Content uploaded by Mark A Lau
Author content
All content in this area was uploaded by Mark A Lau on Jul 23, 2015
Content may be subject to copyright.
Journal of Consulting and Clinical Psychology Copyright 2000 by the American Psychological Association, Inc.
2000, Vol. 68, No. 4, 615-623 0022-006X/00/$5.00 DOI: 10.1037//0022-006X.68.4.615
Prevention of Relapse/Recurrence in Major Depression
by Mindfulness-Based Cognitive Therapy
John D. Teasdale
Medical Research Council Cognition and Brain Sciences Unit
Zindel V. Segal
Centre for Addiction and Mental Health, Clarke Division, and
University of Toronto
J. Mark G.
Williams
University of Wales
Valerie A. Ridgeway
Medical Research Council Cognition and Brain Sciences Unit
Judith M. Soulsby
University of Wales
Mark A. Lau
Centre for Addiction and Mental Health, Clarke Division, and
University of Toronto
This study evaluated mindfuiness-based cognitive therapy (MBCT), a group intervention designed to train
recovered recurrently depressed patients to disengage from dysphoria-activated depressogenic thinking that
may mediate relapse/recurrence. Recovered recurrently depressed patients (n = 145) were randomized to
continue with treatment as usual or, in addition, to receive MBCT. Relapse/recurrence to major depression was
assessed over a 60-week study period. For patients with 3 or more previous episodes of depression (77% of
the sample), MBCT significantly reduced risk of relapse/recurrence. For patients with only 2 previous
episodes, MBCT did not reduce relapse/recurrence. MBCT offers a promising cost-efficient psychological
approach to preventing relapse/recurrence in recovered recurrently depressed patients.
Relapse and recurrence following successful treatment of major
depressive disorder (MDD) is common and often carries massive
social cost (Mintz, Mintz, Arruda, & Hwang, 1992). Reviewing
John D. Teasdale and Valerie A. Ridgeway, Medical Research Council
Cognition and Brain Sciences Unit, Cambridge, United Kingdom; Zindel
V. Segal, Centre for Addiction and Mental Health, Clarke Division, To-
ronto, Ontario, Canada, and Departments of Psychology and Psychiatry,
University of Toronto, Toronto, Ontario, Canada; J. Mark G. Williams and
Judith M. Soulsby, Institute for Medical and Social Care Research, Uni-
versity of Wales, Bangor, United Kingdom; Mark A. Lau, Centre for
Addiction and Mental Health, Clarke Division, Toronto, Ontario, Canada,
and Department of Psychiatry, University of Toronto.
This research was supported in part by Grant RA 013 from the Wales
Office of Research and Development for Health and Social Care and by
Grant MH53457 from the National Institute of Mental Health.
We are most grateful to Jon Kabat-Zinn, Saki Santorelli, Ferris Ur-
banowski, Elana Rosenbaum, and the staff of the Center for Mindfulness in
Medicine, Health Care and Society, University of Massachusetts Medical
Center, for invaluable guidance and support in treatment development. We
express appreciation to Sally Cox, Susan Williams, Neff Rector, and Michael
Gemar for assistance with data management and analysis; to Keith Evans, who
independently allocated patients to treatment groups; and to John Hedges for
invaluable help with patient recruitment. We thank Surbala Morgan and Isabel
Hargreaves for their contributions to treatment development and Leyland
Sheppard and Alison Jenaway for independent psychiatric diagnostic ratings.
Finally, our sincere thanks to Christina Feldman for her considerable contri-
butions to the instruction of the instructors.
Correspondence concerning this article should be addressed to John D.
Teasdale, Medical Research Council Cognition and Brain Sciences
Unit, 15 Chaucer Road, Cambridge CB2 2EF, United Kingdom. Electronic
mail may be sent to john.teasdale@mrc-cbu.cam.ac.uk.
615
studies of lifetime course of depression, a recent commentary
concluded that "it has been established that unipolar major depres-
sive disorder is a chronic, lifelong illness, the risk for repeated
episodes exceeds 80%, patients will experience an average of 4
lifetime major depressive episodes of 20 weeks duration each"
(Judd, 1997, p. 990). Such data suggest that the prevention of
relapse and recurrence poses a central challenge in the overall
management of MDD. Currently, maintenance pharmacotherapy is
the best validated and most widely used approach to prophylaxis in
depression, the lowest rates of recurrence occurring when patients
are continued at the dosage of antidepressant medication used to
achieve remission (Kupfer et al., 1992).
Maintenance psychotherapy may also be helpful. The pioneer-
ing work of Frank, Kupfer, and colleagues (e.g., Frank et al., 1990;
Frank, Kupfer, Wagner, McEachran, & Comes, 1991; Kupfer et
al., 1992) has shown that continuation of a psychological treatment
(interpersonal psychotherapy) in maintenance form can also sig-
nificantly extend survival time following recovery. Cognitive-
behavioral therapy (CBT) for depression (Beck, Rush, Shaw, &
Emery, 1979), administered during depressive episodes, appears to
he effective in reducing subsequent rates of relapse and recurrence.
Studies comparing the long-term outcome of patients who recov-
ered following treatment of acute depression by CBT with the
outcome of patients who recovered following treatment with an-
tidepressant medication and who were then withdrawn from med-
ication have consistently found less relapse or need for further
treatment in the CBT group (Blackburn, Eunson, & Bishop, 1986;
Evans et al., 1992; Shea et al., 1992; Simons, Murphy, Levine, &
Wetzel, 1986). Such findings suggest that CBT may be a treatment
for acute depression that has long-term effects in reducing risk of
future relapse and recurrence, presumably through patients acquir-
616 TEASDALE ET AL.
ing skills, or changes in thinking, that confer some degree of
protection against future onsets.
A recent novel approach to the prevention of relapse and recur-
rence in depression, for which there is encouraging preliminary
evidence, is to combine pharmacotherapy for the acute episode
with psychological prophylactic interventions administered fol-
lowing recovery. Fava and colleagues (e.g., Fava, Grandi,
Zielezny, Canestrari, & Morphy, 1994; Fava, Grandi, Zielezny,
Rafanelli, & Canestrari, 1996; Fava, Rafanelli, Grandi, Conti, &
Belluardo, 1998) have reported successful use of such an ap-
proach, combining treatment of the acute episode by antidepres-
sant medication with provision of CBT, following recovery, while
antidepressant medication is gradually withdrawn. For example,
Fava et al. (1998) described the results of a trial comparing the
long-term outcome of 40 patients with recurrent major depression
(three or more episodes) successfully treated with antidepressant
medication and then randomized to clinical management or a
combination of (a) CBT for residual symptoms, (b) lifestyle mod-
ification, and (c) well-being therapy, while antidepressant medi-
cation was withdrawn. Over a 2-year follow-up, the CBT group
showed significantly less relapse/recurrence (25%) than the clini-
cal management group (80%).
