Content uploaded by Shian-Ling Keng
Author content
All content in this area was uploaded by Shian-Ling Keng
Content may be subject to copyright.
COGNITIVE
PSYCHOTHER APY
JOURNAL OF
www.springerpub.com/jcp
SPECIAL ISSUE
Evidence-Based Mentoring
GUEST EDITOR
Jennifer Block-Lerner, PhD
With the Compliments of Springer Publishing Company, LLC
Journal of Cognitive Psychotherapy: An International Quarterly
Volume 26, Number 3 • 2012
270
© 2012 Springer Publishing Company
http://dx.doi.org/10.1891/0889-8391.26.3.270
Mechanisms of Change in
Mindfulness-Based Stress Reduction:
Self-Compassion and Mindfulness as
Mediators of Intervention Outcomes
Shian-Ling Keng, MA
Department of Psychology and Neuroscience, Duke University, Durham, North Carolina
Moria J. Smoski, PhD
Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham, North Carolina
Clive J. Robins, PhD
Department of Psychology and Neuroscience, Duke University
Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham, North Carolina
Andrew G. Ekblad, PhD
Centre for Addiction and Mental Health, University of Toronto, Ontario, Canada
Jeffrey G. Brantley, MD
Duke Integrative Medicine, Duke University Medical Center, Durham, North Carolina
Research has demonstrated support for the efficacy of mindfulness-based stress reduction
(MBSR) in alleviating psychological distress and symptoms. Less is known, however, about
the mechanisms through which MBSR achieves its outcomes. This study examined mindful-
ness and self-compassion as potential mediators of MBSR’s effects on several processes and
behaviors related to emotion regulation, using data from a randomized trial of MBSR versus
waitlist (WL), in which MBSR participants demonstrated significantly greater improvements
in worry, fear of emotion, difficulties in emotion regulation, suppression of anger, and aggres-
sive anger expression. Mediation analysis using bootstrap resampling indicated that increases
in self-compassion mediated MBSR’s effects on worry, controlling for change in mindfulness.
Increases in mindfulness mediated the intervention’s effects on difficulties in emotion regula-
tion, controlling for change in self-compassion. Both variables mediated MBSR’s effects on fear
of emotion. These findings highlight the importance of mindfulness and self-compassion as
key processes of change that underlie MBSR’s outcomes.
Keywords: mindfulness; self-compassion; mechanisms of change; mindfulness-based stress reduction
In recent years, mindfulness-based interventions have gained increasing popularity in both
clinical and research settings. These interventions all share common emphasis on cultivating
mindfulness, a form of awareness that arises through “paying attention in a particular way: on
Journal of Cognitive Psychotherapy: An International Quarterly
26
3
2012
© 2012 Springer Publishing Company
http://dx.doi.org/10.1891/0889-8391.26.3
Copyright © Springer Publishing Company, LLC
271Mechanisms of Change in MBSR
purpose, in the present moment, and nonjudgmentally” (Kabat-Zinn, 1994, p. 4). Another con-
struct that has been proposed to be an important process and outcome in mindfulness-based
interventions (Shapiro, Astin, Bishop, & Cordova, 2005) is self-compassion, which has been
defined as
being open to and moved by one’s own suffering, experiencing feelings of caring and kindness
toward oneself, taking an understanding, nonjudgmental attitude toward one’s inadequacies
and failures, and recognizing that one’s own experience is part of the common human
experience. (Neff, 2003, p. 24)
Although there is conceptual overlap between mindfulness and self-compassion, most re-
search on their psychological correlates and effects has been conducted independently, and little
work has attempted to tease apart their independent contribution to psychological well-being
or to the effects of mindfulness-based clinical interventions. The goal of this research was to
examine the independent roles of mindfulness and self-compassion in mediating the effects of
mindfulness-based stress reduction (MBSR).
MBSR is a group-based psychosocial intervention in which participants are taught principles
of mindfulness and various techniques and exercises aimed at cultivating mindfulness in daily life.
To date, much research has examined MBSR’s effects, and found the intervention to have a wide
range of beneficial psychological and physical outcomes among psychiatric, medical, and non-
clinical populations. Its effects are multidimensional, ranging from reduced depressive symptoms
(Anderson, Lau, Segal, & Bishop, 2007; Sephton et al., 2007; Shapiro, Schwartz, & Bonner, 1998),
anxiety (Shapiro et al., 1998), and medical symptoms (Williams, Kolar, Reger, & Pearson, 2001),
to increased empathy (Shapiro et al., 1998), positive affect (Bränström, Kvillemo, Brandberg, &
Moskowitz, 2010; Jain et al., 2007; Nyklícek & Kuijpers, 2008), sense of spirituality (Astin, 1997;
Shapiro et al., 1998), and quality of life (Koszycki, Benger, Shlik, & Bradwejn, 2007). For a review
of the effects of MBSR on psychological health, see Keng, Smoski, and Robins (2011). Results
of a meta-analysis indicated a moderate effect size of d ,0.50 for the intervention’s effects on
general physical and psychological well-being (Grossman, Niemann, Schmidt, & Walach, 2004).
