The Effects of Mindfulness Meditation on Cognitive Processes and Affect in Patients with Past Depression

Abstract

This study describes the effects of an 8-week course in Mindfulness-Based Stress Reduction (MBSR; J. Kabat-Zinn, 1982, 1990) on affective symptoms (depression and anxiety), dysfunctional attitudes, and rumination. Given the focus of mindfulness meditation (MM) in modifying cognitive processes, it was hypothesized that the primary change in MM practice involves reductions in ruminative tendencies. We studied a sample of individuals with lifetime mood disorders who were assessed prior to and upon completion of an MBSR course. We also compared a waitlist sample matched with a subset of the MBSR completers. Overall, the results suggest that MM practice primarily leads to decreases in ruminative thinking, even after controlling for reductions in affective symptoms and dysfunctional beliefs.

Full text

Cognitive Therapy and Research, Vol. 28, No. 4, August 2004 (
C
2004), pp. 433–455
The Effects of Mindfulness Meditation on Cognitive
Processes and Affect in Patients With Past Depression
Wiveka Ramel,
1,2,3,4
Philippe R. Goldin,
2,3
Paula E. Carmona,
3
and John R. McQuaid
2,3
This study describes the effects of an 8-week course in Mindfulness-Based Stress
Reduction (MBSR; J. Kabat-Zinn, 1982, 1990) on affective symptoms (depression
and anxiety), dysfunctional attitudes, and rumination. Given the focus of mindfulness
meditation (MM) in modifying cognitive processes, it was hypothesized that the
primary change in MM practice involves reductions in ruminative tendencies. We
studied a sample of individuals with lifetime mood disorders who were assessed prior
to and upon completion of an MBSR course. We also compared a waitlist sample
matched with a subset of the MBSR completers. Overall, the results suggest that MM
practice primarily leads to decreases in ruminative thinking, even after controlling for
reductions in affective symptoms and dysfunctional beliefs.
KEY WORDS: meditation; cognitive processes; rumination; affective disorders; treatment outcomes.
INTRODUCTION
Dysfunctional attitudes and rumination are cognitive attributes associated with
a vulnerability to developing, maintaining, and relapsing into a depressive disorder
(for reviews, see Beck, 1967; Beck, Rush, Shaw, & Emery, 1979; Ingram, Miranda,
& Segal, 1998; Nolen-Hoeksema, 1991). Dysfunctional attitudes are characterized
by negative, rigid, and extreme assumptions and beliefs about self-worth and
typically involve conditional standards in areas of evaluation, perfectionism, and
interpersonal approval (Weissman & Beck, 1978; Zuroff, Blatt, Sanislow, Bondi,
& Pilkonis, 1999). Elevated dysfunctional attitudes are commonly reported in
currently depressed individuals (for a review, see Haaga, Dyck, & Ernst, 1991),
but recovered depressed individuals tend not to differ from healthy controls in the
1
Department of Psychology, San Diego State University, San Diego, California.
2
Department of Psychiatry, University of California San Diego, San Diego, California.
3
Department of Veterans Affairs, San Diego Healthcare System, San Diego, California.
4
Correspondence should be directed to Wiveka Ramel, Department of Psychology, Jordan Hall,
Building 420, 450 Serra Mall, Stanford University, Stanford, California 94305; e-mail: wramel@
ucsd.edu.
433
0147-5916/04/0800-0433/0
C

2004 Springer Science+Business Media, Inc.

434 Ramel, Goldin, Carmona, and McQuaid
number and intensity of dysfunctional attitudes they endorse, unless challenged by
sad mood or a stressor (for reviews, see Ingram, Miranda, & Segal, 1998; Segal &
Ingram, 1994). This suggests that dysfunctional attitudes are mood state-dependent.
Thus, while extreme and dysfunctional attitudes are likely to be available in the
minds of vulnerable individuals, they are generally not accessible unless activated by
a personally relevant environmental demand or by an increase in negative mood (for
examples of research studies demonstrating this, see Miranda, 1992; Miranda, Gross,
Persons, & Hahn, 1998; Miranda & Persons, 1988; Miranda, Persons, & Byers,
1990).
While dysfunctional attitudes reflect the content of the mind (i.e., what a person
thinks about), rumination is associated with the processes of the mind, that is, how a
person relates to the content of the mind. Rumination has been defined as passively
focusing one’s attention on a negative emotional state like depression, its symptoms,
and thinking repetitively about the causes, meanings, and consequences of that state
(Nolen-Hoeksema, 1991). Individuals who ruminate report they believe this will
increase their understanding of themselves and solve their problems, but studies
suggest ruminators are ineffective in active, interpersonal problem-solving and
show an inflexible, perseverative cognitive style on a traditional neuropsychological
test of novel problem solving (Davis & Nolen-Hoeksema, 2000; Lyubomirsky &
Nolen-Hoeksema, 1995; Watkins & Baracaia, 2002). Both laboratory and field
studies have demonstrated that ruminating in response to negative moods is
associated with maintenance of depression and exacerbated sad affect (Morrow
& Nolen-Hoeksema, 1990; Nolen-Hoeksema & Morrow, 1991; Nolen-Hoeksema,
Morrow, & Fredrickson 1993). Rumination has also been shown to increase the
risk of developing depressive episodes in healthy participants who were followed
prospectively (Just & Alloy, 1997; Robinson & Alloy, 2003; Spasojevi
´
c & Alloy,
2001). Moreover, as reported by Spasojevi
´
c and Alloy (2001), rumination mediated
several other hypothesized risk factors that prospectively predicted number of
depressive episodes, including dysfunctional attitudes, neediness, self-criticism, and
history of past depression. Thus, a ruminative response style appears to be a factor
in both the onset and maintenance of depression and a relevant target for treatments
aimed at reducing current and future affective symptoms.
Self-regulatory strategies based on meditation practice, used alone or as
adjuncts to other behavioral or medication regimens, may provide a set of tech-
niques for modifying depressogenic cognitive variables such as rumination and
dysfunctional beliefs. Mindfulness meditation (MM) is one such strategy that has
been used in clinical practice, and existing research studies suggest that it may be
a promising form of treatment for several physical and psychological conditions,
including stress and mood symptoms in general, anxiety disorders, depression
relapse prevention, chronic pain, fibromyalgia, binge eating, substance abuse, and
skin related diseases (e.g., Astin, 1997; Goldenberg et al., 1994; Kabat-Zinn, 1990;
Kabat-Zinn et al., 1992, 1998; Kabat-Zinn, Lipworth, & Burney, 1985; Kristeller
& Hallett, 1999; Marlatt, 2002; Speca, Carlson, Goodey, & Angen, 2000; Teasdale
et al., 2000). Results from a recent meta-analytic review corroborate the utility of
mindfulness-based treatments for a variety of disorders while highlighting the need
for research with sounder methodology (Baer, 2003). The latter point is echoed in a

Mindfulness, Affective Symptoms, and Cognitions 435
review by Bishop (2002), who emphasized the need for construct validation and the
paucity of randomized controlled studies on Mindfulness-Based Stress Reduction
(MBSR; Kabat-Zinn, 1982, 1990), which limits inferences regarding the efficacy of
this approach. Some studies indicate that MM is a useful adjunct to traditional forms
of psychotherapy (Kutz et al., 1985; Kutz, Borysenko, & Benson, 1985) and it is an
essential component of several more recently developed and empirically validated
psychotherapy interventions (e.g., Acceptance and Commitment Therapy, Hayes,
Strosahl, & Wilson, 1999; Dialectic Behavior Therapy for borderline personality
disorder, Linehan, 1993; Mindfulness-Based Cognitive Therapy for depression,
Segal, Williams, & Teasdale, 2002; Relapse prevention for substance abuse, Marlatt
& Gordon, 1985).
An ancient Buddhist meditation practice often characterized as the heart of
the Buddha’s teachings, MM is aimed at reducing mental anguish. Mindfulness
has been described as “paying attention in a particular way: on purpose, in the
present moment, and non-judgmentally” (Kabat-Zinn, 1994, p. 4). One of the goals
of MM is to learn how to become aware of, observe and react less habitually to
sensations, thoughts, and feelings. MM involves training in deployment of attention
to maintain awareness on a designated object, such as the breath or physical
sensations in the body, without ignoring other aspects of internal and external
stimuli. Thus, participants are instructed to notice the thoughts and feelings that
arise without becoming absorbed in their content (Kabat-Zinn, 1982). As is more
extensively described by Kabat-Zinn (1990, 1994), Santorelli (1999), and Segal et al.
(2002), with repeated training, this exercise is intended to highlight the habitual and
automatic patterns of the mind and cultivate a more decentered and nonjudgmental
perspective to cognitions, emotions, and sensations (e.g., viewing thoughts and
feelings as merely passing events of the mind rather than necessarily accurate
reflections of reality). One of the possible consequences of MM practice is a more
flexible notion of the self and a more fluid relationship to the content of the mind.
Segal, Teasdale, and Williams have proposed that MM may be beneficial for
patients with a history of depression because it is a systematic method of enhancing
attentional awareness and allocation (Segal et al., 2002; Teasdale, Segal, & Williams,
1995). As noted above, previously depressed patients may have specific cognitive
vulnerabilities, including dysfunctional attitudes and a t endency toward rumination,
which increase the risk for further depressive episodes. Segal and colleagues suggest
that MM teaches individuals to (a) identify destructive contents and habitual
patterns of the mind at an early stage, and (b) relate and process this information in a
nonjudgmental way that facilitates the individual choosing between various options.
This approach increases flexibility over cognitive activities and allows the individual
the possibility to reduce rumination, overgeneralization, and self-critical evaluation
and increase constructive cognitive processes such as nonjudgmental observation
of the content of the mind. Previously depressed patients utilizing these MM skills
may be less susceptible to mood fluctuations and reactivation of pervasive and
destructive patters of thinking and feeling, which, if not contained, may escalate
into full-blown depressive episodes (Teasdale, 1988). In support for this theory, a
randomized clinical trial found a significant reduction in relapse/recurrence to major
depression in recovered patients with three or more previous episodes of major

