Mindfulness-Based Cognitive Therapy to Prevent Relapse in Recurrent Depression

ArticleinJournal of Consulting and Clinical Psychology 76(6):966-78 · January 2009with751 Reads
Impact Factor: 4.85 · DOI: 10.1037/a0013786 · Source: PubMed

For people at risk of depressive relapse, mindfulness-based cognitive therapy (MBCT) has an additive benefit to usual care (H. F. Coelho, P. H. Canter, & E. Ernst, 2007). This study asked if, among patients with recurrent depression who are treated with antidepressant medication (ADM), MBCT is comparable to treatment with maintenance ADM (m-ADM) in (a) depressive relapse prevention, (b) key secondary outcomes, and (c) cost effectiveness. The study design was a parallel 2-group randomized controlled trial comparing those on m-ADM (N = 62) with those receiving MBCT plus support to taper/discontinue antidepressants (N = 61). Relapse/recurrence rates over 15-month follow-ups in MBCT were 47%, compared with 60% in the m-ADM group (hazard ratio = 0.63; 95% confidence interval: 0.39 to 1.04). MBCT was more effective than m-ADM in reducing residual depressive symptoms and psychiatric comorbidity and in improving quality of life in the physical and psychological domains. There was no difference in average annual cost between the 2 groups. Rates of ADM usage in the MBCT group was significantly reduced, and 46 patients (75%) completely discontinued their ADM. For patients treated with ADM, MBCT may provide an alternative approach for relapse prevention.


Available from: Barbara Barrett
Mindfulness-Based Cognitive Therapy to Prevent Relapse in
Recurrent Depression
Willem Kuyken
University of Exeter
Sarah Byford
King’s College London
Rod S. Taylor
Peninsula Medical School
Ed Watkins, Emily Holden, and Kat White
University of Exeter
Barbara Barrett
King’s College London
Richard Byng
Peninsula Medical School
Alison Evans
Devon Primary Care Trust
Eugene Mullan
University of Exeter
John D. Teasdale
Cambridge, Cambridgeshire, United Kingdom
For people at risk of depressive relapse, mindfulness-based cognitive therapy (MBCT) has an additive benefit
to usual care (H. F. Coelho, P. H. Canter, & E. Ernst, 2007). This study asked if, among patients with recurrent
depression who are treated with antidepressant medication (ADM), MBCT is comparable to treatment with
maintenance ADM (m-ADM) in (a) depressive relapse prevention, (b) key secondary outcomes, and (c) cost
effectiveness. The study design was a parallel 2-group randomized controlled trial comparing those on
m-ADM (N 62) with those receiving MBCT plus support to taper/discontinue antidepressants (N 61).
Relapse/recurrence rates over 15-month follow-ups in MBCT were 47%, compared with 60% in the m-ADM
group (hazard ratio 0.63; 95% confidence interval: 0.39 to 1.04). MBCT was more effective than m-ADM
in reducing residual depressive symptoms and psychiatric comorbidity and in improving quality of life in the
physical and psychological domains. There was no difference in average annual cost between the 2 groups.
Rates of ADM usage in the MBCT group was significantly reduced, and 46 patients (75%) completely
discontinued their ADM. For patients treated with ADM, MBCT may provide an alternative approach for
relapse prevention.
Keywords: mindfulness-based cognitive therapy, randomized controlled trial, depression, antidepres-
sants, health economics
Depression is a major public health problem, in part because like
other chronic conditions it tends to run a relapsing course (Judd,
1997b; Keller et al., 1984). Without treatment, people suffering
recurrent depression experience relapse at rates as high as 80%
(Frank et al., 1990; Kupfer et al., 1992; Prien & Kupfer, 1986).
Currently the majority of those with depression are treated in
Willem Kuyken, Ed Watkins, Emily Holden, Kat White, and Eugene Mullan,
Mood Disorders Centre, University of Exeter, Exeter, Devon, United Kingdom;
Sarah Byford and Barbara Barrett, Centre for the Economics of Mental Health,
King’s College London, London; Rod S. Taylor, Peninsula Technology Assess-
ment Group, Peninsula Medical School, Exeter, Devon, United Kingdom; Richard
Byng, Primary Care Research group, Peninsula Medical School; Alison Evans,
Devon Primary Care Trust, Barnstable, Devon, United Kingdom; John D. Teas-
dale, Cambridge, Cambridgeshire, United Kingdom.