The strategy of combining acute pharmacotherapy with psycho-
logical prophylaxis offers the possibility of (a) capitalizing on the
cost-efficiency of antidepressant medication to reduce acute symp-
tomatology while (b) avoiding the need for patients to remain
indefinitely on maintenance medication to reduce future relapse
and recurrence. In this article, we describe a multicenter trial
evaluating the effectiveness of this strategy using a novel, theory-
driven approach to psychological prophylaxis, mindfulness-based
cognitive therapy (MBCT). To increase the potential cost-
efficiency of this strategy, MBCT was designed as a group skills-
training approach rather than as an individual psychological ther-
apy. In contrast to Fava et al. (1998), we (a) focused on a group
intervention rather than an individual intervention, (b) studied
more than a single therapist, (c) used a larger sample size, and (d)
administered the psychological intervention at least 3 months after,
rather than during, withdrawal of antidepressant medication.
The theoretical background to MBCT (referred to previously
[Teasdale, Segal, & Williams, 1995] as attentional control [mind-
fulness] training) has been described in detail elsewhere (Segal,
Williams, Teasdale, & Gemar, 1996; Teasdale et al., 1995). It is
assumed that vulnerability to relapse and recurrence of depression
arises from repeated associations between depressed mood and
patterns of negative, self-devaluative, hopeless thinking during
episodes of major depression, leading to changes at both cognitive
and neuronal levels. As a result, individuals who have recovered
from major depression differ from individuals who have never
experienced major depression in the patterns of thinking subse-
quently activated by dysphoria.
Specifically, it is suggested that, in recovered depressed pa-
tients, the thinking activated by dysphoria will show similarities to
the thinking patterns previously present in episode. These reacti-
vated patterns of thinking can act to maintain and intensify the
dysphoric state through escalating and self-perpetuating cycles of
ruminative cognitive-affective processing (Teasdale, 1988, 1997).
In this way, in those with a history of major depression, states of
mild dysphoria will be more likely to progress to more intense and
persistent states, thereby increasing risk of further onsets of epi-
sodes of major depression.
Studies that have compared the patterns of thinking activated by
mild dysphoria in those with and without a history of major
depression support this account (Ingram, Miranda, & Segal, 1998;
Segal, Gemar, & Williams, 1999). This analysis provides a parallel
explanation, at the cognitive level, to more biological accounts of
episode sensitization and kindling in recurrent affective disorder
(Post, 1992). Accounts at both biological and cognitive levels are
consistent with the finding that, with repeated experiences of
episodes of major depression, less environmental stress is required
to provoke relapse/recurrence (Post, 1992). That is, the processes
mediating relapse/recurrence appear to become progressively more
autonomous with increasing experience of episodes of depression.
The above account suggests that risk of relapse and recurrence
will be reduced if patients who have recovered from episodes of
major depression can learn, first, to be more aware of negative
thoughts and feelings at times of potential relapse/recurrence and,
second, to respond to those thoughts and feelings in ways that
allow them to disengage from ruminative depressive processing
(Nolen-Hoeksema, 1991). MBCT was designed to achieve those
alms (Teasdale et al., 1995). MBCT is based on an integration of
aspects of CBT for depression (Beck et al., 1979) with components
of the mindfulness-based stress reduction program (MBSR) devel-
oped by Kabat-Zinn and colleagues (e.g., Kabat-Zinn, 1990).
There is preliminary evidence for the effectiveness of MBSR in the
treatment of generalized anxiety disorder (GAD) and panic
(Kabat-Zinn et al., 1992) and chronic pain (Kabat-Zinn, Lipworth,
Burney, & Sellers, 1986). Unlike CBT, there is little emphasis in
MBCT on changing the content of thoughts; rather, the emphasis
is on changing awareness of and relationship to thoughts. Aspects
of CBT included in MBCT are primarily those designed to facil-
itate "decentered" views, such as "Thoughts are not facts" and "I
am not my thoughts."
The focus of MBCT is to teach individuals to become more
aware of thoughts and feelings and to relate to them in a wider,
decentered perspective as "mental events" rather than as aspects of
the self or as necessarily accurate reflections of reality. It is
assumed that the cultivation of a detached, decentered relationship
to depression-related thoughts and feelings is central in providing
individuals with skills to prevent the escalation of negative think-
ing patterns at times of potential relapse/recurrence (Teasdale,
1997; Teasdale et al., 1995). Because, unlike CBT, there is little
explicit emphasis in MBCT on changing the content or specific
meanings of negative automatic thoughts, in MBCT training can
occur in the remitted state, using everyday experience as the object
of training.
We report an initial multicenter randomized clinical trial eval-
uating the efficacy of MBCT in reducing relapse and recurrence in
patients with recurrent depressive disorder. Patients entered the
trial in remission, following treatment of previous episodes by
antidepressant medication. Choice of an appropriate design for the
initial evaluation of a novel intervention, such as MBCT, is influ-
enced by a number of factors. At the time this trial was planned,
there was no published evidence that any psychological interven-
tion, initially administered in the recovered state could, prospec-
tively, reduce risk of future recurrence in major depression. Given
this situation, the first priority for research was to evaluate whether
MBCT was of any benefit in reducing relapse/recurrence; if ben-
efits were observed, subsequent research could compare MBCT
with other psychological interventions, including controls for
MINDFULNESS-BASED COGNITIVE THERAPY
617
attention-placebo factors, and with alternative approaches to pre-
vention, such as maintenance pharmacotherapy.
We used a design in which patients who continued with treat-
ment as usual (TAU) were compared with patients who, addition-
ally, received training in MBCT. Such a design does not aim to
compare MBCT with the best available alternative preventive
intervention. Nor does it allow any reduction in rates of relapse and
recurrence for patients receiving MBCT to be attributed unambig-
uously to the specific components of MBCT rather than to non-
specific factors, such as therapeutic attention or group participa-
tion. However, this design is the most appropriate to answer the
question that was of primary interest in this initial evaluation of
MBCT: Does this intervention, when offered in addition to TAU,
reduce rates of relapse and recurrence compared to TAU alone?
Method
Design
At three treatment sites, 145 patients, currently in remission or recovery
from major depression at the time of the baseline assessment, were ran-
domized to continue with TAU or, additionally, to receive MBCT training.
Following an initial treatment phase, patients entered a 1-year follow-up
phase; a period of 1 year was selected because it has been a follow-up
reported in earlier studies (e.g., Simons et al., 1986) and because it was not
considered appropriate to defer the possibility for patients allocated to
TAU to participate in the MBCT program for a longer time (all of the
patients initially allocated to TAU were offered the possibility of MBCT on
completion of the follow-up year). Thus, the total 60-week study period
comprised an initial 8-week treatment phase followed by a 52-week
follow-up phase.
