Professional Documents
Culture Documents
Interventions based on training in mindfulness skills are (Goleman & Schwartz, 1976; Kabat-Zinn, 1982). The cur-
becoming increasingly popular. Mindfulness involves in- rent mindfulness literature describes numerous meditation
tentionally bringing ones attention to the internal and ex- exercises designed to develop mindfulness skills (Hanh,
ternal experiences occurring in the present moment, and 1976; Kabat-Zinn, 1990, 1994; Linehan, 1993b). Many
is often taught through a variety of meditation exercises. encourage individuals to attend to the internal experiences
This review summarizes conceptual approaches to mind-
occurring in each moment, such as bodily sensations,
thoughts, and emotions. Others encourage attention to
fulness and empirical research on the utility of mindfulness-
aspects of the environment, such as sights and sounds
based interventions. Meta-analytic techniques were in-
(Kabat-Zinn, 1994; Linehan, 1993b). All suggest that
corporated to facilitate quantification of findings and
mindfulness should be practiced with an attitude of non-
comparison across studies. Although the current empir-
judgmental acceptance. That is, phenomena that enter the
ical literature includes many methodological flaws, find- individuals awareness during mindfulness practice, such as
ings suggest that mindfulness-based interventions may perceptions, cognitions, emotions, or sensations, are ob-
be helpful in the treatment of several disorders. Method- served carefully but are not evaluated as good or bad, true
ologically sound investigations are recommended in order or false, healthy or sick, or important or trivial (Marlatt &
to clarify the utility of these interventions. Kristeller, 1999). Thus, mindfulness is the nonjudgmental
Key words: mindfulness, meditation, meta-analysis, observation of the ongoing stream of internal and external
treatment outcome. [Clin Psychol Sci Prac 10: 125143, stimuli as they arise.
2003] Until recently, mindfulness has been a relatively unfa-
miliar concept in much of our culture (Kabat-Zinn, 1982),
Mindfulness is a way of paying attention that originated in perhaps because of its origins in Buddhism. Kabat-Zinn
Eastern meditation practices. It has been described as (2000) suggests that mindfulness practice may be benecial
bringing ones complete attention to the present experi- to many people in Western society who might be un-
ence on a moment-to-moment basis (Marlatt & Kris- willing to adopt Buddhist traditions or vocabulary. Thus,
teller, 1999, p. 68) and as paying attention in a particular Western researchers and clinicians who have introduced
way: on purpose, in the present moment, and nonjudg- mindfulness practice into mental health treatment pro-
mentally (Kabat-Zinn, 1994, p. 4). The ability to direct grams usually teach these skills independently of the reli-
ones attention in this way can be developed through the gious and cultural traditions of their origins (Kabat-Zinn,
practice of meditation, which is dened as the intentional 1982; Linehan, 1993b). In the current empirical literature,
self-regulation of attention from moment to moment clinical interventions based on training in mindfulness skills
are described with increasing frequency, and their popu-
larity appears to be growing rapidly. According to Salmon,
Send correspondence to Ruth A. Baer, Department of Psychol- Santorelli, and Kabat-Zinn (1998), over 240 hospitals and
ogy, 115 Kastle Hall, University of Kentucky, Lexington, KY clinics in the United States and abroad were oering stress
40506-0044. E-mail: rbaer@uky.edu. reduction programs based on mindfulness training as of
DOI: 10.1093/clipsy/bpg015
2003 AMERICAN PSYCHOLOGICAL ASSOCIATION D12 125
1997. Mindfulness training is also a central component of described in this review often are directed toward the in-
dialectical behavior therapy (Linehan, 1993a, 1993b), an ner experiences of the individual (e.g., thoughts, emo-
increasingly popular approach to the treatment of border- tions) and emphasize a less goal-directed, nonjudgmental
line personality disorder. The empirical literature on the observation. Langer (1989) has cautioned against drawing
eects of mindfulness training contains many method- unwarranted parallels between the two forms of mindful-
ological weaknesses, but it suggests that mindfulness inter- ness, noting that they are derived from very dierent his-
ventions may lead to reductions in a variety of problematic torical and cultural backgrounds.
conditions, including pain, stress, anxiety, depressive re-
lapse, and disordered eating (e.g., Kabat-Zinn, 1982;Kabat- INTERVENTIONS BASED ON MINDFULNESS TRAINING
Zinn et al., 1992; Kristeller & Hallett, 1999; Shapiro, Mindfulness-Based Stress Reduction
Schwartz, & Bonner, 1998; Teasdale et al., 2000). The most frequently cited method of mindfulness training
This review summarizes the recent literature on mind- is the mindfulness-based stress reduction (MBSR) pro-
fulness training as a clinical intervention. First, current gram, formerly known as the stress reduction and relax-
methods for teaching mindfulness skills are described. ation program (SR-RP; Kabat-Zinn, 1982, 1990). It was
Next, conceptual approaches that articulate how mindful- developed in a behavioral medicine setting for populations
ness skills may be helpful in treating clinical conditions are with a wide range of chronic pain and stress-related disor-
summarized. Finally, the empirical literature on the eects ders. The program is conducted as an 8- to 10-week course
of mindfulness training is reviewed. for groups of up to 30 participants who meet weekly for
This review does not address transcendental meditation 22.5 hr for instruction and practice in mindfulness med-
(TM) and other concentration-based approaches, which itation skills, together with discussion of stress, coping, and
have been reviewed elsewhere (Delmonte, 1985; Smith, homework assignments. An all-day (78-hr) intensive
1975). Concentration-based approaches train participants mindfulness session usually is held around the sixth week.
to restrict the focus of attention to a single stimulus, such Several mindfulness meditation skills are taught. For ex-
as a word (e.g., a mantra), sound, object, or sensation. ample, the body scan is a 45-min exercise in which atten-
When attention wanders, it is redirected to the object of tion is directed sequentially to numerous areas of the body
meditation. No attention is paid to the nature of the dis- while the participant is lying down with eyes closed. Sen-
traction. Mindfulness meditation, in contrast, involves ob- sations in each area are carefully observed. In sitting med-
servation of constantly changing internal and external itation, participants are instructed to sit in a relaxed and
stimuli as they arise. wakeful posture with eyes closed and to direct attention to
This review also does not address Langers (1989, 1997) the sensations of breathing. Hatha yoga postures are used
cognitive model of mindfulness, which includes alertness to teach mindfulness of bodily sensations during gentle
to distinctions, context, and multiple perspectives, open- movements and stretching. Participants also practice mind-
ness to novelty, and orientation in the present (Sternberg, fulness during ordinary activities like walking, standing,
2000). Mindfulness interventions studied by Langer and and eating.