Authors are listed in the following order: principal investigator (Willem
Kuyken), coinvestigators (Sarah Byford, Rod S. Taylor, Ed Watkins), research
staff (Emily Holden and Kat White), study therapist and collaborators (Barbara
Barrett, Richard Byng, Alison Evans, Eugene Mullan, and John D. Teasdale).
Contributors are listed alphabetically within each section.
No financial or other conflicts of interest exist. This trial was registered
(ISRCTN12720810) and was funded by the UK Medical Research Council
(TP 72167). We are grateful to the patients who participated in the trial; the
physicians and other health care staff who enabled the trial; Becca Crane
and Trish Bartley at the University of Bangor Centre for Mindfulness
Research and Practice for their input to the MBCT therapist training; the
independent members of the Trial Steering Committee (John Campbell
[Chair], Emer O’Neill, Richard Moore, Paul Lanham, and Andy Richards);
and Rachel Hayes and Vicky Green for research assistance.
Correspondence concerning this article should be addressed to
Willem Kuyken, Mood Disorders Centre, School of Psychology, Uni-
versity of Exeter, Exeter EX4 4QG, United Kingdom. E-mail:
Journal of Consulting and Clinical Psychology Copyright 2008 by the American Psychological Association
2008, Vol. 76, No. 6, 966 –978 0022-006X/08/$12.00 DOI: 10.1037/a0013786
Page 1
primary care (Katon & Schulberg, 1992), and maintenance anti-
depressant medication (m-ADM) is the mainstay approach (Ged-
des et al., 2003; National Institute for Clinical Excellence [NICE],
2004). To stay well, people with a history of recurrent depression
are recommended to continue antidepressant medication (ADM)
for at least 2 years. However, many experience unpleasant side
effects, rates of adherence tend to be low, and many express a
preference for psychosocial interventions (Cooper et al., 2007;
NICE, 2004; Olfson, Marcus, Tedeschi, & Wan, 2006; van Schaik
et al., 2004). User group and professional consensus recommends
as priorities for future research (a) the development of psychoso-
cial interventions to prevent depressive relapse and (b) the use of
nontraditional delivery systems, such as group interventions, to
significantly expand the accessibility of cost-effective therapies
(Hirschfield et al., 1997; Hollon et al., 2002).
In response to this challenge, mindfulness-based cognitive ther-
apy (MBCT) was developed with a specific focus on preventing
relapse/recurrence of depression (Segal, Williams, & Teasdale,
2002). MBCT is a relatively brief, 8-week group program. With
8 –15 patients per group, MBCT has the potential to help a large
number of people in primary care settings at relatively low cost
compared with individual therapies. In two randomized controlled
trials, MBCT plus usual care halved the rates of relapse compared
with usual care over a 60-week follow-up period among people
who had experienced three or more previous episodes of major
depression: Approximately two thirds of the sample relapsed in the
usual care arm versus approximately one third relapsed in the
MBCT plus usual care arm (Ma & Teasdale, 2004; Teasdale et al.,
2000). It is noteworthy that an exclusion criterion for these trials
was current ADM treatment, even though this is the most common
treatment approach for recurrent depression. Moreover, there are
currently no trials comparing MBCT with an active treatment
(Coelho et al., 2007).
The rationale for this trial was as follows. First, given that
m-ADM over at least 2 years is recommended to prevent relapse/
recurrence (Geddes et al., 2003; NICE, 2004) and many patients
express a preference for an alternative approach, it is important to
establish whether MBCT enables patients to taper/discontinue
ADM. Second, the majority of depression presents in primary care,
and the generalizability of MBCT to real-world primary care
settings has not been established. Third, the two trials to date were
conducted by members of the group that developed MBCT, and
there is a need for independent replication. Fourth, in health and
mental health trials there is a growing acknowledgement of the
need for a more sophisticated and patient-centered approach to
outcome assessment that extends beyond the assessment of depres-
sive relapse/recurrence to include measures of the nature of de-
pressive relapses/recurrences, comorbidity, and quality of life
(e.g., Zimmerman et al., 2006). Finally, MBCT has not yet been
compared with another active treatment.