436 Ramel, Goldin, Carmona, and McQuaid
depression following an 8-week course in Mindfulness-Based Cognitive Therapy
(MBCT), which trained the patients to “disengage from dysphoria-activated de-
pressogenic thinking” (Teasdale et al., 2000, p. 615). Moreover, Williams, Teasdale,
Segal, and Soulsby (2000) found that compared to a treatment-as-usual control
group, recovered depressed patients reduced the number of overgeneral autobi-
ographical memories—a characteristic of depressive cognition—and increased the
specificity of their memories following MBCT.
Although there is increased research investigating the efficacy of MM, few
studies have examined the mechanisms, especially from a cognitive perspective, by
which this particular form of meditation operates. One study that did examine the
effect of meditation on attention compared concentrative or focused meditators,
who aim to restrict their attention to a single point and ignore extraneous stimuli,
to mindfulness or receptive meditators, who seek to distribute their attention and
incorporate extraneous stimuli as observational events in their practice (Valentine
& Sweet, 1999). The authors found that both groups of meditators demonstrated
improved sustained attention in comparison to controls on an auditory perception
test, in which the task was to count bleeps presented at slower or faster rates.
Although the two groups of meditators were equally accurate at estimating the
numbers of bleeps when the stimulus was slower, the mindfulness meditators
showed superior performance of accuracy compared to concentrative meditators
when the stimulus was faster. The authors argued that beeps at a slower rate
were expected and beeps at a faster rate were less expected. They explained their
findings in light of Posner and Snyder’s (1975) theory of expectancy and suggested
that mindfulness meditators are less likely to get absorbed or “caught up” in the
presenting stimuli and can more easily shift between unexpected stimuli. The results
from this study are consistent with the theory of the mechanism involved in MM
as proposed by Teasdale et al. (1995), but the study did not address psychiatric
populations or symptoms.
To examine the potential impact of MM on affect and negative cognitive
patterns in a psychiatrically vulnerable population, we studied a sample of indi-
viduals with current or lifetime diagnosis of a mood disorder before and after
participation in an MBSR course (Kabat-Zinn, 1982, 1990). We complemented this
within-subject design with a smaller but well-matched comparison group who were
waitlisted for the MBSR course. Given the focus of MM in modifying cognitive and
affective processes (i.e., the relationship to cognitions, emotions, and sensations), we
hypothesized that MM practice would decrease ruminative tendencies. MM practice
also emphasizes acceptance and a nonjudgmental observation of the content of
the mind, thus, we hypothesized that MM would lead to significant reductions in
dysfunctional attitudes and affective symptoms. However, we expected that these
would be reduced when controlling for changes in rumination, as altering the
relationship or response style to emotions, cognitions, and sensations is at the heart
of MM practice.
As might be predicted, the association between affective symptomatology and
a ruminative response style tends to be fairly high, making it hard to dissoci-
ate components of mood versus rumination. Recent advances in the rumination
literature have attempted to separate items on a commonly used rumination

Mindfulness, Affective Symptoms, and Cognitions 437
scale derived from the Response Style Questionnaire (RSQ; Nolen-Hoeksema
& Morrow, 1991) that overlap with measurement of depressive symptomatology.
Two factors, reflection/pondering and brooding, have been identified and appear
to be differentially related to affective distress (Fresco, Armey, Mennin, Turk,
& Heimberg, 2004; Treynor, Gonzalez, & Nolen-Hoeksema, 2003). As reported
by Treynor and colleagues, the reflection factor consists of neutrally valenced
items (e.g., “go away by yourself and think about why you feel this way”) and
indicates contemplation and pondering, “a purposeful turning inward to engage
in cognitive problem solving to alleviate one’s depressive symptoms” (Treynor
et al., 2003, p. 256). The brooding factor consists of more negatively valenced
items (e.g., “think, Why can’t I handle t hings better”), and reflects “a passive
comparison of one’s current situation with some unachieved standard” (Treynor
et al., 2003, p. 256). Both Treynor et al. and Fresco et al. find the brooding factor
to be more related to depressive symptoms than the pondering/reflection factor
and is therefore considered more maladaptive. To delineate potential changes in
rumination following training in MM, we conducted follow-up analyses on brooding
and reflection, predicting a significant decline in both following the MBSR course.
METHOD
Participants
The study was initially intended to be a within-subject design assessing par-
ticipants before and after the MBSR course. Thus, all subjects initially recruited
participated in the course immediately upon enrollment in the study. This sample
consisted of 27 participants, 13 recruited at the VA San Diego Healthcare System
(VASDHS) and 14 at the University of California San Diego (UCSD) Department
of Psychiatry. To increase generalizability of this study, the only exclusion criteria
were symptoms that could prevent participation or interfere with learning, such
as psychosis, current alcohol or substance dependence, or substantial cognitive
impairment such as dementia (assessed via the Shipley, described below, and
neuropsychological evaluations at the VASDHS). These criteria excluded two
participants (both veterans). Two veteran participants dropped out of the course
and did not come in for follow-up assessments. The final within-subject design
sample consisted of 23 participants and will be referred to as the treated sample.
After this data collection was completed, we had an opportunity to modify
the design to also include a between-subject comparison condition. Eleven subjects
(3 veterans and 8 nonveterans), meeting the same inclusion criteria as described
above, were recruited and assessed twice while on a waitlist to start the MBSR
course. All of the waitlist participants were offered participation in the MBSR
course after the waitlist period. For between-subject comparisons, we matched
these 11 waitlist participants on age, gender, and intake Beck Depression Inventory
scores with 11 subjects from the treated sample who had completed four or more
sessions of the MBSR course. Thus, although random assignment of subjects to
waitlist or MBSR condition would clearly have been preferable, the development

438 Ramel, Goldin, Carmona, and McQuaid
of our study design prevented us from conducting randomization, and instead we
chose to carefully match the waitlist sample with MBSR completers. We reasoned
that rather than comparing the 11 waitlist subjects with t he entire MBSR-treated
sample (n = 23), a carefully matched sample would limit confounds by reducing
the number of potential baseline differences between the groups and make in-
terpretation of between-subject comparisons less problematic. As can be seen in
Table I, the matched sample turned out to be very similar on variables that were
not explicitly used as matching criteria, such as years of education, estimated IQ,
ethnicity, previous cognitive behavioral therapy experience, psychiatric diagnoses,
and veteran and psychiatric medication status. For the sake of clarity, the subset of
treated patients who were matched with the waitlist sample will be referred to as
the matched completer sample.
Demographic and diagnostic information of the treated as well as the matched
samples are presented in Table I. All of the 34 participants met a minimum of
8th-grade education requirement and were able to complete the assessment battery
unassisted. All participants met lifetime diagnostic criteria for a mood disorder
and 17 (50%) participants met lifetime criteria for one or more anxiety disorder.
Nine (26%) participants met criteria for a current major depressive episode (MDE)
as part of either major depressive or bipolar disorder. Current diagnostic criteria
was met for four (12%) participants for panic disorder, three (9%) for social
phobia, three (9%) for specific phobia, one (3%) for obsessive–compulsive disorder,
one (3%) for posttraumatic disorder, and three (9%) for generalized anxiety
disorder.
Twenty-two percent of the treated sample, 18% of the matched completer
sample, and 27% of the waitlist sample reported a change in their psychotropic
medication or psychotherapy treatment between the two assessments, and the
differences between groups were not significant.
Procedure
Veteran participants were recruited from among patients referred by VA San
Diego health care practitioners to the MBSR course for mental or physical health
related concerns such as mood and anxiety symptoms or chronic pain. Eleven
(48%) subjects in the treated sample, three (27%) of the waitlist sample, and
two (18%) of the matched completer sample were veterans (see Table I). Upon
contacting the veterans to enroll them in the course, the instructor (PEC) informed
the veterans about the research study, emphasizing that participation in the research
was voluntary and not necessary for participation in the course itself. Veterans who
expressed an interest in being part of the research study were then contacted by one
of the investigators (primarily WR) and assessed approximately 1–2 weeks prior to
the first MBSR session.
Nonveteran participants were recruited at University of California San Diego
and enrolled in the MBSR course either immediately or after a waitlist period.
These participants had previously been recruited via advertisements in local
newspapers and enrolled in either a cognitive behavioral therapy or medication
treatment study for major depression. Whereas the MBSR course was part of