colleagues (e.g., Langer, 1989; Langer & Moldoveanu, Participants in MBSR are instructed to practice these
2000) often include teaching participants to consider in- skills outside group meetings for at least 45 min per day, six
formation or situations from multiple perspectives or days per week. Audiotapes are used early in treatment, but
within new contexts in order to increase learning or cre- participants are encouraged to practice without tapes after
ativity. Although this concept of mindfulness shares with a few weeks. For all mindfulness exercises, participants are
meditative approaches an emphasis on exible awareness in instructed to focus attention on the target of observation
the present, several important dierences can be noted. (e.g., breathing or walking) and to be aware of it in each
Langers (1989) mindfulness interventions usually involve moment. When emotions, sensations, or cognitions arise,
working with material external to the participants, such as they are observed nonjudgmentally. When the participant
information to be learned or manipulated, and often in- notices that the mind has wandered into thoughts, mem-
clude active, goal-oriented cognitive tasks, such as solving ories, or fantasies, the nature or content of them is briey
problems. In contrast, the meditation-based approaches noted, if possible, and then attention is returned to the
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V10 N2, SUMMER 2003 126
present moment. Thus, participants are instructed to no- view, which postulates that reality consists of opposing
tice their thoughts and feelings but not to become absorbed forces. The synthesis of these forces leads to a new reality,
in their content (Kabat-Zinn, 1982). Even judgmental which in turn consists of opposing forces, in a continual
thoughts (e.g., this is a foolish waste of time) are to be process of change. In DBT, the most central dialectic is the
observed nonjudgmentally. Upon noticing such a thought, relationship between acceptance and change. Clients are
the participant might label it as a judgmental thought, or encouraged to accept themselves, their histories, and their
simply as thinking, and then return attention to the pres- current situations exactly as they are, while working in-
ent moment. An important consequence of mindfulness tensively to change their behaviors and environments in
practice is the realization that most sensations, thoughts, order to build a better life. The synthesis of this apparent
and emotions uctuate, or are transient, passing by like contradiction is a central goal of DBT.
waves in the sea (Linehan, 1993b, p. 87). DBT includes a wide range of cognitive and behavioral
treatment procedures, most of which are designed to
Mindfulness-Based Cognitive Therapy change thoughts, emotions, or behaviors. Mindfulness
Teasdale, Segal, and Williams (1995) proposed that the skills are taught in DBT within the context of synthesizing
skills of attentional control taught in mindfulness medita- acceptance and change. Although the skills taught are sim-
tion could be helpful in preventing relapse of major de- ilar to those targeted in MBSR, including nonjudgmental
pressive episodes. Their information-processing theory of observation of thoughts, emotions, sensations, and envi-
depressive relapse suggests that individuals who have ex- ronmental stimuli, the concepts are organized somewhat
perienced major depressive episodes are vulnerable to re- dierently. For example, Linehan (1993a, 1993b) describes
currences whenever mild dysphoric states are encountered, three mindfulness what skills (observe, describe, par-
because these states may reactivate the depressive thinking ticipate) and three mindfulness how skills (nonjudg-
patterns present during the previous episode, or episodes, mentally, one-mindfully, eectively). DBT clients learn
thus precipitating a new episode. Mindfulness-based cog- mindfulness skills in a year-long weekly skills group, which
nitive therapy (MBCT) is a manualized (Segal, Williams, & also covers interpersonal eectiveness, emotion regulation,
Teasdale, 2002) 8-week group intervention based largely on and distress tolerance skills. Clients work with their indi-
Kabat-Zinns (1990) MBSR program. It incorporates ele- vidual therapists on applying skills learned in group to their
ments of cognitive therapy that facilitate a detached or de- daily lives.
centered view of ones thoughts, including statements such Linehan (1994) notes that some severely impaired in-
as thoughts are not facts and I am not my thoughts. dividuals may be unable or unwilling to meditate as
This decentered approach also is applied to emotions and extensively as Kabat-Zinns (1990) MBSR program rec-
bodily sensations. MBCT is designed to prevent depressive ommends. Thus, DBT does not prescribe a specic fre-
relapse by teaching formerly depressed individuals to ob- quency or duration of mindfulness practice. Instead, goals
serve their thoughts and feelings nonjudgmentally, and to for mindfulness practice are established by individual
view them simply as mental events that come and go, clients and their therapists. DBT oers numerous mind-
rather than as aspects of themselves, or as necessarily accu- fulness exercises from which clients may choose (some
rate reections of reality. This attitude toward depression- adapted from Hanh, 1976). In one example, clients imag-
related cognitions is believed to prevent the escalation of ine that the mind is a conveyor belt. Thoughts, feelings,
negative thoughts into ruminative patterns (Teasdale et al., and sensations that come down the belt are observed, la-
1995). beled, and categorized. In another exercise, clients imag-
ine that the mind is the sky, and that thoughts, feelings, and
INTERVENTIONS INCORPORATING sensations are clouds that they watch passing by. Several
MINDFULNESS TRAINING variations on observing the breath are taught, including
Dialectical Behavior Therapy following the breath in and out, counting breaths, coordi-
Dialectical behavior therapy (DBT) is a multifaceted ap- nating breathing with footsteps while walking, and fol-
proach to the treatment of borderline personality disorder lowing the breath while listening to music. Some exercises
(Linehan, 1993a, 1993b). It is based on a dialectical world- encourage mindful awareness during everyday activities,
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V10 N2, SUMMER 2003 128
duce symptoms of panic through exercises such as hyper- decentered view of ones thoughts encouraged by mind-
ventilation and aerobic activity, and to practice tolerating fulness training may interfere with ruminative patterns
these sensations until they subside. In contrast, however, believed to be characteristic of depressive episodes (Nolen-
mindfulness training does not include the deliberate in- Hoeksema, 1991). That is, mindfulness training may en-
duction of panic symptoms. Instead, participants are in- able formerly depressed individuals to notice depressogenic
structed to observe these sensations nonjudgmentally when thoughts and to redirect attention to other aspects of the
they naturally arise. present moment, such as breathing, walking, or environ-
Linehan (1993a, 1993b) describes individuals with bor- mental sounds, thus avoiding rumination. Teasdale has
derline personality disorder as emotion phobic. That is, described this perspective on ones thoughts as meta-
they are often afraid of experiencing strong negative aec- cognitive insight. Teasdale et al. (1995) also note that a
tive states. This fear is understandable, because their neg- practical advantage of mindfulness skills in encouraging
ative aective states typically are very intense. However, cognitive change is that they can be practiced at any time,
their attempts to avoid these states often have maladaptive including during periods of remission, when depresso-
consequences. Linehan (1993a, 1993b) suggests that pro- genic thinking may be occurring too rarely to permit reg-
longed observation of current thoughts and emotions, ular practice of traditional cognitive therapy exercises, such
without trying to avoid or escape them, can be seen as an as identication and disputing of cognitive distortions.
