated (threefold) in the adult autistic brain (Perry et al., 2001). This increase is also apparent in the frontal cortex, where, although normal BDNF levels were lower than in the basal forebrain, levels were over fivefold higher in six autistic brains when compared with five controls (Perry et al., in preparation). These findings need to be interpreted with caution, because they are based on elisa assay alone and the contribution of nonspecific peptide activity cannot be excluded. If validated, however, increased BDNF in the basal forebrain and cortex could be interpreted variously. BDNF plays a role in sculpting synaptic connections. Because BDNF is upregulated by cholinergic activity in developing rat hippocampus (da Penha Berzaghi et al., 1993), it is possible that the abnormality is due to a transient developmental cholinergic hypertrophy. Another possible explanation for the elevation is that it reflects a regional compensatory mechanism (Hohmann and Berger-Sweeney, 1998; Hashimoto et al., 1999). It is also possible that overexpression of BDNF is an intrinsic component of the atypical brain development reported in autism. The findings of elevated blood levels of BDNF in autistic individuals aged up to adulthood (Nelson et al., 2001; Miyasaki et al., 2004) may indicate an intrinsic rather than compensatory mechanism. With regard to cholinergic receptors, binding to the M2 muscarinic receptor (which is located presynaptically on a variety of neuronal types) has been found to be normal in the frontal and parietal cortex in adult autism (Perry et al., 2001). In contrast, cortical muscarinic M1 receptor binding in the same areas has been reported to be modestly decreased (Perry et al., 2001). M1 receptor loss was apparent in frontal and parietal cortex and in both outer and inner cortical layers, reaching significance in the parietal cortex. This finding may be specific to autism because it was not observed in nonautistic mentally retarded individuals. Reduced M1 muscarinic receptor binding in the parietal cortex in autism indicates a specific abnormality in cholinoreceptive function because the M1 receptor is located postsynaptically. Nicotinic receptors are ligand-gated ion channels (Na+ and CA++) consisting of a variety of α and β subunits. The principal subtypes in human brain are the α4β2 heteromer and α7 homomer. With respect to nicotinic receptor binding, no alteration of α-bungarotoxin (αBT) binding in the cerebral cortex in autism has been so far reported (Perry et al., 2001). By contrast, in almost all cortical areas examined to date (frontal, parietal, and occipital), significant and extensive reductions of

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 313 Wednesday, February 15, 2006 3:25 PM

Neuroanatomical and Neurochemical Studies of the Autistic Brain

313

epibatidine binding (20 to 30% of the normal) have been apparent throughout the different cortical layers. Epibatidine binding has also been noted to be significantly reduced in a proportion of nonautistic mentally retarded individuals but not in those with Down’s syndrome. Similar to the M1 receptor abnormality, the nicotinic receptor abnormality reflects a reduction in receptor number not affinity. Recent studies suggest that the cortical nicotinic receptor abnormality may occur at the level of gene expression and raises the question of whether these are abnormalities of the alpha 4 subunit gene (CHRNA 4) on chromosome 20, the beta 2 subunit gene (CHRNB 2) on chromosome 1, or the gene promoters in autism (Martin-Ruiz et al., 2004). As in the cerebral cortex, no alteration in cerebellar ChAT or M2 receptor binding (Lee et al., 2002) has been found in the adult autistic brain. In the cerebellum, as in the cerebral cortex, presynatic cholinergic structures appear to be intact in autism, whereas nicotinic receptor changes are likely to reflect abnormalities in cholinoceptive neurons or in noncholinergic presynaptic structures. Muscarinic M1 and M2 receptors are not affected in the cerebellum in autism with the exception of significant elevation in M1 levels (Lee et al., 2002), which are normally extremely low in this brain area. This elevation may represent “vestigial” activity reflecting developmental abnormalities. In contrast to the cerebral cortex, there is a significant reduction of up to 50% in epibatidine binding in autism as compared to normal controls in all layers of the cerebellar cortex (Lee et al., 2002). This receptor-binding loss was not associated with a reduction in α4 mRNA (Martin-Ruiz et al., 2004). Further, αBT binding was elevated in the cerebellum in autism when compared to controls with a significant threefold increase in the granule cell layer. This α7 nicotinic receptor increase appears to be specific to autism, not being apparent to the same extent in a group of nonautistic mentally retarded cases. αBT binding predominantly reflects the α7 subunit. The gene encoding this subunit is located close to q11-15 on chromosome 15 (Chini et al., 1994) and is near to the locus associated with abnormalities in autism (Lamb et al., 2000). However, mRNA levels of α7 subunits in the cerebellum were not significantly altered in autism despite an upward trend (Martin-Ruiz et al., 2004). These data suggest a different etiopathology for the nicotinic receptor abnormalities in the cerebellar and cerebral cortices in autism. Diffuse α7 nAChR subunit immunoreactivity extends throughout the layers of the normal cerebellar cortex and the deep white matter. Purkinje cells have variable α7 immunoreactivity, sometimes extending into the apical dendrites. Granular immunoreactivity occurs in the neuropil surrounding the Purkinje cells and between α7 immunoreactive neurons in the deep cerebellar nuclei. In autistic cases examined, there is a reduction in diffuse α7 immunoreactivity throughout the layers of the cerebellar cortex. Purkinje cells and stellate cells also show decreased immunoreactivity. In contrast, granule cells in certain regions of the autistic brains showed an increase in α7 immunoreactivity. The apparent increase in α7 immunoreactivity in the granule layer may reflect compensatory upregulation during development. The α7 abnormalities in autism may be of interest in relation to improvements in recognition and in social behavior by a specific α7 agonist (Van Kampen et al., 2004).

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 314 Wednesday, February 15, 2006 3:25 PM

314

Understanding Autism: From Basic Neuroscience to Treatment

Epibatidine binding is, as in cortex and cerebellum, also reduced in the thalamus in adult autistic cases. In six cases (mean age 33 years) compared to nine controls (mean age 34 years), there was a 40% reduction (Perry et al., unpublished data). The human thalamus has been identified as a region of high nAChR expression (Spurden et al., 1997), and findings of reduced mean thalamic volume (Tsatsanis et al., 2003) and impairment of dentate-thalamo-cortical pathway (Muller et al., 1998) in autism indicate that it may be a relevant structure in the development of the disorder. In a semiquantitative immunohistochemical study of a α4, β2, and α7 nAChR subunits, reduced neuronal α7 and β2, IR was observed in the paraventricular (PV) and reuniens (Re) nuclei in the individuals with autism (n = 3) compared to control cases (n = 3) (Ray et al., in press). Reduced neuropil α7 IR was also evident in these regions. No changes in α7 or β2 IR were observed in other thalamic nuclei. An increase in the expression of α7-IR astrocytes in PV and Re was found in all three cases with autism, but increased expression of β2-IR astrocytes was only observed in one individual (with autism and epilepsy). Glutamic acid decarboxylase (GAD), the major rate-limiting enzyme in the synthesis of GABA, demonstrated low IR in PV, coexpressed with α7 in control and autistic cases, but was not reduced in autism. These results suggest that α7 and β2 nAChR deficits in PV and Re may contribute to the development of neurofunctional abnormalities in autism. PV and Re form part of the midline thalamus with reciprocal connections to multiple limbic regions. The findings therefore support previous evidence of limbic abnormalities in autism (Bauman, 1991; Haznedar et al., 2000; Raymond et al., 1996; Sweeten et al., 2002), and nAChR deficits in these nuclei may lead to features of dysmodulated sensory processing (Waterhouse et al., 1996). Proliferation of α7-IR astrocytes (in the thalamus and cortex) correlates with previous reports of increased GFAP in autism (Ahlsen et al., 1993; Bailey et al., 1998; Purcell et al., 2001; Rosengren et al., 1992) and suggests that astrocytes may contribute to the neuropathology of the disorder. Colocalization of GAD with α7 was not reduced in autism, indicating that loss of thalamic α7 in this disorder is not caused by loss of GABAergic neurons. As in all other areas so far examined, epibatidine binding to the nicotinic receptor is reduced in adult autism in the striatum, more so in putamen than caudate (Perry et al., in preparation). The nicotinic receptor abnormality thus appears to be part of a systemic, globally distributed pathology. Because nicotinic receptors are thought to play a particular role in regulating synaptic/dendritic plasticity, it is likely that the receptor reduction in autism relates to this aspect of neuronal function. Nicotinic receptors modulate synthesis of neurotrophins (such as NGF, BDNF, and FGF-2) and modify hippocampal plasticity in rodents (McGehee, 2002). In the development of the retinal ganglion cells for example, exposure to the nicotinic antagonist, curare, from early embryonic stages aborts dendritic proliferation (Semagor et al., 2001) and nicotinic cholinergic mechanisms play a critical role in synaptic plasticity in song birds (Salgado-Commissariat et al., 2004). Relationships between the loss of α4 nicotinic receptor subtype and synaptophysin identified in the cerebral cortex in another neurological disorder, Alzheimer’s disease (Sabbagh et al., 1998), suggest that the receptor loss in autism may be associated with abnormal synaptic morphology and function. In addition, loss of

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 315 Wednesday, February 15, 2006 3:25 PM

Neuroanatomical and Neurochemical Studies of the Autistic Brain

315

dendritic function, reflected in reduced MAP-2 IR in adult autism (MukaetovaLadinska et al., 2004) raises the question of further potential links between the nicotinic receptor loss and neuronal connectivity. Given the functions of the nicotinic receptor in normal brain, receptor loss in autism could relate to one or more of several clinical features associated with this disorder, such as attentional abnormalities, pain perception, anxiety, social interaction, epilepsy, or even conscious awareness (Perry and Perry, 2004). So far it does not appear that the receptor loss is associated with epilepsy (Perry et al., 2001). The cholinergic system has long been implicated in attention (reviewed in Sater and Bruno, 2000; Perry et al., 1999) with a specific role for the nicotinic receptor (Mirza and Stoleman, 2000). Nicotine administered in man improves performance in extended vigilance tasks, divided attention, and rapid information processing tasks (Wesnes and Warburton, 1984; Wesnes et al., 1983). Functional MRI indicates that nicotine alters neuronal activity in disturbed neural networks involving the anterior cingulate, superior frontal, superior parietal, and parahippocampal regions related to online task monitoring, attention, and arousal (Kumari et al., 2003). β2 receptor knockouts have shown impaired spatial learning (Zoli et al., 1999) as well as disrupted social and executive behavior believed to be reminiscent of autism (Granon et al., 2003) and sleep and arousal abnormalities (Lena et al., 2004). Reduced pain reactivity has been reported in autism (Tordjman et al., 1999). Nicotinic agents are analgesic (Bannon et al., 1998), and on the basis of a gene knockout model (Table 1 of Marubio et al., 1999) the α4 subunit has been implicated in the pain perception. In addition, α4 knockouts have shown increased levels of anxiety (Ross et al., 2000).

SUMMARY AND FUTURE DIRECTIONS At this point in time, there is accumulating evidence that the underlying biological processes involved in autism may be ongoing, resulting in changes in brain weight, brain volume, microscopic morphology, and probably neurochemical, features with age. Future research in genetics, neuroanatomy, and neurochemistry, as well as developmental neurobiology, will need to take into account and address these factors in order to better understand the underlying processes that form the basis of this complex disorder. A central and critical issue when trying to find genetic and neurochemical associations in neuropsychiatry is the exact phenotype being considered (Gottesmann, 2003). Genetic and neurochemical research in autism increasingly indicate that progress will be most rapid if components or dimensions of behavior are examined (McBride et al., 1996; Skuse, 2003; Leboyer et al., 2001). Much of the prior work needs to be reconsidered in this context, as it is becoming increasingly clear that the neurobiological complexity of autism may have hobbled research that has taken a more broadened categorical approach. There are now several important leads that merit investigation in autism research. Anatomically, the cause of and explanation for apparent brain overgrowth during early development may be an important clue, and deserves careful scrutiny. Does this observation relate to failure of apoptosis, abnormal synaptic structure and plasticity, atypical development of dendritic and/or axonal branching, or abnormal myelin composition and/or layering? Because there appears to be evidence of morphological

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 316 Wednesday, February 15, 2006 3:25 PM

316

Understanding Autism: From Basic Neuroscience to Treatment

changes with age, is the retention of fetal circuitry (as suggested by the changes in cell size and number in the septum and olivocerebellar regions) a viable hypothesis, or might there be alternative mechanisms to be considered? Although one of the most consistently reported brain abnormalities has been the reduced number of Purkinje cells in the cerebellum, it is not clear at this point as to whether these cells fail to migrate to their usual destination in the cerebellum or whether they disappear at a later time. Answers to these questions might provide clues to timing as well as underlying biological mechanisms. Given the current limitations of tissue quality and availability, especially from young autistic subjects, and the technical limitations of working with human material, the availability of animal models could significantly advance the field. In addition, newer biomedical techniques should be explored in order to better address the definition of potential cortical abnormalities, the role of minicolumns, dendritic and myelin structure, and synaptic plasticity. Some advances in technology are emerging, and these newer approaches will need to be applied. Although a better delineation of the neuroanatomical abnormalities in the autistic brain may provide leads to a clearer understanding of the developmental trajectory and some of the clinical characteristics of the ASD, this discipline should also be integrated with work being done in other neurobiological fields including neuroimmunology, neuroimaging, and neurochemistry. As discussed previously, neurochemical research has provided several important leads, the most compelling of which are alterations reported in serotonergic and cholinergic systems. The important roles that serotonin and acetylcholine play in early neurodevelopment, information processing, selective attention, and arousal may serve to further enhance interest in these systems in ASD. Clinically, the leads in these areas, as in other neurobiological systems, may be most fruitfully examined in concert with genetic studies and in a dimensional context. Given the uncertainties regarding the neurobiology of the autism spectrum disorders, exploratory research in a collaborative multidisciplinary environment should also be encouraged. From a practical perspective, neurochemical studies can often be efficiently and productively coupled with clinical trials because these trials may provide large groups of well-characterized individuals for study. In addition, well-designed, effective collaborations now typify the rapidly developing areas of transciptomics (mRNA expression array technology), proteomics, and metabolomics. Although individual fields in the neurosciences have made some substantial progress, especially over the past ten years, given the complexity of this disorder, it is highly likely that well-coordinated, interdisciplinary approaches will offer the best and most effective means of unraveling the many remaining questions related to the underlying neurobiology of ASD.

ACKNOWLEDGMENTS Margaret Bauman would like to thank Dr. Thomas Kemper and Dr. Gene Blatt for their support and collaboration. This work has been supported by a grant from the National Institute for Neurologic Diseases and Stroke (NINDS). Elaine Perry and Melissa Ray would like to thank Lorraine Hood for manuscript preparation. Their research in autism has been supported by Cure Autism Now. George Anderson is grateful for the support of the Korczak Foundation for Autism Research and the Gettner Research Foundation.

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 317 Wednesday, February 15, 2006 3:25 PM

Neuroanatomical and Neurochemical Studies of the Autistic Brain

317

REFERENCES Ahlsen, G., Rosengren, L.E., Belfrage, M. et al., Fibrillary acidic protein in the cerebrospinal fluid of children with autism and other neuropsychiatric disorders, Biol Psychiatry, 1993; 33: 734–743. Anderson, G.M., Monoamines in autism: an update of neurochemical research on a pervasive developmental disorder, Med Biol., 1987; 65: 67–74. Anderson, G.M., Genetics of childhood disorders: XLV. Autism, part 4: serotonin in autism, J Am Acad Child Adolesc Psychiatry, 2002; 12: 1513–1516. Anderson, G.M., Gutknecht, L., Cohen, D.J. et al., Serotonin transporter promoter variants in autism: functional effects and relationship to platelet hyperserotonemia, Mol Psychiatry, 2002; 7: 831–836. Anderson, G.M., Horne, W.C., Chatterjee, D. et al., The hyperserotonemia of autism, Ann N Y Acad Sci., 1990; 600: 331–340. Anderson, G.M., Minderaa, R.B., Cho, S.C. et al., The issue of hyperserotonemia and platelet serotonin exposure: a preliminary study, J Autism Dev Disord., 1989; 19: 349–351. Anderson, G.M., Zimmerman, A.W., Akshoomoff, N. et al., Autism clinical trials: biological and medical issues in patient selection and treatment response, CNS Spectrums, 2004; 9: 57–64. Antzoulatos, E. and Blatt, G.J., Density and distribution of 125I-labelled 5HT 2A/2C receptors in the anterior cingulate cortex in control and autistic individuals, Presented at the 3rd International Meeting for Autism Research (IMFAR), May 7–8, 2004, Sacramento, CA. Arin, D.M., Bauman, M.L., and Kemper, T.L., The distribution of Purkinje cell loss in the cerebellum in autism, Neurology, 1991; 41: 307 (abstract). Bailey, A., Luthert, P., Dean, A. et al., A clinicopathological study of autism, Brain, 1998; 121: 889–905. Bannon, A.W., Decker, M.W., Hilladay, M.W. et al., Broad-spectrum non-opioid analgesic activity by selective modulation of neuronal nicotinic acetylcholine receptors, Science, 1998; 279: 77–81. Barmack, N.H., Baughman, R.W., and Eckenstein, F.P., Cholinergic innervation of the cerebellum of the rat by secondary vestibular afferents, Ann NY Acad Sci., 1992; 656: 566–579. Bauman, M.L. and Kemper, T.L., The brain in infantile autism: a histoanatomic report, Neurology, 1984; 34: 275 (abstract). Bauman, M.L. and Kemper, T.L., Histoanatomic observations of the brain in early infantile autism, Neurology, 1985; 35: 866–874. Bauman, M.L. and Kemper, T.L., Neuroanatomic observations of the brain in autism, in Bauman, M.L. and Kemper, T.L., Eds., The Neurobiology of Autism, Baltimore, MD: Johns Hopkins University Press, 1994, 119–145. Bauman, M.L. and Kemper, T.L., Observations on the Purkinje cells in the cerebellar vermis in autism, J Neuropath Exp Neurol., 1996; 55: 613 (abstract). Bauman, M.L. and Kemper, T.L., Structural brain anatomy in autism: what is the evidence?, in Bauman, M.L. and Kemper, T.L., Eds., The Neurobiology of Autism, 2nd ed., Baltimore, MD: Johns Hopkins University Press, 2005, 121–135. Blatt, G.J., Fitzgerald, C.M., Guptill, J.T. et al., Density and distribution of hippocampal neurotransmitter receptors in autism: an autoradiographic study, J Autism Dev Disord., 2000; 31: 537–543. Carper, R.A., Moses, P., Tigue, Z.D. et al., Cerebral lobes in autism: early hyperplasia and abnormal age effects, Neuroimage, 2002; 16: 1038–1051.

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 318 Wednesday, February 15, 2006 3:25 PM

318

Understanding Autism: From Basic Neuroscience to Treatment

Casanova, M.F., Buxhoeveden, D.P., Switala, A.E. et al., Minicolumnar pathology in autism, Neurology, 2002; 58: 428–432. Chini, B., Raimond, E., Elgoyhen, A.B. et al., Molecular cloning and chromosomal localization of the human alpha 7-nicotinic receptor subunit gene (CHRNA 7), Genomics, 1994; 19: 379–381. Chrousos, G.P. and Gold, P.W., The concepts of stress and stress system disorders: overview of physical and behavioral homeostasis, JAMA, 1992; 267: 1244–1252. Cook, E.H., Leventhal, B.L., and Freedman, D.X., Free serotonin in plasma: autistic children and their first-degree relatives, Biol Psychiatry, 1988; 24: 488–491. Cook, E.H., Autism: review of neurochemical investigation, Synapse 1990; 6: 292–308. Courchesne, E., Karns, C.M., Davis, H.R. et al., Unusual brain growth patterns in early life in patients with autistic disorder, Neurology, 2001; 57: 245–254. Court, J.A., Perry, E.K., Spurden, D. et al., The role of the cholinergic system in the development of the human cerebellum, Dev Brain Res., 1995; 90: 159–167. da Penha Berzaghi, M., Cooper, J., Castren, E. et al., Expression of alpha 7 nicotinic receptors during postnatal development of the cerebellum, Brain Res Dev Brain Res., 1997; 98: 125–133. Eccles, J.C., Ito, M., and Szentagothai, J., The Cerebellum as a Neuronal Machine, New York: Springer-Verlag, 1967. Ernst, M., Zametkin, A.J., Matochik, J.A. et al., Low medial prefrontal dopaminergic activity in autistic children, Lancet, 1997; 350(9078): 638. Fatemi, H., Reelin mutations in mouse and man: from reeler mouse to schizophrenia, mood disorders, autism and lissencephaly, Mol Psychiatry, 2001; 6: 129–133. Gottesmann, I.I. and Gould, T.D., The endophenotype concept in psychiatry: etymology and strategic intentions, Am J Psychiatry, 2003; 160: 636–645. Granon, S., Faure, P., and Changeux, J.P., Executive and social behaviors under nicotinic receptor regulation, Proc Natl Acad Sci U S A, 2003; 100: 9596–9601. Greenfield, J.G., The Spino-Cerebellar Degenerations, Springfield, IL: CC Thomas, 1954. Hashimoto, Y., Abiru, Y., Nishio, C. et al., Synergistic effects of brain-derived neurotrophic factor and ciliary neurotrophic factor on cultured basal forebrain cholinergic neurons from postnatal 2-week old rats, Brain Res Dev Brain Res., 1999; 115: 25–32. Haznedar, M.M., Buchsbaum, M.S., Wei, T.C. et al., Limbic circuitry in patients with autism spectrum disorders studied with positron emission tomography and magnetic resonance imaging, Am J Psychiatry, 2000; 157: 1994–2001. Hellstrom-Lindahl, E. and Court, J.A., Nicotinic acetylcholine receptors during prenatal development and brain pathology in human aging, Behav Brain Res., 2000; 113: 159–168. Hellstrom-Lindahl, E., Gorbounova, O., Seiger, A. et al., Regional distribution of nicotinic receptors during prenatal development of human brain and spinal cord, Brain Res Dev Brain Res., 1998; 108: 147–160. Herbert, M.R., Ziegler, D.A., Makris, N. et al., Localization of white matter volume increase in autism and developmental language disorder, Ann Neurol., 2004; 55: 530–540. Hohmann, C.F. and Berger-Sweeney, J., Cholinergic regulation of cortical development and plasticity: new twists to an old story, Perspect Dev Neurobiol., 1998; 5: 401–425. Holmes, G. and Stewart, T.G., On the connection of the inferior olives with the cerebellum in man, Brain, 1908; 31: 125–137. Janusonis, S., Gluncic, V., and Rakic, P., Early serotonergic projections to Cajal-Retzius cells: relevance for cortical development, J Neurosci., 2004; 24: 1652–1659. Kanner, L., Autistic disturbances of affective contact, Nerv Child, 1943; 2: 217–250.

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 319 Wednesday, February 15, 2006 3:25 PM

Neuroanatomical and Neurochemical Studies of the Autistic Brain

319

Kemper, T.L. and Bauman, M.L., Neuropathology of infantile autism, J Neuropath Exp Neurol., 1998; 57: 645–652. Kemper, T.L., Neuropathology of autism, in Golden, J.A. and Harding, B.N., Eds., Developmental Neuropathology, Basel, Switzerland: International Society of Neuropathology, 2004; 40: 349–352. Koul, O., Myelin and autism, in Bauman, M.L. and Kemper, T.L., Eds., The Neurobiology of Autism, 2nd ed., Baltimore, MD: Johns Hopkins University Press, 2005, pp. 150–163. Kumari, V., Gray, J.A., Fytche, D.H. et al., Cognitive effects of nicotine in humans: and fMRI study, Neuroimage, 2003; 19: 1002–1003. Lainhart, J.E., Piven, J., Wzorek, M. et al., Macrocephaly in children and adults with autism, J Am Acad Child Adolesc Psychiatry, 1997; 36: 282 Lamb, J.A., Moore, J., Bailey, A. et al., Autism: recent molecular genetic advances, Hum Mol Genet., 2000; 4: 861–865. Lan, C.T., Wen, C.Y., Tan, C.K. et al., Multiple origins of cerebellar cholinergic afferents from lower brainstem in the gerbil, J Anat., 1995; 186: 549–561. Lawrence, Y., Kemper, T.L., Bauman, M.L. et al., Increased density of parvalbumin labeled hippocampal interneurons in autism (abstract), IMFAR, Boston, MA, May 5, 2005. Leboyer, M., Bellivier, F., Nosten-Bertrand, M. et al., Psychiatric genetics: search for phenotypes, Trends Neurosci., 1998; 21: 102–105. Lee, M., Martin-Ruiz, C., Graham, A. et al., Nicotinic receptor abnormalities in the cerebellar cortex in autism, Brain, 2002; 125: 1483–1495. Lena, C., Popa, D., Grailhe, R. et al., Beta 2-containing nicotinic receptors contribute to the organization of sleep and regulate putative micro-arousals in mice, J Neurosci., 2004; 24: 5711–5718. Lotspeich, L.J. and Ciaranello, R.D., The neurobiology and genetics of infantile autism, Int Rev Neurobiol., 1993; 35: 87–129. Martin-Ruiz, C.M., Lee, M., Perry, R.H. et al., Molecular analysis of nicotinic receptor expression in autism, Brain Res Mol Brain Res., 2004; 123: 81–90. Martineau, J., Herault, J., Petit, E. et al., Catecholaminergic metabolism and autism, Dev Med Child Neurol., 1994; 8: 688–697 Marubio, L.M., del Mar Arroyo-Jamenez, M., Cordero-Erausquin, M. et al., Reduced antinociception in mice lacking neuronal nicotinic receptor subunits, Nature, 1999; 398(6730): 805–810. McBride, P.A., Anderson, G.M., Hertzig, M.E. et al., Serotonergic responsivity in male young adults with autistic disorder: results of a pilot study, Arch Gen Psychiatry, 1989; 46: 213–221. McBride, P.A., Anderson, G.M., and Shapiro, T., Autism research: bringing together approaches to pull apart the disorder, Arch Gen Psychiatry, 1996; 53: 980–983. McGehee, D.S., Nicotinic receptors and hippocampal synaptic plasticity … it’s all in the timing, Trends Neurosci., 2002; 25: 171–172. Minderaa, R.B., Anderson, G.M., and Volkmar, F.R. et al., Neurochemical study of dopamine functioning in autistic and normal subjects, J Am Acad Child Adolesc Psychiatry, 1989; 28: 200–206. Minderaa, R.B., Anderson, G.M., and Volkmar, F.R. et al., Noradrenergic and adrenergic functioning in autism, Biol Psychiatry, 1994; 36: 237–241. Minshew, N.J., Goldstein, G., Siegel, D.J. et al., Neuropsychological functioning in autism: profile of a complex information processing disorder, J Int Neuropsychol Soc., 1997; 3: 303–316. Mirza, N.R. and Stoleman, I.P., The role of nicotinic and muscarinic acetylcholine receptors in attention, Psychopharmacology (Berl), 2000; 148: 243–250.

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 320 Wednesday, February 15, 2006 3:25 PM

320

Understanding Autism: From Basic Neuroscience to Treatment

Miyazaki, K., Narita, N., Sakuta, R. et al., Serum neurotrophin concentrations in autism and mental retardation: a pilot study, Brain Dev., 2004; 26: 292–295. Mosconi, M., Cody, H., Poe, M. et al., Amygdala and hippocampus enlargement in young children with autism, IMFAR, Boston, MA, May 6, 2005. Mukaetova-Ladinska, E.B., Arnold, H., Jaros, E. et al., Depletion of MAP-2 expression and laminar cytoarchitectonic changes in dorsolateral prefrontal cortex in adult autistic individuals, Neruopathol Appl Neurobiol., 2004; 30: 615–623. Mulder, E.J., Anderson, G.M., Kema, I.P. et al., Platelet serotonin in pervasive developmental disorders and mental retardation: diagnostic group differences, within-group distribution and behavioral correlates, J Am Acad Child Adolesc Psychiatry, 2004; 43: 491–499. Muller, R.A., Chugani, D.C., Behen, M.E. et al., Impairment of dento-thalamo-cortical pathway in autistic men: language activation data from positron emission tomography, Neurosci Lett., 1998; 245: 1–4. Narayan, M., Srinath, S., Anderson, G.M., and Meundi, D.B., Cerebrospinal fluid levels of homovanillic acid and 5-hydroxyindoleacetic acid in autism, Biol Psychiatry, 1993; 33: 630–635. Nelson, K.B., Grether, J.K., Crien, L.A. et al., Neuropeptides and neurotrophins in neonatal blood of children with autism or mental retardation, Ann Neurol., 2001; 49: 597–606. Norman, R.M., Cerebellar atrophy associated with etat marbre of the basal ganglia, J Neurol Psychiatry, 1940; 3: 311–318. Novotny, S., Hollander, E., Allen, A. et al., Increased growth hormone response to sumatriptan challenge in adult autistic disorders, Psychiatry Res., 2000; 94: 173–177. Perry, E.K., Walker, M., Grace, J. et al., Acetylcholine in mind: a neurotransmitter correlate of consciousness? Trends Neurosci., 1999; 22: 273–280. Perry, E.K., Lee, M.L., Martin-Ruiz, C.M. et al., Cholinergic activity in autism: abnormalities in the cerebral cortex and basal forebrain, Am J Psychiatry 2001; 158: 1058–1066. Perry, E.K. and Perry, R.H., Neurochemistry of consciousness: cholinergic pathologies in the human brain, Prog Brain Res., 2004; 145: 287–299. Purcell, A.E., Jeon, O.H., Zimmerman, A.W. et al., Postmortem brain abnormalities of the glutamate neurotransmitter system in autism, Neurology 2001; 57: 1618–1628. Rakic, P. and Sidman, R.L., Histogenesis of the cortical layers in human cerebellum particularly in the lamina dissecans, J Comp Neurol., 1970; 139: 473–500. Rakic, P., Neuron-glia relationship during granule cell migration in developing cerebellar cortex: a Golgi and electron microscopic study in Macacus rhesus, J Comp Neurol., 1971; 141: 282–312. Raymond, G.V., Bauman, M.L., and Kemper, T.L., Hippocampus in autism: a Golgi analysis, Acta Neuropathologica, 1996; 91: 117–119. Ritvo, E.R., Freeman, B.J., Scheibel, A.B. et al., Lower Purkinje cell counts in the cerebella of four autistic subjects: Initial findings of the UCLA-NSAC autopsy research report, Am J Psychiatry, 1986; 146: 862–866. Rodier, P.M., Ingram, J.L., Tisdale, B. et al., Embryological origins of autism: developmental abnormalities of the cranial nerve nuclei, J Comp Neurol., 1996; 36: 351–356. Rosene, D.L. and van Hoesen, G.W., The hippocampal formation in the primate brain, in Cerebral Cortex, Jones, E.G. and Peters, A., Eds., Vol. 6, Plenum Press, New York, 1987, pp. 345–450. Rosengren, L.E., Ahlsen, G., Belfrage, M. et al., A sensitive ELISA for glial fibrillary acidic protein: application in CSF in children, J Neurosci Methods, 1992; 44: 113–119.

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 321 Wednesday, February 15, 2006 3:25 PM

Neuroanatomical and Neurochemical Studies of the Autistic Brain

321

Ross, S.A., Wong, J.Y., Clifford, J.J. et al., Phenotypic characterization of an alpha 4 neuronal nicotinic acetylcholine receptor subunit knock-out mouse, J Neurosci., 2000; 20: 6431–6441. Sabbagh, M.N., Reid, R.T., Corey-Brown, J. et al., Correlation of nicotinic binding with neurochemical markers in Alzheimer’s disease, J Neural Transm., 1998; 105: 709–717. Salgado-Commissariat, D., Rosenfield, D.B., and Helekar, S.A., Nicotine-mediated plasticity in robust nucleus of the archistriatum of the adult zebra fish, Brain Res., 2004; 1018: 97–105. Sarter, M. and Bruno, J.P., Cortical cholinergic inputs mediating arousal, attentional processing and dreaming: differential afferent regulation of the basal forebrain by telencephalic and brainstem afferents, Neuroscience, 2000; 95: 933–952. Schain, R.J. and Freedman, D.X., Studies on 5-hydroxyindole metabolism in autistic and other mentally retarded children, J Pediatr., 1961; 58: 315–320. Schultz, R.T. and Anderson, G.M., The neurobiology of autism and the pervasive developmental disorders, in Neurobiology of Mental Illness, 2nd ed., Eds., Nestler, E. and Charney, D., Oxford University Press, New York, 2004, pp. 954–967 Schumann, C.M., Hamstra, J., Goodlin-Jones, B.L. et al., The amygdala is enlarged in children but not adolescents with autism: the hippocampus is enlarged at all ages, J Neurosci., 2004; 24(28): 6392–6401. Schumann, C. and Amaral, D., No difference in the number of neurons in the amygdala in postmortem cases of autism: a stereologic study, IMFAR, Boston, MA, May 6, 2005. Semagor, E., Eglen, S.J., and Wong, R.O.L., Development of retinal ganglon cell structure and function, Prog Ret and Eye Res., 2001; 20: 139–174. Skuse, D.H., Endophenotypes and child psychiatry, Br J Psychiatry, 2001; 178: 395–396. Sparks, B.F., Friedman, S.D., Shaw, D.W. et al., Brain structural abnormalities in young children with autism spectrum disorder, Neurology, 2002; 59(2): 184–192. Spurden, D.P., Court, J.A., Lloyd, S. et al., Nicotinic receptor distribution in the human thalamus: autoradiographical localization of (3H) nicotine and (125 l) a-bungarotoxin binding, J Chem Neuroanatomy, 1997; 13: 105–113. Sweeten, T.L., Posey, D.J., Shekhar, A. et al., The amygdala and related structures in the pathophysiology of autism, Pharmacol Biochem Behav., 2002; 71: 449–455. Tice, M.A., Hashemi, T., Taylor, L.A. et al., Distribution of muscarinic receptor subtypes in rat brain from postnatal to old age, Brain Res Dev Brain Res., 1996; 92: 70–76. Tordjman, S., Anderson, G.M., McBride, P.A. et al., Plasma B-endorphin, adrenocorticotropin hormone, and cortisol in autism, J Child Psychol Psychiatry, 1997; 38: 705–715. Tordjman, S., Antoine, C., Cohen, D.J. et al., Study of the relationships between self-injurious behavior and pain reactivity in infantile autism, Encephale, 1999; 25: 122–134. Tsatsanis, K.D., Rourke, B.P., Klin, A. et al., Reduced thalamic volume in high-functioning individuals with autism, Biol Psychiatry, 2003; 53: 121–129. Van Kampen, M., Selbach, K., Schneider, R. et al., AR-R 17779 improves social recognition in rats by activation of nicotinic alpha 7 receptors, Psychopharmacology (Berl), 2004; 172: 375–383. Waterhouse, L., Fein, D., and Modahl, C., Neurofunctional mechanisms in autism, Psychol Rev., 1996; 103: 457–489. Wenk, G.L., Naidu, S., Casanova, M.F. et al., Altered neurochemical markers in Rett’s syndrome, Neurology, 1991; 41: 1753–1756.

© 2006 by Taylor & Francis Group, LLC

2732_C014.fm Page 322 Wednesday, February 15, 2006 3:25 PM

322

Understanding Autism: From Basic Neuroscience to Treatment

Wenk, G.L. and Mobley, S.L., Choline acetyltransferase activity and vesamicol binding in Rett syndrome and in rats with nucleus basalis lesions, Neuroscience 1996; 73: 79–84. Wesnes, K., Warburton, D.M., and Matz, B., Effects of nicotine on stimulus sensitivity and response bias in a visual vigilance task, Neuropsychology, 1983; 9: 41–44. Wesnes, K. and Warburton, D.M., Effects of scopolamine and nicotine on human rapid information processing performance, Psychopharmacology (Berl), 1984; 82: 147–150. Yakovlev, P.I. and Lecours, A.R., Myelogenetic cycles of regional maturation of the brain, in Regional Development of the Brain in early Life, Minkowski, A., Ed., 1967, 3–70. Zoli, M., Picciotto, M.R., Ferrari, R. et al., Increased neurodegeneration during aging in mice lacking high-affinity nicotine receptors, Embo J., 1999; 18: 1235–1244.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 323 Wednesday, February 15, 2006 3:25 PM

Social Brain in Autism: 15 The Perspectives from Neuropsychology and Neuroimaging Robert T. Schultz, Katarzyna Chawarska, and Fred R. Volkmar CONTENTS Introduction............................................................................................................323 Social Deficits as the Hallmark of Autism ...........................................................324 Theoretical Understanding of Social Dysfunction in Autism .......................325 Developmental Aspects of Social Deficits in Autism....................................326 Joint Attention and Gaze Monitoring in Autism ......................................327 Brain Mechanisms in Autism ................................................................................328 Studies of Social Perception ..........................................................................329 Face Identity Perception ............................................................................330 Neural Bases of Face Recognition ............................................................331 Facial Expression Perception ....................................................................332 The Neural Basis of Facial Expression Perception ..................................333 Studies of Social Cognition ...........................................................................334 Functional Neuroimaging Studies of Social Cognition ............................334 Social Motivation ...........................................................................................335 Summary and Conclusions....................................................................................337 Acknowledgments..................................................................................................337 References..............................................................................................................338

INTRODUCTION Autism is part of a spectrum of disorders characterized by a triad of symptoms, including deficits in social relatedness and communication and a variety of behavioral problems such as restricted interests, sensory sensitivities, and repetitive behaviors, all with onset before age 3 years.1 Autism is recognized as a heterogeneous disorder and as part of a continuum of disability shared with Asperger syndrome and

323

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 324 Wednesday, February 15, 2006 3:25 PM

324

Understanding Autism: From Basic Neuroscience to Treatment

pervasive developmental disorder not otherwise specified (PDDNOS). Collectively, these syndromes are referred to as the autism spectrum disorders (ASDs), with the belief that the underlying neurobiological bases are shared to some degree and that characterization of individual cases is best conceptualized along one or more continuous dimensions of severity. Asperger syndrome involves autistic-like disturbance in social reciprocity, but it does not share the developmental disability of communication and language; language can in some respects be a source of strength for the person. PDDNOS, on the other hand, is a subsyndromal manifestation of autism. This does not necessarily imply that it is a milder form of autism; indeed, many with PPD NOS will have significant life difficulties.2,3 PDDNOS may well be the most heterogeneous of the subtypes, with a wide range of etiologies, all converging to some extent on the final common pathway involved in core autistic symptomatology.3,4 In his original 1943 description of autism, Leo Kanner emphasized that social deficits and difficulties are a central feature of the disorder.5 Kanner showed, in contrast to typically developing infants, how much more the children he described were interested in the nonsocial world. The subsequent 60 years of research has further strengthened our confidence in Kanner’s original observation. The social dysfunction of autism appears to be the single-most robust defining feature of the condition.6,7 The key diagnostic criteria that define the social deficits include a failure to develop peer relationships appropriate to developmental level; abnormal emotional intonations in voice and speech; impairment in the use of nonverbal behaviors such as eye-to-eye gaze, facial expression, body postures, and gestures to regulate social interaction; and failure to spontaneously seek to share enjoyment, interests, or achievements with other people (e.g., by a lack of showing, bringing, or pointing out objects of interest).1 In this chapter, we review aspects of the growing body of work on the social deficits in autism. We begin with a brief review of the nature of social difficulties in autism and then summarize current psychological and neuropsychological perspectives on these difficulties. Next, we review the development of important skills in infancy that lay the foundation for social relatedness. Finally, we review progress in neuropsychology and neuroimaging that has significantly advanced our knowledge and, for the first time, provided us with the potential for integrating clinical observation and theory relative to specific brain systems and mechanisms.

SOCIAL DEFICITS AS THE HALLMARK OF AUTISM Beginning with Kanner’s original report, social deficits have been a (if not the) defining feature of autism.5 Social difficulties have been emphasized in all of the various official definitions of the disorder since DSM-III.1,5,8 Although some social skills develop over time in individuals with autism, they are invariably delayed as well as deviant.2 Statistical approaches, such as signal-detection methods, have also emphasized the centrality of social deficits.9 In contrast to other symptom domains in the diagnostic criteria, social deficits are the most evident in autism and less shared with other neuropsychiatric disabilities.1,6,9 In addition to the categorical approach exemplified by the DSM, dimensional approaches have been used to

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 325 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

325

characterize the social difficulties in autism. Dimensional approaches have some advantages in terms of providing more detailed information, such as the relative levels of severity.10

THEORETICAL UNDERSTANDING

OF

SOCIAL DYSFUNCTION

IN

AUTISM

Several attempts have been made to elaborate psychological models for understanding social dysfunction in autism. Such models, and the practical attempts to apply them in specific experimental paradigms or assessment instruments, are of considerable interest in that they may provide a basic measure of the disability, which could then potentially help define core endophenotypes. Such endophenotypes are sorely needed in translational neuroimaging and genetics work on autism. The predominant theories have focused on putative “core” (i.e., causative) aspects of the social deficit. This work builds on a body of earlier work in which any number of processes were thought to give rise to the social difficulty in autism. For example, problems in the area of attention,11,12 perception,13 or language14 have each been the focus of interest at various points in history. Often, however, such theoretical approaches are applicable only to a subset of the entire spectrum of autism; for example, about one third of individuals with autism have reasonably good language ability (apart from their pragmatic and prosodic difficulties), and yet, such persons remain significantly socially disabled.15 Three theoretical approaches in particular have been the focus of great interest for the field — executive dysfunction, weak central coherence, and ToM deficits. The executive function model posits that the difficulties in autism arise as a result of disorder in those processes that allow the individual to maintain an appropriate focus to attain a desired goal. This model is particularly relevant to the problems in forward planning, set shifting, and dealing with novelty and unpredictability so characteristic of individuals with autism at all levels of ability (see Ozonoff et al. for a review).16 However, several different observations pose significant obstacles for this hypothesis. For example, difficulties in executive function are not unique to autism but seen in a range of disorders.17 Furthermore, it appears that the degree of social disability is not straightforwardly related to executive function deficits.16,17 However, deficits in executive functioning do appear to predict adaptive behavior difficulties.18 A second major theory has focused on what has been labeled weak central coherence, i.e., the capacity to integrate information meaningfully into a coherent whole.19 This theory emphasizes an autistic cognitive style that is biased toward local rather than global information processing. This bias results in characteristic deficits as well as the special skills that are exhibited in autism.20 Empirical work from this point of view has tended to focus on perceptual processing and grouping, such as the predilection in autism for focusing on local levels of detail at the expense of the global level.21 Moreover, the theory correctly points out that any explanation of autism must address the splinter skill and areas of strength (e.g., the prevalence of savant or savant-like spikes in the cognitive profile), in addition to explaining areas of impairment. The theory explains the social deficits in autism as stemming from difficulties in integrating local-level information with global-level information,

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 326 Wednesday, February 15, 2006 3:25 PM

326

Understanding Autism: From Basic Neuroscience to Treatment

such as the need to appreciate social context to modulate social behavior. Although this is a very influential heuristic model, the experimental literature relevant to this hypothesis is still limited and, at times, equivocal.22–24 Moreover, one could easily argue that the problems in central coherence in autism arise from social deficits, rather than the reverse. For example, in the absence of the early drive for social mastery, children with autism lack the important social “templates” that guide later development.25 Perhaps, the most widely discussed theoretical account of social dysfunction has centered on posited deficits in theory-of-mind (ToM) skills, i.e., the capacity of both self and others to think about social relations.26,27 It has been a very important framework to guide experimentation and theory, as it has stimulated important advances in thinking in the field about the primary deficits in autism, and it has refocused interest on the developmental unfolding of autism. However, problems arise in several respects with this hypothesis. For example, such skills may have an independent relationship to language ability (an obvious area of major deficit in autism), apart from their role in social abilities. In fact, many higher-functioning individuals with autism or Asperger syndrome may be able to solve laboratory-based ToM tasks quite nicely, and yet, they remain significantly impaired socially.28,29 In addition, it would appear that the severity and pervasiveness of the social difficulties in autism are not consistent with what is known about the development of ToM abilities, i.e., these problems arise in very young infants before ToM abilities (as typically defined) are first evident.30 These criticisms aside, it is clear that these and other theories have stimulated an important body of work and that they continue to provide a framework for important new contributions to the literature.

DEVELOPMENTAL ASPECTS

OF

SOCIAL DEFICITS

IN

AUTISM

Early studies of social development in autism were based almost entirely on parental report rather than child observation.31 Subsequent work using direct observation has refined the understanding of the nature of these difficulties, and it is clear that some skills are more impaired than others.32,33 Social difficulties in infants and toddlers may include failure to develop reciprocal eye contact and social engagement; at the same time, the infant may be overly sensitive to the inanimate environment.34 The human face and voice are of much less interest, in contrast to typically developing infants and toddlers.35 Retrospective analysis of home movies and videotapes of infants as young as 6 to 12 months of age who are later diagnosed with autism show that they exhibit significantly fewer social behaviors and appear much less aware of social signals and nonverbal social communications as expressed through tone of voice and facial expression.36,37 To date, there are no prospective, longitudinal studies of young infants at risk for autism (e.g., siblings of a child with autism), and it is clear that these are now sorely needed to clarify the unfolding of the these early social difficulties.34 Infants and very young children with autism may seem to hold themselves apart from social interaction, but some social skills do emerge with increased cognitive ability; e.g., selective attachments to parents as measured in the Ainsworth strangesituation task.38 Increased social interest in school-aged and adolescent individuals

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 327 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

327

with autism is typical, although social skills remain significantly impaired.39 For adults who are independent and self sufficient, social oddities, eccentricity, and significant social deficits remain common.40 Wing has outlined a general progression from social aloofness to passive acceptance of social interaction to, in older and more able individuals, an eccentric social style that she characterizes as “active but odd.”40 Unfortunately, important aspects of this progression remain poorly understood.39 Increasingly, individuals with autism who have received intensive treatment have been more able to attain adult self-sufficiency and independence. Even in such individuals, however, social deficits remain a source of significant disability.41 It is the case that individuals with Asperger syndrome appear, on average, to be much more able as adults to engage in serious and significant social relationships and often marry and form a family.41,42 Increasingly, intervention techniques have centered on social skills development, but relatively few systematic studies of such interventions have been conducted.43 Joint Attention and Gaze Monitoring in Autism Joint Attention A deficit of joint attention (JA) is one of the defining criteria of autism in most diagnostic instruments as well as in measures developed for the screening for autism in infancy. On the behavioral level, JA involves episodes in which two people share attention to an object of mutual interest. These episodes may occur when a child follows the attention of an adult by simply “looking where someone else is looking” (gaze monitoring) or following communicative gestures of others, such as pointing.44–46 The child may also initiate such an episode by directing attention of an adult to objects by shifting gaze between the objects and the adult, showing them or pointing to them to communicate interest (protodeclarative pointing).47 Initiating and responding to the attention of others, although related, appear to reflect distinct psychological processes.44,48 The ability to respond to others’ bids for attention precedes the emergence of the ability to initiate such bids, and, in ontogenesis, each of them predicts a different aspect of language development.48–51 On the conceptual level, JA behaviors reflect an understanding of other people as intentional, goal-oriented entities.52–54 In this sense, JA constitutes a first step in the ontogeny of social cognition and the development of ToM.55,56 Although infants exhibit some rudimentary forms of JA early in the second year of life, these behaviors are at first infrequent and highly context dependent, and their conceptual basis is not fully understood. It is not until about 18 to 19 months of age that these skills become more robust and functional across contexts.52,53 The ability to share attention with others in early development provides the foundation for communicative and social-cognitive development.50,57 JA episodes have high functional significance for early language development, including comprehension of language,48 production of verbal and nonverbal communicative behaviors,58–60 and novel word learning.61 JA has also been implicated in the phenomenon of social referencing in which emotional information about an ambiguous object or event is conveyed from adult to the infant.62

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 328 Wednesday, February 15, 2006 3:25 PM

328

Understanding Autism: From Basic Neuroscience to Treatment

Gaze Monitoring Gaze monitoring, or the ability to follow the visual regard of others, provides a foundation for a range of social-cognitive skills in infancy,50,57 and early language development.48,58–61 On the behavioral level, gaze monitoring (or gaze following) simply signifies “looking where someone else is looking”44; on the conceptual level, it represents understanding of the attentional significance of gaze and the fact that gaze can express others’ thoughts, emotions, and intentions.52 Deficits in spontaneous gaze monitoring in autism are present very early in development and constitute one of the key symptoms that differentiate ASDs from global and specific developmental delays in children of 36 months and younger.37,63–70 This deficit is relatively stable over time, and even older and higher-functioning individuals with ASD continue to have marked difficulties monitoring gaze in naturalistic settings and fail to use gaze direction to infer another’s object of regard.71–76 Development of Gaze Monitoring Faces and eyes have significance to infants from very early in life. Newborn babies preferentially track moving, face-like patterns,77 orient more frequently to face-like stimuli as compared to non-face-like patterns,78 prefer to fixate faces with a direct rather than averted gaze,79,80 and show a rudimentary form of sensitivity to changes in gaze direction.81 Whereas infants in the first month of life pay more attention to the high-contrast edge area of the face, by two months of age, they preferentially scan the region of the eyes.82–84 Although not capable of following the gaze of others spontaneously, four-month-old infants perceive the movement inherent in gaze shift as a directional cue.85–87 That is, when tested in a spatial cueing attention paradigm, infants have shorter saccadic reaction times (SRT) to peripheral targets appearing in locations congruent with the eye-gaze direction of the cue (a stimulus face immediately preceding a target presentation) than to targets appearing in incongruent locations. Eight-month-old infants do not usually follow gaze spontaneously, but when provided with a contingent feedback, they are able to learn that a shift in a person’s direction of head and gaze predicts a location where an interesting event will occur (e.g., a toy will appear from a box on the left if a person in front of the child turns her head and looks in that location).8 By 10 to 11 months, infants follow head and gaze turn spontaneously,88 but it is not until 18 to 19 months of age that the infants exhibit appreciation of the significance of an eye-gaze shift alone.53,89 It has been hypothesized that onset, at this age, of gaze monitoring, which is relatively frequent and independent of the presence or absence of the targets in the visual field, represents a conceptual development toward the understanding that gaze shift signifies attentional shift of focus of another person.53 The conceptual understanding of attentional significance of gaze continues to develop well into preschool age.89

BRAIN MECHANISMS IN AUTISM There is a large and growing body of evidence characterizing the pathophysiology of ASDs. One important issue is whether the neural abnormality is widespread or discrete. Because the triad of symptoms that define autism are rather broad in nature, one would presume that the syndrome affects a widely distributed set of neural

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 329 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

329

systems. It is even possible that autism is caused by some pervasive impairment that affects every neuron, synapse, or functional cortical minicolumn. However, the extreme form of this hypothesis seems unlikely because there are also areas of spared function in autism, for example, basic perceptual skills, as well certain cognitive skills. In fact, severe autism is not incompatible with average or even above-average general intelligence. The brain features mediating individual differences in intelligence, therefore, must be essentially orthogonal to the processes that cause autism; otherwise, it would be impossible to have “high-functioning” autism. In other words, social intelligence and traditional notions of cognitive intelligence are probably mediated by quite separate neural systems. This is important in that it suggests that not every system in the brain needs to be impaired in autism, and in fact, it seems clear that many functional systems are intact. The perspective adopted in this review is that autism is defined on the basis of a select set of disturbances, each of which in principal should map onto discrete functional systems in the brain. This makes functional neuroimaging methodologies well suited for the study of autism and specific neuropsychological processes that have been found to be impaired among those with autism. Functional neuroimaging procedures such as 15O-water positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) are ideally suited for studying in detail the separate neural systems that govern select sensory and cognitive domains. Both techniques rely on the fact that when specific brain areas are engaged (and are more active relative to some baseline set of circumstances), there is an increase in blood flow to those neural regions. fMRI is predicated on the fact that there is a local increase in oxygenated hemoglobin in the capillary beds in the midst of activate neurons, and this increase results in localized increase in MRI signal (known as the blood-oxygen-level-dependent signal or the BOLD signal). fMRI is far and away the most popular functional technique for investigating the processing capacities of discrete brain areas and distributed brain systems. This is probably due to several factors, including a slight advantage for fMRI vs. PET in spatial resolution and the fact that the MRI equipment is widely available. In addition, fMRI is noninvasive and involves only minimal risk. 15O-water PET, on the other hand, requires the use of a radioactive tracer to map brain blood flow, glucose metabolism, or specific neurotransmitter properties. Even though the levels of radiation exposure with PET are small, it is not allowable on a research basis for use with minors, but it nevertheless represents an excellent technique for adults. This section reviews current knowledge from the fMRI and PET literature on neural system dysfunction related to the social impairment in autism. From a neuropsychological perspective, the prevailing belief is that these aberrant social behaviors are due to deficits in (1) social perception, for example, recognizing personal identity from the face and reading facial expressions, (2) social cognition, for example, perspective taking and ToM, and (3) social motivation. Thus, this review will be organized according to this scheme.

STUDIES

OF

SOCIAL PERCEPTION

Social perception entails decoding social communicative inputs in the auditory, somatosensory, and visual processing domains. These nonverbal communications

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 330 Wednesday, February 15, 2006 3:25 PM

330

Understanding Autism: From Basic Neuroscience to Treatment

can take several forms and are regularly used during social interactions to facilitate reciprocal social understanding. For example, one gains an appreciation of another person’s state of mind by decoding patterns of emotional intonation in the voice, by reading emotional displays on the face, and through dynamic changes in body posture, and in more limited circumstances, by interpreting the meaning of touch. Face and prosody perception have both been studied vis-à-vis the social deficits in autism. However, with respect to functional neuroimaging, much more work has been done on face perception, and so this will be the focus of this review. Face perception can be subdivided into (1) recognition of facial identity and (2) recognition of facial expression. Although the importance of decoding the meaning of facial expressions during a reciprocal social interaction is rather obvious, the importance of face identity recognition has not often been appreciated. It is important, for example, to be able to quickly differentiate friends from strangers. It is also important to remember personal information in association with face identity to facilitate friendship formation, mate selection, and the like. In this regard, face identity perception becomes an important signpost for other semantic knowledge that is critical to interpersonal relationships. Facial expression and facial identity perception, although related, appear to be mediated by different neural systems. Face Identity Perception Compared to other classes of objects that we encounter everyday, for example, furniture, houses, cars, etc., faces are usually more homogeneous in their features and in the spatial positioning of the primary features. Given the structural similarity between faces, our easy and immediate recognition of faces that we encounter is remarkable. In fact, our skill in discriminating faces is probably more highly developed than is our skill for any other (nonface) object.90 This has led some to argue that virtually all adults are experts in the recognition of faces.91–93 This type of perceptual expertise involves sensitivity to the configuration of the major features, such that slight distortions of their spatial relationship are quickly recognized by experts but not by novices.94 Studies of the process by which one becomes a perceptual expert show that fundamental changes occur in the perceptual processes, including a shift from piecemeal processing toward holistic and configural processing.95,96 One of the most striking areas of disability in autism is concerned with the problems that individuals with autism have in face-to-face social engagement. Aversion of eye contact, difficulties in mutual gaze and pragmatic communication, and difficulties in face perception have been described repeatedly.97 Whereas people without an ASD appear to be each a “face expert,” a substantial literature now shows that individuals with an ASD have rather profound difficulty with recognizing face identity, in the context of preserved general vision and normal ability to recognize other types of complex objects.98–105 In fact, ongoing work from our laboratories suggests that the magnitude of the deficit in recognizing facial identity is as large as or larger than that in any other neuropsychological domain. Perhaps more interesting in face perception tests is not the impaired accuracy, but rather the type of errors and the contexts that make for relatively better or worse performance.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 331 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

331

Persons with an ASD rely too heavily on feature-level analyses and seem to be less influenced by the overall gestalt of the face, showing, for example, less performance degradation with inverted faces than expected.105,106 This feature-level bias may indicate a preference for high spatial frequency (HSF) information, i.e., sharp changes in brightness (the edges). Behavioral studies have consistently shown that face recognition among typically developing controls is facilitated more by the low spatial frequencies (LSFs, i.e., the position of features in space) than the HSFs.107–109 In contrast, children with an ASD do relatively better on face tasks with the HSF information compared to the LSF information, a pattern that is opposite that of typically developing children.101 Neural Bases of Face Recognition A small region on the underside of the temporal lobe, along the lateral extent of the middle portion of the fusiform gyrus (FG) shows selectivity (i.e., enhanced activation) for faces compared to other complex objects.110,111 Because of the consistency with which this area of face-selective activation has been found and because of the known relationship between lesions to this region of the ventral temporal lobe and prosopagnosia,112,113 this swath of tissue has come to be known as the fusiform face area (FFA).110 The specificity of the FFA is for perceptual identification of the face; it is distinct from other brain areas that are involved in the perceptual recognition of facial expressions, such as the superior temporal sulcus (e.g., Reference 114 to Reference 117). Anatomically, the middle portion of the FG is split along its rostralcaudal extent by a shallow, mid-fusiform sulcus. In fMRI, the center of activation in face-perception tasks is typically offset toward the lateral aspect of the FG, and it is typically greater in the right (vs. the left) hemisphere.118 In the first functional neuroimaging study of face perception among those with an ASD, we showed that the FFA was hypoactive in a group of 14 persons with autism or Asperger syndrome compared to two independent samples of 14 control participants.119 Hypoactivity of the FFA has since been replicated by nine other laboratories.120–128 This now represents the single-best replicated functional MRI marker of autism in the literature (see Schultz6,129 for more detailed reviews of these studies, as well as two failed replications, and a theoretical model for this line of research). The question facing researchers in this area is no longer whether the face selectivity of this area is abnormal, but why is it abnormal. We have presented preliminary data showing that individual differences in autism symptom severity strongly predict individual differences in the degree of FFA activation, such that less FFA activation is associated with greater social disability.102,130 We also find that accuracy in face identity perception tests correlate with the degree of FFA activation. Most recently, Dalton and colleagues131 showed that individual differences in FFA activity were correlated with measures of how much, on average, a study participant focused on the eye region of the face. Thus, those participants with an ASD who were most normal in their percentage looking time at the eye region were also most normal in terms of their FFA activation. This is a very important finding, but it cannot be said from those data that when a person with an ASD fixates on the eye region, the FFA is engaged normally. A within-subject analysis is needed to address

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 332 Wednesday, February 15, 2006 3:25 PM

332

Understanding Autism: From Basic Neuroscience to Treatment

such a question. We do not yet know what factors mediate FFA activity in real time among persons with an ASD, although these studies are certainly under way. Thus, the findings so far about factors that mediate the degree of FFA hypoactivation are each about individual differences, such that the degree of FFA hypoactivation is greatest for those participants with an ASD who have the weakest face identity perception skills, who have the lowest degree of eye fixation, and who are the most socially impaired. It seems likely that all three predictors of individual differences in FFA activation level share something in common, but the nature of those relationships has not yet been spelled out. Moreover, it is not yet clear how the face perception and FFA abnormalities in autism are related to the social skill deficits. Clearly, deficits in face perception should be a hindrance to social functioning. But this does not mean that face perception skills are required for normal social functioning. If this were true, all blind individuals would be autistic, which is clearly not the case, although they do appear to be at an increased risk for autism (reviewed in Schultz6). In addition, it has been known for a long time that lesions in adulthood that result in prosopagnosia generally do not also cause social dysfunction reminiscent of autism. What is still not clear is whether developmental deficits in face perception might play a causal role in producing autistic social deficits6. Some new evidence with a small sample of congenital prosopagnosics, however, shows that one can have early-appearing face identity perception deficits and still be free from the signs and symptoms of autism (Marlene Behrmann, personal communication). If this finding is borne out in larger studies, it would suggest a limited role of face perception and FFA abnormalities in the causal etiology of autism. In summary, it is much more likely that these face perception deficits are instrumental in producing autism only when they co-occur with other deficits. As such, the field needs to start investigating the interaction of FFA abnormalities and other aspects of the social brain in autism. In light of most findings indicating that the FG has a specific role in ASDs, studies that examine its morphology are now beginning to appear. One recent structural MRI study132 of 16 adolescent and young adult males with an ASD used an automated procedure known as voxel brain morphology (VBM) to investigate morphological differences in the gray matter of the brain and found about a dozen brain areas that were specifically enlarged in the ASD group, consistent with the findings of overall brain enlargement that has been reported many times (see Chapter 16). Their second-strongest finding involved a specific enlargement of the right FG with the location of the peak size difference being consistent with peak coordinates found in fMRI studies of the FFA. Thus, there may be anatomical as well functional abnormalities of this specific region of the cortex. Facial Expression Perception Persons with ASDs are also impaired in their ability to perceive, label, and show comprehension of facially expressed emotions.133–145 This relative inability to perceive accurately what others may be feeling is believed to be a fundamental component of the difficulties with social reciprocity that define autism. Some researchers have focused on the eye region as the most critical part of the face for understanding

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 333 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

333

another’s state of mind. Notably, Baron-Cohen and colleagues146,147 have shown that individuals with an ASD have significant difficulty extracting the “language” of complex emotional states from the eyes.146,147 Consistent with this focus on the eye region, our colleague Ami Klin74 has used infrared eye-tracking technology to measure visual scan paths in persons with autism as they try to make sense of interpersonal social interactions. He found that persons with autism focused much more than typical viewers on one feature of the face – the mouth – and less on the rest of the face, particularly the eye region. In fact, the distribution of percent viewing time on the eye region for each group did not overlap at all in this first study, showing that this one behavioral variable could classify participants with 100% sensitivity and specificity. Others148 have found a similar reliance on the mouth region using still photos instead of the movie clips employed by Klin and colleagues.74 The Neural Basis of Facial Expression Perception The FFA is one of the two main social perceptual areas. The other is the superior temporal sulcus (STS). This is a long sulcus spanning as much as 75 mm, which separates the superior from the middle temporal gyrus. Within the banks of the posterior portion of STS, there are groups of neurons specializing in interpreting dynamic social signals, such as direction of eye gaze, gestures, facial expression, and other “changeable” aspects of the face and body.114,116,149 The posterior STS is a highlevel visual processing area that receives polysensory inputs.114 It appears to be primarily concerned with the perception of facial components (e.g., eyes and mouth), head orientation and movement, as well as goal-oriented body movements.150–158 One recent study showed differential activation of the posterior STS depending on the exact type of body movement.159 Perception of mouth movements elicited STS activity along the midposterior STS. Eye movements, on the other hand, elicited activity that was somewhat more posterior along the STS. Thus, there appear to be differentiable subdivisions of the STS, each of which could have a unique relationship to the pathophysiology of autism. Neither the STS nor the FG is an encapsulated module that can function in an autonomous manner; rather, they operate as part of larger neural systems and information processing systems subserving social perception, which probably reflects the interactions between multiple nodes within these systems. The precise interactions between the nodes has not yet been well specified, but as will be described later in the discussion on amygdala, the FFA and amygdala appear to have a special relationship to each other and to form the core of one functional neural network.6,129 One recent fMRI study has examined the functioning of the STS in autism.160 This study found a difference in the way the STS responded in individuals with autism in gaze processing tasks. Specifically, the STS was abnormally modulated by shifts in gaze toward a region of space that was occupied in some trials but empty in others (which violated the experimentally imposed expectations). This result nicely captures the location of neural activity that is abnormal under conditions pulling for joint attention. Using a paradigm focused on perceptual discrimination of different facial expressions, our group has also presented data in a preliminary

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 334 Wednesday, February 15, 2006 3:25 PM

334

Understanding Autism: From Basic Neuroscience to Treatment

form to support an abnormality of STS functioning.102,130 Moreover, in a separate study on the way the STS participates in integrating auditory prosody with visual information on facial expressions, we also find hypoactivation of the STS, but in this case, the abnormality includes some more anterior portions of the STS too.150 Clearly, the STS supports critical social perceptual functions that are abnormal in autism, and thus, its functioning is of great interest. Moreover, there are now also at least two reports that the morphology of the STS is altered in autism.132,161 Thus, both the FFA and the STS show structural and functional abnormalities in autism.