The primary aim of this trial was to examine whether MBCT
provides an alternative approach to m-ADM in preventing depres-
sive relapse/recurrence. We compared the sustained recovery of
people taking m-ADM with that of people who participated in an
MBCT program and were supported in tapering and discontinuing
their m-ADM. Our secondary study aim was to compare MBCT
and m-ADM in terms of residual depressive symptoms, comorbid
psychiatric diagnoses, quality of life, and cost effectiveness. Fi-
nally, we sought to establish if MBCT enabled patients to taper/
discontinue their ADM.
Study participants had a history of three or more previous
episodes of depression, had been treated with a therapeutic dose of
ADM over the last 6 months, and were currently either in full or
partial remission from the most recent episode. They were ran-
domly allocated to participate in either a traditional m-ADM
treatment or an 8-week MBCT class that included support to
taper/discontinue their m-ADM. Block randomization (block
size 4) to the two groups was performed by an independent
statistician using computer-generated quasi-random numbers.
Randomization was stratified according to patients’ symptomatic
status at intake assessment by using the Hamilton Rating Scale for
Depression (HRSD; J. B. Williams, 1988; asymptomatic
HRSD 8; partially symptomatic HRSD 8). The study was
approved by the UK National Health Service North and East
Devon Research Ethics Committee.
Time from randomization to depressive relapse/recurrence was
the primary outcome measure, with patients followed up at
3-month intervals for 15 months. Secondary outcome measures
were severity/duration of relapses/recurrences, severity of residual
depressive symptoms, number of comorbid psychiatric diagnoses,
quality of life, and service use.
The study was conducted in primary care settings across a range
of urban and rural locations in Devon, England. Recruitment was
designed to screen as wide a population as possible in primary care
(White, Holden, Byng, Mullan, & Kuyken, 2007). First, we
searched computerized practice databases to identify patients who
had been prescribed ADM for the previous 6 months. Medical
records were then consulted to establish as far as possible whether
patients met study inclusion and exclusion criteria. Primary care
physicians then screened the list of selected patients and wrote
letters to potential participants describing the study, enclosing the
study information sheet, and stating that unless they decided to opt
out they would be contacted by a member of the study team.
Unless potential participants opted out, a research officer made
contact by telephone to discuss the study, and with potential
participants’ further verbal consent they were screened for eligi-
bility. If eligible and willing, patients attended a meeting for the
study intake assessment, where they were invited to sign a formal
consent form.
One hundred twenty-three people met the study’s criteria and
agreed to participate. Inclusion criteria were as follows: three or
more previous episodes of depression meeting criteria for depres-
sion according to the Diagnostic and Statistical Manual of Mental
Disorders (4th ed.; DSM–IV; American Psychiatric Association,
1994); 18 years of age or older; and on a therapeutic dose of
m-ADM in line with the British National Formulary (British
Medical Association & Royal Pharmaceutical Society of Great
Britain [BMA–RPSGB], 2006) for at least the previous 6 months
and is in either full or partial remission from the most recent
Page 2
episode of depression. Exclusion criteria were as follows: comor-
bid diagnoses of current substance dependence; organic brain
damage; current/past psychosis; bipolar disorder; persistent anti-
social behavior; persistent self-injury requiring clinical manage-
ment/therapy; unable to engage with MBCT for physical, practical,
or other reasons (e.g., very disabling physical problem, unable to
comprehend materials); and formal concurrent psychotherapy.
MBCT and antidepressant tapering/discontinuation. MBCT is
a manualized, group-based skills training program designed to
enable patients to learn skills that prevent the recurrence of de-
pression (Segal, Williams, & Teasdale, 2002). It is derived from
mindfulness-based stress reduction, a program with proven effi-
cacy in ameliorating distress in people suffering chronic disease
(Baer, 2003; Kabat-Zinn, 1990), and cognitive– behavioral therapy
for acute depression (Beck, Rush, Shaw, & Emery, 1979), which
has demonstrated efficacy in preventing depressive relapse/
recurrence (Hollon et al., 2005). MBCT is intended to enable
people to learn to become more aware of the bodily sensations,
thoughts, and feelings associated with depressive relapse and to
relate constructively to these experiences. It is based on theoretical
and empirical work demonstrating that depressive relapse is asso-
ciated with the reinstatement of automatic modes of thinking,
feeling, and behaving that are counterproductive because they
contribute to and maintain depressive relapse and recurrence (e.g.,
self-critical thinking and avoidance; Lau, Segal, & Williams,
2004). Participants learn to recognize these “automatic pilot”
modes, step out of these modes, and respond in healthier ways by
intentionally moving into a mode in which they decenter from
negative thoughts/feelings (e.g., by learning that “thoughts are not
facts”), accept difficulties with a stance of self-compassion, and
use bodily awareness to ground and transform their experience. In
the latter stages of the course, patients develop an “action plan”
that sets out strategies for responding when they become aware of
early warning signs of relapse/recurrence (see J. M. G. Williams,
Teasdale, Segal, & Kabat-Zinn, 2007).