example of exposure, which should encourage the extinc- That is, a mindful perspective about ones thoughts can be
tion of fear responses and avoidance behaviors previously applied to all thoughts.
elicited by these stimuli. Thus, the practice of mindfulness
skills may improve patients ability to tolerate negative Self-Management
emotional states and ability to cope with them eectively. Several authors have noted that improved self-observation
resulting from mindfulness training may promote use of a
Cognitive Change range of coping skills. For example, Kabat-Zinn (1982)
Several authors have noted that the practice of mindfulness suggests that increased awareness of pain sensations and
may lead to changes in thought patterns, or in attitudes stress responses as they occur may enable individuals to en-
about ones thoughts. For example, Kabat-Zinn (1982, gage in a variety of coping responses, including skills not
1990) suggests that nonjudgmental observation of pain and included in their treatment program. Kristeller and Hallett
anxiety-related thoughts may lead to the understanding (1999) suggest that the self-observation skills developed
that they are just thoughts, rather than reections of truth through mindfulness training might lead to improved
or reality, and do not necessitate escape or avoidance recognition of satiety cues in binge eaters, as well as in-
behavior. Similarly, Linehan (1993a, 1993b) notes that ob- creased ability to observe urges to binge without yielding
serving ones thoughts and feelings and applying descrip- to them. Marlatt (1994) suggests a similar eect in patients
tive labels to them encourages the understanding that they recovering from addictions. Teasdale et al. (1995) note that
are not always accurate reections of reality. For example, mindfulness training encourages awareness of all cognitive
feeling afraid does not necessarily mean that danger is im- and emotional events as they occur, including those that
minent, and thinking I am a failure does not make it may be early signs of potential depressive relapse. Thus,
true. Kristeller and Hallett (1999), in a study of MBSR in mindfulness training may promote recognition of early
patients with binge eating disorder, cite Heatherton and signs of a problem, at a time when application of previously
Baumeisters (1991) theory of binge eating as an escape learned skills will be most likely to be eective in prevent-
from self-awareness and suggest that mindfulness training ing the problem. Linehan (1993b) suggests that nonjudg-
might develop nonjudgmental acceptance of the aversive mental observation and description permits recognition
cognitions that binge-eaters are thought to be avoiding, of the consequences of behaviors (e.g., irritating ones boss
such as unfavorable comparisons of self to others and per- with frequent lateness) in place of global judgments about
ceived inability to meet others demands. the self (e.g., I am a bad employee). This recognition
Teasdale (1999) and Teasdale et al. (1995), in their may lead to more eective behavior change, including re-
discussion of MBCT, suggest that the nonjudgmental, duction of impulsive, maladaptive behaviors. Linehan
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V10 N2, SUMMER 2003 130
practice (Segal et al., 2002). Professionals using more tra- sure, Mc = the mean of the comparison group on that mea-
ditional cognitive-behavioral strategies generally are not sure, and SDp = the pooled standard deviation of the two
expected to engage in regular practice of the skills they are groups. If means or standard deviations were not provided,
teaching. eect sizes were calculated from the signicance level (p).
Although the practice of mindfulness generally involves For studies using within-groups designs, eect sizes were
acceptance of current reality, rather than systematic at- calculated from t or F (1 df ), or from the signicance level
tempts to change reality, individuals who practice these when t or F were not reported. Calculations of eect sizes
skills may experience reductions in a variety of symptoms. relied on methods described by Rosenthal (1984).
The empirical literature addressing this issue is reviewed
next. Overview of Treatment Literature
Studies examining the eects of mindfulness-based inter-
EMPIRICAL RESEARCH ON MINDFULNESS-BASED ventions are summarized in Table 1. The studies are
INTERVENTIONS grouped by participant population, beginning with stud-
The empirical literature investigating the eects of mind- ies of chronic pain patients. Next are studies of patients
fulness-based interventions is reviewed here. Meta-analytic with other Axis I disorders (anxiety, eating, and major de-
procedures were incorporated to facilitate quantication of pressive disorders), followed by studies of patients with
ndings and comparisons across studies. To locate relevant other medical problems (bromyalgia, psoriasis, and can-
studies, a computer search (using PsycInfo and Medline cer). Next are studies with mixed populations, including
databases) was conducted of articles and chapters including psychotherapy and medical patients. The last group in-
the terms mindfulness or meditation. Reference lists of all cludes studies of nonclinical populations (students and
articles were searched for additional articles. Studies were other volunteers). Within each group studies are listed in
included if they were published in English and compared order of publication date.
a group of participants trained in mindfulness with a group Sample sizes in these studies have ranged from 16 to
not trained, or a group who provided data before and af- 142. Mean age of participants has ranged from 3850 years,
ter mindfulness training. Unpublished dissertations and with a mean of 45 years. Gender ratio of the samples has
conference papers were excluded, as were studies address- ranged from 0 to 46% male. Education and race/ethnicity
ing nonmindfulness forms of meditation, such as transcen- were rarely reported.