STUDIES

OF

SOCIAL COGNITION

A full model of the social brain includes posterior perceptual inputs, as just reviewed, as well as cognitive outputs. Heuristic models of brain organization go back many decades. The famous Russian neuropsychologist, Aleksandr R. Luria,162 divided the cortex into a posterior sensory unit and an anterior motor unit. Whereas the sensory unit receives and processes perceptual inputs, the motor unit formulates intentions, plans for actions, and then executes behaviors. In fact, the pattern of axonal inputs and projections within the brain confirms this rather simple, but nonetheless powerful observation. From such a model, social cognition, broadly defined as planning and executing social interactions, would be under the purview of the motor unit. Social cognition is inclusive of concepts such as ToM. Therefore, an understanding of the brain bases of social cognition necessarily entails a shift to the frontal lobe, and to its connections with posterior sensory areas as well as limbic regions for the integration of drives and motivation. The anterior social-cognitive areas are thus thought to receive highly processed sensory data from posterior regions such as the FFA and STS, and to integrate that information with motivational data and with past experience prior to the execution of social behaviors. It is this distributed system that is in some manner vulnerable and suboptimally performing in persons with ASDs. Indeed, the involvement of the frontal as well as temporal lobe cortices in the pathogenesis of autistic social behavior seems very likely, as resting blood flow studies have consistently shown both of these global regions to be hypoactive.163 Functional Neuroimaging Studies of Social Cognition Functional neuroimaging studies have highlighted the role of a variety of brain areas in social cognition, including aspects of the orbital and medial prefrontal cortices (PFC) (see Reference 129, Reference 149, and Reference 164 for recent reviews). In addition, aspects of the inferior frontal convexity may have a specific role in empathy165 and imitation learning (i.e., via the mirror neuron network),166 both of which would be relevant to the deficits found in autism. The orbital and medial PFC have dense reciprocal connections with the amygdala,167,168 and, via these connections, these areas are believed to form a system that can regulate social–emotional behavior.129 Within nonhuman primate, lesions to the orbital and medial PFC result in abnormal social responsivity, and loss of social position within the monkey’s social unit.169,170 Similar lesions in humans have been related to ToM task deficits among neurological patients.171

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 335 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

335

The functional imaging literature very consistently highlights the medial PFC, the ring of tissue dorsal to the anterior cingulate, as being involved in processing feelings and intentions.129 This region is probably involved in both ToM type of cognition, as well as self reflection.172 Deficits in ToM ability in autism are well documented, and studies showing medial PFC abnormalities are beginning to appear in the literature. A pilot 15O-water PET study of Asperger syndrome using a ToM task showed specific engagement of the medial PFC, except that the center of activation was displaced below and anteriorly in patients compared with controls.173 More recently, Castelli et al.174 showed reduced dorsomedial PFC activation in ASDs during an adaptation of the social attribution task, involving ToM skills. In addition, PET studies have shown reduced dopaminergic activity in the dorsal medial PFC175 and reduced glucose metabolic activity in adjacent areas of the anterior cingulate gyrus.176 Although frontal cortices are generally believed to have the dominant role in social cognition, the line demarcating the motor and sensory units vis-à-vis social cognition is beginning to become blurred. For example, in addition to its primary role in social perception, the FFA also appears to be involved in select aspects of social cognition. Three studies employing visual ToM type of tasks have now shown the FFA to be active during social judgments in the absence of any presentation of a face or a face-like object.129,177,178 One interpretation of the FG’s activity during social-cognitive tasks is that it is reactivated in service of the semantic system, and that the coding and storage of social-semantic knowledge normally entails reactivation of perceptual representations.6,129 In this context, the FFA’s low activity level during face perception in individuals with an ASD might, in part, reflect a paucity of social ideation in response to a face, as well as deficits in face perception.6,129

SOCIAL MOTIVATION One current theory that appears to be increasingly popular is that autism stems in a large part from a failure of appropriate levels of social motivation, which when deficient from birth, derails a whole host of normal developmental processes.6,97,129,180–182 One example of the developmental consequences of congenitally low levels of social motivation would be a failure to develop normal perceptual expertise for faces.6,182 More broadly, insufficient motivation for social engagement would be expected to lead to less social experience, which would impact social-perceptual and socialcognitive skill development. From the perspective of this model, congenital deficits in social motivation are a lynchpin to the development of autism. The neurobiology of motivation is a complex topic that is often reduced to the neurobiology of reward and punishment for specific behaviors.183 Although reward systems in the brain have been well studied with respect to basic physiological needs (e.g., hunger, thirst, sex, pain avoidance, etc.), not much is known about social motivation, and what is known is usually inferred indirectly. The neural systems for social motivation are generally assumed to be the same reward pathways in the ventral striatum and related structures that mediate reward for simpler behaviors and basic drives.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 336 Wednesday, February 15, 2006 3:25 PM

336

Understanding Autism: From Basic Neuroscience to Treatment

One structure in particular, the amygdala, is often ascribed a key role in the social motivation deficits in autism. The amygdala is known, for example, to be necessary for learning to associate sensory perceptions with reinforcers.183 Early studies by Weiskrantz184 showed that bilateral removal of the amygdala in monkeys created animals that were largely insensitive to reinforcers. For this and other reasons reviewed in more detail in Chapter 11, the amygdala has been a focus of great interest among autism researchers.6,119,129,147,182,185,186 The amygdaloid complex is a small, almond-shaped structure located deep in the medial temporal lobe, composed of more than a dozen nuclei, each with its own afferent and efferent connections, neurochemical makeup, and cytoarchitecture.187,188 Contemporary models of its functions suggest that it adds an emotional tag to incoming sensory stimuli, which is important for triaging information in terms of motivational salience, and linking the sensory information with prior knowledge and experience to formulate thoughts and actions. The amygdala has a reciprocal set of connections with the temporal cortex as well as orbital and medial prefrontal cortices.182 In this way, the amygdala is centrally positioned and capable of modulating and interpreting the emotional significance of data processed in the perceptual cortices as well as assisting with the integration of emotion and cognition for decision making and action in the frontal cortices.114,149,152,189 The healthy amygdala plays a critical role in emotional arousal, in assigning significance to environmental stimuli, and in mediating emotional learning.190,191 Damage to the amygdala causes impairment in recognizing facial expression,192,193 detecting social faux pas,194 judging trustworthiness,195 and attributing social intentions.196 Damage to the amygdala also appears to reduce FFA activation, suggesting that there are direct and active inputs from the amygdala to the FFA that support or prime its computational activities.189 The amygdala is consistently activated by almost any task involving face perception, as well as by many socialcognitive tasks.164 The fMRI literature on autism suggests that the amygdala is underactive during social-perceptual and social-cognitive tasks.126,128,130,147,174,197 It is not clear whether amygdala activation reflects the nature of the sensory inputs to this structure, or whether the activation is reflective of computations by component nuclei, or whether it is a combination of these processes. One perspective is that the amygdala’s role in social-cognitive and social-perceptual processes might largely be one of mediating physiological arousal, and that its activation reflects stimulus intensity.198 Thus, hypoactivation of the amygdala in autism may reflect nonspecific task effects. In this regard, reduced activity on fMRI in persons with an ASD might best be characterized as reflecting generalizable low levels of arousal to a broad range of stimuli, with the common thread being the stimuli’s importance to social perception. From this perspective, amygdala hypoarousal is a likely principal source of low social motivation in autism, which in turn leads to a broad array of social perceptual and social-cognitive deficits across the course of development. Although this model is perhaps an oversimplification of reality, it does provide a structure for future studies, which are clearly needed for testing the precise role of these and other neural substrates of social motivation deficits in autism.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 337 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

337

SUMMARY AND CONCLUSIONS Autism is clearly a disorder of early onset, with neurobiological origins, involving a primary disturbance of social-perceptual and social-cognitive skills, as well as a primary deficit in social motivation. Although social disability remains a hallmark of autism, the underlying pathophysiology of this disability remains poorly understood. Fundamental to the effort of characterizing the pathophysiology is a firm foundation of knowledge from developmental and cognitive sciences. We need a rich description of the natural history of autism, as well as of each of the specific processes that define the core of the disorder to make progress in understanding the nature of the neural systems that are aberrant. Recent progress in research has led to increased focus on specific processes (e.g., joint attention and face perception) as well as the extension of this work to infants and young children. These developments are particularly exciting because they offer the potential for providing a much deeper understanding of the nature of social dysfunction, especially as it concerns the developmental unfolding of the disorder, and the ability to fully characterize the initial state and the precursors to full-blown autism. Over the past 10 years, animal, postmortem, and functional neuroimaging studies have begun to create a clear outline of the neural systems that make up the social brain. These systems include the amygdala, FFA, posterior superior temporal sulcus, as well as dorsal medial and orbital aspects of the frontal lobe. The ventral lateral PFC may also have a special role in empathy, as recent work on the mirror neuron system would seem to suggest. There is growing evidence that these same brain areas are structurally and functionally abnormal in autism. Although the past decade of research has led to exciting descriptions of the social brain and its putative role in autism, work in this area is still literally in its infancy stage. Nevertheless, the boundaries for this area of study appear to be set, and the next few years will no doubt witness exciting progress. Although fMRI promises to lead the way in terms of more precise delineation of the brain–behavior relationships that characterize autism, this imaging modality will need to be augmented and integrated with other data gathered with approaches, such as diffusion tensor imaging and electrophysiological methods. Because autism is conceived of as disorders affecting separable neural systems, multimodal approaches are especially attractive. Abnormalities in cerebral white matter volume are already being related to fMRI data, suggesting alterations in functional connectivity.199,200 This kind of integration across distinct types of data is essential if we are to have a rich understanding of the neural processes that govern the ontogeny and maintenance of the dysfunctional behaviors that define autism.

ACKNOWLEDGMENTS This work was supported by grants from the National Institute for Mental Health (R01 MH073084-01), the National Institute of Child Health and Human Development (grants PO1 HD 03008, PO1 HD/DC35482, and U54 MH066494-01), and James S. McDonnell Foundation (Bridging the Brain, Mind, and Behavior Program).

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 338 Wednesday, February 15, 2006 3:25 PM

338

Understanding Autism: From Basic Neuroscience to Treatment

In addition, we would like to thank our colleagues Ami Klin, Cheryl Klaiman, Isabel Gauthier, Kathy Koenig, David Grelotti, and James Tanaka for many helpful discussions of the ideas presented in this chapter.

REFERENCES 1. American Psychiatric Association, Diagnostic and Statistical Manual, 4th ed., APA Press, Washington D.C., 1994. 2. Volkmar, F.R. and Klin, A., Issues in the classification of autism and related conditions, in Handbook of Autism and Pervasive Developmental Disorders, 3rd ed., Volkmar, F.R., Klin, A. Paul, R., and Cohen, D.J. Eds., John Wiley & Sons, NJ, 2005, pp. 5–41. 3. Towbin, K.E., Pervasive developmental disorder not otherwise specified, in Handbook of Autism and Pervasive Developmental Disorders, 3rd ed., Volkmar, F.R., Klin, A. Paul, R., and Cohen, D.J. Eds., John Wiley & Sons, NJ, 2005, pp. 165–200. 4. Rutter, M., Genetic influences and autism, in Handbook of Autism and Pervasive Developmental Disorders, 3rd ed., Volkmar, F.R., Klin, A., Paul, R., and Cohen, D.J. Eds., John Wiley & Sons, NJ, 2005, pp. 425–452. 5. Kanner, L., Autistic disturbances of affective contact, Nervous Child, 2, 1943, pp. 217–250. 6. Schultz, R.T., Developmental deficits in social perception in autism: the role of the amygdala and fusiform face area, International Journal of Developmental Neuroscience, 23, 125, 2005. 7. Klin, A., McPartland, J., Volkmar, F.R., Asperger syndrome, in Handbook of Autism and Pervasive Developmental Disorders, 3rd ed., Volkmar, F.R., Klin, A., Paul, R., and Cohen, D.J. Eds., John Wiley & Sons, NJ, 2005, 88–125. 8. American Psychiatric Press, Diagnostic and Statistical Manual, APA Press, Washington, D.C., 1980. 9. Volkmar, F.R. et al., Field trial for autistic disorder in DSM-IV, American Journal of Psychiatry, 151(9), 1361–1367, 1994. 10. Lord, C. and Corsello, C., Diagnostic instruments in autistic spectrum disorders, in Handbook of Autism and Pervasive Developmental Disorders, Volkmar, F.R., Paul, R., Klin, A., and Cohen, D. Eds., John Wiley & Sons, NJ, 2005, p. 312. 11. Constantino, J.N. and Todd, R.D., Genetic structure of reciprocal social behavior, American Journal of Psychiatry, 157(12), 2043–2045, 2000. 12. Leekam, S.R., Lopez, B., Moore, C., Attention and joint attention in preschool children with autism, Developmental Psychology, 36(2), 261–273, 2000. 13. O’Neill, M. and Jones, R.S.P., Sensory-perceptual abnormalities in autism: a case for more research?, Journal of Autism and Developmental Disorders, 27(3), 283–293, 1997. 14. Lord, C. and Paul, R., Language and communication in autism, In Cohen D.J. and Volkmar, F.R. (Eds.), Handbook of Autism and Pervasive Developmental Disorders, 2nd ed., John Wiley & Sons, New York,1997, pp. 195–225. 15. Tager-Flusberg, H., Paul, R., and Lord C., Language and communication in autism, in Handbook of Autism and Pervasive Developmental Disorders, 3rd ed., Volkmar, F.R., Klin, A., Paul, R., and Cohen, D.J. Eds., John Wiley & Sons, NJ, 2005, pp. 335–364.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 339 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

339

16. Ozonoff, S., South, M., and Provencal, S., Executive function, in Handbook of Autism and Pervasive Developmental Disorders, 3rd ed., Volkmar, F.R., Klin, A., Paul, R., and Cohen, D.J., Eds., John Wiley & Sons, NJ, 2005, pp. 606-627 17. Pennington, B.F. and Ozonoff, S., Executive functions and developmental psychopathology, Journal of Child Psychology and Psychiatry, 37(1), 51–87, 1996. 18. Ozonoff, S., Cook, I., Coon, H., Dawson, G., Joseph, R.M., Klin, A., McMahon, W.M., Minshew, N., Munson, J.A., Pennington, B.F., Rogers, S.J., Spence, M.A., TagerFlusberg, H., Volkmar, F.R., Wrathall, D., Performance on Cambridge neuropsychological test automated battery subtests sensitive to frontal lobe function in people with autistic disorder: evidence from the Collaborative Programs of Excellence in Autism network, Journal of Autism and Developmental Disorders, 34(2): 139–150, 2004. 19. Frith, U., Autism: Explaining the Enigma, Basil Blackwell, London, 1989. 20. Happe, F., Briskman, J., and Frith, U., Exploring the cognitive phenotype of autism: weak “central coherence” in parents and siblings of children with autism: I. Experimental tests, Journal of Child Psychology and Psychiatry and Allied Disciplines, 42(3), 299–307, 2001. 21. Plaisted, K. et al., Towards an understanding of the mechanisms of weak central coherence effects: experiments in visual configural learning and auditory perception, Philosophical Transactions of the Royal Society of London — Series B: Biological Sciences, 358(1430), 375–386, 2003. 22. Burnette, C.P. et al., Weak central coherence and its relations to theory of mind and anxiety in autism, Journal of Autism and Developmental Disorders, 35(1), 63–73, 2005. 23. Lopez, B. and Leekam, S.R., Do children with autism fail to process information in context?, Journal of Child Psychology and Psychiatry and Allied Disciplines, 44(2), 285–300, 2003. 24. Mottron, L., Peretz, I., and Menard, E., Local and global processing of music in high-functioning persons with autism: beyond central coherence?, Journal of Child Psychology and Psychiatry and Allied Disciplines, 41(8), 1057–1065, 2000. 25. Caron, R.F. et al., Infant sensitivity to deviation in dynamic facial–vocal displays: the role of eye regard, Developmental Psychology, 33, 802–813, 1997. 26. Baron-Cohen, S., Leslie, A.M., and Frith, U., Does the autistic child have a “theory of mind”? Cognition, 21(1), 37–46, 1985. 27. Baron-Cohen, S., Mind Blindness, MIT Press, Cambridge, MA, 1995. 28. Bowler, D.M., “Theory of mind” in Asperger’s syndrome, Journal of Child Psychology and Psychiatry, 33(5), 877–893, 1992. 29. Dahlgren, S.O. and Trillingsgaard, A., Theory of mind in non-retarded children with autism and Asperger’s syndrome: a research note, Journal of Child Psychology and Psychiatry, 37(6), 759–763, 1996. 30. Klin, A., Volkmar, F.R., and Sparrow, S.S., Autistic social dysfunction: some limitations of the theory of mind, Journal of Child Psychology and Psychiatry and Allied Disciplines, July 33(5), 861–876, 1992. 31. Ornitz, E.M., Guthrie, D., and Farley, A.H., The early development of autistic children, Journal of Autism and Childhood Schizophrenia, 7(3), 207–229, 1997. 32. Stone, W.L., Autism in infancy and early childhood, in Cohen, D.J. and Volkmar, F.R. (Eds.), Handbook of Autism and Pervasive Developmental Disorders, 2nd ed., John Wiley & Sons, New York, 1997, pp. 266–282. 33. Werner, E. et al., Brief report: recognition of autism spectrum disorder before one year of age: a retrospective study based on home videotapes, Journal of Autism and Developmental Disorders, 30(2), 157–162, 2000.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 340 Wednesday, February 15, 2006 3:25 PM

340

Understanding Autism: From Basic Neuroscience to Treatment

34. Klin, A. et al., Autism in a 15-month-old child, American Journal of Psychiatry, 161(11), 1981–1988, 2004. 35. Volkmar, F., Chawarska, K., and Klin, A., Autism in infancy and early childhood, Annual Review of Psychology, 56, 315–336, 2005. 36. Maestro, S. et al., Attentional skills during the first 6 months of age in autism spectrum disorder, Journal of the American Academy of Child and Adolescent Psychiatry, 41(10), 1239–1245, 2002. 37. Osterling, J. and Dawson, G., Early recognition of children with autism: a study of first birthday home videotapes, Journal of Autism and Developmental Disorders, 24(3), 247–257, 1994. 38. Capps, L., Sigman, M., and Mundy, P., Attachment security in children with autism, Developmental Psychopathology, 6(2), 249–261, 1994. 39. Carter, A.S. et al., Social development in autism, in Handbook of Autism and Pervasive Developmental Disorders, Volkmar, F.R., Paul, R., Klin, A., and Cohen, D. Eds., John Wiley & Sons, NJ, 2005, pp. 312–334, 40. Wing, L., Language, social, and cognitive impairments in autism and severe mental retardation, Journal of Autism and Developmental Disorders, 11(1), 31–44, 1981. 41. Howlin, P., Outcomes in autism spectrum disorders, in Volkmar, F.R., Klin, A., Paul, R., and Cohen, D.J. (Eds.), Handbook of Autism and Pervasive Developmental Disorders, 3rd ed., John Wiley & Sons, NJ, 2005, (1) 201–222. 42. Klin, A. et al., The enactive mind from actions to cognition: lessons from autism, in Volkmar, F.R., Klin, A., Paul, R., and Cohen, D.J. (Eds.), Handbook of Autism and Pervasive Developmental Disorders, 3rd ed., John Wiley & Sons, NJ, 2005, (1) 682–703. 43. Paul, R., Promoting social communication in high functioning individuals with autistic spectrum disorders, Child and Adolescent Psychiatric Clinics of North America, 12(1), 87–106, vi–vii., 2003. 44. Butterworth, G., Origins of mind in perception and action, in Joint Attention: Its Origins and Role in Development, Moore, C., and Dunham, P., Eds., Erlbaum, Hillsdale, NJ, 1995, pp. 103–130. 45. Guillaume, P., Imitation in Children, University of Chicago Press, Chicago, IL, 1991. 46. Murphy, C.M., and Messer, D.J., Mothers, infants, and pointing: a study of gesture, in Studies in Mother-Infant Interaction, Schaffers, H.R. Ed., Academic Press, London, 1977. 47. Bates, E. et al., The Emergence of Symbols: Cognition and Communication in Infancy, Academic Press, New York, 1979. 48. Mundy, P. and Gomes, A., Individual differences in joint attention skill development in the second year, Infant Behavior and Development, 21(3), 469, 1998. 49. Mundy, P et al., Nonverbal communication and early language acquisition in children with Down syndrome and in normally developing children, Journal of Speech and Hearing Research, 38(1), 157, 1995. 50. Ulvund, S.E. and Smith, L., The predictive validity of nonverbal communication skills in infants with perinatal hazards, Infant Behavior and Development, 19, 441, 1996 51. Morisette, P., Ricard, M., and Decarie, T., Interactions between shared attention during the preverbal period and the development of verbal references, International Journal of Psychology, 30, 481, 1995. 52. Tomasello, M., Joint attention as social cognition, in Moore, C. and Dunham, P., Eds., Joint attention: Its Origins and Role in Development, Lawrence Erlbaum, Englewood Cliffs, NJ, 1995, pp. 103–130.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 341 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

341

53. Moore, C. and Corkum, V., Infant gaze following based on eye direction, British Journal of Developmental Psychology, 16, 495, 1998. 54. Moore, C., Gaze following and the control of attention, in Rochat, P. Ed., Early Social Cognition, Erlbaum, Hillsdale, NJ, 1999. 55. Mundy, P. et al. Defining the social deficits of autism: the contribution of non-verbal, Journal of Child Psychology and Psychiatry and Allied Disciplines, 27(5), 63, 1995. 56. Lagattuta, K.H. and Wellman, H.G., Thinking about the past: early knowledge about links between prior experience, thinking, and emotion, Child Development, 72(1), 82, 2001. 57. Tomasello, M., Savage-Rumbaugh, S., and Kruger, A.C., Imitative learning of actions on objects by children, chimpanzees, and enculturated chimpanzees, Child Development, 64(6), 1688, 1993. 58. Bakeman, R. and Adamson, L.B., Coordinating attention to people and objects in mother-infant and peer-infant interaction, Child Development, 8, 1, 1984. 59. Baldwin, D.A., Understanding the link between joint attention and language, in Moore, C. and Dunham, P., Eds., Joint Attention: Its Origins and Roles in Development, Erlbaum, Hillsdale, NJ, 1995, pp. 103–130. 60. Dunham, P. et al., The nonreciproating robot: effects on verbal discourse, social play, and social referencing at two years of age, Child Development, 62(6), 1489, 1991. 61. Tomasello, M. and Farrar, M.J., Joint attention and early language, Child Development, 57(6), 1454, 1986. 62. Sorce, J.F. et al., Maternal emotional signaling: its effects on the visual cliff behavior of 1-year-olds, Developmental Psychology, 21, 195, 1985. 63. Chawarska, K., and Volkmar, F., Autism in infancy and early childhood, in Volkmar, F., Klin, A., Paul, R., and Cohen, D.J., Eds., Handbook of Autism and Pervasive Developmental Disorders, 3rd ed., John Wiley & Sons, New York, 2005, pp. 223–246. 64. Lord, C., Follow-up of two-year-olds referred for possible autism, Journal of Child Psychology and Psychiatry, 36(8), 1365, 1995. 65. Stone, W.L. et al., Nonverbal communication in two- and three-year-old children with autism, Journal of Autism and Developmental Disorders, 27(6), 677, 1997. 66. Volkmar, F.R. et al., Social development in autism, in Cohen, D.J and Volkmar, F.R., Eds., Handbook of Autism and Pervasive Developmental Disorders, 2nd ed., John Wiley & Sons, New York, 1997, pp. 173–194. 67. Baron-Cohen, S. et al., Psychological markers in the detection of autism in infancy in a large population, British Journal of Psychiatry, 165(5), 640, 1994. 68. Charman, T., The relationship between joint attention and pretend play in autism, Developmental Psychopathology, 9(1), 1, 1997. 69. Swettenham, J. et al., The frequency and distribution of spontaneous attention shifts between social and nonsocial stimuli in autistic, typically developing, and nonautistic developmentally delayed infants, Journal of Child Psychology and Psychiatry and Allied Disciplines, 39(5), 747, 1998. 70. Sigman, M. et al., Continuity and change in the social competence of children with Down syndrome, and developmental delays, Monograph of the Society for Research in Child Development, 64(1), 1, 1999. 71. Baron-Cohen, S., Are autistic children “behaviorists”? An examination of their mental-physical and appearance-reality distinctions, Journal of Autism and Developmental Disorders, 19(4), 579, 1989. 72. Baron-Cohen, S., Baldwin, D.A., and Crowson, M., Do children with autism use the speaker’s direction of gaze strategy to crack the code of language?, Child Development, 68(1), 48, 1997.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 342 Wednesday, February 15, 2006 3:25 PM

342

Understanding Autism: From Basic Neuroscience to Treatment

73. Klin, A. et al., Defining and quantifying the social phenotype in autism, American Journal of Psychiatry, 159(6), 895, 2002. 74. Klin, A. et al., Visual fixation patterns during viewing of naturalistic social situations as predictors of social competence in individuals with autism, Archives of General Psychiatry, 59(9), 809, 2002. 75. Leekam, S.R., Lopez, B., and Moore, C., Attention and joint attention in preschool children with autism, Developmental Psychology, 36(2), 261, 2000. 76. Leekam, S.R., Hunnisett, E., and Moore, C., Targets and cues: gaze-following in children with autism, Journal of Child Psychology and Psychiatry and Allied Disciplines, 39(7), 951, 1998. 77. Johnson, M.H. et al., Newborns’ preferential tracking of face-like stimuli and its subsequent decline, Cognition, 40, 1, 1991. 78. Valenza, E. et al., Face preference at birth, Journal of Experimental Psychology: Human Perception and Performance, 22, 892, 1996. 79. Batiki et al., 2001 . 80. Farroni, R. et al., Eye contact detection in humans from birth, Proceedings of the National Academy of Sciences, 9602, 2002. 81. Farroni, T. et al., Gaze following in newborns, Infancy, 5(1), 39, 2004. 82. Haith, M.M., Bergman, T., and Moore, M., Eye contact and gaze scanning in early infancy, Science, 218, 179, 1977. 83. Hainline, L., Developmental changes in visual scanning of face and nonface patterns by infants, Journal of Experimental Child Psychology, 25, 90, 1978. 84. Maurer, D. and Salapatek, P., Developmental changes in the scanning of faces by young infants, Child Development, 47, 523, 1976. 85. Farroni, T. et al., Infants use of gaze direction to cue attention: the importance of perceived motion, Visual Cognition, 7, 705, 2000. 86. Farroni, T. et al., Infants perceiving and acting on the eyes: tests of an evolutionary hypothesis, Journal of Experimental Child Psychology, 85(3), 199, 2003. 87. Farroni, Hood, Willen, and Drive, 1998. 88. Corkum, V. and Moore, C., The origins of joint visual attention in infants, Developmental Psychology, 34(1), 28, 1998. 89. Butterworth, G. and Jarrett, N., What minds have in common is space: spatial mechanisms serving joint visual attention in infancy, British Journal of Developmental Psychology, 9, 55, 1991. 90. Diamond, R. and Carey, S., Why faces are and are not special: an effect of expertise, Journal of Experimental Psychology: General, 115, 107, 1986. 91. Carey, S., Becoming a face expert, Philosophical Transactions of the Royal Society of London — Series B: Biological Sciences, 335, 95, 1992. 92. Carey, S. and Diamond, R. 1977, From piecemeal to configurational representation of faces, Science, 195(4275), 312, 1986. 93. Tanaka, J.W. and Gauthier, I., Expertise in object and face recognition, in Goldstone, R.L., Schyns, P.G., and Medin, D. L. Eds., Psychology of Learning and Motivation, Vol. 36, Academic Press, San Diego, CA, 1997, pp. 83, 125. 94. Gauthier, I. and Tarr, M.J., Unraveling mechanisms for expert object recognition: bridging brain activity and behavior, Journal of Experimental Psychology: Human Perception and Performance, 28(2), 431, 2002. 95. Gauthier, I. and Tarr, M.J., Becoming a “greeble” expert: exploring mechanisms for face recognition, Vision Research, 37, 1673, 1997. 96. Gauthier, I. et al., Training “greeble” experts: a framework for studying expert object recognition processes, Vision Research, 38, 2401, 1998.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 343 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

343

97. Grelotti, D., Gauthier, I., and Schultz, R.T., Social interest and the development of cortical face specialization: what autism teaches us about face processing, Developmental Psychobiology, 40, 213, 2002. 98. Boucher, J. and Lewis, V., Unfamiliar face recognition in relatively able autistic children, Journal of Child Psychology and Psychiatry, 33(5), 843, 1992. 99. Davies, S. et al., Face perception in children with autism and Asperger’s syndrome, Journal of Child Psychology and Psychiatry, 35(6), 1033, 1994. 100. De Gelder, B., Vroomen, J., and Van der Heide, L., Face recognition and lip reading in autism, European Journal of Cognitive Psychology, 3, 69, 1991. 101. Deruelle, C. et al. Spatial frequency and face processing in children with autism and Asperger syndrome, Journal of Autism and Developmental Disorders, 34(2), 199, 2004. 102. Schultz, R.T., Task design and patient characteristics that influence degree of FFA activation to faces, Presentation at the 3rd annual International Meeting for Autism Research (IMFAR), Sacramento, CA, May 7–8, 2004. 103. Joseph, R.M. and Tanaka, J.R., Holistic and part-based face recognition in children with autism, Journal of Child Psychology and Psychiatry, 44, 529, 2002. 104. Klin, A. et al., A normed study of face recognition in autism and related disorders, Journal of Autism and Developmental Disorders, 29(6), 499, 1999. 105. Langdell, T., Recognition of faces: an approach to the study of autism, Journal of Child Psychology and Psychiatry, 19, 255, 1978. 106. Teunisse, J. and de Gelder, B., Face processing in adolescents with autistic disorder: the inversion and composite effects, Brain and Cognition, 52(3), 285, 2003. 107. Costen, N.P., Parker, D.M., and Craw, I., Effects of high-pass and low-pass spatial filtering on face identification, Perceptual Psychophysiology, 58(4), 602, 1996. 108. Morrison, D. and Schyns, P.G., Use of spatial scales for the categorization of faces, objects, and scenes, Psychonomic Bulletin and Review, 8, 254, 2001. 109. Parker, D.M., Lishman, J.R., and Hughes, J., Role of coarse and fine spatial information in face and object processing, Journal of Experimental Psychology: Human Perception and Performance, 22(6), 1448, 1996. 110. Kanwisher, N., McDermott, J., and Chun, M.M., The fusiform face area: a module in human extrastriate cortex specialized for the perception of faces, Journal of Neuroscience, 17, 4302, 1997. 111. Puce, A. et al., Face-sensitive regions in human extrastriate cortex studied by functional MRI, Journal of Neurophysiology, 74, 1192, 1995. 112. Damasio, A.R., Tranel, D., and Damasio, H., Face agnosia and the neural substrates of memory, Annual Review of Neuroscience, 13, 89, 1990. 113. Tranel, D., Damasio, A.R., and Damasio, H., Intact recognition of facial expression, gender, and age in patients with impaired recognition of facial identity, Neurology, 38, 690, 1988. 114. Allison, T., Puce, A., and McCarthy, G., Social perception from visual cues: role of the STS region, Trends in Cognitive Science, 4(7), 267, 2000. 115. Grill-Spector, K., Knouf, N., and Kanwisher, N., The fusiform face area subserves face perception, not generic within-category identification, Nature Neuroscience, 7(5), 555, 2004. 116. Haxby, J.V., Hoffman, E.A., and Gobbini, M.I., The distributed human neural system for face perception, Trends in Cognitive Science, 4, 223, 2000. 117. Winston, J.S. et al., fMRI-adaptation reveals dissociable neural representations of identity and expression in face perception, Journal of Neurophysiology, 92(3), 1830, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 344 Wednesday, February 15, 2006 3:25 PM

344

Understanding Autism: From Basic Neuroscience to Treatment

118. Haxby, J.V. et al., The effect of face inversion on activity in human neural systems for face and object perception, Neuron, 22(1), 189, 1999. 119. Schultz, R.T. et al., Abnormal ventral temporal cortical activity during face discrimination among individuals with autism and Asperger’s syndrome, Archives of General Psychiatry, 57, 331, 2000. 120. Aylward, E. et al., Normal activation of fusiform gyrus in adolescents and adults with autism during viewing of familiar, but not unfamiliar faces, Paper Presented at the CPEA/STAART Annual Meeting, Bethesda, MD, 2004. 121. Curby, K.M. et al., Face-selective fusiform activation in Asperger’s syndrome: a matter of tuning to the right (spatial) frequency, Poster Presented at Cognitive Neuroscience, New York, 2003. 122. Critchley, H. et al., Explicit and implicit neural mechanisms for processing of social information from facial expressions: a functional magnetic resonance imaging study, Human Brain Mapping, 9(2), 93, 2000. 123. Davidson, R. and Dalton, K., Dysfunction in the neural circuitry of emotional face processing in individuals with autism, Psychophysiology, 40, S3, 2003. 124. Hall, G.B.C., Szechtman, H., and Nahmias, C., Enhanced salience and emotion recognition in autism: a PET study, American Journal of Psychiatry, 160, 1439, 2003. 125. Hubl, D. et al., Functional imbalance of visual pathways indicates alternative face processing strategies in autism, Neurology, 61(9), 1232, 2003. 126. Pierce, K. et al., Face processing occurs outside the fusiform “face area” in autism: evidence from functional MRI, Brain, 124, 2059, 2001. 127. Piggot, J. et al., Emotional attribution in high functioning individuals with autistic spectrum disorder: a functional imaging study, Journal of the American Academy of Child and Adolescent Psychiatry, 43(4), 473, 2004. 128. Wang, T.A. et al., Neural correlates of facial affect processing in children and adolescents with autism spectrum disorder, Journal of the American Academy of Child and Adolescent Psychiatry, 43(4), 481, 2004. 129. Schultz, R.T., Grelotti, D.J., Klin, A., Kleinman, J., Van der Gaag, C., Marois, R., and Skudlarski, P., The role of the fusiform face area in social cognition: implications for the pathobiology of autism, Philosophical Transactions of the Royal Society of London — Series B: Biological Sciences, 358, 415–427, 2003. 130. Schultz, R.T., Attention, perceptual expertise and semantic representations in perceptual cortices: models to explain the hypoactivation of the FFA and other social nodes, Presentation to the Annual Meeting of the NIH CPEA and STAART Networks, Bethesda, MD, May 19, 2004. 131. Dalton, K.M et al., Gaze fixation and the neural circuitry of face processing in autism, Nature Neuroscience, 8(4), 519, 2005. 132. Waiter, G.D. et al., A voxel-based investigation of brain structure in male adolescents with autistic spectrum disorder, Neuroimage, 22(2), 619, 2004. 133. Capps, L., Yirmiya, N., and Sigman, M., Understanding of simple and complex emotions in nonretarded children with autism, Cognitive Neuropsychology, 13, 699, 1992. 134. Dawson, G. et al., Affective exchanges between young autistic children and their mothers, Journal of Abnormal Child Psychology, 18, 335, 1990. 135. Fein, D. et al., Comprehension of affect in context in children with pervasive developmental disorders, Journal of Child Psychology and Psychiatry, 33, 1157, 1992. 136. Jaedicke, S., Storoschuk, S., and Lord, C., Subjective experience and causes of affect in high-functioning children and adolescents with autism, Development and Psychopathology, 6, 273, 1994.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 345 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

345

137. Hobson, R.P., The autistic child’s appraisal of expressions of emotion: a further study, Journal of Child Psychology and Psychiatry, 2, 671, 1986. 138. Hobson, R.P., The autistic child’s appraisal of expressions of emotion, Journal of Child Psychology and Psychiatry, 27, 321, 1986. 139. Hobson, R.P. and Lee, A., Emotion-related and abstract concepts in autistic people: evidence from the British picture vocabulary scale, Journal of Autism and Developmental Disorders, 19, 601, 1989. 140. Hobson, R.P., Ouston, J., and Lee, A., What’s in a face? The case of autism, British Journal of Psychology, 79, 441, 1988. 141. Hobson, R.P., Ouston, J., and Lee, A., Emotion recognition in autism: coordinating faces and voices, Psychological Medicine, 18, 911, 1988. 142. Macdonald, H. et al., Recognition and expression of emotional cues by autistic and normal adults, Journal of Child Psychology and Psychiatry, 30, 865, 1989. 143. McGee, G.G., Feldman, R.S., and Chernin, L., A comparison of emotional facial display by children with autism and typical preschoolers, Journal of Early Intervention, 15, 237, 1991. 144. Yirmiya, N. et al., Empathy and cognition in high-functioning children with autism, Child Development, 63, 150, 1992. 145. Snow, M.E., Hertzig, M.E., and Shapiro, T., Expression of emotion in young autistic children, Journal of the American Academy of Child and Adolescent Psychiatry, 26, 836, 1987. 146. Baron-Cohen, S. et al., The amygdala theory of autism, Neuroscience and Biobehavioral Reviews, 24, 355, 2000. 147. Baron-Cohen, S. et al., Social intelligence in the normal and autistic brain: an fMRI study, European Journal of Neuroscience, 11, 1891, 1999. 148. Pelphrey, K.A. et al., Visual scanning of faces in autism, Journal of Autism Developmental Disorders, 32(4), 249, 2002. 149. Adolphs, R., Cognitive neuroscience of human social behavior, Nature Reviews, 4, 165, 2003. 150. Robins, D.L., Hunyadi, E.T., and Schultz, R.T., Unique brain activations to integration of audio-visual emotional cues, Paper Presented at the 3rd Annual Meeting of the International Meeting for Autism Research (IMFAR), Sacremento, CA, May 7–8, 2004. 151. Hasselmo, M.E., Rolls, E.T., and Baylis, G.C., The role of expression and identity in the face-selective responses of neurons in the temporal visual cortex of the monkey, Behavioral Brain Research, 32(3), 203, 1989. 152. Hoffman, E.A. and Haxby, J.V., Distinct representations of eye gaze and identity in the distributed human neural system for face perception, Nature Neuroscience, 3(1), 80, 2000. 153. Perrett, D.I. et al., Visual cells in the temporal cortex sensitive to face view and gaze direction, Proceedings of the Royal Society of London — Series B: Biological Sciences, 223(1232), 293, 1985. 154. Perrett, D.I. et al., Organization and functions of cells responsive to faces in the temporal cortex, Philosophical Transactions of the Royal Society of London — Series B: Biological Sciences, 335(1273), 23, 1992. 155. Puce, A. et al., Face-sensitive regions in human extrastriate cortex studied by functional MRI, Journal of Neurophysiology, 74, 1192, 1995. 156. Wicker, B. et al. Brain regions involved in the perception of gaze: a PET study, Neuroimage, 8(2), 221, 1998. 157. Rizzolatti, G. et al., Localization of grasp representations in humans by PET: 1. Observation versus execution, Experimental Brain Research, 111(2), 1996.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 346 Wednesday, February 15, 2006 3:25 PM

346

Understanding Autism: From Basic Neuroscience to Treatment

158. Grafton, S.T. et al., Functional anatomy of pointing and grasping in humans, Cerebral Cortex, 6(2), 226, 1996. 159. Pelphrey, K.A. et al., Functional anatomy of biological motion perception in posterior temporal cortex: an fMRI study of eye, mouth, and hand movements, Cerebral Cortex, in press. 160. Pelphrey, K.A., Morris, J.P., and McCarthy, G., Neural basis of eye gaze processing deficits in autism, Brain, 128(5), 1038, 2005. 161. Boddaert, N. et al., Superior temporal sulcus anatomical abnormalities in childhood autism: a voxel-based morphometry MRI study, Neuroimage, 23(1), 364, 2004. 162. Luria, R.A., The Working Brain, Penguin, Harmondsworth, U.K., 1973. 163. Boddaert, N. and Zilbovicius, M., Functional neuroimaging and childhood autism, Pediatric Radiology, 23, 1, 2002. 164. Schultz, R.T. and Robbins, D.L., Functional neuroimaging studies of autism spectrum disorders, in Handbook of Autism and Pervasive Developmental Disorders, 3rd ed., Volkmar, F.R., Klin, A., Paul, R., and Cohen, D.J., Eds., John Wiley & Sons, NJ, pp. 515–533. 165. Leslie, K.R., Johnson-Frey, S.H., and Grafton, S.T., Functional imaging of face and hand imitation: towards a motor theory of empathy, Neuroimage, 21(2), 601, 2004. 166. Rizzolatti, G. and Craighero, L., The mirror-neuron system, Annual Review of Neuroscience, 27, 169, 2004. 167. Carmichael, S.T. and Price, J.L., Limbic connections of the orbital and medial prefrontal cortex in macaque monkeys, Journal of Comparative Neurology, 363, 615, 1995. 168. Price, J.L., Carmichael, S.T., and Drevets, W.C., Networks related to the orbital and medial prefrontal cortex: a substrate for emotional behavior? Progress in Brain Research, 107, 523, 1996. 169. Bachevalier, J. and Mishkin, M., Visual recognition impairment follows ventromedial but not dorsolateral prefrontal lesions in monkeys, Behavioral Brain Research, 20(3), 249, 1986. 170. Butter, C.M., McDonald, J.A., and Snyder, D.R., Orality, preference behavior, and reinforcement value of nonfood object in monkeys with orbital frontal lesions, Science, 164, 1306, 1969. 171. Stone, V.E., Baron-Cohen, S., and Knight, R.T., Frontal lobe contributions to theory of mind, Journal of Cognitive Neuroscience, 10(5), 640, 1998. 172. Gusnard, D.A. et al., Medial prefrontal cortex and self-referential mental activity: relation to a default mode of brain function, Proceedings of the National Academy of Sciences of the USA, 98(7), 4259, 2001. 173. Happe, F. et al., “Theory of mind” in the brain: evidence from a PET scan study of Asperger’s syndrome, Neuroreport, 8, 197, 1996. 174. Castelli, F. et al., Autism, Asperger’s syndrome and brain mechanisms for the attribution of mental states to animated shapes, Brain, 125, 1839, 2002. 175. Ernst, M. et al., Reduced medial prefrontal dopaminergic activity in autistic children, Lancet, 350, 1997. 176. Haznedar, M.M. et al., Limbic circuitry in patients with autism spectrum disorders studied with positron emission tomography and magnetic resonance imaging, American Journal of Psychiatry, 157, 1994, 2000. 177. Castelli, F. et al., Movement and mind: a functional imaging study of perception and interpretation of complex intentional movement patterns, Neuroimage, 12, 314, 2000.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 347 Wednesday, February 15, 2006 3:25 PM

The Social Brain in Autism

347

178. Martin, A. and Weisberg, J., Neural foundations for understanding social and mechanical concepts, Cognitive Neuropsychology, 20, 575, 2003. 179. Schultz, R.T. et al., fMRI studies of the fusiform face area in autism spectrum disorders, Presentation at the Annual Meeting of the Society for Psychophysical Research, Chicago, IL, 2003. 180. Dawson, G. et al., Neuropsychological correlates of early symptoms of autism, Child Development, 69(5), 1276, 1998. 181. Klin, A. et al., The enactive mind — from actions to cognition: lessons from autism, Philosophical Transactions of the Royal Society, Series B, 358, 345, 2003. 182. Schultz, R.T., Romanski, L., and Tsatsanis, K., Neurofunctional models of autistic disorder and Asperger’s syndrome: clues from neuroimaging, in Asperger’s Syndrome, Klin, A., Volkmar, F.R., and Sparrow, S.S., Eds., Plenum Press, New York, 2000, pp. 179–209. 183. Rolls, E.T., The Brain and Emotion, Oxford University Press, Oxford, 1999. 184. Weiskrantz, L., Behavioral changes associated with the ablation of the amygdaloid complex in monkeys, Journal of Comparative and Physiological Psychology, 49, 381, 1956. 185. Bachevalier, J., Medial temporal lobe structures and autism: a review of clinical and experimental findings, Neuropsychologia, 32, 627, 1994. 186. Bauman, M.L. and Kemper, T.L., Neuroanatomic observations of the brain in autism, in The Neurobiology of Autism, Bauman, M.L. and Kemper, T.L., Eds., Johns Hopkins Press, Baltimore, MD, 1994, pp. 119–145. 187. Amaral, D.G. et al., Anatomical organization of the primate amygdaloid complex, in The Amygdala: Neurobiological Aspects of Emotion, Memory and Mental Dysfunction, Aggleton, J. Ed., Wiley-Liss, New York, 1992, pp. 1–66. 188. Amaral, D.G. and Price, J.L., Amygdalo-cortical projections in the monkey (Macaca fascicularis), Journal of Comparative Neurology, 230, 465, 1984. 189. Vuilleumier, P. et al., Distant influences of amygdala lesion on visual cortical activation during emotional face processing, Nature Neuroscience, 7(11), 1271, 2004. 190. Gaffan, E.A., Gaffan, D., and Harrison, S., Disconnection of the amygdala from visual association cortex impairs visual-reward association learning in monkeys, Journal of Neuroscience, 8, 3144, 1988. 191. LeDoux, J.E., The Emotional Brain, Simon and Schuster, New York, 1996. 192. Adolphs, R., Social cognition and the human brain, Trends in Cognitive Science, 3, 469, 1999. 193. Calder, A.J., Lawrence, A.D., and Young, A.W., Neuropsychology of fear and loathing, Nature Reviews Neuroscience, 2, 352, 2001. 194. Stone, V.E. et al., Acquired theory of mind impairments in patients with bilateral amygdala lesions, Neuropsychologia, 41, 209, 2003. 195. Adolphs, R., Tranel, D., and Damasio, A.R., The human amygdala in social judgment, Nature, 393(6684), 470, 1998. 196. Heberlein, A.S. and Adolphs, R., Impaired spontaneous anthropomorphizing despite intact perception and social knowledge, Proceedings from the National Academy of Sciences of the USA, 101(19), 7487, 2004. 197. Critchley, H.D. et al., The functional neuroanatomy of social behavior: changes in cerebral blood flow when people with autistic disorder process facial expressions, Brain, 123, 2203, 2000. 198. Anderson, A.K. and Sobel, N., Dissociating intensity from valence as sensory inputs to emotion, Neuron, 39(4), 581, 2003.

© 2006 by Taylor & Francis Group, LLC

2732_C015.fm Page 348 Wednesday, February 15, 2006 3:25 PM

348

Understanding Autism: From Basic Neuroscience to Treatment

199. Schultz, R.T., Hunyadi, E., Conners, C., and Pasley, B., fMRI Study of facial expression perception in autism: The amygdala, fusiform face area and their functional connectivity, Presentation at the annual meeting of the Organization for Human Brain Mapping, Toronto, CA, June 12–16, 2005. 200. Just, M.A., Cherkassky, V.L., Keller, T.A., and Minshew, N.J., Cortical activation and synchronization during sentences comprehension in high-functioning autism: evidence of underconnectivity, Brain, 127, 1811, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 349 Friday, March 10, 2006 2:25 PM

16

Structural Neuroimaging Ruth A. Carper, Graham M. Wideman, and Eric Courchesne

CONTENTS Introduction............................................................................................................349 Review of Recent Structural Imaging Literature ..................................................350 Brain Size and Head Size in Autism .............................................................350 Cerebral Findings ...........................................................................................352 Early Childhood.........................................................................................352 Late Childhood through Adulthood ..........................................................354 Specific White Matter Findings.................................................................358 Neuropathological Findings ......................................................................358 Cerebellum......................................................................................................359 Limbic System................................................................................................360 Basal Ganglia..................................................................................................361 Mechanisms ...........................................................................................................361 Possible Abnormalities of Neuroproliferation ...............................................362 Possible Inflammatory Processes ...................................................................364 Future Directions in Neuroimaging ......................................................................365 Advances in MRI Scanner Hardware and Software......................................366 The Challenge of Automated Morphological Processing..............................367 VBM — A Controversial Quantitation Technique ...................................367 Surface Reconstruction and Morphology..................................................370 Diffusion-Weighted Imaging, Diffusion Tensor Imaging, White Matter Orientation, and Tractography.................................................371 Conclusion .............................................................................................................372 References..............................................................................................................373

INTRODUCTION This chapter will focus on structural neuroimaging studies of autism, which are aimed at determining the neuroanatomical abnormalities that characterize the disorder. The information that can be derived from in vivo structural neuroimaging is highly complementary to that gained from studies of neuropathology in postmortem tissue (Chapter 14). Neuropathological studies of autism are limited in sample size, with only about 50 postmortem cases reported across a dozen published papers, and these cases vary in age, quality, and availability. Noninvasive in vivo MRI studies, 349

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 350 Friday, March 10, 2006 2:25 PM

350

Understanding Autism: From Basic Neuroscience to Treatment

on the other hand, may have sample sizes as large as 20 or 40 subjects per group in a single publication, carefully controlled for age and other variables, providing greater confidence in reported findings. MRI studies also allow inspection of age effects (through examination of various ages of subject and through longitudinal studies), a factor that cannot be controlled in postmortem studies (there are probably fewer than a dozen postmortem autism cases available for study that were under age 10 years at time of death). Finally, in vivo studies allow us to assess important correlations between structural brain abnormalities and the constellation of atypical behaviors and cognitive abilities that they produce. Each of these features is vital to developing a robust and comprehensive knowledge of the neurophenotype of autism, and to tying that phenotype to the complex and nonunique collection of genetic alleles that no doubt underlies most cases. Of course, microscopic abnormalities remain obscure without neuropathological studies. Traditional structural neuroimaging studies focus on structure sizes, largescale morphology, and limited estimates of tissue content. We can draw inferences about the meaning of such abnormalities, e.g., whether they are due to changes in cell size or number, in expanse of dendritic arbors, or glial content, etc., but we can only prove these assumptions under a microscope. However, structural MRI allows us to identify important targets, both spatially and temporally, for time-consuming (and resource-limited) microscopic studies. Additionally, as we will discuss in a later section of this chapter, new imaging techniques, engineering advances, and analysis approaches are allowing us to perform more detailed studies in living subjects. In this chapter we will review important structural imaging findings from the autism literature that are pertinent to the neural systems discussed in preceding chapters. These findings will be presented in the context of related neuropathological findings from the postmortem literature. We will focus primarily on imaging research from the last 5 years, but will endeavor to include the most relevant findings from earlier years. We will also discuss some neurodevelopmental mechanisms that might explain some of these neuroanatomical findings. Finally, we will provide a brief overview of some of the newer imaging and analysis techniques that are on the horizon, and what we hope they will tell us.

REVIEW OF RECENT STRUCTURAL IMAGING LITERATURE BRAIN SIZE

AND

HEAD SIZE

IN

AUTISM

Studies of brain size in autism come from two primary sources, in vivo imaging studies of brain volume and postmortem evaluations of brain weight. Approximations of brain size in early life can also be derived from measures of head circumference, which correlate well with brain size in young children.1 The latter measurement is particularly useful for elucidating abnormal brain growth in the very earliest years of development through retrospective studies of medical records. As children rarely receive autism diagnoses before about 18 months of age, this is one of the few methods that allows us to characterize brain growth in the months and years prior to diagnosis. A meta-analysis of data from all of these sources (brain volume, weight,

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 351 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

351

15 2 10 2

Best fit curve

4 2

HC & MRI studies

5

Percent difference

2

10

7

5

8 9 8

0

4

3

4

12

14

11

8

13 6

−5

15 −10 1 2

−15 0

5

10

15

20 Age (yrs)

25

30

35

FIGURE 16.1 A recent meta-analysis compared data from studies of brain volume and head circumference in individuals with autism. To allow comparison of data from different measurement sources, data were converted to measures of percentage difference from control samples. Some studies examined subjects at multiple age points, which are indicated separately. Results indicate that brain sizes in autism are substantially larger than normal during early childhood, with a peak enlargement around 1 to 3 yr of age. 1, Gillberg 20025; 2, Courchesne 20033; 3, Lainhart 19974; 4, Courchesne 20016; 5, Sparks 20028; 6, Kates 2004 69; 7, Herbert 200311; 8, Aylward 200270; 9, Hardan 200348; 10, Piven 199571; 11, Tstasanis 200372; 12, Aylward 199945; 13, Carper, unpublished65; 14, Haznedar 200044; 15, Rojas 2002.14 (From Redcay, E. and Courchesne, E., When is the brain enlarged in autism? A meta-analysis of all brain size reports, Biol Psychiatry 58(1), 1–9, 2005.2)

and head circumference) has recently been published.2 As illustrated in that study (Figure 16.1), data collected from 15 MRI and head circumference studies indicate that brain sizes in autism are substantially larger than normal during early childhood, with a peak enlargement of about 10% around 1 to 3 yr of age. The accelerated growth appears to occur postnatally, largely during the first year of life. Three independent studies3–5 all reported head circumferences that were in the normal average range or significantly below normal at birth in children who later developed autism and, in one of these studies, the same children had head circumferences that were in the 84th percentile when measured between 6 and 14 months of age.3 This extremely rapid growth is not maintained, however. Brain volumes in 2- and 3-yearolds with autism differ little from 12- to 16-year-olds with autism, a period when the normal brain increased by about 20%.6 Because of this slower rate of growth after the initial burst, brain volumes differ little from normal in adolescence and adulthood. As demonstrated in the meta-analysis, most studies are consistent in reporting brain sizes that are, on average, normal or only a little larger than normal

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 352 Friday, March 10, 2006 2:25 PM

352

Understanding Autism: From Basic Neuroscience to Treatment

(0 to 5%) after about 15 years of age. Data for ages 5 to 15 years are not as consistent as those at other ages, however. But, given that this is a time when normal brain growth is still rapid, and the age of puberty onset can be quite variable across individuals, intersubject variability in brain growth may be quite high. In addition, one study of 7- to 15-year-olds with a broader autism spectrum diagnosis showed that a significant brain enlargement among the ASD patients may have been due to a difference in pubertal status between the two groups.7 This variability may necessitate larger sample sizes for consistent results when effect sizes are small.

CEREBRAL FINDINGS Early Childhood The cerebral cortex and its affiliated white matter make up about 80% of normal overall brain volume, so cerebral overgrowth is clearly likely to be a contributor to abnormal increases in brain and head size. As with overall brain volume, there is evidence of overgrowth during the earliest years of life in autism. Two independent studies have examined cerebral volume in children in the important time period before 5 years of age,6,8 and both found that significant enlargement was already present — one in 2and 3-year-olds (Courchesne et al.6) and the other in 3- and 4-year-olds (Sparks et al.8). This was true in both the gray matter (12%) and the white matter (18%).6 Clearly, at some time prior to these early toddler years, the cerebrum has undergone an abnormal schedule of growth. When spatial distribution of these abnormalities was examined, the greatest degree of enlargement in 2- and 3-year-olds was found in the frontal lobe (15% parenchymal enlargement) followed by the temporal and parietal lobes (14 and 9% parenchymal enlargement respectively)9. Again, both gray matter and white matter volumes contributed to these enlargements. The distribution of these abnormalities is consistent with the behavioral deficits seen in autism that affect language and social functions so strongly. The question then arises whether these areas are selectively affected whereas others, such as the primary visual cortex, are unchanged. It is not surprising to see that the areas of more extreme abnormality are also the cerebral areas bearing the greatest proportion of higher-order multimodal cortex (Figure 16.2). Again, the most noticeable deficits in autism are related to higher-order cognitive functions. These areas, particularly the frontal lobe, are normally the slowest to develop ontogenetically. We must consider the possibility that all cerebral regions, including the occipital lobe, may have shown a limited period of accelerated growth, but at an earlier age. In such a scenario, it may be possible to identify additional areas of abnormality if we look at subjects even younger than 2 years. In a further analysis of the frontal lobe, it was found that dorsolateral and medial aspects of prefrontal gray matter were significantly enlarged, but the precentral gyrus and orbitofrontal cortex were not.10 The precentral gyrus contains primary and secondary motor areas and develops earlier than other frontal areas. The orbital area is an association area, but one with strong connections to limbic structures. It is likely that this area normally develops more rapidly than dorsal and medial aspects

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 353 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

353

FIGURE 16.2 (A color version of this figure follows page 236.) Cerebral enlargement in 2- and 3-year-olds with autism was greatest in the frontal lobe (green) followed by the parietal (yellow) and temporal (blue) lobes.9 Striping in frontal areas indicates primary and secondary motor cortex. Striping in other areas indicates primary sensory (somatosensory, auditory, and visual), and assorted unimodal association areas. The ranking of overgrowth in autism also corresponds with the relative proportions of higher-order multimodal association cortex (unpatterned areas).

of the frontal lobe. As with the occipital lobe, this earlier development in the normal brain may help explain why overgrowth was not detected in that area. Of course, it is also quite possible that higher-order association areas really are selectively affected in autism. The fact that these regions are phylogenetically newer, and develop later ontogenetically, suggests that there will be some overlap in the proteins involved in their development and that some of these will be distinct from proteins active during earlier stages of development. Changes in these genes could cause alterations specific to later-developing regions. The hurried development of the autistic brain may mean that, at a point when experience-driven plasticity becomes biologically essential for prototypical development of higher-order areas, lower-level primary or secondary sensory areas may not have fully matured. This would mean that the input (experience) provided by these lower level areas may be noisy or information poor. With inadequate input, normal experience-driven construction of association areas will be derailed. At each level of additional abstraction (i.e., the farther we get from basic unimodal input), the computational puzzles become more difficult, the amount of imperfect, noisy

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 354 Friday, March 10, 2006 2:25 PM

354

Understanding Autism: From Basic Neuroscience to Treatment

information coming in increases, and the resulting inefficiency further limits the precision of subsequent experience-driven development. Late Childhood through Adulthood Studies of patients in later childhood and adolescence are not quite as consistent as those from early childhood, but nevertheless corroborate this atypical schedule of growth. Four studies examined cerebral volumes across fairly comparable age ranges (roughly 7 to 18 years): one showed an enlargement of white matter but not gray matter11; two, an enlargement of gray matter7,12 but not white matter; and one, no differences in either measure.6 This suggests that any differences that are truly present are likely to be in the form of an enlargement but with a relatively small effect size. We have already seen from the meta-analysis of overall brain size that an enlargement seen in early childhood is later diminished in magnitude or even hidden as the brains of normal comparison groups continue to grow. It would seem that by late adolescence the two groups have similar cerebral volumes, though they may still be slightly larger than normal in autism. Similarly, it remains to be seen whether effect sizes found in the two initial studies of young children will be replicated in future studies by more research groups or whether overgrowth will be determined to be more or less severe. One of these studies also examined volumes of individual cerebral lobes, reporting significant enlargement of anterior lobes but not the occipital lobe, although this enlargement was limited to gray matter.7 Studies looking at more detailed regions of interest in late childhood (age 7 to 11 years) have reported abnormal asymmetries in the inferior frontal gyrus, planum temporale, and posterior fusiform gyrus, areas related to language and social function.13 Another study examined adults (age 19 to 47 years) and also found a shift in asymmetry of the planum temporale, but in the opposite direction.14 Many of the studies just described used manual, expert-based measurement techniques. These methods generally offer easy-to-interpret, anatomically valid results but can be very time consuming to perform. As the field of quantitative neuroimaging expands, new techniques are being developed to deal with these issues. One that has seen frequent use in the last few years is referred to as voxel-based morphometry or VBM. This is a relatively new, automated measurement technique that is somewhat controversial.15–18 Details of the method and the issues relevant to the controversy are described in the section titled “Future Directions in Neuroimaging” later in this chapter. For the moment, we will simply describe some of the results. Five studies have been published using VBM to examine cerebral cortical gray matter in autism. All of these examined similar age groups, roughly 7 to 20 years, an age range that we already know shows inconsistent results or reduced effect sizes when using more traditional methods. In addition, some of these studies mixed autistic disorder patients with patients diagnosed with Asperger disorder, sometimes failing to specify the number of subjects with each diagnosis, making it more difficult to compare them to earlier reports. The results of these studies were again inconsistent, in spite of using the same or similar analysis algorithms. Looking at Table 16.1, we see that the regions found to differ from normal varied substantially across studies. In the rare regions where multiple studies identified abnormalities, notably

© 2006 by Taylor & Francis Group, LLC

Patients

Abell et al., 199968

Boddaert et al., 200467

Kwon et al., 200420

McAlonan et al., 200519

Waiter et al. 2004

Asperger

21 Autism

11 Asperger 9 HFA 20M

17 Autism

16 ASD

12M; 3F

16M; 5F

16M; 1F

Unspecified

Controls

12M; 3F IQ matched

7M; 5F

13M IQ matched

16M; 1F IQ matched

16 unspecified IQ matched

Subject age

Average 28 yr

7–15 yr

10–18 yr

8–14 yr

12–20 yr

Software

SPM

SPM

SPM99

BAMM

SPM2

Subjects may be severely retarded

Findings where Ctrl vs. Asp

Software specific to this research group

Notes

Frontal lobe

R↑

Medial L↓ Superior (ant.) Superior (post.) Middle Inferior Orbital

Parietal lobe

Medial Lingual

L↓ (sulcus)

L↓R↓ R↓ L↓R↓

L↑ L↑ L↑ †L ↑

L↓R↓ †L ↑ (continued )

355

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 355 Friday, March 10, 2006 2:25 PM

Study

Structural Neuroimaging

TABLE 16.1 Regions of Abnormality Detected in Studies Using Voxel-Based Morphometry

Study

Abell et al., 199968

Boddaert et al., 200467

Kwon et al., 200420

Waiter et al. 2004

L↓

L ↑ †R ↑

L↓

L↑

Superior lobule Postcentral Temporal lobe

Heschl’s gyrus Planum temporale Superior gyrus STS Middle gyrus Inferior Ventral Fusiform Parahippocampal cortex Entorhinal cortex Occipitotemporal cortex White matter temporal pole

© 2006 by Taylor & Francis Group, LLC

L↓R↓ L↑ R↑

L↓ L↓R↓

L↓ R↓ (rostral)

R↑ L↑

R↓ L↓ R↓

Understanding Autism: From Basic Neuroscience to Treatment

McAlonan et al., 200519

2732_C016.fm Page 356 Friday, March 10, 2006 2:25 PM

356

TABLE 16.1 (Continued) Regions of Abnormality Detected in Studies Using Voxel-Based Morphometry

Other

Thalamus (VLN) Cingulate (body) Cingulate (posterior) Paracingulate sulcus Amygdala/periamygdaloid cortex Hippocampus Int. capsule and fornix Caudate

Cerebellum

Gray White

†L ↑ †R ↓ ↓ R↑ R↓ ↑

↓ L↓R↓ L↑R↑ L↓

L↓R↓

Note: Upward arrows (red) indicate reported increases in the patient group, downward (blue) arrows indicate decreases; L and R indicate hemisphere, whereas an arrow without such an indicator signifies a bilateral effect; regions of abnormality listed are based on the nominal descriptions given in the texts; † = nonsignificant trend; HFA = high-functioning autism; ASD = autism spectrum disorder (may include both autism and PDDNOS patients).19–21,67,68 *differences between Ctrl and HFA were not significant

357

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 357 Friday, March 10, 2006 2:25 PM

Inferior

Structural Neuroimaging

Occipital lobe

2732_C016.fm Page 358 Friday, March 10, 2006 2:25 PM

358

Understanding Autism: From Basic Neuroscience to Treatment

the left middle and superior temporal gyri, right fusiform gyrus, and left middle frontal gyrus, the direction of findings was not consistent.19–21 Both decreases and increases in gray matter concentration were found. The discrepancies could again be signs of a reduced effect size in this age range, but could also be due to the added confounds of mixed or undefined diagnostic groups, as well as the challenges intrinsic to the VBM method. Specific White Matter Findings A few of the preceding studies evaluated cortical gray matter and white matter separately. Among these, Herbert et al.11 and Courchesne et al.6 reported a greater magnitude of enlargement for white matter than for gray matter. This could be interpreted as an indication of specific white matter abnormalities such as excessive myelination, larger-than-usual numbers of oligodendrocytes, or increased density of axonal connections. However, this white matter effect was not reported in all studies. Additionally, an enlargement of white matter volume, even if the magnitude is greater than that in gray matter, may simply be a result of an increased number of cortical neurons. The normal relationship between gray and white matter volumes may not be linear. In cross-species studies, larger cerebral volumes generally involve a disproportionate increase of white matter compared to gray matter volume.22 If the same is true across individuals or, more importantly, across human development, this may be a simpler explanation for the greater magnitude of white matter increase in some studies of autism. Regardless, a study using diffusion tensor imaging to assess cerebral white matter in a small sample of autistic children and adolescents found a number of areas of reduced anisotropy.23 Anisotropy indicates how well axonal fibers are aligned (this concept is explained in greater detail in the section titled “Future Directions in Neuroimaging”). A reduction in this measure indicates some form of abnormality, perhaps in the linear organization of axonal connections or in the formation of the surrounding myelin structure. Reduced anisotropy was found beneath motor and premotor areas, temporoparietal regions, and parts of the frontal lobe, as well as other regions. Although the findings were not very specific spatially, they do suggest the presence of structural abnormality within cerebral white matter tracts. Neuropathological Findings Most neuropathological studies of the cerebrum in autism have reported results qualitatively. Findings have included thickened cortex, areas of laminar disorganization,24 and areas of decreased pyramidal cell density25 variably affecting the frontal lobe (most frequently), the parietal lobe, and the temporal lobe. The anterior cingulate gyrus has shown increased cell packing density, smaller cells and, again, abnormalities in laminar organization.26 More recently, a preliminary quantitative study, using objective stereological methods, has reported an abnormal increase in the number of cerebral neurons when compared to control brains.27 The greatest degree of abnormal increase was found in the youngest cases (as young as 4 years), much as it was in the volumetric MRI studies described earlier. However, it has

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 359 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

359

yet to be determined whether these increases will be evident throughout the cortex, or will be more prominent in anterior areas, as was seen in the MRI studies. Another abnormality reported in the neuropathological literature is a reduced size of minicolumns in higher-order areas of the autistic cortex. A minicolumn (sometimes referred to as a microcolumn) consists of a string of 80 to 100 pyramidal cells in some degree of vertical alignment, which extends across 5 or 6 layers of cortex in the postnatal human brain. Although these minicolumns are thought to function as a single computational unit, they should not be confused with the functional columns that we often think of in the visual cortex and other notable areas. A number of individual minicolumns can be found within a single such functional (e.g., orientation-specific) column. The pyramidal cells within a minicolumn are surrounded by neuropil space, which is made up of interneurons, dendrites, and synapses, and the volume of this neuropil space is believed to affect the overall computational complexity of the minicolumn. For example, minicolumns in the human frontal association cortex have substantially greater volume than those of the primary visual cortex.28 In the first examination of this structure in autism, it was found that the neuropil space associated with each minicolumn was reduced in the three cortical areas examined, one frontal association area (Brodmann area 9) and two temporal association areas (Brodmann areas 21 and 22).29 A further study30 examined distinct areas of the frontal cortex and compared them to a site in the primary visual cortex. Initial results again found decreased neuropil space in the association cortices but did not find abnormalities in the primary visual cortex. In combination, data from the two studies indicate that the greatest abnormality is in the dorsal and orbital frontal cortex, followed by the medial frontal and temporal cortex, with no abnormality in the primary occipital cortex. With the exception of the orbital frontal cortex, this hierarchy of abnormalities is remarkably similar to that seen in cortical gray matter in volumetric MRI studies described previously.9,10

CEREBELLUM For many years now, one of the most consistent neuropathological findings in autism has been a reduction in the numbers of Purkinje cells in the cerebellum.24,26,31–34 This has been described as “patchy” in some cases and has variably affected the vermis and the cerebellar hemispheres. Preliminary results using more statistically reliable stereological methods have further confirmed this, reporting on average a 30% decrease in the numbers of Purkinje cells among cases ranging in age from 4 to 67 years.35 Other reported cerebellar abnormalities have included a reduction of the antiapoptotic protein Bcl-2, which could cause increased cell death36 and an imbalance of nicotinic receptor types in the same subjects.37 Although the neuropathological findings are quite convincing, the earliest MRI studies of the cerebellum in autism appeared to be mixed in their results, despite the fact that the largest studies showed hypoplasia of regions of the vermis,38,39 consistent with the decreased number of Purkinje cells. Most of these early studies measured only the cross-sectional area of the cerebellar vermis because reliable volume measures were more difficult in the relatively low-resolution scans of the day. Precise measurement required a good understanding of cerebellar anatomy and

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 360 Friday, March 10, 2006 2:25 PM

360

Understanding Autism: From Basic Neuroscience to Treatment

a carefully specified, artifact-free image. More recently, however, volumetric measures of the entire cerebellum have become more practicable. Surprisingly, studies of children under age 5 with autism report a significant enlargement of the cerebellum. An enlargement of about 7% was found in 3- and 4-year-olds8 and in 2and 3-year-olds.6 This effect disappeared, however, in older children, in which the volume of cerebellar gray matter, in particular, was significantly smaller than normal.6 Contrary to previous thought, this suggests that Purkinje cell loss in the cerebellum may occur later in life, perhaps in late childhood. Along these lines, a study that examined six postmortem cases found that the five adult subjects, all in their 20s at time of death, had reduced density of Purkinje cells, but the youngest subject (age 4 years at death) did not show alterations in density but did show unidentified abnormal inclusions within those cells.24 However, this case also showed a number of other striking abnormalities that distinguished it from most cases in the literature, suggesting that it may not be fully representative of the disorder. Bauman and Kemper have also reported abnormalities of the cerebellar nuclei and inferior olive that appear to be age related. Cells in these structures were abnormally large but normal in number in the two youngest cases (9 and 11 years) but decreased in number and were either normal or small in size in three cases in their 20s.40 Finally, recent examination of neuroinflammatory responses in postmortem cases indicates that such abnormal responses are present even in the child cases examined.63

LIMBIC SYSTEM The limbic system has long been a system of interest in autism owing to its important role in emotion processing and learning. In addition, medial temporal areas are involved in certain types of seizure disorders, which have a high rate of co-occurrence in autism patients. Neuropathological studies do report consistent abnormalities in the limbic system affecting nine out of nine cases, including decreased cell size and increased cell packing density in the amygdala, hippocampus, entorhinal cortex, medial septal nucleus, mammillary body, and anterior cingulate.26 These authors also reported decreased dendritic complexity in the hippocampus in two cases examined with Golgi staining.41 Decreased complexity would likely result in less neuropil volume and would therefore be consistent with the higher density of neurons per unit volume. However, preliminary results from another group, using more methodologically objective quantitative stereological techniques, did not find any difference in overall neuron numbers in the hippocampus.35 Limbic structures, specifically the amygdala and hippocampus, have also been examined in MRI studies, again providing access to larger sample sizes and more complete sampling of the population. Studies of the amygdala indicate that there is an important effect of age in the development and maturation of this region, just as there is in the cerebrum. Looking across various studies, it appears that the amygdala is abnormally enlarged in both early (2 to 4 years42 or 3 to 5 years8) and late (7.5 to 12.5 years43 ) childhood. But in adolescence and adulthood, the volume is either normal or smaller than normal,43–45 suggesting an early overgrowth followed by either an arrest of growth or an atrophy of tissue in later years, much as we saw in the cerebrum (see the preceding text).