Whereas many studies have demonstrated MBSR’s positive effects on psychological symptoms
and well-being, few studies have investigated the effects of MBSR on clinically relevant psycholog-
ical processes, particularly processes relevant to emotional experience, regulation, and expression.
MBSR may reduce individuals’ tendency to engage in maladaptive cognitive and emotional ten-
dencies, such as worry and fear of emotion, by enhancing the ability to decenter from sensations,
thoughts, and emotions. MBSR may also improve individuals’ ability to regulate emotions and
express emotions in more adaptive ways. Our recently completed randomized trial of MBSR, on
which this study is based, demonstrated that MBSR leads to significant improvements in worry,
fear of emotion, difficulties in emotion regulation, aggressive anger expression, and suppression
of anger (Robins, Keng, Ekblad, & Brantley, 2012).
Although the efficacy of MBSR is relatively well established, less is known regarding the
mechanisms through which MBSR achieves its psychological effects. Of a number of processes
proposed to be potential mechanisms of effects of MBSR, increases in mindfulness have been
noted as a potentially crucial mechanism of change. Research has found that MBSR leads to
increased self-reported mindfulness (Anderson et al., 2007; Bränström et al., 2010; Campbell,
Labelle, Bacon, Faris, & Carlson, 2012; Carmody, Reed, Kristeller, & Merriam, 2008; Dobkin &
Zhao, 2011; Garland, Gaylord, & Fredrickson, 2011; Gaylord et al., 2011; Gayner et al., 2012;
Hölzel et al., 2011; Jensen, Vangkilde, Frokjaer, & Hasselbach, 2012; Kilpatrick et al., 2011;
Kimbrough, Magyari, Langenberg, Chesney, & Berman, 2010; Lau et al., 2006; Robins et al.,
2012; Schmidt et al., 2011; Shapiro, Brown, & Biegel, 2007; Shapiro, Oman, Thoresen, Plante, &
Flinders, 2008; Vøllestad, Sivertsen, & Nielsen, 2011), and that intervention-related increases in
mindfulness predicted increases in positive states of mind, self-compassion, and spirituality, and
Copyright © Springer Publishing Company, LLC
272 Keng et al.
decreases in perceived stress, psychological distress, depression, anxiety, rumination, and worry
(Anderson et al., 2007; Bränström et al., 2010; Campbell et al., 2012; Carmody et al., 2008; Dobkin
& Zhao, 2011; Garland et al., 2011; Gayner et al., 2012; Gross et al., 2010; Hölzel et al., 2011; Jensen
et al., 2012; Lau et al., 2006; Shapiro et al., 2007; Vøllestad et al., 2011). Other studies have shown
that increases in mindfulness statistically mediated the effects of mindfulness interventions on
outcomes including reductions in rumination, anxiety, perceived stress, avoidance behavior,
and cognitive reactivity (Bränström et al., 2010; Nyklícek & Kuijpers, 2008; Raes, Dewulf, Van
Heeringen, & Williams, 2009; Shapiro et al., 2008; Vøllestad et al., 2011), and increases in positive
states of mind and quality of life (Bränström et al., 2010; Nyklícek & Kuijpers, 2008). In addi-
tion, changes in self-reported mindfulness have been found to partially mediate the relationship
between the amount of time spent in formal mindfulness homework practice during the MBSR
program and changes in perceived stress, psychological symptoms, and psychological well-being
(Carmody & Baer, 2008).
Although these studies point to mindfulness as a mediator of mindfulness-based interven-
tions’ effects, none have taken into account other potential constructs that may mediate the effects
of mindfulness-based interventions. One construct that has been identified as a potentially im-
portant process and outcome of mindfulness-based interventions is self-compassion, which has
been associated with positive psychological outcomes, including lower depression and anxiety,
and increased satisfaction with life (Neff, 2003). Like mindfulness, self-compassion is a concept
prominent in Buddhist psychology, and it involves adopting a caring and compassionate attitude
toward oneself, recognizing one’s experience as part of the larger human experience, and bringing
nonjudgmental awareness to one’s painful thoughts and experiences rather than overidentify-
ing with them. There is some conceptual overlap between the constructs of mindfulness and
self-compassion, in that both involve an attitude of nonjudgment toward one’s experiences. It is
thus perhaps unsurprising that research has found that MBSR led to significant increases in self-
compassion, and that increases in self-compassion predicted decreases in perceived stress result-
ing from the intervention (Shapiro et al., 2005).
To date, there has been some preliminary work examining the relative role of mindful-
ness and self-compassion in psychological well-being. In a cross-sectional, correlational study
by Van Dam, Sheppard, Forsyth, and Earleywine (2011), self-compassion was found to be a
stronger correlate of symptom severity and quality of life of individuals with mixed anxiety
and depression than mindfulness. In a randomized controlled trial of mindfulness-based
cognitive therapy (MBCT), an adaptation of MBSR was developed for recurrently depressed
patients; the effects of MBCT on depressive symptoms were found to be mediated by changes
in both mindfulness and self-compassion (Kuyken et al., 2010). One limitation of this study
is that changes in mindfulness and self-compassion were each included as a mediator in sepa-
rate analyses, which does not control for possible shared variance between the two constructs.