Randomization involved treatment sites faxing patient initials, date of
birth, gender, date of assessment, and details of number and recency of
previous episodes of depression to a central independent allocator. Infor-
mation was sent for groups of eligible patients at a time. The central
allocator randomly allocated patients to treatment condition, gave each a
study number, and faxed the allocations and study numbers back to
treatment sites.
Patients were stratified on two baseline variables--recency of recovery
from last episode of depression (within 0-12 months prior to randomiza-
tion vs. within 13-24 months prior to randomization) and number of
previous episodes of MDD (two vs. more than two)--and randomized by
strata within each site. Both of these variables have been found to be
related to risk of relapse/recurrence in previous studies (e.g., see Evans et
al., 1992; Post, 1992). A 1-year cutoff for recency of recovery meant that
all those in the less recent stratum were clearly recovered from their last
episode and all those who satisfied criteria for remission from episode, but
did not yet satisfy criteria for recovery, fell in the more recent stratum
(Frank, Prien, et al., 1991). A cutoff between those with only two episodes
and those with more than two episodes meant that those in the latter
stratum were broadly comparable with patient samples studied in other
trials of psychological treatments for recurrent depression (e.g., Fava et al.,
1998; Frank et al., 1990).
Sample size was calculated on the basis that a sample of 120 patients (60
per group), would have 80% power to detect at p < .05 a reduction in
relapse/recurrence rates from 50% in the TAU group to 28% in the MBCT
group on a directional hypothesis (Cohen, 1988).
Participants
Patients were recruited from community health care facilities and by
media announcements at three different sites: a predominantly rural,
Welsh-speaking area of north Wales centered on the small city of Bangor
(population 20,000); an area centered on and including the city of Cam-
bridge, England (population 110,000), together with surrounding small
towns, villages, and rural area; and the metropolitan area of Toronto,
Ontario, Canada (population 3 million). Although Cambridge is a well-
known university city, no participants at that site were actually academic
staff or students of the University of Cambridge.
Inclusion criteria were (a) 18 to 65 years of age; (b) meeting enhanced
Diagnostic and Statistical Manual of Mental Disorders
(3rd ed.;
DSM-
III-R;
American Psychiatric Association, 1987) criteria for a history of
recurrent major depression (these normally require a history of two or more
previous episodes of
DSM-II1-R
major depression in the absence of a
history of mania or hypomania; in addition, we required that at least two
episodes of major depression occurred within the past 5 years and that at
least one of those episodes was within the past 2 years); (c) a history of
treatment by a recognized antidepressant medication, but off antidepressant
medication, and in recovery/remission, at the time of baseline assessment
and for at least the preceding 12 weeks (it was not possible to determine the
adequacy of treatment by antidepressant medication; rather, this criterion
was used as an indicator that, in the naturalistic course of service delivery,
patients had been judged as appropriate for pharmacotherapy by a treating
physician); and (d) at baseline assessment, a 17-item Hamilton Rating
Scale for Depression (HRSD; Hamilton, 1960) score of less than 10.
Exclusion criteria were (a) history of schizophrenia or schizoaffective
disorder; (b) current substance abuse, eating disorder, or obsessive-
compulsive disorder (OCD); (c) organic mental disorder, pervasive devel-
opmental delay, or borderline personality disorder (BPD); (d) dysthymia
before age 20; (e) more than four sessions of cognitive-behavioral treat-
ment ever; (f) current psychotherapy or counseling more frequently than
once per month; and (g) current practice of meditation more than once per
week or yoga more than twice per week. Patients with eating disorders
were excluded because they frequently experience depression secondary to
those disorders and the MBCT program was not designed to deal with the
primary eating disorder. Patients with OCD were excluded because the
obsessional quality of their thoughts might have rendered the implemen-
tation of mindfulness strategies particularly difficult. Patients with dysthy-
mia before the age of 20 were excluded because of the possible charac-
terological nature of tbeir depression. Patients who currently practiced
yoga more than twice a week were excluded because yoga overlaps
considerably with mindfulness training and is, indeed, a component of the
MBCT program.
Informed Consent
Patients meeting the inclusion criteria, and willing to participate in the
study after it had been explained to them, gave written informed consent on
forms approved by local research ethics committees prior to randomization.
Measures
HRSD.
As part of the assessment of inclusion criteria, the baseline
assessment interview included the 17-item HRSD (Hamilton, 1960), a
widely used interview-based measure of severity of depressive symptom-
atology that covers a range of affective, behavioral, and biological symp-
toms. Scores can range from 0 to 52. This measure, administered by
doctoral-level psychologists or an experienced psychiatric social worker,
was also repeated at each subsequent follow-up assessment. The HRSD has
acceptable psychometric properties that have been reviewed elsewhere (see
Rabkin & Klein, 1987). A sample of 41 interviews from the follow-up
period were second-rated for the HRSD by an independent psychiatric rater
to yield an interrater correlation of r(39) = .963, p < .001.
Beck Depression Inventory. (BDI).
The BDI (Beck, Ward, Mendelson,
Mock, & Erbaugh, 1961), a widely used 21-item self-report measure of
severity of depressive symptoms, was completed by patients at the baseline
assessment and at each follow-up assessment. The BDI covers affective,
cognitive, motivational, behavioral, and biological symptoms of depression
and yields scores ranging from 0 to 63. The BDI has acceptable psycho-
618
TEASDALE ETAL.
metric properties that have been reviewed elsewhere (Rabkin & Klein,
1987).
Relapse~recurrence.
The primary-outcome variable was the occurrence
of relapse or recurrence meeting
DSM-III-R
criteria for major depressive
episode (American Psychiatric Association, 1987), as assessed by the
Structured Clinical Interview for
DSM-III-R
(SCID; Spitzer, Williams,
Gibbon, & First, 1992) administered at bimonthly assessments through the
follow-up period and covering the period from the previous assessment.
Assessments were made by doctoral-level psychologists and an experi-
enced psychiatric social worker. To maintain blindness of assessors to
treatment condition, we instructed patients not to reveal whether they were
receiving MBCT or any details that might prejudice blindness. Nonethe-
less, assessors occasionally became aware of a patient's treatment, condi-
tion. To overcome such occasional unblinding, and to examine interrater
reliability, interviews were audiotaped and all 133 occasions on which
patients met the screening criteria for major depression were evaluated by
an independent, blind, experienced research psychiatrist (any information
potentially revealing patients' treatment allocation was excluded from the
taped interview presented to the blind assessor). Only patients responding
positively to the screening question were included in this analysis. The
kappa for interrater agreement on categorization of presence/absence of
major depression was .74, which is indicative of good/excellent agreement.