Mindfulness, Affective Symptoms, and Cognitions 439
Table I. Demographic and Diagnostic Information for the Treated and Matched Samples
Matched sample
Total treated sample (n = 23) Completers (n = 11) Waitlist (n = 11)
Variable Mean (SD) N % Mean (SD) N % Mean (SD) N %
Demographics
Age 50.87 (8.87) 50.27 (7.39) 50.18 (10.25)
Education (estimated
years)
16.13 (2.20) 15.36 (2.29) 16.27 (4.94)
Males 15 65 6 55 6 55
Caucasian 23 100 11 100 10 91
Estimated WAIS-R IQ 109.04 (10.05) 107.18 (10.90) 108.0 (14.40)
Veteran status 11 48 2 18 3 27
Previous CBT 17 74 7 64 9 82
Psychiatric medication 13 57 6 55 6 55
Psychiatric diagnoses: Current Lifetime Current Lifetime Current Lifetime
N % N % N % N % N % N %
Mood disorder 8 35 23 100 2 18 11 100 1 9 11 100
Major Depressive
Episode
5 22 23 100 2 18 11 100 1 9 10 91
Two or fewer MDEs 11 48 6 54 7 64
Three or more MDEs 12 52 5 45 4 36
Dysthymia 0 2 9 0 0 0 2 18
Bipolar I & II 3
a
13 5 22 1
a
93 270 0
Anxiety disorder 9 39 13 57 4 36 5 45 2 18 5 45
Current comorbid mood
& anxiety diagnosis
313190
Note. CBT: Cognitive Behavioral Therapy; MDE: Major Depressive Episode; WAIS-R: Wechsler Adult Intelligence Scale—Revised.
a
All current Bipolar I and II patients were in a depressed episode.

440 Ramel, Goldin, Carmona, and McQuaid
standard care offered by the VASDHS to all eligible veterans, it was only available
to nonveterans as part of this research study.
After the study was explained to the participant, he or she read and signed an
informed consent form. A trained clinician then conducted a psychiatric screen and
a structured clinical interview of past and current diagnosis of mood and anxiety dis-
orders. Next, the participants completed a battery of questionnaires and a computer
task. The participants then enrolled in the MBSR course or were placed on a waitlist
before they enrolled in the course. Following the course or waitlist period the
intake assessment battery was readministered. The mean interval between intake
and follow-up assessments was 10.18 weeks, SD = 1.60.
Intervention
The manualized 8-week MBSR course was modeled on the MBSR course
that was developed at the University of Massachusetts Medical Center’s Stress
Reduction Clinic by Kabat-Zinn (1982, 1990). The course consisted of weekly
2-hr classes, one half-day meditation, and daily homework of 45 min of guided
meditation to a tape or 30 min of meditation on their own if a tape was not used.
A psychiatric nurse specialist (PEC), who has participated in a professional training
program under the direction of Kabat-Zinn, led the course each time together with
a doctoral student in clinical psychology. Participants were asked not to practice
other forms of meditation, yoga, or related stress reductions during the course of
the MBSR program. Patients were not asked to change or discontinue any ongoing
psychiatric treatment. An average of eight research participants attended each
mindfulness course. Participants learned experiential MM techniques that use the
breath, body sensations, and yoga to develop nonjudgmental, moment-to-moment
awareness, attentional monitoring, and acceptance.
Instruments
Participants completed an extensive battery including questionnaires, inter-
views, and a computer task. Only those relevant to the current report are described
below.
Diagnostic Assessment
Structured Clinical Interview for the DSM-IV (SCID; Spitzer, Williams, Gibbon,
& First, 1992). The SCID is a standardized semistructured clinician administered
interview for diagnosing DSM-IV mental disorders. It has been shown to have
adequate reliability and validity for most diagnoses, and considered the “gold
standard” for diagnostic assessment. SCID interviews were conducted by the first
and second author (WR and PRG), clinical psychology graduate students, and
research nurses who were trained to criterion by research staff at the UCSD Clinical
Research Center. Diagnoses were reviewed and agreed upon in consensus meetings
together with a licensed clinical psychologist (JM), who is experienced in making
psychiatric diagnoses based on the SCID.

Mindfulness, Affective Symptoms, and Cognitions 441
Affective Symptom Questionnaires
Beck Depression Inventory (BDI; Beck et al., 1979; Beck, Ward, Mendelson,
Mock, & Erbaugh, 1961). The BDI is a 21-item, standard self-report measure of
depressive symptomatology that is widely used in research. The BDI has been found
to correlate .68 with interview-based assessments of depression (Rabkin & Klein,
1987) and has good internal consistency (α = .84) and moderate test-retest reliability
(r = .69; Yin & Fan, 2000). The internal consistency at intake in the current sample
was high (α = .92).
Spielberger State–Trait Anxiety Inventory (STAI; Spielberger, Gorush,
Lushene, Vagg, & Jacobs, 1983). The STAI is composed of two 20-item scales, one
assessing state anxiety and the other assessing trait anxiety. The STAI generally
has high internal consistency for both the trait (α = .91 for working adults) and the
state (α = .93 for working adults) scales, and a relatively low STAI-state test-retest
correlation (Spielberger et al., 1983). The internal consistency at intake in the
current sample was high and in the expected range for both scales, especially the
state form (α = .96 for STAI-state and .82 for STAI-trait).
Cognitive Measures
Dysfunctional Attitudes Scale (DAS; Weissman, 1979; Weissman & Beck, 1978).
The DAS is a self-report questionnaire designed to measure typical depressogenic
beliefs. There are several versions, and the 40-item version (Form A) was used
in the current study. Each item is rated on a 7-point Likert scale ranging from
totally agree to totally disagree. Factor analytic studies have suggested two factors:
Perfectionism/Performance Evaluation and Need for Approval (Cane, Olinger,
Gotlib, & Kuiper, 1986; Imber et al., 1990). The two subscales have been shown
to have good internal consistency (α = .91 for Need for Approval and α = .82
for Perfectionism) and be moderately correlated with each other (r = .59) as well
as with self-report measures of depression (Zuroff et al., 1999). In the current
sample, the two factors were moderately to highly correlated (r = .65 at intake,
r = .74 at follow-up, both p < .0001) and showed good internal consistency at intake
(approval: α = .87; perfectionism: α = .88).
Response Style Questionnaire (RSQ; Nolen-Hoeksema & Morrow, 1991). The
RSQ is a 71-item self-report measure that assesses the level at which individuals
engage in various cognitive coping styles. It has four subscales (Rumination,
Distraction, Problem-solving, and Dangerous activities), although only rumination
responses (22 items) were analyzed in this study. The internal consistency of the
Rumination scale was reported to be .89 (Cronbach’s α), and subjects’ responses to
this scale have been shown to correlate significantly with depressed mood (Nolen-
Hoeksema & Morrow, 1991).
In addition to the original subscales, scores were also calculated for reflection
and brooding based on recent refinements of the RSQ (Fresco et al., 2004; Treynor
et al., 2003). The scales have adequate internal consistency (for brooding α =
.77–.80; Fresco et al., 2004; Treynor et al., 2003; for reflection α = .72; Treynor
et al., 2003) and moderate test-retest reliability (r = .60 for reflection and r = .62
for brooding; Treynor et al., 2003).

