Forearm EMG response activity during motor performance in [PDF]

AMERICAN JOURNAL OF INDUSTRIAL MEDICINE 41:406–419 (2002)

Forearm EMG Response Activity During Motor Performance in Individuals Prone to Increased Stress Reactivity Gerard P. Van Galen,1
Martijn L.T.M. Mu ¨ller,2 Ruud G.J. Meulenbroek,1 and Arend W.A. Van Gemmert3

Background Work-related Upper Extremity Disorders (WRUEDs) are conceived of as a multifactorial syndrome caused by the effects of excessive repetitive motions, sustained static postures, and muscular stiffness. Our aim is to test an etiological model derived from a theory by Van Galen and Van Huygevoort [2000] Biol Psychol 51:151–171. The theory holds that physical, emotional, and psychosocial stressors enhance muscular stiffness as a compensatory filtering of impoverished signal-to-noise ratios in the motor system. High individual levels of arousal, as measured by Spielberger et al. [1970], State and Trait Anxiety Test would further enhance a subject’s predisposition to react with stiffness responses in conditions of stress. Methods Ten participants with a high- and 10 with a low trait-anxiety score performed a computer task involving series of fast but well-dosed accelerations of the forearm along the surface of a digitizer. To induce cognitive stress a tone had to be remembered simultaneously with the aiming task. Pen-tip displacements and surface electromyographic (EMG) signals were recorded from four forearm muscles. Results Memory load did not affect error rates but produced shorter reaction times and prolonged movement times. EMG data show that under stress overall levels of neuromotor activation were enhanced. High-anxious participants exhibited higher cocontraction levels than low-anxious participants. Conclusions The findings support the view that stress and muscular tension are closely related and may provide a clue to the origin of WRUEDs. Am. J. Ind. Med. 41:406–419, 2002. ß 2002 Wiley-Liss, Inc. KEY WORDS: anxiety; surface EMG; motor control; muscular cocontraction; musculoskeletal diseases etiology; neuromotor activation; neuromotor noise; psychophysiology stress

INTRODUCTION

1

University of Nijmegen, Nijmegen Institute for Cognition and Information (NICI), Nijmegen, The Netherlands 2 University of Pittsburgh, Pittsburgh, Pennsylvania 3 Arizona State University, Motor Control Laboratory,Tempe, Arizona *Correspondence to: Gerard P.Van Galen, Nijmegen Institute for Cognition and Information (NICI), P.O. Box 9104, 6500 HE Nijmegen,The Netherlands. E-mail: [email protected] Accepted 28 November 2001 DOI 10.1002/ajim.10051. Published online in Wiley InterScience (www.interscience.wiley.com)


2002 Wiley-Liss, Inc.

At present, in Western industrialized countries, musculoskeletal disease is responsible for a considerable loss of labor due to work absence, sickness, and disablement; this situation has serious economic, social, and individual implications. Among other symptoms characteristic of musculoskeletal disease, complaints about pains in the neck, shoulder, arm, and hand region have increasingly caused concern in our society, especially since they seem to be related to major industrial changes and innovations [Dembe,

Forearm EMG Response and Stress Reactivity

1999]. The terminology most often used for this group of complaints is work-related upper extremity disorders (WRUEDs) although the term repetitive strain injuries (RSI) is also used. In the Netherlands, WRUEDs have caused functional complaints and pain in 20–40 percent of the working population, depending on the occupational category and affected anatomical subregion [De Zwart et al., 1997; Otten et al., 1998; Blatter and Bongers, 1999; Picavet et al., 2000]. Although a comparison of the demographic statistics of different countries is difficult due to the diverse social systems, other studies paint a comparable or even bleaker picture [Kuorinka et al., 1995; Harvey and Peper, 1997; Smith et al., 1997; Silverstein et al., 1998; Levenstein, 1999]. Since an early study by Ferguson [1971a,b] on pain in the shoulders and arms of telegraphists, similar studies have been published on the risks for WRUEDs in specific professions such as musicians [Fry, 1986; Moran, 1992; Moulton and Spence, 1992], postal workers [Gomer et al., 1987], visual display workers [Bergqvist et al., 1995; Schreibers et al., 1995; De Ridder et al., 1997], and industrial workers [Bernard, 1997]. The complaints most frequently reported are a deep, burning pain in the hands, arms, shoulders, and/or neck accompanied by either hypersensitivity or the opposite, viz. numbness. Patients further complain of tingling sensations, spasms, cramps, increased muscular tension, subcutaneous swelling, dermatographia, diminished passive motility of arms and hands, fatigue, function loss, and, in advanced states, loss of control in the execution of everyday routines. Symptoms mostly develop gradually and may persist for years. Until a few years ago, there was a tendency in most countries to mainly or even exclusively reserve the diagnosis WRUED for those cases for which an objective medical or physical criterion was available. According to Kiesler and Finholt [1988], however, only a minority of WRUED cases (5%) can be attributed to unambiguous physiological causes, like carpal tunnel syndrome, epicondylitis, tenosynovitis, muscle tissue inflammation, nerve compression, tendinitis or cellutis. The majority of cases lack objective radiological, vascular, or electrodiagnostic signs of pathology. We, therefore, propose to describe WRUEDs as a complaint syndrome rather than as a disease. Many suggestions about the etiology of WRUED find their origin in early studies of writer’s cramp [see e.g., Ferguson, 1971a,b; Bindman and Tibbets, 1977; Hughes and McLellan, 1985; Sheehy and Marsden, 1982; Nakashima et al., 1989; Marsden and Sheehy, 1990; Tempel and Perlmutter, 1993]. Since then, a great variety of potential risk factors for or mechanisms that could lead to WRUED have been identified. In 1996 the U.S. Bureau of Labor Statistics reported 439,000 new cases, which constituted 64% of all occupational illnesses. In the majority of review articles on WRUEDs in the USA [Tyrer, 1994], Australia [Kiesler and Finholt, 1988; Quintner, 1995], and

407

the UK [Reilly, 1995] a comparative stance on the multifactorial etiology of WRUED was adopted. Interestingly, well-controlled prospective and longitudinal studies have contributed to the ergonomically valid findings that the exposure to frequency and force levels of repetitive labor are the main causes of WRUEDs, together with sustained and uneasy postures of limbs and trunk. What is perhaps equally important, is that the latter studies have also furnished substantial evidence for what was suspected long ago [Gowers, 1892 in Tyrer, 1994], namely that psychosocial factors have a mediating role in the pathogenesis of WRUEDs [Bongers et al., 1993; Bernard, 1997; Bongers and Houtman, 2000]. Besides excessive motion in combination with static postures and psychosocial risks, a further category of risk factors relates to individual factors. Research into the extent to which these factors contribute to WRUEDs is more sparse than studies into general epidemiological and ergonomic issues. Dembe [1999] mentioned that the role of gender is ambiguous. As to the personality traits of WRUED patients, in most studies only retrospective measures of individual risk factors have been collected. Until now, none of the studies have uncovered major psychiatric illnesses as contributors to WRUEDs [Windgassen and Ludolph, 1991]. However, it is interesting to note that the latter authors reported that patients showed heightened levels of muscular activity as compared to controls. Other authors that have investigated psychological and personality features of patients with WRUEDs are Helliwell et al. [1992], Hughes and McLellan [1985], and Moulton and Spence [1992]. None of them found major psychological deviations from normalcy that could be linked to WRUED. The only factor that is mentioned in the literature as having a relationship with WRUED is anxiety [Helliwell et al., 1992].