The MBCT relapse prevention intervention was delivered in
primary care settings with MBCT groups of 9 –15 patients follow-
ing the treatment protocol (Segal, Williams, & Teasdale, 2002):
2-hr sessions over 8 consecutive weeks, followed by four
follow-up sessions in the following year. Session content included
guided mindfulness practices (i.e., body scan, sitting meditation,
yoga); inquiry into patients’ experience of these practices; review
of weekly homework (i.e., 40 min of mindfulness practice per day
and generalization of session learning); and teaching/discussion of
cognitive– behavioral skills. In line with the two previous MBCT
trials, an adequate dose of MBCT was defined as participation in
at least four of the eight MBCT group sessions.
All trial groups were videotaped with digital cameras for ther-
apist supervision, checks on therapist competence, and checks on
treatment adherence. The five trial groups were instructed by either
a clinical psychologist or an occupational therapist (three and two
groups, respectively). Both therapists had undergone a training
program taught and supervised by one of the developers of MBCT
(John D. Teasdale), had experience of running at least two super-
vised pilot groups, and had an ongoing personal mindfulness
practice. An independent check on therapist competency was es-
tablished before therapists progressed to running trial groups: An
experienced MBCT therapist independent of the trial rated at least
two videotapes of MBCT therapy sessions and made an overall
judgment as to whether the therapists were competent.
Patients in the MBCT arm were supported in tapering and
discontinuing their ADM by their primary care physician. Patients
and physicians were initially prompted to begin discussing a
tapering/discontinuation regime after 4 –5 weeks of the MBCT
groups. At the end of the MBCT groups, they were reminded to
ensure a tapering/discontinuation regime was in place. We rea-
soned that because our sample reported three or more previous
episodes of depression and had received at least 6 months of
m-ADM, their tapering/discontinuation scheme would need to be
conducted with great care. Tapering/discontinuation regimes were
determined by physicians and patients, although the research team
asked that patients consider tapering/discontinuing their medica-
tion as soon following MBCT as they deemed appropriate and
within 6 months of the MBCT group ending. This allowed (a)
tapering to be conducted at a pace determined by physicians and
patients and (b) a substantial window to the study’s end when
patients had discontinued m-ADM to monitor the primary and
secondary outcomes. The study team provided guideline informa-
tion to physicians and patients about typical tapering/discontinu-
ation regimes and possible withdrawal effects. If at any time the
study team became aware of difficulties with medication tapering/
discontinuation, the MBCT therapist first contacted the patient to
understand the difficulty and then wherever appropriate encour-
aged the patient together with their physician to review the taper-
ing/discontinuation regime.
Maintenance antidepressant treatment. The m-ADM relapse
prevention intervention comprised maintenance of the ADM treat-
ment that was an inclusion criterion for the study. Patients were
monitored and treated by their physicians in primary care settings.
During the maintenance phase, physicians were asked to manage
m-ADM in line with standard clinical practice and the British
National Formulary. Primary care physicians were asked to meet
with patients regularly to review their medication treatment.
Changes in medication sometimes occurred during the mainte-
nance treatment stage, but physicians and patients were asked to
ensure the dose remained within therapeutic limits.
Medication adherence. Medication adherence was monitored
through patients’ self-report at follow-ups every 3 months, practice
databases, and the Morisky Medication Adherence Scale (MMAS;
Morisky, Green, & Levine, 1986). Scores of 0 –1 on the MMAS
are considered to indicate high levels of adherence; scores of 2 or
more in our study were taken to signal a possible adherence
problem requiring action. If problems were identified at any as-
sessment point, these were resolved through dialogue between a
member of the research team not blind to treatment condition, the
prescribing physician, and the patient, but we ensured that the
research officer conducting follow-ups remained blind to treatment
condition. Normally this addressed any problems with the
m-ADM. However, if there were ongoing problems with adher-
ence, these were addressed on a case-by-case basis with the goal of
encouraging patients to continue taking a therapeutic dose of
m-ADM for the duration of the follow-up period. Protocol adher-
ence was defined as continuing to take m-ADM at a therapeutic
maintenance dose for the duration of the trial.