Some of the disagreements arose from the fact that the first raters had wider
knowledge of the patients who they were ~ating and so were more able to
place the specific information elicited in the SCID interview in a wider
context that sometimes altered the significance of that specific information.
Also, of course, the second rater did not have access to the nonauditory
information that was available to the rater making the live rating. In cases
of disagreement, the blind ratings of the independent psychiatric rater were
used for analysis.
Following baseline assessment, interviews were scheduled at points
corresponding to the completion of the initial eight MBCT training ses-
sions and bimonthly thereafter over the course of the follow-up year.
Treatment
TAU.
Patients were instructed to seek help from their family doctor, or
other sources, as they normally would, should they encounter symptomatic
deterioration or other difficulties over the course of the study. The treat-
ment that patients in both the TAU and MBCT groups actually received
was monitored at the bimonthly assessment sessions and is described in the
Results section.
MBCT.
MBCT is a manualized group skills-training program (Segal,
Williams, & Teasdale, in press). MBCT is based on an integration of
aspects of CBT for depression (Beck et al., 1979) with components of the
MBSR program developed by Kabat-Zinn and colleagues (e.g., Kabat-
Zinn, 1990). It is designed to teach patients in renlkssion from recurrent
major depression to become more aware of, and to relate differently to,
their thoughts, feelings, and bodily sensations (e.g., relating to thoughts and
feelings as passing events in the mind rather than identifying with them or
treating them as necessarily accurate readouts on reality). The program
teaches skills that allow individuals to disengage from habitual ("auto-
matic") dysfunctional cognitive routines, in particular depression-related
ruminative thought patterns, as a way to reduce future risk of relapse and
recurrence of depression.
After an initial individual orientation session, the MBCT program is
delivered by an instructor in eight weekly 2-hr group training sessions
involving up to 12 recovered recurrently depressed patients. During that
period, the program includes daily homework exercises. Homework invari-
ably includes some form of guided (taped) or unguided awareness exer-
cises, directed at increasing moment-by-moment nonjudgmental awareness
of bodily sensations, thoughts, and feelings, together with exercises de-
signed to integrate application of awareness skills into daily life. Key
themes of the program include empowerment of participants and a focus on
awareness of experience in the moment. Participants are helped to cultivate
an open and acceptant mode of response, in which they intentionally face
and move in to difficulties and discomfort, and to develop a decentered
perspective on thoughts and feelings, in which these are viewed as passing
events in the mind.
A core feature of the program involves facilitation of an aware mode of
being, characterized by freedom and choice, in contrast to a mode domi-
nated by habitual, ovedearned, automatic patterns of cognitive-affective
processing. For patients, this distinction is often illustrated by reference to
the common experience, when driving on a familiar route, of suddenly
realizing that one has been driving for miles "on automatic pilot," unaware
of the road or other vehicles, preoccupied with planning future activities or
ruminating on a current concern. By contrast, "mindful" driving is asso-
ciated with being fully present in each moment, consciously aware of
sights, sounds, thoughts, and body sensations as they arise. When one is
mindful, the mind responds afresh to the unique pattern of experience in
each moment instead of reacting "mindlessly" to fragments of a total
experience with old, relatively stereotyped, habitual patterns of mind.
Increased mindfulness is relevant to the prevention of relapse/recurrence of
depression as it allows early detection of relapse-related patterns of nega-
tive thinking, feelings, and body sensations, thus allowing them to be
"nipped in the bud" at a stage when this may be much easier than if such
warning signs are not noticed or are ignored. Further, entering a mindful
mode of processing at such times allows disengagement from the relatively
automatic ruminative thought patterns that would otherwise fuel the relapse
process. Formulation of specific relapse/recurrence prevention strategies
(such as involving family members in an "early warning" system, keeping
written suggestions to engage in activities that are helpful in interrupting
relapse-engendering processes, or looking out for habitual negative
thoughts) are also included in the later stages of the initial 8-week phase.
Following the initial phase of weekly group meetings, four follow-up
meetings were scheduled at intervals of 1, 2, 3, and 4 months.
MBCT sessions were video- or audiotaped, with patients' permission, to
allow monitoring of treatment integrity.
Instructors
The three instructors were all experienced cognitive therapists who had,
jointly, developed the MBCT program. Each had previously led at least one
cohort of recovered depressed patients through the MBCT program.
Results
Intent-to-Treat and Per-Protocol Samples
Results were analyzed separately for an intent-to-treat sample
(n --- 145), comprising all of the patients included in the random
allocation, and a per-protocol sample (n = 132), comprising (a) all
of the patients allocated to the TAU condition (n = 69) and (b)
those patients allocated to MBCT who received a predetermined
"minimum effective dose" of MBCT (at least four of the eight
weekly MBCT sessions; n = 63). The results from these two
samples are complementary: The intent-to-treat sample provides a
stringent test of whether the MBCT and TAU groups differed in
outcome, reducing possible artifactual selective effects of differ-
ential attrition from the two treatment conditions, and the per-
protocol sample provides an estimate of the benefits of MBCT
among those who actually experienced at least a minimally ade-
quate exposure to that treatment program.
Patient Flow
One hundred forty-nine patients met the inclusion criteria at a
baseline screening interview and were invited to participate in the
study. Of these, 4 declined, leaving 145 patients to be randomized.
Of the 13 patients allocated to MBCT not included in the per-
MINDFULNESS-BASED COGNITIVE THERAPY 619
protocol sample, 6 failed to attend any training sessions and 7 (9%
of those allocated to MBCT) dropped out after attending fewer
than four sessions.
Complete data on relapse or recurrence were available for 137
(95%) of the 145 patients in the intent-to-treat sample and 128
(97%) of the 132 patients in the per-protocol sample; data were
incomplete for 3 TAU patients, 4 "insufficient treatment" MBCT
patients, and 1 "adequate treatment" MBCT patient.
Patient Characteristics
Baseline characteristics of the intent-to-treat sample are given in
Table 1.
The TAU and MBCT treatment groups were closely similar on
each of the baseline variables, with the exception of age. Given the
size of this difference in means in relation to standard deviations,
age was included as a covariate in all of the comparisons of
treatment group outcome. For the sample as a whole, social class
distribution (Office of Population Censuses and Surveys, 1991)
was as follows (percentages for the general population of England
and Wales are given in parentheses for comparison): for Class 1
(e.g., general managers of large corporations), 5% (4%); for
Class 2, 40% (21%); for Class 3, 45% (46%); for Class 4, 7%
(17%); for Class 5 (e.g., road sweepers), 3% (8%); and for armed
services/unclassified, 0% (5%). Class distribution was very similar
in the TAU (M = 2.7,
SD
= 0.9) and MBCT (M = 2.6,
SD
= 0.8)
groups. Basic patient characteristics across the three sites were as
follows: for Bangor (n = 45), mean age was 44.0 years
(SD
= 9.5)
and 73% were female; for Cambridge (n = 54), mean age was 44.5
years
(SD
= 10.6) and 78% were female; and for Toronto (n =
46), mean age was 41.3 years (10.6) and 76% were female.
