Idea Transcript
Mindfulness and Psychological Wellbeing
1
Running head: MINDFULNESS AND PSYCHOLOGICAL WELLBEING
A Longitudinal Multilevel Analysis of the Intra-Individual Relationship Between Mindfulness and Psychological Wellbeing Author: R.A. Overmeijer First Instructor: Dr. I. Nyklíček Second Instructor: F. B. Tekle Master Thesis: Clinical Health Psychology Department of Medical Psychology Specialization: Clinical Health Psychology University of Tilburg
Mindfulness and Psychological Wellbeing
2
Abstract The main objective of this study was to investigate the intra-individual association between mindfulness and psychological wellbeing among participants undergoing mindfulness-based stress reduction (MBSR) intervention. Eighteen participants suffering from general distress completed two shortened versions of questionnaires measuring five facets of mindfulness and five facets of psychological wellbeing, respectively. Measurements were conducted daily, throughout the course of eight consecutive weeks. Longitudinal multilevel analysis established several results. First, a positive association was established between all mindfulness facets and psychological wellbeing. Second, multivariate analysis established that variation in psychological wellbeing throughout the course of MBSR intervention was partially mediated by variation in the mindfulness facets observing, describing, nonjudging and nonreactivity. Third, current levels of all mindfulness facets were not predictive of future levels of psychological wellbeing. In contrast, current level of psychological wellbeing was predictive of future levels of observing, describing, nonjudging and nonreactivity. These results suggest the possibility of an upward spiral of mindfulness and psychological wellbeing throughout the course of MBSR intervention.
Key words: meditation, MBSR intervention, mindfulness, psychological wellbeing, longitudinal multilevel analysis, observing.
Mindfulness and Psychological Wellbeing
3
A Longitudinal Multilevel Analysis of the Intra-Individual Relationship Between Mindfulness and Psychological Wellbeing At the present time, interventions based on the practice of meditation are becoming increasingly popular in Western society. They have been applied in a wide variety of settings ranging from commercial companies to health-care institutions. Specifically, within the system of health-care, interest in the potential benefits of the practice of meditation has greatly increased (Baer, Smith, Hopkins, Krietemeyer & Toney, 2006; Bishop et al., 2004; Grossman, Niemann, Schmidt & Walach, 2004; Grossman, 2008). Several interventions, developed to help patients with medical and/or mental-health problems have successfully incorporated meditation techniques into their programs. The most noteworthy among these interventions are acceptance and commitment therapy (ACT), mindfulness-based stress reduction (MBSR), dialectical behavior therapy (DBT), and mindfulness-based cognitive therapy (MBCT) (Baer et al., 2006; Grossman, 2008). Whereas some interventions place a certain emphasis on meditation in conjunction with other components, other interventions focus wholeheartedly on the practice of meditation itself. For instance, MBCT concentrates on a combination of meditation and cognitive therapy. In contrast, MBSR focuses on the practice of meditation as the central aspect of intervention (Baer et al., 2006). While the above mentioned interventions may differ in their procedures, they all aim for the mitigation of psychological and psychopathological problems through the cultivation of mindfulness. It is difficult to define the concept of mindfulness. This is partly due to the wide variety of contexts in which it is applied (Baer et al., 2006; Grossman, 2008). For example, mindfulness is explained differently in eastern religious traditions (e.g., Buddhism) than in western scientific traditions (e.g., psychology). Eastern culture tends to favor a more fluid picture of mindfulness as opposed to a more categorical view favored by Western culture. Moreover, some critics argue that mindfulness, essentially an eastern concept derived
Mindfulness and Psychological Wellbeing
4
from the practice of Buddhism, does not lend itself very easily to Western theoretical interpretation (Baer et al., 2006). In their view, true comprehension of mindfulness is realised only through rigorous, introspective practice. Despite these arguments, there are several aspects of mindfulness that practitioners and scientists generally agree upon. Mindfulness is commonly defined as the capacity to direct and sustain attention in a non-judgmental, accepting and non-reactive way to each successive moment of experience (Baer et al., 2006; Ludwig & Kabat-Zinn, 2008; Nyklíček & Kuijpers, 2008; Shapiro, Schwartz & Bonner, 1998). The underlying assumption of mindfulness training is that during large parts of the day, most people function on auto-pilot, seemingly unaware of their moment-to-moment experiences (Grossman, 2008; Kabat-Zinn, 2002). By gradually learning to increase their ability to focus and sustain attention to current experiences and by learning to develop a nonevaluative and non-reactive attitude towards these experiences, individuals will eventually cultivate a greater intimacy with the present moment, which in turn will lead to an increase in quality of life. (Kabat-Zinn, 1998; Kabat-Zinn, 1990; Kabat-Zinn, 2002; Ludwig & KabatZinn, 2008). Studies have repeatedly demonstrated the favorable effects of mindfulness-based interventions on the treatment and prevention of a considerable amount of psychological and medical problems (Bishop, 2002; Nyklíček & Kuijpers, 2008; Ostafin et al., 2006; Shapiro et al., 1998). For instance, a number of studies have demonstrated the positive influence of mindfulness on the alleviation of general distress (e.g., excessive worry and rumination) and more severe psychiatric symptoms (e.g., clinical depression and general anxiety) (Grossman et al., 2004; Hofmann, Sawyer, Witt & Oh, 2010; Nyklíček & Kuijpers, 2008; Ostafin et al., 2006; Shapiro et al., 1998). In addition, several studies have established a substantial increase in quality of life for individuals suffering from various physical ailments including chronic
Mindfulness and Psychological Wellbeing
5