This study’s sample consisted of patients with a history of recurrent depression, and it is un-
known the extent to which the findings are generalizable to other populations, for example, a
nonclinical population. This study is a preliminary attempt at addressing the present knowl-
edge gap regarding the relative role of mindfulness and self-compassion in MBSR’s treatment
outcome by examining the extent to which each construct independently mediates the effects
of MBSR on several psychological health-related outcomes, including maladaptive cognitive
and overt behaviors, using data from a randomized, waitlisted controlled trial conducted by
our group (Robins et al., 2012). This study also aims to broaden the literature concerning
mechanisms of change of mindfulness-based interventions by assessing mindfulness using
a frequently used measure, the Five Facet Mindfulness Questionnaire (FFMQ; Baer, Smith,
Hopkins, Krietemeyer, & Toney, 2006), and by investigating the mechanisms of MBSR in a
nonclinical sample. Given the exploratory nature of this work, we made no specific predictions
Copyright © Springer Publishing Company, LLC
273Mechanisms of Change in MBSR
whether changes in mindfulness or self-compassion would more strongly mediate the effects
of MBSR on a particular outcome.
Me t h o d
Participants and Procedure
The sample, recruitment procedures, treatment conditions, and measures used in the current
analyses have been described in detail in Robins et al. (2012). We were interested in studying the
effects of MBSR in a nonclinical sample, similar to the population for whom the MBSR program
typically is provided. In brief, participants met the following study criteria: (a) aged at least
18 years, (b) no prior participation in an MBSR program, (c) no active psychotic symptoms or
suicidal ideation, (d) no psychiatric hospitalizations within the past 6 months, (e) able to attend
MBSR classes at designated times, (f) able to commit to daily home practice of mindfulness
exercises, and (g) no regular mindfulness meditation practice (or any other form of meditative
practice such as yoga or contemplative prayer) for more than an average of 20 min a week within
the past 6 months. The recruitment materials stated the duration and content of the program
(i.e., examples of types of mindfulness exercises taught) and procedures that participants would
go through if they enrolled in the study. Fifty-six individuals met the study criteria and were
randomly assigned to MBSR (n 5 28) or waitlist (WL) (n 5 28). Participants completed assess-
ments prior to the beginning of MBSR and immediately after the intervention group’s comple-
tion of MBSR (Time 2; 2 months after Time 1). Participants were paid $10 for completing each
assessment session, and the WL group was offered a course of MBSR after the MBSR group
completed their course. Twenty MBSR participants completed both the program, defined as
having attended at least five out of the eight regular class meetings, and the Time 2 (post-MBSR)
assessment, and 21 WL participants completed the Time 2 assessment. Participants’ mean age
was 46.25 years (SD 5 12.97; range 5 21–87). The sample was 84% female and 91% White.
Fifty-eight percent were married, living with a partner, or in an intimate relationship, and 51%
had a graduate degree.
In this study, mindfulness was measured using the FFMQ (Baer et al., 2006), which consists
of 39 items in five subscales: nonreactivity to inner experience, observing (noting or attending to
sensations/perceptions/thoughts/feelings), acting with awareness, describing (noting or labeling
experiences with words), and nonjudging of experiences. According to Baer et al. (2006), the
internal consistencies of the subscales range between .75 and .91. Self-compassion was assessed
using the Self-Compassion Scale (Neff, 2003), a 26-item questionnaire consisting of six subscales:
self-kindness, self-judgment, common humanity, isolation, mindfulness, and overidentification.
The scale’s internal consistency and test–retest reliability were .92 and .93, respectively (Neff,
2003). The Affective Control Scale provides a measure of fear of the experience of emotions and
fear of loss of control over internal and behavioral reactions to emotions (a 5 .94 and test–retest
r 5 .78; Williams, Chambless, & Ahrens, 1997). Worry was measured using the Penn State Worry
Questionnaire (Meyer, Miller, Metzger & Borkovec, 1990), a 16-item self-report inventory
designed to assess generality, excessiveness, and uncontrollability of pathological worry. Meyer et
al. (1990) reported that the questionnaire’s internal consistency and test–retest reliability were .93
and .92, respectively. The Difficulties in Emotion Regulation Scale was used as a measure of dif-
ficulties with emotion regulation (a 5 .93 and test–retest r 5 .88; Gratz & Roemer, 2004). Lastly,
the two subscales (anger-out and anger-in) of the Spielberger Anger Expression Scale (Spielberger
et al., 1985) provided a measure of expression of anger as aggressive behavior and suppression
of anger respectively. Internal reliabilities for each of the subscales were reported as .73 and .70,
respectively (Knight, Chisholm, Pauling, & Waal-Manning, 1988).