Comparison of the 13 "insufficient treatment" patients in the
MBCT group, who either attended no treatment sessions or
dropped out before completing at least four sessions, with the 63
patients who completed four or more sessions revealed no statis-
tically significant differences between these groups on baseline
characteristics (smallest p = .17).
TA U
The treatment for depression actually received by patients in the
TAU condition was monitored at the bimonthly assessment inter-
views over the follow-up period and is summarized in Table 2. The
corresponding data for patients in the MBCT condition are also
shown for comparison. There were no statistically significant
differences between the TAU and MBCT conditions for any of
these measures of treatment received (all ps > . 10).
Outcome Analysis: Relapse/Recurrence to Major
Depression
Time to onset of relapse or recurrence (in weeks) was compared
between treatment groups using Cox proportional hazards regres-
sion models (SPSS, 1994, pp. 291-328), with treatment condition
as a categorical (indicator) variable and TAU as the reference
condition. In the results that follow, 95% confidence intervals
(CIs) for hazard ratios are provided following Wald and hazard
ratio statistics.
To examine whether effects of treatment condition were mod-
erated by either of the stratifying variables used in randomization,
it was necessary to conduct preliminary Cox regression analyses
that included, separately, each of these variables (recency of last
episode of depression [0-12 months vs. 13-24 months] and num-
ber of previous episodes of MDD [two vs. more than two]) and its
interaction with treatment condition, as covariates, together with
treatment condition (MBCT vs. TAU). These analyses revealed a
significant effect of the interaction of number of previous episodes
and treatment condition in both the intent-to-treat sample, Wald(1)
Table 1
Baseline Characteristics of Treatment as Usual (TAU) and Mindfulness-Based
Cognitive Therapy (MBCT) Samples
Variable TAU (n = 69) MBCT (n = 76)
Female (%) 78 74
White (%) 100 97
Age (years) 46.2 --- 9.6 40.7 ___ 10.3
Marital status (%)
Single 12 18
Married/cohabiting 57 55
Divorced/separated/widowed 32 26
Years of education 14.3 --- 3.3 14.9 --- 3.1
Depression
Median HRSD score (IQR) 3.0 (4.3) 4.0 (5.0)
Median BDI score (IQR) 10.0 (10.0) 10.0 (10.0)
Median previous episodes (IQR) 3.0 (3.8) 3.5 (2.0)
Age of first onset (years) 28.1 --- 10.4 25.7 --- 9.9
Median duration of episodes (weeks)
Last (IQR) 15.0 (19.0) 17.5 (16.3)
Penultimate (IQR) 22.0 (32.0) 16.0 (22.0)
Previous treatment for depression (%)
Antidepressant medication 100 100
Hospitalization 17 11
Psychotherapy/counseling 68 73
Note.
HRSD = Hamilton Rating Scale for Depression; IQR = interquartile range; BDI = Beck Depression
Inventory.
620 TEASDALE ET AL.
Table
2
Treatment for Depression From Other Sources Received by
Patients in Treatment as Usual (TAU) and Mindfulness-Based
Cognitive Therapy (MBCT) Over the 60-Week Study Period
Variable TAU MBCT
52 58
One or more depression-related visits
to general practitioner (%)
Psychiatric treatment (%)
Outpatient 8 10
Day patient 2 0
Inpatient 2 0
Counseling/psychotherapy/professional 34 49
mental health support (%)a
Other mental health contacts (%)b 21 17
Medication for depression (ADM; %) 40 45
Mean
(+_SD)
duration (weeks) 32.7 -+ 21.2 23.3 -+ 17.9
Mean
(+-SD)
reported dosage SSRI c 20.1 _+ 8.6 18.2 --_ 3.8
Note.
ADM = antidepressant medication.
"Includes psychiatric social worker, community psychiatric nurse, com-
munity mental health team worker, counselor, psychotherapist, group ther-
apy/support, and marital/family therapy, b Includes voluntary mental
health organizations (e.g., Samaritans) and health visitor, c SSRIs (selec-
tive serotonin reuptake inhibitors) were the most commonly prescribed
antidepressants; reported dosage is expressed in milligrams of fluoxetine
daily dose equivalents.
= 4.32, p < .05, and the per-protocol sample, Wald(1) = 4.32, p
< .05. That is, differences in outcome between treatment condi-
tions were not the same in participants with three or more previous
episodes as in participants with only two previous episodes, thus
mandating separate analyses for these two groups.
Figure 1 shows survival (i.e., nonrelapse/nonrecurrence) curves
comparing relapse/recurrence over the 60-week study period for
MBCT and TAU in participants with a history of three or more
episodes of depression. These participants composed 77% (105/
137) of the intent-to-treat sample for whom relapse/recurrence data
were available and 77% (99/128) of the per-protocol sample for
whom relapse/recurrence data were available. Cox regression anal-
yses showed significantly less hazard of relapse/recurrence in
MBCT participants, compared with TAU participants, for both the
intent-to-treat sample, Wald(1) = 6.65, p < .01, hazard ratio =
.473, CI = .267-.836, and the per-protocol sample, Wald(1) =
7.97, p < .005, hazard ratio = .419, CI = .229-.766. These
treatment effects remained significant when baseline values of the
HRSD or the BDI were also entered as covariates. Over the total
study period, in the intent-to-treat sample, 40% (22/55) of MBCT
participants experienced relapse/recurrence compared with 66%
(33/50) of TAU participants, )(2(1, N = 105) = 7.10, p < .01, a
39% reduction in risk of relapse/recurrence in the MBCT condi-
tion. The difference between 66% relapse/recurrence and 40%
relapse/recurrence yields an h value of .53, which Cohen (1988, p.
185) described as indicating a medium effect size. In the per-
protocol sample, corresponding figures were 37% (18/49) relapse/
recurrence for the MBCT group and 66% (33/50) relapse/recur-
rence for the TAU group, )(2(1, N = 99) = 8.49, p < .005, a 44%
reduction in risk of relapse/recurrence in the MBCT condition. The
difference between 66% relapse/recurrence and 37% relapse/re-
currence yields an h value of .59, which Cohen (1988, p. 185)
described as indicating a medium effect size.