dental meditation (TM). Studies of mindfulness as dened Nine studies used pre-post designs with no control
by Langer (1989, 1997) also were excluded, for reasons group. Nine used between-groups designs with Treatment
described earlier. Twenty-one studies meeting these cri- As Usual (TAU) or waiting-list control groups. Most stud-
teria were found. ies used the 810 week MBSR group intervention (Kabat-
For each study, several demographic and methodolog- Zinn,1982, 1990), or a variation of this intervention
ical variables were coded, including number, type, and tailored to the population under study. Two studies (Teas-
characteristics of participants, research design, the nature of dale et al., 2000; Williams, Teasdale, Segal, & Soulsby,
the mindfulness intervention, type of comparison group, 2000) examined MBCT. Dependent variables have in-
whether participants were randomly assigned to interven- cluded a variety of self-report measures of pain, other med-
tion or comparison groups, the dependent variables re- ical symptoms, anxiety, depression, eating behaviors, and
ported, and follow-up intervals and data. general psychological functioning, as well as objective
Eect sizes (Cohens d) were calculated for all studies measures such as analysis of urine chemistry.
that provided sucient data. Cohens d expresses eect size For each study eect sizes were calculated separately
in standard deviation units; thus, an eect size of 1.0 on a for each dependent measure completed at the conclusion
given dependent measure indicates that the treatment of treatment and at all reported follow-up intervals. Post-
group scored one standard deviation better, on average, treatment eect sizes then were averaged across dependent
than the comparison group on that measure. For studies measures within studies, yielding a single posttreatment
using between-groups designs, eect sizes were calculated eect size for each study. Similarly, eect sizes for all de-
with the following formula: d = (Mt Mc)/SDp, in which pendent measures completed at all follow-up intervals were
Mt = the mean of the treatment group on a specic mea- averaged, yielding a single follow-up eect size for each
Kabat-Zinn 51 Chronic 46 35 Pre-post MBSR None No 12 Pain Rtg Index 0.70 0.35
(1982) pain patients Body Parts PA (27)
Interfere rtg
MSCL
POMS total
SCL-90-R GSI
Kabat-Zinn 90 Chronic 44 33 Pre-post MBSR None No 15 Same as above 0.36 0.51
et al. (1985) pain patientsa (215)
Part 1
Kabat-Zinn 42 Chronic 48 24 Between group MBSR TAU No Same as above 0.26
et al. (1985) pain patientsb (n = 21) (n = 21)
Part 2
post-MBSR MSCL
SCL-90-R GSI
Randolph 78 Chronic 50 31 Pre-post MBSR None No Pain rtgs 0.15 0.08
et al. (1999) pain patients BSI-GSI (212)
POMS total
Pain beliefs
Kabat-Zinn 22 Anxiety 38 23 Pre-post MBSR None No 8 Hamilton Anx 0.88 1.35
et al. (1992) patients Hanilton Dep (3)
BDI; BAI; FSS
Mobility Inv
Miller et al. 18 Anxiety 3-year follow-up MBSR None No Hamilton Anx 1.10
(1995) patients (patients from Hamilton Dep (36)
132
Goldenberg 121 Fibro-myalgia 46 7 Between group MBSR Wait list No 9 Vass pain, sleep, etc. 0.42
et al. (1994) patients (n = 79) (n = 18); Fibro Impact Q
declined SCL-90-R GSI
MBSR
(n = 24)
Kabat-Zinn 37 Psoriasis 43 46 Between group Mindfulness TAU Yes Days to clearing 0.63
et al. (1998) patients tapes during of psoriasis
receiving light tx
MINDFULNESS TRAINING
light therapy
Speca et al. 90 Cancer 51 19 Between group Variant of Wait list Yes 13 POMS 0.60
(2000) patients MBSR SOSI
Carlson et al. 54 Cancerf 51 19 6-month follow- Variant of None No POMS NS
(2000) patients up in Speca et al. MBSR SOSI (6)
BAER
(2000)
Kutz et al. 20 Long-term 38 Pre-post MBSR None No SCL-90-R 0.72 0.52
(1985) dynamic POMS (6)
therapy clients Interfere rtg
Thera rtgs
Roth & Creasor 86 Outpatients 43 16 Pre-post MBSR None NO 40 SCL-90-R 0.69
(1997) inner city BAI
Latino/Latina MSCL
Mixed dx Self-esteem
Reibel et al. 121 Medical 47 29 Pre-post MBSR None No 11 SF-36 0.56 NS
(2001) patients, MSCL (12)
Mixed dx SCL-90-R GSI
Massion et al. 16 Healthy 42 0 Between group Regular Non- No Melatonin metabolite 0.98
(1995) women mediators mediators in urine
Astin (1997) 19 College 5 Between group MBSR Wait list Yes 14 SCL-90-R GSI 1.51
students (n = 12) (n = 7) Control Inv
INSPIRIT
Shapiro et al. 73 Premed and 44 Between group MBSR Wait list Yes 3 SCL-90-R GSI 0.50
(1998) med students SCL-90-R Dep
STAI
Empathy rtg
INSPIRIT
Williams 75 Community 43 28 Between group MBSR Wait list Yes 15 DSI 0.67 0.67
et al. (2001) volunteers and info SCL-90-R GSI (3)
MSCL
Notes. rand assn = random assignment; att = attrition from treatment group; post = posttreatment; foll = follow-up; MDD = major depressive disorder; medicn = medication; tx = treatment; dx = diagnosis;
MBSR = Mindfulness-Based Stress Reduction; MBCT = Mindfulness-Based Cognitive Therapy; TAU = Treatment As Usual; rtg = rating; PA = problem assessment; MSCL = Medical Symptom Checklist;
POMS = Profile of Mood States; SCL-90-R = Symptom Checklist-90 Revised; GSI = General Severity Index; Anx = anxiety; Dep = depression; BDI = Beck Depression Inventory; BAI = Beck Anxiety Inven-
tory; FSS = Fear Survey Schedule; Inv = inventory; Autobiog = autobiographical; VAS = visual analog scale; Strat = strategies; Q = questionnaire; fibro = fibromyalgia; SOSI = Symptoms of Stress Inventory;
SF-36 = Short Form 36; INSPIRIT = Index of Core Spiritual Experiences; STAI = State-Trait Anxiety Inventory; DSI = Daily Stress Inventory; NS = no significant difference between posttest and follow-up.
a
Includes 51 patients from Kabat-Zinn (1982).
b
Includes 21 patients from Part 1 of this study.
c
Includes patients from Kabat-Zinn (1982) and Kabat-Zinn et al. (1985).
d
Subset of patients in Teasdale et al. (2000).
e
Insufficient data to calculate.
f
Subset of patients in Speca et al. (2000).