pain, fibromyalgia, heart disease, and cancer (Carlson, Speca, Patel & Goodey, 2003; Gregg, Callaghan & Hayes, 2007; Rosenzweig et al., 2007). Despite evidence of the beneficial value of mindfulness-based interventions on psychological wellbeing, the actual mechanisms and processes of change that are responsible, have not been widely researched. Specifically, the constituents of mindfulness and their positive relationship with psychological wellbeing are not fully comprehended. Several studies have been conducted however, that have investigated these constituents and their relationship with psychological wellbeing (Baer et al., 2006; Baer et al., 2008; Carmody, Baer, Lykins & Olendzki, 2009; Hayes, Strosahl & Wilson, 1999; Shapiro, Carlson, Astin & Freedman, 2006). For instance, Baer et al. (2006) have examined the facet structure of mindfulness by means of analyzing five mindfulness questionnaires completed by two large samples of undergraduate students. Factor analyses of all mindfulness questionnaires yielded five separate facets of mindfulness: (1) the capacity to observe (i.e., focus and sustain attention to internal as well as external stimuli), (2) the ability to describe (i.e., label observed internal phenomena), (3) the means to act with awareness (i.e., to be fully aware of the present moment), (4) the capacity to be nonjudgmental towards thoughts and experiences ( i.e., acceptance of inner experience in a nonjudgmental way), (5) and the ability to be nonreactive towards thoughts and experiences (i.e., maintain a nonreactive attitude towards personal inner experiences). These facets were combined to form the Five Facet Mindfulness Questionnaire (FFMQ; Baer et al., 2006) and collectively represent the constituents of mindfulness (i.e., skills that can be cultivated by the practice of meditation). In a subsequent study, Baer et al. (2008) have demonstrated a positive relationship between mindfulness (assessed with the FFMQ) and psychological wellbeing. For example, in a sample of experienced meditators, each individual facet of mindfulness correlated
Mindfulness and Psychological Wellbeing
6
significantly with psychological wellbeing. However, in a sample of inexperienced meditators, no association was found between observing and psychological wellbeing. These results suggest that among inexperienced meditators, observation of internal and external stimuli does not imply increased psychological wellbeing. Thus, in comparison with other mindfulness skills, it may require more experience in the practice of meditation to utilize this skill to increase psychological wellbeing (Baer et al., 2008). Furthermore, in a combined sample of experienced and inexperienced meditators, a positive association was established between meditation experience and four of five facets of mindfulness. No association was found between meditation experience and acting with awareness. These findings suggest that the capacity to act with awareness is not cultivated by the practice of meditation. Moreover, in a combined sample of experienced and inexperienced meditators, it was found that the relationship between meditation experience and psychological wellbeing was fully mediated by the mindfulness facets: observing, describing, nonjudging and nonreactivity (Baer et al., 2008). Overall, these findings suggest that the practice of meditation is positively associated with mindfulness skills, and that mindfulness skills have a positive relationship with psychological wellbeing. However, differences have been found among facets in their association with meditation experience and psychological wellbeing (Baer et al., 2008). Based on these findings, the central aim of this study is to confirm the relationship between mindfulness and psychological wellbeing in samples of participants undergoing MBSR intervention. All participants undergoing MBSR intervention are relatively inexperienced in meditation. In contrast to the findings of Baer et al. (2008), we expect to find a positive association between every facet of mindfulness and psychological wellbeing, including the mindfulness facet observing. We expect this facet to be associated with psychological wellbeing due to the fact that MBSR intervention will last for eight weeks and meditation practices will be required daily. We expect that this will provide participants
Mindfulness and Psychological Wellbeing
7
sufficient experience to cultivate the capacity to observe and utilize this skill to increase psychological wellbeing. In addition, in correspondence with the findings of Baer et al. (2008), we expect to find that variation in psychological wellbeing over time (i.e., the relationship between time and psychological wellbeing) will be fully mediated by variation in the mindfulness facets observing, describing, nonjudging and nonreactivity. Moreover, during MBSR intervention, we expect increases in levels of psychological wellbeing to be preceded by increases in levels of mindfulness. Specifically, we expect that current level of mindfulness will be predictive of future levels of psychological wellbeing. Throughout MBSR intervention, measurements of mindfulness and psychological wellbeing will be performed on a daily basis. Inspection of the relationship between mindfulness and psychological wellbeing will be conducted by means of longitudinal multilevel analysis. This will provide the opportunity to investigate this relationship on an intra-individual level. Furthermore, by means of constructing lagged effects, we will be able to inspect whether current level of mindfulness is predictive of future levels of psychological wellbeing. Hence, we predict that, at the within-person level, mindfulness will be positively associated with psychological wellbeing. Furthermore, we predict that, at the within-person level, variation in psychological wellbeing over time will be fully mediated by variation in the mindfulness facets observing, describing, nonjudging and nonreactivity. Conclusively, we predict that, at the within-person level, current level of mindfulness will be predictive of future levels of psychological wellbeing. Method Participants As of January 2010 until March 2010, we recruited participants via a website and an advertisement in the Dutch papers near the city of Tilburg. In the advertisement and the
Mindfulness and Psychological Wellbeing
8