Neuromotor Noise Theory and WRUEDs Summarizing the foregoing, it may be stated that excessive repetitive motions by the upper limbs in combination with static postures of the more proximal musculature together with psychosocial stressors such as time pressure, emotional threat, and external locus of control are the primary ingredients for the etiology of WRUED. At the same time, it should be recognized that as yet it is unclear which biological and psychophysiological mechanisms lead to the interaction between muscular overuse and psychosocial stress. Here, we propose an etiological model which encompasses such seemingly divergent factors as muscle strain, task stress, and trait anxiety. Before introducing the model in more detail, we will briefly summarize the theory on stress and human performance that led us to formulate the model. Details on the theory may be found in Van Galen et al. [1990], Van Galen and Schomaker [1992],

408

Van Galen et al.

Van Gemmert and Van Galen [1997, 1998], and more recently, Van Galen and Van Huygevoort [2000]. .

.

.

.

Human movement is an inherently noisy process. Or more precisely, endpoint accuracy of movement is a dynamic outcome of a stochastic recruitment process of muscle force which is inherently contaminated by nonintentional noise sources such as stochastic variation of the current recruitment signal, reflex-induced oscillations of force output (e.g., stretch reflexes and physiological tremor), feedback and feedforward servo mechanisms, and mechanical oscillations of muscle tissue, tendons, and bones. The signal-to-noise ratio of the neuromotor signal is not only the summed effect of biophysical processes but is also determined by concurrent cognitive control processes. In other words, the prevalent signal-to-noise ratio is the summed outcome of the intentional recruitment of motor signals and the neuromotor activity caused by all simultaneously active stressors. The spatial outcome of a move is the net outcome over time of the stochastic neuromotor signal to the muscles and the filtering capacities of the body. These filtering capacities are supposed to be subject to movement strategies set by the motor control system. Stiffness (amongst other factors) is considered to be a suitable degree of freedom for noise filtering strategies [Van Galen and De Jong, 1995; Van Galen et al., 1996]. Individual state and trait variables may further determine effective signal-to-noise ratios and may require an individual to apply more or less extended filtering of neuromotor signals. The latter statement was specifically developed in the context of forensic research of the dynamics of handwriting when subjects try to imitate signatures and the handwriting of other persons [Van Galen and Van Gemmert, 1996]. The literature on the effects of anxiety on human motor control also suggests that anxiety and external threat are reflected by higher levels of neuromotor noise. Goldstein [1964] speaks, in this context, of generalized muscular tension as an expression of immobilization and defense. More recent research has articulated this view in the sense that electromyographic (EMG) patterns of high-anxious persons are particularly characterized by uncorrelated response bursts [Fridlund et al., 1986]. The latter of the elements mentioned, i.e., non-specific activation or arousal contributing to neuromotor noise, will play a role in our etiological model of WRUEDs as well.

Towards an Etiological Model of WRUEDs We have stressed four risk factors for WRUEDs: (1) excessive motion, (2) static postures, (3) psychosocial and

task stress, and (4) high levels of arousal due to personality traits. We think that all factors must be considered to contribute to the mechanisms through which WRUED develops. In our model it is argued that in the primary phase of their manifestation WRUEDs are caused by inefficient movement strategies due to enhanced levels of cocontraction of agonist and antagonist muscles. These continuously high levels of cocontraction may be the combined result of excessive motion with high spatial and temporal demands [Rempel et al., 1997]. Others have also suggested that failing or disturbed co-activation patterns are the primary cause for WRUEDs [see e.g., Sheehy and Marsden, 1982; Nakashima et al., 1989; Marsden and Sheehy, 1990]. In our model, however, increased cocontraction is in essence viewed as a normal stiffness response of the neuromotor system. According to our neuromotor noise theory task conditions such as environmental stress, dual task load and psychosocial stress provoke a deterioration in overall signal-to-noise levels in the motor control system which, in turn, leads to higher stiffness levels through cocontraction. Because individual state and trait variables such as fatigue and anxiety may enhance base levels of arousal even further, more excessive noise filtering and thus increased cocontraction will be elicited in easily aroused individuals. Especially in combination with continued static postures, this condition will lead to hampered blood flow, poor energy supply, poor carry-off of metabolic products, nerve entrapment, sensations of pain, fatigue, and somesthetic abnormalities. If such conditions occur repeatedly over prolonged periods of time, a secondary phase of the development of the syndrome, most likely provoked by central nervous system processes, is held responsible for the persistence of pain, sensory loss, and dysfunction even if the original triggering activities have been suspended. In our model it is assumed that during the primary phase of the syndrome, the origin of heightened cocontraction and thus inefficient movement is to be found in the very nature of motor control, i.e., an increased stiffness response in situations of physical and psychological strain. According to the theory, the motor system copes with the concomitant impoverished signal-to-noise ratios by increased low-pass filtering of the neuromotor signal which is effected by an enhancement of muscle and limb stiffness through cocontraction. For the extrapolation of the neuromotor noise theory to the field of WRUEDs we think corroborating muscular evidence is necessary. In the present research, therefore, we devised an experiment that allowed us to take direct measurements of the neuromuscular signal by using surface EMG measures in an experimental design with repeated movements in a precise motor control task with a concomitant dual-task load. The task was implemented as a computer game in which precise accelerations of the forearm were elicited by having the participants move a pen along the surface of a digitizer.

Forearm EMG Response and Stress Reactivity

A further ingredient of our model of WRUEDs is the base level of arousal that is characteristic of each individual. To provide more conclusive evidence for the role of personality factors we needed to demonstrate that such factors do indeed contribute to inefficient movement strategies and that these strategies are likely to provoke WRUEDs. To this end, we decided to differentiate between participants with a high score and those with a low score on Spielberger’s State and Trait Anxiety test [Spielberger et al., 1970]. In our experiment we measured chronometric, kinematic as well as neuromuscular variables related to task performance. In the chronometric and kinematic domain, we measured reaction times, movement times, acceleration, and task precision (hits and misses). To test whether the task load (dual-task memory load) would influence neuromuscular activation we used surface EMG measures of two antagonistic muscles of the forearm and two of the wrist to directly assess neuromotor activation and cocontraction during successive reaction-time and movement-time phases of an aiming task.

409

voluntary base, a Dutch version of Spielberger’s test. This inventory, the so-called ‘Zelfbeoordelingvragenlijst’ (ZBV) [Van der Ploeg et al., 1980], is a Dutch-language version of the Spiel-berger State-Trait Anxiety Inventory [Spielberger et al., 1970]. The ZBV consists of two separate questionnaires (20 items each), measuring state anxiety and trait anxiety, respectively. The trait anxiety scores were taken as the criterion for participation in the present experiment. On the basis of this pre-experimental measurement, 20 students were invited to participate in the experiment. They were divided into two groups: ‘high-anxious’ participants with a decile score of 9 or 10 (mean 9.9), and a ‘low-anxious’ group with a decile score of 1 or 2 (mean 1.3). Each group consisted of nine-female and one-male student. The average age for the high-anxious group and low anxious group was 20.1 and 19.7 years, respectively. Each participant received payment or course credits for his or her participation. All participants were right-handed and had normal or correctedto-normal vision.