The data from the per-protocol sample displayed in Figure 1
yield the following cumulative relapse rates (the TAU figure is
presented first, the MBCT figure second): for 10 weeks, 28%
versus 8%; for 20 weeks, 38% versus 20%; for 30 weeks, 44%
versus 26%; for 40 weeks, 60% versus 31%; and for 50 weeks,
66% versus 35%. These data appear to suggest that the differences
in relapse rates between TAU and MBCT become established
within the first 10 weeks of the study period, remain much the
same until 30 weeks, and then increase again. However, these
apparent trends should be interpreted with caution because (a) the
relapses from the TAU group are from smaller surviving popula-
tions than in the MBCT group so that numerical relapse underes-
timates probability of relapse in the TAU group and (b) the sample
sizes in the two groups mean that estimates of risk have apprecia-
ble margins of error.
Participants with a history of two episodes of depression com-
posed 23% (32/137) of the intent-to-treat sample for whom re-
lapse/recurrence data were available and 23% (29/128) of the
per-protocol sample for whom relapse/recurrence data were avail-
able. Cox regression analyses showed no significant differences in
hazard of relapse/recurrence between MBCT participants and
TAU participants for either the intent-to-treat sample, Wald(1) =
0.82, p > .10, or the per-protocol sample, Wald(1) = 0.67, p >
.10. Over the total study period, in the intent-to-treat sample, 56%
a.
I I .... Mindfulness-based CT (Intent-to-treat
1.0 I-];"--, I-- Treatment-as-usual
• 9t ~
--:--::.,
"rJ
~-~ ,~ ...... :-,
G. Z '----,
o~ .6 q . "
...........
~ .0
,~'6.4
.3
0 10 20 30 40 50 60
Weeks of Study
b.
1.0 °
Ol
.9
.8
~. .7
In_ .6
.3
- - "" Mlndfulneml-based CT (4+ lenlons)~
imt-alPusual
" """" '--,. ...... '..,...., _...,.. _ _,....,. ...........
0 10 20 30 40 50 60
Weeks of Study
Figure 1.
Survival (nonrelapse/nonrecurrence) curves comparing relapse/
recurrence to
Diagnostic and Statistical Manual of Mental Disorders
(3rd
ed.; American Psychiatric Association, 1987) major depression for treat-
ment as usual and mindfulness-based cognitive therapy in patients with
three
or more previous episodes of major depression: (a) intent-to-treat
sample and (b) per-protocol sample. CT = cognitive therapy.
MINDFULNESS-BASED COGNrFIVE THERAPY 621
(9/16) of MBCT participants experienced relapse/recurrence com-
pared with 31% (5/16) of TAU participants, X2(1, N = 32) = 2.03,
p >.10. In the per-protocol sample, corresponding figures were
54% (7/13) relapse/recurrence for the MBCT group and 31%
(5/16) relapse/recurrence for the TAU group, )(2(1, N = 29) =
1.51, p > .10.
To examine further the effects of number of previous episodes
on differential response to TAU and MBCT, we examined the
relationship between number of previous episodes (two vs. more
than two) and hazard of relapse/recurrence by separate Cox re-
gression analyses in the TAU and MBCT groups. In the TAU
group, there was a significant relationship between number of
previous episodes and relapse/recurrence, Wald(1) = 4.08, p <
.05. Further examination revealed a positive linear relationship
between number of previous episodes and risk of relapse/recur-
rence over the follow-up period: for two episodes, 31% relapse/
recurrence (5/16); for three episodes, 56% relapse/recurrence (10/
18); and for four or more episodes, 72% relapse/recurrence (23/
32), Mantel-Haenszel test for linear association, xZ(l, N = 66) =
7.06, p < .025. In the MBCT group, there was no significant
relationship between number of previous episodes and hazard of
relapse/recurrence in either the intent-to-treat sample, Wald(1) =
0.38, p > .10 (9 of 16 [56%] relapsed in the fewer-than-three-
episodes group, and 22 of 55 [40%] relapsed in the more-than-
two-episodes group), or the per-protocol sample, Wald(1) = 0.53,
p > .10 (7 of 13 [54%] relapsed in the fewer-than-three-episodes
group, and 18 of 49 [37%] relapsed in the more-than-two-episodes
group).
In summary, the main finding was that, in participants with three
or more previous episodes of depression (who composed 77% of
the sample), an "adequate dose" of MBCT almost halved relapse/
recurrence rates over the follow-up period compared with TAU.
Clinical Significance of Outcomes
The observed reduction in rates of relapse/recurrence for pa-
tients with more than two previous episodes of major depression
was statistically significant, but was it clinically significant? Ken-
dall, Marrs-Garcia, Nath, and Sheldrick (1999) have recently de-
scribed the use of normative comparisons as a method to evaluate
the clinical significance of the changes produced by therapeutic
interventions. This approach is particularly useful when applied to
patient populations that begin treatment with abnormally elevated
symptom scores and are reassessed on those measures following
treatment. In this situation, comparison of patients' posttreatment
scores with those from normative samples provides a valuable
indicator of the clinical significance of the extent of therapeutic
gains achieved.
Unfortunately, this elegant method is not applicable in the
present study. Unusual among clinical treatment trials, the key
outcome of interest in this study was the prevention of a future
event (relapse/recurrence) rather than reduction of symptoms
present at baseline assessment. Indeed, because it was assumed
that depression-related difficulties in concentration would interfere
with the implementation of MBCT, selection criteria for the trial
were deliberately chosen to exclude patients who were not largely
recovered or remitted. For example, at baseline assessment 86% of
patients fell in the asymptomatic range on the HRSD (Frank, Prien,
et al., 1991). In this situation,, it is clearly inappropriate to assess
the clinical significance of the outcomes in terms of the numbers
of patients falling in the asymptomatic range on posttreatment
assessments of severity of depressive symptomatology.
The relapse/recurrence rate in patients with three or more pre-
vious episodes treated with "adequate" MBCT (37%) was clearly
substantially above the expected annual incidence rate of MDD
among those with no prior history of major depression in general
population samples. On this basis, it is clear that the intervention
did not reduce risks of major depression to the "normal" range.
Nonetheless, the halving of relapse/recurrence rates in a group at
high risk for relapse/recurrence would appear to be a clinically
useful outcome. On this basis, we suggest that the benefits of
MBCT to patients with three or more previous episodes were both
statistically and clinically significant.
Use of Medication for Depression
To examine whether the reduction in relapse and recurrence in
patients with three or more episodes receiving MBCT was sec-
ondary to increased use of medications for depression, we com-
pared the proportions of patients in the two treatment groups using
such medications at any time over the follow-up period. This
procedure showed no significant differences between groups: for
the intent-to-treat sample, 40% (19/47) in the MBCT group and
46% (20/44) in the TAU group, x2(l, N = 91) = 0.24,p > .10; for
the per-protocol sample, 33% (14/42) in the MBCT group and 46%
(20/44) in the TAU group, X2(1, N = 86) = 1.32,p > .10. (These
figures differ from those in Table 2: The table shows figures for
the total TAU and MBCT samples, whereas these figures are for
patients with more than two previous episodes of depression.) The
lack of significant differences between the TAU and MBCT
groups in use of medications for depression or other forms of
treatment (see Table 2) in the presence of significantly less relapse/
recurrence in MBCT is open to a number of possible explanations.