Copyright © Springer Publishing Company, LLC
274 Keng et al.
Statistical Analyses
The mediation analyses were conducted using a bootstrap resampling procedure that allows
for the simultaneous examination of multiple mediators (Preacher & Hayes, 2008). This pro-
cedure is a multivariate extension of the Sobel test that does not assume normality of the sam-
pling distribution of the indirect effect (ab product), an assumption that is typically violated in
small study samples. The bootstrapping approach has been recommended over the Sobel test
or the traditional causal steps approach by several researchers, given that the former approach
has been demonstrated to have higher power with reasonably controlled Type I error rate
(MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002; MacKinnon, Lockwood, & Williams,
2004). In this study, bootstrapping was accomplished by taking 3,000 random samples of the
original sample size and computing the ab product for each mediator in each sample. The
point estimate of the indirect effect is the mean of the ab product over 3,000 samples. The
procedure yields a 95% bias-corrected confidence interval. If the upper and lower limits of the
confidence interval do not contain zero, the indirect effect is significant. The procedure allows
for examination of the total indirect effect and the individual effect of a specified mediator,
over and above the effects of other mediators. In the current analyses, changes in mindfulness
and changes in self-compassion were entered as candidate mediators of MBSR’s effects on
changes in worry, fear of emotion, anger suppression, aggressive anger expression, and diffi-
culties in emotion regulation. The analyses were conducted using SPSS macros for mediation
analysis provided online at http://www.afhayes.com/spss-sas-and-mplus-macros-and-code.
html#indirect
Re s u l t s
At baseline, there were no significant group differences on any demographic or dependent vari-
ables, except that the MBSR group included a higher number of participants with some prior
meditation, yoga, or contemplative practice experience, 2 (1, N 5 41) 5 4.08, p 5 .04. This
variable was therefore included as a covariate in the mediation analyses.
Table 1 presents the means and standard deviations of the candidate mediators and depen-
dent variables for both groups at both time points. Robins et al. (2012) reported that MBSR
TABLE 1. Me a n s a n d st a n da R d deviat i o n s o f t h e Me d i a t o R s a n d ou t c o M e v a R i a b l e s a t bo t h
ti M e Po i n t s
Intervention (N 5 20) Waitlist (N 5 21)
Time 1 Time 2 Time 1 Time 2
Variable M (SD) M (SD) M (SD) M (SD)
Mindfulness 121.6 (20.31) 138.35 (13.27) 125.10 (22.13) 125.76 (21.99)
Self-compassion 49.20 (12.80) 60.95 (10.57) 51.19 (13.87) 51.81 (14.81)
Worry 58.14 (13.96) 45.50 (14.53) 51.14 (15.63) 48.33 (15.15)
Fear of emotion 129.31 (35.08) 105.60 (26.28) 121.73 (38.20) 115.48 (30.99)
Aggressive anger 12.45 (2.76) 11.55 (1.88) 15.33 (4.94) 15.71 (4.54)
expression
Suppression of anger 18.15 (4.26) 14.80 (3.19) 17.52 (5.30) 17.43 (4.11)
Difficulties with 89.66 (22.72) 72.00 (12.93) 82.89 (25.21) 83.38 (19.27)
emotion regulation
Copyright © Springer Publishing Company, LLC
275Mechanisms of Change in MBSR
TABLE 2. bo o t s t R a P Re s a M P l i n g Po i n t es t i M at e s a n d 95 % co n f i d e n c e in t e R v a l s (ci) f o R t h e
to t a l in d i R e c t ef f e c t o f MbsR o n ea c h ou t c o M e va R i a b l e , t h e in d i v i d u a l in d i R e c t ef f e c t s
o f ea c h Me d i a t o R , a n d co n t R a s t s be t w e e n t h e Me d i a t o R s
Bootstrapping BCa 95% CI
Point Estimate SE Lower Upper
Worry M 22.09 2.97 28.96 2.83
SCa 6.89 3.29 1.57 15.27
Total indirect 4.80 2.89 2.44 10.97
M vs. SC 28.98 5.56 223.15 .10
Fear of emotion Ma 8.56 5.13 2.00 25.06
SCa 6.20 3.26 .59 13.67
Total indirecta 14.76 4.42 8.24 26.45
M vs. SC 2.36 7.37 28.68 21.56
Aggressive anger M 2.37 .62 21.52 1.01
expression SC 2.65 .61 22.27 .27
Total indirect 21.02 .77 22.96 .20
M vs. SC .27 .96 21.27 2.56
Suppression of anger M 2.63 .88 22.24 1.29
SC 1.26 .84 2.12 3.25
Total indirect .63 .88 21.19 2.32
M vs. SC 21.88 1.48 24.82 1.07
Difficulties in Ma 13.30 4.59 5.59 24.34
emotion regulation SC .72 2.03 22.97 5.36
Total indirecta 14.02 3.76 7.80 23.45
M vs. SCa 12.58 6.03 2.17 26.06
Notes. Confidence intervals that do not contain zero indicate that the point estimate is
statistically significant. M 5 mindfulness; SC 5 self-compassion; BCa 5 bias-corrected and
accelerated confidence intervals; SE 5 standard error.