Task Predictions We predicted that introducing a memory load during an aiming task would lead to enhanced levels of neuromuscular activity together with higher levels of cocontraction but that this would not, due to biomechanic compensatory mechanisms, affect the overall percentage of hits. In the chronometric domain we did not expect an immediate deterioration of task performance. In line with findings of many earlier studies, reaction times were expected to become shorter in the dual-task load condition. In the kinematic domain, however, we expected to find manifestations of increased stiffness as an effect of the dual-task load which would be expressed in prolonged movement times and lower acceleration values. As to increased neuromuscular activity, we predicted that a dual-task memory load would increase agonistic as well as antagonistic muscular activity. Accordingly, measures of cocontraction were also expected to be higher. And finally, with regard to individual personality measures, we expected that under a condition of task load high-anxious participants would manifest higher levels of cocontraction than low-anxious participants would.

MATERIALS AND METHODS Participants Twenty participants, 10 with a high trait anxiety score and 10 with a low trait anxiety score on Spielberger’s Trait and State Anxiety test were selected from a larger population of students (N ¼ 100). All participants were first-year students in psychology. They were requested to fill out, on a

The participants performed a computer game requiring them to make series of fast movements with an electronic pen along the surface of a digitizer. The movements (always to the left) had to be made to a fixed location on the digitizer which corresponded with the position of a ball that was presented on the computer screen positioned in front of the participants. Supported by real-time feedback of the pen-tip position that was presented on the computer screen, it was the participants’ task to kick the ball into a basket by means of their stylus movement. When the pen-tip reached the position of the ball, the ball moved in reaction to the participant’s movement characteristics at the moment of virtual impact. The ball was given a virtual mass and distance from the basket. The on-line measured velocity of the pen-tip over the digitizer was entered into a Newtonian displacement equation that determined the visually presented trajectory on the screen. The participants had to produce fast but well-dosed accelerations of the forearm in order to kick the ball into the basket. The lay-out of the visual information on the screen as presented to the participants is illustrated in Figure 1. The cognitive stressor was a tone-discrimination condition that required the participant to recall, after each trial, the serial position of a high-pitch tone within a group of two low-pitch (750 Hz) and one high-pitch (1,500 Hz) tones that was presented at the start of every trial. The serial position of the high-pitch tone in the sequence of three tones was determined at random. The serial position of the high-pitch tone was to be reported in a requester that appeared on the computer screen after each trial. In the no-load condition, participants were presented with a similar tone sequence to which they did not need to respond.

410

Van Galen et al.

FIGURE1. Layoutofthe visual presentation onthecomputerscreen of the currentposition of thepentip (1),targetball (2),basket(3), and virtual platform (4) at the start of an experimental trial.

Apparatus The task was implemented using the OASIS software package, version 6.47 [De Jong et al., 1996]. The package allows manipulation of experimental conditions through visual and auditory stimulus control, and simultaneous recording of horizontal and vertical pen-displacements on a digitizer and pen-force data, as well as analysis of the data. The hardware consisted of the following elements: Pentium 90 MHz PC, Philips Brilliance 15 A monitor, 8-channel physiological amplifier with a sampling frequency of 1,000 Hz (TD 90087 custom-built for NICI), DAS 1202 ADDA card, 8-channel interface for physiological amplifier and ADDA card (custom-built), Sensor Medics Ag/AgCl 2.5 mm skin electrodes, Wacom UD-1218 digitizer working at a sampling frequency of 200 Hz, and a NICI 111 modified digitizer pen.

Procedure The experiment consisted of four blocks of 50 trials each and a training block of 30 trials presented before the experimental measurements. The digitizer was placed directly in front of the participant. The monitor was placed behind the digitizer, in such a manner that the participant, the digitizer, and the screen were in line. Each trial was structured as follows: To start a new trial, the participant had to place the electronic pen in a randomly chosen position in the lower area of the digitizer surface, at approximately 2/3 of the total width of the digitizer. Immediately after the participant put his pen on the digitizer, the screen was activated and showed the pen cursor, the basket and the platform (see Fig. 1). After a random delay of 0.3–0.6 s,

the three tones were presented. Each tone lasted 0.5 s and the interval between the tones was 0.2 s. Immediately following the tone-presentation period, the target ball appeared on the screen and the reaction-time period started. The appearance of the ball was the ‘go’ signal for the participant to which he or she had to respond as quickly as possible. When the pen was moved in a horizontal direction on the digitizer, the pen cursor on the screen moved towards the ball. The pen cursor had to be moved towards the target ball in a straight line over a distance of 11.8 cm. At impact of the ball it was launched. The velocity and spatial features of the flight of the ball were determined by the computer according to Newtonian movement equations. Given the width of the basket, the ball had to be hit with a minimum actual velocity of 62.93 cm/s and a maximum of 82.47 cm/s. The mean ‘contact’ velocity measured was 72.04 cm/s. As soon as the ball had landed, the participants in the tone discrimination condition were required to respond to the requester, after which, just as in the no-tone discrimination condition, the screen turned blank. The next trial was started as soon as the participant returned the pen to the starting position. Each participant was tested individually. Prior to the experiment, participants had to fill out the ZBV once again. The average decile scores on the re-test were 9.3 for the high-anxious group and 1.5 for the low-anxious group. Next, the participant was attached to the EMG equipment which was then calibrated. The participant proceeded to read the instruction and queries, if any, were responded to by the experimenter. Subsequently, the participant performed the 30-trial practice block with memory load and feedback on performance present. The experiment consisted of four blocks of 50 trials each. Two of the four blocks had

Forearm EMG Response and Stress Reactivity

the tone-discrimination condition. The four trial blocks were presented in a completely counter-balanced ABBA design, with half of the participants starting with the tonediscrimination condition and half of the participants starting with the no-load condition. EMG measures for Maximum Voluntary Contraction (MVC) and resting values were taken before the start of the experiment, between each block, and at the end of the experiment. After the second block of 50 trials a short break was inserted.

Measurements Overall performance and chronometric data If the ball landed within the basket it was recorded as a hit. If it stopped short of the basket it was marked as an undershoot, and if it went past the trial was scored as an overshoot. All subsequent analyses were done for the hits, the misses and the total of trials separately. The percentage of hits was taken as an overall indication of performance. Chronometric measures were reaction time and movement time. Reaction time was defined as the period between the appearance of the ball on the edge of the virtual table (immediately following the tone-presentation period) and the first measurable increase of velocity of the pen on the digitizer (threshold: 0.57 cm/s). Movement time was the time between the start of the movement and the moment of hitting the ball. We also measured mean acceleration during the movement trajectory, and the curvedness of the trajectory, i.e., {(Trajectory length/vector length)–1.0}.

EMG-recordings Simultaneously with the recording of movement kinematics, surface EMG measures were taken from M. Biceps (primary agonist), M. Triceps (primary antagonist), M. Flexor Carpi Ulnaris, and M. Extensor Carpi Radialis Longus. The first two muscles represented the functional agonist and antagonist in starting and stopping the forearm movement. The latter two muscles played an active role in stabilizing the hand posture relative to the forearm because participants were instructed to make the move by bending the elbow joint while keeping a fixed wrist angle. Given between brackets is the electrode position [Delagi and Perotto, 1980]: biceps brachii (participant supine with the arm extended); triceps long head (four finger breadths distal to posterior axillary fold); flexor carpi ulnaris (two finger breadths volar to ulna at the junction of the upper and middle thirds of the forearm); and extensor carpi radialis longus (two finger breadths distal to lateral epicondyle). Maximum voluntary contraction (MVC) and rest values of the muscles were taken for normalization purposes.