133
study that reported follow-up data. Mean posttreatment Axis I Disorders. Kabat-Zinn et al. (1992) examined a
and follow-up eect sizes are presented in the nal two sample of 22 patients with generalized anxiety and panic
columns of Table 1. disorders, and found signicant improvements in several
No studies of DBT, ACT, or RP were included, be- measures of anxiety and depression, both at posttreatment
cause none were found that examined the mindfulness and at 3-month follow-up. A no-treatment control group
component independently of the behavior change strate- was not included. Miller, Fletcher, and Kabat-Zinn (1995)
gies also included in these treatment approaches. Thus, reported a 3-year follow-up of the same participants and
although empirical studies support the ecacy of these found that treatment gains had been maintained.
treatments (Curry, Marlatt, Gordon, & Baer, 1988; Ito, Kristeller and Hallett (1999) examined the eects of
Donovan, & Hall, 1988; Koons et al., 2001; Linehan, Arm- MBSR on binge eating disorder. In a pre-post design with
strong, Suarez, Allmon, & Heard, 1991; Linehan, Heard, & no control group, 18 female patients showed statistically
Armstrong, 1993; Linehan, Tutek, Heard, & Armstrong, signicant improvements in several measures of eating and
1994; Strosahl, Hayes, Bergan, & Romano, 1998; Zettle & mood.
Raines, 1989), the relative contribution of mindfulness Teasdale et al. (2000) examined the eects of MBCT on
training to these treatment eects has not been investi- rates of depressive relapse in a large sample of patients
gated. In contrast, studies of MBCT were included, be- whose major depressive disorder (MDD) had remitted af-
cause mindfulness training appears to be the central focus ter treatment with medication. All participants had dis-
of this approach, although some cognitive techniques have continued their medications at least 12 weeks before the
been incorporated. study began. Patients were randomly assigned to either
MBCT (8-week manualized group treatment) or TAU and
General Findings then followed for 1 year. For patients with three or more
Chronic Pain. Four studies have examined the eects of previous depressive episodes, results showed much lower
MBSR on patients with chronic pain. The rst study relapse rates for MBCT patients (37% of patients relapsed)
(Kabat-Zinn, 1982) describes pre-post data for 51 patients. than for the TAU group (66% of patients relapsed) during
The second study (Kabat-Zinn, Lipworth, & Burney, the 1-year follow-up period. However, relapse rates for
1985), has two parts. Part 1 presents pre-post data for a the MBCT and TAU groups did not dier for patients with
sample of 90 patients, including the 51 patients from only one or two previous episodes.
Kabat-Zinn (1982). In Part 2, 21 of these 90 patients are Using a subset of the participants from Teasdale et al.
compared to 21 other pain patients who had received TAU (2000), J. M. G. Williams et al. (2000) found that those
in the pain clinic but had not participated in MBSR. Parts who had completed MBCT produced fewer general and
1 and 2 of this study are entered separately in Table 1. The more specic memories when asked to recall events from
third study (Kabat-Zinn, Lipworth, Burney, & Sellers, their pasts in response to cue words. The authors speculate
1987) is an extensive series of follow-up evaluations of that mindfulness training may modify the overgeneral
chronic pain patients who had completed MBSR over the autobiographical memory believed to be characteristic of
preceding several years, including patients in the previous individuals with depression (Kuyken & Brewin, 1995).
two studies. Thus, the rst four entries in Table 1 are de-
rived from three published articles with overlapping par- Other Medical Disorders. Two studies have investigated
ticipant samples. Finally, Randolph, Caldera, Tacone, and eects of MBSR on bromyalgia. Both reported improve-
Greak (1999) investigated the eects of MBSR in an in- ments in a variety of symptoms. In a study of psoriasis pa-
dependent sample of 78 chronic pain patients. tients, Kabat-Zinn et al. (1998) found that patients who
In general, ndings for chronic pain patients show sta- listened to mindfulness audiotapes during individual light-
tistically signicant improvements in ratings of pain, other therapy sessions showed quicker clearing of their skin
medical symptoms, and general psychological symptoms. (Mdn = 65 days) than did patients who received light ther-
Many of these changes were maintained at follow-up eval- apy alone (Mdn = 97 days). Speca, Carlson, Goodey, and
uations. Most of these comparisons used pre-post designs Angen (2000) examined the eects of MBSR in a group
with no control group. of cancer patients and reported signicant reductions in
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V10 N2, SUMMER 2003 134
mood disturbance and stress levels. Carlson, Ursuliak, eect sizes in this calculation, the eect sizes obtained from
Goodey, Angen, and Speca (2001) reported that these Kabat-Zinn (1982) and Parts 1 and 2 of Kabat-Zinn et al.
changes were maintained at 6-month follow-up. (1985) rst were averaged, because these comparisons have
overlapping participant samples. Similarly, the mean eect
Mixed Clinical Populations. Kutz et al. (1985) studied a sizes obtained for Teasdale et al. (2000) and J. M. G.
sample of long-term psychodynamic therapy patients with Williams et al. (2000) were averaged, because these two
diagnoses including anxiety and obsessive neuroses, and studies also have overlapping participant samples. After
narcissistic and borderline personality disorders. They these preliminary calculations, 15 independent posttreat-
completed a 10-week MBSR program while continuing ment mean eect sizes, each from a separate sample, were
with their individual psychotherapy and showed statisti- available for analysis. Their mean was 0.74 (SD = 0.39).
cally signicant improvements in a variety of self- and When each of these 15 eect sizes was weighted by sample
therapist-rated symptoms. Roth and Creasor (1997) stud- size, overall mean eect size was 0.59.
ied outpatients from a low-income, primarily Latino pop-
ulation attending an inner city health clinic and showed Mean Effect Size at Follow-Up
statistically signicant improvements on several measures of Follow-up data were reported less often. Eect sizes at
medical and psychological functioning. Reibel, Greeson, follow-up ranged from 0.08 to 1.35. Before an overall
Brainard, and Rosenzweig (2001) studied medical patients mean of these eect sizes was calculated, mean eect sizes
with a variety of medical and psychiatric diagnoses and obtained from studies with overlapping participant samples
found signicant improvements in medical and psycho- were averaged. The overall mean of these independent
logical symptoms. None of these studies used control follow-up eect sizes was 0.59 (SD = 0.41).
groups. Cohen (1977) has described eect sizes of d = 0.2, d =
0.5, and d = 0.8 as small, medium, and large, respectively.