aIndicates a significant effect.
participants showed significantly greater increases in trait mindfulness and self-compassion, and
significantly greater decreases in fear of emotions, suppression of anger, aggressive anger expres-
sion, worry, and difficulties regulating emotions. Self-compassion and mindfulness were corre-
lated r 5 .55, p , .001 at pretreatment and r 5 .60, p , .001 posttreatment. Changes in both
variables from pretreatment to posttreatment are correlated at r 5 .53, p , .001.
The bootstrap resampling procedure produced the point estimate of each indirect effect
and their associated 95% confidence intervals. These data are presented in Table 2. The anal-
yses indicated that increases in self-compassion independently (accounting for mindfulness)
mediated MBSR’s effects on worry. Increases in mindfulness independently mediated the
effects of MBSR on difficulties in emotion regulation. Finally, both variables were signifi-
cant mediators of MBSR’s effects on fear of emotion, even when controlling for each other.
Neither mindfulness nor self-compassion mediated the effects of MBSR on aggressive anger
expression or suppression of anger. Pairwise contrasts between the two mediators showed that
changes in mindfulness and self-compassion did not differ significantly in their mediating
effects on worry or fear of emotions, but the mediating effect of mindfulness was significantly
Copyright © Springer Publishing Company, LLC
276 Keng et al.
greater than the mediating effect of self-compassion on difficulties in emotion regulation.
Given that mindfulness significantly mediated the effects of MBSR on fear of emotion and
difficulties in emotion regulation, we additionally examined whether the distinct facets of
mindfulness assessed by the FFMQ differentially mediated the effects of MBSR on these two
variables. Controlling for the effects of other facets of mindfulness, none of the mindfulness
facets uniquely mediated the effects of MBSR on fear of emotion and difficulties in emotion
regulation.
di s c u s s i o n
This study aimed to examine the relative contribution of changes in mindfulness and self-
compassion as mediators of the effects of MBSR on maladaptive cognitive and behavioral tenden-
cies, including worry, fear of emotion, aggressive anger expression and suppression of anger, and
difficulties in emotion regulation. Results indicated that changes in mindfulness independently
mediated the effects of MBSR on difficulties in emotion regulation, controlling for changes in
self-compassion, whereas changes in self-compassion mediated the effects of the intervention on
worry, controlling for changes in mindfulness. Both variables mediated the effects of MBSR on
fear of emotion, and neither mediated MBSR’s effects on aggressive anger expression nor sup-
pression of anger. Controlling for changes in other facets of mindfulness, none of the individual
facets of mindfulness assessed via the FFMQ uniquely mediated MBSR’s effects on difficulties in
emotion regulation and fear of emotion. These findings are consistent with previous intervention
studies that indicate that changes in self-compassion and/or mindfulness each mediated some
aspects of outcome of mindfulness-based interventions, including perceived stress (Bränström et
al., 2010; Nyklícek & Kuijpers, 2008; Shapiro et al., 2005; Shapiro et al., 2008), depressive symp-
toms (Kuyken et al., 2010), and quality of life (Nyklícek & Kuijpers, 2008), with the added advan-
tage of a direct comparison of the relative contribution of each of these two constructs controlling
for the other construct.
This study’s findings highlight the importance of changes in both self-compassion and
mindfulness as mediators of the effects of the intervention, and suggest that there are unique pro-
cesses in MBSR that are responsible for specific outcomes. Changes in self-compassion, above and
beyond changes in mindfulness, mediated reductions in worry; a cognitive behavior that is a key
feature of anxiety disorders (Borkovec, Shadick, Hopkins, & Rapee, 1991). This finding is consis-
tent with a previous cross-sectional study’s finding that self-compassion, relative to mindfulness,
was more strongly related to anxious and depressive symptomatology of individuals with mixed
anxiety and depression (Van Dam et al., 2011). Changes in both self-compassion and mindfulness
significantly mediated the effects of MBSR on fear of emotion, suggesting that improvements in
both processes may help reduce avoidance and fear of emotion. The finding that self-compassion
mediated the effects of the intervention on both worry and fear of emotion is worth noting, given
that previously found effects of MBSR on outcomes related to maladaptive cognitive tendencies,
such as rumination (Shapiro et al., 2008), have been primarily linked to changes in mindfulness
rather than self-compassion. These findings suggest that enhancing focus on developing self-
compassion in MBSR, or other mindfulness-based interventions, may bring about direct ben-
efit in terms of reducing maladaptive cognitive coping tendencies and increasing willingness to
accept and experience emotions.