442 Ramel, Goldin, Carmona, and McQuaid
In the current sample brooding and reflection were uncorrelated at both time
points as were reflection and rumination (although r = .29, p < .09 at intake; r = .10,
p = .57 at follow-up). However, brooding and rumination were highly correlated
at both intake and follow-up (r = .79 and .73, respectively, both p < .0001). The
internal consistency at intake for the Rumination scale was relatively high (α =
.87), but as other researchers have found, the coefficient alphas for brooding and
reflection were relatively low (.70 and .72, respectively), which may be due to the
small number of items.
Additional Questionnaires
General Follow-Up Questionnaire (GFQ). At the follow-up assessment the
research participants were given a questionnaire designed specifically for this study.
The brief questionnaire assessed any significant changes in other treatments the
patient was receiving (medication or psychosocial) that occurred since the intake,
and the amount and frequency of meditation practiced since the beginning of the
course. While participants were given a form to track their meditation and yoga
practice on a weekly basis during the MBSR course, only about half the participants
did so in a reliable manner. Correlational analysis indicated a medium to high
association (r = .66, p = .04) between amount of meditation practice reported on the
GFQ and on the weekly meditation tracking form. As every participant estimated
the total amount of meditation practiced at the follow-up assessment, these data
were used in the statistical analyses.
Institute of Living Scale (Shipley; Zachary, 1986). The Shipley consists of a test
of 40 vocabulary items and a test of 20 abstraction pattern recognition tasks (e.g.,
AB BC CD D
). These tests assess verbal and logical reasoning abilities and provide
an estimate of general intellectual aptitude (IQ). The Shipley correlates highly with
the revised Wechsler Adult Intelligence Scale (WAIS-R). The Shipley was given to
participants once, either at the intake or follow-up session.
RESULTS
Treated Sample
Descriptive and Preliminary Analyses
Twenty-three individuals were studied before and after completion of the
MBSR course and make up the treated sample. A ll reported analyses examine the
treated sample unless otherwise specified. Means and standard deviation for the
intake and follow-up affective symptoms (BDI and STAI) and cognitions (RSQ and
DAS) are presented in Table II. Eighteen participants completed four or more of
the eight weekly MBSR sessions (completers) and five participants completed three
or fewer sessions (noncompleters). There were no significant differences between
the completers and noncompleters on any of the demographic variables or baseline
measures, so the primary analyses were conducted on the entire treated sample.

Mindfulness, Affective Symptoms, and Cognitions 443
Table II. Treated Sample Intake and Follow-Up Affective Symptoms and
Cognitive Measures
Variable Intake mean (SD) Follow-up mean (SD)
Affective symptoms
BDI 16.41 (12.04) 11.50 (10.44)
∗
STAI-state 46.77 (16.15) 44.48 (13.81)
STAI-trait 52.34 (15.30) 46.58 (11.35)
∗
Cognitive measures
RSQ-Rumination 56.43 (9.00) 49.57 (8.33)
∗∗
RSQ-Brooding 13.43 (3.29) 11.61 (3.22)
∗∗
RSQ-Reflection 11.26 (3.08) 9.83 (2.12)
∗
DAS-Approval 45.78 (13.05) 41.43 (12.64)
∗
DAS-Perfectionism 50.91 (16.46) 45.48 (17.20)
DAS-Total 142.65 (36.24) 127.61 (35.22)
∗
Note. BDI: Beck Depression Inventory; DAS: Dysfunctional Attitude
Scale; RSQ: Response Style Questionnaire; STAI: Spielberger State–Trait
Anxiety Inventory. A total of 23 participants are included in the treated
sample. The BDI and STAI-state measures were missing from one pa-
tient’s follow-up questionnaire packet, and therefore only 22 participants
had these measures at follow-up.
∗
p < .05 and
∗∗
p < .005 for within-sample paired t test.
Within-Subject Analyses: Change in Affective Symptoms and Cognition Over Time
Paired t tests compared intake and follow-up outcome measures (BDI, STAI-
state, STAI-trait, RSQ-Rumination, DAS-Approval, and DAS-Perfectionism) to
assess whether they changed over time in the treated sample.
Both the BDI [t(21) = 2.43, p < .024] and STAI-trait [t(22) = 2.52, p < .02]
significantly decreased following the MBSR course and yielded medium effect sizes
(Cohen’s D = .52 for both BDI and STAI-trait). Among the cognitive measures,
RSQ-Rumination [t(22) = 3.82, p < .001] and DAS-approval [t(22) = 2.17, p < .05]
significantly decreased from intake to follow-up assessments, with effect sizes in the
large range for RSQ-Rumination (Cohen’s D = .80) and in the medium range for
DAS-approval (Cohen’s D = .45). Neither STAI-state [t(22) = .94, p = .36] nor
DAS-perfectionism [t(22) = 1.68, p = .11.] changed significantly over time.
To test whether decrease in affective and DAS symptoms accounted for the
reduction in rumination and whether rumination decreases accounted for reductions
in affective and DAS symptoms, the analyses were rerun using changes in BDI,
STAI, RSQ-Rumination, and DAS as covariates. Changes in rumination remained
significant when controlling for intake to follow-up changes in BDI, STAI-trait,
and DAS-approval [F(18) = 11.12, p < .004]. As predicted, changes in BDI,
STAI-trait, and DAS-approval were no longer significant when controlling for
changes in rumination. This indicates that change in affective symptoms and need
for approval do not explain the reductions in rumination following the MBSR
course. However, changes in rumination explain a significant part of t he decrease
in affective symptoms and DAS-approval following MM practice.
To further examine the decrease in rumination while separating out items that
are redundant with depressive symptomatology, the analyses were repeated using
the brooding and reflection scales. The results revealed significant decreases in

444 Ramel, Goldin, Carmona, and McQuaid
both reflection [t(22) = 3.86, p < .001] and brooding [t(22) = 2.51, p < .02], even
after controlling for intake to follow-up changes in BDI and STAI-trait [reflection:
F(19) = 8.44, p < .009; brooding: F(19) = 7.32, p < .02].
Additional Analyses
Completers Vs. Noncompleters. Repeated measures ANOVA with one
within-subject variable (intake and follow-up) and one between-subject variable
(completer and noncompleter) revealed no significant interaction between com-
pleters’ and noncompleters’ change in either affective symptoms (BDI and STAI)
or cognitions (RSQ and DAS) across the two assessments.
Current Depression Status and History. While there was no significant differ-
ence between currently and formerly depressed participants in their change in either
rumination or dysfunctional attitudes before and after the MBSR course, currently
depressed participants demonstrated a significantly greater decline in BDI [F(20) =
4.48, p < .05], STAI-state [F(20) = 4.70, p < .05], and STAI-trait [F(21) = 6.42, p
< .02] than formerly depressed participants. We also followed up on Teasdale and
colleagues’ finding on differential effects of relapse prevention for individuals with
two or fewer versus three or more MDEs (Teasdale et al., 2000). Of the affective
symptoms and cognitive measures, the only variable that approached a significant
interaction was the STAI-trait [F(21) = 3.72, p = .067], indicating that individuals
with two or fewer MDEs reported a smaller reduction in trait anxiety symptoms (1.4
points) than did participants with three or more MDEs (9.75 points) from intake to
follow-up.
Gender Effects. Because women are more likely than men to engage in
rumination (Nolen-Hoeksema, 1987), we tested potential gender interactions. We
did not find any significant differences between men and women on their changes in
affective and cognitive symptoms following the MBSR course.
Within-Subject Analyses: Regression
Hierarchical linear regression analyses tested if amount of meditation prac-
ticed, as reported by participants on the GFQ, predicted affective symptoms, or
cognitions at the completion of the MBSR course. The average amount of medita-
tion practiced during the 8 weeks of the course was 11.46 hr, or about 1.4 hr a week
(SD = 15.35 hours, range 0–72 hours). Amount of meditation practiced significantly
predicted follow-up rumination values after controlling for intake rumination [t(20)
=−2.21, p < .04, β =−.38], indicating that the more meditation practiced, the less
ruminative cognitions were reported at follow-up, as depicted in Fig. 1. Amount
of meditation practiced uniquely accounted for 15% of the variance in follow-up
rumination. As can be seen in Fig. 1, most participants engaged in very little practice,
and the effect appears to be strongly influenced by one participant who practiced a
considerable amount of meditation (about 1.3 hr per day). However, the amount
of MM practiced by this participant fell within the range of what were normal
expectations for practice. Amount of meditation practiced remained significant in
predicting tendencies to ruminate at the follow-up assessment when controlling
for intake to follow-up changes in BDI and STAI-trait [t(17) =−2.14, p < .05,