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 361 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

361

A similar age effect may also occur in the hippocampus, but results are slightly more variable. For the most part, results were similar to those seen in the amygdala, with enlargement during childhood8,43 and either normal volume or decreased volume in adolescence and adulthood.44–46 However, one study included this entire age range, from 2 years to 42 years, and reported a different effect.47 Specifically, they found an abnormal decrease in the cross-sectional area of the dentate gyrus in 2- to 4-yearolds but did not find differences in later childhood or adulthood, nor did they find an overall difference in hippocampus size. Considering the early overgrowth reported throughout the cerebrum, there is currently no structural evidence to indicate that autistic features are due to any specific abnormalities of the limbic system. Although neuropathological abnormalities are present, they may reflect atypical developmental processes at work throughout the brain, and the greatest functional impact could occur elsewhere. In addition, the findings of increased cell packing density at postmortem are not necessarily developmental effects, but might instead be results of postnatal factors secondary to autism or to seizure disorders. In any case, higher-resolution MRI images now becoming available will hopefully help resolve any remaining discrepancies about the timing of limbic abnormalities.

BASAL GANGLIA Only a small number of studies have examined the basal ganglia volumetrically, including various combinations of the caudate, putamen, and globus pallidus.11,48–50 All of these indicated normal or enlarged volume of one or another of these ganglia, and none reported reduced size. The two largest studies of the caudate48,49 did not find enlargement when overall brain volume was taken into account, indicating that any enlargement present was in proportion to overall brain overgrowth. This suggests that overgrowth of the basal ganglia is influenced by the same mechanisms that produce overgrowth of other brain regions. Still, these findings are difficult to reconcile with a new preliminary report of substantially decreased neuron numbers in the nucleus accumbens (49% decrease) and elsewhere in the basal ganglia.35 This report referred to the same cases that showed reduced neuron numbers in the cerebellum35 and an increase in the number of neurons in the cerebrum.27

MECHANISMS We have described a number of neuroanatomical abnormalities that have been found in autism, either through neuroimaging techniques or neuropathological techniques. With this view of the neurostructural landscape, we can formulate some hypotheses about possible neurodevelopmental alterations that might lead to this state — one of the ultimate goals of clinical research. While we do not suggest a role for any specific genes (and remember that autism is almost certainly a polygenic, multiplefactor human disorder), theoretical mechanisms such as these can be further tested with specific neurostructural studies. They can also be considered by neurogeneticists when identifying genes of interest in the disorder.

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 362 Friday, March 10, 2006 2:25 PM

362

Understanding Autism: From Basic Neuroscience to Treatment

POSSIBLE ABNORMALITIES

OF

NEUROPROLIFERATION

One of the main findings described earlier was an abnormally large volume of various brain regions in very young children with autism. The most obvious possible source of this enlargement is an increase in the number of neurons. Although neuroproliferation is completed prior to birth in the human (with the very restricted exceptions of the hippocampus and olfactory bulb), an increase in neuron numbers might still help explain the timing of rapid brain overgrowth in autism. Based on retrospective head circumference data, head size at birth is at or below normal but increases to the 84th percentile by about 1 year.3 An increase in the number of pyramidal neurons would likely produce an increase in the overall volume of dendrites, and much of dendritogenesis occurs postnatally. Although we would expect some increase in volume simply due to the cell soma, this would be relatively minimal compared to the resulting increase in dendritic volume. The neocortex has two dimensions of organization: laminar organization (from the white matter to the pial surface) and lateral organization (with different areas of the cortical sheet having different characteristic cytoarchitechtonic structures). An increase in cortical volume may be due to an increase in laminar thickness or in overall surface area. Reflective of these two types of organization, neuroproliferation has two stages: symmetric division (in which each cycle of mitosis produces two daughter cells that both continue to divide) and asymmetric division (in which one of the daughter cells exits the mitotic cycle permanently and begins migration to the cortical plate to form a neuron). Increased symmetric division is believed to contribute to lateral expansion of cortex during evolution,51 whereas asymmetric division has a greater influence on cortical thickness. Although each progenitor cell switches from symmetric to asymmetric division only once, if we look at the entire population of mitotic cells in the ventricular zone collectively, the transition between these two stages is gradual — a decreasing proportion of the mitotic cells participates in symmetric division, and therefore an increasing proportion participates in asymmetric division (Figure 16.3A). Because each round of symmetric division produces two progenitor cells for each initial progenitor whereas asymmetric division produces only one, a delay in the population’s average time of transition from symmetric to asymmetric division will ultimately lead to more neurons. In other words, if most cells wait until the very end of the neuroproliferative interval to switch to asymmetric division, there would be vastly more cells than if the majority had switched very early in the process, assuming that the rate of apoptosis is unchanged. In a series of research studies by Takahashi, Nowakowski, and Caviness examining interrelated mechanisms of normal neuroproliferation, it was hypothesized that a delay in the symmetric-to-asymmetric transition would also result in an increase in cortical thickness because of the increased number of neurons. An initial presentation and test of this hypothesis52, 52a, 52b demonstrated that an experimentally induced delay in the transition did, in fact, result in a thicker cerebral cortex. This could be relevant to the autism postmortem literature, in which patches of increased cortical thickness have been reported in several cases,24 and recent reports have found increased numbers of neurons in the cortex.27

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 363 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

363

B 100%

Proportion of cell population

Asymmetric division

Symmetric division

Asymmetric division

Proportion of cell population

A 100%

Symmetric division

0%

0%

Neuroproliferative interval (time)

Neuroproliferative interval (time)

Predicted changes in laminar proportions

Predicted changes in cortical thickness

D 100%

120% 100%

% of total cortex

Percent of normal

C

80% 60% 40% 20% 0%

Normal

Lower Q

Raise Q

90% 80% 70% 60% 50% 40% 30% 20% 10% 0

III/II IV V VI

Normal

Lower Q

Raise Q

FIGURE 16.3 During the neuroproliferative interval, a decreasing proportion of mitotic cells participates in symmetric division (A, light gray) and an increasing proportion in asymmetric division (A, dark gray). Caviness, Takahashi, and Nowakowski conjectured that an initial delay in the transition from symmetric to asymmetric division (B) would increase the overall thickness of cortex (C) and alter the relative thicknesses of the individual cortical lamina (D). “Lower Q” is the terminology those authors used to refer to a delay in transition. (From Caviness et al. 2003,52 and Takahashi et al. 199652a and 1999.52b With permission.)

An additional aspect of their hypothesis proposed that a delayed transition would also produce changes in the relative thickness of the individual cortical lamina (Figure 16.3B and Figure 16.3D). This would occur because the increase in neuron numbers would have the greatest impact toward the end of the neuroproliferative interval, when the superficial cortical layers (e.g., layers II and III) were being produced. Thus, these layers would be thicker relative to the deeper layers (e.g., V and VI). They demonstrated this hypothesized effect first with a computer model and then with experimental manipulations of the symmetric-to-asymmetric transition. Although their hypothesized mechanism for normal development awaits further confirmation, it provides some interesting possibilities for relating some of the diverse neuropathological findings in autism. Several researchers have suggested that some of the functional abnormalities in autism may be due to an abnormal decrease in the ratio of long-distance corticocortical connections relative to shorter local connections.53–57 As described earlier, delay in the transition from symmetric to asymmetric division during neurogenesis

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 364 Friday, March 10, 2006 2:25 PM

364

Understanding Autism: From Basic Neuroscience to Treatment

could produce thicker superficial cortical layers relative to the deeper layers. Longdistance cortico-cortical projections tend to begin in layer III, whereas local connections (e.g., U-fibers) are more predominantly based in layer II. Although the degree of abnormality in each of these two layers would probably differ only slightly, this might be a mechanism that could contribute to the suspected imbalance in shortand long-distance connections. However, callosal fibers also tend to project from layers II and III, which suggests that the size of the corpus callosum would be larger in this model. But several studies have reported the opposite in autism: a decreased thickness of all or part of the corpus callosum.58–60 Similarly, thicker cortical layers II and III might also be expected to result in more fibers lying between minicolumns as the fibers pass from the superficial layers of cortex to the white matter. An increase in the size of these bundles would likely produce a relative increase in space between columns, but again, this is the opposite of what has been reported: Casanova et al.29 found a decrease in the space between minicolumns in autism. It is interesting to note, however, that part of the space between minicolumns is taken up by inhibitory interneurons, many of which do not arise from the population of neuroproliferative cells in the ventricular zone but from the basal forebrain. They would not be directly affected by the same processes. Thus, an increase in the number of ventricular pyramidal neurons would be relative to the population of interneurons and might, therefore, appear as a relative decrease in the size of the space intervening between minicolumns. This is in line with a hypothesis by Rubenstein and Merzenich,61 who have suggested that an increase in the ratio of excitatory to inhibitory cortical activity is a possible common pathway in the development of autism. An abnormal ratio of excitatory pyramidal neurons derived from the ventricular zone to inhibitory interneurons derived from the basal forebrain could be one mechanism by which such a functional imbalance could be produced. As mentioned previously, the increased cerebral volume seen in autism could also be due to a lateral expansion of the cortical surface area. This would suggest an alteration in the earliest cycles of neuroproliferation, when cell division still occurs symmetrically. One study has examined the gyrification of frontal cortex in individuals with autism and reported an increase in this measure.62 This suggests that the cortical sheet may have expanded in this area, forcing the increased convolution.

POSSIBLE INFLAMMATORY PROCESSES A number of indicators of neuroinflammation were reported in a study that examined both postmortem tissue from patients with autism and CSF sampled from a separate group of living patients with the disorder.63 Tissue samples were taken from middle frontal and anterior cingulate gyri and from posterior cerebellar hemispheres, and they all showed astroglial and microglial activation. Microglial activation was the greatest in the cerebellum, where it was associated with degenerating Purkinje cells, granule cells, and axons. The in vivo CSF samples also showed increased expression of proinflammatory cytokines. Although all these findings indicated activation of the innate immune system, there was no evidence of involvement of the adaptive immune system (e.g., lymphocyte- and antibody-related responses).

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 365 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

365

The full import of these findings is not yet known. These immune-related responses may indicate that an intrinsic abnormality in the immune system has had a detrimental influence throughout life, most importantly, during neurodevelopment. Alternatively, this may be a sign of a secondary effect, a response to neural damage caused by other factors. For example, abnormalities caused by neuroproliferative changes or any other developmental abnormality or insult is likely to change the function and survival of neurons. If apoptosis is increased or if cells or neural branches are lost due to disuse (similar to pruning), the innate immune system will play a role in the cleanup of these cells. When neurons die due to an insult, microglia are normally activated. Thus, the activation reported by Vargas could be a response to other abnormalities or could be a biological problem in itself. However, the finding of actively degenerating cerebellar neurons is particularly intriguing. Decreased numbers of Purkinje cells are a very consistent finding in the autism neuropathology literature, but it has generally been thought that this was due to either a failure of normal cell numbers to develop initially or to a prenatal loss of the neurons. The main arguments for this were a lack of both surrounding gliosis and empty basket cells. Dendritic arbors of basket cells normally surround Purkinje cell bodies very closely, and they are generally left behind when Purkinje cells are lost during adulthood (the remaining cells are thus referred to as empty baskets). Clearly, both this new report of glial activation and an earlier report by Bailey et al.24 that described gliosis and increases in glial fibrillary acidic protein (GFAP) in cerebella of some cases brings the former argument into question. It is therefore possible that Purkinje cell loss is an ongoing process in autism. Vargas et al.6 note similarities between their findings in autism and findings typically seen in some neurodegenerative diseases, again suggesting that there may be an ongoing process. In this regard, it is interesting to note evidence of cortical loss in adults with autism. In two studies (one qualitative64 and one quantitative65) performed on separate samples of adults with autism, we identified an apparent loss of parietal lobe tissue. In both studies, increased amounts of CSF surrounding the parietal lobe indicated a loss of parenchyma that had once been present rather than hypoplasia during development. In addition, a preliminary neuropathological study examining cases from 4 to 67 years of age at death found increased numbers of neurons in the cerebrum, but the magnitude of this abnormality declined with age.27 This too may indicate an ongoing loss of excess neurons in autism. It is not yet known whether inflammation is present in the parietal lobe as it is in the frontal lobe and cerebellum.

FUTURE DIRECTIONS IN NEUROIMAGING The structural neuroimaging studies described in the section titled “Review of Recent Structural Imaging Literature” of this chapter provide substantial insight into the neuroanatomical phenotype of autism. The field does face certain limitations with regard to what we can visualize and in what detail we can assess the brain. However, imaging technology and methods of analysis are improving rapidly. With this, we may see great advances in the science from in vivo neurostructural studies in the coming years. In the following text, we provide an overview of just a few of the new techniques and technologies that are currently being developed, how they work

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 366 Friday, March 10, 2006 2:25 PM

366

Understanding Autism: From Basic Neuroscience to Treatment

(for those who are not familiar) and, more importantly, what we hope they can tell us about autism.

ADVANCES

IN

MRI SCANNER HARDWARE

AND

SOFTWARE

Traditionally, MRI scanners have been oriented toward the clinical priority of simply having a picture with which to make a qualitative assessment (e.g., is there a tumor?), with only secondary concerns about fine quantification and image uniformity, or reduction of the spatial distortion that can affect this process. Scanner manufacturers are now moving toward supporting more quantitative clinical applications, such as characterization of the size, shape, and localization of structures (e.g., for imagingguided surgery), that parallel research interests.66 Scanner hardware continues to improve incrementally, and scanner control software makes better use of it. These trends bring a combination of better image quality, less spatial distortion, and finer resolution of voxels. At the same time, the researcher can now choose scans that complete more quickly, thus reducing motion artifacts in difficult-to-image child populations, or allowing more types of images to be collected in a particular scanning session (e.g., collecting a set of scans that includes multiple protocols which can be combined for better classification of tissue type or content). Improved voxel resolution is a priority in autism in which some of the structures of neuroscientific interest are small, such as the amygdala or the cross section of the cerebral cortical lamina. When an imaged voxel falls on the boundary of two tissue types, it appears on the image with some intermediate brightness (this is known as the partial volume problem). Analysis software must estimate the proportion of each tissue at that location. Inaccuracies in the estimate may be significant if the overall structure is small. The cerebellum, where neuropathological abnormalities have been identified in autism, is a particular challenge with its fine interleaving of white and gray matter features at or below typical MRI resolutions. Figure 16.4 compares a typical contemporary MRI with a photo of a postmortem sample. The fine cortical convolutions of the cerebellum frequently result in partial volumes for many or most of the voxels, making separate assessment of overall gray matter and white matter volumes difficult. Ongoing improvements in image resolution may prove vital to the further study of the cerebellum and other small structures, such as the basal ganglia, amygdala, and hippocampus, where abnormal neuron numbers or packing density have been reported in postmortem studies.26,35 More quantitation-oriented scanners with less spatial distortion and, therefore, more consistent representation of anatomy will also be of interest to autism researchers, who are increasingly interested in using multiple scanning sites for access to greater numbers of subjects, and in longitudinal studies that last several years and require access to the same scanner for the entire duration. Also related to resolution is the trend of MRI centers acquiring one or more scanners with more powerful magnets, for example 3 T (whereas 1.5 T has been the norm for the past several years). These higher-field magnets afford higher spatial resolution (or faster scans), but are more susceptible to spatial distortion and signal inhomogeneity. For structural imaging, these problems will need to be overcome, or bypassed by scanning only portions of the brain.

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 367 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

MRI

367

Post mortem photo 8

7

6

5

4 (a)

(b)

FIGURE 16.4 (A color version of this figure follows page 236.) A typical anatomical MRI, with resolution around 1 mm, is unable to show the detail actually present in the cerebellum. Where the convolutions of the cerebellum are smaller than the voxel size, the MRI voxels must necessarily contain part white and part gray matter, resulting in some intermediate intensity level.

THE CHALLENGE

OF

AUTOMATED MORPHOLOGICAL PROCESSING

We have already mentioned the move toward more automated methods of defining neuroanatomical regions or identifying areas of abnormality. This move is driven largely by the desire to minimize the time involved in more traditional manual methods, which can be quite time consuming. Another goal is to make designation of regions less subjective and less susceptible to the individual variability in the morphology of the human brain. A number of approaches are being developed to try to accomplish this extremely challenging task. Two important examples are highlighted in the following text. VBM — A Controversial Quantitation Technique One automated technique that has been frequently used recently is voxel based morphometry or VBM. However, as we mentioned previously, this technique is somewhat controversial. VBM was first introduced as part of the SPM (statistical parametric mapping) software package that is widely used in functional MRI (fMRI) analysis. SPM’s core technique performs voxelwise statistics on groups of MRIs, producing a 3-D voxel array displayed as an image of the brain whose voxel values are a statistical parameter (such as t or F). Taken as a whole, the resulting 3-D voxel array thus forms a map of the statistical parameter (hence the name), showing

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 368 Friday, March 10, 2006 2:25 PM

368

Understanding Autism: From Basic Neuroscience to Treatment

regions where differences between groups of subjects are significant, lighting them up on a 3-D display. As a preparatory step, MRIs for all subjects (or the several images for each subject) are rotated and warped into anatomical alignment with a reference MRI to some level of precision. Taking the fMRI case, alignment allows the comparison of voxels from like regions of different subjects’ brains, allowing direct comparisons of the activation (blood oxygen level signal) in the different groups of subjects. Better alignment increases the likelihood that legitimate signal will be separable from noise, but even unsophisticated alignment produces worthwhile results. In the VBM application of the SPM idea (Figure 16.5), the goal is not to contrast two subject groups on their activation within a certain brain region, but to contrast the morphology itself. Similar to the fMRI case, a warping operation is employed. In the statistics step, the voxel values being compared represent the measure tissue concentration, an index of the amount of a certain tissue (for instance, gray matter) in the neighborhood of each voxel (in practice, a 12-mm smoothing or “blurring” kernel). The output map is thus claimed to show regions in which tissue concentration is significantly different between groups. This map may be the final product, or it might be used to preselect the regions investigated in a subsequent experiment, reducing the work and allowing less strict multiple comparison correction to be used in the second procedure’s statistics. Note that VBM itself does not provide a quantitative measurement of the amount of volume difference (for example, number of milliliters by which a structure is larger in one group compared to the other) but only a statistical significance at each voxel. Unfortunately, the precise meaning of performing voxelwise statistics on the warped and blurred values of tissue concentration is unclear. The core issue is alignment: Perfect alignment between subject brain and reference template is impossible (indeed, it is quite common for some brains to lack morphological features present in others), and even “good” alignment is computationally challenging. Therefore, there is always some degree of remaining misalignment, almost certainly varying between subjects and possibly between groups, that is hard to characterize. Although the smoothing step blurs this issue, it does not remove it. Thus, whereas effects discovered by the voxelwise statistics are evidently sensitive to differences in tissue concentration as intended, they are also sensitive to the remaining misalignment, which will be present in every case. Individual brain regions vary in their sensitivity to these alignment problems depending on the size of the brain feature relative to the neighborhood kernel used and relative to the fineness of alignment. Therefore, voxelwise statistics are likely to change simply due to their proximity to difficult-to-align areas. The likely stumbling blocks for VBM have been discussed in several papers, which should be of interest to all those considering use of this method.15–18 Critics argue that VBM may well highlight some genuinely different regions, but possibly for reasons other than true gray concentration difference, with no way to distinguish which differences are relevant. Mixed results in VBM studies of autism may, in fact, reflect these challenges: As seen earlier in Table 16.1, these studies have identified diverse areas of abnormality and have even reported opposite directions of abnormality. As a specific example, the frontal lobe, which has shown the greatest cerebral overgrowth

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 369 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

369

Deform calc

T1 MRI

Reference MRI Deform Deformation field

Deformed “spatially normalized”

Segment

CSF

Grey

Other subjects, two or more groups

White

Blur

“Grey concentration”

Voxelwise statistics (SPM)

Significant differences map

FIGURE 16.5 (A color version of this figure follows page 236.) A simplified conceptual view of VBM, incorporating deformation (“warp”), tissue segmentation, and blur, followed by statistics across multiple subjects and groups. More recent versions of VBM integrate the warp and segment processing so that the registration process benefits from segmentation. Recent versions are also more sophisticated about how voxel contributions are copied from original volume to warped volume, weighting for the degree to which each voxel is stretched or shrunk.

with more traditional methods,7,9 was found to be abnormal in only two of five VBM studies, and these found opposite directions of abnormality. Thus, although VBM’s goal of rapid automated morphological processing is a commendable one, the approach must be further refined and explained before its results will be truly convincing.

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 370 Friday, March 10, 2006 2:25 PM

370

Understanding Autism: From Basic Neuroscience to Treatment

Surface Reconstruction and Morphology Another emerging analysis technique that is starting to see implementation in autism research involves software that reconstructs surfaces from the MRI voxel data. Rather than quantifying brain anatomy by viewing and measuring a series of cross-sectional images of the brain or ”slices,” a representation is created of complex 3-D surfaces, such as the outer surface of the cerebral cortex. Simplifying somewhat, one representative version of this software starts with a spherical mesh (consisting of 10,000s of triangles, similar to a miniature Buckminster Fuller geodesic dome) positioned within a cerebral hemisphere. The software “inflates” this balloon through many small iterations, moving each vertex outward until it reaches, for example, a position in the MRI voxels at which it can detect a shift from white matter to gray matter (Figure 16.6). Proceeding gradually, and with some constraints on the “stiffness” of the balloon surface, the result is a surface that fits the white–gray boundary. A second copy of the surface can be further inflated to reach the gray–CSF boundary, thus bounding the cortex. There are a number of challenges to getting this to work reliably — regions of the brain with finer curves or poorer MRI contrast and noise are more difficult for surface-finding algorithms to work with. Some developers of this kind of method are moving toward specifying two or more different MRI protocols as input to provide more robust discrimination of surface boundaries. Nonetheless, having

Reconstructed grey-white surface

Closeup view showing mesh

(a)

(b)

FIGURE 16.6 (A color version of this figure follows page 236.) Surface reconstruction software, such as Freesurfer (http://surfer.nmr.mgh.harvard.edu), calculates a surface that is represented as a mesh of triangles. (A) A view of the surface where white matter meets gray matter. Green indicates convex surfaces, and red indicates concave surfaces. (B) A magnified view of the same reconstruction with the mesh visible on the surface.

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 371 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

371

cortical surface meshes or meshes surrounding subcortical structures of interest presents several opportunities: •

• •

•

•

Characteristics of the mesh, such as curvature or convexity of the surface, can be measured at each vertex on the mesh. These can be portrayed in color in a 3-D rendition of the mesh. With inner and outer surface meshes, thickness can be calculated at each mesh vertex. Either automatically or aided by manual marking of landmarks on the surface, lengths and curvature of salient sulci or intersections of sulci can be measured or otherwise categorized. Shapes of 3-D subcortical structures can be characterized, such as the angle of long and short axes, ratios of axes, and higher-order characterizations of structures. As an additional step, with the gyral/sulcal pattern of curvature attached as data to its vertices, software can create a version of the mesh that is inflated to a sphere and then warped, similar to rubber, so that its gyral pattern aligns to a reference one. Because this alignment is a 2-D surface problem, and has a wealth of detail available (gyral/sulcal curvature), it is thought to be of significantly higher quality than more traditional 3-D voxel-based warping algorithms. With vertices thus stretched into anatomical correspondence with each other, other data from the vertices (such as cortical thickness or fMRI activation data) may be combined across multiple subjects.

This surface reconstruction process is scarcely represented in the recent autism literature. Software implementing these techniques has become available and usable only in the last couple of years. The technique requires some experience to use and is sensitive to various imaging parameters. However, we can expect to see more studies using it in the near future as a number of relevant questions could be addressed. As described in the section titled “Mechanisms” of this chapter, there are data to suggest that cortical thickness might be affected in autism. Another study has suggested an alteration in the morphology of the planum temporale.14 These could be quantified objectively and relatively easily using surface models. If the technique proves effective in spatially registering and combining scans from multiple subjects, this could accelerate even traditional measurement processes.

DIFFUSION-WEIGHTED IMAGING, DIFFUSION TENSOR IMAGING, WHITE MATTER ORIENTATION, AND TRACTOGRAPHY Diffusion-weighted imaging and diffusion tensor imaging (DTI) employ a particular MRI protocol that is sensitive to the diffusion of molecules (typically water) within each voxel. Depending on the exact protocol and number of images acquired, the data can indicate either simple uniformity of direction of diffusion (fractional anisotropy) or include information on the actual direction of diffusion.

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 372 Friday, March 10, 2006 2:25 PM

372

Understanding Autism: From Basic Neuroscience to Treatment

Much of the diffusion in the brain is constrained in its travel by the surfaces of axons and dendrites, with greater diffusion occurring parallel to these fibers. Diffusion can therefore provide information about the uniformity of organization and the orientation of these microstructures. In white matter regions, some neighborhoods feature a predominance of axons aligning in a particular direction, as in the major fiber tracts such as the corpus callosum and arcuate fasciculus, whereas others are occupied by axons oriented in many different directions. The simplest form of DTI will have brighter voxels where orientation is anisotropic, brightest where all fibers are parallel, and darker where it is randomly oriented. Voxelwise statistics may be applied to this variety of DTI to detect whether there are differences in white matter organization in regions of interest. More elaborate DTI uses a larger number of scans to determine more precisely the direction at each voxel. At a first level, an average direction can be determined. With more scans, the information at each voxel becomes more elaborate, potentially showing several predominant fiber directions. For the simple case, one may visualize an arrow attached to each voxel pointing the average direction of its white matter. By joining these arrows head to tail, the DTI analysis software can piece together lengthy pathways, usually white matter tracts. This produces very compelling tractography images, for example, of pathways from cortex through corpus callosum. However, this technique is somewhat controversial as the voxel resolution is coarse relative to the size of axons. Because of this, the average direction may be a misleading combination of two or more quite different directions, and misjudgments as to how the arrows should be joined head to tail could occur and lead to an erroneous tract map, particularly where fibers cross. Local vs. long-distance connectivity, being a matter of interest in autism, motivates consideration of DTI’s ability to shed some light on white matter orientation. Simple DTI (looking only at anisotropy) might at least be able to identify nonnormality. It will also be interesting to see whether full DTI tractography, which is so visually appealing, is a fine enough tool to discern differences in autism. (Such techniques complement other connectivity-assessing techniques in fMRI and EEG.) In addition, there is some promise that fine-voxel DTI images may be able to distinguish cortical layers. Work will need to be done to correlate the MRI or DTI view of living cortex layers with layers as perceived in stained postmortem tissue.

CONCLUSION In this chapter, we have described structural imaging evidence of abnormalities of brain development in autism. This includes a substantial enlargement of the forebrain, affecting the cerebrum, amygdala and, possibly, the hippocampus, as well as more complex effects on the cerebellum. The few studies that have examined very young children with autism suggest that this enlargement is already present by 2 to 3 years of age, although much of the overgrowth appears to take place postnatally. This implies a very rapid rate of growth in the first postnatal years. We have also seen that this rapid growth is not maintained, so that by late childhood or adolescence, enlargement, if present, is not as marked as it is in the toddler years. The overgrowth

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 373 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

373

in the cerebrum is greatest in areas of greatest multimodal connectivity, and examination of postmortem tissue suggests that this overgrowth may be due to an overproduction of neurons. There is also evidence of a loss of certain types of neurons, certainly in the cerebellum, and possibly in the cerebrum later in life. Reconciling all of the diverse data into a cogent explanation of mechanisms will be challenging. A number of possibilities exist, including atypical neuroproliferation or neuroimmune effects as well as mechanisms discussed elsewhere in this text such as abnormalities of synaptogenesis or of the serotonin system. The neuroanatomical subtleties of this disorder, such as the regional specificity of effects, and how abnormalities in distinct cortical layers or of specific cell types might differentially contribute to these measures, must be better characterized before we can fully identify the true implications of results. New advances in brain imaging, such as higher-resolution images, surface reconstruction, and diffusion imaging, together with access to larger sample sizes, may help us address some of these issues.

REFERENCES 1. Bartholomeusz, H.H., Courchesne, E., and Karns, C.M., Relationship between head circumference and brain volume in healthy normal toddlers, children, and adults, Neuropediatrics 33(5), 239–241, 2002. 2. Redcay, E. and Courchesne, E., When is the brain enlarged in autism? A meta-analysis of all brain size reports, Biol Psychiatry 58(1), 1–9, 2005. 3. Courchesne, E., Carper, R., and Akshoomoff, N., Evidence of brain overgrowth in the first year of life in autism, J Am Med Assoc 290, 337–344, 2003. 4. Lainhart, J.E., Piven, J., Wzorek, M., Landa, R., Santangelo, S.L., Coon, H., and Folstein, S.E., Macrocephaly in children and adults with autism, J Am Acad Child Adolesc Psychiatry 36(2), 282–290, 1997. 5. Gillberg, C. and de Souza, L., Head circumference in autism, Asperger syndrome, and ADHD: a comparative study, Dev Med Child Neurol 44(5), 296–300, 2002. 6. Courchesne, E., Karns, C., Davis, H.R., Ziccardi, R., Carper, R.A., Tigue, Z.D., Chisum, H.J., Moses, P., Pierce, K., Lord, C., Lincoln, A.J., Pizzo, S., Schreibman, L., Haas, R.H., Akshoomoff, N.A., and Courchesne, R.Y., Unusual brain growth patterns in early life in patients with autistic disorder: an MRI study, Neurology 57(2), 245–254, 2001. 7. Palmen, S.J., Hulshoff Pol, H.E., Kemner, C., Schnack, H.G., Durston, S., Lahuis, B.E., Kahn, R.S., and Van Engeland, H., Increased gray-matter volume in medicationnaive high-functioning children with autism spectrum disorder, Psychol Med 35(4), 561–570, 2005. 8. Sparks, B., Friedman, S., Shaw, D., Aylward, E., Echelard, D., Artru, A., Maravilla, K., Giedd, J., Munson, J., Dawson, G., and Dager, S., Brain structural abnormalities in young children with autism spectrum disorder, Neurology 59, 184–192, 2002. 9. Carper, R.A., Moses, P., Tigue, Z.D., and Courchesne, E., Cerebral lobes in autism: early hyperplasia and abnormal age effects, NeuroImage 16, 1038–1051, 2002. 10. Carper, R.A. and Courchesne, E., Localized enlargement of the frontal cortex in early autism, Biol Psychiatry 57(2), 126–133, 2005. 11. Herbert, M.R., Ziegler, D.A., Deutsch, C.K., O’Brien, L.M., Lange, N., Bakardjiev, A., Hodgson, J., Adrien, K.T., Steele, S., Makris, N., Kennedy, D., Harris, G.J., and

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 374 Friday, March 10, 2006 2:25 PM

374

12.

13.

14.

15.

16. 17. 18. 19.

20.

21.

22.

23.

24.

25. 26. 27.

28.

Understanding Autism: From Basic Neuroscience to Treatment Caviness, V.S., Jr., Dissociations of cerebral cortex, subcortical and cerebral white matter volumes in autistic boys, Brain 126(Pt. 5), 1182–1192, 2003. Lotspeich, L.J., Kwon, H., Schumann, C.M., Fryer, S.L., Goodlin-Jones, B.L., Buonocore, M.H., Lammers, C.R., Amaral, D.G., and Reiss, A.L., Investigation of neuroanatomical differences between autism and Asperger syndrome, Arch Gen Psychiatry 61(3), 291–298, 2004. Herbert, M.R., Harris, G.J., Adrien, K.T., Ziegler, D.A., Makris, N., Kennedy, D.N., Lange, N.T., Chabris, C.F., Bakardjiev, A., Hodgson, J., Takeoka, M., Tager-Flusberg, H., and Caviness, V.S., Jr., Abnormal asymmetry in language association cortex in autism, Ann Neurol 52, 588–596, 2002. Rojas, D.C., Bawn, S.D., Benkers, T.L., Reite, M.L., and Rogers, S.J., Smaller left hemisphere planum temporale in adults with autistic disorder, Neurosci Lett 328(3), 237–240, 2002. Good, C.D., Johnsrude, I.S., Ashburner, J., Henson, R.N.A., Friston, K.J., and Frackowiak, R.S.J., A voxel-based morphometric study of ageing in 465 normal adult human brains, NeuroImage 14, 21–36, 2001. Bookstein, F.L., “Voxel-based morphometry” should not be used with imperfectly registered images, Neuroimage 14(6), 1454–1462, 2001. Ashburner, J. and Friston, K.J., Voxel-based morphometry — the methods, Neuroimage 11(6 Pt. 1), 805–821, 2000. Ashburner, J. and Friston, K.J., Why voxel-based morphometry should be used, Neuroimage 14(6), 1238–1243, 2001. McAlonan, G.M., Cheung, V., Cheung, C., Suckling, J., Lam, G.Y., Tai, K.S., Yip, L., Murphy, D.G., and Chua, S.E., Mapping the brain in autism: a voxel-based MRI study of volumetric differences and intercorrelations in autism, Brain 128(Pt. 2), 268–276, 2005. Kwon, H., Ow, A.W., Pedatella, K.E., Lotspeich, L.J., and Reiss, A.L., Voxel-based morphometry elucidates structural neuroanatomy of high-functioning autism and Asperger syndrome, Dev Med Child Neurol 46(11), 760–764, 2004. Waiter, G.D., Williams, J.H., Murray, A.D., Gilchrist, A., Perrett, D.I., and Whiten, A., A voxel-based investigation of brain structure in male adolescents with autistic spectrum disorder, NeuroImage 22(2), 619–625, 2004. Zhang, K. and Sejnowski, T.J., A universal scaling law between gray matter and white matter of cerebral cortex, Proceedings of the National Academy of Sciences of the United States of America 97(10), 5621–5626, 2000. Barnea-Goraly, N., Kwon, H., Menon, V., Eliez, S., Lotspeich, L., and Reiss, A.L., White matter structure in autism: preliminary evidence from diffusion tensor imaging, Biol Psychiatry 55(3), 323–326, 2004. Bailey, A., Luthert, P., Dean, A., Harding, B., Janota, I., Montgomery, M., Rutter, M., and Lantos, P., A clinicopathological study of autism, Brain 121(Pt. 5), 889–905, 1998. Belichenko, P.V., Hagberg, B., and Dahlstrom, A., Morphological study of neocortical areas in Rett syndrome, Acta Neuropathologica 93(1), 50–61, 1997. Kemper, T.L. and Bauman, M., Neuropathology of infantile autism, J Neuropathol Exp Neurol 57(7), 645–652, 1998. Palmen, S., Hof, P., Heinsen, H., Steinbusch, H., van Engeland, H., and Schmitz, C., Cortical neurons are more numerous in autism, in International Meeting for Autism Research, Boston, MA, 2005, p. 48. Buxhoeveden, D.P. and Casanova, M.F., The minicolumn hypothesis in neuroscience, Brain 125, 935–951, 2002.

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 375 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

375

29. Casanova, M.F., Buxhoeveden, D.P., Switala, A.E., and Roy, E., Minicolumnar pathology in autism, Neurology 58, 428–432, 2002. 30. Buxhoeveden, D.P., Semendeferi, K., Schenker, N., and Courchesne, E., Decreased cell column spacing in autism, in Society for Neuroscience, Washington, D.C., 2004. 31. Williams, R.S., Hauser, S.L., Purpura, D.P., DeLong, G.R., and Swisher, C.N., Autism and mental retardation: Neuropathologic studies performed in four retarded persons with autistic behavior, Arch Neurol 37(12), 749–753, 1980. 32. Guerin, P., Lyon, G., Barthelemy, C., Sostak, E., Chevrollier, V., Garreau, B., and Lelord, G., Neuropathological study of a case of autistic syndrome with severe mental retardation, Dev Med Child Neurol 38, 203–211, 1996. 33. Fehlow, P., Bernstein, K., Tennstedt, A., and Walther, F., Early infantile autism and excessive aerophagy with symptomatic megacolon and ileus in a case of EhlersDanlos syndrome, Padiatr Grenzgeb 31(4), 259–267, 1993. 34. Ritvo, E.R., Freeman, B.J., Scheibel, A.B., Duong, T., Robinson, H., Guthrie, D., and Ritvo, A., Lower purkinje cell counts in the cerebella of four autistic subjects: initial findings of the UCLA-NSAC autopsy research report, Am J Psychiatry 143(7), 862–866, 1986. 35. Wegiel, J., Subcortical neuropathology, in Integrating the Clinical and Basic Sciences of Autism: A Developmental Biology Workshop, National Alliance for Autism Research, Ft. Lauderdale, FL, 2004. 36. Fatemi, S.H., Halt, A.R., Stary, J.M., Realmuto, G.M., and Jalali-Mousavi, M., Reduction in anti-apoptotic protein Bcl-2 in autistic cerebellum, NeuroReport 12(5), 929–933, 2001. 37. Lee, M., Martin-Ruiz, C., Graham, A., Court, J., Jaros, E., Perry, R., Iversen, P., Bauman, M., and Perry, E., Nicotinic receptor abnormalities in the cerebellar cortex in autism, Brain 125(Pt. 7), 1483–1495, 2002. 38. Courchesne, E., Townsend, J., and Saitoh, O., The brain in infantile autism: posterior fossa structures are abnormal, Neurology 44(2), 214–223, 1994. 39. Hashimoto, T., Tayama, M., Murakawa, K., Yoshimoto, T., Miyazaki, M., Harada, M., and Kuroda, Y., Development of the brainstem and cerebellum in autistic patients, J Autism Dev Disord 25(1), 1–18, 1995. 40. Bauman, M.L., Microscopic neuroanatomic abnormalities in autism, Pediatrics 87(5), 791–796, 1991. 41. Raymond, G.V., Bauman, M.L., and Kemper, T.L., Hippocampus in autism: a Golgi analysis, Acta Neuropathol 91, 117–119, 1996. 42. Morgan, J.T. and Courchesne, E., The amygdala is enlarged in toddlers with autism: a structural MRI study, in Society for Neuroscience, Washington, D.C., 2005. 43. Schumann, C.M., Hamstra, J., Goodlin-Jones, B.L., Lotspeich, L.J., Kwon, H., Buonocore, M.H., Lammers, C.R., Reiss, A.L., and Amaral, D.G., The amygdala is enlarged in children but not adolescents with autism; the hippocampus is enlarged at all ages, J Neurosci 24(28), 6392–6401, 2004. 44. Haznedar, M.M., Buchsbaum, M.S., Wei, T.C., Hof, P.R., Cartwright, C., Bienstock, C.A., and Hollander, E., Limbic circuitry in patients with autism spectrum disorders studied with positron emission tomography and magnetic resonance imaging, Am J Psychiatry 157(12), 1994–2001, 2000. 45. Aylward, E.H., Minshew, N.J., Goldstein, G., Honeycutt, N.A., Augustine, A.M., Yates, K.O., Barta, P.E., and Pearlson, G.D., MRI volumes of amygdala and hippocampus in non-mentally retarded autistic adolescents and adults, Neurology 53(9), 2145–2150, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 376 Friday, March 10, 2006 2:25 PM

376

Understanding Autism: From Basic Neuroscience to Treatment

46. Piven, J., Bailey, J., Ranson, B.J., and Arndt, S., No difference in hippocampus volume detected on magnetic resonance imaging in autistic individuals [published erratum appears in J Autism Dev Disord June 1998; 28(3):271], J Autism Dev Disord 28(2), 105–110, 1998. 47. Saitoh, O., Karns, C., and Courchesne, E., Development of the hippocampal formation from 2 to 42 years: MRI evidence of smaller area dentata in autism, Brain 124, 1317–1324, 2001. 48. Hardan, A.Y., Kilpatrick, M., Keshavan, M.S., and Minshew, N.J., Motor performance and anatomic magnetic resonance imaging (MRI) of the basal ganglia in autism, J Child Neurol 18(5), 317–324, 2003. 49. Sears, L.L., Vest, C., Mohamed, S., Bailey, J., Ranson, B.J., and Piven, J., An MRI study of the basal ganglia in autism, Prog Neuropsychopharmacol Biol Psychiatry 23(4), 613–624, 1999. 50. Hollander, E., Anagnostou, E., Chaplin, W., Esposito, K., Haznedar, M.M., Licalzi, E., Wasserman, S., Soorya, L., and Buchsbaum, M., Striatal volume on magnetic resonance imaging and repetitive behaviors in autism, Biol Psychiatry, 58(3), 226–232, 2005. 51. Rakic, P., A small step for the cell, a giant leap for mankind: a hypothesis of neocortical expansion during evolution, Trends Neurosci 18(9), 383–388, 1995. 52. Caviness, V.S., Jr., Goto, T., Tarui, T., Takahashi, T., Bhide, P.G., and Nowakowski, R.S., Cell output, cell cycle duration and neuronal specification: a model of integrated mechanisms of the neocortical proliferative process, Cereb Cortex 13(6), 592–598, 2003. 52a. Takahashi, T., Nowakowski, R.S., and Caviness V.S. Jr., The leaving or Q fraction of the murine cerebral proliferative epithelium: a general model of neocortical neuronogenesis. J Neurosci, 16(19), 6183–6196, 1996. 52b. Takahashi, T., Goto, T., Miyama, S., Nowakowski, R.S., and Caviness V.S. Jr., Sequence of neuron origin and neocortical laminar fate: relation to cell cycle of origin in the developing murine cerebral wall, J Neurosci, 19(23), 10357–10371, 1999. 53. Just, M.A., Cherkassky, V.L., Keller, T.A., and Minshew, N.J., Cortical activation and synchronization during sentence comprehension in high-functioning autism: evidence of underconnectivity, Brain 127(Pt. 8), 1811–1821, 2004. 54. Belmonte, M.K., Allen, G., Beckel-Mitchener, A., Boulanger, L.M., Carper, R.A., and Webb, S.J., Autism and abnormal development of brain connectivity, J Neurosci 24, 9228–9231, 2004. 55. Herbert, M.R., Ziegler, D.A., Makris, N., Filipek, P.A., Kemper, T.L., Normandin, J.J., Sanders, H.A., Kennedy, D.N., and Caviness, V.S.C., Jr., Localization of white matter volume increase in autism and developmental language disorder, Ann Neurol 55(4), 530–540, 2004. 56. Courchesne, E., Brain development in autism: early overgrowth followed by premature arrest of growth, Ment Retard Dev Disabil Res Rev 10(2), 106–111, 2004. 57. Courchesne, E. and Pierce, K., Why the frontal cortex in autism might be talking only to itself: local over-connectivity but long-distance disconnection, Curr Opin Neurobiol 15(2), 225–230, 2005. 58. Manes, F., Piven, J., Vrancic, D., Nanclares, V., Plebst, C., and Starkstein, S.E., An MRI study of the corpus callosum and cerebellum in mentally retarded autistic individuals, J Neuropsychiatry Clin Neurosci 11(4), 470–474, 1999. 59. Hardan, A.Y., Minshew, N.J., and Keshavan, M.S., Corpus callosum size in autism., Neurology 55(7), 1033–1036, 2000.

© 2006 by Taylor & Francis Group, LLC

2732_C016.fm Page 377 Friday, March 10, 2006 2:25 PM

Structural Neuroimaging

377

60. Egaas, B., Courchesne, E., and Saitoh, O., Reduced size of corpus callosum in autism, Arch Neurol 52(8), 794–801, 1995. 61. Rubenstein, J.L. and Merzenich, M.M., Model of autism: increased ratio of excitation/inhibition in key neural systems, Genes Brain Behav 2(5), 255–267, 2003. 62. Hardan, A.Y., Jou, R.J., Keshavan, M.S., Varma, R., and Minshew, N.J., Increased frontal cortical folding in autism: a preliminary MRI study, Psychiatry Res 131(3), 263–268, 2004. 63. Vargas, D.L., Nascimbene, C., Krishnan, C., Zimmerman, A.W., and Pardo, C.A., Neuroglial activation and neuroinflammation in the brain of patients with autism, Ann Neurol 57(1), 67–81, 2005. 64. Courchesne, E., Press, G.A., and Yeung-Courchesne, R., Parietal lobe abnormalities detected with MR in patients with infantile autism, Am J Roentgenol 160(2), 387–393, 1993. 65. Carper, R. and Courchesne, E., Brain volume in autism from 2 years to 42 years, unpublished results. 66. Dale, A.M., Fischl, B., and Sereno, M.I., Cortical surface-based analysis I: segmentation and surface reconstruction, NeuroImage 9(2), 179–194, 1999. 67. Boddaert, N., Chabane, N., Gervais, H., Good, C.D., Bourgeois, M., Plumet, M.H., Barthelemy, C., Mouren, M.C., Artiges, E., Samson, Y., Brunelle, F., Frackowiak, R.S., and Zilbovicius, M., Superior temporal sulcus anatomical abnormalities in childhood autism: a voxel-based morphometry MRI study, NeuroImage 23(1), 364–369, 2004. 68. Abell, F., Krams, M., Ashburner, J., Passingham, R., Friston, K., Frackowiak, R., Happe, F., Frith, C., and Frith, U., The neuroanatomy of autism: a voxel-based whole brain analysis of structural scans, Neuroreport 10(8), 1647–1651, 1999. 69. Kates, W.R., Burnette, C.P., Eliez, S., Strunge, L.A., Kaplan, D., Landa, R., Reiss, A.L., and Pearlson, G.D., Neuroanatomic variation in monozygotic twin pairs discordant for the narrow phenotype for autism, Am J Psychiatry 161(3), 539–546, 2004. 70. Aylward, E., Minshew, N., Field, K., Sparks, B., and Singh, N., Effects of age on brain volume and head circumference in autism, Neurology 59, 175–183, 2002. 71. Piven, J., Arndt, S., Bailey, J., Havercamp, S., Andreasen, N.C., and Palmer, P., An MRI study of brain size in autism, Am J Psychiatry 152(8), 1145–1149, 1995. 72. Tsatsanis, K.D., Rourke, B.P., Klin, A., Volkmar, F.R., Cicchetti, D., and Schultz, R.T., Reduced thalamic volume in high-functioning individuals with autism, Biol Psychiatry 53(2), 121–129, 2003.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 379 Friday, March 10, 2006 2:27 PM

17

Neuropsychology and Neurophysiology of Autism Spectrum Disorders Nancy J. Minshew, Sara J. Webb, Diane L. Williams, and Geraldine Dawson

CONTENTS Introduction............................................................................................................380 The Complex Information Processing Model.......................................................381 Further Articulation of What “Complex Information Processing” Means..............................................................................................................385 Neuroimaging Validation of the Cognitive Profile ........................................387 Confirmation of Reduced Information Processing Capacity.........................387 Underconnectivity and Overconnectivity of Neocortical Systems................388 Another Aspect of the Information Processing Impairment: Local-Global Processing........................................................................................389 Interrelationships between Local-Global Processing, Object Processing, and Face Processing........................................................391 Extending the Local-Global Processing Account to High-Level Tasks .......................................................................................391 Executive Function and Abstraction: The Nonsocial Impairments in Autism................................................................................................................392 Abstract Reasoning.........................................................................................393 Deficits in Part-Whole Processing and in Abstraction: A Basis for Restricted and Repetitive Behavior............................................394 Other Models for Autism: Social Cognition, a Core Diagnostic Phenotype...............................................................................................................395 Early Social Impairments ...............................................................................395 Face Processing: Early Emerging Impairments.............................................396 Evidence for Early Impairment in Facial Emotion Processing and Memory .................................................................................398

379

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 380 Friday, March 10, 2006 2:27 PM

380

Understanding Autism: From Basic Neuroscience to Treatment

Explanations for Face Processing Impairments.............................................398 Theory of Mind (ToM)...................................................................................400 Oculomotor and Postural Physiology: Beyond Neuropsychological Tests ..........402 Oculomotor Physiology..................................................................................402 Postural Physiology ........................................................................................404 Conclusion .............................................................................................................405 Concluding Remarks Regarding Social Deficits ...........................................406 Closing Comments .........................................................................................406 Perspectives for Future Research ...................................................................407 Acknowledgments..................................................................................................407 References..............................................................................................................408

INTRODUCTION This chapter will provide a selected overview of some key recent findings in the neuropsychology and neurophysiology of autism. The neuropsychology of autism is a very broad topic. Rather than attempt a comprehensive review of the literature on all aspects of this topic, we have chosen to articulate selected models and hypotheses and selected issues of the most relevance to each of these models. This is followed by a review of recent findings in the area of neurophysiology. Two perspectives on the neuropsychology of autism are discussed in some detail. One perspective, offered by Minshew and colleagues, is to consider that there is a constellation of co-occurring deficits and co-occurring intact abilities and that this entire constellation is the primary outcome of the underlying brain abnormality. This is based on the general observation that a brain abnormality, whether of developmental or acquired origin, rarely (if ever) results in a single sign or symptom but rather a constellation of related symptoms. This neurologic perspective then results in simultaneous investigation of all areas of cognitive and neurologic function in the same participants to ascertain the common denominator of the impairments and of intact skills. The common denominator of the deficits is expected to reveal essential or fundamental characteristics about what is altered about cognition and the brain. Examining intact skills confirms the boundaries of the deficits and verifies that identified deficits are not secondary to deficits in other skills. In the case of autism, the intact skills are at times also enhanced and reveal something important about another aspect of developmental neurobiology. The second perspective, articulated by Dawson and colleagues, emphasizes early “primary” deficits that result in downstream consequences to multiple neural systems. This perspective does not preclude multiple deficits but considers the origin of such deficits as related to specific earlyemerging neural systems. We begin the chapter by describing the perspective offered by Minshew and colleagues. It is important to keep in mind that all models are temporary and inaccurate constructs that organize existing knowledge into working hypotheses that hopefully guide research in the most scientifically plausible and fruitful directions to acquire additional knowledge. The differences between these two models and others not discussed here largely reflect what is unknown about autism, which is still quite

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 381 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

381

substantial. In such situations, there is room for speculation, which can take many forms. A second major limitation of models is that the majority of current models of autism are based on the clinical impairments, whereas the disorder arises from a genetically influenced developmental neurobiological disturbance that has a molecular basis. The myriad and diverse ways in which this disturbance may be expressed in cognition and neurologic function are likely to seem inexplicable until the molecular basis is identified. The research process entails progressive iterations during which findings at one level are used to inform findings at all other levels until molecular answers are finally achieved.

THE COMPLEX INFORMATION PROCESSING MODEL As a result of several studies of the profile of neuropsychological functioning in nonretarded individuals with autism and observations of language development in preschool children with autism, Minshew and colleagues posit that autism is most accurately conceptualized as a disorder of information processing that disproportionately impacts complex, high level, or integrative processing. According to this model, overall information processing capacity is constrained or reduced in individuals with autism compared to what is expected of their general ability level,* and within that constraint, the abilities and tasks most impacted are those placing the highest demands on information processing. Furthermore, simpler abilities within the same domains as impairments are intact or enhanced, and basic information acquisition abilities are intact. The second defining feature of the model proposed by Minshew and colleagues is that the disturbance in information processing is generalized, extending beyond the neural systems related to the diagnostic triad of autism to involve other areas such as the motor, sensory, memory, oculomotor, and postural systems. The third key feature of Minshew’s model is that automatic or nonconscious processes are impaired. That is, there are many things that typical people automatically or “naturally know” even as infants but individuals with autism do not know and must cognitively discover or be taught using compensatory, verbally mediated, conscious strategies. Finally, Minshew and colleagues stress that although it is often thought that autism is a disorder largely or exclusively of social function or social cognition, there is abundant evidence demonstrating that nonsocial cognition is equally impaired and that the pronounced limitation with automatic processing and problem solving is a serious impediment in life functioning. The model as articulated above is based on data acquired through a number of studies, some of which are described in the following sections. It is important to review the data that gave rise to the model in order to understand the concepts that the model represents. The first comprehensive study of neuropsychological functioning in autism involved 33 individually matched pairs of typical controls, and adolescents and adults with Full Scale and Verbal IQ scores > 80 who met Autism Diagnostic Interview and * The term general ability level is used in lieu of intellectual level, intellectual quotient, developmental level, or mental age. As these various terms indicate, general ability level is measured by a variety of tests and expressed in various formats, i.e., intellectual quotient using the Wechsler Intelligence Scales or mental age using the Bayley Scales of Infant Development.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 382 Friday, March 10, 2006 2:27 PM

382

Understanding Autism: From Basic Neuroscience to Treatment

Autism Diagnostic Observation Schedule criteria and expert clinical opinion for autism.1,2,3 These individuals were administered a comprehensive battery of tests in the following domains: attention (encoding, focusing, sustaining, and shifting), sensory perception (elementary and higher cortical sensory perception), motor (elementary and skilled motor movements), basic associative memory (word pairs, word list, and short maze), complex memory (story recall, long maze, and complex visual figures), formal language (spelling, reading decoding, vocabulary, and fluency), complex or interpretive language (inferences, metaphors, and idioms), reasoning (attribute identification, rule learning, self-initiated problem solving, and flexible use of strategies with changing contexts), and visuospatial processing (Block Design, Picture Completion, and Object Assembly subtests from the Wechsler intelligence scales). At the time this study was initiated, there were no tests of social cognition appropriate for high-functioning individuals with autism, nor were there time-efficient tests of prosody and facial expression. Thus, some domains were not included in the test battery though deficits were expected. Wilk’s stepwise discriminant analysis was used (see Table 17.1). A kappa score in the .40 to .75 range is considered to be in the fair to good range of agreement. A few findings were conspicuous by their unexpected presence, e.g., impairments in skilled motor movements and memory for complex material, whereas sensory impairments were conspicuous by their absence. The impairment in skilled motor movements was of moderate clinical severity and, prior to this study, was not considered the integral part of the autism syndrome that it is today. The importance of the motor impairment was the implication that the neurobiological abnormality causing autism had to extend to neural systems beyond those involved in the diagnostic triad, especially because the nature or pattern of the abnormality in the motor system

TABLE 17.1 Discriminant Analysis Results by Domain and by Order of Entry Domain

Tests Passing Tolerance Test

Attention Letter Cancellation; Number Cancellation Sensory perception Fingertip Writing; Luria-Nebraska Sharp/Dull Tactile Scale Item Motor Grooved Pegboard; Trail Making A Simple language K-TEA Reading Decoding; K-TEA Spelling; WRMTR Word Attack; Controlled Oral Word Association Complex language K-TEA Reading Comprehension; Verbal Absurdities; Token Test Simple memory CVLT Trial 1 Complex memory NVSRT-Consistent Long Term Retrieval; WMS-R Logical Memory — Delayed Recall; Rey Figure — Delayed Recall Reasoning 20 Questions; Picture Absurdities; Trail Making B Visual-spatial WAIS-R Block Design

© 2006 by Taylor & Francis Group, LLC

Percentage Correct

Kappa Score

66.7 64.6

.33 .29

75.8 71.2

.52 .42

72.7

.45

65.2 77.3

.30 .55

75.8 56.1

.52 .12

2732_C017.fm Page 383 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

383

was the same as in the other affected domains. The memory impairment involved the failure to automatically use cognitive organizing strategies to facilitate recall when there were many elements, such as the word list from the California Verbal Learning Test, and to detect structure or inherent organization in complex material, such as the Rey Osterrieth Complex Figure (ROCF) or stories, and use it for recall. The presence of a memory impairment associated with information complexity was a particular surprise, as it was not a deficit included in DSM-IV criteria or even thought to be associated with autism. As with the motor impairment, the memory impairment implied that there were deficits beyond the traditional triad and that the brain disturbance in autism had to be much broader than suspected. The lack of a significant deficit in higher cortical sensory perception was surprising, given the frequency of clinical complaints about sensory sensitivities in individuals with autism. This domain included tests of both elementary and higher cortical sensory perception.* There was a significant between-group difference for graphesthesia (fingertip number writing) in the higher cortical sensory component of the battery, but this was not enough to carry the domain to significance. There was a trend toward a significant difference for stereognosis, another higher cortical sensory function, but it was not significant in the adult study. The tests used in this study were coarse and not particularly sensitive and primarily assessed sensory deficits not distortions, which are more commonly reported in autism. Nonetheless a standard neuropsychological instrument was used, and the findings were suggestive of abnormalities in higher cortical perception, which would be consistent with symptoms. Aside from these specific features, the neuropsychological profile had a number of overall features. The most obvious feature was that basic information acquisition in terms of attention, elementary sensory perception, associative memory, and basic language encoding and decoding were all intact. In fact, formal language skills (spelling and word decoding) were superior to the age- and IQ-matched controls. In addition, attention shifting and attention to extrapersonal space were intact. Hence, autism was not the consequence of impaired acquisition of information. Impairments were found in skilled motor movements, memory for complex information that required an organizing strategy or detection of inherent organization in material for optimal recall, comprehension of higher-level language meaning, and self-initiated concept formation and problem solving. Thus, impairments were present in the highlevel processing or integration of information, whether this involved motor skills, memory, language, or reasoning. Impairments, when present, were in the same domains as intact skills, producing dissociations within domains.** The dissociations * Sensory perception can be divided neurologically into elementary and higher cortical sensory perceptual abilities to correspond to the spinal thalamic and posterior column sensations of touch, pain, position, vibration, and temperature and to the cortically dependent sensations of stereognosis, graphesthesia, and two-point discrimination. There are more sophisticated assessments of sensory processing but these are the ones targeted with neuropsychological batteries. ** If autism were a general information processing disorder, then both simple and complex abilities would be impacted. The selective impact on complex information processing and sparing of simpler abilities in the same domains was striking and unusual. It is also the converse of the simple information processing profile that is sometimes seen in selective language impairment.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 384 Friday, March 10, 2006 2:27 PM

384

Understanding Autism: From Basic Neuroscience to Treatment

appeared to revolve around the complexity of the information to be analyzed or processed and the number of cognitive processes needed. In the memory domain, one of the tests (maze tasks) confirmed the direct relationship between increasing stimulus complexity and progressive worsening of the deficit. In the case of the language domain, there was a double dissociation. The autism group was significantly better than controls in basic or formal language skills and significantly worse in higher-order language comprehension. In the reasoning domain, the autism group had intact attribute identification and rule-learning abilities but was impaired in the flexible use of concepts in changing contexts and in self-initiated concept formation. When the results were considered as a whole, a number of overall concepts emerged. The first was that the basic abilities involved in information acquisition were intact (e.g., elementary sensory perception, basic associative memory, and formal language). The second was that there were deficits across many domains that selectively involved higher-order abilities that pertained to the processing of complex information; simpler abilities and the processing of simpler information in the same domains was intact or superior. This constellation suggested that there was a problem with the brain’s fundamental mechanisms for processing complex information across neural systems, perhaps involving a common cytoarchitectural feature. The enhanced skills suggested that there might be a developmental neurobiological mechanism whereby a disturbance in the development of the higher-order circuitry involved in the processing of complex information and emergence of higher-order abilities might result in overelaboration of local circuitry underlying simple cognitive, language, and perceptual abilities. In arriving at a characterization of the profile in autism as a complex information processing disorder, consideration was given to the ways in which complexity was defined. Within cognitive theory, complexity is defined as an increase in the number of elements in the stimulus material or an increase in the cognitive demands involved in task performance. The latter definition involves emergent abilities that are not directly reducible to simpler elements of cognitive function. Thus, the cognitive capacity to comprehend extended blocks of language is not simply reducible to vocabulary and grammar skills but requires another level of language abilities in order to comprehend the meanings beyond those implicit to vocabulary and the arrangement of words in sentences. The model proposed here does not distinguish between these two definitions of complexity because they are related in the sense that, as the number of elements increases, there is an increase in the number of cognitive processes needed for task performance. The conclusions drawn from this neuropsychological profile from nonretarded individuals with autism were therefore that autism is a disorder of complex information processing, that the capacity to process information is reduced or constrained, and that it is a generalized limitation of many cognitive and neurologic domains.* * It is acknowledged that information processing begins as soon as information triggers sensory receptors. However, there is a long history in autism of theories proposing core deficits in elementary sensory perception, attention to extrapersonal space, and amnesia. “Complex information processing” is used to draw a distinction from these theories and abilities.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 385 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

385

Signs and symptoms are the most prominent in the domains and systems that place the highest demands on complex information processing and call on the integration of the functions of many brain regions. The identification of this common denominator for the social, language, and reasoning impairments that are relied upon for diagnosis, which also explains less prominent associated symptoms, provides a plausible explanation for this syndrome. Extrapolation of the complex information processing profile to lower functioning individuals with autism would predict that the total information processing capacity would be progressively reduced as the severity of autism increases with a continuation at all levels of the disproportionate impairment of integrative abilities relative to IQ expectations. There would be a progressive shrinkage of higher-order circuitry and eventual reliance on a shrinking lower-order circuitry until, in the severest cases, there would be no processing of information whatsoever. Such a conceptualization readily explains the co-occurrence of mental retardation and autism, which is far too common to be explained by chance. A major clinical distinction between high-functioning children with autism and children in a general child neurology clinic is the absence of dyslexia, spelling, arithmetic, and visuospatial problems in children with autism. This is reflected in the intact simple information processing abilities found in this study. The lack of these basic-skill problems also suggested that autism is a distributed neural network or neural systems disorder rather than a focal or modular brain disorder. It is also the reason that verbal children with autism can score well or “super-well” relative to their adaptive functioning on IQ tests.