The finding that the effects of MBSR on difficulties in emotion regulation were mediated by
improvements in mindfulness, but not self-compassion (after controlling for changes in mind-
fulness), was surprising given that self-compassion has previously been positively correlated
with some aspects of emotional intelligence, including clarity of individuals’ experience of their
Copyright © Springer Publishing Company, LLC
277Mechanisms of Change in MBSR
emotions and individuals’ ability to regulate their mood states (Neff, 2003). This finding suggests
that in the context of MBSR, relative to self-compassion, improvements in mindfulness may play
a stronger role in reducing difficulties in emotion regulation. It is unclear why neither self-com-
passion nor mindfulness mediated the effects of MBSR on aggressive anger expression and sup-
pression of anger. We suspect that the effects of MBSR on these outcomes may be mediated by
processes that are unexamined in this study, such as interpersonal functioning and the ability to
regulate expression of emotions.
Although the study was designed to highlight the unique contributions of mindfulness and
self-compassion to the benefits of MBSR, it should be noted that the two constructs are not
unrelated, as we highlighted in the introduction section. In this study, self-compassion and mind-
fulness were correlated at preintervention and postintervention. We do not suggest that these
constructs operate fully independent of each other. In fact, we posit that mindfulness allows for
greater clarity in developing self-compassion, whereas self-compassion “clears the way” for mind-
fulness by reducing attention-interfering cognitions such as negative rumination. Future studies
can examine these hypotheses by incorporating additional assessment time points and testing a
path model in which changes in mindfulness lead to changes in self-compassion, or vice versa.
Even though mindfulness and self-compassion might work together to produce improvements,
this study shows that these constructs are in fact differentiable in their impact on clinical and
psychological outcomes, and thus may be emphasized to a greater or lesser degree in mindfulness-
based interventions.
This study’s strengths include an experimental design that involves randomized assignment
to treatment or a WL control group, inclusion of assessment of constructs that have not been
examined in previous MBSR trials (such as fear of emotion and difficulties in emotion regu-
lation), and use of data analytic strategies (bootstrap resampling) that are appropriate for the
sample size and purpose of the study. This study has several limitations. First, we analyzed only
data of completers because of the relatively small sample and lack of interim assessment points,
which precludes statistically rigorous approximation of missing data points. Future studies
should include a larger sample and interim assessment points and perform intent-to-treat anal-
yses of treatment effects. The lack of interim assessment points also precludes a strong conclusion
regarding mediation. Because changes in the proposed mediators and outcomes were calculated
from the same two time points, we cannot assess whether changes in the outcome were tempo-
rally preceded by changes in the mediators. Future studies that include multiple assessment time
points during the course of the intervention will permit stronger conclusions regarding media-
tion. Another limitation of this study is that the sample recruited was a self-selected sample that
was predominantly female, White, and highly educated, which limits generalizability of the find-
ings. Future studies should aim to recruit participants from more diverse ethnic, educational, and
socioeconomic backgrounds.
To the authors’ knowledge, this study is the first to assess the unique roles of mindfulness
and self-compassion as mediators of outcomes of MBSR by using a statistical model that enables
examination of one mediator’s effect while accounting for effects of other mediators. The find-
ings of this study illustrate the differential effects of both constructs in mediating MBSR’s out-
come, and underscore the importance of self-compassion as a mechanism of change for a range
of outcomes. Although the findings must be interpreted with caution given previously noted
limitations of the study, they point to the value of simultaneously investigating multiple key vari-
ables as mediators of effects of mindfulness-based interventions. Future studies should continue
to examine the mediating effects of mindfulness and self-compassion, as well as those of other
potential processes of change, in the interest of broadening our understanding of the mecha-
nisms of mindfulness-based interventions and eventually, refining and enhancing their treatment
specificity.
Copyright © Springer Publishing Company, LLC
278 Keng et al.
Re f e R e n c e s
Anderson, N. D., Lau, M. A., Segal, Z. V., & Bishop, S. R. (2007). Mindfulness-based stress reduction and
attentional control. Clinical Psychology & Psychotherapy, 14, 449–463.
Astin, J. A. (1997). Stress reduction through mindfulness meditation. Effects on psychological symptom-
atology, sense of control, and spiritual experiences. Psychotherapy and Psychosomatics, 66, 97–106.
Baer, R. A., Smith, G. T., Hopkins, J., Krietemeyer, J., & Toney, L. (2006). Using self-report assessment meth-
ods to explore facets of mindfulness. Assessment, 13, 27–45.
Borkovec, T., Shadick, R. N., Hopkins, M., & Rapee, R. M. (1991). The nature of normal and pathological
worry. In D. Barlow (Ed.), Chronic anxiety: Generalized anxiety disorder and mixed anxiety-depression
(pp. 29–51). New York: Guilford Press.
Bränström, R., Kvillemo, P., Brandberg, Y., & Moskowitz, J. (2010). Self-report mindfulness as a mediator
of psychological well-being in a stress reduction intervention for cancer patients: A randomized study.
Annals of Behavioral Medicine, 39, 151–161.