Nonclinical Populations. Massion, Teas, Hebert, Wert- Thus, on the average, the literature reviewed here suggests
heimer, and Kabat-Zinn (1995) analyzed urine levels of a that mindfulness-based interventions have yielded at least
melatonin metabolite in two groups of women. Levels medium-sized eects, with some eect sizes falling within
were signicantly higher in women previously trained in the large range. Many of the eect sizes calculated for these
MBSR who continued to meditate regularly than in wo- studies are probably conservative, because several studies
men who had never been trained and did not meditate. did not present means, standard deviations, or t values,
The authors cite previous ndings suggesting that mela- making it necessary to calculate d from the p value. In many
tonin level may be related to immune function (Bartsch et cases exact p values were not reported. Instead, for ex-
al., 1992;Guerrero & Reiter, 1992), and suggest that mind- ample, a p value between .01 and .05 might have been re-
fulness meditation may inuence health status through its ported as .05, which was then used to compute d. Larger
eects on melatonin. Astin (1997) and Shapiro et al. (1998) ps yield smaller ds. In addition, when ndings were re-
studied student populations who completed group MBSR, ported only as nonsignicant, eect sizes of zero were
reporting signicant eects on psychological symptoms, recorded. If means, SDs, or t values had been reported in
empathy ratings, and spiritual experiences. Both of these these cases, the calculated eect size might have been larger
studies used waiting-list control groups. Williams, Kolar, than zero.
Reger, and Pearson (2001) studied community volunteers
who completed MBSR to reduce their stress levels, re- Relationships Between Mean Effect Size at Posttreatment and
porting signicant improvements in medical and psycho- Study Characteristics
logical symptoms. Relationships between mean eect sizes at posttreatment
and selected methodological variables can be seen in Table
Mean Effect Size at Posttreatment 2. (Follow-up eect sizes are not included in this table.)
Posttreatment eect sizes ranged from 0.15 to 1.65. An The small number of studies available and the noninde-
overall mean of these eect sizes, collapsed across studies, pendence of some of the eect sizes make statistical analy-
was calculated. In order to include only independent mean ses of these dierences impractical. Thus, these ndings
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V10 N2, SUMMER 2003 136
ies, some of which used uncontrolled pre-post designs. scores (r = .66) and in BDI scores (r = .59). Astin (1997),
However, they suggest that mindfulness training, on aver- in a sample of college students, reported that participants
age, may bring participants with mild to moderate psy- practiced meditation for an average of 30 min per day, 3.5
chological distress into or close to the normal range. days per week. Reported practice time and improvement
on the GSI of the SCL-90-R were not signicantly corre-
Attrition, Adherence, and Maintenance of Mindfulness Practice lated. Reibel et al. (2001) reported that 90% of their mixed
Thirteen studies reported both the number of individuals sample of medical patients practiced three times per week
who agreed to participate in mindfulness training and the or more and 57% practiced nearly every day, most for 15
number who completed it. Program completion usually 30 min each time.
was dened as attendance at a minimum number of ses- Four studies reported the extent to which participants
sions, or was undened. Percentage of enrolled partici- trained in mindfulness skills continued to practice these
pants who completed treatment ranged from 60 to 97, skills after treatment had ended. In a series of follow-up
with a mean of 85% (SD = 8.91). The lowest completion studies of former MBSR patients, Kabat-Zinn et al. (1987)
rate (60%) was noted by Roth & Creasor (1997), who stud- noted that 75% of former patients reported that they still
ied an inner city health clinic population. The highest com- practiced meditation (averaged across follow-up intervals
pletion rate (97%) was reported by Shapiro et al. (1998), of 648 months). Of these patients, 43% meditated regu-
whose participants were premedical and medical students. larly ( three times weekly, 15 min each time), whereas
The most extensive analysis of program completion was 19% meditated sporadically (one or two times weekly, 15
provided by Kabat-Zinn and Chapman-Waldrop (1988), min each time, or three times weekly, 15 min each
who reported completion rates for the MBSR program (at time), and 38% were classied as marginal meditators
that time known as the Stress Reduction and Relaxation (< one time weekly for any length of time, or < three
Program) at the University of Massachusetts Medical Cen- times weekly, < 15 min each time). Practice of yoga two
ter. (This study is not included in Table 1, because it did or more times per week was reported by 31% of respon-
not examine treatment eects.) During the 2-year period dents, and 49% reported using awareness of breathing in
examined (19821984), 1,155 patients were referred to daily life often.
the program, mostly by their physicians. Of these patients, Kabat-Zinn et al. (1992), at 3-month follow-up of 22
75% completed an intake interview, and 90% of those patients with anxiety disorders, found that 84% reported
interviewed enrolled in the program. Of the 784 patients practicing meditation or yoga three or more times per
who enrolled, 76% completed the program, whereas 15% week, for 1545 min each time. Mindfulness of breathing
dropped out after beginning and 9% never attended a ses- in daily life was practiced by 95% (77% often and 18% some-
sion. Regression analyses showed that patients with stress- times). Miller et al (1995) contacted 18 of these patients for
related problems (hypertension, anxiety, sleep disorders, a 3-year follow-up evaluation and reported that 10 (56%)
etc.) were signicantly more likely to complete the pro- still practiced meditation: 4 regularly, 3 sporadically, and 3
gram than those with chronic pain complaints (lower back, marginally (as dened above). Sixteen of 18 (89%) reported
headache, etc.). Completers also had somewhat higher that they used awareness of breathing in daily life (4 often,
pretreatment scores than noncompleters on the GSI and 11 sometimes, and 1 rarely).
the Obsessive-Compulsive (OC) scales of the SCL-90-R. K. A. Williams et al. (2001), in a sample of community
Within the chronic pain group, women were slightly more volunteers self-identied as stressed out, reported that at
likely than men to complete the program. 3-month follow-up 81% of MBSR participants were prac-
Only three studies reported the extent to which partic- ticing either meditation, yoga, or awareness of breathing in
ipants completed their assigned home practice during the daily life.