FURTHER ARTICULATION OF WHAT “COMPLEX INFORMATION PROCESSING” MEANS A model of autism that argues for difficulty with complex information processing needs to operationalize what is meant by complexity. Complexity is a proxy for the level of demands placed on the brain’s processing system by tasks or situations. Cognitive or neurologic function breaks down in individuals with autism when the processing demands placed on the brain’s systems exceed their capacity and this typically occurs when the information to be handled is complex either inherently, by virtue of its amount, or because of time constraints. Performance difficulty is apparent on tasks or in situations in which age- and IQ-matched peers have no difficulty. At the same time, individuals with autism are able to perform simpler skills in these same skill areas as well as or even better than peers. Hence, the concept of complexity has more to do with the demands placed on the brain’s mechanisms for dealing with the processing of information than it does with the content of the information per se. Increases in processing demands can occur from a number of factors. The task may be challenging because it is beyond the level of brain development of the child, requiring neural resources that have not yet matured or requiring a speed of processing that cannot yet be accomplished. Complexity may occur because a task demands a high coordination between distant brain areas. Complexity may occur

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 386 Friday, March 10, 2006 2:27 PM

386

Understanding Autism: From Basic Neuroscience to Treatment

because a task requires the coordination of a large number of brain areas, meaning that in autism, some of the areas are brought online to accomplish the task but not the full complement needed to complete the task well. Complexity may occur because of the amount of information that must be processed. Individuals with autism may have difficulty with this type of task because their neural systems cannot handle such large amounts of information rapidly. This problem may be conceptualized as one of “bandwidth,” that is, the neural connections in the brain with autism transmit the information at either a slower rate or in a degraded form (the result of the inability to process all of the information). These differences in the way in which individuals with autism process information that demand increasing neural resources or speed of processing are theoretically attributable to differences in brain functioning, differences that are now beginning to be defined in autism. Functional magnetic resonance imaging (fMRI) studies have begun to demonstrate a reduced capacity to bring regions online for processing information in autism.4 Dual tasks have demonstrated remarkable constraints on processing capacity.5 As more such studies are published, the neural counterpart for the difficulty in processing complex information will become increasingly clear, as will the basis for the integrity or even enhanced capacity in other areas. Other aspects of complexity are also important. Online processing changes in relation to the factors that are inherent in any particular scenario. Important factors that are to be considered are the developmental age of the individual, the overall cognitive ability of the individual (a reflection of general processing strength), individual characteristics (pattern of cognitive strengths and weaknesses that the individual would have had even if he or she did not have autism, that is, some individuals are better at mathematics, others at spatial knowledge, art, or language, etc.), the demands of the “to-be-processed” information, whether external aids to reduce processing demands are provided, and other competing demands on processing. To predict whether or not an individual with autism would have difficulty processing particular information, all of these factors must be delineated. In other words, an individual with autism might have difficulty processing information because it demands cognitive resources that have not yet developed. On the face of it, this is no different from other children. However, in autism, the difficulty occurs for information that other children matched for age and cognitive ability can handle without difficulty. Therefore, age by itself is not a good predictor of what an individual with autism can handle with regard to cognitive processing. A second factor is prior exposure to the information. Individuals with autism are renowned for reliance on scripts for conversation and answering questions. Response speed and accuracy can be remarkably fast when scripted knowledge is used. However, when such knowledge does not apply and new knowledge must be processed, response and processing speed can drop substantially. In developing this model of autism, it is also necessary to explain why individuals with autism are adept at processing certain types of information. For example, our neuropsychological studies of adults and children have indicated a relative strength in the area of spelling. One might argue that spelling is a complex skill involving the holding of information online (the auditory production of the word to be spelled), while the speller retrieves previously stored information about the word and the

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 387 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

387

relationship between phonemes and graphemes. However, the very nature of spelling may lend itself well to the way in which the brain with autism handles information. There is only one way to spell a word (“judgment” and “judgement” notwithstanding, or English vs. American spellings). Therefore, decisions do not have to be made between multiple meanings nor does situational and cultural information have to be considered. The most complex part of spelling is the phoneme–grapheme relationship. In English there are a number of graphemic representations for a particular phoneme (“read” and “red,” for instance). This might necessitate a level of analysis that would support an argument that spelling is “complex information.” However, if the spelling is accomplished by retrieval of a visual representation of the word without consideration of the phoneme–grapheme representation, the task becomes much less complex. One word, spelled one way, has one visual representation. All of this processing can be accomplished within a fairly restricted area of the brain alleviating the need for the coordination of diverse brain regions to accomplish the task.

NEUROIMAGING VALIDATION

OF THE

COGNITIVE PROFILE

In support of this neuropsychological profile study, Luna and colleagues6 published an fMRI study demonstrating an activation pattern analogous to the aforementioned profile. The fMRI study involved a spatial working memory task (oculomotor delayed response task), which typically activates left dorsolateral prefrontal cortex. In the control group, there was substantial activation of left frontal cortex and left posterior visual and parietal cortices. However, in the autism group there was almost no activation in left frontal cortex, but there was bilateral activation of posterior cortical regions indicating near total reliance on visuospatial areas for performance of this task. The activation patterns for both groups are shown in Figure 17.1. Thus, there was a failure of the autism group to activate higher-order circuitry underlying executive functions and a reliance on lower-order circuitry of visual cortex for task performance, which was activated bilaterally.

CONFIRMATION

OF

REDUCED INFORMATION PROCESSING CAPACITY

Validation for the above profile was sought in several other ways. One prediction of the complex information processing model was that there was a reduced capacity for processing information in individuals with autism. One manifestation of that would be a reduced capacity for dual task processing or multitasking. A review of the available literature on autism yielded a relevant study. Individuals with autism and a matched group of typical individuals performed individually and simultaneously a digit recall task and motor tracking task.5 When the autism group performed the tasks individually, their performance was equivalent to that of controls. For the controls, task performance was similar whether tasks were performed individually or simultaneously. For the autism group, a 40% decrement in performance occurred when performing the tasks simultaneously compared to individually. These are relatively simple tasks and even under these simple demands, the average

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 388 Friday, March 10, 2006 2:27 PM

388

Understanding Autism: From Basic Neuroscience to Treatment

FIGURE 17.1 (A color version of this figure follows page 236.) Activation during a spatial working memory task. (Courtesy of Dr. John Sweeney and Dr. Beatriz Luna.)

IQ individuals with autism experienced a serious decline in performance, demonstrating a significant constraint in their information processing capacity.

UNDERCONNECTIVITY

AND

OVERCONNECTIVITY

OF

NEOCORTICAL SYSTEMS

A second effort to validate or extend the understanding of the altered pattern of information processing in autism was through fMRI studies of tasks corresponding to those in the neuropsychology profile battery. One of the first of these involved sentence comprehension.4 This study of 17 high-functioning adults with autism and matched controls demonstrated increased activity in Wernicke’s area, which processes word meaning, and decreased activity in Broca’s area, which plays an integrating role in sentence comprehension. In addition, reliable different reductions in the functional correlations (the degree of synchronization or correlation of the time series of the activation) were present between ten cortical language region pairs in the autism group as compared to the control group. The alterations in cortical activity corresponded to the superior performance on tests of formal language and impaired performance on language comprehension tests reported in the neuropsychological profile study.* The functional underconnectivity among cortical language regions is consistent with abnormalities in white matter connections, whereas the alterations in cortical activity suggest changes in local connectivity within gray matter. Analogous alterations in functional connectivity have also been demonstrated in fMRI studies utilizing problem solving,7 social cognition,8 and a working memory task with face * This pattern also corresponds to the focus on parts (words) and failure to appreciate the global whole (sentences) discussed below in the local-global processing account and in the central coherence account. Thus multiple terms have been used for the same alteration in information processing in autism.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 389 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

389

stimuli,9 thus documenting functional underconnectivity underlying high-level abilities in all elements of the diagnostic triad of autism.

ANOTHER ASPECT OF THE INFORMATION PROCESSING IMPAIRMENT: LOCAL-GLOBAL PROCESSING Alterations in information processing in autism have also been characterized in terms of a local-processing bias and a global-processing deficit. This phenomenon originated with observations of disturbances in perception. Perceptual processing abnormalities in autism have focused mainly on two areas — deriving organized wholes from perceptual parts and face processing. Face-processing impairments have involved perception of face identity, memory for faces, face emotion, and face gender; these will be discussed in a separate section that follows. The relationship between impairments in part-whole (local-global) processing and face perception has not been investigated to any extent, but a recent paper by Behrmann and colleagues10 suggests that perceptual impairments in face perception, object perception, and deriving wholes from parts may be related phenomena relying on a common mechanism. This is an important issue as Brosnan and colleagues11 concluded in a recent report on autism by posing the question of the relationship of “low-level” perceptual impairments to high-level executive function impairments, as did Happé in an earlier paper about low- and high-level central coherence.12 The local-global model of information processing proposes that individuals with autism are predisposed to seeing details of a stimulus rather than the global whole, whereas typical individuals are predisposed to seeing the whole before the details. That is, the circuitry of the brain is wired to enable typical individuals to first see the global whole (global precedence) then the details, whereas in autism the circuitry of the brain remains in an immature state that is biased to perceive details (local bias) with difficulty or inability to see the global whole. In typical individuals, global precedence also results in global interference when the person is asked to focus on local details. In individuals with autism, this model predicts an absence of global precedence, a lack of global interference, and a global-processing deficit. Debates in autism have focused on whether the focus on details arises from a local-processing bias or a global-processing deficit or both, and vice versa. The evidence to support a local-processing bias — to focus more on the parts of a stimulus and to not see the whole — in individuals with autism has included evidence of resistance to visual illusions (which has not replicated),13,14 failure to perceive impossible geometric figures,15 enhanced capacity to detect local targets in visual-search tasks,16,17,18,19 superior performance on Embedded-Figure tasks,12,20 an inability to inhibit processing of irrelevant details,21 and peak performance on the Block Design and Object Assembly subtests of the Wechsler Intelligence Scales.22 Not all studies have reported findings consistent with a local bias. Those that have not usually included individuals with Autism Spectrum Disorders (ASD) who were less affected than those with autism.23,24 Variations in test versions and presentation formats may also have contributed to variability in results.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 390 Friday, March 10, 2006 2:27 PM

390

Understanding Autism: From Basic Neuroscience to Treatment

Some, but not all, studies have also shown a lack of global precedence in autism.15,25,26,27,19,21 Along the same lines, a recent study has reported an impairment in processing inter-element relationships and a bias away from using gestalt grouping principles, which is akin to, if not identical to, the concept of global processing.11 Two hypotheses have often been placed in opposition in the above studies: (1) there is a local-processing bias in autism that interferes with and results in a globalprocessing deficit or (2) there is a global-processing deficit that results in a focus on local elements. The third alternative is that the local-processing bias and the global-processing deficit coexist and reflect a common and interrelated developmental neurobiological phenomenon. That is, immature circuitry is predisposed to encoding details and only with maturity does neural circuitry develop that allows concepts to emerge, but this occurs at the sacrifice of attending to details. It could be that the disturbance in the development of neural circuitry in autism results, not only in a failure of the higher-level circuitry to mature, but overgrowth of lowerlevel circuitry that results in enhanced perception of details. A recent fMRI study provided an interesting perspective on the local vs. global debate. Koshino and colleagues28 completed a study of verbal individuals with autism and normal IQ compared to matched controls using a letter working memory task. Task performance of the autism and control groups was comparable, but surprisingly, the autism group had increased activation of the right frontal executive and posterior visual areas, whereas the typical control group activated left frontal executive and language areas as expected for this task. Despite having average verbal IQ scores, the autism group appeared to process letters as graphic images using elementary visual areas rather than as letter names using language areas. The autism group had reduced this task to its most elementary level from a cognitive standpoint, although it was not apparent from a behavioral level that they were doing the task any differently as compared to the controls. This has several implications. If the letters had been from the Russian alphabet and all participants were unfamiliar with Russian, then the individuals with autism might have had a task advantage and performed better than controls. Second, investigators could have inferred from the behavioral performance of the autism group that their dorsolateral prefrontal cortical function was superior or at least spared. Thus, caution must be used in drawing anatomic conclusions from behavioral data as individuals with autism may be performing a task using a different cognitive and brain basis than what is typical.* Third, investigators might have inferred intact comprehension of letters from intact reading when in fact the autism group was not seeing the letters as phonologic representations but strictly as graphics without linguistic associations. This study also reported a factor analysis of brain structure suggesting that the brain is organized differently in the autism group than in the control group.

* For example, it might be inferred from intact performance on the n-back letter task of working memory that there is only selective involvement of the frontal lobes. However, an fMRI study of this task has shown that nonretarded individuals with autism are relying more on the temporal occipital lobe (see NeuroImage 2005, 24: 810–821). Eye movement studies can be used to provide a more reliable assessment of frontal function than neuropsychological or cognitive tests. Their localization is well established and sensitivity to impairments is high.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 391 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

391

INTERRELATIONSHIPS BETWEEN LOCAL-GLOBAL PROCESSING, OBJECT PROCESSING, AND FACE PROCESSING Face processing as well as object processing, in general, rely on the abstraction of both local and global features for identification. The degree to which local and global features may be important may differ depending on the type of discrimination needed. For example, perceptual similarity for two faces within the same gender is greater than for two faces of different genders. Moreover, decisions about similarity for two items of different categories, for example, a face and a car, can be done at the feature or local level (e.g., does the item have a tire), whereas decisions about two items at the subordinate level (e.g., is this a sedan or a coupe) or exemplar level (e.g., is this a Honda Accord or a Toyota Camry) involve precise discrimination based on both featural and configural knowledge. Behrmann and colleagues10 found that adults with autism compared to matched controls were more impaired on tasks that involved discrimination of items at the exemplar level (discriminating between two female faces) than at the basic level (discriminating a car from a face), both for faces and for objects. They then demonstrated on two different tasks that the autism group had a local bias and was impaired in their capacity to derive global wholes and that this impairment correlated with their impairments in discriminating between faces and discriminating between objects. Taken together, the results suggest that the impairment in processing faces also extends to objects (nonsocial stimuli) and that the bias for local processing might adversely impact the capacity for processing objects and faces.

EXTENDING THE LOCAL-GLOBAL PROCESSING ACCOUNT TO HIGH-LEVEL TASKS The local-global processing issue was ultimately extended to include studies of the reproduction of the ROCF, which requires executive planning to produce the global elements, as opposed to the perceptual tasks described in the above studies. In an initial study,29 10- to 17-year-old children with autism with Full Scale IQ scores between 76 and 109 were asked to copy and reproduce the ROCF from memory after a 3-minute delay. The children with autism perseverated on lines and produced more details than global elements. However, this study was criticized because the scoring method was not objective. A second study of adults aged 18 to 49 years with IQ scores of 85 to 135 reported no differences in details as opposed to global elements after a 3-minute delay.20 The authors commented on a “trend for the group with autism to draw in a more fragmented way using more lines to complete the figure” (p. 531). However, this study used a modified version of the ROCF that was considerably less complex than the one used in the original study. In a third study, Minshew and Goldstein30 reported recall deficits under immediate and delayed copy conditions of the ROCF in 52 nonretarded adolescents and adults with autism, which was attributed to failure to use organizational strategies. A fourth study failed to find differences on the ROCF but used observers to rate performance rather than an objective scoring system.31

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 392 Friday, March 10, 2006 2:27 PM

392

Understanding Autism: From Basic Neuroscience to Treatment

Complex figures other than the ROCF have been used in an attempt to reconcile these diverse results. Mottron and collegues32 used impossible and possible figures and reported that individuals with autism favored local features and produced fewer global features than controls. They could not decide whether this was the result of impairment at the perceptual level or at the conceptual level. Booth and collegues33 attempted to distinguish between perceptual and executive function deficits by asking children with autism to add features when they reproduced a figure they had previously drawn, which would require planning. The study revealed both a detail-oriented approach to the drawing as well as poor planning in adding the requested new features. Thus, although there appear to be parallel or analogous “part-whole” impairments in visual perceptual tasks and executive tasks, it is not clear how these are related. Are these analogous but separate impairments at different levels, as this study suggests? Or are the low-level impairments causing or contributing to the high-level impairments? Or are they the consequence of a common neurodevelopmental process as suggested above? These questions are unanswered at present but are important for the design of new cognitive intervention programs.

EXECUTIVE FUNCTION AND ABSTRACTION: THE NONSOCIAL IMPAIRMENTS IN AUTISM One prominent model of autism has emphasized the widely documented impairments in executive function as the primary deficit.34,24,35,36 Executive functions are a collection of cognitive skills thought to be mediated by prefrontal regions of the brain. Typical executive functions include planning, working memory, inhibition of prepotent responses, impulse control, shifting set, and monitoring of action.37 Individuals with autism have been shown to have deficits in planning tasks such as the Tower of London or Tower of Hanoi.24,38,39,40,41 However, mixed results have been obtained in studies of other executive functions in autism, such as inhibition of prepotent responses42,43,27,44 and working memory.45,46,47 Additional processing requirements such as a problem-solving element or the requirement to follow arbitrary rules appear to be, in particular, the key to the occurrence of relative difficulty on these tasks by individuals with autism. Nonretarded individuals with autism successfully perform executive function tasks that do not have a problem-solving element but consistently demonstrate impaired performance on tasks that have a flexibility or problem-solving component.1,48,2 Thus, while some aspects of executive functioning are impaired in autism, other aspects are not. Although some models have tried to characterize deficits in executive function as the central processing deficit of autism, it is hard to argue for this level of specificity. Executive dysfunction occurs in clinical populations other than autism such as attention deficit hyperactivity disorder and Tourette’s syndrome49,50,51; therefore, executive dysfunction is not a cognitive impairment that is specific to autism. In response to this issue, an attempt has been made to develop a differential profile of executive impairment that would be distinctive for autism40 (but see Sergeant et al.52). Other studies have also indicated that executive dysfunction may not be a universal feature of autism.53,54,55 Although dysfunction in executive processing may

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 393 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

393

be part of the cognitive profile of autism, a model based solely on this cognitive construct is incomplete. How, for example, are the posterior hemispheric aspects of language, prosody, and pursuit eye movements explained by deficits in executive function and frontal lobe dysfunction? How are postural physiologic abnormalities and mental retardation in 70% of individuals with autism explained?

ABSTRACT REASONING While acknowledging that impairments occur in some aspects of executive function in individuals with autism, Minshew and colleagues have also focused on other areas of cognition, particularly abstract reasoning.48 Abstract reasoning is the process of considering and manipulating information about events, objects, and concepts not in the immediate environment or removed in time. The exploration of abstract reasoning abilities in autism allows for the further characterization of the difficulty individuals with autism have with higher-order integration of information. Although not an ideal construct for neurobehavioral characterization, abstract reasoning is based on previously established models of cognition leading to particular measures of integrative functioning. Minshew and colleagues argue that an inability to engage in abstract reasoning underlies the deficits in pretend play, behavioral inflexibility, and generalization that characterize autism. Abstract reasoning incorporates some of the elements of processing that are generally placed under the umbrella of executive functioning such as planning and decision making; however, it also incorporates the inability of individuals with autism to acquire knowledge from concrete instances and to generalize this knowledge to new experiences. The recognition that individuals with high-functioning autism often have prominent deficits in the areas of conceptual reasoning and problem solving is not new.56,57,58 However, Minshew and colleagues have proposed that only certain aspects of this area of cognitive function may be affected in autism. In a study with 90 non–mentally retarded individuals with autism and 107 normal controls, significant group differences were obtained on all abstract reasoning tasks. However, with the exception of the Wisconsin Card Sorting Test, differences on concept identification tasks were not clinically significant.48 In concept identification, the abstraction to be understood is established within the task and only has to be identified. The classic example of concept identification is the Halstead Category Test.59 In concept formation tasks, the concept is not inherent in the test material but has to be generated by the individual who must decide where to set the boundaries of the conceptual relationship. In this study, concept identification and concept formation loaded on separate factors in the autism group but not in the control group. In addition, stepwise discriminant function analyses revealed that two tests of concept formation correctly classified 78.4% of the cases, whereas concept identification tasks did not pass the tolerance test. Minshew and colleagues concluded that in nonretarded individuals with autism, a dissociation occurs between rule learning or concept identification and higher-order conceptual integration; therefore, these might be separable cognitive abilities with distinct biological bases. Although concept identification, concept formation, and rule learning may all require abstract conceptual reasoning, they may be differentially affected such that high-functioning individuals with autism learn

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 394 Friday, March 10, 2006 2:27 PM

394

Understanding Autism: From Basic Neuroscience to Treatment

rules and identify concepts much better than they form concepts and solve problems. In fact, at higher IQ levels, individual with autism may do as well on rule-learning tasks as normal individuals. The proposal of a dissociation between rule learning and concept formation in autism is consistent with Sloman’s60 proposed two systems of reasoning — one rulebased and one associative. These two systems are differentiated on their underlying computations. The rule-based system follows principles of formal, logistical, or deductive reasoning (if X, then Y). Deductive reasoning relies on factual knowledge, formal rules, and mental models.61 The associative system is based on computations of underlying relationships such as similarity structures and temporal contiguity. Minshew and colleagues suggest that individuals with autism can perform reasoning tasks in which the application of rules is explicit. They have more difficulty when they have to decide whether to apply a rule and have even more difficulty when they have to extract the rule from provided information. The latter requires the same cognitive process that underlies mental inferencing,60 a skill that individuals with autism have difficulty with across cognitive domains. As with executive dysfunction in autism, not all aspects of abstraction are impaired in autism. Attribute identification and rule learning may be intact in high-functioning individuals. What appears to be unique to autism is the inability to flexibly apply rules to changing circumstances or contexts, that is, to fully understand the conceptual meaning of the rule and to apply the rule to novel problems. The challenge for the complex information processing model is to explain the apparent dissociation between rule learning and concept formation. At first consideration, these would both appear to be tasks that require higher-order integration. However, the consistent nature of the rules, once learned, may actually place less demand on the brain’s information processing system and may require the coordination of a limited number of cortical centers. Concept formation, due to the need to consider a wider range of information, may place higher demands on the information processing system. These speculations still need to be confirmed through studies of brain function.

DEFICITS IN PART-WHOLE PROCESSING AND IN ABSTRACTION: A BASIS FOR RESTRICTED AND REPETITIVE BEHAVIOR DSM-IV criteria for autism specify three symptom categories for autism — qualitative impairment in social interactions, qualitative impairments in communication, and restricted, repetitive and stereotyped patterns of behavior, interests, and activities with onset prior to 3 years of age. A reading of the symptoms and of the related literature reveals that there is a relatively sound and extensive understanding of the cognitive and language basis of the symptoms in the first two categories. However, the third symptom category lists four behaviors, without reference to an underlying cognitive mechanism: stereotyped or repetitive motor mannerisms, persistent preoccupation with parts of objects, inflexible adherence to specific nonfunctional routines or rituals, and encompassing preoccupations with one or more stereotyped and restricted patterns of interests that is abnormal in intensity or focus. With the

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 395 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

395

exception of the first, which occurs in mental retardation in general, the other three behaviors share a focus on details and a failure to appreciate the whole either at the perceptual or conceptual level.48 These behavioral problems are likely to be attributable to the problems with reasoning and problem solving in autism. Not surprisingly, the impairments in reasoning correspond to well-known behavioral intervention programs that use thousands of massed trials to train basic attributes of objects in very impaired young children with autism. With some progress, children with autism learn rules or scripts, but may not generalize this knowledge outside of the original context in which they were taught (inflexible use of concepts is consistent with incomplete attainment of the concept). With further developmental progress, children with autism attain rules but apply them rigidly, or they accommodate ordinary exceptions but not unordinary exceptions and cannot cope with situations where known rules are insufficient. In addition, individuals with autism frequently have difficulty self-initiating the structure necessary to organize or negotiate life independently.

OTHER MODELS FOR AUTISM: SOCIAL COGNITION, A CORE DIAGNOSTIC PHENOTYPE Impaired social cognition is a distinctive element or hallmark of autism. Any characterization of autism must explain the clinical presentation of the disorder that includes the lack of attention to social stimuli and a dissociation between overall cognitive level and adaptive functioning. Even in milder cases of autism, individuals are distinguished by their social “awkwardness” (impaired reciprocity) and lack of ability to take the perspective of others (theory of mind deficit).

EARLY SOCIAL IMPAIRMENTS Dawson and colleagues have proposed a social motivation theory of autism. This theory argues that the primary or causative deficit in autism is an early lack of motivation to socialize and this deficit leads to the other deficits that characterize the disorder. Early social impairments in children with autism include deficits in social orienting, joint attention, emotion perception, affective sharing, and imitation. A failure to orient to social stimuli (“social orienting”) has been proposed as one of the earliest and most basic impairments in autism.62,63,64,65,66 In typical infants, the early attentional preferences for social stimuli provide opportunities for the infant to engage in social exchanges and facilitate the acquisition of later social and communication skills.67 Young children with autism both fail to orient to social stimuli such as human sounds (clapping, calling names, etc.) and fail to prefer human vs. nonhuman speech.68,69 Similarly, impairments in joint attention can also be observed earlier than other core impairments70 and are correlated with later language ability in children with autism.71 Mundy expands on this by suggesting that joint attention skills increase incidental social learning.65 The pervasiveness of social-cognitive impairments early in the course of autism have resulted in a number of hypotheses including the following: (1) failure in

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 396 Friday, March 10, 2006 2:27 PM

396

Understanding Autism: From Basic Neuroscience to Treatment

attention shifting,72 (2) impairments in representational capacity,73 and (3) deficits in social motivation, which deprive the system of experience critical to the formation of neural systems that support social processing.74,75 The attentional hypothesis proposes that the ability to rapidly shift attention between different stimuli and different modalities is compromised in autism. If children with autism are not able to rapidly shift attention or time direction of attention, then they may be unable to participate in social exchanges. A failure to attend to early, simple social and affective interactions would result in a failure to develop the basic mechanisms needed for more complex interchanges. In contrast, Meltzoff and colleagues have proposed that imitation impairments might lead to a failure to correctly represent others.76,73 Meltzoff and Decety73 have argued that typical infants distinguish themselves from others but intrinsically represent others as “like me.” This representational capacity forms the basis for imitation. Lastly, Dawson and others posit that the social impairments result from, or at least are compounded by, a fundamental deficit in social motivation.77,78,65,75 Decreased social motivation might be based on abnormalities in neural systems such as the dopaminergic reward system,79 the amygdala,80,75 or the neuropeptides vasopressin and oxytocin.81 Supporting the amygdala hypothesis, Dager and colleagues found that 3- to 4-year-old children with autism have significantly increased amygdala volume in excess of increased cerebral volume, compared to typical and developmentally delayed children82; larger amygdala volume was found to be associated with slower development of social skills during the preschool period.83

FACE PROCESSING: EARLY EMERGING IMPAIRMENTS Face processing has been posited as critical to the development of social relationships and theory of mind, or the understanding that other individuals have thoughts and inferences about what those thoughts might be.84–88 Many of the early social impairments seen in toddlers and young children with autism, such as eye contact, joint attention, responses to emotional displays, and face recognition, involve the ability to attend to and use information from the face. Retrospective video and case studies of early autism provide the earliest reports of what is known about face processing in children who later are diagnosed with autism. Dawson et al.89 reported that a young infant who was diagnosed with autism at 1 year of age had relatively good eye contact until about 6 months of age, but poor eye contact by 13 months of age. Similarly, using retrospective videotapes, failure to look at others discriminated 1-year-old infants who later received a diagnosis of autism from those with typical development90 (see also Adrien et al.91). Both early and late components of the ERP have been studied during face and object perception and memory in children and adults with autism. In typical adults, faces evoke a specific ERP component that is negative going and peaks at approximately 170 msec.92 McPartland and colleagues93 found an altered N170 pattern in adolescents and adults with ASD. High-functioning individuals with ASD exhibited a slowed N170 to faces compared to control stimuli (furniture) and failed to show a face inversion effect. The authors concluded that early

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 397 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

397

structural encoding of faces was disrupted in autism. Slowed processing of faces and atypical cortical representation was also found in parents of children with autism using ERPs.74 In addition to early stage processing of faces, face and object memory differentially evoked later stage components such as the Nc, P400, and slow wave in children. Dawson et al.77 evaluated children with ASD during ERP tests of face and object recognition. Participants (ages 3 to 4 years) were shown a picture of their mother (familiar face), a picture of an unfamiliar female face, a picture of a familiar object (favorite toy), and a picture of an unfamiliar toy. Unlike the comparison children, children with ASD did not show amplitude differences in P400 or Nc when viewing the familiar vs. the unfamiliar face. However, similar to the control children, ASD children did show amplitude differences in both P400 and Nc when they were looking at a picture of a familiar object (a favorite toy) vs. an unfamiliar one. This was interpreted as indicating a social processing deficit.77 By middle childhood, children with autism demonstrated marked deficits in face processing when compared to both mental and chronological age matched peers. This includes tests of face discrimination,94 face recognition,95,96,97,98 and face memory.99 In comparison to typically developing children who show better memory performance for faces than nonface visual stimuli, children and adults with autism perform comparably on face and nonface tasks100 or show better performance on nonface tasks (e.g., memory for buildings) than on face tasks.95,101 Several studies suggest that individuals with autism process faces using abnormal strategies. Children with autism demonstrate a pattern of performance in which they are better at recognizing isolated facial features and partially obscured faces than typical children102,94 and show better performance on memory for the lower half of the face than the upper half, during childhood.103 Some studies of visual attention to faces suggest that individuals with autism exhibit reduced attention to the upper region of the face, such as the eyes and nose, relative to typical individuals.104,105,106 One study found that, when viewing emotionally expressive faces, children with autism may exhibit more typical patterns of visual attention, fixating more on the eyes and mouth than other parts of the face.107 Individuals with autism recognize inverted faces better than control participants102,103 and spend equal time looking at inverted faces compared with upright faces.107 It has been suggested that this pattern of deficits represents a failure to process faces configurally,108 with an emphasis on local detail rather than global patterns.12 In summary, several studies have documented early-emerging impairments in face discrimination and recognition in individuals with autism. Studies of young children, adolescents, and adults with autism, consistently find slower speed of processing of faces, a failure to show the expected speed advantage of processing faces vs. nonface stimuli, atypical scalp topography suggesting abnormal cortical specialization for faces in autism, and relations between processing of face stimuli and performance on social-cognition tasks. The affected ERP components suggest that both “early stage” processing of faces is impaired as well as later-stage higherorder cognitive processing that reflects the use of face information. As a caveat, “early stage” processing most likely reflects processing within the ventral stream

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 398 Friday, March 10, 2006 2:27 PM

398

Understanding Autism: From Basic Neuroscience to Treatment

and is altered by both bottom-up and top-down influences. Furthermore, there is evidence that individuals with autism use atypical strategies for processing faces characterized by reduced attention to the eyes and piecemeal rather than holistic strategies.

EVIDENCE FOR EARLY IMPAIRMENT PROCESSING AND MEMORY

IN

FACIAL EMOTION

Very little is known about emotion processing and memory in young children with autism. Young children with autism fail to attend to the emotional expressions of others.68,109,110 Children with autism, aged 3 to 4 years, exhibited significantly slower early (N300) brain responses to fear as compared to typically developing children.111 Children with autism also failed to show larger-amplitude negative slow-wave responses to the fear faces as was found in typical children, regardless of mental age. The delay in response to fear faces suggests that information processing speed is compromised, and the abnormal topography suggests failure of cortical specialization or atypical recruitment of cortical areas. In addition, children with ASD who displayed a faster N300 latency to the fear face exhibited better joint attention, fewer social orienting errors, and more time spent looking at an experimenter expressing distress. These findings suggest that slower information processing speed for emotional stimuli is associated with more severe social attention impairments in children with autism. Individuals with autism often demonstrate impairments on behavioral tasks that involve matching or categorizing pictures of facial expressions of emotions,112,113,114 recognizing facial expressions in photos,115 and cross-modal matching of photographs of facial emotions and their corresponding vocalizations.116,117 This deficit extends to finding facial expressions in an array, and labeling facial expressions.94 No deficit on sorting and matching facial expressions has been found when individuals with autism are matched to controls on verbal mental age118 or matched to mentally retarded controls.97 Further, two reports suggest that performance within the autism group was similar on perception of emotional and nonemotional expressions119 or on emotional facial, nonemotional facial and nonface stimuli,120 suggesting that the deficit was not specific to the emotion expression. In addition, matching of facial expression has been correlated with verbal abilities.118,121,122,123

EXPLANATIONS

FOR

FACE PROCESSING IMPAIRMENTS

The domain of face processing represents one area in which the characterization of a social cognitive impairment has resulted in theories that are both domain specific (i.e., applicable to face processing only) and domain general (i.e., represent a failure of general processes). The explanation of face-processing impairments in autism includes hypotheses that autism involves the following: (1) a fundamental general problem in perceptual binding; (2) an early-emerging general higher-order perceptual/cognitive deficit, as described above in the model proposed by Minshew and colleagues, that prevents infants with autism from extracting perceptually relevant information from faces (e.g., prototype formation124,125);

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 399 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

399

(3) dysfunction of the specific neural mechanism that supports face processing, namely, the fusiform gyrus; (4) disruption in configural or global processing that supports face processing10; and/or (5) a deficit in social motivation, which deprives the system of experience critical to the formation of neural systems that support social processing. The perceptual binding hypothesis is derived from the weak central coherence hypothesis and specifies that individuals with autism fail to process items as complex or meaningful wholes.126,127 Both Frith128 and Brock et al.126 suggest that the features associated with weak central coherence result from a failure to integrate information at a “higher level”; this is consistent with the description of Minshew et al. that autism is a disorder of complex information processing.1,129 One mechanism may be the failure of the brain to bind information temporally, a process that involves distinct regions activating in synchrony. When the brain needs to encode information from a face (such as the parts, as well as their configuration), this information needs to be bound together. In a small study, Grice et al.130 found decreased gamma activity (a neural mechanism of feature binding) in individuals with autism, compared to controls, while processing faces. This hypothesis is not exclusive to deficits in face processing but may account for deficits in information processing in general. Face-processing deficits may also arise from a failure to extract relevant information from faces (e.g., prototype formation124,125). Strauss125 suggests that individuals with autism fail to extract similarities across exemplars in the same manner. In a prototype theory, items within a category share a set of features that vary along a dimension. Over time, these features are abstracted from a limited number of examples and are used to build up a representation. In autism, children may not be abstracting the features that are important for the category and/or creating feature continuums that have an average value and a deviation from typicality. Preliminary results from Strauss suggest atypical ratings of facial typicality and attractiveness of faces as well as a failure to show the “other race” effect.125 Similar to the perceptual-binding theory, impairments in categorical or prototype formation would not be limited to face processing. In the fusiform hypothesis, face-processing impairment would represent a primary deficit in autism. The dysfunction of the fusiform gyrus might be a neural deficit that was present at birth and thus could be screened for during early infancy. This deficit, early on, would be specific to faces, and it would be expected that the “typical” pathway for processing faces would be inherently deficient. It is unclear, however, if impairments to the fusiform gyrus would disrupt specific face processing or domain general processes such as individuation of exemplars, expert individuation, or configural processing.131 Lastly, Dawson and Webb77,74 and others78 have proposed a social motivation hypothesis, namely, the hypothesis that the behavioral and electrophysiological indices of face-processing deficits, described above, are secondary to a primary deficit in social motivation. The notion that individuals with autism lack social motivation is based partly on clinical observation; DSM criteria for autism include “a lack of spontaneous seeking to share enjoyment, interests, or achievements with other people” and “lack of social or emotional reciprocity.” This difficulty might

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 400 Friday, March 10, 2006 2:27 PM

400

Understanding Autism: From Basic Neuroscience to Treatment

stem from abnormalities in either (1) the reward system per se79 or (2) neural systems that might be important for the perception of social reward, such as ability to form representations of others as “like me,” i.e., similar to self in some way.132,133,134,88

THEORY

OF

MIND (TOM)

There is a marked shift in ToM functioning in typical children between the ages of 3 and 4 years of age, with continuous development through childhood. The recognition that individuals with autism have a deficit in ToM, the ability to be aware of and take into consideration the thoughts of others, was an important development that promoted understanding of the social and communicative difficulties of these individuals (see Baron-Cohen135 for a review). However, although the recognition of the theory of mind deficit has helped to make autism more understandable, the underlying neuropsychological mechanism that results in the lack of formation or use of a theory ToM has not been delineated.136 Several studies have indicated that the failure of individuals with autism to develop a ToM is not a distinct cognitive deficit but may be related to underlying deficits in areas such as language137 and/or reasoning.138 Recent research on children with language impairment suggests that difficulty with ToM may be related to difficulty with processing complex language. Miller139 examined the understanding of false belief in children of ages 4 to 7 years with specific language impairment (SLI) as compared to two groups — age-matched and language-comprehension-matched controls. Four conditions of a standard false belief task that varied with respect to linguistic complexity were administered to the children. When the linguistic complexity was low, the SLI group performed similarly to the age-matched control group. However, when the linguistic complexity was high, the children with SLI performed similarly to the younger, languagecomprehension-matched control group. Miller interpreted these results as indicating that linguistic competence affects performance on false belief tasks for children with SLI. The results of this study on children with SLI suggest that performance on measures of ToM may be affected by the individual’s language comprehension abilities. In fact, studies on individuals with autism have reported a connection between performance on measures of ToM and higher-level language tasks. Happé140 suggested that the difficulty autistic individuals have with comprehension of metaphoric and ironic statements is related to one of the core features of autism — the lack of a ToM. Individuals with autism who had difficulty on first- and second-order ToM tasks had more difficulty completing metaphoric sentences and answering binary choice questions about metaphors than individuals with autism, who only had difficulty on second-order ToM tasks, or matched controls. These same participants (poor performers on the first- and second-order ToM tasks) and a second group of individuals with autism, who failed second-order ToM tasks, had difficulty answering questions about ironic statements. Happé interpreted these results as indicating that the understanding of mental states, particularly intention,

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 401 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

401

was important for understanding metaphors and irony. An alternate explanation is that comprehension of metaphoric and ironic language and the formation of a ToM are dependent on the same underlying cognitive process and that it is this cognitive process that is deficient in individuals with autism. For individuals with autism, performance on ToM tasks may be influenced by language comprehension abilities or both of these behaviors may be influenced by a third factor. One such factor may be the effect of the complexity of the stimuli to be processed, a scenario which would be consistent with the complex information processing model of autism proposed by Minshew et al. At least one ToM study with normally developing 3-, 4-, and 5-year-olds suggests that the formation of a ToM is related to changes in capacity to “hold in mind.”141 Individuals with autism may be able to perform ToM tasks and language comprehension tasks when the tasks are at a lower level of complexity. However, as the complexity level of the tasks is increased relative to a person’s cognitive functioning level, individuals with autism may demonstrate increasing difficulty. A small number of studies have been completed in which the associated neurophysiological processing was measured while participants with autism completed ToM tasks. The first reported study was a PET study conducted with individuals with Asperger syndrome.142 The Asperger group demonstrated less activation in the medial prefrontal region as compared to a group of normal controls. Baron-Cohen et al.53 completed an fMRI study in which participants judged a person’s emotional states from photographs of the eye region. The ability to make these judgments was assumed to be related to the formation of a ToM. The results indicated that individuals with autism had less extensive activation in the frontal regions and no activation in the amygdala. Castelli et al.143 completed a PET study of ten adults with autism or Asperger syndrome while watching animated sequences with two triangles that symbolically depicted social interactions. The group with autism demonstrated less activation in the cortical regions of the previously identified mentalizing network (medial prefrontal cortex, superior temporal sulcus at the temporo-parietal junction and temporal poles) and showed reduced functional connectivity with the superior temporal sulcus at the temporoparietal junction, an area associated with the processing of biological motion. The authors interpreted these results as suggesting that the mentalizing dysfunction in autism may have occurred because information was not transmitted from an area of the visual cortex responsive to form and motion to the superior temporal sulcus, reflecting a lack of top-down modulation. Although the notion of a lack of ToM is perhaps clinically useful, in looking for the brain basis of this behavior it is more useful to focus on what processing function is served by the “mentalizing network.” The functions of cortical networks cannot be directly reduced to their psychological equivalents. What is unknown at this point is if there is something unique or special in autism about this network or if the problems seen are simply indicative of a more pervasive underlying problem of recruitment of highly synchronized networks in response to the processing demands of dynamic, complex stimuli (which social information clearly is).

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 402 Friday, March 10, 2006 2:27 PM

402

Understanding Autism: From Basic Neuroscience to Treatment

OCULOMOTOR AND POSTURAL PHYSIOLOGY: BEYOND NEUROPSYCHOLOGICAL TESTS* The integrity of neural pathways in the brain can be investigated with several methods, most commonly through studies of evoked potentials and oculomotor and postural physiology. These methods provide information about neural pathways at multiple levels of the nervous system and about selected aspects of sensory, motor, and cognitive functions. Concurrent with the neuropsychological profile in the preceding text, the neurophysiological profile in autism is characterized by abnormalities in cognitive processing and neocortical circuitry, with intact early information processing, simple cognitive abilities, and relatively subtle disturbances in posterior fossa circuitry (e.g., brain stem and cerebellum). Whereas a comprehensive review of the neurologic aspects of autism is beyond the scope of this chapter (see Minshew et al.144), studies of oculomotor and postural abilities provide evidence of the basic functioning of a number of the neural systems proposed as underlying autism deficits.

OCULOMOTOR PHYSIOLOGY Eye movement studies can provide important information about cortical and subcortical regions that control eye movement activity. Such eye movement studies can be viewed as complementary to sensory-evoked responses, in that they provide information primarily about the motor system and sensorimotor integration. Responses to lights moving abruptly from one point to another can be measured (latency, accuracy, and peak velocity) to assess the integrity of saccadic eye movements, and the tracking of slowly but steadily moving targets can be evaluated to assess smooth pursuit eye movements.145 With regard to the cerebral cortex, saccadic eye movements report on frontal regions and pursuit eye movements report on posterior regions though other regions of the neuraxis are certainly involved in the generation of these eye movements. Kemner and colleagues found that children with autism made more saccades in a passive viewing task and during and between stimulus presentations than did control children.146,147 Using a gap-overlap paradigm, van der Geest et al.148 found that unlike the control group, the autism group showed no differences in performance on the gap or overlap conditions. This was interpreted as a deficiency in attentional engagement, which may be related to dysfunction in several areas of the brain, including the frontal eye fields, the superior colliculus, and/or parietal cortex. To investigate cerebellar function in autism, Takarae et al.149 investigated visually guided * The chapter shifts at this point from studies of cognition to a review of studies of neurophysiology. These studies provide information about how the brain is handling or transferring information and about localization of impairments. For example, the two types of saccadic eye movements can be used to probe frontal cortex and brain stem–cerebellum, and voluntary saccades and pursuit eye movements can be used to assess whether both anterior and posterior regions of the cerebral hemispheres are involved. It is helpful to understand the question being asked and the rationale for these studies in order to understand the meaning of the results for autism.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 403 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

403

saccades in 46 high-functioning individuals with autism with and without early developmental language delay and 104 typical individuals. They found reduced saccade accuracy in the presence of normal saccade latency, which suggested a subtle deficit in motor function rather than visual attention. Only those participants without a history of delayed language development exhibited significantly lower saccade gain. This pattern is consistent with reported histopathologic abnormalities in the cerebellar vermis, which is thought to control saccade adaptation,150 or its output through the fastigial nuclei. In a study of voluntary saccadic eye movements known to be subserved by discrete regions in frontal and parietal cortex,* Minshew et al.151 found significant abnormalities in cortically controlled eye movements in 26 adults with autism compared to 26 typical controls (see also Goldberg et al.152). The individuals with autism demonstrated significant impairments both in the frontally mediated ability to suppress saccades to unpredictable targets (context inappropriate response) and to shift gaze to remembered target locations (spatial working memory). These findings, together with findings of intact visually guided saccades (reflex shifts of attention) subserved by the cerebellum and brain stem, indicate that behavioral abnormalities in shifting attention in autism are secondary to abnormal executive processes and frontal dysfunction. No abnormalities in the basic capacity to shift attention supported by posterior fossa structures were found. Koczat et al.153 found deficits in unaffected family members using the same memory-guided saccade task (oculomotor delayed response task). Pursuit eye movements are dependent on neural circuitry that includes extrastriate areas of visual cortex involved in processing motion information, cortical eye fields and cerebellum involved in translating sensory information to motor commands, and striatum and brain stem that initiate motor commands. Because of the high demands for functional integration across brain areas, studies of pursuit eye movements provide an excellent method for studying functional brain connectivity in autism. In a study of 60 high-functioning children and adults with autism and 94 typical controls, bilateral pursuit deficits were observed in the autism group on closed-loop pursuit gain when tracking oscillating and ramp targets.154 The most apparent difference occurred for individuals with autism who were past midadolescence. The individuals with autism also performed more poorly on the initial open-loop stage of pursuit but only when the target moved into the right visual field (indicating a possible problem in left extrastriate area extraction of visual motion information or in transfer of information to sensorimotor areas for transformation to visual information into oculomotor commands). There were modest correlations between these oculomotor impairments and manual motor performance but no correlations with visual attention. These associations suggest that impairment * The previous paragraph discussed the findings with visually guided saccades, which are not subject to frontal control. The topic of this paragraph is voluntary saccades, which are subject to volitional and frontal control.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 404 Friday, March 10, 2006 2:27 PM

404

Understanding Autism: From Basic Neuroscience to Treatment

in sensorimotor integration is not specific to types of visual information or types of motor output. This study provided additional evidence of functional connectivity deficits in another domain outside the diagnostic triad — the sensorimotor domain. Although the brain areas that have been implicated in autism do overlap with the neural substrate of the pursuit system, no intrinsic abnormalities in any single area could result in the eye movement abnormalities documented in this study. The findings are most consistent with an intrinsic reduction in functional connectivity within the visual pursuit system as a result of a developmental disturbance in brain maturation. This study adds further support to the model of abnormal maturation of distributed neural networks involved in the integration of information as a fundamental characteristic of autism. The vestibular ocular reflex (VOR) keeps the eyes still when the head moves. The VOR consists of involuntary eye movements (nystagmus) with increasing slowvelocity component (SVC) during acceleration, a decaying SVC and, finally, a reversal of nystagmus direction. Lesions of cerebellar lobes IX and X cause the eyes to move involuntarily without head movement. In children with autism, Ornitz et al.155 found that the VOR gain (peak SVC/head acceleration) was normal; however, the time for SVC to fade was prolonged. The significant result appeared to be the result of one outlier. In contrast, Goldberg et al.156 found that tilt suppression of the VOR in high-functioning children with autism did not differ from controls. As tilt suppression of the VOR is thought to involve the nodulus and uvula (lobes X and IX, respectively), the authors suggested that this region of the cerebellum was spared. Overall, the small amount of available data would suggest the VOR is normal in autism.

POSTURAL PHYSIOLOGY Studies of postural functions are another method for providing direct and specific evidence of the physiologic integrity of the vestibular system including the cerebellum. Although posterior fossa circuitry contributes significantly to postural function, contributions from more widespread regions are also important. Kohen-Raz et al.157 conducted a study of 91 mentally retarded 6- to 20-year-olds with autism compared to 166 normal 4- to 11-year-olds, and 18 mentally retarded 7- to 16-year-olds. The individuals with autism and those with mental retardation had significantly lower postural stability than the control participants, performing at the level of preschool children even as adolescents. The autism group, but not the group with mental retardation without autism, showed paradoxically better stability when vision was occluded or somatosensory input restricted by standing on pads. Minshew et al.158 tested the postural stability of 79 nonretarded high-functioning individuals with autism 5- to 52-year-olds compared to 61 typical controls matched for age, IQ, and gender. Individuals with autism were found to have reduced postural stability and delayed development of postural stability. Postural stability was reduced under all conditions but was only clinically significant when somatosensory input was disrupted alone or in combination with other sensory challenges. Postural control did not begin to improve in the individuals with autism until 12 years of age and never achieved adult levels. The abnormalities were not consistent with an

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 405 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

405

impairment in the motor system or abnormalities arising from the reduction in Purkinje cells or cerebellar vermis, nor were they consistent with vestibular dysfunction. The abnormalities in postural control were indicative of deficits in multimodal sensory integration, which is dependent on a widely distributed multineuronal system that typically involves the basal ganglia, supplementary motor cortex, anterior cingulate cortex, and subcortical connections more generally. The paradoxical effect reported in the study of mentally retarded individuals with autism was not replicated. The abnormalities in multimodal sensory integration giving rise to postural instability provide additional evidence of abnormalities outside the diagnostic triad of autism and further evidence of abnormalities in neural connectivity and the dependence of deficits on high demands for integration of information. The abnormality in postural stability also explains in part the odd gait that is so commonly apparent in individuals with autism.

CONCLUSION This chapter has presented some of what we consider to be the latest and most salient research and concepts about the neuropsychology and neurophysiology of autism. Within this framework, two overall models for the cognitive and neural basis for autism were presented, the first being that autism is a disorder of neural connectivity and of information processing that disproportionately affects complex or integrative processing and the second that it is triggered by a disorder of social motivation that leads to a downstream cascade of deficits. The evidence for deficits in memory (as a function of information complexity), motor sequences, and sensory integration as integral parts of the autism syndrome suggests that the neural abnormalities responsible for autism are not restricted to the neural systems involved in social, language, and reasoning abilities but extend more generally. The common denominator is not involvement of a few neural systems but rather the demands for integration of information and, thus, a specialized cytoarchitectural feature of brain organization required for higher-level integration of information. The perceptual disturbance in part-whole processing further suggests an abnormality in a developmental neurobiological process involved in the interrelated development and maturation of local and distant connections of neocortex and neocortical systems. Overdevelopment of local connections results in enhanced perception of details and undergrowth of distant connections results in underdevelopment of higher-level cognitive abilities and impaired perception of the whole. fMRI studies have provided evidence of functional underconnectivity in autism.4 Structural imaging studies and studies of head circumference and brain weight report evidence consistent with overgrowth of brain, certainly cerebral white matter if not also cortical gray matter, in early childhood. The overgrowth of the brain in autism appears to coincide with, if not precede, the onset of symptoms. Hence, increased brain volume is not indicative of improved brain function but rather impaired function. The finding of functional underconnectivity is not therefore incompatible with a brain that is oversized, as the circuitry may fail to make needed connections, make faulty connections, or make highly inefficient connections all of which would result

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 406 Friday, March 10, 2006 2:27 PM

406

Understanding Autism: From Basic Neuroscience to Treatment

in functional underconnectivity. Thus, there is a failure at the cognitive, neurologic, and neural levels in the integration of information and in the development of the underlying integratory circuitry of the brain.

CONCLUDING REMARKS REGARDING SOCIAL DEFICITS Dawson and Webb have proposed an alternative model to account for the core social deficit and for autism. In the social motivation hypothesis, decreased social motivation in young children with autism leads to a failure to attend to social information in the environment.74 This failure to attend to faces and voices during the first few months of development, significantly alters the infant’s experience with the environment. Greenough et al.159 propose that the brain develops and matures based on several types of experiences. Experience-expectant processes reflect the incorporation of information that is “ubiquitous” or common to all members of the species. Typical human rearing conditions offer extensive exposure to social information and interactions. However, if an infant does not find these social interactions rewarding or interesting, then the infant may not actively attend to this information. This failure to attend to socially important stimuli has important implications for the development of cortical specialization and efficiency of information processing. Although this theory has specific implications for the core deficits in autism, it is not incompatible with the theory of information processing deficits proposed by Minshew and colleagues. As Minshew and colleagues have proposed, abnormalities in neural circuitry will impact how social information is encoded and processed; thus, the social impairment results from the abnormalities in circuitry. Decreased social motivation found in children with autism may result from the inability to adequately encode the variability in social information. A second hypothesis is that failure in social motivation affects the child’s ability to establish social contingencies that are present in very early face-to-face interaction, such that these social contingencies are not encoded fully and are not reinforced. This drives attention and exploration, further altering the child’s experience and development.

CLOSING COMMENTS Much has been learned about the cognitive and neural basis of autism in the past 10 years. Indeed, the understanding of this disorder has changed radically as a result of the active research that has taken place. Research in progress will undoubtedly lead to further major advances. Improved understanding of the cognitive and neural basis of behavior from imaging and experimental cognitive paradigms will guide the development of new cognitive intervention programs aimed at promoting the growth of underdeveloped brain circuitry and higher-level skills. An accurate description of the cognitive and perceptual basis of impairments is essential if new interventions are to be theory driven, rather than behavior based. Such interventions will have the greatest chance of providing advances over existing methods. From the long-term perspective, a precise definition of the cognitive and neural basis of autism is critical if links between behavior and genome are to be established opening the way to medical prevention and cures for autism.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 407 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

PERSPECTIVES

FOR

407

FUTURE RESEARCH

The past 10 years has demonstrated that remarkable advances can be made in autism and with these will come the potential for dramatic improvements in treatment. In the face of the growing number of identified individuals with ASD, it is imperative that even more rapid strides be made toward more definitive treatments and, eventually, cures and prevention. Current findings suggest the need to focus on defining the disturbances in neural circuitry and determining ways to stimulate the growth of underdeveloped aspects and avoiding the stimulation of those aspects that become overdeveloped. The first round of functional magnetic resonance studies are just completing the initial definition of the alterations in connectivity. It is all too tempting to skip confirmation and further delineation of these alterations, and focus on treatment. This would be a grave mistake, as sample sizes are ranging from 10 to 18 per group and individual variability has not even been a consideration. Thus, only the coarsest of analyses have been completed. There is much yet to be learned. The investigation of structural connectivity, an entirely different parameter, is far behind, but early results are exciting. It remains to be seen if functional underconnectivity will be accompanied by structurally fewer fibers, excessive but inefficiently functioning fibers, or fibers that are misconnected and thus inefficiently functioning. The third anatomic parameter is volume, and an understanding of brain growth in autism awaits coordination of multicenter efforts to amass very large sample sizes so that the growth of individual structures and the relationships between the growth of structures over a large age and IQ range can be defined. It is likely that some regions are differentially affected in autism. This may also be the case with the functional and structural connections. All of these issues will need to be defined carefully. The second major area of future work is to begin to explore interindividual variability in cognitive and neurologic performance and determine how this relates to variability in behavior. Most studies to date have relied on group differences and have failed to examine individual performance, yet we know there is a tremendous amount of variability in the behavior between individuals. Third, cognitive and neurologic performance also needs to be compared to functional and structural connectivity studies. The fourth need is to begin to search for developmental neurobiological mechanisms responsible for the alterations in brain circuitry disturbed in autism. From an intervention perspective, there have been some findings that suggest that it may be possible to target cortical circuitry with behavioral paradigms to test if the circuitry will respond with improvement in function. Such studies should be carefully designed and be based on data. Above all, it is important that the field be allowed to continue to acquire critical data to define the brain abnormality that will enable the eventual identification of the disturbed developmental neurobiological mechanisms and their genetic control that will lead to cures and prevention.

ACKNOWLEDGMENTS Nancy Minshew wishes to acknowledge support of the National Institute of Child Health and Human Development (NICHD) and the National Institute of Deafness and other Communication Disorders (NIDCD) (U19HD35469), a Collaborative

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 408 Friday, March 10, 2006 2:27 PM

408

Understanding Autism: From Basic Neuroscience to Treatment

Program of Excellence in Autism. Diane Williams wishes to acknowledge the support of NIDCD (K23DC006691) in the writing of this chapter. Geraldine Dawson and Sara Jane Webb wish to acknowledge support by NICHD and NIDCD (U19HD35465), a Collaborative Program of Excellence in Autism, and the National Institute of Mental Health (U54MH066399) STAART Center in the writing of this chapter. We also wish to acknowledge the dedication of families everywhere to the pursuit of new knowledge that will bring about earlier recognition, improved outcome, and better treatment for people with autism.