411

EMG signal processing Before the statistical analysis of the EMG data, the raw EMG signals were normalized according to the following procedure. Prior to the start of the experiment, between every experimental condition, and after the final trial block a maximum voluntary contraction (MVC), and a rest EMG measure was taken of each individual muscle. The raw MVC signal of each muscle was first centered around zero and was rectified by converting it to absolute values. The mean value of the rectified MVC signal of the second 1-s interval of the MVC recording was used to normalize the EMG task and rest signals. Signals were converted to percentages of the MVC signal and they were filtered by means of a non-zero phase lag, third-order low-pass Butterworth filter with a cut-off frequency of 12 Hz.

Description of the relevant EMG variables The EMG-data records of each trial of the experiment were segmented into three consecutive periods, i.e., the Tone Presentation Period, the Reaction Time Period, and the Movement Time Period. For each of these three consecutive periods in time, comparable EMG variables were chosen for further analysis. The variables concerning the biceps– triceps, i.e., forearm flexor-extensor combination, are described in Table I. Phasic EMG-activity is defined as the net activity of the functional agonist. During acceleration the functional agonist is the biceps and during deceleration it is the triceps. Static EMG-activity is the average activity of the functional antagonist during the same periods of acceleration and deceleration, respectively. For the two wrist stability muscles, extensor carpi radialis longus and flexor carpi ulnaris, an analogous definition of phasic and static activity is given in Table II. Since we sought to measure phasic neuromuscular activity and static cocontraction as independent measures, we first determined for each data segment (tone presentation, reaction time, and movement time period) in which samples of the record the functional agonist muscle was more active than the functional antagonist. For the movement time period the biceps was considered to be the functional agonist during acceleration and the triceps the functional agonist during deceleration. For the tone presentation period and the reaction time period, of course, these muscle activity measures reflected preparatory activation.

Statistical analysis All variables were analyzed for the three movement outcomes (hits, undershoots, and overshoots) and for the total of shots separately. To keep a clear view on the analyses, the figures represent the hits and the total of shots. The

412

Van Galen et al.

TABLE I. Description of the EMG Biceps and TricepsVariables Related to Forearm Movement Variable Mean__bic Mean__tric Phas__(bic-tric) j bic > tric

Description

Definition

Tot__phas__bic n tric

Average biceps activity Average triceps activity Phasic net activity of biceps for EMG samples during which biceps activity is larger than triceps activity Static antagonistic activity of triceps for EMG samples during which biceps activity is larger than triceps activity Phasic net activity of triceps for EMG samples during which triceps activity is larger than biceps activity Static antagonistic activity of biceps for EMG samples during which triceps activity is larger than biceps activity Total phasic activity for biceps and triceps

Tot__stat__bic n tric

Total static activity for biceps and triceps

Stat__(tric) j bic > tric Phas__(tric-bic) j tric > bic Stat__(bic) j tric > bic

tables only reproduce values for the total of shots. It appeared that the experimental variables showed the same tendencies in the case of undershoots, overshoots, hits, and total of shots. For each participant the data were averaged per condition. These averaged data were subsequently subjected to multivariate analysis of variance, with tone discrimination (TD) as within-subject variable and anxiety (A) as betweensubjects variable.

RESULTS Prior to the start of the experimental measurements participants had to fill out the ZBV once again. The average decile scores on the re-test were 9.3 for the high-anxious

Biceps activity
biceps rest activity Triceps activity
triceps rest activity Biceps activity
triceps activity given that, biceps activity > triceps activity Triceps activity, given that, biceps activity > triceps activity Triceps activity
biceps activity given that,triceps activity > biceps activity Biceps activity, given that, triceps activity > biceps activity Phas__(bic-tric) j bic > tric þ Phas__(tric-bic) j tric > bic Stat__(tric) j bic > tric þ Stat__(bic) j tric > bic

group and 1.5 for the low-anxious group. The re-test scores confirm the reliability of the selection procedure.

Performance and Kinematic Data Participants learned the task with ease during the training block and were found to make few errors in the experimental blocks. The average percentage of misses (undershoots and overshoots) was 4.34% in the control condition and even slightly less (3.33%) in the tone discrimination condition. The two groups did not significantly differ as to their overall accuracy. The results further revealed that memory load had differential effects on task speed. Reaction times were slightly faster in the tone-discrimination condition. Mean

TABLE II. Description of the Phasic and Static EMG Variables Related to Wrist Stabilization Variable

Description

Definition Flexor activity
flexor rest activity Extensor activity
extensor rest activity Flexor activity
extensor activity given that,flexor activity > extensor activity Extensor activity, given that,flexor activity > extensor activity

Tot__phas__flex n ext

Average flexor carpi ulnaris activity Average extensor carpi radialis longus activity Phasic net activity of flexor for EMG samples during which flexor activity is larger than extensor activity Static antagonistic activity of extensor for EMG samples during which flexor activity is larger than extensor activity Phasic net agonistic activity of extensor for EMG samples during which extensor activity is larger than flexor activity Static antagonistic activity of flexor for EMG samples during which extensor activity is larger than flexor activity Total phasic activity for flexor and extensor

Tot__stat__flex n ext

Total static activity for flexor and extensor

Mean__flex Mean__ext Phas__(flex-ext) j flex > ext Stat__(ext) j flex > ext Phas__(ext-flex) j ext > flex Stat__(flex) j ext > flex

Extensor activity
flexor activity given that, extensor activity > flexor activity Flexor activity, given that, extensor activity > flexor activity Phas__(flex-ext) j flex > ext þ Phas__(ext-flex) j ext > flex Stat__(ext) j flex > ext þ Stat__(flex) j ext > flex

Forearm EMG Response and Stress Reactivity

reaction times were 312 ms in the control condition and 298 ms in the tone discrimination condition, F (1, 18) ¼ 4.1, P < 0.06. The effect is in accordance with the predicted activating role of stress. There was no interaction of tone discrimination with anxiety. Movement times, however, were slower and were 310 ms in the no-tone and 324 ms in the tone-discrimination conditions, F (1, 18) ¼ 9.4, P < 0.01. The effect was predicted by the loss of gain through increased limb stiffness, which was also reflected by a significant lower acceleration. The latter variable was 233.03 and 224.49 cm/s2 in the no-tone and tone-discrimination conditions, respectively, F (1, 18) ¼ 5.42, P < 0.05. Finally, there was a significant main effect for anxiety on the trajectory shapes (F (1, 18) ¼ 5.31, P < 0.05. Lowanxious participants produced slightly more curved trajectories (0.0019) as compared to high-anxious participants (0.0013). Apparently, the low-anxious participants move less directly to the target, which suggests that they use fewer coarticulations by the wrist and fingers or less shoulder rotation.

413

EMG Data Means and significance of P-values for the effects of tone discrimination and trait anxiety, for the EMG variables for the forearm movers (biceps and triceps), and the wrist stabilization muscles (flexor carpi ulnaris and extensor carpi radialis longus) are given in Tables III and IV, respectively. The tables give data for the total of shots and are vertically organized so that the effects for the three consecutive periods of the task, i.e., tone presentation period, reaction time period, and movement time period, can be inspected separately.