Mindfulness, Affective Symptoms, and Cognitions 445
Fig. 1. Partial regression plot controlling for intake RSQ-Rumination. Amount of
meditation practiced significantly predicts follow-up rumination [t(20) =−2.21,
p < .04, β =−.38]. Amount of meditation practiced uniquely accounted for 15%
of the variance in follow-up rumination. The data on the x-axis are shifted to the
left because values are estimates after partialling out intake RSQ-Rumination
scores, which is why there are values less than 0.
β =−.37]. Amount of meditation practiced did not significantly predict follow-up
depressive or anxious symptoms, or DAS values after controlling for intake values
in the combined sample.
Linear regressions were also conducted on the rumination subscales, brood-
ing and reflection. The r esults indicated a significant decrease in reflection with
increased meditation practice after controlling for intake reflection scores [t(20) =
−2.40, p < .03, β =−.41], but not for brooding [t(20) =−1.65, p = .115]. Amount
of meditation practiced uniquely accounted for 17% of the variance in follow-up
reflection scores. Meditation practice significantly predicted decreased reflection
when controlling for changes in BDI and STAI-trait [t(17) =−2.10, p = .05].
Matched Sample
Descriptive and Preliminary Analyses
In order to examine our hypotheses in a between-subject design, 11 participants
were recruited and tested twice while on a waitlist to participate in the MBSR
course. As noted above, the waitlist participants were matched on age, gender, and
intake BDI scores with 11 completers from the treated sample (see Table I). An
average of 10.48 weeks (SD = 1.52, range = 7.7–12.6 weeks) elapsed between the

446 Ramel, Goldin, Carmona, and McQuaid
Table III. Matched Sample Intake and Follow-Up Affective Symptoms and Cognitive Measures
Matched sample
Completers (n = 11) Waitlist (n = 11)
Variable Intake mean (SD) FU mean (SD) Intake mean (SD) FU mean (SD)
Affective symptoms
BDI 12.54 (10.61) 9.45 (6.73) 13.09 (7.26) 11.55 (7.83)
STAI-state 39.82 (14.24) 39.86 (10.51) 41.16 (13.69) 42.00 (10.44)
STAI-trait 46.28 (15.94) 42.39 (11.37) 50.00 (12.13) 52.00 (12.03)
Cognitive measures
RSQ-Rumination 56.82 (6.81) 46.00 (8.07)
∗∗
48.18 (9.11) 49.36 (11.21)
RSQ-Reflection 12.45 (3.24) 10.00 (2.28)
∗
10.10 (2.59) 9.82 (2.44)
RSQ-Brooding 13.55 (3.53) 11.00 (3.10)
∗
11.64 (2.20) 11.45 (3.30)
DAS-Approval 41.64 (12.89) 39.36 (11.97) 43.91 (12.75) 45.55 (10.60)
DAS-
Perfectionism
47.27 (17.96) 42.18 (20.32) 53.36 (15.83) 54.82 (19.51)
DAS-Total 131.45 (37.84) 119.45 (38.56) 146.45 (34.63) 147.18 (40.74)
Note. BDI: Beck Depression Inventory; DAS: Dysfunctional Attitude Scale; FU: follow-up; RSQ:
Response Style Questionnaire; STAI: Spielberger State–Trait Anxiety Inventory.
∗
p < .005 and
∗∗
p < .001 for within-sample paired t test.
initial and follow-up assessments in the waitlist sample and 10.08 weeks (SD = 1.69,
range = 8–13.6) in the completer sample, a nonsignificant difference (p = .51).
Descriptive statistics on the affective symptom and cognitive measures at intake
and follow-up of the matched sample are displayed in Table III. Overall, the waitlist
and the completer samples were similar and did not significantly differ on any of
the demographic and intake diagnostic, affective or cognitive variables except for
intake rumination (and its subscales), where the completers reported higher values
[t(20) = 2.52, p <.02].
Between-Subject Comparisons: Completers Vs. Waitlist
Repeated measures ANOVAs with one within-subject factor (time) and one
between-subject factor (completers vs. waitlist) were conducted to examine changes
in affective symptoms (BDI, STAI-state, and STAI-trait) and cognitions (DAS-
Approval, DAS-Perfectionism, and RSQ-Rumination) from intake to follow-up in
the matched MBSR waitlist and completer sample. We expected to find significant
interaction effects indicating decreases in affective symptoms and cognitions for the
MBSR completer sample and no changes in the waitlist sample (i.e., interaction
effects). The results demonstrated a significant interaction effect for the RSQ-
Rumination scale [F(20) = 10.78, p < .004], and this effect held up also after control-
ling for BDI and STAI-trait intake to follow-up change values [F(18) = 10.04, p <
.005]. The effect size for this interaction was large, Cohen’s D = 1.47. As can be seen
in Fig. 2, the completers’ rumination levels decreased considerably following the
MBSR course (from an intake mean of 56.82 to a follow-up mean of 46, a reduction
of 10.8 points), while the waitlist participants’ rumination levels increased slightly
(from an intake mean of 48.18 to a follow-up mean of 49.36, an increase of 1.18
points). No significant interaction effects between the waitlist and completers were
found for the DAS and affective symptoms measures (BDI and STAI).

Mindfulness, Affective Symptoms, and Cognitions 447
Fig. 2. RSQ-Rumination by group interaction effect for MBSR completers vs.
waitlist participants at intake and follow-up [F(20) = 10.78, p < .004]. The results
remained significant when controlling for changes in depressive and anxious symp-
toms.
Again, we followed up the finding on rumination by analyzing the reflection
and brooding subscales of rumination. We found a significant interaction for
reflection [F(20) = 5.52, p < .03; Cohen’s D = 1.05], and an interaction that
approached significance for brooding [F(20) = 3.88, p = .063; Cohen’s D = .88].
These interaction effects remained when controlling for intake to follow-up changes
in BDI and STAI-trait [reflection: F(18) = 5.38, p < .04; brooding: F(18) = 4.06, p
= .059].
DISCUSSION
The goals of this study were to examine three variables that, based on prior
research and theory, were predicted to be modified by training in mindfulness
meditation (MM): affective disturbances, dysfunctional attitudes, and rumination.
While we expected to see a decline in all these f actors, drawing upon the theory
derived from Segal, Teasdale, and Williams (Segal et al., 2002; Teasdale et al., 1995),
we predicted that the primary area of change would be the manner in which MM
practitioners learned to relate to cognitions, emotions, and sensations. As such, we
expected to find the most reliable reductions across our analyses in rumination and
predicted change in rumination to account for reductions in affective disturbance
and dysfunctional attitudes.
Our results were generally consistent with our predictions. Reductions in
ruminative tendencies were found across our three methods of analyses (within-
subject repeated measures ANOVA, multiple regression, and between-subject
analyses). Reductions in affective symptoms and dysfunctional attitudes were only
found in our within-subject repeated measures analyses, and were not significant
when rumination was used as a covariate.

448 Ramel, Goldin, Carmona, and McQuaid
The between-subject analyses based on a matched sample found significantly
greater reductions in ruminative tendencies for the MM participants compared to
the waitlist participants (Fig. 2). This result remained when accounting for changes
in depressive and anxious affect. We found no significant interaction effects for
affect or dysfunctional attitudes between the waitlist and active MM participants.
Given the significantly different intake rumination values, an interpretation of
regression toward the mean can be posited. However, the pattern of crossing lines
and switching of mean differences from intake to follow-up make this and other
alternative causal explanations, such as scaling or ceiling/floor artifacts, unlikely
(Cook & Campbell, 1979, pp. 111–112). As elaborated by Cook and Campbell, if a
statistical regression were to explain this finding, the completers’ grand mean (51.4)
would have to be lower than the grand mean of the waitlist group (48.8); this is not
the case.
Also in support of our hypothesis, our regression analyses indicated that the
more MM practiced, the less rumination was reported at the follow-up assessment,
even after controlling for intake levels of rumination as well as changes in affective
disturbance (BDI and STAI-trait). Amount of meditation uniquely accounted for
15% of the variance in follow-up rumination levels after controlling for intake
rumination. However, there was a notable leverage point (outlier) in our sample. It
is worth pointing out, that while this outlier participant practiced MM as homework
more than anybody else, she practiced only slightly more than instructed (about
an hour and 15 min rather than 45 min per day). In contrast, the treated sample
reported practicing on average about one and a half hour per week. This, in addition
to the general regression finding, speaks to the potential importance of a dedicated
MM practice to obtain the benefit of reduced rumination. A larger sample with more
careful tracking of participants’ MM practice time is needed to probe this hypothesis
further.
The relatively low level of MM practice overall is concerning, and likely limited
the overall outcomes. We do not know why the compliance in the sample was
this poor, but we speculate that part of the explanation may be the nature of our
sample, that is, participants with a history of relatively significant psychopathology.
However, few studies on MBSR report the average number of hours practiced per
week during the MBSR course, so it is unclear whether the homework adherence of
our sample is unusually low.
Overall, our results suggest that MM practice based on an 8-week stress
reduction course is primarily effective at decreasing rumination, and that changes
in rumination account for reductions in maladaptive cognitive content and affective
symptoms, specifically depressive and anxious symptoms and dysfunctional beliefs
relating to need for approval. Although we have emphasized the relationship
between rumination and depression, as every participant in our sample met a
lifetime diagnosis of a mood disorder, it is worth noting that half of the sample
also met lifetime criteria for an anxiety disorder. Moreover, given the high level
of comorbidity between depressive and anxious disorders, it is not surprising that
cognitive processes similar to rumination are also central features of anxiety. As
shown in a recent study, depressive rumination (dwelling on the negative) and
worry are moderately correlated (r = .46), and both demonstrate similar positive