REFERENCES 1. Minshew, N.J., Goldstein, G., and Siegel, D.J., Neuropsychologic functioning in autism: profile of a complex information processing disorder, J. Int. Neuropsychol. Soc., 3, 303, 1997. 2. Minshew, N.J., Williams, D.L., and Goldstein, G., A further characterization of complex cognitive abilities in high functioning autism [abstract], Int. Meet. Autism Res., 3, 71, 2004. 3. Minshew, N.J., Goldstein, G., and Williams, D.L., Neuropsychologic functioning in children with autism [abstract], Int. Meet. Autism Res., 4, 153, 2005. 4. Just, M.A. et al., Cortical activation and synchronization during sentence comprehension in high-functioning autism: evidence of underconnectivity, Brain, 127, 1811, 2004. 5. Garcia-Villamisar, D. and Della Sala, S., Dual-task performance in adults with autism, Cognit. Neuropsychiatry, 7, 63, 2002. 6. Luna, B. et al., Neocortical system abnormalities in autism: an fMRI study of spatial working memory, Neurology, 59, 834, 2002. 7. Just, M.A. et al., Functional and anatomical cortical underconnectivity in autism: evidence from an executive function task, in review. 8. Kana, R., Keller, T., Williams, D., Minshew, N., and Just, M., Attribution of mental states in high functioning autism: evidence for cortical underconnectivity [abstract], Int. Meet. Autism Res., 4, 117, 2005. 9. Koshino, H., Kana, R., Minshew, N., and Just, M., Brain activation to face stimuli in a working memory task in high functioning autism [abstract], Int. Meet. Autism Res., 4, 118, 2005. 10. Behrmann, M. et al., Configural processing in autism and its relationship to face processing, Neuropsychologia, 44, 110, 2005. 11. Brosnan, M.J. et al., Gestalt processing in autism: failure to process perceptual relationships and the implications for contextual understanding, J. Child Psychol. Psychiatry, 4, 5459, 2004. 12. Happé, F.G.E., Autism: cognitive deficit or cognitive style?, Trends in Cognit. Sci., 3, 216, 1999. 13. Happé, F.G.E., Studying weak central coherence at low levels: children with autism do not succumb to visual illusions, J. Child Psychol. Psychiatry, 37, 873, 1996. 14. Ropar, D. and Mitchell, P., Are individuals with autism and Asperger syndrome susceptible to visual illusions?, J. Child Psychol. Psychiatry, 40, 1283, 1999. 15. Mottron, L. and Belleville, S., A study of perceptual analysis in a high-level autistic subject with exceptional graphics abilities, Brain Cognit., 23, 279, 1993. 16. O’Riordan, M.A., Superior visual search in adults with autism, Autism, 8, 229, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 409 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

409

17. O’Riordan, M.A. et al., Superior visual search in autism, J. Exp. Psychol., 54, 961, 2001. 18. Plaisted, K., O’Riordan, M., and Baron-Cohen, S., Enhanced discrimination of novel, highly similar stimuli by adults with autism during a perceptual learning task, J. Child Psychol. Psychiatry, 39, 765, 1998. 19. Plaisted, K., Sweetenham, J., and Rees, L., Children with autism show local precedence in a divided attention task and global precedence in a selective attention task, J. Child Psychol. Psychiatry, 40, 733, 1999. 20. Jolliffe, T. and Baron-Cohen, S., Are people with autism and Asperger syndrome faster than normal on the Embedded Figures Test?, J. Child Psychol. Psychiatry, 38, 527, 1997. 21. Rinehart, N.J. et al., Atypical interference of local detail on global processing in high-functioning autism and Asperger’s disorder, J. Child Psyc. Psychiatry, 41, 769, 2000. 22. Siegel, D.J., Minshew, N.J., and Goldstein, G., Wechsler IQ profiles in diagnosis of high-functioning autism, J. Autism Dev. Disord., 26, 389, 1996. 23. Brian, J.A. and Bryson, S.E., Disembedding performance and recognition memory in autism/PDD, J. Child Psychol. Psychiatry, 37, 865, 1996. 24. Ozonoff, S., Pennington, B.F., and Rogers, S.J., Executive function deficits in highfunctioning autistic individuals: relationship to theory of mind, J. Child Psychol. Psychiatry, 32, 1081, 1991. 25. Mottron, L. et al., Locally oriented perception with intact global processing among adolescents with high-functioning autism: evidence from multiple paradigms, J. Child Psychol. Psychiatry, 44, 904, 2003. 26. Mottron, L. et al., Perceptual processing among high-functioning persons with autism, J. Child Psychol. Psychiatry, 40, 203, 1999. 27. Ozonoff, S. et al., Executive function abilities in autism and Tourette syndrome: an information processing approach, J. Child Psychol. Psychiatry, 35, 1015, 1994. 28. Koshino, H. et al., Functional connectivity in an fMRI working memory task in highfunctioning autism, Neuroimage, 24, 810, 2005. 29. Prior, M. and Hoffman, W., Brief report: neuropsychological testing of autistic children through an exploration with frontal lobe tests, J. Autism Dev. Disord., 20, 581, 1990. 30. Minshew, N.J. and Goldstein, G., The pattern of intact and impaired memory functions in autism, J. Child Psychol. Psychiatry, 42, 1095, 2001. 31. Ropar, D. and Mitchell, P., Susceptibility to illusions and performance on visuospatial tasks in individuals with autism, J. Child Psychol. Psychiatry, 42, 539, 2001. 32. Mottron, L., Belleville, S., and Menard, E., Local bias in autistic participants as evidenced by graphic tasks: perceptual hierarchization or working memory deficit?, J. Child Psychol. Psychiatry, 40, 743, 1999. 33. Booth, R. et al., Disentangling weak coherence and executive dysfunction: planning drawing in autism and attention-deficit/hyperactivity disorder, Philos. Trans. Royal Soc. B., 358, 387, 2003. 34. Hughes, C., Executive dysfunction in autism: its nature and implications for the everyday problems experienced by individuals with autism, in The Development of Autism, Burack, J.et al., Eds., Lawrence Erlbaum Associates, Mahwah, NJ, 2001, p. 255. 35. Pennington, B.F., The working memory function of the pre-frontal cortices: implications for developmental and individual differences in cognition, in The Development of Future-Oriented Processes, Haith, M.M. et al., Eds., University of Chicago Press, Chicago, IL, 1994, p. 243.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 410 Friday, March 10, 2006 2:27 PM

410

Understanding Autism: From Basic Neuroscience to Treatment

36. Russell, J., How executive disorders can bring about an adequate “theory of mind,” in Autism as an Executive Disorder, Russell, J., Ed., Oxford University Press, Oxford, 1997, p. 256. 37. Stuss, D.T. and Knight, R.T., Introduction, in Principles of Frontal Lobe Function, Stuss, D.T. and Knight R.T., Eds., Oxford University Press, Oxford, 2002. 38. Hughes, C., Russell, J., and Robbins, T.W., Evidence for executive dysfunction in autism, Neuropsychologia, 32, 477, 1994. 39. Bennetto, L., Pennington, B.F., and Rogers, S.J., Intact and impaired memory functions in autism, Child Dev., 67, 1816, 1996. 40. Ozonoff, S. and Jensen, J., Brief report: specific executive function profiles in three neurodevelopmental disorders, J. Autism Dev. Disord., 29, 171, 1999. 41. Ozonoff, S. and McEvoy, R.E., A longitudinal study of executive function and theory of mind development in autism, Dev. Psychopathol., 6, 415, 1994. 42. Brian, J.A. et al., Inhibitory mechanisms in autism spectrum disorders: typical selective inhibition of location versus facilitated perceptual processing, J. Child Psychol. Psychiatry, 44, 552, 2003. 43. Ozonoff, S. and Strayer, D.L., Inhibitory function in nonretarded children with autism, J. Autism Dev. Disord., 27, 59, 1997. 44. Russell, J., Jarrold, C., and Hood, B., Two intact executive capacities in children with autism: implications for the core executive dysfunctions in the disorder, J. Autism Dev. Disord., 29, 103, 1999. 45. Ozonoff, S. and Strayer, D.L., Further evidence of intact working memory in autism, J. Autism Dev. Disord., 31, 257, 2001. 46. Russell, J., Jarrold, C., and Henry, L., Working memory in children with autism and with moderate learning difficulties, J. Child Psychol. Psychiatry, 37, 673, 1996. 47. Williams, D.L., Goldstein, G., Carpenter, P.A., and Minshew, N.J., Verbal and spatial working memory in autism, J. Autism Dev. Disord., 35, 747, 2005. 48. Minshew, N.J., Meyer, J., and Goldstein, G., Abstract reasoning in autism: a dissociation between concept formation and concept identification, Neuropsychology, 16, 327, 2002. 49. Barkley, R.A., Behavioral inhibition, sustained attention, and executive functions: constructing a unifying theory of ADHD, Psychol. Bull., 121, 65, 1997. 50. Clark, C., Prior, M., and Kinsella, G.J., Do executive function deficits differentiate between adolescents with ADHD and oppositional defiant/conduct disorder? A neuropsychological study using the six elements test and hayling sentence completion test, J. Abnorm. Child Psychol., 28, 403, 2000. 51. Schuerholz, L.J. et al., Neuropsychological status of children with Tourette’s syndrome with and without attention deficit hyperactivity disorder, Neurology, 46, 958, 1996. 52. Sergeant, J.A., Geurts, H., and Oosterlaan, J., How specific is a deficit of executive functioning for attention-deficit/hyperactivity disorder?, Behav. Brain Res., 130, 3, 2002. 53. Baron-Cohen, S. et al., Social intelligence in the normal and autistic brain: an fMRI study, Eur. J. Neurosci., 11, 1981, 1999. 54. Liss, M. et al., Executive functioning in high-functioning children with autism, J. Child Psychol. Psychiatry, 42, 261, 2001. 55. Russell, J. and Hill, E.L., Action-monitoring and intention reporting in children with autism, J. Child Psychol. Psychiatry, 42, 317, 2001.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 411 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

411

56. Adams, W.V. and Sheslow, D.V., A developmental perspective of adolescence, in Autism in Adolescents and Adults, Schopler, E. and Mesibov, G., Eds., Plenum Press, New York, 1983, p. 11. 57. Rumsey, J., Conceptual problem-solving in highly verbal, nonretarded autistic men, J. Autism Dev. Disord., 15, 23, 1985. 58. Rutter, M., Cognitive deficits in the pathogenesis of autism, J. Child Psychol. Psychiatry, 24, 513, 1983. 59. Reitan, R.M. and Wolfson, D., The Halstead-Reitan Neuropsychological Test Battery: Theory and Clinical Interpretation, 2nd ed., Neuropsychology Press, Tucson, AZ, 1993. 60. Sloman, S.A., The empirical case for two systems of reasoning, Psychol. Bull., 119, 3, 1996. 61. Johnson-Laird, P.N., Deductive reasoning, Annu. Rev. Psychol., 50, 109, 1999. 62. Bernier, R., Webb, S., and Dawson, G., Understanding impairments in social engagement in autism, in The Development of Social Engagement: Neurobiological Perspectives, Marshall, P. and Fox, N., Eds., Oxford University Press, Oxford, 2005. 63. Dawson, G. et al., Children with autism fail to orient to naturally occurring social stimuli, J. Autism Dev. Disord., 28, 479, 1998. 64. Dawson, G. et al., Neurocognitive function and joint attention ability in young children with autism spectrum disorder versus developmental delay, Child Dev., 73, 345, 2002. 65. Mundy, P. and Neal, R., Neural plasticity, joint attention, and autistic developmental pathology, in International Review of Research in Mental Retardation, Vol. 23, Glidden, L., Ed., Academic Press, New York, 2001, p. 139. 66. Tantam, D., Characterizing the fundamental social handicap in autism, Acta Paedopsychiatr., 55, 83, 1992. 67. Rochat, P. and Striano, T., Social-cognitive development in the first year, in Early Social Cognition: Understanding Others in the First Months of Life, Rochat, P., Ed., Lawrence Erlbaum, Hillsdale, NJ, 1999, p. 3. 68. Dawson, G. et al., Defining the broader phenotype of autism: genetic, brain, and behavioral perspectives, Dev. Psychopathol., 14, 581, 2002. 69. Kuhl, P. et al., Link between social and linguistic processing of speech in preschool children with autism: behavioral and electrophysiological measures, Dev. Sci., 8, 9, 2005. 70. Mundy, P., Sigman, M., and Kasari, C., Joint attention, developmental level, and symptom presentation in autism, Dev. Psychopathol., 6, 389, 1994. 71. Sigman, M. and Ruskin, E., Continuity and change in the social competence of children with autism, down syndrome, and developmental delays, Monogr. Soc. Res. Child., 64, 1, 1999. 72. Courchesne, E. et al., Impairment in shifting attention in autistic and cerebellar patients, Behav. Neurosci., 108, 848, 1994. 73. Meltzoff, A., and Decety, J., What imitation tells us about social cognition: a rapprochement between developmental psychology and cognitive neuroscience, Philos. Trans. Royal Soc. B., 358, 491, 2003. 74. Dawson, G., Webb, S., and McPartland, J., Understanding the nature of face processing impairment in autism: insights from behavioral and electrophysiological studies, Dev. Neuropsychol., 27, 403, 2005. 75. Schultz, R., Developmental deficits in social perception in autism: the role of the amygdala and fusiform face area, Int. J. Dev. Neurosci., 23, 125, 2005.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 412 Friday, March 10, 2006 2:27 PM

412

Understanding Autism: From Basic Neuroscience to Treatment

76. Meltzoff, A. and Gopnick, A., The role of imitation in understanding persons and developing a theory of mind, in Understanding Other Minds, Baron-Cohen, S., Tager-Flusberg, H., and Cohen, D., Eds., Oxford University Press, Oxford, 1993, p. 335. 77. Dawson, G. et al., Neural correlates of face and object recognition in young children with autism spectrum disorder, developmental delay, and typical development, Child Dev., 73, 700, 2002. 78. Grelotti, D., Gauthier, I., and Schultz, R., Social interest and the development of cortical face specialization: what autism teaches us about face processing, Dev. Psychopathol., 40, 213, 2002. 79. Schultz, W., Predictive reward signal of dopamine neurons, J. Neurophysiol., 80, 1, 1998. 80. Baron-Cohen, S. et al., The amygdala theory of autism, Neurosci. Biobehav. Rev., 24, 355, 2000. 81. Insel, T., O’Brien, D., and Leckman, J., Oxytocin, vasopressin and autism: is there a connection?, Biol. Psychiatry, 45, 145, 1999. 82. Sparks, B. et al., Brain structural abnormalities in young children with autism spectrum disorders, Neurology, 59, 184, 2002. 83. Dager, S. et al., Neuroimaging relationships to behavioral performance and clinical course in young children with ASD [abstract], Int. Meet. Autism Res., 2, 29, 2002. 84. Baron-Cohen, S., Mindblindness: An Essay on Autism and Theory of Mind, Bradford/MIT Press, Cambridge, MA, 1995. 85. Brothers, L., Ring, B., and Kling, A., Responses of neurons in the macaque amygdala to complex social stimuli, Behav. Brain Res., 41, 199, 1990. 86. Perrett, D. et al., Social signals analyzed at the cell level: someone is looking at me, something touched me, something moved!, Int. J. Comp. Psychol., 4, 25, 1990. 87. Perrett, D. et al., Organization and functions of cells responsive to faces in the temporal cortex, Philos. Trans. Royal Soc. B., 335, 23, 1992. 88. Williams, J. et al., Imitation, mirror neurons and autism, Neurosci. Biobehav. Rev., 25, 287, 2001. 89. Dawson, G. et al., Case study of the development of an infant with autism from birth to two years of age, J. Appl. Dev. Psychol., 21, 299, 2000. 90. Osterling, J. and Dawson, G., Early recognition of children with autism: a study of first birthday home videotapes, J. Autism Dev. Disord., 24, 247, 1994. 91. Adrien, J. et al., Autism and family home movies: preliminary findings, J. Autism Dev. Disord., 21, 43, 1991. 92. Bentin, S. et al., Electrophysiological studies of face perception in humans, J. Cognit. Neurosci., 8, 551, 1996. 93. McPartland, J. et al., Event-related brain potentials reveal anomalies in temporal processing of faces in autism spectrum disorder, J. Child Psychol. Psychiatry, 45, 1235, 2004. 94. Tantam, D. et al., Autistic children’s ability to interpret faces: a researchers note, J. Child Psychol. Psychiatry, 30, 623, 1989. 95. Boucher, J. and Lewis, V., Unfamiliar face recognition in relatively able autistic children, J. Child Psychol. Psychiatry, 33, 843, 1992. 96. Boucher, J., Lewis, V., and Collis, G., Familiar face and voice matching and recognition in children with autism, J. Child Psychol. Psychiatry, 39, 171, 1998. 97. Gepner, B., de Gelder, B., and de Schonen, S., Face processing in autistics: evidence for a generalized deficit?, Child Neuropsychol., 2, 123, 1996.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 413 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

413

98. Klin, A. et al., A normed study of face recognition in autism and related disorders, J. Autism Dev. Disord., 29, 499, 1999. 99. de Gelder, B., Vroomen, J., and Van der Heide, L., Face recognition and lip-reading in autism, Eur. J. Cognit. Psychol., 3, 69, 1991. 100. Serra, M. et al., Face recognition in children with a pervasive developmental disorder not otherwise specified, J. Autism Dev. Disord., 33, 303, 2003. 101. Blair, R. et al., Fractionation of visual memory: agency detection and its impairment in autism, Neuropsychologia, 40, 108, 2002. 102. Hobson, R., Ouston, J., and Lee, A., What’s in a face? The case of autism, Br. J. Psychol., 79, 441, 1988. 103. Langdell, T., Recognition of faces: an approach to the study of autism, J. Child Psychol. Psychiatry, 19, 255, 1978. 104. Klin, A. et al., Visual fixation patterns during viewing of naturalistic social situations as predictors of social competence in individuals with autism, Arch. Gen. Psychiatry, 59, 809, 2002. 105. Pelphrey, K.A. et al., Visual scanning of faces in autism, J. Autism Dev. Disord., 32, 249, 2002. 106. Trepagnier, C., Sebrechts, M., and Peterson, R., Atypical face gaze in autism, Cyberpsychol. Behav., 5, 213, 2002. 107. van der Geest, J. et al., Gaze behavior of children with pervasive developmental disorder toward human faces: a fixation time study, J. Child Psychol. Psychiatry, 43, 135, 2002. 108. Elgar, K. and Campbell, R., Annotation: the cognitive neuroscience of face recognition: implications for developmental disorders, J. Child Psychol. Psychiatry, 42, 705, 2001. 109. Sigman, M. et al., Responses to the negative emotions of others by autistic, mentally retarded, and normal children, Child Dev., 63, 796, 1992. 110. Toth, K. et al., Defining the early social impairments in autism: social orienting, joint attention, and response to emotions, Presented at the Meeting for the Society for Research in Child Development, Minneapolis, MN, April 2001. 111. Dawson, G. et al., Young children with autism show atypical brain responses to fearful versus neutral facial expressions, Dev. Sci., 7, 340, 2004. 112. Hobson, R., The autistic child’s concept of people, Communication, 20, 12, 1986. 113. Hobson, R., The autistic child’s appraisal of expressions of emotion, J. Child Psychol. Psychiatry, 27, 321, 1986. 114. Celani, G., Battacchi, M., and Arcidiacono, L., The understanding of the emotional meaning of facial expressions in people with autism, J. Autism Dev. Disord., 29, 57, 1999. 115. Gepner, B., de Schonen, S., and Buttin, C., Face processing in young autistic children, Infant Behav. Dev., 17, 661, 1994. 116. Hobson, R., Ouston, J., and Lee, A., Emotion recognition in autism: coordinating faces and voices, Psychol Med., 18, 911, 1988. 117. Loveland, K. et al., Intermodal perception of affect in persons with autism or Down syndrome, Dev. Psychopathol., 7, 409, 1995. 118. Ozonoff, S., Pennington, B., and Rogers, S., Are there emotion perception deficits in young autistic children?, J. Child Psychol. Psychiatry, 31, 343, 1990. 119. Gepner, B., Deruelle, C., and Grynfeltt, S., Motion and emotion: a novel approach to the study of face processing by young autistic children, J. Autism Dev. Disord., 31, 37, 2001.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 414 Friday, March 10, 2006 2:27 PM

414

Understanding Autism: From Basic Neuroscience to Treatment

120. Davies, S. et al., Face perception in children with autism and Asperger’s syndrome, J. Child Psychol. Psychiatry, 35, 1033, 1994. 121. Buitelaar, J. et al., Verbal memory and performance IQ predict theory of mind and emotional recognition ability in children with autistic spectrum disorder and in psychiatric control children, Dev. Psychopathol., 11, 39, 1999. 122. Happé, F., Wechsler I.Q. profile and theory of mind in autism: a research note, J. Child Psychol. Psychiatry, 35, 1461, 1994. 123. Hobson, R., Ouston, J., and Lee, A., Naming emotion in faces and voices: abilities and disabilities in autism and mental retardation, Br. J. Dev. Psychol., 7, 237, 1989. 124. Klinger, L. and Dawson, G., Prototype formation in autism, Dev. Psychopathol., 13, 111, 2001. 125. Strauss, M., Facial expertise: the perception and recognition of faces by infants, preschoolers, and adults with autism, Presented at International Meeting for Autism Research, Sacramento, CA, May, 2004. 126. Brock, J. et al., The temporal binding deficit hypothesis of autism, Dev. Psychopathol., 14, 209, 2002. 127. Webb, S.J., Two faced: dual theories of face processing in autism, Presented at International Meeting for Autism Research, Sacramento, CA May, 2004. 128. Frith, U., Autism: Explaining the Enigma, Oxford, Blackwell Publishers, 1989. 129. Minshew, N., Johnson, C., and Luna, B., The cognitive and neural basis of autism: a disorder complex information processing and dysfunction of neocortical systems, in International Review of Research in Mental Retardation, Vol. 23, Glidden, L., Ed., Academic Press, New York, 2001, p. 111. 130. Grice, S. et al., Disordered visual processing and oscillatory brain activity in autism and Williams Syndrome, Neuroreport, 12, 2697, 2001. 131. Kanwisher, N., Domain- and process-specificity in visual recognition, Presented at Annual Meeting for the Society for Cognitive Neuroscience, New York, April, 2005. 132. Meltzoff, A. and Brooks, R., “Like me” as a building block for understanding other’s minds: bodily acts, attention, and intention, in Intentions and Intentionality: Foundations for Social Cognition, Malle B.F. et al., Eds., MIT Press, Cambridge, MA, 2001, p. 171. 133. Decety, J. et al., Rapid communication: a PET exploration of the neural mechanisms involved in reciprocal imitation, NeuroImage, 15, 265, 2002. 134. Decety, J. and Sommerville, J., Shared representations between self and other: a social cognitive neuroscience view, Trends Cognit. Sci., 7, 527, 2003. 135. Baron-Cohen, S., Autism: deficits in folk psychology exist alongside superiority in folk physics, in Understanding Other Minds: Perspectives from Autism and Developmental Cognitive Neuroscience, Baron Cohen, S., Tager-Flusberg, H., and Cohen, D., Eds., Oxford University Press, New York, 2000, p. 73. 136. Happé, F., Brownwell, H., and Winner, E., Acquired “theory of mind” impairments following stroke, Cognition, 70, 211, 1999. 137. Abell, F., Happé, F., and Frith, U., Do triangles play tricks? Attribution of mental states to animated shapes in normal and abnormal development, Cognit. Dev., 15, 1, 2000. 138. Grant, C.M., Riggs, K.J., and Boucher, J., Counterfactual and mental state reasoning in children with autism, J. Autism Dev. Disord., 34, 177, 2004. 139. Miller, C.A., False belief understanding in children with specific language impairment, J. Commun. Disord., 34, 73, 2001.

© 2006 by Taylor & Francis Group, LLC

2732_C017.fm Page 415 Friday, March 10, 2006 2:27 PM

Neuropsychology and Neurophysiology of Autism Spectrum Disorders

415

140. Happé, F.G., Communicative competence and theory of mind in autism: a test of relevance theory, Cognition, 48, 101, 1993. 141. Gordon, A.C. and Olson, D.R., The relation between acquisition of a theory of mind and the capacity to hold in mind, J. Exp. Child Psychol., 68, 70, 1998. 142. Happé, F. et al., “Theory of mind” in the brain: evidence from a PET scan study of Asperger syndrome, Neuroreport, 8, 197, 1996. 143. Castelli, F. et al., Autism, Asperger syndrome and brain mechanisms for the attribution of mental states to animated shapes, Brain, 125, 1839, 2002. 144. Minshew, N.J. et al., Neurologic aspects of autism, in Autism and Pervasive Developmental Disorders: Vol. 1. Neurological and Medical Issues, Volkmar, F.R. et al., Eds., John Wiley & Sons, New Jersey, 2005, p. 473. 145. Leigh, R.J. and Zee, D.S., The Neurology of Eye Movements, F.A. Davis Company, Philadelphia, PA, 1991. 146. Kemner, C. et al., In search of neurophysiological markers of pervasive developmental disorders: smooth pursuit eye movements?, J. Neural Transm., 111, 1617, 2004. 147. Kemner, C. et al., Abnormal saccadic eye movements in autistic children, J. Autism Dev. Disord., 28, 61, 1998. 148. van der Geest, J.N. et al., Eye movements, visual attention, and autism: a saccadic reaction time study using the gap and overlap paradigm, Biol. Psychiatry, 50, 614, 2001. 149. Takarae Y. et al., Oculomotor abnormalities parallel cerebellar histopathology in autism, J. Neurol. Neurosurg. Psychiatry, 75, 1359, 2004. 150. Thier, P. et al., Encoding of movement time by populations of cerebellar Purkinje cells, Nature, 405, 72, 2000. 151. Minshew, N.J., Luna, B., and Sweeney, J., Oculomotor evidence for neocortical systems but cerebellar dysfunction in autism, Neurology, 52, 917, 1999. 152. Goldberg, M. et al., Deficits in the initiation of eye movements in the absence of a visual target in adolescents with high-functioning autism, Neuropsychologia, 40, 2039, 2002. 153. Koczat, D.L. et al., Eye movement abnormality suggestive of a spatial working memory deficit is present in parents of autistic probands, J. Autism Dev. Disord., 32, 513, 2002. 154. Takarae Y. et al., Pursuit eye movement deficits in autism, Brain, 127, 2584, 2004. 155. Ornitz, E.M. et al., Brain-stem dysfunction in autism: results of vestibular stimulation, Arch. Gen. Psychiatry, 42, 1018, 1985. 156. Goldberg, M.C. et al., Evidence of normal cerebellar control of the vestibulo-ocular reflex (VOR) in children with high-functioning autism, J. Autism Dev. Disord., 30, 519, 2000. 157. Kohen-Raz, R., Volkmar, F.R., and Cohen, D.J., Postural control in children with autism, J. Autism Dev. Disord., 22, 419, 1992. 158. Minshew, N.J. et al., Underdevelopment of the postural control system in autism, Neurology, 63, 2056, 2004. 159. Greenough, W., Black, J., and Wallace, C., Experience and brain development, Child Dev., 58, 539, 1987.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 417 Wednesday, February 15, 2006 4:34 PM

18 Pharmacological Treatments Christopher J. McDougle, David J. Posey, and Kimberly A. Stigler CONTENTS Introduction............................................................................................................417 Motor Hyperactivity and Inattention.....................................................................418 Psychostimulants ............................................................................................418 Alpha2 Adrenergic Agonists ...........................................................................421 Interfering Stereotypical and Repetitive Behavior................................................422 Clomipramine .................................................................................................423 Fluvoxamine ...................................................................................................424 Fluoxetine .......................................................................................................425 Other SSRIs ....................................................................................................425 Aggression and Self-Injurious Behavior ...............................................................426 Haloperidol .....................................................................................................426 Clozapine ........................................................................................................428 Risperidone .....................................................................................................428 Olanzapine ......................................................................................................429 Quetiapine.......................................................................................................430 Ziprasidone .....................................................................................................431 Aripiprazole ....................................................................................................431 Core Social Impairment.........................................................................................432 Drugs Affecting Glutamate Function.............................................................432 D-Cycloserine .................................................................................................433 Summary and Future Directions ...........................................................................434 Coactive Pharmacological Treatment Strategies............................................435 Acknowledgments..................................................................................................436 References..............................................................................................................436

INTRODUCTION The psychopharmacology of autistic disorder (autism) and other pervasive developmental disorders (PDDs) can be reviewed from different perspectives. One approach is to present drug treatments based upon the neurochemical mechanism of action of 417

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 418 Wednesday, February 15, 2006 4:34 PM

418

Understanding Autism: From Basic Neuroscience to Treatment

the agent. Another is to discuss agents from different general classes of drugs, such as psychostimulants, antidepressants, and antipsychotics. For this chapter, we have chosen to review pharmacological treatment based on specific target symptom domains of behavior. We believe that this format is useful for clinicians and is representative of the way they consider pharmacotherapy in the real-world setting. Rather than making recommendations about which particular drug or drugs to use for specific target symptoms, we have chosen to review the available literature so that readers can make their own informed decisions. We will discuss drug treatment strategies directed toward the following target symptom domains: motor hyperactivity and inattention, interfering stereotypical and repetitive behavior, aggression and self-injurious behavior (SIB), and the core social impairment of autism and other PDDs. In this chapter, we will not be discussing all potential symptom areas. For example, mood and sleep disturbances and interfering anxiety are often targets of pharmacotherapy in individuals with PDDs. Controlled studies directed toward these specific symptom areas, however, have not yet been published. Also, drugs that have had putative efficacy in autism, but which have been shown to be largely ineffective, such as megavitamins, fenfluramine, naltrexone, and secretin, will not be discussed in any detail. Recent reports have shown that medications are more frequently being prescribed for individuals with PDDs in the community, often without good empirical evidence. We have attempted to highlight what is known scientifically rather than summarize what may necessarily be occurring in clinical practice. For these reasons, more emphasis is placed on published, randomized, controlled trials with an effort to address efficacy or effectiveness, as well as common adverse effects of the different drugs.

MOTOR HYPERACTIVITY AND INATTENTION Motor hyperactivity and inattention frequently cause significant impairment in children with PDDs, particularly younger children within the school setting. Interestingly, in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Addition (DSM-IV), attention-deficit/hyperactivity disorder (ADHD) is not diagnosed if the symptoms of inattention and hyperactivity occur exclusively during the course of a PDD.1 This is an important point to emphasize that not all clinical presentations of inattention and hyperactivity should be equated with a diagnosis of ADHD. It follows, then, that not all pharmacological treatments that are effective for symptoms of ADHD will necessarily be helpful and well tolerated in individuals with other disorders associated with inattention and hyperactivity, including those with PDDs. Selected published placebo-controlled studies of drugs for motor hyperactivity and inattention in PDDs are shown in Table 18.1.

PSYCHOSTIMULANTS Psychostimulants are first-line agents for the treatment of hyperactivity and inattention in patients diagnosed with ADHD.2 In addition to their efficacy in patients with ADHD, stimulants have been studied in individuals with ADHD and comorbid mental retardation. Although Handen et al.3 cautioned that children with ADHD and

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 419 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

419

TABLE 18.1 Selected Published Placebo-Controlled Studies of Drugs for Motor Hyperactivity and Inattention Subjects Study Quintana et al., 1995 Handen et al., 2000 Jaselskis et al., 1992

Fankhauser et al., 1992

Drug

N

Age

Design

Results

Ref.

Methylphenidate

10

7–11

4-weeks crossover

Methylphenidate > PLA

14

Methylphenidate

13

5–11

3-weeks crossover

15

Clonidine

8

5–13

14-weeks crossover

Clonidine (transdermal)

9

5–33

10-weeks crossover

Methylphenidate > PLA 8/13 (62%) responders Clonidine > PLA (rated by teacher and parent, but not clinician) 6/8 (75%) responders 2/8 (25%) responders at 1-yr follow-up Clonidine > PLA 6/9 (67%) responders

24

25

Note: All ages are in years; all studies involved subjects with autistic disorder, all were double-blind, and placebo-controlled; PLA = placebo; N = number.

mental retardation may be more likely to develop adverse effects, research shows that this population responds positively to stimulants albeit, at times, at lower rates than in patients without mental retardation.4,5 Although less is known about the use of stimulants in patients with PDDs, they are frequently prescribed for hyperactivity and inattention.6,7 Early reviews suggested that stimulants were generally ineffective and associated with adverse effects in patients with PDDs.8,9 Although preliminary research supported this viewpoint, the subjects studied had diverse diagnoses. Campbell et al.,10 in their placebo-controlled comparison study of triiodothyronine and dextroamphetamine (mean dosage 4.8 mg/d, range 1.25 to 10 mg/d) in 16 children (mean age 4.3 years, range 3 to 6 years) with diagnoses of schizophrenia, organic brain syndrome, and autism, found that dextroamphetamine resulted in a clinical worsening in all diagnostic groups. Adverse effects were common and included irritability, hyperactivity, worsened stereotypy, and loss of appetite. The effects of levodopa and levoamphetamine (mean dosage 13.4 mg/d, range 3.5 to 42 mg/d) were compared in a double-blind crossover study conducted in 12 children (mean age 5.4 years, range 3 to 12 years) with schizophrenia and autistic features.11 Levoamphetamine resulted in overall worsening of preexisting symptoms and was poorly tolerated. The most frequent adverse effect reported was an increase or emergence of stereotypies in 9 (81.8%) of 11 children.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 420 Wednesday, February 15, 2006 4:34 PM

420

Understanding Autism: From Basic Neuroscience to Treatment

In contrast, other trials have suggested that stimulants may be effective in this population. Methylphenidate (dosage range 5 to 10 mg/d, 0.3 to 1.0 mg/kg/d) given for 6.5 months (range 2 to 60 weeks) resulted in improvement in 9 (60%) of 15 patients (mean age 7 years, range 2 to 13 years) diagnosed with early infantile autism.12 The medication was found to decrease hyperactivity and impulsivity. Adverse effects included anorexia, irritability, and insomnia. Birmaher et al.,13 in their 2-week, openlabel study of nine children (age range 4 to 16 years) diagnosed with autism, found that methylphenidate (dosage range 10 to 50 mg/d) resulted in significant improvement in the Conners teacher questionnaire. Two subjects received haloperidol (4 and 5 mg/d, respectively) in conjunction with methylphenidate during the trial. A doubleblind, crossover study of 10 autistic children (age range 7 to 11 years) revealed a statistically significant improvement in irritability and hyperactivity.14 Subjects received methylphenidate 10 mg twice daily or placebo for 2 weeks. A second 2-week treatment phase with methylphenidate 20 mg twice a day concluded the study. The authors reported an overall modest benefit of methylphenidate over placebo. Two subjects required the addition of haloperidol at the end of the study because of behavioral problems. Although there were no statistically significant differences in adverse effects between methylphenidate and placebo, decreased appetite, insomnia, and irritability were among those reported. Another double-blind, placebo-controlled crossover study of methylphenidate (0.3 and 0.6 mg/kg/d) found a 50% decrease on the Conners hyperactivity index in 8 of 13 autistic children (age range 5 to 11 years).15 Adverse effects were most commonly reported at the 0.6 mg/kg/d dosage and included social withdrawal and irritability. Our group recently completed a large, retrospective chart review involving all patients with PDDs treated with a stimulant within our clinic.16 Among 195 patients, (174 males, 21 females; mean age ± SD = 7.3 ± 3.5 years, range 2 to 19 years), 61 had more than one trial, resulting in a total of 274 separate stimulant trials. We learned that 24.6%, 23.2%, and 11.1% of patients with a history of one, two, or three stimulant trials, respectively, responded to their first stimulant trial. Among first-trial nonresponders, 6 (14.0%) of 43 patients responded to a second trial. Of those who did not respond to their first or second stimulant trial, 2 (14.3%) of 14 patients responded to a third trial. Patients with Asperger’s disorder, in contrast to those with autism or PDD not otherwise specified (NOS), were significantly more likely to respond to a stimulant. Use of concomitant medication positively affected response, whereas no association was found between stimulant type and IQ and response. Adverse effects, including agitation, dysphoria, and irritability, often occurred (154 [57.5%] of 268 trials, with 6 missing values). Overall, stimulants appeared ineffective and poorly tolerated for most patients in the study. The National Institute of Mental Health (NIMH)-sponsored Research Units on Pediatric Psychopharmacology (RUPP) Autism Network recently completed the largest controlled trial of a psychostimulant in PDDs to date.17 The study involved a 4-week, double-blind, placebo-controlled crossover with three doses of methylphenidate followed by an 8-week, open-label continuation. Seventy-two drug-free children (ages 5 to 14 years) with PDDs accompanied by moderate to severe hyperactivity entered the study. Prior to randomization, subjects entered a one-week test-dose phase in which placebo and three doses (low, medium, and high) of methylphenidate

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 421 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

421

were administered. Subjects who tolerated the test dose (n = 66) were assigned to receive 1 week each of placebo and the three methylphenidate doses in random order during a double-blind crossover phase. Methylphenidate responders then entered an 8-week open-label treatment with the individually determined best dose. Methylphenidate was superior to placebo on the primary outcome measure, the teacherrated hyperactivity subscale of the Aberrant Behavior Checklist (ABC).18 Of the 72 enrolled subjects, 35 (49%) responded to methylphenidate. Adverse effects led to discontinuation of study medication in 13 (18%) of 72 subjects. These results are consistent with the 13-subject study in PDD reported by Di Martino et al.19 They gave a single test dose of methylphenidate (0.4 mg/kg), between our medium and high doses, and found that five subjects could not tolerate it, showing increased hyperactivity, stereotypy, dysphoria, or tics within 1 hour. Of the remaining eight subjects, six improved when treated, constituting an overall response rate of 46%. In contrast, in the NIMH collaborative multisite Multimodal Treatment study of children with ADHD (the MTA), 69% of subjects randomized to methylphenidate were rated as responders, and only 1.4% of subjects discontinued medication because of adverse events.20 Thus, it appears that methylphenidate is less efficacious and associated with more frequent adverse effects in children with PDDs than in typically developing children with ADHD.

ALPHA2 ADRENERGIC AGONISTS Nonstimulant drugs traditionally used to treat ADHD include tricyclic antidepressants (TCAs), bupropion, and alpha2 adrenergic agonists. TCAs and bupropion have not been used frequently in the treatment of children with PDDs, perhaps because of their propensity to lower the seizure threshold or cause anticholinergic-associated cognitive impairment, both of which are common in PDD that is often comorbid with mental retardation. The potential cardiotoxicity of TCAs has also been a concern.21 Clonidine, an alpha2 adrenergic agonist, has been used in the treatment of several neuropsychiatric disorders, including ADHD and Tourette’s disorder.22,23 It has also been demonstrated to be efficacious in two small, controlled studies of subjects with autism. In a double-blind, placebo-controlled crossover study of clonidine in eight children with autism (aged 5.0 to 13.4 years), improvement was seen in hyperactivity, irritability, stereotypies, inappropriate speech, and oppositional behavior.24 Subjects received either clonidine (4 to 10 µg/kg/d, 0.15 to 0.20 mg/d) or placebo for 6 weeks before being crossed over to the other treatment. Symptomatic improvement was found on some, but not all, of the assessment instruments utilized. Statistically significant improvement was recorded on the teacher-rated ABC and on the parent ratings of the Connors scales. No significant drug–placebo difference was identified on clinician ratings. Adverse effects included hypotension, sedation, and irritability. Six of the eight subjects continued open-label treatment following the study. Four of those subjects relapsed, however, within the next few months, with only two continuing on clonidine for as long as 1 year. Transdermal clonidine was reported to be efficacious in a double-blind, placebocontrolled crossover study (4 weeks in each treatment phase) involving nine males

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 422 Wednesday, February 15, 2006 4:34 PM

422

Understanding Autism: From Basic Neuroscience to Treatment

with autism (aged 5 to 33 years).25 Significant improvement was seen on the Clinical Global Impressions (CGI) scale26 and in abnormal sensory responses, affectual reactions, and social relationships to people on the Ritvo-Freeman Real-Life Rating Scale.27 There was also a reduction in hyperactivity and anxiety. The dose of clonidine was approximately 5 µg/kg daily and ranged from 0.1 to 0.3 mg/d. The most common adverse effects were sedation and fatigue. Guanfacine, another alpha2 adrenergic agonist, has a longer half-life than clonidine, which allows for less frequent dosing and lowers the risk of rebound hypertension. In addition, there is some evidence that guanfacine is less sedating, with less pronounced hypotensive effects.28 Because of the potential advantages of guanfacine over clonidine and other agents discussed earlier, we undertook a systematic review of the effectiveness of guanfacine in our autism clinic population.29 Eighty subjects with PDDs (10 females, 70 males) (mean age 7.7 ± 3.5 years, range 3 to 18 years) were treated with guanfacine within our clinic. Charts were reviewed to determine the response of specific target symptoms, including hyperactivity, inattention, and impulsivity. The relationship between treatment response and age, diagnosis, level of cognitive impairment, and symptom severity was determined. Adverse effects were also evaluated. Guanfacine (mean daily dose 2.6 ± 1.7 mg, range 0.25 to 9 mg, mean duration of treatment 334 to 374 days, range 7 to 1776 days) treatment was effective in 19 (23.8%) of 80 subjects. Subjects with PDDNOS (11 of 28 responders, 39.3%) and Asperger’s disorder (2 of 6 responders, 33.3%) showed a greater rate of global response than those with autism (6 of 46 responders, 13.0%). There was a trend for subjects without comorbid mental retardation (9 of 24 subjects, 37.5%) to respond at a greater rate than those with mental retardation (10 of 56 subjects, 17.9%). Symptom improvement was seen in hyperactivity, inattention, insomnia, and tics. Guanfacine was well tolerated, and did not lead to significant changes in blood pressure or heart rate. Randomized controlled studies of guanfacine in autism and other PDDs have not yet been conducted. Atomoxetine is a selective norepinephrine (NE) reuptake inhibitor recently released for the treatment of ADHD. Studies of atomoxetine are currently ongoing in subjects with autism and other PDDs. Of some caution is that an earlier controlled study of the relatively selective NE reuptake inhibitor desipramine in youth with autism found the drug to be associated with increased irritability, although the effects on hyperactivity were beneficial.21

INTERFERING STEREOTYPICAL AND REPETITIVE BEHAVIOR Since Kanner’s30 original description of autism, repetitive thoughts and behavior have been considered integral and core components of the syndrome. In that original report, verbal and motor rituals, obsessive questioning, a rigid adherence to routine, a preoccupation with details, and an anxiously obsessive desire for the maintenance of sameness and completeness were all noted among the 11 autistic children described. Religious and somatic obsessions, repetitive handwashing, and tics were identified in family members.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 423 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

423

In a case-controlled study, we previously showed that the types of repetitive thoughts and behaviors differed between adults with autism and those with obsessivecompulsive disorder (OCD).31 However, repetitive phenomena can interfere with the level of functioning of individuals within either diagnostic group. Based upon the efficacy of serotonin reuptake inhibitors (SRIs) in OCD, the high prevalence of interfering repetitive thoughts and behaviors in patients with PDDs, and evidence indicating that a dysregulation in serotonin (5-hydroxytryptamine, i.e., 5-HT) neurotransmission may contribute to the pathophysiology of some patients with autism,32 our group and others have been studying the clinical response and side effect profile of SRIs in children, adolescents, and adults with PDDs.33,34 To date, two larger-scale placebo-controlled studies, a blinded comparison with placebo and haloperidol, and a small placebo-controlled crossover trial of SRIs in autism have been published, and a fifth has been completed. Selected published placebo-controlled studies of drugs for interfering stereotypical and repetitive behavior in PDDs are shown in Table 18.2.

CLOMIPRAMINE The first published placebo-controlled trial of an SRI in autism investigated clomipramine, a TCA and potent but nonselective inhibitor of 5-HT uptake.21 Following a 2-week single-blind phase, 12 subjects completed a 10-week double-blind,

TABLE 18.2 Selected Published Placebo-Controlled Studies of Drugs for Interfering Stereotypical and Repetitive Behavior Subjects Study Gordon et al., 1993 Remington et al., 2001 McDougle et al., 1996 Buchsbaum et al., 2001 McDougle et al., 2005

Drug

N

Age

Dx

Design

Clomipramine 28

6–23

AUT

10-weeks crossover

Clomipramine 36

10–36 AUT

23-weeks crossover

Fluvoxamine

30

18–53 AUT

Fluoxetine

6

Risperidone

101 5–17

12-weeks parallel groups 30.5 AUT, 16-weeks ± 8.6 ASP crossover AUT

8-weeks parallel groups

Results Clomipramine > PLA Clomipramine > DMI 19/28 (68%) responders Clomipramine = PLA

Ref. 21

38

Fluvoxamine > PLA 8/15 (53%) responders

39

Fluoxetine > PLA 3/6 (50%) responders

41

Risperidone > PLA 34/49 (69%) responders

53

Note: All ages are in years; all studies were double-blind and placebo-controlled; AUT = autistic disorder; ASP = Asperger’s disorder; DMI = desipramine; Dx = diagnosis; PLA = placebo; N = number.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 424 Wednesday, February 15, 2006 4:34 PM

424

Understanding Autism: From Basic Neuroscience to Treatment

crossover comparison of clomipramine and placebo, and 12 different subjects completed a similar comparison of clomipramine and desipramine. Clomipramine (mean dose 152 mg/d) was superior to both placebo and desipramine (mean dose 127 mg/d) on ratings of autistic symptoms (including stereotypies), anger and compulsiveness, ritualized behaviors, with no difference between desipramine and placebo. Clomipramine was as good as desipramine, and both drugs were superior to placebo for reducing motor hyperactivity. One child developed prolongation of the corrected QT interval (0.45 sec) on electrocardiogram (ECG), and another developed severe tachycardia (resting heart rate, 160 to 170 beats per min) during clomipramine treatment. A third child experienced a grand mal seizure. Subsequent open-label studies have generally reported that clomipramine is more effective and better tolerated in adults35 than children36,37 with autism. Clomipramine (mean daily dose 128.4 mg/d, range 100 to 150 mg/d) was also compared with haloperidol (mean daily dose 1.3 mg, range 1 to 1.5 mg) and placebo in a 7-week study involving 36 subjects with autism (mean age 16.3 years, range 10 to 36 years). In those subjects who were able to complete a full therapeutic trial, clomipramine proved comparable to haloperidol in terms of improvement with respect to the baseline. However, significantly fewer individuals receiving clomipramine as opposed to haloperidol were able to complete the trial (37.5% vs. 69.7%, respectively) for reasons related to both side effects and efficacy or behavior problems. In the intent-to-treat sample, only haloperidol proved superior to the baseline on a global measure of autistic symptom severity, as well as specific measures for irritability and hyperactivity. Clomipramine did not seem more effective on a measure of stereotypy, nor was it better tolerated.38 Because of their better side effect profile compared with clomipramine, selective SRIs (SSRIs) have been receiving increased attention as a treatment for interfering symptoms associated with autism.

FLUVOXAMINE As of this writing, the only published larger-scale placebo-controlled study of an SSRI in autism involved fluvoxamine.39 Fluvoxamine (276.7 mg/d) or placebo was given to 30 adults with autism for 12 weeks. Eight (53%) of 15 subjects who received fluvoxamine vs. none given placebo were categorized as responders on the CGI. Fluvoxamine was significantly more effective than placebo for reducing repetitive thoughts and behavior, maladaptive behavior, and aggression. In addition, fluvoxamine reduced inappropriate repetitive language usage. Adverse effects included nausea and sedation, which were transient and minor in severity. In contrast to the encouraging results from this study of fluvoxamine in adults with autism, a 12-week double-blind, placebo-controlled study in children and adolescents with autism and other PDDs found the drug to be poorly tolerated with limited efficacy at best.40 Thirty-four subjects (5 female, 29 male; age range 5 to 18 years, mean age 9.5 years), 12 of whom met the criteria for autism, 8 for Asperger’s disorder, and 14 for PDDNOS, participated. Of the 16 subjects randomized to placebo, none demonstrated any significant change in target symptoms. Two of the placebo-treated subjects showed increased motor hyperactivity, two had insomnia, one had dizziness/vertigo, one had agitation, and one had diarrhea,

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 425 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

425

decreased concentration, and increased self-stimulation. Eighteen of the subjects were randomized to fluvoxamine (25 to 250 mg/d, mean dose 106.9 mg/d). The drug was begun at 25 mg every other day and increased by 25 mg every 3 to 7 d, as tolerated. Only one of the fluvoxamine-treated children demonstrated a significant clinical improvement with the drug. Fourteen of the subjects randomized to fluvoxamine demonstrated adverse effects (insomnia n = 9, motor hyperactivity n = 5, agitation n = 5, aggression n = 5, increased rituals n = 2, anxiety n = 3, anorexia n = 3, increased appetite n = 1, and increased impulsivity n = 1).

FLUOXETINE Results from one small double-blind, placebo-controlled crossover trial of fluoxetine (eight weeks on drugs and placebo) in adults (five with autism, one with Asperger’s disorder; mean age 30.5 ± 8.6 years), completed in the context of a PET neuroimaging study, have been published.41 Five subjects received a dose of 40 mg/d of fluoxetine and one, a dose of 20 mg/d (reduced from 40 mg/d due to a frontal headache). The subjects showed a significant reduction in repetitive behavior and anxiety with fluoxetine, and three out of six subjects were categorized as responders. In contrast to the favorable results obtained with fluoxetine in this small placebocontrolled study involving adults with autism, open-label reports involving younger subjects42 and children43 have described significant amounts of drug-related restlessness, hyperactivity, agitation, insomnia, and aggression, despite meaningful clinical improvement in a number of subjects. In addition to potential behavior-activating side effects, particularly in children and adolescents, fluoxetine has been reported to induce manic symptoms in some patients with PDDs.44

OTHER SSRIS Open-label reports of sertraline and paroxetine in autism have been published. In general, the results from these trials have paralleled those of the studies described earlier in which adults with autism and other PDDs45-47 showed an overall higher rate of response and better tolerability of the drugs than children and adolescents.48,49 A retrospective study of citalopram in 17 children and adolescents (mean age 9.4 ± 2.9 years, range 4 to 15 years) with PDDs found that 59% of subjects were much or very much improved.50 Anxiety and aggression were the most responsive symptoms, whereas social and communication impairment, stereotypies, and preoccupations showed minimal improvement. Citalopram appeared to be better tolerated in the children and adolescents participating in this study compared with the other SRIs discussed earlier; only four subjects experienced treatment-limiting adverse effects. The preceding controlled and open-label studies suggest that SRIs may be less well tolerated and less effective in younger (prepubertal) autistic patients compared to adolescents and adults (postpubertal) with autism. Although this developmental difference in tolerability and response to SRIs may be a dose-related phenomenon, other factors need to be considered. Data indicate that significant changes occur in measures of 5-HT function during puberty in individuals with autism. For example, McBride et al.51 found that mean platelet 5-HT levels were significantly higher in

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 426 Wednesday, February 15, 2006 4:34 PM

426

Understanding Autism: From Basic Neuroscience to Treatment

prepubertal children with autism than in prepubertal normal controls, but no significant difference was found between postpubertal male subjects with autism and normal controls. Furthermore, Chugani et al.52 reported results from a PET neuroimaging study showing that changes in brain 5-HT synthesis capacity that occur normally in developing humans are disrupted in children with autism. Thus, preand postpubertal subjects with autism may have significant differences in brain 5-HT function that influence their ability to tolerate and respond to SRIs. Finally, in the recent large double-blind, placebo-controlled study of the atypical antipsychotic risperidone in children and adolescents with autism by the RUPP Autism Network, the drug was significantly more effective than placebo for reducing interfering stereotypical and repetitive behavior.53

AGGRESSION AND SELF-INJURIOUS BEHAVIOR Aggression toward self, others, or property can be a common presenting problem for patients with autism and other PDDs. The aggression is often impulsive in nature and can result in significant physical injury or destruction of property. The selfinjurious behavior (SIB) may have a repetitive or compulsive quality and may be related to some element of sensory stimulation. At times, the SIB can result in severe damage, including retinal detachment and subdural hematoma. Property destruction can span a range of severity from throwing objects to putting holes in walls and breaking doors and windows. Many times, these aggressive and destructive acts are the result of the individual experiencing frustration with their inability to communicate effectively or due to sensory sensitivities to visual, auditory, or tactile input. Clearly, aggression and SIB can interfere with the educational or vocational environment for the individual and on-site staff, and require effective and often intensive intervention. To date, the most effective drugs for treating aggression and SIB in autism have included typical and atypical antipsychotic agents. Clinically, anticonvulsants and mood stabilizers are sometimes used to treat this target symptom cluster. Controlled studies supporting their use in this population, however, have not been reported. Selected published placebo-controlled studies of drugs for aggression and SIB in PDDs are shown in Table 18.3.

HALOPERIDOL The dopamine (DA) D2 receptor antagonist haloperidol has been extensively studied in controlled trials in children with autism. Campbell et al.54 completed a 12-week study in 40 hospitalized autistic children (32 boys and 8 girls; mean age 4.5 years). Haloperidol (mean optimal dose 1.65 mg/d) was superior to placebo in reducing stereotypies and some elements of social withdrawal. However, 12 children experienced dose-dependent sedation, and 2 had acute dystonic reactions. A number of additional short-term, controlled studies of haloperidol have demonstrated that the drug is more efficacious than placebo for maladaptive behavior.55–57 However, haloperidol was associated with significant side effects, including excessive sedation, acute dystonic reactions, and increased irritability.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 427 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

427

TABLE 18.3 Selected Published Placebo-Controlled Studies of Drugs for Aggression and Self-Injury Subjects Study

Drug

N

Age

Anderson et al., 1984 Anderson et al., 1989 McDougle et al., 1998

Haloperidol

40

2–7

Haloperidol

45

2–7

Risperidone

31

18–43

RUPP Autism Network, 2002

Risperidone

101

5–17

Dx

Design

Results

Ref.

AUT

12-weeks crossover

Haloperidol > PLA

56

AUT

12-weeks crossover

Haloperidol > PLA

57

AUT, PDD

12-weeks parallel groups

69

AUT

8-weeks parallel groups

Risperidone > PLA 8/14 (57%) responders Risperidone > PLA 34/49 (69%) responders

70

Note: All ages are in years; all studies were double-blind and placebo-controlled; Dx = diagnosis; AUT = autistic disorder; PDD = pervasive developmental disorder not otherwise specified; PLA = placebo; RUPP = Research Units on Pediatric Psychopharmacology; N = number.

Because longer-term administration of drugs is often needed in the treatment of moderately to severely affected autistic children, Perry et al.58 studied haloperidol (mean optimal dose 1.23 mg/d) given for six months in 60 children with autism. Twelve children developed haloperidol-related dyskinesias, three during administration and nine upon drug discontinuation. In an attempt to more carefully define the occurrence of drug-related dyskinesias in this subject group, Campbell et al.59 conducted a prospective study of haloperidol (mean daily dose 1.75 mg) in 118 autistic children. The mean duration of treatment was 708.4 d. Forty (33.9%) of the children developed withdrawal dyskinesias, and nine developed tardive dyskinesia. In summary, controlled studies of the DA receptor antagonist haloperidol and other typical antipsychotics have shown this class of drugs to be effective for reducing many of the maladaptive behaviors associated with autism. Because of the high percentage of drug-induced and withdrawal-related dyskinesias associated with haloperidol and other typical antipsychotics, however, safer agents were needed. The newer “atypical” antipsychotics (e.g., clozapine, risperidone, olanzapine, quetiapine, ziprasidone, and aripiprazole), which modulate both 5-HT and DA neurotransmission with purported significantly lower risks of acute extrapyramidal symptoms (EPS) and tardive dyskinesia, have been studied in children, adolescents, and adults with autism and other PDDs, in this regard. In addition, they have been shown to improve the “negative” symptoms of schizophrenia, along with beneficial effects on “positive”

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 428 Wednesday, February 15, 2006 4:34 PM

428

Understanding Autism: From Basic Neuroscience to Treatment

symptoms, such as hallucinations and delusions. The negative symptoms of schizophrenia include blunted affect, emotional and social withdrawal, disinterest in interpersonal relationships, difficulty in abstract thinking, lack of spontaneity and flow of conversation, and stereotyped thinking.60 A number of investigators, among them Fisman and Steele,61 have suggested that the negative symptoms of schizophrenia are comparable to those that characterize the social impairment of autism.

CLOZAPINE Clozapine was the first atypical antipsychotic to be introduced in the U.S.62 The drug’s ability to block 5-HT2A, 5-HT2C, 5-HT3 and DA D1-D4 receptors has been proposed as its mechanism of action. To our knowledge, there have been only three reports describing the use of clozapine in youth with autism. Zuddas et al.63 treated three children who displayed marked hyperactivity, fidgetiness, or aggression for up to 8 months with doses ranging from 200 to 450 mg/d. Two of the three children showed sustained improvement, although the third had a return of symptoms to baseline levels after an initial response. Chen et al.64 reported the case of a 17-year-old male with autism and severe mental retardation who showed a significant reduction in signs of “overt tension,” hyperactivity, and repetitive motions in response to clozapine 275 mg/d during a 15-d hospitalization. In the third report, a 32-year-old man with autism and profound mental retardation showed marked improvement of aggressiveness and social interaction after 2 months of treatment with clozapine 300 mg/d.65 The patient had been refractory to numerous prior medication trials and had been admitted to the hospital frequently for self-inflicted injuries and to various institutions for harming his parents and destroying household items. The patient showed progressive improvement over a 5-year period. The scarcity of reports describing clozapine treatment in autism might reflect concern regarding the risks of agranulocytosis and seizures associated with the drug in children and adolescents. Additionally, the necessary frequent blood draws are not ideal for children in general or for those with autism in particular.

RISPERIDONE Risperidone has high affinities for DA D2-D4, 5-HT1D, 5-HT2A, 5-HT2C, alpha1 adrenergic, and H1 histaminic receptors, with negligible affinities for muscarinic receptors.66 Multiple open-label reports and case series,67,68 as well as double-blind, placebo-controlled trials in adults69 and children and adolescents,70 have described the beneficial effects of risperidone in individuals with autism and other PDDs. Our group completed a prospective 12-week study of risperidone (mean dose 1.8 mg/d) in 18 children and adolescents with autism and other PDDs.71 Based on the CGI, 12 (67%) of 18 patients responded with significant decreases in repetitive behavior, aggression, and impulsivity. The most significant side effect was weight gain; 12 of the 18 patients gained weight (mean 17.8 ± 7.5 lb, range 10 to 35 lb). Results from two double-blind, placebo-controlled studies of risperidone in autism involving adults and children and adolescents, respectively, have been published. In the adult study, 31 subjects (mean age 28.1 ± 7.3 years) with autism (n = 17)

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 429 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

429

or PDDNOS (n = 14) received either risperidone (mean dose 2.9 ± 1.4 mg/d) or placebo for 12 weeks.69 Overall, risperidone was significantly more effective than placebo with 8 (57%) of 14 subjects being categorized as responders on the CGI as against none of 16 in the placebo group. Specifically, risperidone was effective for reducing interfering repetitive behavior, as well as aggression toward self, others, and property. Significant differences between risperidone and placebo were not captured on scales measuring social relationships to people and language. Of the 15 subjects randomized to risperidone, 13 (87%) had at least one adverse effect, although this included only mild, transient sedation in 5 subjects, compared with 5 (31%) of 16 subjects given placebo (agitation in all 5 cases). Importantly, the weight gain often observed with risperidone in children and adolescents with PDDs, as described earlier, did not occur to the same degree in this study of adults. The double-blind, placebo-controlled study of risperidone in children and adolescents with autism was completed by the RUPP Autism Network.70 A total of 101 children (82 boys, 19 girls; mean age 8.8 ± 2.7 years) were randomly assigned to receive 8 weeks of risperidone (n = 49) or placebo (n = 52) with target symptoms including tantrums, aggression, or SIB. Treatment with risperidone for eight weeks (dose range, 0.5 to 3.5 mg/d) resulted in a 56.9% reduction in the Irritability subscale score of the ABC, as compared with a 14.1% decrease in the placebo group. Of the risperidone-treated subjects, 69% (vs. 12% of those given placebo) were categorized as responders. Risperidone was associated with an average weight gain of 5.9 ± 6.4 lb, as compared with 1.8 ± 4.8 lb with placebo. Also, increased appetite, fatigue, drowsiness, and drooling were more common in the risperidone group. A core companion study to the initial 8-week acute risperidone trial by the RUPP Autism Network has also been completed.72 In this study, 63 subjects who responded to 8 weeks of acute treatment continued on open-label risperidone for an additional 16 weeks. During this open-label continuation phase, the mean risperidone dose remained stable, and there was no significant worsening of target symptoms. Only 8% of subjects discontinued the drug because of loss of efficacy, and one because of adverse effects. Subjects gained an average of 11.2 lb of body weight with extended risperidone treatment. Thirty-two subjects who continued to be classified as responders after the 16-week extension were then randomized to continued risperidone vs. gradual substitution with placebo (over the course of 4 weeks). Of 16 subjects gradually switched to placebo, 10 (62.5%) relapsed compared to 2 (12.5%) of 16 subjects who continued on risperidone, a statistically significant difference.

OLANZAPINE Olanzapine has high affinity for DA D1, D2, and D4 receptors, 5-HT2A, 5-HT2C, and 5-HT3 receptors, alpha1 adrenergic and H1 histaminic receptors, and five muscarinic receptor subtypes.73 Case reports,67,68 an open-label case series,74 and a prospective open-label comparison with haloperidol75 have described positive responses to the atypical antipsychotic olanzapine in subjects with autism and other PDDs. In a case series evaluating olanzapine monotherapy in children, adolescents, and adults with autism and other PDDs, Potenza et al.74 reported that six of seven patients who completed a 12-week open-label trial were responders based on the CGI.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 430 Wednesday, February 15, 2006 4:34 PM

430

Understanding Autism: From Basic Neuroscience to Treatment

Significant improvements in overall symptoms of autism, motor restlessness or hyperactivity, social relatedness, affectual reactions, sensory responses, language usage, SIB, aggression, irritability or anger, anxiety, and depression were observed. Significant changes in repetitive behaviors did not occur for the group. The mean dose of olanzapine was 7.8 ± 4.7 mg/d (range 5 to 20 mg/d). The drug was well tolerated, the most significant adverse effects being increased appetite and weight gain in six patients and sedation in three. With respect to weight gain, the mean weight for the group increased from 137.5 ± 55.8 lb at baseline to 155.9 ± 55.1 lb after 12 weeks of treatment. In a study employing parallel groups design, 12 children with autism (mean age 7.8 ± 2.1 years) were randomized to 6 weeks of open-label treatment with olanzapine or haloperidol.75 Mean final dosages were 7.9 ± 2.5 mg/d for olanzapine and 1.4 ± 0.7 mg/d for haloperidol. Five of six subjects in the olanzapine group and three of six in the haloperidol group were rated as responders. Weight gain from the baseline to the end of treatment was significantly higher, and at times quite substantial in the olanzapine group (range 5.9 to 15.8 lb, mean 9.0 ± 3.5 lb) when compared to the haloperidol group (range −5.5 to 8.8 lb, mean 3.2 ± 4.9 lb).

QUETIAPINE Quetiapine binds to several neurotransmitter sites, including DA D1 and D2, 5-HT2A, 5-HT1A, and histamine H1 receptors.76 There are currently four known reports of quetiapine in the treatment of individuals with PDDs. Martin and colleagues77 conducted a 16-week open-label study of quetiapine (dosage range 100 to 350 mg/d) in six children and adolescents (age range 6 to 15 years) with autism. Two of the six subjects were judged responders, as determined by the CGI. The remaining four subjects discontinued treatment prematurely. Three subjects withdrew because of sedation or lack of response, and one subject dropped out after a possible seizure. Increased appetite and weight gain (range 2.0 to 18.0 lb) were also reported. The investigators concluded that quetiapine was poorly tolerated and generally ineffective in this diagnostic group. Hardan et al.78 reported on an open-label study of quetiapine (mean dosage 477 mg/d, range 265 to 689 mg/d) in 14 youth (mean age 12 years, range 7 to 17 years) with diagnoses of both PDD and mental retardation (n = 10) or mental retardation only (n = 4) over a mean duration of 22 weeks (range 10 to 48 weeks). Subjects receiving concomitant medications were included in the study if their dosages were held constant during the trial. In the PDD group, a significant improvement was found in symptoms of hyperactivity and inattention as measured by the Conners parent scale. Six of ten subjects were judged responders as determined on the CGI. Adverse effects were mild and included sedation (n = 2) and sialorrhea (n = 1). Mean weight gain for these ten subjects was 2.2 lb (range –21.1 to 16.1 lb). Among the four subjects in the mental retardation group, three were judged responders based on the preceding criteria. This subgroup of four subjects lost weight during treatment with quetiapine (mean –4.6 lbs, range –20.0 to 4.0 lbs). More recently, a 12-week open-label study of quetiapine (mean dosage 291.7 mg, range 100 to 450 mg/d) in nine adolescent males (mean age 14.6 years, range 10 to

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 431 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

431

17 years) with autism was published.79 Six of nine subjects completed the trial. Two (22%) of nine subjects were judged responders on the CGI. Two subjects discontinued quetiapine because of sedation (n = 1) and agitation or aggression (n = 1), respectively. Overall, adverse effects reported for the group included sedation (n = 7), weight gain (n = 5), agitation (n = 4), and aggression (n = 2). Our group recently reviewed the medical records of all patients with PDDs in our clinic who had received quetiapine for at least 4 weeks and were not being concurrently treated with another antipsychotic or mood stabilizer.80 Twenty patients (16 males, 4 females; mean age 12.1 ± 6.7 yr, range 5 to 28 yr) met the inclusion criteria and received a quetiapine trial (mean dosage 248.7 ± 198.4 mg/d, range 25 to 600 mg/d) over a mean duration of 59.8 ± 55.1 weeks (range 4 to 180 weeks). Eight (40%) of twenty patients were judged responders to quetiapine. Adverse effects occurred in 50% of patients and led to drug discontinuation in 15% of cases.