course of the mindfulness intervention. Kristeller and Hal-
lett (1999), in a sample of women with binge eating disor- Patients Reactions to Treatment
der, noted that participants reported engaging in a mean of In their follow-up study of former MBSR patients, Kabat-
15.82 hr of meditation (SD = 3.15) across the 6-week in- Zinn et al (1987) found that the majority of those who
tervention program. Reported practice time was signi- considered themselves improved since completing MBSR
cantly correlated with improvements in Binge Eating Scale attributed 50100% of their improvement to the MSBR
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V10 N2, SUMMER 2003 138
not well dened. Many of the studies reviewed here were whether observed eects are due to mindfulness training or
conducted in the program developed by the originator of to confounding factors such as placebo eects or passage of
MBSR (Kabat-Zinn, 1982, 1990). Similarly, the MBCT time (Chambless & Hollon, 1998). Outcome studies using
groups in Teasdale et al. (2000) were led by the developers waiting-list or no-treatment controls might shed more light
of the treatment. In these cases it seems likely that thera- on the eects of mindfulness training as a treatment pack-
pists conducted the treatment competently. However, be- age, but more rigorous tests would compare mindfulness-
cause mindfulness-based interventions are relatively new based interventions to established treatments. Dismantling
and may be less familiar than more established cognitive- studies of treatment packages that include both mindfulness
behavioral interventions, descriptions of the training and and behavior change strategies, such as DBT, ACT, and
supervision of the therapists conducting the mindfulness RP, could clarify the relative contributions of acceptance-
treatment might increase condence in the ndings from based and change-based strategies in these packages.
future studies. Whether the eectiveness of established treatment pro-
grams may be increased by adding mindfulness training is
Clinical Signicance. The clinical signicance of the eects also an important question. Additional research could in-
of an intervention can be evaluated in several ways ( Ja- vestigate the eects of mindfulness practice on a broader
cobson & Revensdorf, 1988; Jacobson & Truax, 1991; range of outcomes, such as subjective well-being and qual-
Kazdin, 1994). For example, after patients have completed ity of life, as well as symptom reduction. The mechanisms
the experimental treatment, the extent to which they fall through which mindfulness training may create clinical
within the normal range on relevant dependent measures change, such as exposure, relaxation, and cognitive change,
can be examined. Alternatively, their diagnostic status can also should be examined.
be reevaluated to determine whether they continue to The Division 12 Task Force on Promotion and Dis-
meet criteria for the disorder for which they sought treat- semination of Psychological Procedures (1995) proposed
ment. The studies reviewed here do not explicitly address denitions for well-established and probably ecacious
the clinical signicance of their ndings in either of these treatments. Well-established treatments have been shown
ways. Increased attention to the issue of clinical signi- to be superior to a placebo or alternative treatment, or
cance would contribute substantially to the utility of future equivalent to an already established treatment, in group-
studies. design studies with adequate sample sizes and conducted by
dierent investigators. Alternatively, they have demon-
CONCLUSION strated ecacy in a large series of single case designs that
In spite of signicant methodological aws, the current compare the intervention to another treatment. In all cases,
literature suggests that mindfulness-based interventions well-established treatments have been investigated for
may help to alleviate a variety of mental health problems specic disorders, with use of treatment manuals and well-
and improve psychological functioning. These studies also specied samples.
suggest that many patients who enroll in mindfulness-based Designation as probably ecacious requires two stud-
programs will complete them, in spite of high demands for ies showing the treatment to be more eective than a wait-
homework practice, and that a substantial subset will con- ing-list control group, or than another treatment (but
tinue to practice mindfulness skills long after the treatment conducted by the same investigator), or two studies demon-
program has ended. Mindfulness-based interventions ap- strating eectiveness in heterogeneous client samples.
pear to be conceptually consistent with many other em- Five studies of MBSR using group designs with random
pirically supported treatment approaches and may provide assignment are reviewed here (Astin, 1997; Kabat-Zinn et
a technology of acceptance to complement the technology al., 1998; Shapiro et al., 1998; Speca et al., 2000; K. A.
of change exemplied by most cognitive-behavioral pro- Williams et al., 2001). All show MBSR to be more eec-
cedures (Linehan, 1993a). tive than a waiting-list or TAU control group. Samples in-
Thus, it appears that methodologically sound studies of clude students (two studies), psoriasis patients, cancer
mindfulness-based interventions would be very informa- patients, and community volunteers complaining of high
tive. Randomized clinical trials are needed to clarify stress levels. Thus, MBSR may meet criteria for the prob-
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V10 N2, SUMMER 2003 140
Delmonte, M. M. (1985). Meditation and anxiety reduction: A fulness meditation: Theoretical considerations and prelimi-
literature review. Clinical Psychology Review, 5, 91102. nary results. General Hospital Psychiatry, 4, 3347.
Derogatis, L. R. (1983). SCL-90-R: Administration, scoring, and Kabat-Zinn, J. (1990). Full catastrophe living: Using the wisdom
procedures manual-II. Towson, MD: Clinical Psychometric Re- of your body and mind to face stress, pain, and illness. New
search. York: Delacorte.
Goldenberg, D. L., Kaplan, K. H., Nadeau, M. G., Brodeur, C., Kabat-Zinn, J. (1994). Wherever you go, there you are: Mindfulness
Smith, S., & Schmid, C. H. (1994). A controlled study of a meditation in everyday life. New York: Hyperion.
stress-reduction, cognitive-behavioral treatment program in Kabat-Zinn, J. (2000). Indras net at work: The mainstreaming of
bromyalgia. Journal of Musculoskeletal Pain, 2, 5366. Dharma practice in society. In G. Watson & S. Batchelor
Goleman, D. J., & Schwartz, G. E. (1976). Meditation as an in- (Eds.), The psychology of awakening: Buddhism, science, and our
tervention in stress reactivity. Journal of Consulting and Clinical day-to-day lives (pp. 225249). Nork Beach, ME: Weiser.
Psychology, 44, 456466. Kabat-Zinn, J., & Chapman-Waldrop, A. (1988). Compliance
Guerrero, J., & Reiter, R. (1992). A brief survey of penial gland with an outpatient stress reduction program: Rates and pre-
immune system interrelationships. Endocrinology Research, 18, dictors of program completion. Journal of Behavioral Medicine,
91113. 11, 333352.