Tone presentation period Although no actual movement was made in the tonepresentation period there was, of course, muscular activity during this interval. This activity is deployed in order to keep the pen positioned on the digitizer and to balance the

TABLE III. MeanValues in Percentages of MVC for EMG Variables for the Biceps and Triceps Muscles and forTotal of Shots TD

TPP

RTP

MTP

a

A

TD*A

EMG-variable

NoTD

TD

Low anx

High anx

Low anx/noTD

Low anx/TD

High anx/noTD

High anx/TD

Mean__bic Mean__tric Phas__(bic-tric) j bic > tric Stat__(tric) j bic > tric Phas__(tric-bic) j tric > bic Stat__(bic) j tric > bic Tot__phas__bic n tric Tot__stat__bic n tric Mean__bic Mean__tric Phas__(bic-tric) j bic > tric Stat__(tric) j bic > tric Phas__(tric-bic) j tric > bic Stat__(bic) j tric > bic Tot__phas__bic n tric Tot__stat__bic n tric Mean__bic Mean__tric Phas__(bic-tric) j bic > tric Stat__(tric) j bic > tric Phas__(tric-bic) j tric > bic Stat__(bic) j tric > bic Tot__phas__bic n tric Tot__stat__bic n tric

2.82 0.77 7.21 7.99 5.69 6.78 3.17 7.48 3.79 0.77 15.44 15.54 25.36 26.14 3.19 7.76 5.46 2.65 21 21.63 21.59 22.91 4.99 10.54

2.91 1.02 9.04c 9.74c 6.56 7.31 3.74a 7.64 3.92 1.09 19.76c 19.68c 25.38 25.75 3.49c 7.95 5.8 3.48 28.3a 28.69a 26.86 27.51 6.24a 11.87

3.21 0.41 8.53 8.4 8.06 8.68 3.99 7.91 3.84 0.4 15.09 14.71 28.39 28.67 3.11 8.07 5.09 2.32 22.87 22.69 27.9 28.12 5.3 10.34

2.52 1.38 7.71 9.33 4.19 5.41 2.91 7.21 3.86 1.46c 20.11 20.5 22.34 23.22 3.57 7.64 6.17 3.82 26.44 27.63 20.55 22.3 5.93 12.07

3.29 0.16 7.85 7.76 7.54 8.42 3.80 8.13 3.97 0.14 13.21 12.84 30.03 30.42 3.22 8.19 5.17 1.8 19.62 19.79 24.81 25.54 4.96 10.56

3.14 0.67 9.21 9.04 8.58 8.94 4.18 7.7 3.72 0.65 16.97 16.59 26.75 26.92 3 7.95 5 2.84 26.11 25.58 30.98 30.7 5.64 10.12

2.35 1.39 6.57 8.22 3.84 5.14 2.53 6.84 3.6 1.39 17.67 18.23 20.69 21.86 3.16 7.33 5.76 3.51 22.37 23.46 18.36 20.28 5.03 10.52

2.68 1.37 8.86 10.44 4.55 5.69 3.3 7.59c 4.13 1.53 22.55 22.77 24 24.58 3.98a 7.95c 6.59 4.13 30.5 31.8 22.74 24.31 6.84 13.61b

0 < P  0.05. 0.05 < P  0.06. c 0.06 < P < 0.10. TPP, tone presentation period; RTP, reaction period; MTP, movement period; TD, main effect of tone discrimination; A, main effect of anxiety; TD*A, interaction effect of tone discrimination and anxiety. Effects with P < 0.10 are printed in bold. b

414

Van Galen et al.

TABLE IV. MeanValues in Percentages of MVC for EMG Variables for the Wrist Stabilization Muscles and forTotal of Shots TD

TPP

RTP

MTP

A

TD*A

EMG-variable

NoTD

TD

Low anx

High anx

Low anx/noTD

Mean__flex Mean__ext Phas__(flex-ext) j flex > ext Stat__(ext) j flex > ext Phas__(ext-flex) j ext > flex Stat__ (flex) j ext > flex Tot__phas__flex n ext Tot__stat__flex n ext Mean__flex Mean__ext Phas__(flex-ext) j flex > ext Stat__(ext) j flex > ext Phas__(ext-flex) j ext > flex Stat__ (flex) j ext > flex Tot__phas__flex n ext Tot__stat__flex n ext Mean__flex Mean__ext Phas__(flex-ext) j flex > ext Stat__(ext) j flex > ext Phas__(ext-flex) j ext > flex Stat__ (flex) j ext > flex Tot__phas__flex n ext Tot__stat__flex n ext

8.26 12.58 10.35 10.56 11.74 11.87 10.4 22.26 8.82 15.05 25.17 25.39 23.17 22.06 8.74 23.86 13.5 23.73 30.05 34.94 24.89 24.23 19.86 39.11

8.19 11.35 8.43a 10.69 9.33a 11.93 9.75 20.87 8.26 14.08 23.3 25.07 19.79 20.92 8.81 21.77 13.48 23.56 28.78 34.62 24 24.3 17.55 33.97

10.08 15.07 10.06 13.2 11.62 13.81 11.7 27.19 10.82 18.84 27.11 29.45 22.3 22.62 11.47 30.59 15.89 28.85 30.02 38.47 21.39 23.13 20.15 44.68

6.37 8.86 8.72 8.04c 9.44 9.99 8.45 15.94c 6.26 10.29 21.36 21.01 20.65 20.35 6.07 15.04c 11.09 18.44 28.81 31.09 27.5 25.4 17.26 28.4

9.45 15.4 10.87 12.77 12.82 13.84 11.51 28.31 10.38 18.88 26.4 28.17 24.75 24.17 11.36 23.82 14.91 26.87 27.3 36.18 18.81 21.74 20.84 47.64

Low anx/TD High anx/noTD 10.7 14.73 9.24 13.64 10.42 13.79 11.9 26.07 11.27 18.8 27.82 30.72 19.86 21.08 11.59 28.36 16.87 30.84 32.73 40.76 23.96 24.52 19.46 41.72

7.07 9.76 9.82 8.34 10.65 9.9 9.29 16.21 7.27 11.21 23.94 22.6 21.59 19.95 6.11 14.9 12.09 20.6 32.8 33.69 30.96 26.73 18.88 30.58

High anx/TD 5.67 7.97 7.62 7.75 8.24 10.06 7.6 15.68 5.25 9.37 18.78c 19.43 19.71 20.76 6.03 15.18 10.09 16.29 24.82a 28.48 24.04c 24.07 15.65 26.22

a

0.0 < P  0.05. 0.05 < P  0.06. c 0.06 < P < 0.10. TPP, tone presentation period; RTP, reaction time period; MTP, movement time period; TD, main effect of tone discrimination; A, main effect of anxiety; TD*A, interaction effect of tone discrimination and anxiety. Effects with P < 0.10 are printed in bold. b

pen. The data showed that already in this stationary phase of the task the experimental variables had several distinct effects. For the forearm movers there was a consistent tendency towards more activation in the condition of tone-discrimination load. This effect was marginally significant when phasic activity of the biceps and static activity of the triceps are considered but it was highly significant when total phasic activities were combined, (Tot_phas_ bicntric, F (1, 18) ¼ 9.1, P < 0.01). Interestingly, we observed the opposite for the wrist stabilization musculature. Now overall neuromuscular activation was smaller in the condition of tone discrimination and this effect was significant for the net phasic activity of both flexor carpi ulnaris (Phas_flex-ext) j flex > ext, F (1, 18) ¼ 4.55 P < 0.05) and extensor carpi radialis longus (Phas_(ext-flex) j ext > flex, F (1, 18) ¼ 4.47, P < 0.05. Trait anxiety also showed a consistent pattern of effects on EMG. For forearm movers six of the eight measures of the low-anxious participants

reached a higher EMG activation than for those of the highanxious participants. For the wrist stabilizers this was true for all eight measures and the effect was marginally significant for two static measures. For the total static activity of the biceps and the triceps there was also an interesting trend towards an interaction for tone discrimination and anxiety, (F(1,16) ¼ 3.66, P < 0.08. The highanxious group showed an increase of total static activity (6.84–7.59%), whereas the low-anxious group showed a decrease (8.13–7.7%). Although in this phase of the task the effects were not as pronounced as in the later periods, it may already be concluded that dual task load had an overall increasing effect on phasic activation, i.e., on the muscles that prepared the acceleration of the oncoming movement but also on the activation of the functional antagonist, i.e., the triceps. The increased static activity was more pronounced for the high-anxious group although they displayed lower activation levels throughout this