website, we included the following solicitation: “in winter/spring, a mindfulness-based training aimed at reducing feelings of distress will take place in Tilburg. Since the effectiveness of the training will be investigated, 10% discount is offered to those willing to participate in the study”. When participants called for an application, we asked if they suffered from symptoms of distress by asking the question „how often do you feel distressed?‟ Participants were deemed eligible if they replied that they felt this way at least regularly or often. Participants were excluded from the study if they: (1) were suffering from a psychological disorder (i.e., personality disorder, psychoticism, severe anxiety, depressive disorder, or acute burnout), and (2) had an insufficient understanding of the Dutch language. Throughout the course of four months, three MBSR interventions were conducted. In total, 40 participants took part in these interventions. However, 22 participants did not complete the daily questionnaires in a consistent manner. Hence, these participants were excluded from analysis. Analysis was conducted on the remaining participants. Prior to MBSR training, participants were instructed to complete a questionnaire containing questions regarding: age, gender, education, job status, previous treatment, and medication. The mean age of the sample was 38.70 years (SD = 9.04, range = 24-55) and 67 % were women. An overview of demographic sample characteristics is presented in Table 1. Daily Measures. For the assessment of mindfulness, the Five Facet Mindfulness Questionnaire (FFMQ) was selected (Baer et al., 2006). The FFMQ is a 39-item measure assessing multiple facets of mindfulness, including the facets observing, describing, acting with awareness, nonjudging of inner experience, and nonreactivity to inner experience. It is scored by using a five-point Likert scale ranging from 1 (never or almost never) to 5 (always or almost always). The FFMQ has internal consistency (Cronbachs‟ alpha) coefficients ranging from .75 to .91. (Baer et al., 2009; Carmody et al., 2009). For the assessment of psychological wellbeing, the short
Mindfulness and Psychological Wellbeing
9
form of the Profile of Mood States (POMS-SF) was selected (Baker, Denniston, Zabora, Polland & Dudley, 2002). The POMS-SF consists of 32 self-report items using a 5-point Likert Scale. Participants can choose from one (not at all) to five (extremely). The internal consistency (Cronbachs‟ alpha) coefficients range between 0.78 and 0.91 (Baker et al., 2002). The POMS-SF is a commonly used questionnaire for assessing individuals‟ mood states. It has been shown to measure the totality of an individuals‟ mood disturbance which consists of five states: depression, vigor, irritability, tension, and fatigue (Chen, Snyder & Krichbaum, 2001). In order to measure mindfulness and psychological wellbeing on a daily basis throughout the course of MBSR intervention which lasted eight consecutive weeks, we constructed two unique questionnaires. Each evening (after 17:00), participants had to complete both questionnaires. For the questionnaire measuring mindfulness, we selected 10 items from the FFMQ. Scales were constructed for each of the five facets by selecting two items with high applicability in daily life and high loadings on their respective facets. Likewise, for the questionnaire measuring psychological wellbeing, we selected 11 items from the POMS-SF. For each of the five facets, scales were constructed by selecting items with high applicability in daily life and high loadings on their corresponding facets (Baer et al., 2006; Baker et al., 2002). We selected two items corresponding to depression, three items corresponding to vigor, two items corresponding to irritability, two items corresponding to tension and two items corresponding to fatigue. In addition, one item was added reflecting positive affect, which we considered to be unrepresented in the scale of the original POMSSF. Hence, these 12 items were selected to represent psychological wellbeing. Intervention Every participant followed MBSR training at a meditation centre in Tilburg. The training that was provided followed the standard MBSR training developed by Kabat-Zinn
Mindfulness and Psychological Wellbeing
10
(1990) with the exception of the session of practice in silence in the sixth week. Participation in the practice of silence was optional after the intervention. The training consisted of eight weekly group sessions lasting for 150 minutes. The intervention was divided into three components: (1) instruction of theoretical background information (e.g., meditation, bodymind connections), (2) actual practice of meditation during group meetings and at home on a daily basis, and (3) group discussions aimed at solving problems regarding obstacles to effective, day-to-day practice. Participants were asked to continue their practice on a daily basis for at least 40 minutes. Statistical Analysis Prior to actual analysis of the relationship between mindfulness and psychological wellbeing, several preliminary analyses were conducted. First, to test for adequate internal consistencies, we conducted reliability analyses on the shortened versions of the FFMQ and the POMS-SF. Second, we conducted exploratory factor analyses on the shortened versions of both questionnaires. These analyses were conducted in order to investigate whether each pair of items that were selected for the construction of the daily questionnaires (i.e., the shortened versions) did indeed represent individual interpretable facets. Subsequent to the putative confirmation of the facet structure of the daily questionnaire measuring mindfulness, these facets were applied as predictor variables in longitudinal multilevel analysis. One outcome variable representing psychological wellbeing was constructed by calculating a mean score of all wellbeing items loading on the overall wellbeing factor. The first hypothesis predicted that after adjustment for age, gender, education, job status, previous treatment and medication, at the within-person level, mindfulness was positively associated with psychological wellbeing. In order to confirm the first hypothesis a longitudinal multilevel analysis was conducted. Specifically, repeated measures were applied using linear mixed models. With use of repeated measures as linear mixed models it is
Mindfulness and Psychological Wellbeing
11