Mindfulness, Affective Symptoms, and Cognitions 449
associations with anxious and depressive symptoms (rs range from .30 to .59;
Fresco, Frankel, Mennin, Turk, & Heimberg, 2002). Segerstrom and colleagues
demonstrated that the shared variance between worry and rumination can be
explained by a single latent variable, repetitive thought, which is related to both
anxiety and depression (Segerstrom, Tsao, Alden, & Craske, 2000). While results
from these studies indicate that anxious and depressive thought processes, such as
worry and rumination, are similar, anxious and depressive thought content may
not be. There is evidence supporting a content-specificity hypothesis regarding
anxious and depressive self-talk, with rumination containing thoughts of past losses,
incompetence, rejection and failures, and worry involving future-oriented and
“questioning” thoughts of threat, harm, and uncertainties (Kendall & Ingram, 1989;
Safren et al., 2000). We did not include a worry questionnaire in our study, but it
is possible that the reduction in rumination following the MBSR course primarily
reflects a decrease in repetitive thoughts. If so, this alteration in thought process
may be a general mechanism by which MBSR similarly impacts worry in anxiety
disorders and rumination in depressive disorders, while disorder specific cognitive
content may not be directly modified.
It is likely that the emphasis in MM training on observing and noting thoughts,
feelings, and sensations as passing and impermanent events of the mind facilitates
a less habitual pattern of reacting to and ruminating on arising thoughts and
feelings. This alternative and more “decentered” (Segal et al., 2002; Segal, Williams,
Teasdale, & Gemar, 1996, p. 379) way of relating to the contents of the mind may be
incompatible with a prolonged dwelling on thoughts and feelings, and may be the
mechanism by which training in MM reduces ruminative tendencies. Consistent with
this theory, MM training was associated with a significant reduction in our treated
sample in both brooding and reflection, two subfactors of the RSQ-Rumination
scale in which items with obvious overlap with depressive thinking have been re-
moved. Compared to a matched waitlist sample, reflection was significantly reduced
in the MBSR course participants, while the reductions in brooding approached
significance (the latter finding may be due to limited statistical power). However,
the findings from our regression analyses suggest that reflection is indeed a critical
component of rumination that is modified by MM training—the more meditation
practiced, the less reflection was reported at follow-up. (A similar relationship was
found for brooding, but again, it did not reach statistical significance.)
Although reflection appears to be more adaptive than brooding, in the sense
that it correlates less with anxious and depressive symptoms than does brooding
(Fresco et al., 2003; Treynor et al., 2003), its scale items indicate a tendency to
ponder, mull over, or s peculate about one’s actions, thoughts, feelings, and attempts
to understand cognitions and emotions by analyzing them. While there is less
emphasis on making derogatory self-judgments or comparing self to others in the
reflection scale as opposed to the brooding scale, the very act of reflecting and
contemplating in this manner may be a crucial initial step of what places a vulnerable
individual at risk for depression relapse or recurrence. As pointed out by Segal
et al., what remains when depression is over is a “tendency to react to small
changes in mood with large changes in negative thinking” (Segal et al., 2002, p. 33).
Although it may appear to be a rational and helpful problem-solving technique

450 Ramel, Goldin, Carmona, and McQuaid
and way of understanding the current states of mind, the process of monitoring
and contemplating thoughts and emotions is likely to instigate further dwelling,
brooding, questioning, and judgments about current versus desired state of self,
which, for a vulnerable population, may perpetuate destructive emotions. Thus,
reflecting on why or how one is feeling a certain way may be exactly what starts
a delicate and potentially harmful process that may escalate to a major depressive
episode.
The more robust association of reductions in rumination following MM rather
than dysfunctional attitudes or affective symptoms supports the focus on altering the
relationship to cognitions in MM practice, rather than changing the content of the
thoughts themselves, as is emphasized in other treatments (e.g., cognitive therapy).
It would be intriguing to test whether, after accounting for psychiatric symptoms, the
primary mechanism of change in cognitive therapy involves a reduction in beliefs
in dysfunctional attitudes (i.e., thought content), while the primary mechanism of
change in MBSR entails a decrease in ruminative tendencies (i.e., relationship to
thoughts), or if both forms of intervention affect similar mechanisms. Of note,
about two thirds (74%) of the treated sample in the current study had previously
participated in cognitive behavioral therapy for depression and 57% were on psy-
chotropic medication. Thus, the MBSR course was offered as a secondary or tertiary
treatment and the fact that we found a significant decrease in rumination and in
BDI in our treated sample (significantly greater in currently depressed compared
to formerly depressed participants), which appears to be primarily accounted for
by decreases in rumination, suggests that MBSR is a worthwhile intervention for
individuals with residual mood symptoms. In addition, that mindfulness appears
to reduce rumination even for patients who have been treated with cognitive
behavioral therapy suggests that either cognitive behavioral therapy did not reduce
rumination, or that mindfulness can provide further benefit above and beyond that
of cognitive behavioral therapy in terms of reducing rumination. This makes some
sense theoretically. The cognitive behavioral therapy intervention offered at the
VA San Diego and UCSD, which participants in this study received, has a strong
behavioral activation emphasis, and the cognitive interventions tended to focus
on content (e.g., accuracy of automatic thoughts) rather than the thought process
(e.g., were patients ruminating). This form of CBT may not have been particularly
suited to reducing rumination, and the MBSR intervention may therefore have
complimented the treatment by addressing thinking from a new perspective.
As conceptualized by Kabat-Zinn ( 1990) and Segal et al. (2002), mindfulness
contrasts with a ruminative, evaluative, and reflective mind set because i t involves
noting states of mind and bodily sensations with an accepting and nonjudgmental
attitude without elaborating on the content of thoughts and feelings. In trying to
prevent a potentially detrimental cognitive-affective process to unfold, mindfulness
practice can help to anchor a person in the present moment by identifying, even
labeling, what arises in the stream of consciousness without becoming engrossed
in or judging it (e.g., upon noticing a thought such as “I am useless at this,”
the practitioner may simply register “judgmental thought” or “thinking”). Thus,
rather than avoiding or becoming absorbed in the content of the body and mind,
MM teaches attentional skills that balance these extremes while remaining alert

Mindfulness, Affective Symptoms, and Cognitions 451
and observing the patterns of the mind. With time, t his awareness and pattern
recognition can contribute to a decentered perspective of thoughts, sensations, and
feelings—a sort of metacognitive awareness—which can help reduce the potency of
the literal meaning of thoughts and feelings and the tendency to become absorbed
in a ruminative state of mind. The very incompatibility of these alternate states
of mind, a ruminative versus a decentered, may be what contributes to improved
regulatory skills in the face of mood swings and decreases the risk of relapse.
As shown by an increasingly large literature on prevention of relapse, reducing
residual symptoms with continued interventions following primary treatment can
significantly reduce recurrence and relapse of major depression (e.g., Fava, Ruini,
Rafanelli, & Grandi, 2002; Hollon, DeRubeis, Shelton, & Amsterdam, 2001; Jarrett
et al., 2001; Paykel et al., 1999; Teasdale et al., 2000). The current study did not con-
duct long-term follow-up assessments with participants and therefore cannot assess
MBSR’s potential efficacy as a relapse prevention intervention. It is hypothesized,
however, that the addition of MBSR following cognitive behavioral therapy may
facilitate further relapse-prevention effects of regular cognitive behavioral therapy
by reducing residual rumination.
Our study has several design considerations that are worth noting, as they
restrict the generalizability and affect interpretation of our results. First of all, we
used a small, nonrandomized sample of individuals. The small sample size cautions
us in interpreting a lack of a significant relationship between our independent
and dependent variables, as a null finding may simply reflect a lack of statistical
power. The self-selecting nature of the sample calls into question how acceptable
and effective a mindfulness-based intervention would be to the general population,
and thus restricts the external validity of the study. The lack of random assignment
of participants t o the waitlist versus treatment groups limits the study’s internal
validity. Because of this, we cannot rule out nonmeasured differences between the
samples that may contribute to the rumination outcome. However, we matched our
groups on relevant demographic and baseline variables and did not find baseline
group differences on nonmatched factors other than rumination and its subscales;
thus, we can infer some degree of pretest–posttest change due to the MBSR
treatment. Further, our between-sample findings replicate the within-sample results,
both of which were derived from a priori predictions.
Throughout this paper we have used change scores (intake to follow-up) for
the BDI and STAI-trait as control factors when testing the effect of changes in
rumination and dysfunctional attitudes following the MBSR course. Although it
could be argued that follow-up BDI and STAI-trait values would be better suited
as covariates, we reasoned that change scores were preferable as they controlled for
baseline symptoms while also addressing change over time.
Participants in our study were not restricted in obtaining new or changing their
ongoing health care treatments while enrolled in the MBSR course, but changes
were reported at the follow-up assessment. These changes were not taken into
account in our statistical analyses, primarily because of limited power. However,
a greater proportion of the waitlist sample reported changes in their treatment from
intake and follow-up than the matched completer as well as the treated sample.
Thus, it is unlikely that the differences we found were due to additional treatments