Page 3
Outcome Measures
Patients were assessed by research staff blind to treatment
allocation at intake and then again every 3 months up to 15 months
Primary outcome measure: relapse/recurrence. The primary
outcome measure was time to relapse/recurrence of depression,
using the depression module of the Structured Clinical Interview
for DSM–IV (SCID; First, Spitzer, Gibbon, & Williams, 1995) to
assess retrospectively the 3-month period between assessments.
Relapse/recurrence was defined as an episode meeting DSM–IV
criteria for major depressive disorder. An experienced clinical
psychologist with formal training in the use of the SCID trained
the two research staff. To examine interrater reliability, we fol-
lowed the method described in the first MBCT randomized control
trial (RCT; Teasdale et al., 2000), which has the added benefit of
guaranteeing all assessments were blind to treatment condition.
For every first actual, borderline, or probable relapse/recurrence,
an independent, blind, and experienced rater second-rated an audio
recording of the SCID interview. The kappa coefficient for agree-
ment between the study interviewer and blind rater was 0.84,
suggesting excellent agreement. Where there were disagreements
between the first and independent rater, consensus was reached
through discussion. If a relapse/recurrence was considered mar-
ginal, a conservative position of no relapse was recorded. Once a
judgment about relapse was made, the onset of relapse was dated
from randomization to the point at which criteria were met.
Secondary outcome measures. To broaden the measurement of
outcomes (Zimmerman et al., 2006), we included several second-
ary outcome measures. For any relapses/recurrences, the severity
of the relapse/recurrence was assessed using the DSM–IV specifi-
ers: “mild,” “moderate,” “severe without psychotic features,” and
“severe with psychotic features” (scale range 1– 4; American Psy-
chiatric Association, 1994). Through the follow-ups, we assessed
the duration of any relapse/recurrence (i.e., period of time in
months that a person met SCID criteria) and the associated dis-
tress, which was rated by patients on a 1–100-point scale ranging
from 0 (the least distressing episode of depression I have ever
experienced) to 100 (the most distressing episode of depression I
have ever experienced).
Residual depressive symptoms cause impairment and predict
future relapse/recurrence (Judd et al., 1999). To assess residual
symptoms, we used the observer-rated interviewer-administered,
17-item version of the HRSD (J. B. Williams, 1988) and the
21-item self-report Beck Depression Inventory (2nd ed.; BDI–II;
Beck, Steer, & Brown, 1996). Psychiatric comorbidity was as-
sessed at study intake, and all comorbid diagnoses identified at
intake were reassessed at the study’s end using the relevant SCID
modules (First et al., 1995).
To assess quality of life, we used the 26-item, self-report, short
version of the World Health Organization Quality of Life instru-
ment (WHOQOL-BREF), which assesses subjective quality of life
in four domains: physical (e.g., “How satisfied are you with your
sleep?”), psychological (e.g., “How much do you enjoy life?”),
social (e.g., “How satisfied are you with your personal relation-
ships?”), and environment (e.g., “How satisfied are you with your
access to health services?”
; Harper & Power, 1998; World Health
Organization, 2004). Data are reported on only the first three
Service Use, Productivity Losses, and Cost
The economic evaluation took a broad perspective, including all
hospital (inpatient, outpatient, emergency department) and com-
munity health and social services (e.g., primary care, social work,
complementary therapies), plus productivity losses resulting from
time off work due to illness. Economic data were collected in
interview at baseline and then in 3-month intervals up to 15
months postrandomization using the Adult Service Use Schedule
(AD-SUS), an instrument designed on the basis of previous studies
of adult mental health populations (Bower et al., 2000; Byford et
al., 2003). The AD-SUS asks recipients for the number and length
of contacts with various services and professionals relevant to the
disease of interest over the previous 3 months. To ensure that the
AD-SUS covered all services relevant to the current population,
we checked the schedule against information from a recent sys-
tematic review of economic evaluations in depression (Barrett,
Byford, & Knapp, 2005). Studies were reviewed for service use
categories included, and any items missing from the AD-SUS were
added. Complementary therapies were the main addition. Data on
MBCT contacts was collected from therapist records to avoid
having patients reveal their treatment group to the research
All unit costs were for the financial year 2005–2006, the most
recent financial year over which the trial data were collected.