Forearm EMG Response and Stress Reactivity

phase. It seems that a shift of attention due to the tonediscrimination condition resulted in an increase of stiffness in musculature that drives the movement although, simultaneously, a decrease in hand stabilization could be observed.

Reaction time period The general picture for the reaction time period is very similar to the tone presentation period. Again, and especially for the forearm movers, tone-discrimination stress resulted in a trend towards increased phasic activity of the biceps (Phas_(bic-tric) j bic > tric, F (1, 18) ¼ 3.828, P < 0.08) but also in an increase of the corresponding static activity of the triceps (Stat_(tric) j bic > tric, F (1, 18) ¼ 3.652, P < 0.08). Again, interactions between tone discrimination and anxiety were manifest, which demonstrated that high-anxious participants reacted with an increase of activation whereas the low-anxious group exhibited lower activation in the tone-discrimination condition. This was highly significant for the total phasic activity of biceps and triceps (Tot_phas_bic n tric, F (1, 16) ¼ 9.41, P < 0.01), and marginally significant for total static activity (Tot_ stat_bic n tric, F (1, 16) ¼ 3.51, P < 0.08). For the wrist stabilizers the effect of tone discrimination was not significant, but anxiety again had the effect of lower activation levels, an effect opposite to the interaction effect for the forearm movers.

Movement time period The movement time period was, of course, characterized by high phasic activity of the forearm movers, which particularly held for the biceps which was responsible for the acceleration during the initial movement phase. Notwithstanding this overall biomechanic effect there were also consistent and significant effects of tone discrimination. The net phasic activity of the biceps increased from 21.0–28.3% MVC (Phas_(bic-tric) j bic > tric, F (1, 18) ¼ 5.721, P < 0.05). At the same time, and this is a clear manifestation of cocontraction, the static activity of the triceps increased to the same extent from 21.63% MVC to 28.69% MVC (Stat_(tric) j bic > tric, F (1, 18) ¼ 5.321, P < 0.05). The effects of tone discrimination on phasic biceps and static triceps activity are illustrated in Figures 2 and 3. The total phasic activity also demonstrates the higher, overall neuromotor activation in the condition of dual task load (Tot_phas_bic n tric, F (1, 17) ¼ 11.25, P < 0.01). For the biceps-triceps musculature an interaction was found for the total static activity. The latter measure is an estimate of overall cocontraction and it shows that, with tone discrimination, for the low-anxious participants the measure decreased from 10.56–10.12% MVC, whereas it increased for the high-anxious group from 10.52–13.61% MVC (Tot_stat_bic n tric, F (1, 17) ¼ 4.26, P < 0.06, see Fig. 4). The wrist stabilizers, of course, also become more active during movement but, in line with the results from the

FIGURE 2. Netphasic activity of the bicepsduring movement acceleration for thenon-loaded (notd) andloaded (td) tone-discrimination conditions.The two left bars represent the hits and the two right bars stand for the total of shots. [Color figure can be viewed in the online issue which is available at www.interscience.wiley.com].

415

416

Van Galen et al.

FIGURE 3. Static activity of the triceps during movement acceleration for the non-loaded (no td) and loaded (td) tone-discrimination conditions.The two left bars represent the hits and the two rightbars the total ofshots. [Color figure can be viewedin the online issue which is available at www.interscience.wiley.com].

FIGURE 4. Total static EMG activity forbicepsand triceps as a function of tone discrimination andanxiety, and forhits (left fourbars) and total of shots (right four bars), respectively. Light gray bars represent the no-tone discrimination condition, dark gray bars refer to tone discrimination. [Color figure can be viewed in the online issue which is available at www.interscience.wiley.com].

Forearm EMG Response and Stress Reactivity

preceding intervals, here tone discrimination alone showed no significant effects. However, the interaction between tone discrimination and anxiety was significant for the net phasic flexor activity (Phas_(flex-ext) j flex > ext, F (1, 18) ¼ 4.62, P < 0.05; increase from 27.3–32.73% MVC for the low-anxious, and decrease from 32.8–24.82% MVC for the high-anxious), and marginally significant for the phasic extensor activity (Phas_(ext-flex) j ext > flex,F (1, 18) ¼ 3.722, P < 0.08, increase from 18.81–23.96% MVC for the low-anxious, and decrease from 30.96–24.04% MVC for the high-anxious). Again, as in the reaction time period, high-anxious participants reacted with a decrease of wrist stabilization whereas the low-anxious group reacted with an increase. It was found that this interaction effect of tone discrimination and anxiety was reversed for the forearm flexor-extensor musculature.

DISCUSSION The main objective of this study was to provide evidence for the theory that task stress and individual trait anxiety measures have an intricate and combined effect on muscle tension. The experiment was designed to measure task performance measures, chronometric measures, and EMG measures of phasic and static muscle activity by a close scrutiny of the different phases of a motor task. The results of the experiment showed that task load had much more pronounced effects on muscle variables and movement kinematics than on performance in terms of hits and misses. Whereas the percentage of hits remained constant across conditions, movement times increased and overall muscle activation also increased as a result of the tone discrimination dual-task load. We also found several meaningful interactions between tone discrimination and anxiety. For the forearm movers our findings show that the high-anxious participants reacted with a substantial increase of static muscle activity, i.e., of antagonistic activation. This effect was absent in the low-anxious participants. For the wrist stabilization musculature, however, the effect was reversed. Now the low-anxious group increased total muscle activity whereas the high-anxious participants exhibited decreased values. Apparently, there are differential effects of anxiety on phasic and on tonic musculature. An alternative interpretation is a differential response to task stressors by proximal and distal musculatures. This may have implications for future research on muscle activation patterns in WRUEDs because the role of phasic and postural musculature in the origin of the complaints is not yet clear. Another interesting finding is that the EMG measures demonstrated that already before the start of movement overall levels of neuromotor activation were enhanced in the condition of memory load. This finding can possibly be explained by considering that it is not movement per se that is the sole provocative factor but that sustained levels of