possible to model variations within participants over successive periods of time (i.e., level 1) and model variations between participants (i.e., level 2). Specifically, repeated measures as linear mixed models model participants‟ scores with an intercept and a regression line. Therefore, every participant has a specific intercept (i.e., baseline score) and a specific slope (i.e., development of scores over time). In order to determine the best model, several models were considered. In model 0 (i.e., intercepts-only model), every participant had an intercept that varied at baseline and one regression line which represented average level of psychological wellbeing (i.e., the model had random intercepts and no slopes). The dependent variable (i.e., psychological wellbeing) was included in the model. Participants were specified as random effects. In subsequent models (i.e., random intercepts models), every participant had an intercept that varied at baseline and a regression line that represented a fixed change over time (i.e., subsequent models had random intercepts and fixed slopes). In model 1, the dependent variable and demographic variables were included. Demographic variables were specified as fixed factors. Participants were specified as random effects. In model 2, the dependent variable, demographic variables, and predictor variables (i.e., mindfulness facets) were included. Demographic variables were specified as fixed factors. Mindfulness variables were specified as fixed covariates. Participants were specified as random effects. The most suitable model was determined by calculating deviance (i.e., -2-log likelihood ratios). The second hypothesis predicted that after adjustment for age, gender, education, job status, previous treatment and medication, at the within-person level, variation in psychological wellbeing over time, would be fully mediated by variation in the mindfulness facets observing, describing, nonjudging and nonreactivity. In order to test the second hypothesis, we conducted several mediation analyses. According to Baron and Kenny (1986), mediation only occurs when several conditions are met. Specifically, the predictor variable
Mindfulness and Psychological Wellbeing
12
has to be significantly associated with the mediator variable. Furthermore, the mediator variable has to be significantly associated with the dependent variable. Moreover, the association between the predictor variable and the dependent variable has to become nonsignificant (full mediation) or weaker (partial mediation) after inclusion of the mediator variable. Conclusively, Sobel z-tests have to be applied in order to confirm the significance of mediation. Hence, mediation between time and psychological wellbeing due to the proposed mindfulness facets was inspected by means of this procedure. Conclusively, demographic variables, the time variable and all variables that were accountable for mediation were included in one model (i.e., model 4). This model was compared with several models: an intercepts-only model (i.e., model 0), a random intercepts model including demographic variables as control variables (i.e., model 1) and a random intercepts model including demographic variables as control variables and time as a predictor variable (i.e., model 3). The most suitable model was determined by calculating deviance (i.e., -2-log likelihood ratios). The third hypothesis predicted that after adjustment for age, gender, education, job status, previous treatment, medication and current level of psychological wellbeing, at the within-person level, current level of mindfulness would be predictive of future levels of psychological wellbeing. In order to confirm the third hypothesis, lagged variables were created so that mindfulness on the current day was predictive of psychological wellbeing on the subsequent day, after controlling for current days‟ psychological wellbeing. Specifically, mindfulness during the first 55 days was used as a predictor of psychological wellbeing on the subsequent 55 days, thus with a continuous difference of 1 day. Mindfulness during the last day (day 56) was treated as missing in our analysis. We considered several models: an intercepts-only model (i.e., model 0), a random intercepts model including demographic variables and current days‟ psychological wellbeing as control variables (i.e., model 5) and a
Mindfulness and Psychological Wellbeing
13
random intercepts model including demographic variables and current days‟ psychological wellbeing as control variables and current days‟ mindfulness facets as predictor variables (i.e., model 6). The most suitable model was determined by calculating deviance (i.e., -2-log likelihood ratios). In contrast, we also wanted to inspect the predictive value of current level of psychological wellbeing on future levels of mindfulness. In order to inspect these relationships, lagged variables were created so that psychological wellbeing on the current day was predictive of mindfulness on the subsequent day, after controlling for current days‟ mindfulness. Since five mindfulness facets were applied in our analyses as outcome variables, we constructed five different models. Each model (i.e., model 8, 10, 12, 14 and 16) was compared with an intercept-only model (i.e., each model referred to as model 0) and a random intercepts model including demographic variables and current days‟ mindfulness facet as control variables (i.e., model 7, 9, 11, 13 and 15). The most suitable models were determined by calculating deviance (i.e., -2-log likelihood ratios). Results Preliminary Analyses Throughout the course of MBSR intervention, measurements were conducted on 56 occasions for each participant (SD = 16.17, range = 1-56). Of all participants, 55 % were excluded from analysis. Reliability analysis of the shortened version of the FFMQ established an internal consistency coefficient (Cronbach‟s alpha) of .86. In order to investigate whether the shortened version of the FFMQ consisted of 5 interpretable facets, an exploratory factor analysis was conducted. Inspection of the correlation matrix yielded 46 % of coefficients above .3. Bartlett‟s Test of Sphericity was significant (p < .001) and the value of the KaiserMeyer-Olkin Measure of Sampling Adequacy was adequate (KMO > .6). These findings indicated that use of factor analysis was appropriate. Subsequent to direct oblimin rotation,
Mindfulness and Psychological Wellbeing
14