ZIPRASIDONE Ziprasidone is a potent 5-HT2A and DA D2 antagonist, and a 5-HT1A agonist that also inhibits 5-HT and NE reuptake sites. In addition, ziprasidone is a potent antagonist at the 5-HT2C and 5-HT1D receptors.81 Our group published a preliminary report on the safety and effectiveness of ziprasidone in children, adolescents, and young adults with autism.82 Twelve patients (mean age 11.6 ± 4.4 years, range 8 to 20 years) with autism (n = 9) or PDDNOS (n = 3) received open-label treatment with ziprasidone (mean daily dose 59.2 ± 34.8 mg, range 20 to 120 mg) for at least 6 weeks (mean duration 14.2 ± 8.3 weeks, range 6 to 30 weeks). Six (50%) of the twelve patients were responders based on the CGI. Transient sedation was the most common side effect. No cardiovascular side effects, including chest pain, tachycardia, palpitations, dizziness, or syncope, were observed or reported. The mean change in body weight for the group was –5.8 ± 12.5 lb (range −35 to 6 lb). Five patients lost weight, five had no change, one gained weight, and one had no follow-up weight beyond the baseline measure. Ziprasidone appeared to have the potential for improving symptoms of aggression, agitation, and irritability in this population without significant weight gain in this short-term trial. The U.S. Food and Drug Administration (FDA) has raised some concerns about the potential for QTc interval prolongation with ziprasidone on the ECG,83 hence the drug should not be given without careful monitoring to individuals with cardiac disease or to those who are taking other medications that can prolong the QTc interval.

ARIPIPRAZOLE Aripiprazole is a partial DA D2 and 5-HT1A agonist and a 5-HT2A antagonist.84 Our group recently completed a pilot study of aripiprazole in youth with autism.85 This prospective, open-label case series is the first known investigation of the effectiveness and tolerability of aripiprazole in children and adolescents with autism. In this study, subjects (mean age 12.2 years, range 5 to 18 years) received the drug (mean dosage 12.0 mg/d, range 10 to 15 mg/d) for a minimum of 8 weeks (mean duration 12.8 weeks, range 8 to 16 weeks). Of the five subjects, all were deemed responders

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 432 Wednesday, February 15, 2006 4:34 PM

432

Understanding Autism: From Basic Neuroscience to Treatment

on the CGI. Significant improvement was noted in a variety of interfering behavioral symptoms, including aggression, SIB, and irritability. Aripiprazole was well tolerated. No acute EPS, or changes in heart rate or blood pressure were recorded. Two of the five subjects experienced mild transient somnolence. Two subjects lost weight, two had no change, and one subject gained 1 lb (mean change −8.2 lbs, range −30 to 1lb). We believe this relative weight loss was secondary to discontinuation of prior atypical antipsychotic treatments that had led to significant weight gain.

CORE SOCIAL IMPAIRMENT The impairment in reciprocal social interaction in autism is severe and persistent. There may be significant deficits in the use of multiple nonverbal behaviors, such as eye-to-eye gaze, facial expression, body posture and gestures, to regulate social interaction and communication. There may be failure to develop peer relationships appropriate to developmental level that may take different forms at different ages: younger individuals may have little or no interest in establishing friendships; older individuals may have an interest in friendship but lack understanding of the conventions of social interaction. There may be a lack of spontaneous seeking to share enjoyment, interests, or achievements with other people. Lack of social or emotional reciprocity may be present. Often an individual’s awareness of others is markedly impaired. Individuals with autism may be oblivious to other children, may have no concept of the needs of others, or may not notice another person’s distress.1 Published placebo-controlled studies of drugs for the core social impairment of PDDs are shown in Table 18.4. A number of drug treatment studies have attempted to impact the core social impairment of autism. Despite encouraging preliminary reports on the prosocial effects of fenfluramine (an indirect 5-HT partial agonist)86,87 and naltrexone (an opioid receptor antagonist)88 subsequent placebo-controlled studies of these drugs89–91 have failed to show consistent improvement in the social impairment of autism. Openlabel reports of SRIs43 and atypical antipsychotics71 have suggested that a few persons with autism show improvement in some aspects of social relatedness following treatment with these drugs. However, this has yet to be shown consistently in placebo-controlled trials with these agents.39,69,70

DRUGS AFFECTING GLUTAMATE FUNCTION Recently, a role for glutamatergic dysfunction in the pathophysiology and treatment of autism has been proposed.92,93 Glutamate, the primary excitatory amino acid neurotransmitter in the brain, is thought to be crucial in neuronal plasticity and higher cognitive functioning.94 A few reports have appeared that described the treatment of autistic individuals with drugs that affect the glutamate neurotransmitter system. Lamotrigine, an anticonvulsant that attenuates glutamate release, resulted in improvement in “autistic symptoms” in 8 of 13 autistic children and adolescents during a study for intractable epilepsy.95 Another report described improvement in SIB, irritability, and disturbed sleep in an 18-year-old female with profound mental retardation and a generalized seizure disorder given lamotrigine.96 Interestingly, this

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 433 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

433

TABLE 18.4 Published Placebo-Controlled Studies of Drugs for the Core Social Impairment Subjects Study

Drug

N

Age

Design 12-weeks double-blind parallel groups 4-weeks double-blind parallel groups

Belsito et al., 2001 King et al., 2001

Lamotrigine

28

3–11

Amantadine

39

5–19

Posey et al., 2004

D-cycloserine

10

5–27

8-weeks single-blind ascending dose

Results Lamotrigine = PLA

Amantadine = PLA (for primary outcome) Amantadine > PLA (on clinician-rated hyperactivity and inappropriate speech) 9/19 (47%) responders D-cycloserine > PLA 4/10 (40%) responders

Ref. 97

102

112

Note: All ages are in years; all studies involved subjects with autistic disorder and were placebocontrolled; PLA = placebo; N = number.

patient also showed improvement in withdrawal and emotional responsivity. Belsito et al.97 conducted a double-blind, placebo-controlled trial of lamotrigine in 28 children with autism. In this study, lamotrigine (mean dose 5.0 mg/kg/d) was no better than placebo on any of the outcome measures, including the autism behavior checklist,98 the ABC, the autism diagnostic observation schedule,99 or the childhood autism rating scale.100 Amantadine (an antagonist at the N-methyl-D-aspartate [NMDA] subtype of glutamate receptor) has also been studied in autism. In an open-label case series of eight children with developmental disabilities, a marked response was seen in four children treated with amantadine (dose range 3.7 to 8.2 mg/kg/d).101 In a subsequent double-blind, placebo-controlled study, 39 subjects with autism (ages 5 to 19 years) were given amantadine (5.0 mg/kg/d) or placebo.102 Clinician ratings of hyperactivity and inappropriate speech showed statistically significant improvement, and there was a trend towards greater response in the amantadine group, based on ratings on the CGI. There was no statistical difference between amantadine and placebo on parent ratings; amantadine was well tolerated.

D-CYCLOSERINE D-Cycloserine is an FDA-approved antibiotic used for the treatment of tuberculosis in children and adults that has been known to have neuropsychiatric effects for decades.103 It acts as a partial agonist at the NMDA subtype of glutamate receptor.104 © 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 434 Wednesday, February 15, 2006 4:34 PM

434

Understanding Autism: From Basic Neuroscience to Treatment

It is thought to act primarily on the glycine-B recognition site on subunit 1 of the NMDA glutamate receptor subtype. D-Cycloserine has been safely administered to a large number of children with primary tuberculosis. At doses as high as 20 mg/kg/d, D-cycloserine was not associated with side effects, even after several months of treatment in children (ages 12 months to 8 years).105 Several studies in adults with schizophrenia have found that D-cycloserine is beneficial for the treatment of the negative symptoms when administered alone106 or when added to typical antipsychotics107,108 or the atypical antipsychotics risperidone and olanzapine109,110 but not clozapine.111 As mentioned earlier, the negative symptoms of schizophrenia may provide an analogy for the treatment of the social impairment seen in autism. Our group recently published results of a single-blind study of D-cycloserine directed toward the core social impairment of subjects with autism.112 Following a 2-week, single-blind placebo lead-in phase, drug-free subjects with autism were given three ascending doses (30 mg/d, 50 mg/d, and 85 mg/d) of D-cycloserine during each of three 2-week periods. Two subjects withdrew from the study after completing only the 2-week placebo lead-in phase. The remaining ten subjects (eight male, two female) (mean age 10.0 ± 7.7 years, range 5.1 to 27.6 years) completed all 8 weeks of the study. Response rates on the global CGI for the placebo, low, medium, and high dose phases of D-cycloserine were 0%, 30%, 40%, and 40%, respectively. A statistically significant improvement was seen on the ABC Social Withdrawal subscale. Two subjects experienced adverse effects (a transient motor tic and increased echolalia) at the highest dose they received. A large double-blind, placebo-controlled study of D-cycloserine monotherapy is currently ongoing.

SUMMARY AND FUTURE DIRECTIONS To date, drug treatment studies in autism have focused on the use of one drug to target one symptom domain or a group of associated symptoms. As reviewed in this chapter, some studies have found the psychostimulant methylphenidate and the alpha2 adrenergic agonist clonidine efficacious for the motor hyperactivity and inattention of autism. Other studies have demonstrated the efficacy of typical (haloperidol) and atypical (risperidone) antipsychotics for the treatment of aggression and SIB. Similarly, the interfering stereotypical and repetitive symptom domain has been shown to be somewhat responsive to SRIs, particularly in postpubertal subjects. Recent preliminary research into drugs affecting the core symptom domain of social impairment, such as D-cycloserine, have been encouraging. Importantly, however, coactive pharmacological treatment strategies utilizing more than one drug to target more than one symptom domain have not been studied in autism. Considering the clinical complexity of the syndrome and the well-characterized symptom domains, this is somewhat surprising. It is unrealistic to expect that a single agent will result in improvement in motor hyperactivity and inattention, reduce the interfering stereotypical and repetitive behavior, decrease the associated aggression and SIB, and improve the social impairment of autism.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 435 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

435

COACTIVE PHARMACOLOGICAL TREATMENT STRATEGIES Coactive pharmacological treatment strategies have proven to be effective for a number of treatment-refractory psychiatric disorders, as well as complex neuropsychiatric syndromes consisting of different target symptom domains.113 One strategy, pharmacological “augmentation,” involves the addition of one or more drugs to the ongoing therapy in order to enhance the specific pharmacodynamic and therapeutic effects of the current drug. An example of this strategy is “lithium augmentation” for treatment-refractory depression. Here, lithium is added to an antidepressant, such as an SRI, after the initial agent has been administered for an adequate amount of time and at an adequate dosage, with only a partial or no clinical response. In this context, lithium facilitates the presynaptic release of 5-HT, which, combined with 5-HT reuptake inhibition by the SRI, results in increased 5-HT in the synaptic cleft. This approach results in the conversion of approximately 50% of the partial responders or nonresponders in a primary antidepressant trial to responder status within 2 to 4 weeks of adding lithium. A second coactive treatment strategy is pharmacological “combination.” This approach involves the simultaneous use of two or more different drugs (one of which may be the current drug), each with a unique and different mechanism of action. Here, two drugs that affect different neurochemical systems are combined in order to target different symptom complexes within a clinical syndrome. Examples of this strategy include the pharmacotherapy of psychotic depression,114 the pharmacotherapy of SRI-refractory OCD with comorbid chronic tic disorder,115,116 and the pharmacotherapy of “positive” and “negative” symptoms of schizophrenia.107 A standard treatment for psychotic depression involves combining an antidepressant that enhances 5-HT and/or NE function with an antipsychotic that antagonizes DA D2 receptors. Typically, neither drug alone would be sufficient to improve both the depressive and psychotic symptoms of the syndrome. For refractory OCD, adding typical or atypical antipsychotics with prominent DA D2 antagonism to ongoing SRI therapy has proved effective for further reducing obsessive-compulsive symptoms and tics in patients with this diagnostic comorbidity. More recently, the combination treatment strategy has been studied in schizophrenia. Typical and atypical antipsychotics are effective agents for reducing the positive symptoms of schizophrenia. Although these drugs also have some effect on negative symptoms, the degree of improvement is often inadequate. In an attempt to more fully address the negative symptoms of schizophrenia, investigators have been adding drugs that modulate glutamate function, including glycine, D-serine and D-cycloserine117 to antipsychotics. This combination treatment strategy has resulted in a more complete reduction in the two primary symptom domains of the syndrome of schizophrenia. We are currently conducting a coactive pharmacological treatment study in autism. As addressed in detail earlier in this chapter, studies of typical and atypical antipsychotics have demonstrated significant improvement in aggression, SIB, and irritability in autism. However, the core symptom domain of social impairment has remained largely unaffected with these agents. Our pilot data with the novel antipsychotic aripiprazole suggests that a similar pattern of symptom response may be anticipated. Children and adolescents with autism who show a significant reduction

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 436 Wednesday, February 15, 2006 4:34 PM

436

Understanding Autism: From Basic Neuroscience to Treatment

in aggression, SIB, and irritability following 6 months of aripiprazole monotherapy without significant improvement in social relatedness will subsequently undergo an 8-week open-label trial of D-cycloserine addition targeted toward the social impairment symptom domain. We hypothesize that these children will show significant reductions in aggressive behavior and improved social relatedness when treated with the combination of aripiprazole and D-cycloserine. We are optimistic that this combination treatment approach will contribute to ongoing advances in the rational pharmacotherapy of autism as we continue to look for safer and more effective treatments for our patients.

ACKNOWLEDGMENTS This work was supported in part by a Daniel X. Freedman Psychiatric Research Fellowship Award (Posey), a National Alliance for Research in Schizophrenia and Depression (NARSAD) Young Investigator Award (Posey), a Research Unit on Pediatric Psychopharmacology — Psychosocial Intervention Grant (U10-MH6676602) from the National Institute of Mental Health (NIMH) to Indiana University (McDougle, Stigler, and Posey), a Research Career Development Award (K23MH068627-01) from the NIMH (Posey), a National Institutes of Health Clinical Research Center Grant to Indiana University (M01-RR00750), and a Department of Housing and Urban Development (HUD) Grant No. B-01-SP-IN-0200 (McDougle).

REFERENCES 1. American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV), 1994. 2. Greenhill, L. et al., Guidelines and algorithms for the use of methylphenidate in children with attention-deficit/hyperactivity disorder, J. Atten. Disord., 6, S89, 2002. 3. Handen, B.L. et al., Adverse side effects of methylphenidate among mentally retarded children with ADHD, J. Am. Acad. Child Adolesc. Psychiatry, 30, 241, 1991. 4. Handen, B.L. et al., Efficacy of methylphenidate among mentally retarded children with attention deficit hyperactivity disorder, Pediatrics, 86, 922, 1990. 5. Aman, M.G. et al., Methylphenidate and thioridazine in the treatment of intellectually subaverage children: effects on cognitive-motor performance, J. Am. Acad. Child Adolesc. Psychiatry, 30, 816, 1991. 6. Aman, M.G. et al., Psychotropic and anticonvulsant drugs in subjects with autism: prevalence and patterns of use, J. Am. Acad. Child Adolesc. Psychiatry, 34, 1672, 1995. 7. Martin, A. et al., Higher-functioning pervasive developmental disorders: rates and patterns of psychotropic drug use, J. Am. Acad. Child Adolesc. Psychiatry, 38, 923, 1999. 8. Campbell, M., Pharmacotherapy in early infantile autism, Biol. Psychiatry, 10, 399, 1975. 9. Aman, M.G., Stimulant drug effects in developmental disorders and hyperactivitytoward a resolution of disparate findings, J. Autism Dev. Disord., 12, 385, 1982. 10. Campbell, M. et al., Response to triiodothyronine and dextroamphetamine: a study of preschool schizophrenic children, J. Autism Child Schizophr., 2, 343, 1972.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 437 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

437

11. Campbell, M. et al., Levodopa and levoamphetamine: a crossover study in young schizophrenic children, Curr. Ther. Res. Clin. Exp., 19, 70, 1976. 12. Hoshino, Y. et al., The effects of methylphenidate on early infantile autism and its relation to serum serotonin levels, Folia Psychiatr. Neurol. Jpn., 31, 605, 1977. 13. Birmaher, B., Quintana, H., and Greenhill, L.L., Methylphenidate treatment of hyperactive autistic children, J. Am. Acad. Child Adolesc. Psychiatry, 27, 248, 1988. 14. Quintana, H. et al., Use of methylphenidate in the treatment of children with autistic disorder, J. Autism Dev. Disord., 25, 283, 1995. 15. Handen, B.L., Johnson, C.R., and Lubetsky, M., Efficacy of methylphenidate among children with autism and symptoms of attention-deficit hyperactivity disorder, J. Autism Dev. Disord., 30, 245, 2000. 16. Stigler, K.A. et al., A naturalistic retrospective analysis of psychostimulants in pervasive developmental disorders, J. Child Adolesc. Psychopharmacol., 14, 49, 2004. 17. Research Units on Pediatric Psychopharmacology Autism Network, randomized, controlled, crossover trial of methylphenidate in pervasive developmental disorders with hyperactivity, Arch. Gen. Psychiatry, 62, 1266, 2005. 18. Aman, M.G. et al., The aberrant behavior checklist: a behavior rating scale for the assessment of treatment effects, Am. J. Ment. Defic., 89, 485, 1985. 19. Di Martino, A. et al., Methylphenidate for pervasive developmental disorders: safety and efficacy of acute single dose test and ongoing therapy: an open-pilot study, J. Child Adolesc. Psychopharmacol., 14, 207, 2004. 20. Greenhill, L.L. et al., Practice parameter for the use of stimulant medications in the treatment of children, adolescents, and adults, J. Am. Acad. Child Adolesc. Psychiatry, 41, 26S, 2001. 21. Gordon, C.T. et al., A double-blind comparison of clomipramine, desipramine, and placebo in the treatment of autistic disorder, Arch. Gen. Psychiatry, 50, 441, 1993. 22. Hunt, R.D., Minderaa, R.B., and Cohen, D.J., Clonidine benefits children with attention deficit disorder and hyperactivity: report of a double-blind placebo-crossover therapeutic trial, J. Am. Acad. Child Adolesc. Psychiatry, 24, 617, 1985. 23. Leckman, J.F. et al., Clonidine treatment of Gilles de la Tourette’s syndrome, Arch. Gen. Psychiatry, 48, 324, 1991. 24. Jaselskis, C.A. et al., Clonidine treatment of hyperactive and impulsive children with autistic disorder, J. Clin. Psychopharmacol., 12, 322, 1992. 25. Fankhauser, M.P. et al., A double-blind, placebo-controlled study of the efficacy of transdermal clonidine in autism, J. Clin. Psychiatry, 53, 77, 1992. 26. Guy, W., ECDEU Assessment Manual for Psychopharmacology, U.S. Department of Health, Education, and Welfare, Publication 76, Washington, D.C., 1976. 27. Freeman, B.J. et al., A scale for rating symptoms of patients with the syndrome of autism in real life settings, J. Am. Acad. Child Psychiatry, 25, 130, 1986. 28. Arnsten, A.F.T., Cai, J.X., and Goldman-Rakic, P.S., The alpha-2 adrenergic agonist guanfacine improves memory in aged monkeys without sedative or hypotensive side effects: evidence for alpha-2 receptor subtypes, J. Neurosci., 8, 4287, 1988. 29. Posey, D.J. et al., Guanfacine treatment of hyperactivity and inattention in pervasive developmental disorders: a retrospective analysis of 80 cases, J. Child Adolesc. Psychopharmacol., 14, 233, 2004. 30. Kanner, L., Autistic disturbances of affective contact, Nerv. Child, 2, 217, 1943. 31. McDougle, C.J. et al., A case-controlled study of repetitive thoughts and behavior in adults with autistic disorder and obsessive-compulsive disorder, Am. J. Psychiatry, 152, 772, 1995.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 438 Wednesday, February 15, 2006 4:34 PM

438

Understanding Autism: From Basic Neuroscience to Treatment

32. McDougle, C.J., Posey, D.J., and Potenza, M.N., Neurobiology of serotonin function in autism, in Autism Spectrum Disorders, Hollander, E., Ed., Marcel Dekker, New York, 2003, chap. 10. 33. McDougle, C.J., Repetitive thoughts and behavior in pervasive developmental disorders: phenomenology and pharmacotherapy, in Asperger Syndrome or High Functioning Autism?, Schopler, E., Mesibov, G.B., and Kunce, L.J., Eds., Current Issues in Autism Series, Plenum, New York, 1998, chap. 13. 34. McDougle, C.J., Kresch, L.E., and Posey, D.J., Repetitive thoughts and behavior in pervasive developmental disorders: treatment with serotonin reuptake inhibitors, J. Autism Dev. Disord., 30, 427, 2000. 35. Brodkin, E.S. et al., Clomipramine in adults with pervasive developmental disorders: a prospective open-label investigation, J. Child Adolesc. Psychopharmacol., 7, 109, 1997. 36. Sanchez, L.E. et al., A pilot study of clomipramine in young autistic children, J. Am. Acad. Child Adolesc. Psychiatry, 35, 537, 1996. 37. Brasic, J.R. et al., Behavioral effects of clomipramine on prepubertal boys with autistic disorder and severe mental retardation, CNS Spectrums, 3, 39, 1998. 38. Remington, G. et al., Clomipramine versus haloperidol in the treatment of autistic disorder: a double-blind, placebo-controlled, crossover study, J. Clin. Psychopharmacol., 21, 440, 2001. 39. McDougle, C.J. et al., A double-blind, placebo-controlled study of fluvoxamine in adults with autistic disorder, Arch. Gen. Psychiatry, 53, 1001, 1996. 40. McDougle, C.J. et al., unpublished data, 2004. 41. Buchsbaum, M.S. et al., Effect of fluoxetine on regional cerebral metabolism in autistic spectrum disorders: a pilot study, Int. J. Neuropsychopharmacol., 4, 119, 2001. 42. Cook, E.H. et al., Fluoxetine treatment of children and adults with autistic disorder and mental retardation, J. Am. Acad. Child Adolesc. Psychiatry, 31, 739, 1992. 43. DeLong, G.R., Teague, L.A., and Kamran, M.M., Effects of fluoxetine treatment in young children with idiopathic autism, Dev. Med. Child Neurol., 40, 551, 1998. 44. Damore, J., Stine, J., and Brody, L., Medication-induced hypomania in Asperger’s disorder, J. Am. Acad. Child Adolesc. Psychiatry, 37, 248, 1998. 45. Hellings, J.A. et al., Sertraline response in adults with mental retardation and autistic disorder, J. Clin. Psychiatry, 57, 333, 1996. 46. McDougle, C.J. et al., Sertraline in adults with pervasive developmental disorders: a prospective open-label investigation, J. Clin. Psychopharmacol., 18, 62, 1998. 47. Branford, D., Bhaumik, S., and Naik, B., Selective serotonin re-uptake inhibitors for the treatment of perseverative and maladaptive behaviors of people with intellectual disability, J. Intellect. Disabil. Res., 42, 301, 1998. 48. Steingard, R.J. et al., Sertraline treatment of transition-associated anxiety and agitation in children with autistic disorder, J. Child Adolesc. Psychopharmacol., 7, 9, 1997. 49. Posey, D.J. et al., Paroxetine in autism, J. Am. Acad. Child Adolesc. Psychiatry, 38, 111, 1999. 50. Courturier, J.L. and Nicolson, R., A retrospective assessment of citalopram in children and adolescents with pervasive developmental disorders, J. Child Adolesc. Psychopharmacol., 12, 243, 2002. 51. McBride, P.A. et al., Effects of diagnosis, race, and puberty on platelet serotonin levels in autism and mental retardation, J. Am. Acad. Child Adolesc. Psychiatry, 37, 767, 1998.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 439 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

439

52. Chugani, D.C. et al., Developmental changes in brain serotonin synthesis capacity in autistic and nonautistic children, Ann. Neurol., 45, 287, 1999. 53. McDougle, C.J. et al., Risperidone for the core symptom domains of autism: results from the RUPP Autism Network Study, Am. J. Psychiatry, 162, 1142, 2005. 54. Campbell, M. et al., A comparison of haloperidol and behavior therapy and their interaction in autistic children, J. Am. Acad. Child Psychiatry, 17, 640, 1978. 55. Cohen, I.L. et al., Behavioral effects of haloperidol in young autistic children: an objective analysis using a within-subjects reversal design, J. Am. Acad. Child Psychiatry, 19, 665, 1980. 56. Anderson, L.T. et al., Haloperidol in the treatment of infantile autism: effects on learning and behavioral symptoms, Am. J. Psychiatry, 41, 1195, 1984. 57. Anderson, L.T. et al., The effects of haloperidol on discrimination learning and behavioral symptoms in autistic children, J. Autism Dev. Disord., 19, 227, 1989. 58. Perry, R. et al., Long-term efficacy of haloperidol in autistic children: continuous versus discontinuous drug administration, J. Am. Acad. Child Adolesc. Psychiatry, 28, 87, 1989. 59. Campbell, M. et al., Neuroleptic-related dyskinesias in autistic children: a prospective, longitudinal study, J. Am. Acad. Child Adolesc. Psychiatry, 36, 835, 1997. 60. Kay, S.R., Opler, L.A., and Lindenmayer, J.P., Reliability and validity of the positive and negative symptom scale for schizophrenia, Psychiatry Res., 23, 99, 1987. 61. Fisman, S. and Steele, M., Use of risperidone in pervasive developmental disorders: a case series, J. Child Adolesc. Psychopharmacol., 6, 177, 1996. 62. Baldessarini, R.J. and Frankenburg, F.R., Clozapine: a novel antipsychotic agent, N. Engl. J. Med., 14, 746, 1991. 63. Zuddas, A. et al., Clinical effects of clozapine on autistic disorder, Am. J. Psychiatry, 153, 738, 1996. 64. Chen, N.C. et al., Clozapine in the treatment of aggression in an adolescent with autistic disorder, J. Clin. Psychiatry, 62, 479, 2001. 65. Gobbi, G. and Pulvirenti, L., Long-term treatment with clozapine in an adult with autistic disorder accompanied by aggressive behavior, J. Psychiatry Neurosci., 26, 340, 2001. 66. Leysen, J.E. et al., Biochemical profile of risperidone, a new antipsychotic, J. Pharmacol. Exp. Ther., 247, 661, 1988. 67. McDougle, C.J. et al., Research units on pediatric psychopharmacology (RUPP) autism network: background and rationale for an initial controlled study of risperidone, in Child and Adolescent Psychiatric Clinics of North America, Martin, A., Scahill, L., Eds., WB Saunders, Philadelphia, PA, 2000, chap. 9. 68. Posey, D.J. and McDougle, C.J., Use of atypical antipsychotics in autism, in Autism Spectrum Disorders, Hollander, E., Ed., Marcel Dekker, New York, 2003, chap. 13. 69. McDougle, C.J. et al., A double-blind, placebo-controlled study of risperidone in adults with autistic disorder and other pervasive developmental disorders, Arch. Gen. Psychiatry, 55, 633, 1998. 70. Research Units on Pediatric Psychopharmacology (RUPP) Autism Network, Risperidone in children with autism and serious behavioral problems, N. Engl. J. Med., 347, 314, 2002 71. McDougle, C.J. et al., Risperidone treatment of children and adolescents with pervasive developmental disorders: a prospective open-label study, J. Am. Acad. Child Adolesc. Psychiatry, 36, 685, 1997.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 440 Wednesday, February 15, 2006 4:34 PM

440

Understanding Autism: From Basic Neuroscience to Treatment

72. Research Units on Pediatric Psychopharmacology Autism Network, Risperidone treatment of autistic disorder: longer term benefits and blinded discontinuation after six months, Am. J. Psychiatry, 162, 1361, 2005. 73. Bymaster, F.P. et al., Neurochemical evidence for antagonism by olanzapine of dopamine, serotonin, alpha 1-adrenergic and muscarinic receptors in vivo in rats, Psychopharmacology (Berl), 124, 87, 1996. 74. Potenza, M.N. et al., Olanzapine treatment of children, adolescents, and adults with pervasive developmental disorders: an open-label pilot study, J. Clin. Psychopharmacol., 19, 37, 1999. 75. Malone, R.P. et al., Olanzapine versus haloperidol in children with autistic disorder: an open pilot study, J. Am. Acad. Child Adolesc. Psychiatry, 40, 887, 2001. 76. Arnt, J. and Skarsfeldt, M., Do novel antipsychotics have similar pharmacological characteristics: a review of the evidence, Neuropsychopharmacology, 18, 63, 1998. 77. Martin, A. et al., Open-label quetiapine in the treatment of children and adolescents with autistic disorder, J. Child Adolesc. Psychopharmacol., 9, 99, 1999. 78. Hardan, A., Jou, R., and Handen, B., Quetiapine open-label trial in children and adolescents with developmental disorders, in New Research Program and Abstracts of the 156th Annual Meeting of the American Psychiatric Association, San Francisco, CA, 2003, p. 163. 79. Findling, R.L. et al., Quetiapine in nine youths with autistic disorder, J. Child Adolesc. Psychopharmacol., 14, 287, 2004. 80. Corson, A.H. et al., A retrospective analysis of quetiapine in the treatment of pervasive developmental disorders, J. Clin. Psychiatry, 65, 1531, 2004. 81. Tandon, R., Harrigan, E., and Zorn, S.H., Ziprasidone: a novel antipsychotic with unique pharmacology and therapeutic potential, J. Serotonin Res., 4, 159, 1997. 82. McDougle, C.J., Kem, D.L., and Posey, D.J., Case series: use of ziprasidone for maladaptive symptoms in youths with autism, J. Am. Acad. Child Adolesc. Psychiatry, 41, 921, 2002. 83. Psychopharmacological Drugs Advisory Committee, FDA Background Document on Zeldox (PDAC 0700.M01), Rockville, MD, Food and Drug Administration, 2000. 84. Burris, K.D. et al., Aripiprazole, a novel antipsychotic, is a high-affinity partial agonist at human dopamine D2 receptors, J. Pharmacol. Exp. Ther., 302, 381, 2002. 85. Stigler, K.A., Posey, D.J., and McDougle, C.J., Aripiprazole for maladaptive behavior in pervasive developmental disorders, J. Child Adolesc. Psychopharmacol., 14, 455, 2004. 86. Geller, E. et al., Preliminary observations on the effect of fenfluramine on blood serotonin and symptoms in three autistic boys, N. Engl. J. Med., 307, 165, 1982. 87. Ritvo, E.R. et al., Fenfluramine treatment of autism: UCLA collaborative study of 81 patients at nine medical centers, Psychopharmacol. Bull., 22, 133, 1986. 88. Campbell, M. et al., Naltrexone in autistic children: an acute dose range tolerance trial, J. Am. Acad. Child Adolesc. Psychiatry, 28, 200, 1989. 89. Campbell, M. et al., Efficacy and safety of fenfluramine in autistic children, J. Am. Acad. Child Adolesc. Psychiatry, 27, 434, 1988. 90. Campbell, M. et al., Naltrexone in autistic children: behavioral symptoms and attentional learning, J. Am. Acad. Child Adolesc. Psychiatry, 32, 1283, 1993. 91. Willemsen-Swinkels, S.H.N., Buitelaar, J.K., and van Engeland, H., The effects of chronic naltrexone treatment in young autisitic children: a double-blind placebocontrolled crossover study, Biol. Psychiatry, 39, 1023, 1996.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 441 Wednesday, February 15, 2006 4:34 PM

Pharmacological Treatments

441

92. Carlsson, M.L., Hypothesis: is infantile autism a hypoglutamatergic disorder? Relevance of glutamate-serotonin interactions for pharmacotherapy, J. Neural. Transm. 105, 525, 1998. 93. McDougle, C.J., Current and emerging therapeutics of autistic disorder and related pervasive developmental disorders, in Neuropsychopharmacology — The Fifth Generation of Progress, Davis, K.L., Charney, D., Coyle, J.T., and Nemeroff, C., Eds., Lippincott, Williams and Wilkens, Philadelphia, PA, 2002, chap. 42. 94. Cotman, C.W. et al., Excitatory amino acid neurotransmission, in Psychopharmacology: The Fourth Generation of Progress, Bloom, F.E. and Kupfer, D.J., Eds., Raven Press, New York, 1995, chap. 7. 95. Uvebrant, P. and Bauziene, R., Intractable epilepsy in children: the efficacy of lamotrigine treatment, including non-seizure-related benefits, Neuropediatrics, 25, 284, 1994. 96. Davanzo, P.A. and King, B.H., Open trial of lamotrigine in the treatment of selfinjurious behavior in an adolescent with profound mental retardation, J. Child Adolesc. Psychopharmacol., 6, 273, 1996. 97. Belsito, K.M. et al., Lamotrigine therapy for autistic disorder: a randomized, doubleblind, placebo-controlled trial, J. Autism Dev. Disord., 31, 175, 2001. 98. Krug, D.A., Arick, J., and Almond, P., Autism Behavior Checklist Record Form, Austin, TX: PRO-ED, 1993. 99. Lord, C. et al., Autism diagnostic observation schedule: a standardized observation of communicative and social behavior, J. Autism Dev. Disord., 19, 185, 1989. 100. Schopler, E., Reichler, R., and Renner, B., Childhood Autism Rating Scale (CARS), Los Angeles, CA: Western Psychological Services, 1988. 101. King, B.H. et al., Case series: amantadine open-label treatment of impulsive and aggressive behavior in hospitalized children with developmental disabilities, J. Am. Acad. Child Adolesc. Psychiatry, 40, 654, 2001. 102. King, B.H. et al., Double-blind, placebo-controlled study of amantadine hydrochloride in the treatment of children with autistic disorder, J. Am. Acad. Child Adolesc. Psychiatry, 40, 658, 2001. 103. Simeon, J. et al., D-cycloserine therapy of psychosis by symptom provocation, Comp. Psychiatry, 11, 80, 1970. 104. D’Souza, D.C., Charney, D., and Krystal, J., Glycine site agonists of the NMDA receptor: a review, CNS Drug Rev., 1, 227, 1995. 105. Battaglia, B. et al., Toxicity of cycloserine combined with isoniazid in the treatment of tuberculosis in children, Am. Rev. Respir. Dis., 83, 751, 1961. 106. van Berckel, B. et al., Efficacy and tolerance of D-cycloserine in drug-free schizophrenic patients, Biol. Psychiatry, 40, 1298, 1996. 107. Goff, D. et al., Dose-finding study of D-cycloserine added to neuroleptics for negative symptoms in schizophrenia, Am. J. Psychiatry, 152, 1213, 1995. 108. Goff, D.C. et al., A placebo-controlled trial of D-cycloserine added to conventional neuroleptics in patients with schizophrenia, Arch. Gen. Psychiatry, 56, 21, 1999. 109. Heresco-Levy, U. et al., Placebo-controlled trial of D-cycloserine added to conventional neuroleptics, olanzapine, or risperidone in schizophrenia, Am. J. Psychiatry, 159, 480, 2002. 110. Evins, A.E. et al., D-cycloserine added to risperidone in patients with primary negative symptoms of schizophrenia, Schizophr. Res., 56, 19, 2002. 111. Goff, D. et al., A placebo-controlled crossover trial of D-cycloserine added to clozapine in patients with schizophrenia, Biol. Psychiatry, 45, 512, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C018.fm Page 442 Wednesday, February 15, 2006 4:34 PM

442

Understanding Autism: From Basic Neuroscience to Treatment

112. Posey, D.J. et al., A pilot study of D-cycloserine in autistic disorder, Am. J. Psychiatry, 161, 2115, 2004. 113. Price, L.H., Carpenter, L.L., and Rasmussen, S.A., Drug combination strategies, in Refractory Mood Disorders, Amsterdam, J.D., Hornig-Rohan, M., and Nierenberg, A.A., Eds., Cambridge University Press, New York, 2000, chap. 10. 114. Nelson, J.C. and Bowers, M.B., Delusional unipolar depression, Arch. Gen. Psychiatry, 35, 1321, 1978. 115. McDougle, C.J. et al., Haloperidol addition in fluvoxamine-refractory obsessive compulsive disorder: a double-blind, placebo-controlled study in patients with and without tics, Arch. Gen. Psychiatry, 51, 302, 1994. 116. McDougle, C.J. et al., A double-blind, placebo-controlled study of risperidone addition in serotonin reuptake inhibitor-refractory obsessive-compulsive disorder, Arch. Gen. Psychiatry, 57, 794, 2000. 117. Krystal, J.H. et al., NMDA agonists and antagonists as probes of glutamatergic dysfunction and pharmacotherapies for neuropsychiatric disorders, Harv. Rev. Psychiatry, 7, 125, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 443 Wednesday, February 15, 2006 6:05 PM

19

Behavioral, Educational, and Developmental Treatments for Autism Sally J. Rogers and Sally Ozonoff

CONTENTS Introduction............................................................................................................444 Language Interventions .........................................................................................444 Studies Using a Didactic Behavioral Approach ............................................444 The Naturalistic Behavioral Language Interventions ....................................446 Developmental Language Approaches...........................................................448 Social Interventions ...............................................................................................450 Interventions with Younger or Less Verbal Children.....................................450 Adult Use of Dyadic Engagement ............................................................450 Self-Management Techniques....................................................................451 Games with Objects...................................................................................451 Pivotal Response Training .........................................................................451 Peer-Mediated Interactions ........................................................................451 Role-Playing Games ..................................................................................452 Peers as Tutors...........................................................................................452 Peers Using PRT........................................................................................453 Integrated Playgroup..................................................................................453 Interventions for Older and More Verbal Children...............................................453 Social Skills Training .....................................................................................453 Special Interest Games ...................................................................................455 Repetitive and Restrictive Behavioral Repertoire .................................................455 Comprehensive Intervention Approaches for Preschoolers ..................................458 The Work of Ivar Lovaas and Colleagues .....................................................458 Developmentally Oriented Treatments...........................................................460 New Approaches ....................................................................................................462 Conclusions............................................................................................................463 Acknowledgments..................................................................................................464 References..............................................................................................................464

443

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 444 Wednesday, February 15, 2006 6:05 PM

444

Understanding Autism: From Basic Neuroscience to Treatment

INTRODUCTION Of the many views of autism that have been radically revised by research published over the past 20 years, one of the most compelling is that behavioral, or psychosocial, interventions can change the course of the disorder. One of the great contrasts of autism as a clinical condition is that, on the one hand, autism profoundly changes the most ordinary and very primitive aspects of social and communicative behavior and of interactions with the nonsocial environment, whereas, on the other hand, persons with autism at all ages and all levels of severity show positive responses to treatments that focus on changing these key behavioral deficiencies or differences. The focus of this chapter will be on reviewing treatments that target the three key symptom sets that define autism: (1) impairments in social relatedness and reciprocity, (2) impairments in language and communication, and (3) a reduced and repetitive repertoire of behaviors. In addition, this chapter will cover comprehensive preschool intervention approaches, given the intensity and importance of these interventions in treatment planning (NAS Committee report). Review of interventions will be restricted to those that involve children and adolescents.

LANGUAGE INTERVENTIONS The abnormal development and use of spoken language is one of autism’s most unusual and striking features. Kanner’s (1943) first description of autism painted a very detailed picture of the two main subgroups of language symptoms in children with autism. One pattern involved children who did not develop spoken language. Unlike other diagnostic groups of children who for one reason or another do not develop speech, these children did not develop an alternative communication system using distal signals either, and instead moved people around physically, manipulating hands, pushing and pulling on others’ limbs and bodies, when the child needed adult help. These children appeared to understand little speech, and long-term follow-up revealed little change in their relatively noncommunicative status over many years (Kanner, 1971). The second pattern involved children who produced speech that was markedly atypical. These children tended to mimic speech rather than generate their own sentences, with patterns of both immediate and delayed echolalia. A second atypical feature was the lack of communicative content of their speech. These children did not use language to share their experiences with others or to gain information by asking questions. Other atypical features of the children’s verbal language included unusual prosody and intonation, unusually mature syntax patterns given the children’s age, and the use of unusual words, or neologisms (Kanner, 1943).

STUDIES USING

A

DIDACTIC BEHAVIORAL APPROACH

Applied behavior analysis involves the use of principles largely derived from the research of B. F. Skinner (1957) on learning processes to change behavior, and treatment of autism using this approach dates back to Hewett (1965), who reported the first use of operant teaching to develop language in autism. The approach applies

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 445 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

445

specific behavioral contingencies to increase or decrease the frequency of specific behaviors and the use of shaping, fading, and chaining procedures to teach new behaviors. The use of applied behavior analysis takes several forms. One method of delivering behavior therapy to teach new skills involves introducing a new antecedent-behavior chain via specific prompts or cues for the new behavior in the presence of a specific stimulus and then carrying out a large number of practice trials during which the prompts are systematically reduced. In this chapter the term didactic behavioral teaching will be used to refer to the massed trial approach in which the teacher signals the start of the teaching episode and conducts a number of repeated trials in close proximity in a carefully controlled environment. The didactic behavioral approach has been supported by many studies as being effective in teaching a wide range of skills, and the power of its teaching lies in several of its elements, among them are the careful management of reinforcer strength, trial-by-trial data collection with adjustment of teaching strategies to enhance learning, high level of environmental control with resulting control of child attention, use of massed practice trials to build new skills quickly, and the very precise application of shaping, fading, and chaining procedures that can be gained in a highly controlled situation. Didactic teaching approaches have been effective in teaching speech to nonverbal children with autism and increasing the use and complexity of speech and vocabulary size in verbal children with autism. Children who, at the start of treatment were completely nonverbal, lacking even consonant sounds, have been taught to speak and to use increasingly complex syntax and vocabulary, including response to questions, use of gestures, and supersegmentals such as intonation and volume. Various pragmatic functions have been taught, such as requesting, commenting, negation, and asking questions. For examples, see Krantz et al., 1981, for a discussion on teaching complex sentences; Ross and Greer, 2003, for vocal imitation; Williams et al., 2003, for use of questions; and Yoder and Layton, 1988, for development of first words. The early reports involved children in hospital settings, who could receive treatment all day long. When these children left the training setting, they lost many of their newly learned language skills. The fragility of their gains demonstrated the importance of ongoing environmental supports for speech (Lovaas et al., 1973). This led to a shift away from hospital-like treatment settings with specially trained therapists and towards explorations of treatment delivery in natural settings, where children’s new learning could be constantly supported in their natural environments by the people in those settings. Several papers examined whether parents could master the didactic approach and deliver the treatment at home. Howlin and Rutter (1989) and Harris (Harris et al., 1983; Harris et al., 1981) reported that a combination of weekly group training classes for parents combined with individual weekly or biweekly home visits resulted in the child’s gains, with two important caveats. First, language gains were almost completely confined to children who already had some speech at the start of treatment. Nonverbal children made minimal gains, perhaps because the shaping of speech in nonspeaking children requires more intensity, clinical training, or treatment precision than parents can easily implement. Second, most gains occurred in the first six months, during which time parents received both group and individual

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 446 Wednesday, February 15, 2006 6:05 PM

446

Understanding Autism: From Basic Neuroscience to Treatment

instruction, with little progress beyond this. Undergraduate college students have also been trained to carry out this type of teaching at high levels of fidelity, resulting in child improvement (Lovaas, 1987; Smith et al., 2000) as long as ongoing training and support are provided to the students. Thus, the didactic approach can be used successfully to teach children with autism spoken language by people without advanced educational degrees, as long as ongoing monitoring, training, and oversight of nonprofessionals are provided by sophisticated therapists who are directing the individualized treatment of the child. Most of the studies cited in the preceding text have involved single-subject designs, which tend to demonstrate short-term changes in very specific behaviors. However, group studies examine the child’s progress over somewhat longer periods of time, use more general measures of progress, and use comparison groups to determine the effect of the treatment. Several group studies have also demonstrated the ability of the didactic method to increase language development. Lovaas (1987) reported significantly greater gains in expressive and receptive language by children with autism who had received his treatment model at a high level of intensity during the preschool period than a comparison group who received the same model at a low level of intensity. Similarly, Eikeseth et al. (2002) reported greater language progress on standardized tests by a group of somewhat older children receiving Lovaas’ treatment model rather than an eclectic model involving the same intensity of treatment. Although Smith et al., (2000) originally reported that the children receiving Lovaas’ method in their study had significantly improved language compared to randomly assigned controls, they later indicated that the difference was not significant (Smith et al., 2001).

THE NATURALISTIC BEHAVIORAL LANGUAGE INTERVENTIONS The “naturalistic” application of principles of applied behavioral analysis involves a very different type of interaction surrounding the discrete teaching episodes. The use of naturalistic behavioral language interventions in autism was pioneered by Robert Koegel and Laura Schreibman in their intervention approach: pivotal response training (PRT) (R.L. Koegel and Frea, 1993; R.L. Koegel et al., 1987; L.K. Koegel, R.L. Koegel, Hurley, and Frea, 1992; Schreibman, 1988; Schreibman and Pierce, 1993). For over 30 years, these two treatment experts have published single-subject design papers demonstrating the effectiveness of naturalistic applications of behavioral teaching approaches to develop a number of social, communicative, play, and self-management skills, as cited in the following text. Gail McGee has demonstrated a very similar approach with good success in several studies using incidental teaching (McGee et al., 1983). Rapid acquisition of skills resulted in evidence of maintenance and generalization of skills, enhanced motivation for learning, and lessened problems with unwanted behaviors. Key elements include: (1) teaching episodes are initiated by the child’s behavior via requests or gestures for preferred items, (2) teaching takes place in the context of ongoing activities with items of the child’s interest as part of the naturally occurring stimuli in the environment, (3) the teaching stimuli are selected by the child and contingent access to them is the reinforcement, and (4) prompt strategies for elaborated language vary according to the child’s initiating

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 447 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

447

behavior (McGee et al., 1983). Warren and Kaiser (1988) suggest that the common elements included “an orientation toward teaching the form and content of language in the context of normal use, using developmentally typical techniques like dispersed trials, following the child’s lead, and teaching in the context of normal communicative exchanges.” They attributed the rapid development of this approach to the convergence of behaviorists, who were following an inductive approach to achieving improved generalization and initiation of speech, with developmentalists, who followed a deductive approach, to apply current theories and research on typical language development to children with atypical communication development. A number of papers, virtually all using single-subject designs, have demonstrated the efficacy of a naturalistic behavioral approach in teaching many linguistic responses to children with autism. Work by R.L. Koegel et al. (1987) demonstrated that completely nonverbal children could learn to speak using this approach. As described in the excellent reviews by Delprato (2001), Lynn Koegel (2000), and Howard Goldstein (2000), children with autism have demonstrated increases in frequency, spontaneity, and syntactic sophistication of their language when using naturalistic approaches. Examples of the use of naturalistic approaches to successfully teach various aspects of language include: teaching prepositions (McGee et al., 1983), using grammatical morphemes (Carter, 2001), improving articulation (R.L. Koegel, Camarata, L.K. Koegel, Ben-Tall, and Smith, 1998), teaching use of yes and no (Neef et al., 1984), teaching joint attention behavior (Whalen and Schreibman, 2003), and using questions appropriately (L.K. Koegel, Camarata, Valdez-Menchaca, and R.L. Koegel, 1998). All aspects of language, pragmatics, semantics, syntax, and phonation, have been successfully treated using naturalistic approaches. Furthermore, comparisons between didactic and naturalistic approaches have demonstrated some advantages, including improved behavior (L.K. Koegel, R.L.Koegel, Hurley, and Frea, 1992; R.L. Koegel, L.K. Koegel, and Surratt, 1992) and greater progress (especially maintenance and generalization of newly learned skills) when using naturalistic strategies rather than didactic strategies (see Delprato, 2001, for a review of the comparative studies). Although all the aforementioned studies involve individual treatment, several studies have also demonstrated delivery of this approach in group treatment. Both Peck (1985) and McBride and Schwartz (2003) trained classroom teaching staff to increase the opportunities for child initiation and response embedded inside ongoing classroom activities, resulting in increases in children’s communications including speech. Similarly, McGee’s current use of incidental teaching is being delivered in the integrated, inclusive Walden preschool at Emory University, although the outcome data have been provided in descriptive papers rather than empirical studies (McGee, Morrier, and Daly, 1999; Strain et al., 2001). The effectiveness of this approach most likely results from four aspects of the technique. First, child language functions to achieve chosen goals and reinforcers, which apparently strengthens the power of the reinforcers. Second, the teaching style heavily emphasizes child initiation, either verbal or nonverbal, and the exchanges involve reciprocity of the child and adult. Third, the social functions of language that lead to reinforcers are highlighted, so the core pragmatic aspects of language permeate the intervention. Finally, the emphasis on child motivation and natural

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 448 Wednesday, February 15, 2006 6:05 PM

448

Understanding Autism: From Basic Neuroscience to Treatment

reinforcers adds a positive affective element to the interactions, which may enhance memory for learning. Whether naturalistic techniques are always preferred over didactic teaching is a matter of current debate. Although several literature reviews have maintained that naturalistic teaching is always advantageous (Delprato, 2001; L.K. Koegel, 2000), others have suggested that comparative data are insufficient to support this conclusion (Goldstein, 2000; Smith, 2001). The actual teaching techniques used in the two methods involve the same behavioral teaching repertoire — prompting, shaping and chaining, as well as reliance on imitation once it is learned — to shape new behaviors. Two other well-known and empirically supported behavioral approaches need to be mentioned. Bondy and Frost’s (1994) picture exchange system (PECS) focuses quite heavily on augmentative or alternative communication supports. The PECS approach uses careful shaping techniques to teach children to hand picture icons to adults to initiate requests. The use of PECS is associated with acquisition of verbal skills in a number of children who have used it (Bondy and Frost 1994; CharlopChristy et al., 2002). The approach known as verbal behavior (Sundberg and Michael, 2001) is solidly based in Skinnerian theory of language acquisition (Skinner, 1957), but the intervention differs from both the massed trial approach of Lovaas’s and PRT in its delivery. Verbal behavior emphasizes the child’s production of communicative behavior in highly motivating functional situations. Child gestural and verbal communications are shaped using reinforcers that are intrinsically reinforcing for the child. Verbal behavior is applied to both vocal and nonverbal communications, and it uses the effective teaching strategies of applied behavior analysis, including errorless learning, to shape children’s natural and conventional gestures and vocal productions, and echoed speech to build spontaneous, functional, fluent communicative repertoires. The approach is well supported in the single-subject research literature. Although both PECS and verbal behavior involve a naturalistic application of applied behavior analysis, the didactic and naturalistic teaching approaches should probably not be considered as mutually exclusive but rather as integrative. Every treatment technique with demonstrated utility represents an additional tool in the tool chest of interventionists.

DEVELOPMENTAL LANGUAGE APPROACHES The finding that nonverbal communication was a better predictor than early speech for predicting later verbal ability (Mundy et al., 1990) led to the third main paradigm for language intervention discussed here: a model built directly from the literature on normal language development, referred to here as the developmental pragmatics approach. Excellent recent descriptions of this line of thinking have been provided by Wetherby and colleagues (2000) and Peter Mundy (Mundy and Crowson, 1997; Mundy and Markus, 1997). The most elaborated description of this kind of treatment has been provided by Prizant et al. (2000) as the SCERTS model of intervention, with its focus on social communication, emotional regulation, and transactional support as both the major components of the intervention and the treatment priorities. This approach emphasizes the development of the full range of interpersonal communicative

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 449 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

449

behaviors, including eye contact, shared affect, intentional vocalization, and manual gestures, as well as speech, to achieve reciprocal communicative exchanges regarding interactions involving objects and social games. Functional communication, rather than speech, is the primary goal. This approach bears many resemblances to the behavioral naturalistic teaching methods described earlier, and provides the potential for an unusual convergence between developmentalists and behaviorists. Kaiser and her colleagues have demonstrated this merged developmental–behavioral approach in two single-subject design papers concerning the efficacy of their model “enhanced milieu teaching” for children with autism (Hancock and Kaiser, 2002; Kaiser et al., 2000). Empirical studies of the efficacy of interventions based on developmental pragmatics approach are few in number, and typically involve pre-post rather than controlled group designs (Chandler et al., 2002; Mahoney and Perales, 2003; Rogers and Lewis, 1989; Rogers and DiLalla, 1991). Lack of comparison or control groups and of blind raters limit the conclusions that can be drawn. Other wellknown developmental pragmatic models include the work of Greenspan, Weider and colleagues (1997) — the Developmental Individual Relationship Based Model (DIR) and the Hanen approach (Sussman, 1999). However, these approaches have not yet published data-based empirical studies involving children with autism. Finally, augmentative or alternative supports for communication provide access to functional and symbolic communication to nonspeaking children (and adults), and these are an extremely important part of any intervention for children who do not have a functional means of communicating. Concerns about whether the use of augmentative systems inhibit development of speech have been alleviated by findings that some users develop speech while using them, but it is not yet known to what extent augmentative systems assist with acquisition of spoken language. Bondy and Frost’s (1994) data suggesting that children using PECS spontaneously acquired some speech has been confirmed by a recent replication by Charlop-Christy et al. (2002). However, whether it was the use of the visual support or the carefully delivered context of the instruction that was responsible for the child’s language acquisition cannot be determined in those studies. The only experiment in the literature that examined that specific question is the elegant experimental design by Yoder and Layton (1988), who demonstrated that there was no advantage to speech acquisition when a visual system was included — in this case, sign language. Given the widespread use of PECS and other visual systems in programs for young children with autism, this is a very important empirical question and needs scientific attention. The importance of the pragmatics of communication is unmistakable, and all the approaches described here have focused on teaching children with autism to use spoken language to accomplish a variety of interpersonal functions. What is not clear is whether nonverbal communication skills such as joint attention, gesture, and eye contact need to be a precursor to other kinds of language teaching for nonverbal children with autism. It may be that simultaneous interventions in pragmatics of nonverbal communication, understanding of spoken language, and development and shaping of spoken language can be carried out and may result in more rapid acquisition of useful communicative speech.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 450 Wednesday, February 15, 2006 6:05 PM

450

Understanding Autism: From Basic Neuroscience to Treatment

A final point to make in this section involves the foundational skill of imitation. Some teaching approaches target imitation training in a very specific and didactic way, as seen in Lovaas’ (1981) curriculum. Verbal behavior also focuses on imitation as one aspect of its curriculum. The naturalistic and developmental approaches develop imitation incidentally as a means of learning from others, and not as an end in itself, and imitation skills are prompted and shaped as part of the larger repertoire of skills that are taught during communicative and play activities in those approaches. Imitation is a core deficit area in early autism and may be one of the most impairing aspects of this disorder early on, and most effective interventions use imitation as a key teaching tool, either directly or incidentally.

SOCIAL INTERVENTIONS Social dysfunction is the single most defining (Kanner, 1943) and arguably its most handicapping feature of autism. A variety of social interventions have been designed, empirically examined, and published in the autism literature. These interventions differ in a variety of ways: the age group of the people with autism involved, the target behavior of the intervention (initiation, response, and maintenance), the kind of social partner involved — peer or adult, the intervention strategy used, and the characteristics of the intervener. Although earlier efforts in the field involved adult-directed teaching, with demonstrated effectiveness, the field has moved to more careful attention being paid to the ecology of children’s social interactions in natural settings, with a concurrent shift to a greater focus on social interactions with peers. As Simpson and colleagues have pointed out, adult-directed instruction in social skills has not been grounded in natural contexts and use of typical stimuli and reinforcers (Simpson et al., 1997).

INTERVENTIONS

WITH

YOUNGER

OR

LESS VERBAL CHILDREN

Interventions for preschoolers or children with limited verbal ability have largely focused on increasing attention, initiative, responsivity, and play with others. Studies have focused on increasing interactions with adults who are in the teaching role, as well as interventions focused on increasing interactions with peers. A number of different intervention techniques have been successful; these will be reviewed in turn in the following subsections. Adult Use of Dyadic Engagement Several papers have focused on teaching adults to focus on mirroring and imitating preschool children’s play acts, stimulating dyadic interactions, and providing contingent and responsive interaction styles (Dawson and Galpert, 1990; Field et al., 2001; Rogers et al., 1986). Dawson and Rogers’ studies involved noncontrolled group designs without reversals or other methods that unequivocally demonstrate a causal relationship. Follow-up after two weeks and after six months demonstrated increased responsiveness and positive affect, initiations, and imitations of the children to adult interactions. The work by Field et al. (2001) demonstrated causal relationships between adult use of imitation and mirroring and increase in child interest, attention, and initiation.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 451 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

451

Self-Management Techniques Several groups have used visual systems to teach children with autism to cue themselves to engage socially. Krantz and McClannahan (1993, 1998) used a visual cueing system of word cards embedded in a child’s schedule that prompted the child to approach an adult or a peer and initiate a joint attention request (look, watch me, etc.) or another social initiation. Each child increased the number of initiations dramatically, maintaining them when the cues were faded. By using a visual cue, potential difficulties seen in adult prompting paradigms, such as disruptions in ongoing social interactions, constant presence of cueing adult, and difficulties with fading adult prompts, were avoided. L.K. Koegel, R.L. Koegel, Hurley, and Frea (1992) reported a multiple baseline across settings and four subjects focused on increasing appropriate verbal responses to others’ social initiations. Training was conducted across several settings. Children used a wrist counter to tally frequencies, which were converted to points and exchanged at intervals for small rewards, mainly edibles. The reinforcement schedule was thinned drastically (1:30) within the first few training sessions. Each of the children demonstrated rapid improvement in appropriate responding that remained at high levels across the rest of the study, as well as in collateral decreases in inappropriate language and disruptive behavior. Withdrawal of the procedure for two subjects resulted in decreases in responding. Games with Objects Coe et al. (1990) reported a direct-instruction procedure using multiple baselines across three 6-year-old children, two with autism, to play ball with an adult. Four steps in the chain were taught using primary reinforcers; three (pick up, throw, and initiate) taught at the same time, and the last (praise) taught after acquisition of the others. All three children increased both verbal and nonverbal behaviors associated with ball play, with initiation being the hardest to acquire. However, no data involving maintenance or generalization was reported. Pivotal Response Training Adult use of the PRT techniques to improve social play skills has demonstrated social improvements in autistic children involving responsiveness to adult cues and increased initiation and interaction with adults (Stahmer, 1995; Thorp et al., 1995). Effective PRT techniques included children selecting preferred materials, adults modeling, shaping procedures used, approximations reinforced with task-related reinforcers, and high levels of success assured, with evidence of maintenance and generalization. Peer-Mediated Interactions In this approach, typical peers are taught to initiate certain kinds of interactions with perseverance: sharing, helping, giving affection, and praise. Peers role-play with adults until they have learned the strategies successfully, and then are cued by adults

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 452 Wednesday, February 15, 2006 6:05 PM

452

Understanding Autism: From Basic Neuroscience to Treatment

to begin to interact with the target children around typical play materials and activities. The peers are reinforced by adults for their efforts, and the reinforcements are systematically and carefully reduced. These strategies are powerful in increasing social interactions of young children with autism, and both generalization and maintenance have been demonstrated in inclusive preschool classes, as reported in many published multiple-baseline studies (Goldstein et al., 1992; McGee et al., 1992; Odom et al., 1999; Odom and Strain,1986; Strain et al., 1979; Strain et al., 1977). Furthermore, untrained peers have been found to increase their interactions with the children with autism as well when in the presence of trained peers, demonstrating generalization across peers (Shafer et al., 1984). Lord and Hopkins (1986) examined the several variables related to peer interactions for six school-aged children with autism. They demonstrated that same-age peers elicited more social behavior than did younger peers, and that daily exposure in peer play sessions significantly increased a number of social behaviors in children with autism, including proximity, appropriate play, time spent looking at a peer, and time engaged socially, effects that generalized to new peers. Social initiations were not particularly affected by the intervention, however. Dewey and colleagues (1988) demonstrated that rule-governed games facilitated the most complex social interactions, were the most fun, and kept the children most involved with the interaction. Construction materials were the next most effective in facilitating more complex interactions compared to dramatic play and functional play. Role-Playing Games Goldstein and colleagues (1988) taught sociodramatic scripts to two trios of preschool children consisting of two typical peers and a child with an autism-related disorder. Each script contained three social roles and each child was trained in each role, with teacher instruction gradually reduced, until each child could carry out 80% of each role’s script. Following training, increases in child interaction and generalization across settings and other behaviors improved during free play periods at preschool. However, the effects depended on continued teacher prompts in roleplaying activities and did not, in general, result in increases in social exchanges across the preschool day. Peers as Tutors Two studies have demonstrated that using typical peers as tutors in academic areas was more effective than teacher-led academic instruction for academic learning and social interaction in children with autism, with unexpected gains in the peer tutors as well (Dugan et al., 1995; Kamps et al., 1994). Blew and colleagues (1985) demonstrated that peers could successfully teach multistep adaptive skills in community settings (e.g., checking out a library book, making a purchase in a store) after the peer learned discrete trial teaching techniques as well as the steps of each adaptive behavior sequence. The procedure required full peer teaching and reinforcements for both the peer-tutor and the learner, but resulted in improved attention and responding of the child with autism to the peer compared to adult teachers.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 453 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

453

Peers Using PRT Pierce and Schreibman (1997) taught eight typical peers to use PRT techniques to involve two children with autism in improved social interaction during toy play. Multiple baselines demonstrated that each target child made rapid increases in maintaining social interactions that generalized across people, settings, and materials and was maintained over a two-month follow-up. There was less effect on social initiations. Integrated Playgroup The final intervention approach to be discussed in this section involves another method of facilitating learning through interactions with typical peers. Wolfberg and Schuler (1993) have demonstrated the effectiveness of this playgroup approach in facilitating language and social behavior in children with autism. The groups involve five or so children, three with typical development and two with autism, for a weekly play hour in a specially prepared play space that fosters interactive play, including pretend play. Adults facilitate interactive play during the session while promoting full child engagement. Single-subject studies have been carried out with both elementary and preschool children and have demonstrated increases in spontaneous language, symbolic and functional play, and joint attention behavior during treatment for extended periods of time, even when adult facilitation is withdrawn (Wolfberg and Schuler, 1993; Zercher et al., 2001).

INTERVENTIONS FOR OLDER AND MORE VERBAL CHILDREN Although more able individuals with autism are generally less severely affected than children with autism who have little language or are nonverbal, they exhibit deficits in social knowledge and skills, have fewer same-age friendships and social group experiences, and endure high rates of teasing and bullying (Tantam, 2003), and remediation is still needed. As with interventions reviewed for less verbal children, techniques used to enhance social adjustment in the more able group vary by teaching strategy, delivery format, social partner, and target behavior.