Campbell, T. S., Labelle, L. E., Bacon, S. L., Faris, P., & Carlson, L. E. (2012). Impact of Mindfulness-Based
Stress Reduction (MBSR) on attention, rumination and resting blood pressure in women with cancer:
A waitlist-controlled study. Journal of Behavioral Medicine, 35(3), 262–271.
Carmody, J., & Baer, R. A. (2008). Relationships between mindfulness practice and levels of mindfulness,
medical, and psychological symptoms and well-being in a mindfulness based stress reduction program.
Journal of Behavioral Medicine, 31, 23–33.
Carmody, J., Reed, G., Kristeller, J., & Merriam, P. (2008). Mindfulness, spirituality, and health-related symp-
toms. Journal of Psychosomatic Research, 64, 393–403.
Dobkin, P. L., & Zhao, Q. (2011). Increased mindfulness—The active component of the mindfulness-based
stress reduction program? Complementary Therapies in Clinical Practice, 17, 22–27.
Garland, E. L., Gaylord, S. A., & Fredrickson, B. L. (2011). Positive reappraisal mediates the stress-reductive
effects of mindfulness: An upward spiral process. Mindfulness, 2, 59–67.
Gaylord, S. A., Palsson, O. S, Garland, E. L., Faurot, K. R., Coble, R. S., Mann, J. D., et al. (2011). Mindfulness
training reduces the severity of irritable bowel syndrome in women: Results of a randomized controlled
trial. The American Journal of Gastroenterology, 106(9), 1678–1688.
Gayner, B., Esplen, M. J., DeRoche, P., Wong, J., Bishop, S., Kavanagh, L., et al. (2012). A randomized con-
trolled trial of mindfulness-based stress reduction to manage affective symptoms and improve quality
of life in gay men living with HIV. Journal of Behavioral Medicine, 35(3), 272–285.
Gratz, K. L., & Roemer, L. (2004). Multidimensional assessment of emotion regulation and dysregulation:
development, factor structure, and initial validation of the Difficulties in Emotion Regulation Scale.
Journal of Psychopathology and Behavioral Assessment, 26, 41–54.
Gross, C. R., Kreitzer, M. J., Thomas, W., Reilly-Spong, M., Cramer-Bornemann, M., Nyman, J. A., et al.
(2010). Mindfulness-based stress reduction for solid organ transplant recipients: A randomized con-
trolled trial. Alternative Therapies in Health and Medicine, 16, 30–38.
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.
Hölzel, B. K., Carmody, J., Vangel, M., Congleton, C., Yerramsetti, S. M., Gard, T., et al. (2011). Mindfulness
practice leads to increases in regional brain gray matter density. Psychiatry Research, 191, 36–43.
Jain, S., Shapiro, S. L., Swanick, S., Roesch, S. C., Mills, P. J., Bell, I., et al. (2007). A randomized controlled
trial of mindfulness meditation versus relaxation training: Effects on distress, positive states of mind,
rumination, and distraction. Annals of Behavioral Medicine, 33, 11–21.
Jensen, C. G., Vangkilde, S., Frokjaer, V., & Hasselbalch, S. G. (2012). Mindfulness training affects attention—or
is it attentional effort? Journal of Experimental Psychology. General, 141(1), 106–123.
Kabat-Zinn, J. (1994). Wherever you go, there you are: Mindfulness meditation in everyday life. New York:
Hyperion.
Copyright © Springer Publishing Company, LLC
279Mechanisms of Change in MBSR
Keng, S. L., Smoski, M. J., & Robins, C. J. (2011). Effects of mindfulness on psychological health: A review of
empirical studies. Clinical Psychology Review, 31, 1041–1056.
Kilpatrick, L. A., Suyenobu, B. Y., Smith, S. R., Bueller, J. A., Goodman, T., Creswell, J. D., et al. (2011).
Impact of mindfulness-based stress reduction training on intrinsic brain connectivity. Neuroimage, 56,
290–298.
Kimbrough, E., Magyari, T., Langenberg, P., Chesney, M., & Berman, B. (2010). Mindfulness intervention for
child abuse survivors. Journal of Clinical Psychology, 66, 17–33.
Knight, R. G., Chisholm, B. J., Pauling, J. M., & Waal-Manning, H. J. (1988). The Spielberger Anger Expression
Scale: Some psychometric data. British Journal of Clinical Psychology, 27, 279–281.
Koszycki, D., Benger, D., Shlik, J., & Bradwejn, J. (2007). Randomized trial of a meditation-based stress
reduction program and cognitive behavior therapy in generalized anxiety disorder. Behavior Research
and Therapy, 45, 2518–2526.
Kuyken, W., Watkins, E., Holden, E., White, K., Taylor, R. S., Byford, S., et al. (2010). How does mindfulness-
based cognitive therapy work? Behaviour Research and Therapy, 48, 1105–1112.