Hanh, T. N. (1976). The miracle of mindfulness. Boston: Beacon Kabat-Zinn, J., Lipworth, L, & Burney, R. (1985). The clinical
Press. use of mindfulness meditation for the self-regulation of
Hayes, S. C. (1987). A contextual approach to therapeutic chronic pain. Journal of Behavioral Medicine, 8, 163190.
change. In N. S. Jacobson (Ed.), Psychotherapists in clinical prac- Kabat-Zinn, J., Lipworth, L., Burney, R., & Sellers, W. (1987).
tice: Cognitive and behavioral perspectives (pp. 327387). New Four-year follow-up of a meditation-based program for the
York: Guilford Press. self-regulation of chronic pain: Treatment outcomes and com-
Hayes, S. C. (1994). Content, context, and the types of psycho- pliance. Clinical Journal of Pain, 2, 159173.
logical acceptance. In S. C. Hayes, N. S. Jacobson, V. M. Fol- Kabat-Zinn, J., Massion, M. D., Kristeller, J., Peterson, L. G.,
lette, & M. J. Dougher (Eds.), Acceptance and change: Content Fletcher, K. E., Pbert, L., et al. (1992). Eectiveness of a
and context in psychotherapy (pp. 1332). Reno, NV: Context meditation-based stress reduction program in the treatment
Press. of anxiety disorders. American Journal of Psychiatry, 149, 936
Hayes, S. C., Jacobson, N. S., Follette, V. M., & Dougher, M. J. 943.
(Eds.). (1994). Acceptance and change: Content and context in Kabat-Zinn, J., Wheeler, E., Light, T., Skillings, Z., Scharf, M. J.,
psychotherapy. Reno, NV: Context Press. Cropley, T. G., et al. (1998). Inuence of a mindfulness
Hayes, S. C., Strosahl, K., & Wilson, K. G. (1999). Acceptance and meditationbased stress reduction intervention on rates of
Commitment Therapy. New York: Guilford Press. skin clearing in patients with moderate to severe psoriasis
Hayes, S. C., & Wilson, K. G. (1993). Some applied implications undergoing phototherapy (UVB) and photochemotherapy
of a contemporary analytic account of verbal events. Behavior (PUVA). Psychosomatic Medicine, 50, 625632.
Analyst, 16, 283301. Kaplan, K. H., Goldenberg, D. L., & Galvin, N. M. (1993). The
Heatherton, T. F., & Baumeister, R. F. (1991). Binge eating as es- impact of a meditation-based stress reduction program on
cape from self-awareness. Psychological Bulletin, 110, 86108. bromyalgia. General Hospital Psychiatry, 15, 284289.
Ito, J. R., Donovan, D. M., & Hall, J. J. (1988). Relapse preven- Kazdin, A. E. (1994). Methodology, design, and evaluation in
tion in alcohol aftercare: Eects on drinking outcome, change psychotherapy research. In A. E. Bergin & S. L. Gareld
process, and aftercare attendance. British Journal of Addiction, (Eds.), Handbook of psychotherapy and behavior change (4th ed.,
83, 171181. pp. 1971). New York: Wiley.
Jacobson, N. S., & Revensdorf, D. (1988). Statistics for assessing Kohlenberg, R. J., Hayes, S. C., & Tsai, M. (1993). Radical be-
the clinical signicance of psychotherapy techniques: Issues, havioral psychotherapy: Two contemporary examples. Clini-
problems, and new developments. Behavioral Assessment, 10, cal Psychology Review, 13, 579592.
133145. Koons, C. R., Robins, C. R., Tweed, J. L., Lynch, T. R., Gon-
Jacobson, N. S., & Truax, P. (1991). Clinical signicance: A sta- zalez, A. M., Morse, J. Q., et al. (2001). Ecacy of dialecti-
tistical approach to dening meaningful change in psycho- cal behavior therapy in women veterans with borderline
therapy research. Journal of Consulting and Clinical Psychology, personality disorder. Behavior Therapy, 32, 371390.
59, 1219. Kristeller, J. L., & Hallett, C. B. (1999). An exploratory study of
Kabat-Zinn, J. (1982). An outpatient program in behavioral med- a meditation-based intervention for binge eating disorder.
icine for chronic pain patients based on the practice of mind- Journal of Health Psychology, 4, 357363.
CLINICAL PSYCHOLOGY: SCIENCE AND PRACTICE V10 N2, SUMMER 2003 142
Strosahl, K. D., Hayes, S. C., Bergan, J., & Romano, P. (1998). cognitive therapy. Journal of Consulting and Clinical Psychology,
Assessing the eld eectiveness of Acceptance and Commit- 68, 615623.
ment Therapy: An example of the manipulated training re- Wallace, R. K., Benson, H., & Wilson, A. F. (1984). A wakeful
search method. Behavior Therapy, 29, 3564. hypometabolic physiologic state. In D. H. Shapiro, Jr., &
Swenson, C. R. (2000). How can we account for DBTs wide- R. N. Walsh (Eds.), Meditation: Classic and contemporary per-
spread popularity? Clinical Psychology: Science and Practice, 7, spectives (pp. 417431). New York: Aldine.
8791. Williams, J. M. G., Teasdale, J. D., Segal, Z. V., & Soulsby, J.
Task Force on Promotion and Dissemination of Psychological (2000). Mindfulness-based cognitive therapy reduces over-
Procedures. (1995). Training in and dissemination of empir- general autobiographical memory in formerly depressed pa-
ically validated psychological treatments: Report and recom- tients. Journal of Abnormal Psychology, 109, 150155.
mendations. Clinical Psychologist, 48, 323. Williams, K. A., Kolar, M. M., Reger, B. E., & Pearson, J. C.
Teasdale, J. D. (1999). Metacognition, mindfulness, and the (2001). Evaluation of a wellness-based mindfulness stress re-
modication of mood disorders. Clinical Psychology and Psycho- duction intervention: A controlled trial. American Journal of
therapy, 6, 146155. Health Promotion, 15, 422432.
Teasdale, J. D., Segal, Z. V., & Williams, M. G. (1995). How does Zettle, R. D., & Raines, J. C. (1989). Group cognitive and con-
cognitive therapy prevent depressive relapse and why should textual therapies in treatment of depression. Journal of Clinical
attentional control (mindfulness training) help? Behaviour Re- Psychology, 45, 438445.
search and Therapy, 33, 2539.
Teasdale, J. D., Williams, J. M., Soulsby, J. M., Segal, Z. V., Received August 27, 2001; revised January 15, 2002; accepted
Ridgeway, V. A., & Lau, M. A. (2000). Prevention of August 27, 2002.
relapse/recurrence in major depression by mindfulness-based