SOCIAL SKILLS TRAINING A commonly used adult-mediated group intervention is social skills training. Such groups teach specific skills and permit opportunities for guided practice using a variety of techniques, including didactic lessons, role-play, videotape review and analysis, collaborative activities, and homework or practice outside the sessions, among others (for a review, see Krasny et al., 2003). Several studies have used repeated-measures designs to examine efficacy of such groups (Marriage et al., 1995; Mesibov, 1984), but only a few have assessed the outcome with standardized objective measures (Howlin and Yates, 1999; Ozonoff and Miller, 1995; Williams, 1989), and only two have used a control group (Ozonoff and Miller, 1995; Solomon

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 454 Wednesday, February 15, 2006 6:05 PM

454

Understanding Autism: From Basic Neuroscience to Treatment

et al., 2004). Gains reported are usually unexceptional, and the few studies that have examined “real-life” social skills exhibited outside the clinic setting find little evidence of generalization (Howlin and Yates, 1999; Marriage et al., 1995; Ozonoff and Miller, 1995). Kamps and colleagues (1992) used social skills instruction in playgroups consisting of both 11 typical peers and 3 high-functioning first graders with autism. Four daily group play sessions were held, each group consisting of one target student, three peers, and some type of planned activity. Social skills training was conducted for the first 10 min, followed by an additional 10 min of play in a central activity. After the entire curriculum was taught, feedback was provided for all students during playgroups. Follow-up consisted of several social skills groups held at the end of the school year. Significant increases in social skills, length of interactions, and consistency of responding were found for all target students with good maintenance across time. Despite only modest evidence of efficacy, social skills groups are widely sought and offered. Future studies may yet find improved outcome with larger samples and both more sensitive and more relevant outcome measures (e.g., examining consumer satisfaction, depression, self-acceptance, etc.). A recent investigation taught similar social skills, but in one-to-one teaching sessions followed by guided practice with a typical peer. Quite impressive and significant gains in social problem-solving, emotion knowledge, social initiations and responses with peers, and teacher-reported social behaviors in the classroom were reported (Bauminger, 2002), although there was no control group and observations and ratings were made by individuals involved in the interventions (e.g., not in a blind fashion). Peer network interventions utilize typical peers and naturalistic settings (e.g., the classroom, lunchroom, and playground) to enhance social skills. Multiple studies, including one with a randomized, controlled, repeated-measures design (Roeyers, 1996), have demonstrated that such peer-mediated approaches can significantly increase social initiations and responses to peers in higher-functioning children with autism (Kamps et al., 1997; Laushey and Heflin, 2000). One well-known approach of this type that is often used for higher-functioning children, the Circle of Friends program, has never been empirically tested. There are other approaches that need to be mentioned. Video modeling, in which a videotaped model presents a behavior to be imitated, has been shown to result in quick acquisition of social behaviors such as affection and greetings (Charlop and Walsh, 1986; LeBlanc et al., 2003). In one study, video modeling was superior to in vivo modeling (Charlop-Christy et al., 2000). The widely used Social Story intervention (Gray, 2000), in which brief written passages highlight social information, explain social cues, and suggest appropriate social behaviors, still lacks solid empirical support. Several small multiple-baseline studies have been carried out but with mixed interventions or lack of replication of effects across baselines (Hagiwara and Myles, 2001; Kuttler et al., 1998; Swaggart et al., 1995; Thiemann and Goldstein, 2001). The ability to appreciate the social perspectives of other people (e.g., theory of mind) can be improved with specific teaching (Bernard-Opitz et al., 2001; Hadwin et al., 1996; Ozonoff and Miller, 1995; Reinecke et al., 1997; Silver and Oakes, 2001; Swettenham, 1996; Wellman et al., 2002), but most of these studies find limited generalization to other social skills.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 455 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

455

SPECIAL INTEREST GAMES One creative adult instruction approach used children’s special interests as the basis for invented games. Baker and colleagues (1998) reported a study of three highfunctioning elementary aged students. An adult created group games from each of these children’s special interests and then taught the game to the child and some typical peers during a playground period at school. Results included dramatic increases in peer interaction that were maintained through the intervention and follow-up period, generalization to other activities, and increases in positive affect in both target children and peers during interactions. Thus, many different approaches have been found to result in short-term increases in children’s social engagement with others. These gains have not been limited to increased responding to others’ cues, but also involve increasing children’s social initiations with other children, sustaining interactions and conversations, and increasing positive affect, eye contact, and other nonverbal communication skills involved in successful social interactions.

REPETITIVE AND RESTRICTIVE BEHAVIORAL REPERTOIRE The behaviors that make up the restricted and repetitive behavior class of symptoms include circumscribed interests, insistence on nonfunctional rituals or routines, stereotyped motor mannerisms, and focus on sensory qualities or nonfunctional aspects of objects. Many children with autism display additional problematic behaviors that are not part of the diagnostic criteria but cause significant morbidity. Problematic behaviors can be defined as any repeated action that is disapproved of by a clear social consensus and/or impedes the learning and social opportunities of the child or other individuals (peers, family, community members, etc.). Such problematic behaviors include acts that are destructive, such as self-injury, aggression toward others, and property destruction, as well as those that are disruptive to the social environment, such as extended and boisterous tantrums or excessive, nonsensical vocalizations. Many children with autism also display a variety of additional behaviors (e.g., noncompliance, inattention, high activity level, distractibility, out-ofcontext comments, repetitive questions) that can significantly limit their participation in school and treatment programs. The class of repetitive and problematic behaviors constitutes the single greatest reason that children with autism spectrum disorders (ASD) are denied services and educational and community participation, and they are significant sources of parental stress, worry, and feelings of inadequacy and social humiliation. Indeed, many family members report that these kinds of behaviors can be the most aggravating and exasperating of all symptoms of autism (Turnbull and Ruef, 1997). These behaviors interfere with learning, social development, and quality of life, and are powerful indicators of negative prognosis. The interference caused by repetitive and problematic behaviors in the lives of many children with ASD has made suitable intervention a high priority. Over the years, many approaches have been explored, with the most efficacious being those

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 456 Wednesday, February 15, 2006 6:05 PM

456

Understanding Autism: From Basic Neuroscience to Treatment

based on the principles of instrumental (operant) learning theory and the methods of applied behavior analysis. Horner et al. (2002) provide an excellent review of this literature, summarizing published research up to the year 2000 on the treatment of repetitive and problem behaviors in autism. They listed five types of intervention relevant to the present review that have been used successfully with children on the autism spectrum during this time period: 1. Stimulus-based interventions altered events antecedent to the problematic behavior. Positive behavior support (PBS) procedures fall into this class of interventions. PBS is predicated on a model of prevention that emphasizes well-organized learning curricula and routines, positive adult–child interactions, predictable schedules, and sound environmental design (see, for example, Fox et al., 2003; Strain and Hemmeter, 1997). Use of visual schedules is another example of this type of intervention, which has been shown to increase on-task behavior and compliance (Bryan and Gast, 2000; MacDuff et al., 1993). 2. Instruction-based procedures included functional communication training (FCT) and self-management programs. In FCT, a more appropriate communicative behavior is taught in order to supplant the aberrant behavior such as reduced hand flapping and body rocking (Durand and Carr, 1987). In self-management programs, individuals are taught to monitor and provide contingent responses to their own behavior such as decreased motor stereotypies (R.L. Koegel and L.K. Koegel, 1990) and repetitive vocalizations (Mancina et al., 2000). 3. Extinction-based procedures, in which the presumed reinforcer of the repetitive or challenging behavior is withheld or the behavior is ignored. Examples include decreased self-injury (Mace et al., 1998). 4. In reinforcement-related procedures the desired behaviors are reinforced. Examples include reduced repetitive speech (Handen et al., 1984). 5. Punishment-based procedures, in which negative contingencies are delivered for engagement in the unwanted behavior. Examples include reduced self-injury (Jenson et al., 1985). Horner et al. (2002) found that the first two types of intervention have become more common in the past decade, whereas the latter three procedures were more common prior to 1990. They summarized evidence that all five types of interventions are effective in reducing unwanted behaviors. All types of repetitive and problematic behaviors were amenable to improvement, with intervention success (defined as reduction of behavior by 90% or more) found in every class of challenging behavior addressed (e.g., aggression, self-injury, motor stereotypy, tantrums, etc.). In the studies published most recently (1996–2000), the mean reduction level of challenging behaviors was 85% (median = 92.3%, mode = 100%), providing substantial evidence of the efficacy of these behavioral interventions. Horner et al. (2002) also found that the likelihood of success of an intervention was greatest in studies in which a functional assessment was conducted and the

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 457 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

457

results used to guide the intervention. Functional assessment is a process of information gathering that leads to an understanding of the relation between problematic behavior and events in the environment (O’Neill et al., 1997). Functional assessment helps identify the environmental antecedents and consequences associated with the behavior and provides hypotheses regarding the purpose (i.e., function or motivation) of the behavior, logically leading to selection of the most appropriate intervention types. The repetitive behavior that is most often seen in higher-functioning children is special or circumscribed interests. Such narrow interests, often pursued to the exclusion of other more appropriate behaviors, can substantially interfere with social functioning and academic success. Howlin (1998) describes an intervention for a child obsessed with Thomas the Tank Engine trains that included clear rules for where and when access to the special topic was permitted, replacement with alternative activities, reinforcement of alternate behaviors, and environmental changes. No published reports of the efficacy of such treatments exist. A widely recommended practice is to use circumscribed interests to the child’s advantage whenever possible. They can be powerful and efficacious reinforcers for other less desirable activities (Charlop-Christy and Haymes, 1998), can be incorporated into educational curricula to increase motivation, and can sometimes be developed into vocations (Grandin, 1995; Kanner, 1973). Executive function deficits (problems with organization, planning, self-monitoring, and flexibility) also fall within this realm of symptoms, but few treatment approaches have been developed to target such deficiencies. One study suggested that teaching children with autism to sort objects according to several different conceptual strategies increases their flexibility and ability to shift cognitive set on untrained materials (Bock, 1994), but such interventions are not widely used in the educational realm yet. A number of widely used interventions using visual systems provide excellent support for executive function problems. Though not specifically described as executive function interventions, the use of picture schedules and related visual systems for time management, work management, and self-management have been very well publicized by the TEACCH program (Mesibov et al., 1994), in data-based papers by MacDuff et al. (1993), and in curriculum materials by Hodgdon (1995). Such techniques support executive functions involving sequencing, goal-directed behavior, working memory, planning, and task completion. What conclusions can be drawn? There is substantial evidence of the efficacy of behavioral approaches for reducing stereotyped, repetitive, and challenging behaviors in autism. The vast majority of the studies have used single subjects or small sample designs. One weakness in this literature is that no group designs using random assignment to treatment and control conditions have yet been conducted. Perhaps even more significantly, none of the single-subject design investigations that have been conducted to date have conducted blind assessment of outcome. These methodological weaknesses notwithstanding, there is substantial reason for optimism for families and professionals faced with a child with severe or multiple unwanted or challenging behaviors, as a host of efficacious interventions have been developed and are described in the literature.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 458 Wednesday, February 15, 2006 6:05 PM

458

Understanding Autism: From Basic Neuroscience to Treatment

COMPREHENSIVE INTERVENTION APPROACHES FOR PRESCHOOLERS Until the late 1980s, our understanding of treatment efficacy was limited to shortterm interventions focused on manipulating particular behaviors. These early papers documented the efficacy of targeted behavioral interventions for children with autism, both for increasing skill development in a wide range of areas and for decreasing rates and severity of unwanted behaviors. However, in the late 1980s, a new approach to treatment began to be reported — comprehensive early intervention programs, similar in concept to the early intervention work developed in the 1970s and 1980s in the U.S. for children from very high-risk backgrounds and those with mental retardation and other disabilities. Such approaches involved very focused and individualized teaching activities that targeted all areas of development, and were delivered for many hours each week. Several different groups, with differing approaches, have reported positive effects of comprehensive early intervention.

THE WORK

OF IVAR

LOVAAS

AND

COLLEAGUES

Major papers by Lovaas and colleagues (Lovaas, 1987; McEachin et al., 1993) had a tremendous effect on the field of autism intervention when they reported that 9 of 19 children who received their treatment were functioning in the average range by 7 to 8 years of age, whereas almost no children in either of the two comparison groups had anywhere near that kind of outcome. These reports, made available to the public through a parent’s autobiographical description of the recovery of both of her children after this treatment (Maurice, 1993), offered new hope to families. These were the first empirical papers to report attainment of typical functioning, and it involved a treatment with a strong empirical base established in single-subject studies spanning more than 20 years. The treatment approach has now been independently examined by several groups. Smith et al. (2000) provided several methodological improvements over Lovaas’ original study, including random assignment to treatment and control groups, a uniform assessment battery delivered at both pre- and post-treatment points, a uniform posttest point, careful diagnosis of autism and differentiation between levels of severity, and objective accounting of the number of treatment hours the children experienced. Their findings replicated Lovaas’ original report of significant IQ gains of the treated group in relation to the comparison group. However, Smith’s treated group still functioned in the IQ range associated with mental retardation, unlike the original group. Only 2 of 15 children in the treated group (and 1 in the comparison group) achieved the “best outcome” status (children who were functioning at age level in all areas and in typical educational classes without support) compared to 9 of 19 in the original study. There were no significant posttreatment group differences in language, adaptive behavior, or intensity of behavior problems. Several other replications of Lovaas’ treatment have also been published in the past 5 years, using comparison groups of convenience. Eikeseth et al. (2002) reported a study of older children (ages 4 to 7) receiving Lovaas’ treatment in a school setting

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 459 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

459

rather than at home. This study is unique in that it involved a well-designed comparison treatment that controlled for intensity of the experimental treatment. The comparison treatment was eclectic, developed from the educational staff’s recommendations of best practices for the individual child, and consisted of techniques from various well-known approaches, including applied behavior analysis, sensory approaches, general special education approaches, the TEAACH approach (Mesibov et al.,1994) and other treatments. The groups did not differ significantly at either the pre- or posttest on any of the measures except for a measure of maladaptive behavior, which demonstrated fewer problem behaviors in Lovaas’ treated group. However, Lovaas’ treated group increased their scores on every measure, and they gained statistically significantly more points than the comparison group on all measures. A significantly greater proportion of children in Lovaas’ treatment attained a posttreatment IQ score in the normal range than the controls. The posttreatment IQ scores of the majority of children in Lovaas’ treatment were above 85, similar to the original 1987 study, although no children are described as “recovered,” the IQ of the group as a whole is still in the borderline range (79) and the mean language scores still were in the impaired range (58 to 67). This study demonstrates beneficial effects of Lovaas’ model delivered to somewhat older children in a school setting compared to an eclectic intervention, although both groups continued to demonstrate impairments. Recent studies have examined the delivery of Lovaas’ treatment model in community settings rather than in research programs. These can be conceived of as addressing the question of effectiveness — does the treatment work when carried out by typical community representatives as opposed to tight experimental conditions? Unfortunately, only two of these have a control condition. Sheinkopf and Siegel (1998) conducted a clinical record review study of 11 children who had received Lovaas’ treatment as delivered by community professionals. Comparison of assessment data suggested positive treatment effects of Lovaas’ intervention model seen in largely nonverbal IQ gain. However, three prospective studies (Bibby et al., 2001; Smith et al., 2000; Takeuchi et al., 2002) did not demonstrate long-term enhancement of IQ or other test scores in the majority of treated children and few attained “best outcome” status. Finally, two recent studies from applied behavior analysis reported on comprehensive programs delivering 30 or more hours a week of 1:1 treatment for young preschoolers with autism (Howard et al., 2005; Sallows and Graupner, 2005). Both sets of investigators used combinations of didactic teaching, naturalistic teaching, much social interaction and play with adults and peers over two or more years. The Sallows study used random assignment to a clinic-supervised model or a parent-supervised model, and the Howard et al. study used two community comparison groups. Both studies reported that a significant number of treated children achieved very good outcomes, including developmental scores in the nonimpaired range. While design difficulties are present in both studies, they nevertheless add to the growing evidence that it is possible for young children with autism to make excellent progress and drastically reduce their level of developmental disability. As in earlier studies, pretreatment characteristics predicted outcomes in the Sallows et al. study.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 460 Wednesday, February 15, 2006 6:05 PM

460

Understanding Autism: From Basic Neuroscience to Treatment

In summary, Lovaas’ original finding of “recovery” of almost half of a group of very young children receiving his particular treatment has thus far failed to be replicated. “Recovery” has not occurred more frequently in the treatment group than in a comparison group in any of these treatment replication studies. What has been replicated is that when Lovaas’ treatment is delivered at a high level of intensity and overseen by experts in the method with rigorous levels of training and supervision, children’s skill levels increase and significant IQ gains may occur. However, in most of the replications these gains have occurred in children who, as a group, continue to show impaired intellectual, language, social, and adaptive functioning. It appears that positive effects can be achieved with somewhat older as well as younger preschoolers, with fewer treatment hours than those used in the original study and in community settings, delivered by well-trained and supervised staff members, as well as tightly controlled research paradigms. Whether Lovaas’ approach itself, independent of intensity, is inherently more effective than other treatment approaches, and for what areas of functioning, has not yet been demonstrated. However, no other name brand treatment has yet been subjected to the rigorous examination that Lovaas’ treatment has.

DEVELOPMENTALLY ORIENTED TREATMENTS These are defined here as those that use typical developmental sequences as the content of their interventions and developmental theory as the guiding principle of their approach. Although developmental treatments for autism have not been studied as rigorously as behaviorally defined treatments, empirical support is beginning to accumulate. Two randomized controlled trials (RCTs) have been reported in the past few years. Jocelyn and colleagues (1998) carried out an RCT of a 12-week developmental or behavioral treatment conducted in inclusive community day care centers by day care aides trained for this project. A total of 35 children (mean age 44 months) who had not received services or childcare before were randomized into treatment and community conditions. Although both groups made developmental progress, the experimental group showed a significant gain in only one area, language development, when compared to controls. There was no change in professional rating of autism severity, and no treatment effect on parent rating of autism severity in either group. Although both groups of parents reported satisfaction with their child’s care, the experimental parents reported significantly greater gains in knowledge of autism and in understanding and meeting their child’s needs. Thus, gains were modest, but the treatment period was very short and carried out by paraprofessionals in an inclusive setting. Drew and colleagues (2002) also provided an RCT that tested the effects of a low-intensity home-based, parent-delivered developmental intervention. The pilot study involved 24 toddlers with autism. The experimental condition involved training the parents in two main areas: the pragmatics of social communication and of behavior management. A speech pathologist visited each home every six weeks for a 3-hour visit, reviewed the child’s progress, taught the parents new facilitation skills, and set goals for the next six weeks. Parents were explicitly taught to carry out joint attention, play, imitation, and turn-taking and to stimulate nonverbal communication in their interactions with their children. Follow-up after 12 months revealed very

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 461 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

461

few group differences. The treated group had fewer nonverbal children than the control group, and there was a trend in the treated group to understand more words. Unfortunately, however, the treated group had a significantly higher nonverbal IQ than the controls at the start of the study (88 vs. 66), which may well have been associated with their greater language gain. The treated group lost 11 IQ points on the nonverbal measure over the course of the treatment, whereas the IQ scores of the controls remained stable. Thus, the evidence does not provide much support for the efficacy of this intervention. The TEACCH approach, classified as a developmental model by its originators, emphasizes a teaching approach that teaches to the strengths of children with autism in visual spatial understanding, object manipulation, and enjoyment of highly structured, independent, and routine activities. Ozonoff and Cathcart (1998) examined the efficacy of a parent-delivered, daily, 1-hour, 1:1 TEACCH intervention for a group of 11 preschoolers with autism already receiving full-day school-based behavioral education. After approximately 16 weeks of treatment, the children who received the TEAACH intervention had made significant gains in a number of developmental areas as compared to 11 well-matched waiting-list controls. In addition to significantly improved overall scores (54) on a developmental test (Schopler et al., 1990), the TEACCH group had significantly improved imitation, nonverbal perception, cognition, and fine and gross motor skills, having made three to four times more progress than the control group, who was getting full-day intensive behavioral education but no TEAACH treatment. The Denver Model emphasizes development of play skills, positive affect, interpersonal relationships, and language development. Outcomes were reported for a large group of children attending a 5-h/d specialized group classroom with six children and three teaching staff per classroom with additional speech or language therapy, occupational therapy, and clinical psychology therapy services. Rogers and colleagues (1986, 1987, 1989) reported that, over a 6- to 12-month intervention period, the group demonstrated statistically significant changes above and beyond the gains expected based on their initial developmental rates in most areas of development, including receptive and expressive language, symbolic play, and responsivity and positive affect in dyadic interactions with their parents. Their developmental rates doubled during the treatment period and attained the normal rate of 1 month of growth per month of treatment, which was sustained for as long as they were enrolled in the program. Rogers and DiLalla (1991) demonstrated that children with more severe delays achieved the same normal developmental rate during intervention as the less severely delayed children and that children with autism progressed at a similar rate as a nonautistic delayed group also enrolled in the treatment program. These findings were replicated by independent groups in several other sites (Rogers et al.,1987). Another early treatment approach in autism that demonstrates a convergence of developmental and behavioral foundations has been the peer-mediated learning approach developed by Phillip Strain, Samuel Odom, and colleagues. Although these researchers have published many single-subject design papers, the approach was developed for a more long-term integrated classroom design known as the LEAP (Learning Experiences: An Alternative Program for Preschoolers and Parents) preschool model.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 462 Wednesday, February 15, 2006 6:05 PM

462

Understanding Autism: From Basic Neuroscience to Treatment

Using a design similar to that of the Denver Model studies, these authors demonstrated accelerations in developmental rates of six treated children to normal levels and significant gains in virtually all areas of development (Hoyson et al., 1984), as well as positive long-term follow-up results, with almost 50% of a larger group of treated children in regular school classes (Strain and Hoyson, 2000). A recent short-term treatment study combining developmental and behavioral approaches to teach joint attention behavior and symbolic play was reported by Kasari and colleagues (2001 and in press). In this three-group RCT, children receiving either joint attention or symbolic play treatment made both specific and general gains in these key areas and generalized them to play at home with parents, compared to a control group not receiving either treatment. Salt et al. (2002) reported a 10-month treatment for a small group of British children with autism (n = 12) who received 8 hr/week of a special developmentally based intervention in addition to their nursery school and other treatments (which occurred at a mean of about 15 h/week). The intervention focused on areas known to be specifically affected in early autism: imitation, joint attention, language, social reciprocity, and play, delivered in a naturalistic child-centered manner. Parents and children attended a small group program for eight 2-hour sessions a month for parent training, and parents delivered additional hours at home. The treated group performed significantly better than the comparison group on several measures: all of the Vineland scales except communication, the imitation measure, joint attention, and social interaction. However, there were no significant group differences on the main language measure or play measure and no indication of acceleration in developmental rates on the Vineland scales. Rather, the experimental group tended to maintain their initial scores, whereas the comparison group’s scores dropped in some areas. The final developmentally based program to be considered is Greenspan and colleagues’ model, which has a foundational type of social communication intervention — “Floortime” and additional professional services (Greenspan et al., 1997). Children received many hours a week of Floortime from their parents, and also speech or language and occupational therapy (OT) therapies weekly. A chart review study of a large group of children demonstrated positive social and developmental gains, although a standard assessment battery was not included in this study (Greenspan and Weider, 1997).

NEW APPROACHES Several new intervention models that address comprehensive developmental needs of young children with autism have appeared in the literature in the past 5 years. Each of these approaches uses a naturalistic or relationship-oriented style of adult interaction. Mahoney and Perales (2003, 2005) describe a developmentally oriented intervention for toddlers with autism built from research in the field of mental retardation over the past 30 years. The intervention focuses on training parents in the kinds of responsive or interactive strategies that have been found to be effective in fostering development in children with other kinds of developmental difficulties. McGee and colleagues (1999) have provided one of the few descriptions of a group approach for toddlers with autism using a full-day inclusive childcare environment, in © 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 463 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

463

which the majority of children have typical development (12 children, 4 of whom have autism). Two main teaching procedures are used: “incidental teaching” and “wait-ask-say-show-do” involving backward chaining and faded guidance for teaching language and independence in center routines and daily living skills. Finally, the use of naturalistic language paradigms is currently being expanded considerably through the work of language scientists such as Ann Kaiser, Steve Warren, and Paul Yoder, who are fusing the sciences of communication, behavior, and autism. Their current intervention studies may be seen as state of the art and best practice in the near future (Hancock and Kaiser, 2002; Yoder and Warren, 2001).

CONCLUSIONS From this chapter, three main points seem the most pertinent. First of all, children with autism respond to well-designed short-term intervention approaches in all the areas of deficits associated with autism. Ability to improve functioning does not seem dependent upon a certain level of intellectual ability or the severity of the autism-specific symptoms. The most profoundly affected children often demonstrate measurable positive changes in short-term interventions that provide teaching for as little as a few hours per week. Furthermore, the design of these effective interventions follows general principles of learning and behavior change. There is every reason to expect that the most limiting symptoms of those children can be successfully treated given high-quality intervention design, implementation, and ongoing evaluation of progress. However, the qualifier “high-quality” needs to be taken seriously. The successful treatment approaches reported in the literature come from experts in applied behavior analysis and early intervention. Successful treatment of symptoms in children with autism may require far more expertise than is provided in general school districts and developmental disability agencies. Second, the degree of change that is possible for children with autism being given optimal treatment is not yet known. It appears that the developmental delays associated with autism can be reduced, at least in some areas, by specific intervention approaches, and Lovaas’ approach meets current criteria such as “probably efficacious” for symptom reduction. “Recovery,” as defined by test scores in the normal range, regular successful school placement and performance, and lack of functional disability, occurs occasionally, both in intensively treated children and in comparison children. However, Lovaas’ initial findings of recovery in almost half of a treated group have not been replicated. At this point, the research data do not indicate that any particular treatment leads to recovery. Third, we do not have data to indicate whether one treatment approach is more effective than others. In terms of the comprehensive preschool approaches, we do not have a single comparative RCT using well-defined and empirically supported treatments. In the single-subject designs, there are few comparative treatment studies. An exception to this is in the language domain, where comparisons of discrete trial teaching and naturalistic teaching have demonstrated some superiority of naturalistic approaches in terms of spontaneity, generalization, and maintenance. However, comparative data that might answer questions about long-term superiority of one vs. the other approach are lacking. The question is probably not which intervention approach is © 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 464 Wednesday, February 15, 2006 6:05 PM

464

Understanding Autism: From Basic Neuroscience to Treatment

best, but rather which intervention approach is best for a particular child and a particular treatment target. Treatment by aptitude interactions are virtually unstudied in the autism treatment research literature, although Laura Schreibman and colleagues have begun to address this question in their work on predictive profiles for PRT responders (Sherer and Schreibman, 2005). The kind of treatment research needed in autism is perhaps unique, given the enormous scale of interventions needed. Treatment research in other conditions focuses on specific symptoms, such as noncompliance, or disorders with a more limited set of symptoms than autism. Autism treatment needs to address every developmental area, all areas of adaptive behavior, and then a whole set of aberrant behavioral responses, involving both positive and negative symptoms. Even treatment for schizophrenia or alcoholism, although required to address multiple aspects of behavior, does not face the need to target every aspect of a person’s life virtually from infancy onward. Interventions for disorders of similar severity, such as addictions and schizophrenia, are often delivered in a protected and restricted setting, unlike autism interventions, which require the least restrictive environments. (All such severe disorders probably require supports and modifications in everyday environments in order to have continued growth and improved functioning while preventing relapse or regression.) Thus, developing treatment packages and manuals that cover all of development, behavior, and relationships, delivered across all environments and for substantial periods of time, is a huge task, and one for which we lack models in other fields. As can be seen from the studies reviewed in this chapter, we are making progress, but the task is large and the obstacles are many.

ACKNOWLEDGMENTS Sally Rogers’ contribution was partially supported by the following grants: U19 HD35468-07 from the National Institute of Child Health and Human Development, R21 DC05574-03 from the National Institute of Deafness and Communication Disorders, and R13 MH70772-01, R21 MH0673631, R01 MH068398-02, and RO1 MH068232-01 from the National Institute of Mental Health. Sally Ozonoff’s contribution was supported in part by U19 HD35468-07 from the National Institute of Child Health and Human Development and R01 MH068398-02 from the National Institute of Mental Health.

REFERENCES Baker, M.J., Koegel, R.L., and Koegel, L.K. (1998), Increasing the social behavior of young children with autism using their obsessive behaviors, The Journal of the Association for Persons with Severe Handicaps, 23, 300–308. Bauminger, N. (2002), The facilitation of social-emotional understanding and social interaction in high-functioning children with autism: Intervention outcomes, Journal of Autism and Developmental Disorders, 32, 283–298. Bernard-Opitz, V., Sriram, N., and Nakhoda-Sapuan, S. (2001), Enhancing social problem solving in children with autism and normal children through computer-assisted instruction, Journal of Autism and Developmental Disorders, 31, 377–384.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 465 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

465

Bibby, P., Eikeseth, S., Martin, N.T., Mudford, O.C., and Reeves, D. (2001), Progress and outcomes for children with autism receiving parent-managed intensive interventions, Research in Developmental Disabilities, 22, 425–447. Blew, P.A., Schwartz, I.S., and Luce, S.C. (1985), Teaching functional community skills to autistic children using nonhandicapped peer tutors, Journal of Applied Behavior Analysis, 18, 337–342. Bock, M.A. (1994), Acquisition, maintenance, and generalization of a categorization strategy by children with autism, Journal of Autism and Developmental Disorders, 24, 39–51. Bondy, A.S. and Frost, L.A. (1994), The picture exchange communication system, Focus on Autistic Behavior, 9, 1–19. Bryan, L.C. and Gast, D.L. (2000), Teaching on-task and on-schedule behaviors to highfunctioning children with autism via picture activity schedules, Journal of Autism and Developmental Disorders, 30, 553–567. Carter, C.M. (2001), Using choice with game play to increase language skills and interactive behaviors in children with autism, Journal of Positive Behavior Interventions, 3, 131–151. Chandler, S., Christie, P., Newson, E., and Prevezer, W. (2002), Developing a diagnostic and intervention package for 2- to 3-year-olds with autism: outcomes of the frameworks for communication approach. Autism, 6, 47–69. Charlop-Christy, M.H., Carpenter, M., Le, L., LeBlanc, L.A., and Kellet, K. (2002), Using the picture exchange communication system (PECS) with children with autism: assessment of PECS acquisition, speech, social-communicative behavior, and problem behavior, Journal of Applied Behavior Analysis, 35, 213–231. Charlop-Christy, M.H. and Haymes, L.K. (1998), Using objects of obsession as token reinforcers for children with autism. Journal of Autism and Developmental Disorders, 28, 189–197. Charlop-Christy, M.H., Le, L., and Freeman, K.A. (2000), A comparison of video modeling with in vivo modeling for teaching children with autism, Journal of Autism and Developmental Disorders, 30, 537–552. Charlop, M.H. and Walsh, M.E. (1986), Increasing autistic children's spontaneous verbalizations of affection: an assessment of time delay and peer modeling procedures, Journal of Applied Behavior Analysis, 19, 307–314. Coe, D., Matson, J., Fee, V., Manikam, R., and Linarello, C. (1990), Training nonverbal and verbal play skills to mentally retarded and autistic children, Journal of Autism and Developmental Disorders, 20, 177–187. Dawson, G. and Galpert, L. (1990), Mothers’ uses of imitative play for facilitating social responsiveness and toy play in young autistic children, Development and Psychopathology, 2, 151–162. Delprato, D.J. (2001), Comparisons of discrete-trial and normalized behavioral language intervention for young children with autism, Journal of Autism and Developmental Disorders, 31, 315–325. Dewey, D., Lord, C., and Magill, J. (1988), Qualitative assessment of the effect of play materials in dyadic peer interactions of children with autism, Canadian Journal of Psychology, 42, 242–260. Drew, A., Baird, G., Baron-Cohen, S., Cox, A., Slonims, V., Wheelwright, S. et al. (2002), A pilot randomized control trial of a parent training intervention for pre-school children with autism: preliminary findings and methodological challenges, European Child and Adolescent Psychiatry, 11, 266–272. Dugan, E.P., Kamps, D.M., Leonard, B.R., Watkins, N., Rheinberger, A., and Stackhaus, J. (1995), Effects of cooperative learning groups during social studies for students with autism and fourth-grade peers, Journal of Applied Behavior Analysis, 28, 175–188.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 466 Wednesday, February 15, 2006 6:05 PM

466

Understanding Autism: From Basic Neuroscience to Treatment

Durand, V.M. and Carr, E.G. (1987), Social influences on “self-stimulatory” behavior: analysis and treatment application, Journal of Applied Behavior Analysis, 20, 119–132. Eikeseth, S., Smith, T., Jahr, E., and Eldevik, S. (2002), Intensive behavioral treatment at school for 4- to 7-year-old children with autism: a 1-year comparison controlled study, Behavior Modification, 26, 49–68. Field, T., Field, T., Sanders, C., and Nadel, J. (2001), Children with autism display more social behaviors after repeated imitation sessions, Autism, 5, 317–325. Fox, L., Dunlap, G., Hemmeter, M.L., Joseph, G.E., and Strain, P.S. (2003), The teaching pyramid: a model for supporting social competence and preventing challenging behavior in young children, Young Children, 48–52. Goldstein, H. (2000), Commentary: interventions to facilitate auditory, visual, and motor integration: “show me the data.” Journal of Autism and Developmental Disorders, 30, 423–426. Goldstein, H., Kaczmarek, L., Pennington, R., and Shafer, K. (1992), Peer-mediated intervention: attending to, commenting on, and acknowledging the behavior of preschoolers with autism, Journal of Applied Behavior Analysis, 25, 289–305. Goldstein, H., Wickstrom, S., Hoyson, M., Jamieson, B., and Odom, S.L. (1988), Effects of sociodramatic play training on social and communicative interaction, Education and Treatment of Children, 11, 97–117. Grandin, T. (1995), Thinking in Pictures, New York: Vintage Books. Gray, C. (2000), The New Social Story Book: Illustrated Edition. Arlington, TX: Future Horizons. Greenspan, S.I., Kalmanson, B., Shahmoon-Shanok, R., Wieder, S., Gordon-Williamson, G., and Anzalone, M. (1997), Assessing and Treating Infants and Young Children with Severe Difficulties in Relating and Communicating, Washington, D.C.: Zero to Three. Greenspan, S.I. and Weider, S. (1997), Developmental patterns and outcomes in infants and children with disorders in relating and communicating: A chart review of 200 cases of children with autistic spectrum disorders, Journal of Developmental and Learning Disorders, 1, 87–141. Hadwin, J., Baron-Cohen, S., Howlin, P., and Hill, K. (1996), Can we teach children with autism to understand emotions, belief, or pretense? Development and Psychopathology, 8, 345–365. Hagiwara, R. and Myles, B. S. (2001), A multimedia social story intervention: teaching skills to children with autism. Focus on Autism and other Developmental Disabilities, 14, 82–95. Hancock, T.B. and Kaiser, A.P. (2002), The effects of trainer-implemented enhanced milieu teaching on the social communication of children with autism, Topics in Early Childhood Special Education, 22, 29–54. Handen, B.L., Apolito, P.M., and Seltzer, G.B. (1984), Use of differential reinforcement of low rates of behavior to decrease repetitive speech in an autistic adolescent, Journal of Behavior Therapy and Experimental Psychiatry, 15, 359–364. Harris, S.L., Wolchik, S.A., and Milch, R.E. (1983), Changes the speech of autistic children and their parents, Child and Family Behavior Therapy, 4, 151–173. Harris, S.L., Wolchik, S.A., and Weitz, S. (1981), The acquisition of language skills by autistic children: can parents do the job?, Journal of Autism and Developmental Disorders, 11, 373–384. Hewett, F.M. (1965), Teaching speech to an autistic child through operant conditioning, American Journal of Orthopsychiatry, 35, 927–936. Hodgdon, L.A. (1995), Visual strategies for Improving Communication, Troy, MI: Quirk Roberts Publishing.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 467 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

467

Horner, R.H., Carr, E.G., Strain, P.S., Todd, A.W., and Reed, H.K. (2002), Problem behavior interventions for young children with autism: a research synthesis, Journal of Autism and Developmental Disorders, 32, 423–446. Howard, J.S., Sparkman, C.R., Cohen, H.G., Green, G., and Stanislaw, H. (2005). A comparison of intensive behavior analytic and eclectic treatments for young children with autism, Research in Developmental Disabilities, 26, 359–383. Howlin, P. (1998), Psychological and educational treatments for autism, Journal of Child Psychology and Psychiatry, 39, 307–322. Howlin, P. and Rutter, M. (1989), Mothers’ speech to autistic children: a preliminary causal analysis, Journal of Child Psychology and Psychiatry, 30, 819–843. Howlin, P. and Yates P. (1999), The potential effectiveness of social skills groups for adults with autism, Autism, 3, 299–307. Hoyson, M., Jamieson, B., and Strain, P.S. (1984), Individualized group instruction of normally developing and autistic-like children: The LEAP curriculum model, Journal of the Division of Early Childhood, 8, 157–172. Jenson, W.R., Rovner, L., Cameron, S., Petersen, B.P., and Kesler, J. (1985), Reduction of self-injurious behavior in an autistic girl using a multifaceted treatment program, Journal of Behavior Therapy and Experimental Psychiatry, 16, 77–80. Jocelyn, L.J., Casiro, O.G., Beattie, D., Bow, J., and Kneisz, J. (1998). Treatment of children with autism: a randomized controlled trial to evaluate a caregiver-based intervention program in community day-care centers, Developmental and Behavioral Pediatrics, 19, 326–334. Kaiser, A., Hancock, T.B., and Nietfeld, J.P. (2000), The effect of parent-implemented Enhanced Milieu Teaching on the social communication of children who have autism, Early Education and Development, 11, 423–446. Kamps, D.M, Barbetta, P.M., Leonard, B.R., and Delquadri, J. (1994), Classwide peer tutoring: An integration strategy to improve reading skills and promote peer interactions among students with autism and general education peers, Journal of Applied Behavior Analysis, 27, 49–61. Kamps, D.M., Leonard, B.R., Vernon, S., Dugan, E.P., and Delquadri, J.C. (1992), Teaching social skills to students with autism to increase peer interactions in an integrated firstgrade classroom, Journal of Applied Behavior Analysis, 25, 281–288. Kamps, D.M., Potucek, M.A., Lopez, A.G., Kravits, T., and Kemmerer, K. (1997), The use of peer networks across multiple settings to improve social interaction for students with autism, Journal of Behavioral Education, 7, 335–357. Kanner, L. (1943), Autistic disturbances of affective contact, Nervous Child, 2, 217–250. Kanner, L. (1971), Follow-up study of eleven autistic children originally reported in 1943, Journal of Autism and Childhood Schizophrenia, 1, 119–145. Kanner, L. (1973), Childhood Psychosis: Initial Studies and New Insights, New York: John Wiley & Sons. Kasari, C., Freeman, S.F., and Paparella, T. (2001), Early intervention in autism: Joint attention and symbolic play, International Review of Research in Mental Retardation, 23, 207–237. Kasari, C., Freeman, S., and Paparella, T. (in press), Joint attention and symbolic play in young children with autism: a randomized controlled intervention study. Journal of Child Psychology and Psychiatry. Koegel, L.K. (2000), Interventions to facilitate communication in autism, Journal of Autism and Developmental Disorders, 30, 383–391. Koegel, R.L., Camarata, S., Koegel, L.K., Ben-Tall, A., and Smith, A.E. (1998), Increasing speech intelligibility in children with autism, Journal of Autism and Developmental Disorders, 28, 241–251.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 468 Wednesday, February 15, 2006 6:05 PM

468

Understanding Autism: From Basic Neuroscience to Treatment

Koegel, L.K., Camarata, S.M., Valdez-Menchaca, M., and Koegel, R.L. (1998), Setting generalization of questing-asking by children with autism, American Journal of Mental Retardation, 102, 346–357. Koegel, R.L. and Frea, W.D. (1993), Treatment of social behavior in autism through the modification of pivotal social skills, Journal of Applied Behavior Analyses, 26, 369–377. Koegel, L.K., Koegel, R.L., Hurley, C., and Frea, W.D. (1992), Improving social skills and disruptive behavior in children with autism through self-management, Journal of Applied Behavior Analysis, 25, 341–353. Koegel, R.L. and Koegel, L.K. (1990), Extended reductions in stereotypic behavior of students with autism through a self-management treatment package, Journal of Applied Behavior Analysis, 23, 119–128. Koegel, R.L., Koegel, L.K., and Surratt, A. (1992), Language intervention and disruptive behavior in preschool children with autism, Journal of Autism and Developmental Disorders, 22, 141–153. Koegel, R.L., O’Dell, M.C., and Koegel, L.K. (1987), A natural language teaching paradigm for nonverbal autistic children, Journal of Autism and Developmental Disorders, 17, 187–200. Krantz, P.J. and McClannahan, L.E. (1993), Teaching children with autism to initiate to peers: effects of a script-fading procedure, Journal of Applied Behavior Analysis, 26, 121–132. Krantz, P.J. and McClannahan, L.E. (1998), Social interaction skills for children with autism: a script-fading procedure for beginning readers, Journal of Applied Behavior Analysis, 31, 191–202. Krantz, P.J., Zalenski, S., Hall, L.J., Fenske, E.C., and McClannahan, L.E. (1981), Teaching complex language to autistic children, Analysis and Intervention in Developmental Disabilities, 1, 259–297. Krasny, L., Williams, B.J., Provencal, S., and Ozonoff, S. (2003), Social skills interventions for the autism spectrum: essential ingredients and a model curriculum, Child and Adolescent Psychiatric Clinics of North America, 12, 107–122. Kuttler, S., Myles, B.S., and Carlson, J.K. (1998), The use of social stories to reduce precursors to tantrum behavior in a student with autism, Focus on Autism and other Developmental Disabilities, 13, 176–182. Laushey, K.M. and Heflin, L.J. (2000), Enhancing social skills of kindergarten children with autism through the training of multiple peers as tutors, Journal of Autism and Developmental Disorders, 30, 183–193. LeBlanc, L.A., Coates, A.M., Daneshvar, S., Charlop-Christy, M.H., Morris, C., and Lancaster, B.M. (2003), Using video modeling and reinforcement to teach perspectivetaking skills to children with autism, Journal of Applied Behavior Analysis, 36, 253–257. Lord, C. and Hopkins, J.M. (1986), The social behavior of autistic children with younger and same-age nonhandicapped peers, Journal of Autism and Developmental Disorders, 16, 249–262. Lovaas, I.O. (1981), Teaching Developmentally Disabled Children: The Me Book. Baltimore, MD: University Park Press. Lovaas, I.O. (1987), Behavioral treatment and normal educational and intellectual functioning in young autistic children, Journal of Consulting and Clinical Psychology, 55, 3–9. Lovaas, I.O., Berberich, J.P., Perloff, B.F., and Schaeffer, B. (1966), Acquisition of imitative speech by schizophrenic children, Science, 151, 705–707. Lovaas, I.O., Koegel, R.L., Simmons, J.Q., and Long, J.S. (1973), Some generalization and follow-up measures on autistic children in behavior therapy, Journal of Applied Behavior Analysis, 6, 131–166.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 469 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

469

MacDuff, G.S., Krantz, P.J., and McClannahan, L.E. (1993), Teaching children with autism to use photographic activity schedules: maintenance and generalization of complex response chains, Journal of Applied Behavior Analysis, 26, 89–97. Mace, A.B., Shapiro, E.S., and Mace, F.C. (1998), Effects of warning stimuli for reinforcer withdrawal and task onset on self-injury, Journal of Applied Behavior Analysis, 31, 679–682. Mahoney, G. and Perales, F. (2003), Using relationship-focused intervention to enhance the social-emotional functioning of young children with autism spectrum disorders, Topics in Early Childhood Special Education, 23, 77–89. Mahoney, G. and Perales, F. (2005), The impact of relationship focused intervention on young children with autism spectrum disorders: a comparative study, Journal of Developmental and Behavioral Pediatrics, 26 (2), 77–85. Mancina, C., Tankersley, M., Kamps, D., Kravits, T., and Parrett, J. (2000), Reduction of inappropriate vocalizations for a child with autism using a self-management treatment program, Journal of Autism and Developmental Disorders, 30, 599–606. Marriage, K.J., Gordon, V., and Brand, L. (1995), A social skills group for boys with Asperger’s syndrome, Australian and New Zealand Journal of Psychiatry, 29, 58–62. Mason, S.A. and Newsom, C.D. (1990), The application of sensory change to reduce stereotyped behavior, Research in Developmental Disabilities, 11, 257–271. Maurice, C. (1993), Let Me Hear Your Voice: A Family's Triumph over Autism, New York: Knopf. McBride, B.J. and Schwartz, I.S. (2003), Effects of teaching early interventionist to use discrete trials during ongoing classroom activities, Topics in Early Childhood Special Education, 23, 5–17. McEachin, J.J., Smith, T., and Lovaas, I.O. (1993), Long-term outcome for children with autism who received early intensive behavioral treatment, American Journal on Mental Retardation, 97, 359–372. McGee, G.G., Almeida, M.C., Sulzer-Azaroff, B., and Feldman, R.S. (1992), Promoting reciprocal interactions via peer incidental teaching, Journal of Applied Behavior Analysis, 25, 117–126. McGee, G.G., Krantz, P.J., Mason, D., and McClannahan, L.E. (1983), A modified incidentalteaching procedure for autistic youth: acquisition and generalization of receptive object labels, Journal of Applied Behavior Analysis, 16, 329–338. McGee, G.G., Krantz, P.J., and McClannahan, L.E. (1999), The facilitative effects of incidental teaching on prepositional use by autistic children, Journal of Applied Behavior Analysis, 18, 17–31. McGee, G.G., Morrier, M.J., and Daly, T. (1999), An incidental teaching approach to early intervention for toddlers with autism, Journal of the Association for Persons with Severe Handicaps, 24, 133–146. Mesibov, G.B. (1984), Social skills training with verbal autistic adolescents and adults: a program model, Journal of Autism and Developmental Disorders, 14, 395–404. Mesibov, G., Schopler, E., and Hearsey, K.A. (1994), Structured teaching, in Schopler, E. and Mesibov, G. (Eds.), Behavioral Issues in Autism, New York: Plenum Press, pp. 195–207. Mundy, P. and Crowson, M. (1997), Joint attention and early social communication; Implications for research on intervention with autism, Journal of Autism and Developmental Disorders, 27, 653–676. Mundy, P. and Markus, J. (1997), On the nature of communication and language impairment in autism, Mental Retardation and Developmental Disabilities Research Reviews, 3, 343–349.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 470 Wednesday, February 15, 2006 6:05 PM

470

Understanding Autism: From Basic Neuroscience to Treatment

Mundy, P., Sigman, M., and Kasari, C. (1990), A longitudinal study of joint attention and language development in autistic children, Journal of Autism and Developmental Disorders, 20, 115–128. Neef, N.A., Walters, J., and Egel, A.L. (1984), Establishing generative yes/no responses in developmentally disabled children, Journal of Applied Behavior Analysis, 17, 453–460. Odom, S.L., McConnell, S.R., McEvoy, M.A., Peterson, C., Ostrosky, M., Chandler, L.K. et al. (1999), Relative effects of interventions for supporting the social competence of young children with disabilities, Topics in Early Childhood Special Education, 19, 75–92. Odom, S.L. and Strain, P.S. (1986), A comparison of peer-initiation and teacher-antecedent interventions for promoting reciprocal social interaction of autistic preschoolers, Journal of Applied Behavior Analysis, 19, 59–71. O’Neill, R.E., Horner, R.H., Albin, R.W., Storey, K., Sprague, J.R., and Newton, J.S. (1997), Functional Assessment of Problem Behavior: A Practical Assessment Guide, Pacific Grove, CA: Brooks/Cole. Ozonoff, S. and Cathcart, K. (1998), Effectiveness of a home program intervention for young children with autism, Journal of Autism and Developmental Disorders, 28, 25–32. Ozonoff, S. and Miller, J.N. (1995), Teaching theory of mind: a new approach to social skills training for individuals with autism, Journal of Autism and Developmental Disorders, 25, 415–433. Peck, C.A. (1985), Increasing opportunities for social control by children with autism and severe handicaps: effects on student behavior and perceived classroom climate, The Journal of The Association for Persons with Severe Handicaps, 10, 183–193. Pierce, K. and Schreibman, L. (1997), Multiple peer use of pivotal response training to increase social behaviors of classmates with autism: Results from trained and untrained peers, Journal of Applied Behavior Analysis, 30, 157–160. Prizant, B.M., Wetherby, A.M., and Rydell, P. (2000), Communication intervention issues for children with autism spectrum disorders, in Wetherby, A.M. and Prizant, B.M. (Eds.), Autism Spectrum Disorders: A Transactional Developmental Perspective, Baltimore, MD: Paul H. Brookes, pp. 193–224. Reinecke, D.R., Newman, B., Kurtz, A.L., Ryan, C.S., and Hemmes, N.S. (1997), Teaching deception skills in a game-play context to three adolescents with autism, Journal of Autism and Developmental Disorders, 27, 127–137. Roeyers, H. (1996), The influence of nonhandicapped peers on the social interactions of children with a pervasive developmental disorder, Journal of Autism and Developmental Disorders, 26, 303–320. Rogers, S.J. and DiLalla, D. (1991), A comparative study of the effects of a developmentally based instructional model on young children with autism and young children with other disorders of behavior and development, Topics in Early Childhood Special Education, 11, 29–48. Rogers, S.J., Herbison, J., Lewis, H., Pantone, J., and Reis, K. (1986), An approach for enhancing the symbolic, communicative, and interpersonal functioning of young children with autism and severe emotional handicaps, Journal of the Division of Early Childhood, 10, 135–148. Rogers, S.J. and Lewis, H. (1989), An effective day treatment model for young children with pervasive developmental disorders, Journal of the American Academy of Child and Adolescent Psychiatry, 28, 207–214. Rogers, S.J., Lewis, H.C., and Reis, K. (1987), An effective procedure for training early special education teams to implement a model program, Journal of the Division of Early Childhood, 11, 180–188.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 471 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

471

Ross, D.E. and Greer, R.D. (2003), Generalized imitation and the mind: inducing first instances of speech in young children with autism, Research in Developmental Disabilities, 24, 58–74. Sallows, G.O. and Graupner, T.D. (2005). Intensive behavioral treatment for children with autism: four-year outcome and predictors, America Journal on Mental Retardation, 110(6) 417–438. Salt, J., Shemilt, J., Sellars, V., Boyd, S., Coulson, T., and McCool, S. (2002), The Scottish Centre for Autism preschool treatment programme. II: The results of a controlled treatment outcome study, Autism, 6, 33–46. Schopler, E., Reichler, R.J., Bashford, A., Lansing, M.D., and Marcus, L.M. (1990), The Psychoeducational Profile Revised (PEP-R), Austin, TX: Pro-Ed. Schreibman (1988), Autism, Newbury Park, CA: Sage. Schreibman, L. and Pierce, K.L. (1993), Achieving greater generalization of treatment effects in children with autism: Pivotal response training and self-management, The Clinical Psychologist, 11, 184–191. Shafer, M.S., Egel, A.L., and Neef, N.A. (1984), Training mildly handicapped peers to facilitate changes in the social interaction skills of autistic children, Journal of Applied Behavior Analysis, 17, 461–476. Sheinkopf, S.J. and Siegel, B. (1998). Home based behavioral treatment of young autistic children. Journal of Autism and Developmental Disorders, 28, 15–24. Sherer, M. and Schreibman, L. (2005), Individual behavioral profiles and predictors of treatment effectiveness for children with autism, Journal of Consulting and Clinical Psychology, 73, 1–14. Simpson, R.L., Myles, B.S., Sasso, G.M., and Kamps, D.M. (1997), Social Skills for Students with Autism, 2nd ed., Reston, VA: Council for Exceptional Children. Silver, M. and Oakes, P. (2001), Evaluation of a new computer intervention to teach people with autism or Asperger syndrome to recognize and predict emotions in others, Autism, 5, 299–316. Skinner, B.F. (1957), Verbal Behavior, Englewood Cliffs, NJ: Prentice Hall. Smith, T. (2001), Discrete trial training in the treatment of autism, Focus on Autism and Other Developmental Disabilities, 16, 86–92. Smith, T., Buch, G.A., and Gamby, T.E. (2000), Parent-directed, intensive early intervention for children with pervasive developmental disorder, Research in Developmental Disabilities, 21, 297–309. Smith, T., Groen, A.D., and Wynn, J.W. (2000), Randomized trial of intensive early intervention for children with pervasive developmental disorder, American Journal on Mental Retardation, 105, 269–285. Smith, T., Groen, A.D., and Wynn, J.W. (2001), Erratum, American Journal on Mental Retardation, 106, 208. Solomon, M., Goodlin-Jones, B.L., and Anders, T.F. (2004), A social adjustment enhancement intervention for high functioning autism, Asperger’s syndrome, and pervasive developmental disorder NOS, Journal of Autism and Developmental Disorders, 34, 649–668. Stahmer, A.C. (1995), Teaching symbolic play skills to children with autism using pivotal response training, Journal of Autism and Developmental Disorders, 25, 123–142. Strain, P.S. and Hemmeter, M.L. (1997), Keys to being successful when confronted with challenging behavior, Young Exceptional Children 1, 2–9. Strain, P.S. and Hoyson, M. (2000), The need for longitudinal, intensive social skill intervention: leap follow-up outcomes for children with autism, Topics in Early Childhood Special Education, 20, 116–122.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 472 Wednesday, February 15, 2006 6:05 PM

472

Understanding Autism: From Basic Neuroscience to Treatment

Strain, P.S., Kerr, M.M., and Ragland, E.U. (1979), Effects of peer-mediated social initiations and prompting/reinforcement procedures on the social behavior of autistic children, Journal of Autism and Developmental Disorders, 9, 41–54. Strain, P.S., McGee, G.G., and Kohler, F.W. (2001), Inclusion of children with autism in early intervention environments, in Guralnick, M.J., Ed., Early Childhood Inclusion: Focus on Change, Baltimore, MD: Paul H. Brookes, pp. 337–363. Strain, P.S., Shores, R.E., and Timm, M.A. (1977), Effects of peer social initiations on the behavior of withdrawn preschool children, Journal of Applied Behavior Analysis, 10, 289–298. Sundberg, M.L. and Michael, J. (2001), The benefits of Skinner’s analysis of verbal behavior for children with autism, Behavior Modification Special, 25, 698–724. Sussman, F. (1999), More than Words, Toronto: The Hanen Centre. Swaggart, B.L., Gagnon, E., Bock, S.J., Earles, T.L., Quinn, C., Myles, B.S. et al. (1995), Using social stories to teach social and behavioral skills to children with autism, Focus on Autistic Behavior, 10, 1–16. Swettenham, J. (1996), Can children with autism be taught to understand false belief using computers? Journal of Child Psychology and Psychiatry and Allied Disciplines, 37, 157–165. Takeuchi, K., Kubota, H., and Yamamoto, J. (2002), Intensive supervision for families conducting home-based behavioral treatment for children with autism in Malaysia, Japanese Journal of Special Education, 39, 155–164. Tantam, D. (2003), Assessment and treatment of comorbid emotional and behavior problems, in Prior, M., Ed., Learning and Behavior Problems in Asperger Syndrome, New York: Guilford, pp. 148–174. Thiemann, K.S. and Goldstein, H. (2001), Social stories, written text cues, and video feedback: effects on social communication of children with autism, Journal of Applied Behavior Analysis, 34, 425–446. Thorp, D.M., Stahmer, A.C., and Schreibman, L. (1995), Effects of sociodramatic play training on children with autism, Journal of Autism and Developmental Disorders, 25, 265–282. Turnbull, A.P. and Ruef, M. (1997), Family perspectives on problem behavior, Mental Retardation, 34, 280–293. Warren, S.F. and Kaiser, A.P. (1988), Research on early language intervention, in Odom, S.L. and Karnes, M.A., Eds., Early Intervention for Infants and Children with Handicaps: An Empirical Base, Baltimore, MD: Brookes, pp. 84–108. Wellman, H.M., Baron-Cohen, S., Caswell, R., Gomez, J.C., Swettenham, J., Toye, E. et al. (2002), Thought-bubbles help children with autism acquire an alternative to a theory of mind, Autism, 6, 343–363. Wetherby, A.M., Prizant, B.M., and Schuler, A.L. (2000), Understanding the nature of communication and language impairments, in Wetherby, A.M. and Prizant, B.M., Eds., Autism Spectrum Disorders: A Transactional Developmental Perspective, Baltimore, MD: Brookes, pp. 109–142. Whalen, C. and Shreibman, L. (2003), Joint attention training for children with autism using behavior modification procedures, Journal of Child Psychology and Psychiatry, 44, 456–468. Williams, G., Perez-Gonzalez, L.A., and Vogt, K. (2003), The role of specific consequences in the maintenance of three types of questions, Journal of Applied Behavior Analysis, 26, 285–296. Williams, T.I. (1989), A social skills group for autistic children, Journal of Autism and Developmental Disorders, 19, 143–155.

© 2006 by Taylor & Francis Group, LLC

2732_C019.fm Page 473 Wednesday, February 15, 2006 6:05 PM

Behavioral, Educational, and Developmental Treatments for Autism

473

Wolfberg, P.J. and Schuler, A.L. (1993), Integrated playgroups: a model for promoting the social and cognitive dimensions of play in children with autism, Journal of Autism and Developmental Disorders, 23, 467–489. Yoder, P.J. and Layton, T.L. (1988), Speech following sign language training in autistic children with minimal verbal language, Journal of Autism and Developmental Disorders, 18, 217–229. Yoder, P.J. and Warren, S.F. (2001), Relative treatment effects of two prelinguistic communication interventions on language development in toddlers with developmental delays vary by maternal characteristics, Journal of Speech, Language, and Hearing Research, 44, 224–237. Zercher, C., Hunt, P., Schuler, A., and Webster, J. (2001), Increase joint attention play and language through peer supported play, Autism, 5, 374–398.

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 475 Wednesday, March 29, 2006 3:54 PM

20

The Costs of Autism Michael L. Ganz

CONTENTS Introduction............................................................................................................476 Overview of Costs .................................................................................................477 Components of Cost .......................................................................................477 Direct Medical Costs .................................................................................478 Direct Nonmedical Costs...........................................................................478 Indirect Costs .............................................................................................478 Sources of Costs and Methods for Constructing Cost Estimates .................478 Perspective ......................................................................................................480 Methods..................................................................................................................480 Review of the Literature and Calculation of Costs ..............................................482 Direct Medical Costs......................................................................................482 Physician, Outpatient, and Clinic Services ...............................................482 Dental.........................................................................................................483 Prescription Medications ...........................................................................483 Complementary and Alternative Therapies...............................................483 Behavioral Therapies .................................................................................484 Hospital and Emergency Department Services.........................................485 Allied Health, Equipment, Supplies, and Home Health...........................485 Medically Related Travel ..........................................................................486 Summary ....................................................................................................486 Direct Nonmedical Costs ...............................................................................486 Child Care..................................................................................................486 Adult Care..................................................................................................487 Respite Care and Family Care ..................................................................487 Home and Car Modifications ....................................................................487 Special Education ......................................................................................487 Supported Employment .............................................................................488 Other ..........................................................................................................489 Summary ....................................................................................................490 Indirect Costs..................................................................................................490 Productivity Losses of People with Autism..............................................490 Productivity Losses of Parents of People with Autism ............................490 Summary ....................................................................................................491

475

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 476 Wednesday, March 29, 2006 3:54 PM

476

Understanding Autism: From Basic Neuroscience to Treatment

Results and Sensitivity Analyses...........................................................................491 Discussion ..............................................................................................................493 Directions for Future Research ......................................................................497 Acknowledgments..................................................................................................497 References..............................................................................................................498

INTRODUCTION Autism can be a very expensive disorder. In any given year, it can cost society about $3.2 million to care for a person with autism over his or her lifetime and about $35 billion (in direct medical, direct nonmedical, and lost productivity costs) to care for all individuals with autism over all of their lifetimes. Although harder to quantify, autism also exacts a large human cost in terms of reduced quality of life for affected individuals and their families. As expenditures for caring for and supporting people with autism and their families have grown at a rapid pace over the past decades, it is important to understand how that money is spent.1 Given the burdens we face and given the increasing variety of options for treatment and possibly for prevention, society will have to make important decisions about the allocation of scarce resources. Competing or complementary treatment and prevention strategies currently available, or yet to be developed, vary in their effectiveness and implementation costs. By understanding how total costs are distributed, we can make better decisions about the distribution of resources, and we can also understand how various stakeholders value the potential outcomes of interventions. The purpose of this chapter is to present a theoretical and methodological outline of how to identify the financial societal costs of autism as well to present estimates of those costs based on publicly available data. The goal is that this information will be useful to cost-effectiveness analysts performing economic evaluations of treatment and prevention options and to policymakers and advocates as a reference source on the costs of autism. This chapter should also be useful to those researchers who are planning to implement systems to prospectively track and document costs of children with autism in order to produce more precise cost estimates. Although this chapter will present a best estimate for the societal costs of autism in the U.S. (along with some reasonable bounds on that estimate), these are, as the name implies, estimates. As will be discussed later in the text, the quality of these final estimates hinges on the quality of the source data. A rather extensive literature review did uncover a number of articles on various aspects of health care for individuals with autism, but I could not identify studies in the peer-reviewed literature that synthesized the various strands in the literature for the U.S. Two notable exceptions are a study of the costs of autism based on a sample of 308 individuals classified as having autism according to DSM-III criteria (out of a sample of 634 individuals with severe developmental disorders) during the mid-1980s by Birenbaum et al.2 and a recent study in Britain by Järbrink and Knapp.3 In addition, because autism is rarer than most other conditions studied and because individuals with autism receive medical and nonmedical services from a variety of sources, each potentially with different financing methods, it is very difficult or almost impossible to rely on one source of data for estimating costs, unlike estimating

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 477 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

477

the costs of more prevalent conditions such as depression, asthma, and high cholesterol, to name a few. Because of these data limitations, it is difficult to easily summarize the costs of autism. In this chapter, estimates of the lifetime and total annual costs of autism are presented. These estimates were constructed by obtaining estimates for various components of care from available published sources. Some studies relied on relatively large samples, whereas others relied on much smaller ones. Inherent in assembling these disparate pieces of information into a national estimate for presentation to a wide audience are certain simplifying assumptions, which are listed in a technical appendix available from the author. This chapter is organized as follows. The next section provides an overview of estimating the costs of illness and reviews the types of costs that are relevant for this analysis and provides definitions and potential sources for those data. The third section presents the methods and assumptions used in the present analyses, and the fourth section provides a review of the literature and a summary of the estimated costs of autism. The fifth section presents the overall results and the results of some sensitivity analyses. The final section concludes and offers suggestions for future work. It is, unfortunately, beyond the scope of this chapter to treat many components of this analysis in great detail (these details are available in a technical appendix).

OVERVIEW OF COSTS In order to understand what exactly comprises the total cost of autism and in order to be able to assess the validity and reliability of such cost estimates, users of these cost data should have detailed information on the components of those estimates. In general, the total cost of autism can be estimated by adding the direct costs of treating autism, both medical and nonmedical, to the indirect costs, which result mainly from associated productivity losses. In this section, those components are defined and discussed. The terms costs and expenditures will be used interchangeably, but do not correspond to charges. A discussion of the differences between costs (expenditures) and charges is beyond the scope of this chapter. Interested readers are referred elsewhere.4

COMPONENTS

OF

COST

The total cost of autism is the sum of the direct and indirect costs associated with it. Direct costs represent the value of actual resources (goods and services) that were consumed and include direct medical and nonmedical costs. Indirect costs represent the value of lost productivity due to autism-related morbidity and mortality and can result from the lost or diminished work time of individuals with autism or members of their families (see Gold et al. for further details4). According to standard economic theory, costs equal the value of forgone opportunities.5,6 In other words, these opportunity costs represent the value of other activities that these resources could have been used for had they not been consumed by or on behalf of people with autism. Only opportunity costs are enumerated in cost-of-illness studies and economic evaluations of treatment or prevention activities.4 From a societal perspective, income transfers, such as disability or welfare payments, are not opportunity costs,

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 478 Wednesday, March 29, 2006 3:54 PM

478

Understanding Autism: From Basic Neuroscience to Treatment

but rather redistributions of money and, hence, do not represent costs due to lost resources. Because income transfers, such as disability payments, are meant to compensate individuals for lost productivity due to injury or illness, inclusion of those transfer payment in the cost of illness calculations will result in double counting. In the discussions and analyses that follow, only opportunity costs are considered. Direct Medical Costs Direct medical costs represent the value of all medical goods and services consumed in the care and support of individuals with autism. These costs include physician and therapist services, inpatient and emergency room costs, long-term care costs, drugs, health care services for the parents or other family members (such as mental health care services), allied health, equipment, eye glasses, other supplies, and can include time costs related to health care such as the direct cost of travel (fuel, parking) and patient waiting times. Direct Nonmedical Costs Similar to direct medical costs, direct nonmedical costs represent the value of nonmedical goods and services used to care for individuals with autism and can include developmental services, out-of-home placement, day programs, camps, transportation, child care and babysitting, respite care, home and vehicle modifications, special education, supported employment, and the direct costs of travel for these nonmedical services. Because some aspects of care for individuals with autism are blurred (for example, many special education programs provide behavioral therapies), there is likely to be some misclassification between direct medical and direct nonmedical costs. Indirect Costs Indirect costs represent the value of lost productivity as a result of an individual having autism, and it also represents the value of lost productivity of those caring for individuals with autism. Lost productivity relates to the value of lost or impaired work time due to morbidity and, when relevant, to mortality. In the case of the individual with autism, it represents the difference between the value of his potential salary income, benefits, and household services as a person with autism and his potential salary income, benefits, and household services as a person without autism. In the case of caregivers, such as parents, it represents the loss in salary income, benefits, and household services due to missed time at work, reduced work hours, switching to a lower-paying but more flexible job, or not working all together.