Lau, M. A., Bishop, S. R., Segal, Z. V., Buis, T., Anderson, N. D., Carlson, L., et al. (2006). The Toronto
Mindfulness Scale: Development and validation. Journal of Clinical Psychology, 62, 1445–1467.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of meth-
ods to test mediation and other intervening variable effects. Psychological Methods, 7, 83–104.
MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence limits for the indirect effect:
Distribution of the product and resampling methods. Multivariate Behavioral Research, 39, 99.
Meyer, T. J., Miller, M. L., Metzger, R. L., & Borkovec, T. D. (1990). Development and validation of the Penn
State Worry Questionnaire. Behaviour Research and Therapy, 28, 487–495.
Neff, K. D. (2003). The development and validation of a scale to measure self-compassion. Self and Identity,
2, 223–250.
Nyklícek, I., & Kuijpers, K. (2008). Effects of mindfulness-based stress reduction intervention on psycholog-
ical well-being and quality of life: Is increased mindfulness indeed the mechanism? Annals of Behavioral
Medicine, 35, 331–340.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing in-
direct effects in multiple mediator models. Behavior Research Methods, 40, 879–891.
Raes, F., Dewulf, D., Van Heeringen, C., & Williams, J. M. (2009). Mindfulness and reduced cognitive reac-
tivity to sad mood: Evidence from a correlational study and a non-randomized waiting list controlled
study. Behaviour Research and Therapy, 47, 623–627.
Robins, C. J., Keng, S. L., Ekblad, A., & Brantley, J. (2012). Effects of mindfulness-based stress reduction on
emotional experience and expression: A randomized controlled trial. Journal of Clinical Psychology, 68,
1–15.
Schmidt, S., Grossman, P., Schwarzer, B., Jena, S., Naumann, J., & Walach, H. (2011). Treating fibromyalgia
with mindfulness-based stress reduction: Results from a 3-armed randomized controlled trial. Pain,
152, 361–369.
Sephton, S. E., Salmon, P., Weissbecker, I., Ulmer, C., Floyd, A., Hoover, K., et al. (2007). Mindfulness med-
itation alleviates depressive symptoms in women with fibromyalgia: Results of a randomized clinical
trial. Arthritis and Rheumatism, 57, 77–85.
Shapiro, S. L., Astin, J. A., Bishop, S. R., & Cordova, M. (2005). Mindfulness-based stress reduction for health
care professionals: Results from a randomized trial. International Journal of Stress Management, 12,
164–176.
Shapiro, S. L., Brown, K. W., & Biegel, G. M. (2007). Teaching self-care to caregivers: Effects of mindfulness-based
stress reduction on the mental health of therapists in training. Training and Education in Professional
Psychology, 1, 105–115.
Shapiro, S. L., Oman, D., Thoresen, C. E., Plante, T. G., & Flinders, T. (2008). Cultivating mindfulness: Effects
on well-being. Journal of Clinical Psychology, 64, 840–862.
Copyright © Springer Publishing Company, LLC
280 Keng et al.
Shapiro, S., Schwartz, G., & Bonner, G. (1998). Effects of mindfulness-based stress reduction on medical and
premedical students. Journal of Behavioral Medicine, 21, 581–599.
Spielberger, C. D., Johnson, E. H., Russell, S., Crane, R., Jacobs, E., & Worden, T. J. (1985). The experience
and expression of anger: Construction and validation of an anger expression scale. In M. A. Chesney &
R. H. Rosenman (Eds.), Anger and hostility in cardiovascular and behavioral disorders (pp. 5–30). New
York: Hemisphere/McGraw-Hill.
Van Dam, N. T., Sheppard, S. C., Forsyth, J. P., & Earleywine, M. (2011). Self-compassion is a better predictor
than mindfulness of symptom severity and quality of life in mixed anxiety and depression. Journal of
Anxiety Disorders, 25, 123–130.
Vøllestad, J., Sivertsen, B., & Nielsen, G. H. (2011). Mindfulness-based stress reduction for patients with
anxiety disorders: Evaluation in a randomized controlled trial. Behaviour Research and Therapy, 49,
281–288.
Williams, K. A., Kolar, M. M., Reger, B. E., & Pearson, J. C. (2001). Evaluation of a wellness-based mindfulness
stress reduction intervention: A controlled trial. American Journal of Health Promotion, 15, 422–432.
Williams, K. E., Chambless, D. L., & Ahrens, A. H. (1997). Are emotions frightening? An extension of the fear
concept. Behaviour Research and Therapy, 35, 239–248.
Acknowledgments. We would like to thank Anita Lungu, Paul Geiger, Christine Jamieson, and Samantha
Lorusso for their assistance in data collection. We are grateful to our participants for their generous contribu-
tion of their time. This project was supported by two grants from the Duke University Undergraduate Research
Support Office awarded to Shian-Ling Keng. Dr. Smoski’s effort was supported by K23 MH087754.
Correspondence regarding this article should be directed to Shian-Ling Keng, MA, Box 3026, Duke University
Medical Center, Durham, NC 27710. E-mail: slk18@duke.edu
Copyright © Springer Publishing Company, LLC