417

muscular activation and cocontraction in a condition of task stress might explain why physiological recovery is suboptimal between movements. It could be argued that the elevations of muscle tension that were found within the time constraint of the present experiment were not sufficient to cause any fatigue or pain and it may be questioned whether they would continue to be elevated over a longer period of time or they were sufficiently high to cause WRUED. Even though we cannot give a definitive answer to this question on the basis of the present findings, it is remarkable that the increase of neuromuscular activation that we observed was substantial. Tone discrimination resulted in 30% higher muscle-activation levels in comparison to the control condition, and the cocontraction effect for the high-anxious group showed a 33% increase. If continued over long periods of time it can be said that these values do indeed fall within the range of values known to restrict natural blood flow and tissue recovery [Jonsson, 1988]. In the presentation of our model we emphasized the earlier observations about the role of enhanced muscle tension in WRUEDs. Some authors have sought the primary cause for WRUEDs in enhanced muscle tension due to a dysfunctioning of agonist-antagonist timing [e.g., Hughes and McLellan, 1985; Nakashima et al., 1989; Marsden and Sheehy, 1990; Deuschl and Hallett, 1998]. Although in many of the studies mentioned increased muscle tension was often observed in patients, it was not made clear how muscle tension might be related to task stress and how it might exacerbate the effects of excessive motion and static postures as such. We think that neuromotor noise theory has provided us with a possible explanation. This is not to say that alternative explanations should not be considered. The literature is rich of explanatory grounds for WRUEDs, whose relevance for our model should be briefly discussed here. Close to our own theorizing is the model of Westgaard [1999] who proposed that physical workload, stress, and individual sensitivity are the three principle components in the origin of muscle pain. Also, Melin and Lundberg [1997] assumed tight interconnections between workload, muscle tension and hormonal stress coping mechanisms. Other authors have stressed the sensorimotor loss as seen in WRUEDs and some have suggested that the higher sensory thresholds in patients may have predisposed them to continue working where others would stop [Byl et al., 1996b]. Experimental research has demonstrated neurological damage that could indeed explain such sensory loss [Byl et al., 1996a]. However, it is not clear whether such sensory loss is the cause or the consequence of excessive high-tensed muscle activity. Others have stressed the changed perception of pain (nociception) in WRUED patients [Littlejohn, 1995]. In the same vein as the former explanation some authors have suggested that patients have heightened thresholds and therefore do not stop in time

418

Van Galen et al.

whereas others have proposed that, after prolonged exposure to pain, patients tend to become hypersensitive to any muscular stimulation due to a central mechanism. They therefore suffer from persistent pain even after complete cessation of motor activity [Cohen et al., 1992, 1995; Helme et al., 1992]. Sensitization is perhaps a cause of the complaints observed in the further development of WRUED. We have stressed that our model has explanatory value for the first phase of WRUEDs, i.e., in explaining their origin when pain and discomfort are still strongly related to excessive work under stressful conditions. The fact that, typically, after a few months of uninterrupted work regimes pain and discomfort become persistent even without muscle activity is still an unexplained element of the chain of events resulting in WRUEDs. New lines of research directed at the relationship between pain and hormonal stress reactions look promising [Clauw and Chrousos, 1997]. Other neural mechanisms, such as plasticity of the dorsal horn of the spine, should be considered as well [Mannion and Woolf, 2000]. Pain is attention demanding and stressful [Eccleston and Crombez, 1999] and has strong and lasting effects on psychophysiological systems [Flor et al., 1985]. Alternatively, initially experienced pain may become a conditioned stimulus for further cocontraction and thus explain the sustained high levels of muscle tension [Van den Bergh et al., 1997]. Such a conditioned stimulus-response chain could prohibit patients from recovering from task strain and make pain and functional loss a persistent feature of the syndrome. Future research into the causal mechanisms of WRUEDs should concentrate on the multifactorial etiology of the complaints and the known and unknown risk factors.

dedifferentation of the representation of the hand in the primary somatosensory cortex in adult monkeys. Neurology 47:508–520. Byl N, Wilson F, Merzenich M, Scott P, Oakes A, McKenzie A. 1996b. Sensory dysfunction associated with repetitive strain injuries of tendinitis and focal hand dystonia: a comparative study. J Orthop Sports Phys Ther 23:234–244. Clauw DJ, Chrousos GP. 1997. Chronic pain and fatigue syndromes: overlapping clinical and neuroendocrine features and potential pathogenic mechanisms. Neuroimmunomodulation 4:134–153. Cohen ML, Arroyo JF, Champion GD, Browne CD. 1992. In search of the pathogenesis of refractory cervicobrachial pain syndrome. A deconstruction of the RSI phenomenon. Med J Aust 156:432–436. Cohen ML, Sheather-Reid RB, Arroyo JF, Champion GD. 1995. Evidence of abnormal nociception in fibromyalgia and repetitive strain injury. J Musculoskeletal Pain 3:49–57. De Jong WP, Hulstijn W, Kosterman BJM, Smits-Engelsman BCM. 1996. OASIS software and its application in experimental handwriting research. In: Simner ML, Leedham CG, Thomassen AJWM, editors. Handwriting and drawing research: basic and applied issues. Amsterdam: IOS Press, p 429–440. De Ridder M, Douwes M, Groothausen J, Osinga D. 1997. Naar het voorkomen van RSI bij beeldschermwerk: het effect van invoer- en aanwijsmiddelen op houding, lichamelijk ongemak en vermoeidheid bij beeldschermwerkers. [On prevention of RSI in visual display workers: the effect of input and pointing devices upon posture, uneasiness and fatigue.] Tijdschr Ergonomie 22(3):80–86. De Zwart BCH, Broersen JPJ, Frings-Dresen MHW, Van Dijk FJH. 1997. Musculosceletal complaints in the Netherlands in relation to age, gender and physically demanding work. Int Arch Occup Environ Health 70:352–360. Delagi EF, Perotto A. 1980. Anatomic guide for the electromyographer: the limbs. Springfield IL: Thomas, 207 p. Dembe AE. 1999. The changing nature of office work: effects on repetitive strain injuries. Occup Med 14:61–72. Deuschl G, Hallett M. 1998. Focal dystonias: from occupational cramp to sensomotor disease that can be treated. Aktuelle Neurol 25:320– 328.

REFERENCES Bergqvist U, Wolgast E, Nilsson B, Voss M. 1995. Musculoskeletal disorders among visual display terminal workers: individual, ergonomic, and work organizational factors. Ergonomics 38:763–776. Bernard BP, editor. 1997. Musculoskeletal disorders and workplace factors: a critical review of epidemiologic evidence for work-related musculoskeletal disorders of the neck, upper extremity, and low back (2nd printing). Cincinnati, OH: National Institute for Occupational Safety and Health. Bindman E, Tibbets RW. 1977. Writer’s cramp—a rational approach to treatment? Br J Psychiatry 131:143–148. Blatter BM, Bongers PM. 1999. Work related neck and upper limb symptoms (RSI): high risk occupations and risk factors in the Dutch working population. Hoofddorp: TNO Work and Employment (report 4070117/r9800293).

Eccleston C, Crombez G. 1999. Pain demands attention: a cognitiveaffective model of the interruptive function of pain. Psychol Bull 125: 356–366. Ferguson D. 1971a. An Australian study of telegraphists’ cramp. Br J Ind Med 28:280–285. Ferguson D. 1971b. Repetition injuries in process workers. Med J Aust 2:408–411. Flor H, Birbaumer N, Turk DC. 1985. Assessment of stress related psychophysiological reactions in chronic back pain patients. J Consult Clin Psychol 53:354–364. Fridlund AJ, Hatfield ME, Cottam GL, Fowler S. 1986. Anxiety and striate-muscle activation: evidence from electromyographic pattern analysis. J Abnorm Psychol 95:228–236. Fry HJH. 1986. Overuse syndrome of the upper limb in musicians. Med J Aust 144:182–185.