SOURCES

OF

COSTS

AND

METHODS

FOR

CONSTRUCTING COST ESTIMATES

In order to collect and compute the lifetime costs of individuals with autism, one would ideally rely on an incidence-based cost-of-illness approach.4,5,7–9 The incidencebased cost-of-illness approach is the most appropriate method for deriving costs of

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 479 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

479

an illness because this method prospectively tracks individuals over their life span. It is most appropriate not only because it incorporates changes in the cost of the illness over individuals’ lifetimes, but as it tracks individuals, changes in the nature of their disability can be captured and the costs of associated comorbid conditions can also be incorporated. Because of this richness and sensitivity to the evolution of an illness or disorder, the cost estimates derived from the incidence-based approach closely estimates the value of prevention. In other words, by preventing autism and its associated comorbidities, society would save the equivalent of the incidence-based cost estimate. However, as it is difficult to derive an incidence-based cost estimate, owing mainly to rather formidable data requirements, many studies use a prevalence-based approach, whereby cross-sectional cohorts of individuals with autism are identified and used to construct cost estimates. Costs derived under the prevalence-based approach represent the costs to care for people with autism at a point in time rather than the lifetime profile of autism itself and, therefore, the prevalence-based cost estimates more closely represent the cost of treatment of and caring for those with autism.7 For health conditions that are declining in importance and incidence (for example, polio, measles, or rubella), the prevalence-based approach will yield costs that are larger than the incidence-based method, and for emerging and increasingly important conditions, the costs based on the prevalence approach will be less than those based on the incidence approach. In certain situations, for example, when incidence rates, cohort sizes, and health care technologies are constant over time, both the prevalence-based and the incidence-based approaches will yield similar estimates. Cost estimates based on the prevalencebased approach are useful for understanding current expenditure patterns and for the allocation of current resources, whereas cost estimates based on the incidencebased approach are more appropriate for understanding dynamic patterns and for making decisions about resource allocation that involve future time periods, as when making decisions about future treatment or research patterns.10 Most of the studies on the costs of autism reviewed in the following text rely on the prevalence-based approach, as does my own analysis of national survey data. In this chapter, I use a prevalence-based approach to approximate an incidence-based cost estimate by applying a lifetime cost estimate based on a synthetic cohort11 of people with autism to the incident population in 2003. The first step in estimating costs includes identifying the resources required and their costs by enumerating the labor and capital inputs.4 Various methods can be used to construct direct cost estimates. One method includes prospectively collecting information on the health care experiences of children with autism as they are diagnosed and start to receive services (the incidence-based approach). In order to accomplish this, children need to be prospectively tracked and all relevant information should be recorded. In theory, this is feasible, especially for direct medical care, in which most services appear in administrative databases such as medical and billing records, and private and public health care claims-processing systems. Direct nonmedical and indirect costs can be ascertained by other means such as expenditure diaries or other methods to track spending.12,13 This is, however, a relatively cumbersome and costly endeavor and, to my knowledge, has not yet been done on a scale large enough to derive stable and representative estimates of costs for autism. Furthermore, depending

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 480 Wednesday, March 29, 2006 3:54 PM

480

Understanding Autism: From Basic Neuroscience to Treatment

on how long costs are to be tracked, such a study may need to be active for many decades. Instead, cost estimates can be constructed from existing data on prevalent cases of autism. In some cases, depending on the sources of data, the resulting cost estimates may be more representative than data collected from smaller prospectively collected panels of individuals. As mentioned earlier the prevalence-based approach combines information on the prevalence rate of autism in different age groups (as an estimate for the incidence rate) with information on the costs of caring for individuals with autism in those age groups. Information on prevalence and costs can come from different sources such as surveillance systems, national surveys, health care claims databases, and other administrative data. Indirect costs, which represent the value of lost productivity, can be measured in two ways. For individuals who are already working and have reduced their work effort to care for a child with autism, for example, the difference between before and after incomes (including benefits) and household services can be calculated. Even when people do not earn an income, their time is nevertheless not worth zero. The value of time of people who are not working can be estimated using published averages for age- and sex-specific wages, which are generally a good approximation of opportunity costs.4 For stay-at-home parents and teenagers, this method may work, but for younger children, unless one can collect willingness-to-pay data, there are no data to use for their time costs.4 In both cases, family members caring for a person with autism may also experience a decrease in leisure time, which may be more difficult to estimate.

PERSPECTIVE In any cost analysis, it is important to consider the perspective from which the analysis is conducted. Families, insurers, taxpayers, and society, all have a stake in paying for the care of people with autism and, depending on the perspective, the enumeration of the relevant costs will differ. For example, a family perspective will exclude, among other things, publicly financed special education, and a state government perspective will exclude, among other items, alternative therapies paid by families. The scope of the perspective is dictated by how narrowly or broadly the problem is defined or who is expected to pay for prevention or treatment or who is expected to benefit from that prevention or treatment. Given the broad nature of autism in terms of its impact not only on families but on schools, social service agencies, and even employers, and because there is considerable public funding for supporting individuals with autism, a societal perspective will be taken, as recommended by the Panel on Cost-effectiveness in Health and Medicine.4

METHODS The following methods were used to construct the estimate of the societal cost of autism in the U.S. An in-depth literature review concentrating on U.S.-based studies was conducted, and studies that reported on direct costs were identified and abstracted. When necessary to obtain otherwise unavailable data, studies conducted in Britain and Canada were also used. Cost estimates as reported in the source

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 481 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

481

materials were then inflated to 2003 U.S. dollars using the all-item consumer price index as published by the Congressional Budget Office (CBO).14 When state-specific costs were reported, they were transformed to national averages using data from Table 699 of the Statistical Abstract of the United States,15 and costs enumerated in foreign currencies were first converted to U.S. costs using the latest available exchange rate for the given year using data from the Federal Reserve.16 When data in a source document were only available as utilization measures (for example, the number of hospitalizations), they were translated to a cost estimate by multiplying the utilization measure by a survey-adjusted average cost for the particular services among all people in the appropriate age group as calculated from the Medical Expenditure Panel Survey (MEPS).17 In addition to identifying costs from the published literature, I also used the National Health Interview Survey (NHIS)18 and the MEPS to identify individuals with autism (or other developmental problems) in order to estimate their direct medical care costs. Individuals with an ICD-9 diagnosis code of 299 (due to confidentiality concerns the MEPS only reports the first three digits), which includes autism diagnoses (299.0x) as well as disintegrative psychoses (299.1x) and early childhood psychoses (299.8x/299.9x), were identified as having autism. As respondents of the MEPS are linkable to the NHIS survey of the previous year, those individuals with an affirmative response to a specific question about autism (available for 1997 to 2000) or with an ICD-9 diagnosis code of 299 in the appropriate NHIS were also identified as having autism. Data on autism status from the NHIS files were then linked to the appropriate MEPS files, and survey-adjusted means for expenditures were then computed as described in the preceding text. When there were multiple cost estimates from multiple sources for a particular category (behavioral therapies, for example), I averaged those costs to obtain a single cost estimate for that category. When possible, cost estimates are derived for higher- and lower-functioning individuals based on how costs in these two categories are presented in the literature. When articles present data for individuals who are semidependent or independent, or for individuals with high-functioning autism, they are classified in the higherfunctioning category. Individuals who are dependent or not classified as having highfunctioning autism are grouped in the lower-functioning category. Based on data presented in Fombonne (2003), we assume that 54% of people with autism are classified as higher functioning (having a lower level of disability) and 46% as lower functioning (having a higher level of disability).19 For summary cost estimates presented in the section titled “Review of the Literature and Calculation of Costs,” population-weighted averages are presented. The age-, sex-, and disability-specific costs, however, were used in computing the overall lifetime and annual estimates presented in the tables. Indirect costs are computed using a human capital approach4,7 and combine data on average earnings, benefits, and household services with information on average work life expectancies and labor force participation rates to derive an estimate of what the average person would have earned. In the case of people with autism, these hypothetical earnings are then adjusted for the fact that some proportion of adults with autism do work. In the case of parents, assumptions are made about reductions in work to derive the value of lost productivity.

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 482 Wednesday, March 29, 2006 3:54 PM

482

Understanding Autism: From Basic Neuroscience to Treatment

The societal costs of autism should reflect those costs that are specific to autism — in other words, the incremental costs of autism. Incremental costs do not include costs that would be incurred regardless of autism. (Regular yearly check-ups would be one example.) When possible, those incremental costs are taken directly from the source materials (for example, behavioral therapies or vitamin supplements for autism). When costs are not reported as incremental costs, I subtracted national average costs data for the appropriate category, using age-group-specific average costs for medical services derived from my analysis of the MEPS data. For example, if a source reports that children less than 18 years old with autism have outpatient costs of $X and the national average for children less than 18 years old is $Y, then the incremental cost due to autism would be $(X − Y). Annual costs per person with autism were computed by summing estimates for direct medical, direct nonmedical, and indirect costs for each age group from age 3 through sex-specific life expectancies. There is some evidence that people with autism have reduced life expectancies.20–23 Standard life expectancy estimates derived from standard life tables24 were reduced according to Shavelle and Strauss.22 The average male life expectancy at age 3 (72.2 yr) was reduced by 6.1 yr to 66 yr and the female life expectancy at age 3 (77.4 yr) was reduced by 12.3 yr to 65 yr. To compute lifetime costs, the costs at each age were discounted to present value using a discount rate of 3% as recommended by the Panel on Cost-Effectiveness in Health and Medicine4 (earnings were first inflated to future values using projected productivity growth rates14 and summed). Finally, the present value lifetime costs were then multiplied by the estimated number of 3-year-olds with autism in the U.S. (11,000), based on the assumed prevalence of 27.5 per 10,000 persons19 and the population of 3-year-olds in the U.S. in 2003 of 4 million.19,25,26 Other assumptions underlying these analyses are also presented in the relevant places in the next section. Sensitivity analyses were conducted and reported based on varying the inputs to these calculations. Summaries of these assumptions, which are presented in a detailed technical appendix, and the computer files necessary to compute the cost estimates are available from the author.

REVIEW OF THE LITERATURE AND CALCULATION OF COSTS DIRECT MEDICAL COSTS This subsection presents the direct medical costs obtained from the literature and from my calculations based on analysis of the NHIS/MEPS. When applicable, the incremental costs are computed and presented as well. Physician, Outpatient, and Clinic Services According to Birenbaum et al., children with autism aged 0 to 5 visited a physician or some type of outpatient clinic six times per year, children aged 6 to 17 and adults 18 to 24 experienced four visits per year for an average cost of $513 ($300 in 1985) for children less than 18 years old and $443 ($259 in 1985) for adults 18 to 24.2 According to my analysis of the NHIS/MEPS, visits to physicians and outpatient

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 483 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

483

settings varied in cost according to age group, but averaged $4,679 per person per year (in 2003 dollars). The annual population-weighted incremental cost for annual physician, outpatient, and clinic services is estimated to be about $1,896 per person (in 2003 dollars). Dental Birenbaum et al2 estimated the annual cost of dental care for children to be $86 ($50 in 1985) and $103 ($60 in 1985) for adults, and I estimate an annual cost of $548 per person per year (in 2003 dollars) based on analysis of the NHIS/MEPS. The annual population-weighted incremental cost for dental care is estimated to be about $134 per person (in 2003 dollars). Prescription Medications Two U.S. studies on the use of psychotropic drugs by individuals with autism reported that approximately 33 to 55% of subjects were using some type of psychotropic drug.27,28 Depending on the level of disability, the most common types were antidepressants (5 to 30%), stimulants (7 to 20%), neuroleptics (7 to 27%), and sedatives (5 to 8%). Based on average expenditure amounts for these classes of drugs from the 1996 to 2001 MEPS, I estimate the annual cost of psychotropic drugs as represented by the Aman27 and Martin28 articles to be between $45 and $97 for individuals with low levels of disability and $170 to $399 for those with higher levels of disabilities (in 2003 dollars); neither article provided data by age and disability level. In addition to those studies, Birenbaum reported average annual prescription drug costs of $171 ($100 in 1985) for children and $274 ($160 in 1985) for adults. The overall best estimate for the cost of prescription drugs is therefore $166 and $295 (in 2003 dollars), depending on the level of disability. As these estimates tend to be smaller than the average national level of expenditures on medications, and as most the drugs reported here were specifically for autism, these estimates are not adjusted to obtain incremental costs. Complementary and Alternative Therapies A number of recent studies have reported on the use of complementary and alternative medicine (CAM) therapies, such as herbs, vitamins, special diets, and other therapies or agents, for the treatment of autism and its symptoms.29–33 Parents can spend a great deal of time and money seeking some type of cure even though many of these agents are of unproven or doubtful efficacy. Nickel reported that 50% of children with autism were using some type of CAM treatment.33 Langworthy-Lam and colleagues reported that 5.7% of members of the Autism Society of North Carolina used vitamin supplements for autism.31 Levy et al. reported that about 32% of children at a hospital-based autism center in Philadelphia used some type of CAM modality (vitamins, gastrointestinal medications, melatonin, special diets, secretin, chelation, and other biological and nonbiological therapies), with approximately 9% having used a potentially harmful therapy and 11% having used multiple therapies.32

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 484 Wednesday, March 29, 2006 3:54 PM

484

Understanding Autism: From Basic Neuroscience to Treatment

Aman et al. reported that 10% of subjects in their survey of individuals with autism used vitamins specifically for autism29 and most recently, Green et al. report that 27% of parents were implementing special diets, 43% were giving their children vitamin supplements, and 26% were using alternative therapies such as aromatherapy and dance therapy.30 These articles, however, do not report on the exact type or intensity of CAM therapies used, so it is difficult to estimate the cost of such treatments. Age is inversely related to the percentage of people using CAM therapies and the number of different agents and therapies used,30,31 but the number used is directly related to the severity of disability.30 Eisenberg et al. reported, based on a national sample of adults, that the average expenditure for CAM providers, dietary supplements, and vitamins was approximately $674 ($478 in 1990) per year.34 Yussman et al. report, using the 1996 MEPS, that among pediatric users of CAM (of which most were in very good to excellent health), the average yearly expenditure on CAM visits and remedies was $101 ($86 in 1996).35 Based on the rates of CAM usage presented in these reports and on the usage rates stratified by severity level in Green et al.,30 I estimate usage rates of 52% for the youngest and 12% for the oldest individuals with low levels of severity, and 67% for the youngest and 16% for the oldest individuals with higher levels of severity, which corresponds to a population-weighted average cost of $84 for low disability and $65 for high-disability individuals. These estimates are likely to understate the cost for individuals with autism, especially for more intensive users of CAM therapies, but little other published data are available on this topic. Behavioral Therapies Although it is not completely clear how effective different types of behavioral interventions are for children with autism,36,37 it is rather well accepted that some type of intervention should be initiated.38–40 Because their use is becoming more pervasive and as more states are legislating that behavioral therapies become covered services as part of health insurance plans, their costs are included here. However, as the estimates of effectiveness, and hence the financial and nonfinancial benefits, are controversial and because the correspondence between improvement in symptoms and the costs of those interventions are not clear, only the costs of the intervention are enumerated here and are not offset by potential benefits.41 According to a cost–benefit analysis of early intensive behavioral intervention (EIBI) by Jacobson et al., the costs are $37,537 ($32,820 in 1996 in Pennsylvania) per child per year starting at age 3 for 3 years using the Lovaas treatment protocol.1,42,43 In preparation for legal action against the provincial government of British Columbia, Canada, a cost–benefit analysis of the Lovaas treatment protocol was prepared by Columbia Pacific Consulting.44 That analysis, which estimated the net costs of the treatment under various outcome assumptions, reported treatment costs of $45,053 ($65,000 in 2000 in Canada) per year for children aged 3 to 6. In addition, for ages 6 through 19 (during which time children would typically be receiving special education services), Hildebrand presented annual costs of $4,140 ($5,810 in 2000 in Canada) and $5,914 ($8,300 in 2000 in Canada), depending on the level of disability, for behavior support.44 I estimate that the annual cost of intensive behavioral

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 485 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

485

therapies and behavior support services to be $41,295 for children aged 3 through 6, and $4,140 and $5,914 per child, depending on level of disability, for ages 6 through 19 (in 2003 dollars) (although I assume age 3 as the age at which autism is diagnosed, most states now provide some type of treatment from diagnosis, which can be earlier than 3 years old). As these services are specifically for autism, there is no need to adjust to obtain incremental costs. Other types of behavioral interventions may be either more or less expensive, but little data are available on those interventions. Hospital and Emergency Department Services According to Birenbaum et al. about 10% of children aged 5 to 17 years with autism experienced one hospital stay per year with an average length of stay of 6 d and about 15% of adults aged 18 to 24 experienced one hospital stay with an average length of stay of 15 d.2 For both age groups less than 1% had more than 1 hospital stay per year, and roughly 15% of children less than 18 years had one emergency department visit, and slightly less than 20% of adults 18 to 24 years old had one emergency department visit. Birenbaum estimated the annual hospital and emergency department care to be $599 ($350 in 1985) for children less than 18 years and $3,307 ($1,775 in 1985) for adults 18 to 24 years. According to Walsh et al., who examined the utilization and expenditure patterns of hospital care for people with developmental disabilities (which includes individuals with autism but also includes those with mental retardation), the annual costs were $4,467 ($3,303 in 1991) for children 0 to 21, $4,672 ($3,454 in 1991) for adults 22 to 55, and $10,792 ($8,112 in 1991) for adults 56 to 85.45 Given that Walsh’s sample contains individuals with mental retardation as well as individuals with autism, these cost estimates are higher than estimates would be if the sample consisted only of people with autism. The ratio of hospital costs for people with autism to people with severe mental retardation in the Birenbaum study is 0.59:1 (=$1,000 ÷ $1,700 in 1985).2 Using that ratio, the Walsh estimates are adjusted downward to $2,636, $2,757, and $6,473 ($1,949, $2,038, and $4,786 in 1991). According to my analysis of the NHIS/MEPS, annual hospital and emergency care costs about $1,063 per person per year (in 2003 dollars). The annual population-weighted incremental cost for hospital and emergency care is estimated to be $1,384 per person. (Because of some concern over the appropriateness of using data on hospital costs from a 20-year-old source, I recomputed the final cost estimate presented in Table 1 after excluding the Birenbaum data. The final estimate differed by less than 1%.) Allied Health, Equipment, Supplies, and Home Health Birenbaum also reported on the cost of allied health, equipment, and supplies.2 Costs for children less than 18 years old were $376 ($220 in 1985) and costs for adults 18 to 24 were $222 ($130 in 1985). According to my analysis of the NHIS/MEPS, the average annual cost for home health is $1,254 per person (in 2003 dollars). The annual population-weighted incremental cost for allied health, equipment, supplies, and home health is estimated to be about $316 per person (in 2003 dollars).

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 486 Wednesday, March 29, 2006 3:54 PM

486

Understanding Autism: From Basic Neuroscience to Treatment

Medically Related Travel Birenbaum also reports that for 0 to 24-year-olds, the average costs for medically related travel is $86 ($50 in 1985) per person per year.2 Measured in 2003 dollars, $86 forms the best estimate for medically related travel. These costs are treated as specific to autism. Summary Combining the estimates of direct medical care costs from the literature reviewed earlier with my analyses of the NHIS/MEPS surveys, I estimate the populationweighted annual per capita costs, measured in 2003 dollars, of direct medical costs for individuals with autism to be, on average, $29,091 for those with lower levels of disability and $29,569 for those with higher levels of disability. These estimates are similar to each other because the estimates for nontherapy costs differed very little by disability level and because the therapy costs were only relevant through age 19 (and differed by less than $2,000 for ages 7 to 19). In the final tabulations presented in Table 20.1, the differences between lower and higher levels of disability are magnified owing to the assumed ratio of lower to higher levels of disability.

DIRECT NONMEDICAL COSTS All of the sources that were reviewed for this chapter on the topic of direct nonmedical reported costs in specific categories. Unlike many of the direct medical costs that are theoretically applicable to all ages, many of the direct nonmedical costs are applicable only for certain ages. I take account of that in presenting the cost estimates in the following text. Most of these costs are specific to autism and hence do not need to be adjusted to obtain incremental costs. Exceptions are noted in the following text. Child Care The Birenbaum report presents estimates of child care costs for people with autism living at home to be $428 ($250 in 1985) per child per year and $86 ($50 in 1985) per adult (ages 18 to 24) per year.2 Hildebrand reported $4,788 ($6,720 in 2000 Canadian) for “semidependent” children and $6,841 ($9,600 in 2000 in Canada) for “very dependent” children.44 Based on data from Jacobson et al.1 and Hildebrand44, I assume children use these services through age 22. In addition, Hildebrand presented a cost estimate of $16,161 ($22,680 in 2000 in Canada) to $23,087 ($32,400 in 2000 in Canada) for placement after age 6 (the costs range from a lower value to a higher value depending on the level of disability).44 After subtracting the average annual household expenditure on child care per year ($3,588),46 I estimate the annual incremental cost of child care services to be $3,509 per person with lower levels of disability and $6,502 per person with higher levels of disability (in 2003 dollars) for ages 3 through 22.

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 487 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

487

Adult Care I assume, starting at age 23, that adults receive services through day programs. Hildebrand estimates the cost of such programs to be $13,168 ($18,480 in 2000 in Canada) and $25,937 ($36,400 in 2000 in Canada), depending on the level of dependence.44 Based on calculations presented in Hildebrand, I further assume that 50% of adults will live at home and the rest will live in a group home. Again, Hildebrand’s report provides estimates of $51,175 ($71,820 in 2000 in Canada) for semidependent adults living in the family home and $73,108 ($102,600 in 2000 in Canada) for very dependent adults living in a group home. Jacobson et al. report that home- and communitybased service costs between $36,394 ($31,818 in 1996 in Pennsylvania) and $53,574 ($46,838 in 1996 in Pennsylvania) per person per year and that institutional services cost $66,448 ($56,775 in 1996 in Pennsylvania) per person per year.1 My best estimate of the annual cost of adult care, based on the data in these articles, is $44,538 per person with lower levels of disability and $52,025 per person with higher levels of disability (in 2003 dollars) for ages 23 through end of life. Respite Care and Family Care Birenbaum’s estimate of $17 ($10 in 1985) per child per year cost for respite care is quite low but reflects the rather low proportion of families in their study using respite services (8%).2 Hildebrand, however, reports a much larger value of $1,846 ($2,590 in 2000 in Canada) and $2,636 ($3,700 in 2000 in Canada) for respite services, depending on the child’s level of disability.44 Jacobson et al. estimated that family support services cost $1,258 ($1,110 in 1996 in Pennsylvania) per person per year for 18 years.1 I use $1,044 and $1,308 per person per year (in 2003 dollars), depending on level of disability, as the cost estimates for respite and family care services for ages 3 through 22. Home and Car Modifications Some modifications to the home environment and to family-owned vehicles may be necessary due to damaging and aggressive behaviors and accidents.23,47–49 Birenbaum reported annual expenditures of about $171 ($100 in 1985) per child and $17 ($10 in 1985) per adult.2 Fujiura et al., in their study of out-of-pocket expenditures, reported that among the extraordinary costs that some families incurred was the cost of architectural and vehicle modifications.49 As those costs were not explicitly presented apart from the other extraordinary costs, they are reported here. These extraordinary costs are, however, discussed in the section on other costs in the following text. The annual population-weighted cost for home and vehicle modifications is estimated to be about $55 per person (in 2003 dollars). Special Education As consequence of the Education of all Handicapped Children Act (PL-94-142) of 1975, currently enacted as the Individuals with Disabilities Education Act (PL-105-17)

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 488 Wednesday, March 29, 2006 3:54 PM

488

Understanding Autism: From Basic Neuroscience to Treatment

that was reauthorized in 1997, many children with autism are enrolled in public school special education programs or other Department of Education programs for children with developmental disorders. Yeargin-Allsopp et al. report that 91% of their sample of children with autism received special education services during the year.50 Clearly, special education is likely to represent a significant portion of the societal costs of autism. Jacobson et al.1 reported that special education costs $15,138 ($12,935 in 1996 in Pennsylvania) and $33,713 ($28,806 in 1996 in Pennsylvania) per child per year depending on the level of a child’s need, and Hildebrand 44 reported $19,702 ($27,650 in 2000 in Canada) and $28,216 ($39,599 in 2000 in Canada) for special education, varying with the level of dependence. According to the Center for Special Education Finance, the nationwide average expenditure per pupil for regular education is $4,854 ($4,745 in 2002) and $8,501 ($8,310 in 2002) per pupil for special education.51 I estimate that children with autism having lower levels of disability cost $8,645 per child per year and children with higher levels of disability cost $18,623 per child per year (in 2003 dollars). Using data presented in Järbrink and Knapp 3, I assume that 100% of children with higher levels of disability utilize (higher costing) special education between the ages of 6 and 21 and of the children with lower levels of disability, 80% utilize (lower costing) special education and 20% utilize regular education (even though there may be some overlap between special education and behavioral therapy services, we treat these as distinct categories and recognize the possibility that these cost estimates may be slightly overstated in this regard as a result). Based on these data and assumptions, I estimate the average annual cost of special education to be $16,128 (in 2003 dollars) for children 6 to 21 years old (in some cases special education costs may be incurred before children reach school age and this can be accounted for in the sensitivity analyses presented in the following text). This may be an underestimate of costs to the extent that special education program also provide more expensive services such as intensive behavioral therapies (see the subsection titled “Behavioral Therapies”). Supported Employment Many adults with autism do not work or cannot work in the competitive job market.52 The value of those lost incomes will be discussed in the section on indirect costs, but I discuss here the cost of supported employment for those that do work. As set out by the Developmental Disabilities Act of 1984, and as amended by the Rehabilitation Act Amendments of 1986 and 1992, supported employment is employment for persons with developmental disabilities (who are otherwise unable to obtain competitive employment at the minimum wage) that is aided by any necessary supervisory, training, or transportation supports.53 Supported employment usually involves a job coach who helps an individual with autism find employment, provides on-the-job training, and for a certain period of time, serves as a liaison between the individual and his or her employer.54 The costs associated with supported employment include the administrative costs of such a program, the costs of the job coach (which tend to be the most expensive part of the program), and any tax credits that employers may receive. Keel et al. reported on the outcomes of the TEACCHsupported employment plan.54 Although they do report that 89% of job candidates

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 489 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

489

retained their jobs and that on average the adults with autism in their study (North Carolina) worked about 29 h per week earning $6.07 ($5.29 in 1997) per h, it is not clear from their study for how long the job coaches worked with these individuals or how much it cost to support them. Rusch et al. performed a cost–benefit analysis of supported employment programs in Illinois during the period 1985 to 1990 for persons with disabilities, and they reported that over the longer term the program’s benefits outweighed its costs.55 They did report that over this time period the supported employment programs cost about $11,469 ($11,500 in 1993 in Illinois dollars) per supported person. Mawhood and Howlin reported a total cost of $4,142 (£ 2,473 in 1999) per program participant.52 Capo suggests that an occupational therapist is most qualified to assume the role of the job coach 56 and, according to the Bureau of Labor Statistics, the median salary of an occupational therapist is $53,186 ($51,990 in 2002).57 Adding fringe and other benefits at 29% of earning income (computed from Table 628 of the Statistical Abstract of the United States15) increases the estimate to $68,610. Assuming that three adults work with a job coach at a time,54 I estimate that the annual (gross) cost of supported employment to be $13,878 per supported person. As Jacobson et al. assume that supported employment earnings are 20% of the average income,1 I offset the costs of supported employment with 20% of the average individual income ($53,641 for men and $36,688 for women15). Assuming that 35% of adults with lower levels of disability and 10% of adults with higher levels of disability with autism are in a supported employment environment,58 and that those who are employed work at 75% effort, I estimate that the net annual cost of supported employment to be $2,041 (men) and $2,931 (women) for those with lower levels of disability, and $874 (men) and $1,56 (women) for those with higher levels of disability per supported person in 2003 dollars for ages 23 through end of working life, which is assumed to be 57 for men and 53 for women.59 Mawhood and Howlin 52 and others (for example, Heal et al.60) note that it may be difficult to fully compute the net costs of supported employment mainly because it is difficult to quantify the benefits of supported employment. Other In their report of the out-of-pocket costs of family care for adults with mental retardation and developmental disabilities (14% of whom were adults with autism), Fujiura et al. reported on nine categories unlikely to be paid by any other source.49 These include groceries, personal care, clothing, formal and informal recreation, durable consumer items, specialized services (such as respite services and day programs), transportation, and extraordinary expenses and were estimated to cost $8,315 ($5,900 in 1990) per person year. Home and environmental modifications, special furniture, and vehicle modifications are included in the extraordinary costs category and Fujuira depreciated these expenses over a 3- to 5-year period, except for home modifications, which were depreciated over a 10-year period. Birenbaum also reported on the costs of special programs (after-school care, day and weekend care, and summer schools and camps) of $342 ($200 in 1990) per person per year.2 I include $2,329 (in 2003 dollars) as an annual estimate of other costs in the calculations of direct nonmedical costs.

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 490 Wednesday, March 29, 2006 3:54 PM

490

Understanding Autism: From Basic Neuroscience to Treatment

Summary Based on the estimates of direct nonmedical care costs presented in the literature that are reviewed in the preceding text, I estimate the population-weighted annual per capita costs, measured in 2003 dollars, of direct nonmedical costs for individuals with autism to be, on average, $38,253 (men) and $38,601 (women) for those with lower levels of disability to $43,679 (men) and $43,828 (women) for those with higher levels of disability.

INDIRECT COSTS As discussed earlier, indirect costs represent the value of lost productivity due to autism. In this section, I calculate the lost earnings and household services of people with autism and their parents, using forgone earnings based on the human capital approach. Although individuals such as other family members or friends may also have suffered productivity losses,3,61 these are not included here because of a lack of data and because of difficulty making assumptions and because most nonmedical care is provided by the family, most often the parents.3,49,61 Productivity Losses of People with Autism In order to calculate the lost productivity of people with autism, I need to assume an average level of earnings and benefits, the proportion of people in the labor force, and a length of working life. Drawing on standard average work life expectancies for all men and all women from the forensic economics literature, I assume that men who start working at age 23 (see Subsection titled “Supported Employment”) will work until age 57 and women will work until age 53.59 I assume that 35% of adults with lower levels of disability and 10% of adults with higher levels of disability work in a supported work environment. For each age from 23 to 57 for men and from 23 to 53 for women, the average income and benefits calculated using data from Table 696 and Table 628 of the Statistical Abstract of the United States15 combined with estimates of age- and sex-specific labor force participation rates62 and estimated productivity growth rates from the Congressional Budget Office14 were used to determine the average total earnings and benefits at each age. Those estimates are adjusted for the fact that some adults with autism are in supported work environments. Finally, an estimate of the lost value of sex-specific household services is added.7,63 These estimates do not account for the effects of taxes or lost leisure time, and vary by age, but for 23-year-olds the average value of lost productivity is $58,043 for men with low levels of disability and $59,498 for those with high levels of disability and $41,779 for women with low levels of disability and $43,612 for those with high levels of disability (in 2003 dollars) over their lifetimes. Productivity Losses of Parents of People with Autism We also make the assumption that parents suffer productivity losses due to caring for their children with autism, which do not necessarily decrease as children grow to adulthood. Based on data in Järbrink and Knapp3 and Birenbaum and Cohen,61

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 491 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

491

I assume that fathers of children with lower levels of disability are unemployed 10% of the time (and work full-time during the remaining 90%) and that mothers are unemployed 55% of time (and are working half-time 25% of the time and full-time 20%). I further assume that fathers of children with higher levels of disability are unemployed 20% (and work full-time during the remaining 80%) of the time and mothers are unemployed 60% of time (and are working half-time 30% of the time and full-time 10%). Based on these assumptions and based on the same average earnings, benefits, productivity growth, and labor force participation rates as those used for individuals with autism, along with the appropriate work life expectances, I estimate that over the course of their lifetimes, parents’ lost productivity is about $39,681 (for parents of children with lower levels of disability) and $46,033 (for parents of children with higher levels of disability), measured in 2003 dollars. These estimates assume a household with both a mother and a father and, of course, estimates will differ based on different family configurations. The costs in 2003 differ in subsequent years owing to the effects of productivity growth and changes in the labor force participation rates as those parents age. Summary Based on the assumptions presented earlier for people with autism and their parents, I estimate the total population-weighted annual indirect costs due to lost productivity to range from a minimum of $39,681 for both men and women with low levels of disability and $46,033 for men and women with high levels of disabilities, to a maximum of $119,290 (low disability) and $129,785 (high disability) for men and $102,677 (low disability) and $113,549 (high disability) for women. As before, these estimates vary from year to year owing to the effects of productivity, growth, and changes in the labor force participation rates.

RESULTS AND SENSITIVITY ANALYSES After combining all of data and assumptions presented previously, the average discounted per capita societal lifetime cost for caring for an individual with autism, in 2003 dollars, is about $3 million for people with lower levels of disability and about $3.3 million for people with higher levels of disability (Table 20.1, column 1 and column 2). The overall discounted total lifetime cost, accounting for the ratio of low to high levels of disability is close to $3.2 million per person (Table 20.1, column 3). The total discounted annual cost of caring for all individuals with autism, assuming a prevalence rate of 27.5 per 10,000 among 3-year-old children (when the diagnosis is usually made) is $34.8 billion (Table 20.1, column 4). In other words, the total societal cost of caring for all children 3 years old over their lifetimes is $34.8 billion measured in 2003 dollars. If the assumptions underlying these estimates remain stable, we as a society will spend $35 billion every year in lost productivity and in providing for the medical and nonmedical needs for each year’s cohort of people with autism. Based on the data and assumptions presented here, indirect costs account for 59% of the total costs (Table 20.1, column 4 and column 5), with the value of own lost

© 2006 by Taylor & Francis Group, LLC

Total Annual Cost ($) (Millions) Percentage of Total (4) (5)

39,034 3,222 4,526 2,391 198,395 36,237 9,738 2,505 296,048

39,034 3,222 8,120 3,074 215,660 36,237 9,738 2,505 317,590

39,034 3,222 6,179 2,705 206,337 36,237 9,738 2,505 305,957

429 35 68 30 2,270 399 107 28 3,366

1.2 0.1 0.2 0.1 6.5 1.1 0.3 0.1 9.7

Direct Nonmedical Costs Child care Adult care Respite care/family support Home/car modifications Special education Supported employment Other Total for direct nonmedical costs

53,865 614,659 16,003 2,389 150,483 26,696 51,530 915,625

99,729 717,989 20,038 2,389 150,483 10,719 51,530 1,052,877

74,962 662,191 17,859 2,389 150,483 19,347 51,530 978,761

825 7,284 196 26 1,655 213 567 10,766

2.4 21.0 0.6 0.1 4.8 0.6 1.6 31.0

961,152 838,371 1,799,523 3,011,196

982,723 982,336 1,965,059 3,335,526

971,075 904,595 1,875,670 3,160,388

10,682 9,951 20,632 34,764

30.7 28.6 59.3 100.0

Indirect Costs Own lost productivity Parents’ lost productivity Total for indirect costs Total

Note: Assumes a prevalence rate of autism of 27.5 per 10,000; a cohort size of 4 million children 3 yr of age; a male-to-female ratio of 4:1; a low disability rate of 54%; average life expectancies of 66 yr for men with autism and 65 yr for women with autism; average work life expectancies of 57 yr for men with autism who work; and average work life expectancies of 53 yr for women with autism who work. See text for further assumptions. © 2006 by Taylor & Francis Group, LLC

Understanding Autism: From Basic Neuroscience to Treatment

Direct Medical Costs Physician services Dental Medications Complementary and alternative therapies Behavioral therapies Hospital/emergency services Allied health, equipment, home health Medically related travel Total for direct medical costs

2732_C020.fm Page 492 Wednesday, March 29, 2006 3:54 PM

Per Capita Lifetime Costs ($) Low Disability High Disability Overall (1) (2) (3)

492

TABLE 20.1 Per Capita Lifetime and Total Annual Costs of Caring for Individuals with Autism in the U.S. (in 2003 Dollars)

2732_C020.fm Page 493 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

493

productivity slightly larger than parental lost productivity. Nonmedical costs account for 31% of the total costs, with adult care and special education being the two highest nonmedical components of cost, and medical costs account for 10% of the total costs, with most categories no larger than 1.2% of total cost and behavioral therapies, supports being the largest medical cost, accounting for 6.5% of the total costs. As the total annual cost of $34.8 billion is an estimate that depends on its underlying assumptions, it is important to explore how sensitive that estimate is to those assumptions. Some key parameters of the model were individually varied and the resulting total annual costs are presented in column 3 and column 4 of Table 20.2. Most of the parameters were varied between a lower and upper bound. For other parameters, only one alternate calculation method or value was used. For example, the prevalence of autism was varied between 10 and 60 cases per 10,000 based on data presented in the literature.19,64 At a prevalence rate of 10/10,000 the total annual cost decreases from $34.8 billion to $12.6 billion, but at a prevalence of 60/10,000 it increases to $75.8 billion. As most of the sources of the underlying data did not present ranges, I vary those costs down and up by 25%, or I halve or double the values (values close to 100% were used if doubling would result in proportions greater than 100%). Average income is varied by ±10% and indirect costs are varied by ±50% as suggested by Glied.65 The bounds of other parameters such as the ratio of low to high levels of disability or of the proportion of individuals with autism that are employed are based on some of the variation I observed in different sources in the literature. Total annual costs were also recomputed assuming full average life and work life expectancies, assuming no wage growth, and after varying the discount rate between 2 and 5% as suggested by the Panel on Cost-Effectiveness in Health and Medicine.4 The total annual costs are generally not sensitive to changes in the direct and indirect components of cost- or labor-related parameters such as the proportion of people that are employed, the average income, or the ratio of low to high disability. However, the total cost estimate is, of course, sensitive to the prevalence rate, the discount rate, and less so to assumptions about wage growth. When simultaneously varying all parameters listed in Table 20.2 between their lower and upper bounds, the total cost of autism ranges from $8.4 billion to $92.7 billion.

DISCUSSION The total annual societal cost of caring for and treating people with autism in the U.S. over their lifetimes is about $35 billion and, depending on underlying assumptions, could range from $13 billion to $76 billion. These are highly conservative estimates and probably best represent lower bounds on the true societal costs. The estimates presented here underestimate and exclude many potential elements such as the following: • •

Legal costs that families may incur to secure services66 The value of lost productivity of individuals other than parents

© 2006 by Taylor & Francis Group, LLC

Value Total Annual Cost ($Millions) Based on… Lower and Upper Bounds (%) Parameter

Proportion of low- disability autism Life expectancy Employment and Economic Factors Rate of employment among low disability Rate of employment among high disability Average income Discount rate Productivity growth Direct Medical Costs Physician services Dental Medications CAM therapies Therapies/behavioral support Hospital services/ER Allied health, equipment, home health Medically related travel

© 2006 by Taylor & Francis Group, LLC

(2)

Lower Bound (3)

Upper Bound (4)

10 per 10,000 25 —

60 per 10,000

12,642

75,849

75 72 for men; 74 for women

35,798

34,015 35,350

20 5 −10 2 0

50 20 +10 5 —

34,280 34,783 32,767 42,173 30,616

34,684 34,755 36,762 25,672

−25 −25 −25 −25 −25 −25 −25 −25

+25 +25 +25 +25 +25 +25 +25 +25

34,657 34,755 34,747 34,757 34,197 34,665 34,373 34,757

34,872 34,773 34,781 34,772 35,332 34,864 34,791 34,771

Understanding Autism: From Basic Neuroscience to Treatment

Epidemiologic Factors Prevalence of autism

(1)

2732_C020.fm Page 494 Wednesday, March 29, 2006 3:54 PM

494

TABLE 20.2 Sensitivity Analysis of Total Annual Costs of Caring for Individuals with Autism in the U.S. (in 2003 Dollars)

+25 +25 75 +25 +25 +25 +25 +25

34,558 32,943 34,972 34,715 34,758 34,350 34,711 34,623

34,970 36,585 34,556 34,813 34,771 35,178 34,817 34,906

−50 −50

+50 +50

29,423 29,789

40,105 39,740

5 25 10 30

20 80 40 80

34,337 33,654 34,036 33,818

35,620 36,034 36,221 35,846

For these sensitivity analyses, mothers’ employment rate (= 1-unemployment rate) is split evenly between full- and part-time.

495

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 495 Wednesday, March 29, 2006 3:54 PM

*

−25 −25 25 −25 −25 −25 −25 −25

The Costs of Autism

Direct Nonmedical Costs Child care Adult care Proportion of adults at home Respite care/family support Home/car modifications Special education Supported employment Other nonmedical Indirect Costs Own lost productivity Parents’ lost productivity Unemployment rate among: Fathers of low-disability children Mothers of low-disability children* Fathers of high-disability children Mothers of high-disability children*

2732_C020.fm Page 496 Wednesday, March 29, 2006 3:54 PM

496

Understanding Autism: From Basic Neuroscience to Treatment

•

• • • •

The value of psychological stress that people with autism and their families may experience,67–69 along with any associated physical and mental health services utilized as a result70 The costs of genetic testing The full costs of alternative therapies that families may pursue, such as special diets or relationship-based therapies The costs associated with adverse outcomes of potentially dangerous treatment modalities Any effects of changes in immunization-avoidance behaviors71

It is also likely to underestimate many of the costs presented in older sources that are based on health care technologies and standards of care that may have changed significantly over time (for example, currently many children with autism may be seen by specialists or in specialized clinics). It is also likely to underestimate the costs in the absence of barriers to care for children with autism and other developmental disorders.72,73 The value of indirect costs, which make up 59% of total costs, are prone to errors, and may be overestimated or underestimated.65 These estimates do not include the value of autism research and advocacy budgets, which are likely to be substantial. For example, the National Institutes of Health research budget for autism has grown from $22 million in fiscal year 1997 to $70 million in fiscal year 2003.74 My estimate of $3.2 million in per capita lifetime costs is roughly consistent with data presented by Järbrink and Knapp,3 and my estimate of $35 billion in total costs is roughly consistent with the value presented in an Autism Society of America newsletter of $31 billion.66 It is also consistent with estimates of other conditions such as mental retardation ($51 billion; prevalence 103/10,000),75 anxiety ($47 billion; prevalence 12.6/10,000),76 and schizophrenia ($33 billion; prevalence 1.1/10,000).76 The societal cost of depression is twice that of autism (about $83 billion) even though the number of Americans suffering from depression is orders of magnitude greater than the number with autism (prevalence 1,620/10,000).77 Gerlai and Gerlai estimated that autism represents a health care market (44 million patient years) almost as large as that for Alzheimer’s disease (54 million patient years),78 yet the total societal cost of Alzheimer’s disease is approximately triple that of autism, at $91 billion (in 2003 dollars calculated from Ernst and Hay79). As many sources of data used for this chapter tended not to differentiate between the different autism spectrum disorders (or simply used the term autism or autistic), in order to reflect that literature, I used the term autism throughout this chapter in an inclusive manner to also mean any and all disorders that are part of the spectrum. However, given the nature of many of the cost categories described earlier, especially for nonmedical and indirect costs, it is likely that those costs reflected individuals at the more disabled end of the spectrum. Furthermore, older sources, such as the Birenbaum volume2, may have been more likely to only include low-functioning children and individuals in their definitions of autism. As such, the cost estimates presented in the section titled “Results and Sensitivity Analyses” and in Table 20.1 are probably high for those with Asperger syndrome, but should reflect the entire population of people with autism spectrum disorders.

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 497 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

497

Varying the proportion of individuals with lower and higher levels of disability in the sensitivity analyses can give some idea as to how sensitive the results are to definitional issues. In this chapter I have attempted to synthesize the available literature on autism and, in some cases, developmental disorders in general, and to derive some reasonable estimates of the costs of autism in the U.S. These estimates represent the best that can be done given the far-from-optimal sources of data available to date. Many sources of data were potentially outdated or based on small or nonrepresentative samples. As such, these estimates should be viewed with some caution. The results of the sensitivity analyses can provide the reader some sense of the reasonableness of the assumptions and sources of data. For example, although behavioral therapies, adult care, special education, and own and parental lost productivity represent the five largest components of cost, the overall estimate of the societal cost of autism is not very sensitive to these inputs. On the other hand, as expected, the discount rate and the assumed prevalence of autism are very important inputs.

DIRECTIONS

FOR

FUTURE RESEARCH

Because the model that produces the estimates given here is built upon simplifying assumptions and some rather incomplete (and sometimes decades old) information, in order to improve upon these estimates it would be important to pursue a research agenda of both carefully and systematically documenting the costs of caring for individuals with autism in the U.S. This can be accomplished by using a variety of methods, including prospectively tracking the life experiences of individuals with autism and their families. As people with autism receive services from a wide variety of sources that are paid by different agencies, special efforts should be directed to enumerating these sources and linking them together into a consolidated data collection system. In addition to tracking and estimating the financial costs of autism, effort should be devoted to using standard methods of economic evaluation to collect data on health-adjusted quality-of-life measures. Financial cost data in conjunction with data on health-related measures can substantially contribute to the health policy debate on how to prioritize health care funding and on how to decide what types of autism treatments are most cost-effective.4 By prospectively collecting cost data (both direct and indirect) and by collecting data on health states, we may be able to assemble a more complete picture of the societal costs, both financial and nonfinancial, of caring for those members of our society with autism.

ACKNOWLEDGMENTS The author would like to thank Sherry Glied, James Heckman, Marie McCormick, Judith Palfrey, Leonard Rappaport, and four anonymous reviewers for helpful comments on earlier drafts of this chapter. All conclusions and errors are those of the author.

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 498 Wednesday, March 29, 2006 3:54 PM

498

Understanding Autism: From Basic Neuroscience to Treatment

REFERENCES 1. Jacobson, J.W., Mulick, J.A., and Green, G., Cost-benefit estimates for early intensive behavioral intervention for young children with autism — general model and single state case, Behavioral Interventions 13, 201–226, 1998. 2. Birenbaum, A., Guyot, D., and Cohen, H.J., Health Care Financing for Severe Developmental Disabilities, American Association on Mental Retardation, Washington, D.C., 1990. 3. Järbrink, K. and Knapp, M., The economic impact of autism in Britain, Autism 5(1), 7–22, 2001. 4. Gold, M.R., Siegel, J.E., Russell, L.B., and Weinstein, M.C., Cost-Effectiveness in Health and Medicine, Oxford University Press, New York, 1996. 5. Rice, D.P., Hodgson, T.A., and Kopstein, A.N., The economic costs of illness: a replication and update, Health Care Financing Review 7(1), 61–80, 1985. 6. Pindyck, R.S. and Rubinfeld, D.L., Microeconomics, 5th ed., Prentice Hall, Upper Saddle River, NJ, 2000. 7. Waitzman, N.J., Scheffler, R.M., and Romano, P.S., The Costs of Birth Defects: Estimates of the Value of Prevention, University Press of America, Lanham, MD, 1996. 8. Hodgson, T.A. and Meiners, M.R., Cost-of-illness methodology: a guide to current practices and procedures, Milbank Memorial Fund Quarterly 60(3), 429–462, 1982. 9. Hodgson, T.A., The state of the art of cost-of-illness estimates, Advances in Health Economics and Health Services Research 4, 129–164, 1983. 10. Rice, D.P., Cost of Illness Studies: Implications for Aging-Related Diseases, New York: International Longevity Center, 2005, http://www.ilcusa.org/_lib/pdf/cost.pdf (accessed September 15, 2005). 11. National Center for Health Statistics, Centers for Disease Control and Prevention, NCHS Defintions: Synthetic Cohort, 2004, http://www.cdc.gov/nchs/datawh/ nchsdefs/syntheticcohort.htm (accessed January 5, 2005). 12. Curran, A.L., Sharples, P.M., White, C., and Knapp, M., Time costs of caring for children with severe disabilities compared with caring for children without disabilities, Developmental Medicine and Child Neurology 43(8), 529–533, 2001. 13. Järbrink, K., Fombonne, E., and Knapp, M., Measuring the parental, service and cost impacts of children with autistic spectrum disorder: a pilot study, Journal of Autism and Developmental Disorders 33(4), 395–402, 2003. 14. Congressional Budget Office, The Budget and Economic Outlook: An Update. 2003. http://www.cbo.gov/showdoc.cfm?index=4493&sequence=3 (accessed January 4, 2005). 15. U.S. Department of Commerce Economics and Statistics Administration, Statistical Abstract of the United States, Bureau of the Census, Washington, D.C., 2004. 16. Board of Governors of the Federal Reserve System, Foreign Exchange Rates Historical Data Series H.10, 2004, http://www.federalreserve.gov/releases/H10/hist/ (accessed January 4, 2005). 17. Agency for Healthcare Research and Quality, The Medical Expenditure Panel Survey, 2005, http://www.ahrq.gov/data/ mepsix.htm (accessed January 3, 2005). 18. Centers for Disease Control and Prevention, National Center for Health Statistics, The National Health Interview Survey, 2004, http://www.cdc.gov/nchs/nhis.htm (accessed January 3, 2005). 19. Fombonne, E., Epidemiological surveys of autism and other pervasive developmental disorders: an update, Journal of Autism and Developmental Disorders 33(4), 365–382, 2003.

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 499 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

499

20. Gillberg, C., Outcome in autism and autistic-like conditions, Journal of the American Academy of Child and Adolescent Psychiatry 30(3), 375–382, 1991. 21. Fombonne, E., The life expectancy of children diagnosed with a pervasive developmental disorder, Journal of Autism and Developmental Disorders 33(3), 361, 2003. 22. Shavelle, R.M. and Strauss, D., Comparative mortality of persons with autism in California, 1980–1996, Journal of Insurance Medicine 30 (4), 220–225, 1998. 23. Shavelle, R.M., Strauss, D.J., and Pickett, J., Causes of death in autism, Journal of Autism and Developmental Disorders 31(6), 569–576, 2001. 24. Arias, E., United States Life Tables, 2000, National Vital Statistics Reports, Vol. 51, No. 3, Hyattsville, MD: National Center for Health Statistics, 2002. 25. Population Division, Annual Estimates of the Population by Sex and Five-Year Age Groups for the United States: April 1, 2000 to July 1, 2003 (NC-EST2003-1). Table 1, U.S. Census Bureau 2004, http://www.census.gov/popest/national/asrh/ NC-EST2003/NC-EST2003-01.pdf (accessed on January 4, 2005). 26. Hamilton, B.E., Sutton, P.D., and Ventura, S.J., National Vital Statistics Reports; Vol. 51, No. 2, Revised birth and fertility rates for the 1990s and new rates for Hispanic populations, 2000 and 2001: Hyattsville, MD: National Center for Health Statistics. 2003. 27. Aman, M.G., Van Bourgondien, M.E., Wolford, P.L., and Sarphare, G., Psychotropic and anticonvulsant drugs in subjects with autism: prevalence and patterns of use, Journal of the American Academy of Child and Adolescent Psychiatry 34(12), 1672–1681, 1995. 28. Martin, A., Scahill, L., Klin, A., and Volkmar, F.R., Higher-functioning pervasive developmental disorders: rates and patterns of psychotropic drug use, Journal of the American Academy of Child and Adolescent Psychiatry 38(7), 923–931, 1999. 29. Aman, M.G., Lam, K.S., and Collier-Crespin, A., Prevalence and patterns of use of psychoactive medicines among individuals with autism in the autism society of Ohio, Journal of Autism and Developmental Disorders 33(5), 527–534, 2003. 30. Green, V.A., Pituch, K.A., Itchon, J., Choi, A., O’Reilly, M., and Sigafoos, J., Internet survey of treatments used by parents of children with autism, Research in Developmental Disabilities, 27(1), 70–84, 2006. 31. Langworthy-Lam, K.S., Aman, M.G., and Van Bourgondien, M.E., Prevalence and patterns of use of psychoactive medicines in individuals with autism in the autism society of North Carolina, Journal of Child and Adolescent Psychopharmacology 12(4), 311–321, 2002. 32. Levy, S.E., Mandell, D.S., Merhar, S., Ittenbach, R.F., and Pinto-Martin, J.A., Use of complementary and alternative medicine among children recently diagnosed with autistic spectrum disorder, Journal of Developmental and Behavioral Pediatrics 24(6), 418–423, 2003. 33. Nickel, R.E., Controversial therapies for young children with developmental disabilities, Infants and Young Children 8, 29–40, 1996. 34. Eisenberg, D.M., Kessler, R.C., Foster, C., Norlock, F.E., Calkins, D.R., and Delbanco, T.L., Unconventional medicine in the United States. Prevalence, costs, and patterns of use, New England Journal of Medicine 328(4), 246–252, 1993. 35. Yussman, S.M., Ryan, S.A., Auinger, P., and Weitzman, M., Visits to complementary and alternative medicine providers by children and adolescents in the United States, Ambulatory Pediatrics 4, 429–435, 2004. 36. Shea, V., A perspective on the research literature related to early intensive behavioral intervention (Lovaas) for young children with autism, Autism 8(4), 349–367, 2004.

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 500 Wednesday, March 29, 2006 3:54 PM

500

Understanding Autism: From Basic Neuroscience to Treatment

37. Filipek, P.A., Accardo, P.J., Baranek, G.T., Cook, E.H., Jr., Dawson, G., Gordon, B., Gravel, J.S., Johnson, C.P., Kallen, R.J., Levy, S.E., Minshew, N.J., Ozonoff, S., Prizant, B.M., Rapin, I., Rogers, S.J., Stone, W.L., Teplin, S., Tuchman, R.F., and Volkmar, F.R., The screening and diagnosis of autistic spectrum disorders, Journal of Autism and Developmental Disorders 29(6), 439–484, 1999. 38. American Academy of Pediatrics, The pediatrician’s role in the diagnosis and management of autistic spectrum disorder in children, Pediatrics 107(5), 1221–1226, 2001. 39. Butter, E.M., Wynn, J., and Mulick, J.A., Early intervention critical to autism treatment, Pediatric Annals 32(10), 677–684, 2003. 40. National Institute of Mental Health, National Insititutes of Health, Autism Spectrum Disorders (Pervasive Developmental Disorders), 2004, http://www.nimh.nih.gov/ publicat/autism.cfm#treatment (accessed January 5, 2005). 41. Marcus, L.M., Rubin, J.S., and Rubin, M.A., Benefit-cost analysis and autism services: a response to Jacobson and Mulick, Journal of Autism and Developmental Disorders 30(6), 595–598, 2000. 42. Lovaas, O.I., Behavioral treatment and normal educational and intellectual functioning in young autistic children, Journal of Consulting and Clinical Psychology 55(1), 3–9, 1987. 43. McEachin, J.J., Smith, T., and Lovaas, O.I., Long-term outcome for children with autism who received early intensive behavioral treatment, Amercan Journal on Mental Retardation 97(4), 359–372, 1993. 44. Hildebrand, D.G., Cost-Benefit Analysis of Lovaas Treatment for Autism and Autism Spectrum Disorder (ASD), Vancouver, British Columbia: Columbia Pacific Consulting, 1999. 45. Walsh, K.K., Kastner, T., and Criscione, T., Characteristics of hospitalizations for people with developmental disabilities: utilization, costs, and impact of care coordination, American Journal on Mental Retardation 101(5), 505–520, 1997. 46. Committee of Ways and Means, Green Book: Background Material and Data on the Programs within the Jurisdiction of the Committee on Ways and Means. Washington, D.C.: U.S. House of Representatives. 2004. http://www.waysandmeans.house.gov/ media/pdf/greenbook2003/Section9.pdf (accessed January 2, 2005). 47. Baghdadli, A., Pascal, C., Grisi, S., and Aussilloux, C., Risk factors for self-injurious behaviours among 222 young children with autistic disorders, Journal of Intellectual Disability Research 47(Pt. 8), 622–627, 2003. 48. Aman, M.G., Management of hyperactivity and other acting-out problems in patients with autism spectrum disorder, Seminars in Pediatrics Neurology 11(3), 225–228, 2004. 49. Fujiura, G.T., Roccoforte, J.A., and Braddock, D., Costs of family care for adults with mental retardation and related developmental disabilities, American Journal on Mental Retardation 99(3), 250–261, 1994. 50. Yeargin-Allsopp, M., Rice, C., Karapurkar, T., Doernberg, N., Boyle, C., and Murphy, C., Prevalence of autism in a U.S. metropolitan area, Journal of the American Medical Association 289(1), 49–55, 2003. 51. Parrish, T., Harr, J., Wolman, J., Anthony, J., Merickel, A., and Esra, P., State Special Education Finance Systems, 1999–2000. Part II: Special Education Revenues and Expenditures, Palo Alto, CA: Center for Special Education Finance. 2004, http://csef. air.org/publications/csef/state/statepart2.pdf (accessed December 27, 2004). 52. Mawhood, L. and Howlin, P., The outcome of a supported employment scheme for high functioning adults with autism or Asperger syndrome, Autism 3(3), 229–254, 1999.

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 501 Wednesday, March 29, 2006 3:54 PM

The Costs of Autism

501

53. Gilson, S.F., Case management and supported employment: a good fit, Journal of Case Management 7(1), 10–17, 1998. 54. Keel, J.H., Mesibov, G.B., and Woods, A.V., TEACCH-supported employment program, Journal of Autism and Developmental Disorders 27(1), 3–9, 1997. 55. Rusch, F.R., Conley, R.W., and McCaughrin, B., Benefit-cost analysis of supported employmental in Illinois, Journal of Rehabilitation 59(2), 31–36, 1993. 56. Capo, L.C., Autism, employment, and the role of occupational therapy, Work 16(3), 201–207, 2001. 57. Bureau of Labor Statistics, Occupational Outlook Handbook, 2004–2005 edition, U.S. Department of Labor, 2004, http://www.bls.gov/oco/ocos078.htm (accessed December 27, 2004). 58. Mank, D.M., National implementation of supported employment, in Supported Employment: Models, Methods, and Issues, Rusch, F.R. Eds., Sycamore, IL, 1990, pp. 289–300. 59. Skoog, G.R. and Ciecka, J.E., The Markov (increment-decrement) model of labor force activity: extended tables of central tendency, variation, and probability intervals, Jounal of Legal Economics Spring/Summer, 23–87, 2001. 60. Heal, L.W., McCaughrin, W.B., and Tines, J.J., Methodological nuances and pitfalls of benefit-cost analysis: a critique, Research in Developmental Disabilities 10(2), 201–212, 1989. 61. Birenbaum, A. and Cohen, H.J., On the importance of helping families: policy implications from a national study, Mental Retardation 31(2), 67–74, 1993. 62. Congressional Budget Office, CBO’s Projections of the Labor Force, 2004, http://www.cbo.gov/showdoc.cfm?index=5803&sequence=0 (accessed January 4, 2005). 63. Tinari, F.D., Household services: toward a more comprehensive measure, Journal of Forensic Economics 11(3), 253–265, 1998. 64. Fombonne, E., Epidemiological trends in rates of autism, Molecular Psychiatry 7(Suppl. 2), S4–6, 2002. 65. Glied, S., Estimating the indirect cost of illness: an assessment of the forgone earnings approach, American Journal of Public Health 86, 1723–1728, 1996. 66. Maltby, J., The costs of autism: More than meets the eye, Advocate 33(6), 12–16, 2000, http://www.autisminfo.com/Advocate.pdf (accessed December 24, 2004). 67. Pilowsky, T., Yirmiya, N., Doppelt, O., Gross-Tsur, V., and Shalev, R.S., Social and emotional adjustment of siblings of children with autism, Journal of Child Psychology and Psychiatry 45(4), 855–865, 2004. 68. Rivers, J.W. and Stoneman, Z., Sibling relationships when a child has autism: marital stress and support coping, Journal of Autism and Developmental Disorders 33(4), 383–394, 2003. 69. McIntyre, L.L., Blacher, J., and Baker, B.L., Behavior/mental health problems in young adults with intellectual disability: the impact on families, Journal of Intellectual Disabilities Research 46(Pt. 3), 239–249, 2002. 70. Hare, D.J., Pratt, C., Burton, M., Bromley, J., and Emerson, E., The health and social care needs of family carers supporting adults with autistic spectrum disorders, Autism 8(4), 425–444, 2004. 71. Folstein, S.E. and Rosen-Sheidley, B., Genetics of autism: complex aetiology for a heterogeneous disorder, Nature Reviews Genetics 2(12), 943–955, 2001. 72. Krauss, M.W., Gulley, S., Sciegaj, M., and Wells, N., Access to specialty medical care for children with mental retardation, autism, and other special health care needs, Mental Retardation 41(5), 329–339, 2003.

© 2006 by Taylor & Francis Group, LLC

2732_C020.fm Page 502 Wednesday, March 29, 2006 3:54 PM

502

Understanding Autism: From Basic Neuroscience to Treatment

73. Schultz, S.T., Shenkin, J.D., and Horowitz, A.M., Parental perceptions of unmet dental need and cost barriers to care for developmentally disabled children, Pediatric Dentistry 23(4), 321–325, 2001. 74. Statement by Stephen L. Foote, Ph.D. Director, Division of Neuroscience and Basic Behavioral Science, National Institute of Mental Health, National Institutes of Health on NIH Autism Research efforts given to House Committee on Government Reform, April 18, 2002, Washington, D.C., http://www.hhs.gov/asl/testify/t020418a.html (accessed on January 5, 2005). 75. Honeycutt, A., Dunlap, L., Chen, H., al Homsi, G., Grosse, S., and Schendel, D., Economic costs associated with mental retardation, cerebral palsy, hearing loss, and vision impairment — United States, 2003, Morbidity and Mortality Weekly Report 53(3), 57–59, 2004. 76. Rice, D.P. and Miller, L.S., Health economics and cost implications of anxiety and other mental disorders in the United States, British Journal of Psychiatry Supplement (34), 4–9, 1998. 77. Kessler, R.C., Berglund, P., Demler, O., Jin, R., Koretz, D., Merikangas, K.R., Rush, A.J., Walters, E.E., and Wang, P.S., The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R), Journal of the American Medical Association 289(23), 3095–3105, 2003. 78. Gerlai, J. and Gerlai, R., Autism: a large unmet medical need and a complex research problem, Physiology and Behavior 79, 461–470, 2003. 79. Ernst, R.L. and Hay, J.W., The U.S. economic and social costs of Alzheimer’s disease revisited, American Journal of Public Health 84(8), 1261–1264, 1994.

© 2006 by Taylor & Francis Group, LLC