MBCT group sessions, each lasting 2 hr, were costed on the basis
of the salary of the trial therapists plus overhead expenses (admin-
istrative, managerial, and capital). Calculation of the indirect time,
including preparation and supervision, was based on information
provided by the trial therapists on the ratio of direct face-to-face
contact to all other MBCT activities. National United Kingdom
unit costs were applied to medication, hospital contacts, and com-
munity health and social services (BMA–RPSGB, 2006; Curtis &
Netten, 2006; Department of Health, 2006). Productivity losses
were calculated using the human capital approach, which involves
multiplying the individual’s salary (mean salary of study partici-
pants $28,248 per annum) by reported days off work due to illness
(Koopmanschap & Rutten, 1996). All costs were converted to
international dollars using a purchasing power parity exchange rate
of 0.6 as recommended by the World Bank (2006 World Devel-
opment Indicators available at http://www.worldbank.org/).
Data Analysis
This is the first trial to directly compare MBCT with an active
therapy arm, and consequently an estimate of the relative treatment
effect of MBCT versus m-ADM was not available. A formal
sample size calculation at the outset was therefore deemed inap-
propriate. Instead, an estimation approach was taken and the
between-group difference in the primary outcome was reported as
a mean and 95% confidence interval. Reanalysis of the two pre-
vious MBCT trials (Ma & Teasdale, 2004; Teasdale et al., 2000)
found an intraclass correlation coefficient for relapse of less than
zero in the treatment groups (J. M. G. Williams, personal commu-
From The World Health Organization Quality of Life (WHOQOL)-
BREF, 2004, Geneva, Switzerland: World Health Organization. Copyright
2004 by the World Health Organization. Reprinted with permission.
Page 4
nication), suggesting adjustments for clustering were not neces-
sary. Nonetheless, we examined heterogeneity across groups.
Time to relapse/recurrence of depression (primary outcome) for
the two treatment groups was compared using Cox regression
proportional hazard survival analysis, with treatment condition
(MBCT or m-ADM) as the independent variable and allowing for
the stratification variable (i.e., asymptomatic vs. partially symp-
tomatic). Individuals who experienced no relapse/recurrence were
considered as censored. The analysis was performed according to
the principle of intention to treat (ITT; i.e., all patients according
to and included in random allocation). A secondary, “per protocol
treatment” (PPT) analysis, comprising all patients who stayed
within key treatment parameters as set out in the protocol, was
undertaken: For the MBCT group, patients attended at least four of
eight MBCT sessions, in line with both previous MBCT trials; for
the m-ADM patients, they continued to take ADM at a therapeutic
maintenance dose for the duration of the trial.
Mixed-models between-groups (ITT and stratification) and re-
peated measures (3, 6, 9, 12, and 15 months follow-ups) analysis
of covariance (baseline) was used to compare the groups with
respect to changes in secondary outcomes.
It was intended that all statistical models be run with and
without adjustment for baseline characteristics where the covariate
selection depended on the baseline comparison of groups. As no
differences in baseline covariates between groups were seen, anal-
yses were performed without adjustment for these variables. We
compared cases with and without missing data on demographic,
psychiatric, and outcome variables, and there were no differences
between these cases at p .05. For the primary survival outcome
analyses, drop out/missing data were handled by censoring. For the
small subset of cases with missing data on secondary outcomes, we
used last variable carried forward to impute missing data. Sensi-
tivity analyses were undertaken to explore the impact of imputa-
tion of data losses on secondary outcome analyses. The analyses
on secondary outcomes were unaffected by data imputation.
Differences in mean costs between MBCT and m-ADM groups
were analyzed using standard parametric t tests, with the validity
of results confirmed using bias-corrected, nonparametric boot-
strapping (repeat resampling; Efron & Tibshirani, 1993). Despite
the skewed nature of cost data, this approach is recommended to
enable inferences to be made about the arithmetic mean (Thomp-
son & Barber, 2000), a more meaningful summary statistic for
costs than is the median or geometric mean. Cost effectiveness was
explored through the calculation of incremental cost-effectiveness
ratios (ICER), defined as the difference in mean costs divided by
the difference in mean effects (Vanhout, Al, Gordon, & Ruten,
1994). Incremental cost per relapse prevented and per depression-
free day is reported.