both items of describing loaded on a separate component, explaining 44 % of the variance, both items of observing loaded on a separate component, explaining 16.3 % of the variance, both items of acting with awareness loaded on a separate component, explaining 11.3 % of the variance, both items of nonreactivity loaded on a separate component, explaining 10.1 % of the variance, and both items of nonjudging loaded on a separate component, explaining 5.4 % of the variance. With the exception of nonjudging, all components yielded an eigenvalue exceeding 1. However, since nonjudging loaded on a separate component, we applied all mindfulness facets as predictor variables in our analysis. Subsequently, an exploratory factor analysis was conducted on the shortened version of the POMS-SF. Inspection of the correlation matrix yielded 44 % of coefficients above .3. Bartlett‟s Test of Sphericity was significant (p < .001) and the value of the Kaiser-Meyer-Olkin Measure of Sampling Adequacy was adequate (KMO > .6). Subsequent to direct oblimin rotation, factor analysis yielded 2 components with eigenvalues exceeding 1. All items of depression, irritability, tension and fatigue loaded on one component explaining 55.4 % of the variance and all items of vigor loaded on a separate component, explaining 20.8 % of the variance. Since factor analysis yielded only 2 components, we constructed one outcome variable of all items of the shortened version of the POMS-SF and the extra item reflecting positive affect. Reliability analysis of this scale yielded an internal consistency coefficient (Cronbachs‟ alpha) of .91. Mindfulness and Psychological Wellbeing In order to examine the first hypothesis, several models were tested by means of longitudinal multilevel analysis. According to – 2 × log criteria model 2 was the best model compared to model 1. All models are represented in Table 2. In model 2, gender was positively associated with psychological wellbeing (γ = .44, p < .05), indicating a higher mean value of psychological wellbeing for women compared to men. Furthermore, observing (γ = .09, p < .001), describing (γ = .18, p < .001), acting with awareness (γ = .05, p < .05),
Mindfulness and Psychological Wellbeing
15
nonjudging (γ = .18, p < .001) and nonreactivity (γ = .18, p < .001) were positively associated with psychological wellbeing. At the within-person level, the proportion of explained variance in psychological wellbeing attributed to all mindfulness facets was 27 % (R2 = .27). Mediation Analyses In order to investigate the second hypothesis, we conducted several mediation analyses according to the previously described procedure of Baron and Kenny (1986). First, the relationship between time and mindfulness facets was inspected. Individual analyses showed that time was positively associated with observing (γ = .01, p < .001), describing (γ = .02, p < .001), acting with awareness (γ = .01, p < .01), nonjudging (γ = .01, p < .01) and nonreactivity (γ = .01, p < .01). Second, the relationship between mindfulness facets and psychological wellbeing was inspected. Individual analyses showed that, observing (γ = .16, p < .001), describing (γ = .28, p < .001), acting with awareness (γ = .21, p < .001), nonjudging (γ = .30, p < .001) and nonreactivity (γ = .27, p < .001) were positively associated with psychological wellbeing. Third, after inclusion of each individual mindfulness facet, the strength of the association between time and psychological wellbeing was inspected. Longitudinal multilevel analysis established an association between time and psychological wellbeing (γ = .02, t= 6.87, p < .001). After inclusion of observing (γ = .01, t = 6.06, p < .001), describing (γ = .01, t = 5.05, p < .001), acting with awareness (γ = .02, t = 6.01, p < .001), nonjudging (γ = .01, t = 5.52, p < .001), and nonreactivity (γ = .01, t = 5.77, p < .001), the strength of the relationship between time and psychological wellbeing decreased. Conclusively, Sobel z-tests indicated that each individual mindfulness facet was accountable for mediation between time and psychological wellbeing (observing: Z= 5.36, p < .001; describing: Z= 4.84, p < .001; acting with awareness: Z= 2.82, p < .01; nonjudging: Z= 6.08, p < .001; nonreactivity: Z= 3.09, p < 001). Conclusively, we included demographic variables, the predictor variable (i.e., time), the mediator variables (i.e., mindfulness facets) and the outcome variable (i.e., psychological
Mindfulness and Psychological Wellbeing
16
wellbeing) in one final model (i.e., model 4). In addition, we compared this model with several models by means of longitudinal multilevel analysis. According to – 2 × log criteria model 4 was the best model compared to model 3. All models are represented in Table 3. In model 4, gender was positively associated with psychological wellbeing (γ = 0.49, p < .05), indicating a higher mean value of psychological wellbeing for women compared to men. Moreover, time (γ = .01, p < .01), observing (γ = .09, p < .001), describing (γ = .17, p < .001), nonjudging (γ = .18, p < .001) and nonreactivity (γ = .17, p < .001) were positively associated with psychological wellbeing. At the within-person level, the proportion of explained variance in psychological wellbeing attributed to the time variable and all mindfulness facets was 35 % (R2 = .35). Lagged Effects To examine the third hypothesis, several models were tested by means of longitudinal multilevel analysis. According to – 2 × log criteria, in comparison with model 5 model 6 was not the best model. This indicated that inclusion of all current days‟ mindfulness facets did not provide a better model fit. All models are represented in Table 4. In model 5, gender was positively associated with next days‟ level of psychological wellbeing (γ = .11, p < .05), indicating a higher mean value for women compared to men. Moreover, current days‟ level of psychological wellbeing was positively associated with next days‟ level of psychological wellbeing (γ = .87, p < .001). At the within-person level, the proportion of explained variance in next days‟ level of psychological wellbeing attributed to all demographic variables and current days‟ level of psychological wellbeing was 62 % (R2= .62). In addition, we conducted univariate analyses to inspect whether each individual mindfulness facet was predictive of future levels of psychological wellbeing. No significant results were obtained. In conclusion, we inspected whether current level of psychological wellbeing was predictive of future levels of mindfulness. All models are represented in Table 5. According