452 Ramel, Goldin, Carmona, and McQuaid
in the MBSR group. It is also worth pointing out that this study does not add
information regarding the effectiveness of the MBSR course as a primary treatment
for depression, because the majority of our sample consisted of individuals already
treated with validated forms of therapy for mood disorder. Moreover, we did not
examine the course leaders’ adherence to the MBSR treatment protocol.
There are several future directions for this work. A randomized design with
an active control condition and larger and more demographically varied samples
would improve inferences of causality and control for self-selection biases. Such a
design could also test whether changes in rumination mediate changes in treatment-
outcome, something we were unable to do due to our limited sample size. It would
also be interesting to contrast a sample of currently versus formerly depressed
individuals. Additionally, a more rigorous and frequent tracking of participants’
practice of MM, life events, and medical and/or psychological interventions would
be desirable in order to better isolate the specific effects of MM practice.
In conclusion, the findings from this preliminary study suggest that the MBSR
course is a good candidate as an intervention for reducing unhelpful thought
processes such as rumination. Longitudinal studies with larger samples that include
an extended follow-up period after the completion of the MBSR course are needed
to provide information whether reductions in rumination i s a key mechanism in
lowering the number of relapses and recurrences of psychiatric disorders such as
depression.
ACKNOWLEDGMENTS
The first author was supported by a training grant T32 MH19934 from the
National Institutes of Mental Health to the University of California San Diego
Intervention Research Center for Late Life Psychosis. This study was also s upported
by the UCSD Mental Health Clinical Research Center (MH30914). Philippe R.
Goldin is now at Department of Psychology, Stanford University. The authors wish
to thank the patients who participated in this research and made the study possible.
Also thanks to Rosella Gacoscos and Jeanette Shaw for help with data collecting
and processing, Kathy Resovsky, Deborah Greenfield, and Laura Sutton for help
in diagnostic training and screening, and Mark G. Williams, Zindel Segal, and Jon
Kabat-Zinn for valuable feedback on earlier drafts of this paper.
REFERENCES
Astin, J. A. (1997). Stress reduction through mindfulness meditation. Psychotherapy and Psychosomatics,
66, 97–106.
Baer, R. A. (2003). Mindfulness training as a clinical intervention: A conceptual and empirical review.
Clinical Psychology: Science and Practice, 10, 125–143.
Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical aspects.NewYork:Hoeber.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive therapy for depression. New York:
Guilford Press.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring
depression. Archives of General Psychiatry, 4, 53–63.

Mindfulness, Affective Symptoms, and Cognitions 453
Bishop, S. R. (2002). What do we really know about Mindfulness-Based Stress R eduction? Psychoso-
matic Medicine, 64, 71–84.
Cane, D. B., Olinger, J., Gotlib, I. H., & Kuiper, N. A. (1986). Factor structure of the dysfunctional
attitude scale in a student population. Journal of Clinical Psychology, 42, 307–309.
Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation: Design & analysis issues for field settings.
Boston: Houghton Mifflin Company.
Davis, R. N., & Nolen-Hoeksema, S. (2000). Cognitive inflexibility among ruminators and nonrumina-
tors. Cognitive Therapy and Research, 24, 699–711.
Fava, G. A., Ruini, C., Rafanelli, C., & Grandi, S. (2002). Cognitive behavior approach to loss of clinical
effect during long-term antidepressant treatment: A pilot study. American Journal of Psychiatry,
159, 2094–2095.
Fresco, D. M., Armey, M., Mennin, D. S., Turk, C. L., & Heimberg, R. G. (2004). Brooding and
pondering: Isolating the active ingredients of depressive rumination with exploratory and confirmatory
factor analysis. Submited for publication.
Fresco, D. M., Frankel, A. N., Mennin, D. S., Turk, C. L., & Heimberg, R. G. (2002). Distinct and
overlapping features of rumination and worry: The relationship of cognitive production to negative
affective states. Cognitive Therapy and Research, 26, 179–188.
Goldenberg, D. L., Kaplan, K. H., Nadeau, M. G., Brodeur, C., Smith, S., & Schmid, C. H. (1994).
A controlled study of a stress-reduction, cognitive-behavioral treatment program in fibromyalgia.
Journal of Musculoskeletal Pain, 2, 53–66.
Haaga, D. A., Dyck, M. J., & Ernst, D. (1991). Empirical status of cognitive theory of depression.
Psychological Bulletin, 110, 215–236.
Hayes, S. C., Strosahl, K. D., & Wilson, K. G. (1999). Acceptance and commitment therapy: An
experiential approach to behavior change. New York: Guilford Press.
Hollon, S. D., D eRubeis, R. J., Shelton, R. C., & Amsterdam, J. (2001). Cognitive therapy and the
prevention of relapse in severely depressed outpatients. Paper presented at the World Congress of
Behavioral and Cognitive Therapy, Vancouver, British Columbia, Canada.
Imber, S. D., Pilkonis, P. A., Sotsky, S. M., Elkins, I., Watkins, J. T., Collins, J. F., et al. (1990). Mode-
specific effects among three treatments of depression. Journal of Consulting and Clinical Psychology,
58, 352–359.
Ingram, R. E., Miranda, J., & Segal, Z. V. (1998). Cognitive vulnerability to depression. New York:
Guilford Press.
Jarrett, R. B., Kraft, D., Doyle, J., Foster, B. M., Eaves, G. G., & Silver, P. C. (2001). Preventing recurrent
depression using cognitive therapy with and without a continuation phase. Archives of General
Psychiatry, 58, 381–388.
Just, N., & Alloy, L. B. (1997). The response styles theory of depression: Tests and an extension of the
theory. Journal of Abnormal Psychology, 106, 221–229.
Kabat-Zinn, J. (1982). An outpatient program in behavioral medicine for chronic pain patients based on
the practice of mindfulness meditation: Theoretical considerations and preliminary results. General
Hospital Psychiatry, 4, 33–47.
Kabat-Zinn, J. (1990). Full catastrophy living: Using the wisdom of your body and mind to face stress,
pain, and illness. New York: Delacorte.
Kabat-Zinn, J. (1994). Wherever you go there you are: Mindfulness meditation in everyday life. New York:
Hyperion.
Kabat-Zinn, J., Lipworth, L., & Burney, R. (1985). The clinical use of mindfulness meditation for the
self-regulation of chronic pain. Journal of Behavioral Medicine, 8(2), 163–190.
Kabat-Zinn, J., Massion, A. O., Kristeller, J., Peterson, L. G., Fletcher, K. E., Pbert, L., et al. (1992).
Effectiveness of a meditation-based stress reduction program in the treatment of anxiety disorders.
American Journal of Psychiatry, 149, 936–943.
Kabat-Zinn, J., Wheeler, E., Light, T., Skillings, A., Scharf, M. J., Cropley, T. G., et al. (1998). Influence
on mindfulness meditation-based stress reduction intervention on rates of skin clearing patients
with moderate to severe psoriasis undergoing phototherapy (uvb) and photochemotherapy (puva).
Psychosomatic Medicine, 60, 625–632.
Kendall, P., & Ingram, R. (1989). Cognitive-behavioral perspectives: Theory and research on depression
and anxiety. In P. Kendall & D. Watson (Eds.), Anxiety and depression: Distinctive and overlapping
features. (pp. 27–53). Orlando, FL: Academic Press.
Kristeller, J. L., & Hallett, C. B. (1999). An exploratory study of a meditation-based intervention for
binge eating disorder. Journal of Health Psychology, 4, 357–363.
Kutz, I., Borysenko, J. Z., & Benson, H. (1985). Meditation and psychotherapy: A rationale for the
integration of dynamic psychotherapy, the relaxation response, and mindfulness meditation. The
American Journal of Psychiatry, 142, 1–8.