The most parsimonious explanation is that these other treatments
contributed equally to the outcomes in the MBCT and TAU
conditions, the lower relapse in MBCT being attributable to the
effects of the MBCT intervention. Alternatively, it is conceivable
that MBCT may have made patients more responsive to the effects
of other treatments.
Comparison of Patients With Two Previous Episodes With
Patients With Three or More Previous Episodes
Exploratory analyses compared patients with two previous ep-
isodes of MDD with those with three or more episodes on a range
of background variables. The only significant differences observed
were on two age-related variables. Those with three episodes or
more were older when admitted into the study (for two episodes,
M = 38.88,
SD
= 9.84; for three or more episodes, M = 44.58,
SD
= 10.11), t(143) = 2.83, p < .01, and were younger when they
experienced their first episode (for two episodes, M = 33.38,
SD
= 8.65; for three or more episodes, M = 25.00,
SD
= 9.84),
t(143) = 4.36, p < .001. The difference in age of onset of first
episode suggests that these two groups may not simply represent
younger and older samples from essentially the same population
but may represent distinct populations, of patients. Combining
these two age-related variables into a single variable ("history")
reflecting the total duration of patients' experience with depression
(history = age at admission to study minus age of first onset)
yielded a mean for those with three or more episodes approxi-
622 TEASDALE ET AL.
rnately four times as great as that for patients with two episodes
(for two episodes, M = 5.50, SD = 4.79; for three or more
episodes, M = 19.58, SD = 10.33), unequal-variances t(l13) =
10.92, p < .001.
Discussion
For patients with recurrent major depression who had experi-
enced three or more previous episodes, MBCT approximately
halved rates of relapse and recurrence over the follow-up period
compared with patients who continued with TAU. This prophy-
lactic effect could not be accounted for in terms of patients who
received MBCT being more likely to use antidepressant medica-
tion. The preventative effect of MBCT was achieved for an aver-
age investment of less than 5 hr of instructor time per patient,
suggesting that offering a group skills-based training program to
recovered depressed patients may be a cost-efficient strategy for
prevention. It is important to note that MBCT was specifically
designed for remitted patients and is unlikely to be effective in the
treatment of acute depression, where factors such as difficulties in
concentration and the intensity of negative thinking may preclude
acquisition of the attentional control skills central to the program.
To our knowledge, the results of the present trial provide the first
demonstration that a group-based psychological intervention, ini-
tially administered in the recovered state, can significantly reduce
risk of future relapse/recurrence in patients with recurrent major
depression.
The finding that MBCT prevented relapse and recurrence in
patients with a history of three or more episodes of depression, but
not in patients With only two previous episodes, is of particular
interest with respect to the theoretical background to MBCT (Segal
et al., 1996; Teasdale et al., 1995). This program was specifically
designed to reduce the contribution of patterns of depressive
thinking reactivated by dysphoria to the processes mediating re-
lapse and recurrence. Such dysphoria-linked thinking, it was as-
sumed, resulted from repeated associations between the depressed
state and characteristic negative thinking patterns within each
depressive episode. The strengthening of these associations with
repeated episodes was assumed to contribute to the increased risk
of subsequent episodes following each episode experienced. In
particular, it was assumed that negative thinking reactivated by
dysphoria contributed to the increasingly autonomous nature of the
relapse/recurrence process with multiple episodes, reflected in the
observation that environmental provoking events appear to play a
progressively less important role in onset with increasing number
of episodes (Post, 1992).
The above account suggests the possibility that, in the present
study, (a) the greater risk of relapse/recurrence in those with three
or more episodes than in those with only two episodes (apparent in
the TAU group) was to a large extent attributable to autonomous
relapse/recurrence processes involving reactivation of depresso-
genic thinking patterns by dysphoria and (b) the prophylactic
effects of MBCT arose, specifically, from disruption of those
processes at times of potential relapse/recurrence. Consistent with
this analysis, MBCT appeared to have no prophylactic effects in
those with only two previous episodes, and the rate to which
relapse/recurrence was reduced by adequate MBCT in those with
three and more episodes (37%) was similar to the rate of relapse/
recurrence in those with only two episodes receiving TAU (31%).
The present findings add to a growing body of evidence (Fava
et al., 1996, 1998; Frank, Kupfer, et al., 1991) that psychological
interventions administered after recovery from the acute symptoms
of a depressive episode can substantially alter the future course of
MDD. These fmdings have considerable potential relevance for
our understanding of the cognitive and biological processes that
mediate the increased vulnerability to subsequent episodes of those
who have already experienced depressive episodes. An effective
prophylactic intervention offers an opportunity to investigate con-
trolled changes in vulnerability processes, with all the consequent
interpretative advantages conferred by experimental, as compared
with correlational, designs. However, the design of the present
study does not allow us to attribute the benefits of MBCT to the
specific skills taught by the program versus nonspecific factors,
such as therapeutic attention and group participation. Equally, the
present study provides no evidence of the extent to which similar
prophylactic effects would be obtained by instructors who had not
been actively involved in the development of the program or in
samples with different ethnic or educational backgrounds.
To our knowledge, this is the first multicenter randomized
clinical trial evaluating a mindfulness-based clinical intervention.
Taken with the results from smaller, or less controlled, evaluations
suggesting the effectiveness of the generic MBSR program in
treating chronic pain, GAD, and panic (Kabat-Zinn et al., 1986,
1992), and the effectiveness of a cognitive-behavioral program
incorporating a substantial mindfulness component in reducing
self-harm in BPD (Linehan, Armstrong, Suarez, Allmon, & Heard,
1991), the present findings suggest that mindfulness-based clinical
interventions may hold considerable therapeutic promise, either
alone or in combination with other forms of intervention.
References
American Psychiatric Association. (1987). Diagnostic and statistical man.
ual of mental disorders (3rd ed., rev.). Washington, DC: Author.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive
therapy of 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 Psychia°
try, 4, 561-571.
Blackburn, I. M., Eunson, K. M., & Bishop, S. (1986). A two-year
naturalistic follow-up of depressed patients treated with cognitive ther-
apy, pharmacotherapy, and a combination of both. Journal of Affective
Disorders, 10, 67-75.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences
(2nd ed.). Hillsdale, NJ: Erlbaum.
Evans, M. D., Hollon, S. D., DeRubeis, R. J., Piasecki, J. M., Grove,
W. M., Garvey, M. J., & Tuason, V. B. (1992). Differential relapse
following cognitive therapy and pharmacotherapy for depression. Ar-
chives of General Psychiatry, 49, 802-808.