Mindfulness and Psychological Wellbeing
17
to – 2 × log criteria, model 8, 10, 14 and 16 were better models compared to model 7, 9, 13 and 15, respectively. Hence, after controlling for current days‟ level of each individual mindfulness facet, current days‟ level of psychological wellbeing was positively associated with next days‟ level of observing (γ = .10, p < .001, R2 = .09), describing (γ = .07, p < .01, R2 = .58), nonjudging (γ = .06, p < .05, R2 = .31), and nonreactivity (γ = .08, p < .01, R2 = .11). Discussion Association Between Mindfulness and Psychological Wellbeing In order to investigate the intra-individual relationship between mindfulness and psychological wellbeing, we conducted a longitudinal multilevel analysis. At the day-level, positive associations were established between all mindfulness facets and psychological wellbeing. Hence, our hypothesis was confirmed. In addition, this included a positive association between observing and psychological wellbeing. This finding was in line with our predictions. We expected that the length of MBSR intervention would provide participants sufficient experience to develop the capacity to observe. In turn, we expected this capacity to be positively related to psychological wellbeing. This finding corresponds to the previous finding of Baer et al. (2008) who established that, within a sample of experienced meditators, all mindfulness facets were positively associated with psychological wellbeing, including the mindfulness facet observing. Mediation Univariate analyses established that, at the day-level, variation in all mindfulness facets mediated variation in psychological wellbeing over time. However, multivariate analysis established that, at the day level, variation in psychological wellbeing over time was mediated by variation in the mindfulness facets observing, describing, nonjudging and nonreactivity. Acting with awareness was not accountable for mediation. These findings
Mindfulness and Psychological Wellbeing
18
indicated that the unique part of variation in acting with awareness did not mediate the variation in psychological wellbeing over time. In contrast, variation in acting with awareness that overlapped with variation in other mindfulness facets was accountable for mediation. Baer et al. (2008) established results that partly correspond to our findings. Specifically, in both univariate and multivariate models, acting with awareness did not mediate the relationship between meditation experience and psychological wellbeing. Conclusively, in our univariate and multivariate models, inclusion of mindfulness facets did not turn the relationship between time and psychological wellbeing into non-significance. The strength of the relationship between time and psychological wellbeing decreased, indicating partial mediation. Overall, these findings partly confirmed our second hypothesis. Predictive Value of Mindfulness on Psychological Wellbeing Longitudinal multilevel analysis of current level of mindfulness on future levels of psychological wellbeing failed to offer support for our final prediction. Specifically, multivariate analysis established that current levels of all mindfulness facets were not predictive of future levels of psychological wellbeing. This suggests that mindfulness has no influence on future levels of psychological wellbeing. Since we established positive associations at the day-level between all mindfulness facets and psychological wellbeing, and partial mediation at the day-level of variation in psychological wellbeing over time by variation in most mindfulness facets, this finding seems contradictory. However, we conducted additional analyses that shed some light on this finding. Predictive Value of Psychological Wellbeing on Mindfulness We investigated whether current level of psychological wellbeing was predictive of future levels of mindfulness. Longitudinal multilevel analysis established that current level of psychological wellbeing was positively associated with future levels of the mindfulness facets observing, describing, nonjudging and nonreactivity. This finding contradicted our
Mindfulness and Psychological Wellbeing
19
expectation that current level of mindfulness was predictive of future levels of psychological wellbeing. In order to explain these results, we draw a parallel to findings obtained by Bakker et al. (2010) who investigated the intra-individual relationship between job resources and work engagement in a sample of 54 Dutch teachers. In this study, it was established that, at the week-level, job resources (e.g., autonomy, social support, opportunities for development, etc) were positively related to work engagement (i.e., a positive, fulfilling work related state of mind). In addition, it was established that momentary work engagement predicted job resources in subsequent weeks. According to Bakker et al. (2010) these findings suggest the occurrence of a positive gain spiral. Specifically, at the week-level, the availability of job resources has a positive influence on the development of work engagement. Consequently, momentary work engagement has a positive influence on the availability of job resources in subsequent weeks. This process of a positive gain spiral corresponds to our own findings. Based on these findings, we propose that, at the day-level, mindfulness skills have a positive influence on the development of psychological wellbeing. Consequently, momentary psychological wellbeing has a positive influence on the cultivation of mindfulness skills on the next day. This process repeats itself, triggering an upward spiral towards mindfulness and psychological wellbeing throughout the course of MBSR intervention. However, since we cannot establish that mindfulness precedes psychological wellbeing at the day-level, inferences regarding causality should be treated with caution. Yet, the assumption that mindfulness precedes psychological wellbeing at the day-level is plausible, since MBSR intervention provides meditation practice. Consequently, at the day-level, meditation practice cultivates mindfulness skills, which in turn are utilized to increase psychological wellbeing. However, in order to provide more evidence for the occurrence of an upward spiral of psychological wellbeing and mindfulness throughout the course of mindfulness-based interventions, more research is needed.