Bongers PM, Houtman ILD. 2000. Psychosocial aspects of musculoskeletal disorders. Leiden: TNO Prevention and Health.

Goldstein IB. 1964. Role of muscle tension in personality theory. Psychol Bull 61:413–425.

Bongers PM, De Winter CR, Kompier MAJ, Hildebrandt VH. 1993. Psychosocial factors at work and musculoskeletal disease. Scand J Work Environ Health 19:297–312.

Gomer FE, Silverstein LD, Berg WK, Lassiter DL. 1987. Changes in electromyographic activity associated with occupational stress and poor performance in the workplace. Hum Factors 29:131–143.

Byl NN, Merzenich MM, Jenkins WM. 1996a. A primate genesis model of focal dystonia and repetitive strain injury: I. learning-induced

Gowers WR. 1982. A manual of diseases of the nervous system. London: Churchill, 710 p.

Forearm EMG Response and Stress Reactivity

Harvey R, Peper E. 1997. Surface electromyography and mouse position. Ergonomics 40:781–789. Helliwell PS, Mumford DB, Smeathers JE, Wright V. 1992. Work related upper limb disorder: the relationship between pain, cumulative load, disability, and psychological factors. Ann Rheum Dis 51; 1325– 1329.

419

Schreibers KBJ, Huppes G, Peereboom KJ, Koningsveld EAP, Osinga DSC. 1995. Een nieuwe aanpak ter preventie van RSI bij beeldschermwerk. [A new approach to prevent RSI in visual display workers.] Tijdschr Ergonomie 20(6):25–29. Sheehy MP, Marsden CD. 1982. Writer’s cramp-a focal dystonia. Brain 105:461–480.

Helme RD, LeVasseur SA, Gibson SJ. 1992. RSI revisited: evidence for psychological and physiological differences from an age, sex and occupation matched control group. Aust N Z J Med 22:23–29.

Silverstein B, Welp E, Nelson N, Kalat J. 1998. Claims incidence of work-related disorders of the upper extremities: Washington State, 1987 through 1995. Am J Public Health 88:1827–1833.

Hughes M, McLellan DL. 1985. Increased co-activation of the upper limb muscles in writer’s cramp. J Neurol Neurosurg Psychiatry 48: 782–787.

Smith AC, Wolf JG, Xie GY, Smith MD. 1997. Musculosceletal pain in ultrasonographers: results of a random survey. J Am Soc Echocardiogr 10:357–362.

Jonsson B. 1988. The static load component in muscle work. Eur J Appl Physiol 57:305–310. Kiesler S, Finholt T. 1988. The mystery of RSI. Am Psychol 43 :1004– 1015. Kuorinka I, Forcier L, Hagberg M, editors. 1995. Work related musculoskeletal disorders (WMSDs): a reference book for prevention. London: Taylor & Francis, 421 p. Levenstein C. 1999. Economic losses from repetitive strain injuries. Occup Med 14:149–161. Littlejohn GO. 1995. Key issues in repetitive strain injury. J Musculoskeletal Pain 3:25–33. Mannion RJ, Woolf CJ. 2000. Pain mechanisms and management: a critical perspective. Clin J Pain 16:S144–S156. Marsden CD, Sheehy MP. 1990. Writer’s cramp. Trends Neurosci 13:148–153. Melin B, Lundberg U. 1997. A biopsychosocial approach to workstress and musculoskeletal disorders. J Psychophysiol 11:238–247. Moran CA. 1992. Using myofascial techniques to treat musicians. J Hand Ther 5:97–100. Moulton B, Spence SH. 1992. Site-specific muscle hyper-reactivity in musicians with occupational upper limb pain. Behav Res Ther 30:375– 386. Nakashima K, Rothwell JC, Day BL, Thompson PD, Shannon K, Marsden CD. 1989. Reciprocal inhibition between forearm muscles in patients with writer’s cramp and other occupational cramps, symptomatic hemidystonia and hemiparesis due to stroke. Brain 112:681–698. Otten F, Bongers P, Houtman I. 1998. De kans op RSI in Nederland. Gegevens uit het permanent onderzoek leefsituatie. [Prevalence of RSI in the Netherlands: data from the permanent survey of life and job situations]. Maandbericht Gezondheid (CBS) 11:5–19. Picavet HSJ, Van Gils HWV, Schouten JSAG. 2000. Klachten van het bewegingsapparaat in de Nederlandse bevolking. Prevalenties, consequenties en risicogroepen. [Musculoskeletal complaints in the Dutch population. Prevalence, impact and risk groups]. Bilthoven: RIVMreport 266807 002. Quintner JL. 1995. The Australian RSI debate: stereotyping and medicine. Disabil Rehabit 17:256–262. Reilly PA. 1995. Repetitive strain injury: from Australia to the UK. J Psychosom Res 39 :783–788. Rempel D, Serina E, Klinenburg E. 1997. The effect of keyboard keyswitch make force on applied force and finger flexor muscle activity. Ergonomics 40:800–808.

Spielberger CD, Gorsuch RL, Lushene RE. 1970. STAI manual for the state-trait anxiety inventory. Palo Alto: Consulting Psychologists Press, 24 p. Tempel LW, Perlmutter JS. 1993. Abnormal cortical responses in patients with writer’s cramp. Neurology 43:2252–2257. Tyrer S. 1994. Repetitive strain injury. J Psychosom Res 38:493–498. Van den Bergh O, Stegen K, Van de Woestijne KP. 1997. Learning to have psychosomatic complaints: conditioning of respiratory behavior and somatic complaints in psychosomatic patients. Psychosom Med 59:13–23. Van der Ploeg HM, Defares PB, Spielberger CD. 1980. Handleiding bij de zelf-beoordelings vragenlijst ZBV [manual for self assessment questionnaire]. Lisse: Swets & Zeitlinger 35 p. Van Galen GP, De Jong WP. 1995. Fitts’ law as the outcome of a dynamic noise filtering model of motor control. Hum Mov Sci 14:539– 572. Van Galen GP, Schomaker LRB. 1992. Fitts’ law as a low-pass filter effect of muscle stiffness. Hum Mov Sci 11:11–22. Van Galen GP, Van Gemmert AWA. 1996. Kinematic and dynamic features of forging another person’s handwriting. J Forensic Document Examination 9:1–25. Van Galen GP, Van Huygevoort MAE. 2000. Error, stress the role of neuromotor noise in space oriented behaviour. Biol Psychol 51:151– 171. Van Galen GP, Van Doorn RRA, Schomaker LRB. 1990. Effects of motor programming on the power spectral density function of finger and wrist movements. J Exp Psychol Hum Percept Perform 16:755– 765. Van Galen GP, Hendriks AW, De Jong WP. 1996. What behavioral benefit does stiffness control have? An elaboration of Smith’s proposal. Behav Brain Sci 19:478–479. Van Gemmert AWA, Van Galen GP. 1997. Stress, neuromotor noise and human performance: a theoretical perspective. J Exp Psychol Hum Percept Perform 23:1299–1313. Van Gemmert AWA, Van Galen GP. 1998. Auditory stress effects on preparation and execution of graphical aiming: a test of the neuromotor noise concept. Acta Psychol 98:81–101. Westgaard RH. 1999. Effects of physical and mental stressors on muscle pain. Scan J Work Environ Health 25 (Suppl 4):19–24. Windgassen K, Ludolph A. 1991. Psychiatric aspects of Writer’s Cramp. Eur Arch Psychiatry Clin Neurosci 241:170–176.