Fava, G., Grandi, S., Zielezny, M., Canestrari, R., & Morphy, M. A.
(1994). Cognitive behavioral treatment of residual symptoms in primary
major depressive disorder. American Journal of Psychiatry, 151, 1295-
1299.
Fava, G., Grandi, S., Zielezny, M., Rafanelli, C., & Canestrari, R. (1996).
Four-year outcome for cognitive behavioral treatment of residual symp-
toms in major depression. American Journal of Psychiatry, 153, 945-
947.
Fava, G., Rafanelli, C., Grandi, S., Conti, S., & Belluardo, P. (1998).
Prevention of recurrent depression with cognitive behavioral therapy:
Preliminary findings. Archives of General Psychiatry, 55, 816-820.
Frank, E., Kupfer, D. J., Perel, J. M., Comes, C., Jarret, J. B., Mallinger,
A. G., Thase, M. E., McEachran, A. B., & Grochocinski, V. J. (1990).
MINDFULNESS-BASED COGNITIVE THERAPY
623
Three-year outcomes for maintenance therapies in recurrent depression.
Archives of General Psychiatry, 47,
1093-1099.
Frank, E., Kupfer, D. J., Wagner, E. F., McEachran, A. B., & Comes, C.
(1991). Efficacy of interpersonal therapy as a maintenance treatment of
recurrent depression.
Archives of General Psychiatry, 48,
1053-1059.
Frank, E., Prien, R. F., Jarrett, R. B., Keller, M. B., Kupfer, D. J., Lavori,
P. W., Rush, A. J., & Weissman, M. M. (1991). Conceptualisation and
rationale for consensus definitions of terms in major depressive disorder:
Remission, recovery, relapse and recurrence.
Archives of General Psy-
chiatry, 48,
851-855.
Hamilton, M. (1960). A rating scale for depression.
Journal of Neurology,
Neurosurgery, and Psychiatry, 23,
56-62.
Ingrain, R. E., Miranda, J., & Segal, Z. V. (1998).
Cognitive vulnerability
to depression.
New York: Guilford Press.
Judd, L. J. (1997). The clinical course of unipolar major depressive
disorders.
Archives of General Psychiatry, 54,
989-991.
Kabat-Zinn, J. (1990).
Full catastrophe living: The program of the Stress
Reduction Clinic at the University of Massachusetts Medical Center.
New York: Delta.
Kabat-Zinn, J., Lipworth, L., Burney, R., & Sellers, W. (1986). Four-year
follow-up of a meditation-based program for the self-regulation of
chronic pain: Treatment outcomes and compliance.
Clinical Journal of
Pain, 2,
159-173.
Kabat-Zinn, J., Massion, A. O., Kristeller, J., Peterson, L. G., Fletcher,
K. E., Pbert, L., Lenderking, W. R., & Santorelli, S. F. (1992). Effec-
tiveness of a meditation-based stress reduction program in the treatment
of anxiety disorders.
American Journal of Psychiatry, 149,
936-943.
Kendall, P. C., Marrs-Garcia, A., Nath, S., & Sheldrick, R. C. (1999).
Normative comparisons for the evaluation of clinical significance.
Jour-
nal of Consulting and Clinical Psychology, 67,
285-299.
Kupfer, D. J., Frank, E., Perel, J. M., Comes, C., Mallinger, A. G., Thase,
M. E., McEachran, A. B., & Grochocinski, V. J. (1992). Five-year
outcomes for maintenance therapies in recurrent depression.
Archives of
General Psychiatry, 49,
769-773.
Linehan, M. M., Armstrong, H. E., Suarez, A., Allmon, D., & Heard H. H.
(1991). Cognitive- behavioral treatment of chronically parasuicidal bor-
derline patients.
Arch~,es of General Psychiatry, 48,
1060-1064.
Mintz, J., Mintz, L. I., Arruda, M. J., & Hwang, S. S. (1992). Treatment of
depression and the functional capacity to work.
Archives of General
Psychiatry, 49,
761-768.
Nolen-Hoeksema, S. (1991). Responses to depression and their effects on
the duration of depressive episodes.
Journal of Abnormal Psychology,
100,
569-582.
Office of Population Censuses and Surveys. (1991).
Standard occupational
classification
(Vol. 3). London: Her Majesty's Stationery Office.
Post, R. M. (1992). Transduction of psychosocial stress into the neurobi-
ology of recurrent affective disorder.
American Journal of Psychiatry,
149,
999-1010.
Rabkin, J. G., & Klein, D, F. (1987). The clinical measurement of depres-
sive disorders. In A. Marsella, R. Hirschfeld, & M. Katz (Eds.), The
measurement of depression
(pp. 30-83). New York: Guilford Press.
Segal, Z. V., Gemar, M., & Williams, S. (1999). Differential cognitive
response to a mood challenge following successful cognitive therapy or
pharmacotherapy for unipolar depression.
Journal of Abnormal Psychol-
ogy, 108,
3-10.
Segal, Z. V., Williams, J. M. G., & Teasdale, J. D. (in press).
Mindfulness
and the prevention of depression: A guide to the theory and practice of
mindfulness-based cognitive therapy.
New York: Guilford Press.
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.
Shea, M. T., Elkin, I., Imber, S. D., Sotsky, F. M., Watkins, J. T., Collins,
J. F., Pilkonis, P. A., Becldaam, E., Glass, R., Dolan, R. T., & Parloff,
M. B. (1992). Course of depressive symptoms over follow-Up: Findings
from the NIMH Treatment of Depression Collaborative Research Pro-
gram.
Archives of General Psychiatry, 49,
782-787.
Simons, A. D., Murphy, G. E., Levine, J. L., & Wetzel, R. D. (1986).
Cognitive therapy and pharmacotherapy for depression: Sustained im-
provement over one year.
Archives of General Psychiatry, 43,
43-50.
Spitzer, R. L., Williams, J. B. W., Gibbon, M., & First, M. B. (1992). The
Structured Clinical Interview for
DSM-III-R
(SCID): I. History, ratio-
nale, and description.
Archives of General Psychiatry, 49,
624-629.
SPSS. (1994).
SPSS advanced statistics 6.1.
Chicago: Author.
Teasdale, J. D. (1988). Cognitive vulnerability to persistent depression.
Cognition and Emotion, 2,
247-274.
Teasdale, J. D. (1997). The relationship between cognition and emotion:
The mind-in-place in mood disorders. In D. M. Clark & C. G. Fairbum
(Eds.),
Science and practice of cognitive behaviour therapy
(pp. 67-93).
Oxford, England: Oxford University Press.
Teasdale, J. D., 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 Ther-
apy, 33,
25-39.
Received June 1, 1999
Revision received December 20, 1999
Accepted December 22, 1999 •