Mindfulness and Psychological Wellbeing
20
Limitations This study has several limitations. First, the data in our study were obtained by shortened versions of two questionnaires. Hence, the scales may not adequately reflect the scales of the original versions of both questionnaires. Second, in order to measure psychological wellbeing on a day to day basis, we selected one questionnaire (i.e., the shortened version of the POMS-SF and one item reflecting positive affect). Since psychological wellbeing covers a wide spectrum of mental states, several studies have chosen to assess psychological wellbeing with multiple measures. Hence, it is plausible that the shortened version of the scale does not cover the whole spectrum of psychological wellbeing. However, participants were asked to complete this questionnaire daily, throughout the course of eight weeks. Hence, in order to increase the chance that participants would continually complete the questionnaires, we only included the shortened version of the scale. Third, we were forced to exclude a considerable amount of participants from our analysis. Thus, analysis was conducted on a small sample size. Therefore, generalization of the obtained results to other populations is not ensured. Fourth, due to the length of our study, we did not include a control group. Hence, inferences regarding causality of mindfulness on psychological wellbeing at the day-level should be treated with caution. Implications and Future Directions The results of this study contribute to the field of research regarding mindfulnessbased interventions. Longitudinal multilevel analysis established results that partly correspond to findings of previous studies. Specifically, our results suggest that at the day-level, meditation practice due to MBSR intervention cultivates mindfulness skills, and these skills are utilized to increase psychological wellbeing. Moreover, at the day-level, we established an association between observing and psychological wellbeing. These findings are of importance since several studies established that inexperienced meditators were unable to utilize this skill
Mindfulness and Psychological Wellbeing
21
to increase their psychological wellbeing. Hence, our results indicate that MBSR intervention provides participants the time and means to cultivate this capacity and utilize it to increase their psychological wellbeing. However, our study also established results that differ from results obtained by previous studies. Specifically, our study suggests the possibility of a positive gain spiral, indicating a highly dynamic interplay between mindfulness and psychological wellbeing throughout MBSR intervention. To the best of our knowledge, studies have never inspected the occurrence of a reciprocal relationship between mindfulness and psychological wellbeing throughout mindfulness-based interventions. Hence, regarding this aspect, our study is unique. However, as previously stated, in order to provide more evidence for the occurrence of this reciprocal relationship, more research is warranted. It is of importance to gain comprehension of the complex relationship between mindfulness and psychological wellbeing. If realized, effectiveness of mindfulness-based interventions can be increased. Therefore, we recommend future studies investigate this relationship by means of longitudinal multilevel analysis. This analysis provides the opportunity to inspect variations of mindfulness and psychological wellbeing within and between individuals over successive periods of time. In conclusion, if possible, designs should include control groups in order to make inferences about the role of causality at the day-level in the relationship between mindfulness and psychological wellbeing throughout mindfulnessbased interventions.