454 Ramel, Goldin, Carmona, and McQuaid
Kutz, I., Leserman, J., Dorrington, C., Morrison, C. H., Borysenko, J. Z., & Benson, H. (1985).
Meditation as an adjunct to psychotherapy: An outcome study. Psychotherapy and Psychosomatics,
43, 209–218.
Linehan, M. M. (1993). Cognitive-behavioral treatment of borderline personality disorder. New York:
Guilford Press.
Lyubomirsky, S., & Nolen-Hoeksema, S. (1995). Effects of self-focused rumination on negative thinking
and interpersonal problem solving. Journal of Personality and Social Psychology, 69, 176–190.
Marlatt, G. A. (2002). Mindfulness treatment of addictive behaviors. Paper presented at the annual
meeting for the Advancement of Behavior Therapy, Reno, NV.
Marlatt, G. A., & Gordon, J. R. (1985). Relapse prevention: Maintenance strategies in the treatment of
addictive behaviors.NewYork:GuilfordPress.
Miranda, J. (1992). Dysfunctional thinking is activated by stressful life event. Cognitive Therapy and
Research, 16, 473–483.
Miranda, J., Gross, J. G., Persons, J. B., & Hahn, J. (1998). Mood matters: Negative mood induction ac-
tivates dysfunctional attitudes in women vulnerable to depression. Cognitive Therapy and Research,
22, 363–376.
Miranda, J., & Persons, J. B. (1988). Dysfunctional attitudes are mood-state dependent. Journal of
Abnormal Psychology, 97, 76–79.
Miranda, J., Persons, J. B., & Byers, C. N. (1990). Endorsement of dysfunctional beliefs depends on
current m ood state. Journal of Abnormal Psychology, 99, 237–241.
Morrow, J., & Nolen-Hoeksema, S. (1990). Effects of responses to depression on the remediation of
depressive affect. Journal of Personality and Social Psychology, 58, 519–527.
Nolen-Hoeksema, S. (1987). Sex differences in unipolar depression: Evidence and theory. Psychological
Bulletin, 101, 259–282.
Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the duration of depressive
episodes. Journal of Abnormal Psychology, 100, 569–582.
Nolen-Hoeksema, S., & Morrow, J. A. (1991). A prospective study of depression and posttraumatic stress
symptoms after a natural disaster: The 1989 Loma Prieta earthquake. Journal of Personality & Social
Psychology, 61, 115–121.
Nolen-Hoeksema, S., Morrow, J., & Fredrickson, B. L. (1993). Response styles and the duration of
episodes of depressed m ood. Journal of Abnormal Psychology, 102, 20–28.
Paykel, E., Scott, J., Teasdale, J., Johnson, A. L., Garland, A., Moore, R., et al. (1999). Prevention of
relapse in residual depression by cognitive therapy. Archives of General Psychiatry, 56, 829–835.
Posner, M. I., & Snyder, C. R. R. (1975). Attentive and cognitive control. In R. L. Solso (Ed.),
Information processing and cognition: The Loyola symposium. Hillsider, NJ: Erlbaum.
Rabkin, J. G., & Klein, D. F. (1987). The clinical measurement of depressive disorders. In A. J. Marsella,
R. M. A. Hirschfeld, & M. M. Katz (Eds.), The measurement of depression (pp. 30–83). New York:
Guilford Press.
Robinson, M. S., & Alloy, L. B. (2003). Negative cognitive styles and stress-reactive rumination interact
to predict depression: A prospective study. Cognitive Therapy and Research, 27, 275–292.
Safren, S. A., Heimberg, R. G., Lerner, J., Henin, A., Warman, M., & Kendall, P. C. (2000). Differ-
entiating anxious and depressive self-statements: Combined factor structure of the Anxious Self-
Statements Questionnaire and the Automatic Thoughts-Questionnaire-Revised. Cognitive Therapy
and Research, 24, 327–344.
Santorelli, S. (1999). Heal thy self: Lessons on mindfulness in medicine.NewYork:BellTower.
Segal, Z. V., & Ingram, R. E. (1994). Mood priming and construct activation in tests of cognitive
vulnerability t o unipolar depression. Clinical Psychology Review, 14, 663–695.
Segal, Z. V., Williams, J. M., Teasdale, J. D., & Gemar, M. (1996). A cognitive science perspective on
kindling and episode sensitization in recurrent affective disorder. Psychological Medicine, 26, 371–
380.
Segal, Z. V., Williams, J. M. G., & Teasdale, J. D. (2002). Mindfulness based cognitive therapy for
depression: A new approach to preventing relapse. New York: Guildford Press.
Segerstrom, S. C., Tsao, J. C. I., Alden, L. E., & Craske, M. G. (2000). Worry and rumination: Repetitive
thought as a concomitant and predictor of negative mood. Cognitive Therapy and Research, 24,
671–688.
Spasojevi
´
c, J., & Alloy, L. B. (2001). Rumination as a common mechanism relating depressive risk factors
to depression. Emotion, 1, 25–37.
Speca, M., Carlson, L. E., Goodey, E., & Angen, M. (2000). A randomized, wait-list controlled clinical
trial: The effect of a mindfulness meditation-based stress reduction program on mood and symptoms
of stress in cancer outpatients. Psychosomatic Medicine, 2, 613–622.

Mindfulness, Affective Symptoms, and Cognitions 455
Spielberger, C. D., Gorush, R. L., Lushene, R. E., Vagg, P. R., & Jacobs, G. A. (1983). Manual for The
State–Trait Anxiety Inventory (Form Y). Palo Alto, CA: Consulting Psychologist Press.
Spitzer, R. L., Williams, J. B.W., Gibbon, M., & First, M. B. (1992). The structured clinical interview
for the DSM-III-R (SCID) I: History, rationale, and description. Archives of General Psychiatry, 49,
624–629.
Teasdale, J. D. (1988). Cognitive vulnerability to persistent depression. Cognition and Emotion, 2, 247–
274.
Teasdale, J. T., Segal, Z. V., & Williams, J. M. G. (1995). How does cognitive therapy prevent depressive
relapse and why should attentional control (mindfulness) training help? Behaviour Research and
Therapy, 33, 25–39.
Teasdale, J. T., Segal, Z. V., Williams, J. M. G., Ridgeway, V. A., Soulsby, J. M., & Lau, M. A. (2000).
Prevention of relapse/recurrence in major depression by mindfulness-based cognitive therapy.
Journal of Consulting and Clinical Psychology, 68, 615–623.
Treynor, W., Gonzalez, R., & Nolen-Hoeksema, S. (2003). Rumination reconsidered: A psychometric
analysis. Cognitive Therapy and Research, 27, 247–259.
Valentine, E. R., & Sweet, P. L. G. (1999). Meditation and attention: A comparison of the effects of
concentrative and mindfulness meditation on sustained attention. Mental Health, Religion & Culture,
2, 59–70.
Watkins, E., & Baracaia, S. (2002). Rumination and social problem-solving in depression. Behaviour
Research and Therapy, 40, 1179–1189.
Weissman, A. N. (1979). The Dysfunctional Attitude Scale: A validation study. Dissertation Abstracts
International, 40, 1389B–1390B. (University Microfilms No. 79-19, 533)
Weissman, A. N., & Beck, A. T. (1978). Development and validation of the Dysfunctional Attitude Scale:
A preliminary investigation. Paper presented at the meeting of the Association for the Advancement
of Behavior Therapy, Chicago.
Williams, J. M. G., Teasdale, J. D., Segal, Z. V., & Soulsby, J. (2000). Mindfulness-based cognitive therapy
reduces overgeneral autobiographical memory in formerly depressed patients. Journal of Abnormal
Psychology, 109, 150–155.
Yin, P., & Fan, X. (2000). Assessing the reliability of Beck Depression Inventory scores: Reliability
generalization across studies. Educational and Psychological Measurement, 60, 201–223.
Zachary, R. (1986). Shipley Institute of Living Scale, Revised Manual (4th ed.). Los Angeles: Western
Psychological Services.
Zuroff, D. C., Blatt, S. J., Sanislow, C. A., III, Bondi, C. M., & Pilkonis, P. A. (1999). Vulnerability to
depression: Reexamining state dependence and relative stability. Journal of Abnormal Psychology,
108, 76–89.