Mindfulness and Psychological Wellbeing
22
References. Baer, R. A. (2009). Self-focused attention and mechanisms of change in mindfulness-based treatment. Cognitive Behaviour Therapy, 38, 15-20. Baer, R. A., Smith, G. T., Hopkins, J., Krietemeyer, J., & Toney, L. (2006). Using self-report assessment methods to explore facets of mindfulness. Assessment, 13, 27-45. Baer, R. A., Smith, G. T., Lykins, E., Button, D., Krietemeyer, J., & Sauer, S. (2008). Construct validity of the Five Facet Mindfulness Questionnaire in meditating and nonmeditating samples. Assessment, 15, 329-342. Baker, F., Denniston, M., Zabora, J., Polland, A., & Dudley, W. N. (2002). A Poms Short Form for cancer patients: Psychometric and structural evaluation. Psycho-Oncology, 11, 273-281. Bakker, A. B., & Bal, M. (2010). Weekly work engagement and performance: A study among starting teachers. Journal of Occupational and Organizational Psychology, 83, 189206. Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Bishop, S. R. (2002). What do we really know about mindfulness-based stress reduction? Psychosomatic Medicine, 64, 71–84. Bishop, R. S., Lau, M., Shapiro, S. L., Carlson L., Anderson, N. D., Carmody, J., . . . Devins, G. (2004). Mindfulness: A proposed operational definition. Clinical Psychology: Science and Practice, 11, 230-241. Carlson, L. E., Speca, M., Patel, K. D., & Goodey, E. (2003). Mindfulness-based stress reduction in relation to quality of life, mood, symptoms of stress, and immune
Mindfulness and Psychological Wellbeing
23
parameters in breast and prostate cancer outpatients. Psychosomatic Medicine, 65, 571-581. Carmody, J., Baer, R. A., Lykins, E. L. B., & Olendzki, N. (2009). An empirical study of the mechanisms of mindfulness in a mindfulness-based stress reduction program. Journal of Clinical Psychology, 65, 613-626. Chen, K. M., Snyder, M., & Krichbaum, K. (2002). Translation and equivalence: The Profile of Mood States Short Form in English and Chinese. International Journal of Nursing Studies, 39, 619-624. Gregg, J. A., Callaghan, G. M., & Hayes, S. C. (2007). Improving diabetes self-management through acceptance, mindfulness, and values: A randomized controlled trial. Journal of Consulting and Clinical Psychology, 75, 336-343. Grossman, P. (2008). On measuring mindfulness in psychosomatic and psychological research. Journal of Psychosomatic Research, 64, 405-408. Grossman, P., Niemann, L., Schmidt, S., & Walach, H. (2004). Mindfulness-based stress reduction and health benefits A meta-analysis. Journal of Psychosomatic Research, 57, 35-43. Hayes, S. C., Strosahl, K. D., & Wilson, K. G., (1999). Acceptance and Commitment Therapy. New York: Guilford Press. Hoffman, S. G., Sawyer, A. T., Witt, A. A., & Oh, D. (2010). The effect of mindfulness-based therapy on anxiety and depression: a meta-analytic review. Journal of Consulting and Clinical Psychology, 78, 169-183. Kabat-Zinn, J. (1990). Full catastrophe living: Using the wisdom of your body and mind to face stress, pain, and illness. New York: Delacourt. Kabat-Zinn, J. (2002). Meditation is about paying attention. Reflections, 3.
Mindfulness and Psychological Wellbeing
24
Kabat-Zinn, J., Wheeler, E., Light, T., Skillings, A., Scharf, M. J., Cropley, T. G., . . . Bernhard, J. D. (1998). Influence of a mindfulness meditation-based stress reduction intervention on rates of skin clearing in patients with moderate to severe psoriasis undergoing phototherapy (UVB) and photochemtherapy (PUVA). Psychosomatic Medicine, 60, 625-632. Nyklíček, I., & Kuijpers, K. F. (2008). Effects of mindfulness-based stress reduction intervention on psychological wellbeing and quality of life: Is increased mindfulness indeed the mechanism? Annals of Behavioral Medicine, 35, 331-340. Ostafin, B. D., Chawla, N., Bowen, S., Dillworth, T. M., Witkiewitz, K., & Marlatt, G. A. (2006). Intensive mindfulness training and the reduction of psychological distress: A preliminary study. Cognitive and Behavioural Practice, 13, 191-197. Rosenzweig, S., Reibel, D. K., Greeson, J. M., Edman, J. S., Jasser, S. A., McMearty, K. D., & Goldstein, M, D. (2007). Mindfulness-based stress reduction is associated with improved glycemic control in type 2 diabetes mellitus: a pilot study. Alternative Therapies, 13, 36-38. Shapiro, S. L., Carlson, L. E., Astin, J. A., & Freedman, B. (2006). Mechanisms of mindfulness. Journal of Clinical Psychology, 62, 373-386. Shapiro, S. L., Schwartz, G. E., & Bonner, G. (1998). Effects of mindfulness based stress reduction on medical and premedical students. Accepted for Publication.
Mindfulness and Psychological Wellbeing 25 Appendices
Table 1 Demographics N
%
6 12
33.3 66.7
M
SD
38.7
9.04
Gender 1= Male 2= Female Age Educational level (finished) University HBO, HBS, MMS, VWO, MBO, MAVO or MULO Job status ≥ 32 hour workweek ≤ 20-32 hour workweek Previous treatment Treatment No treatment Medication Medication No medication
7
38.9
11
61.1
10 8
55.6 44.4
12 6
66.7 33.3
2 16
11.1 88.9
Mindfulness and Psychological Wellbeing 26
Table 2. Multilevel estimates of models predicting a positive association between mindfulness facets and psychological wellbeing Model 0
Model 1
Model 2
Variable
Estimate
SE
Estimate
SE
Estimate
SE
Intercept Gender Age Education Job status Treatment Medication Observe Describe ActAware Nonjudge Nonreact
3.356***
0.137
2.767*** 0.958*** 0.001 -0.212 0.128 -0.176 0.365
0.534 0.191 0.010 0.189 0.203 0.177 0.303
0.987* 0.438* -0.003 -0.163 0.144 -0.099 0.210 0.093*** 0.178*** 0.052* 0.179*** 0.176***
0.468 0.166 0.009 0.162 0.165 0.152 0.261 0.023 0.023 0.026 0.025 0.023
df -2 × log ∆ -2 × log Level 1 Variance Level 2 Variance
4 1233.758
15 909.123 304.655*** 0.409 0.025
0.043 0.029
0.558 0.263
0.058 0.113
*p=