Nonparametric bootstrapping from the costs and effectiveness
data was used to generate a joint distribution of incremental mean
costs and effects for the two treatments. This was used to calculate
the probability that each of the treatments is the optimal choice,
subject to a range of possible maximum values (ceiling ratio) that
a decision maker might be willing to pay for a unit improvement
in outcome. Cost-effectiveness acceptability curves are presented
by plotting these probabilities for a range of possible values of the
ceiling ratio (Fenwick, Claxton, & Sculpher, 2001). These curves
are a recommended decision-making approach to dealing with the
uncertainty that exists around the estimates of expected costs and
expected effects associated with the interventions under investiga-
tion (Claxton, 1999). All economic analyses were adjusted for
baseline costs and the stratification variable. All analyses were
undertaken using SPSS v15 and Stata v. 8.
Patient Flow
Figure 1 shows the patient flow from screening to follow-up.
The three main reasons potentially eligible patients did not partic-
ipate were (a) they declined to participate (533; 36.3%); (b) the
staff team was unable to make contact (451; 30.7%); and (c) they
did not meet the study criteria (362; 24.6%). The three main
reasons patients proved not to be eligible were (a) they had stopped
taking antidepressants (56; 15.5%), (b) they reported fewer than
three episodes of major depressive disorder (53; 15.5%), (c) they
were on a subtherapeutic dose of antidepressants and/or were
intending to reduce their antidepressant dose (52; 14.2%). The
three main reasons they declined were (a) the time commitment
was too much (103; 19.3%); (b) they wanted to stay on antide-
pressants (50; 9.4%), and (c) they did not like the group aspect of
the therapy (28; 5.3%). In summary, the sample can be character-
ized as a group of people with recurrent depression, treated phar-
macologically in primary care, who following a referral from their
primary care physician were interested in a psychological group-
based approach that included tapering/discontinuing their
One hundred twenty-three patients who could be located, agreed
to participate, and met the inclusion criteria were randomized to
either MBCT (N 61) or m-ADM (N 62) treatment. Two
patients in the MBCT arm and 6 in the m-ADM arm were lost to
follow-up. Of these, 5 were lost to follow-up shortly after intake,
and the remaining 3 after a depressive relapse occurred. In the
m-ADM arm, 10 (16%) patients were outside protocol because
they decided to discontinue their medication. In the MBCT arm, 9
(15%) patients fell outside protocol because they attended fewer
than four MBCT sessions (5 patients attended no sessions).
Patient Characteristics
Table 1 shows patient characteristics of the ITT sample. There
were no differences between the MBCT and m-ADM groups on
any of the patient characteristics ( p .10).
Patients who were per protocol were compared with those who
were not. In the m-ADM group, patients choosing to discontinue
ADM were comparable to those continuing their ADM on all
patient characteristics at intake (all ps .10) and on the primary
outcome (relapse/recurrence rates in both groups was identical at
60%) and the secondary outcome variables (all ps .10). In the
MBCT group, the patients who were within protocol were not
significantly different from those outside protocol on any patient
characteristics at intake ( p .05) except number of previous
suicide attempts (within protocol: 0.50, SD 1.21; outside pro-
tocol: 1.78, SD 1.56; Mann–Whitney U 101, p .01, N
61). The within- and outside-protocol MBCT groups were com-
parable on the primary outcome (within protocol: 46% relapsed;
outside protocol: 54% relapsed) and the secondary outcome vari-
Page 5
ables (all ps .05). No adverse events were recorded through the
oversight of the Trial Steering Committee.
Preliminary Analysis
At intake the type of ADM was as follows: selective re-uptake
inhibitor 71 (58%), tricyclic 27 (22%), or combination 25 (20%);
this was not significantly different across the two groups ( p
.10). Adherence to the MBCT protocol was assessed by an expe-
rienced and independent MBCT therapist with the Mindfulness-
Based Cognitive Therapy Adherence Scale (Segal, Teasdale, Wil-
liams, & Gemar, 2002). She watched videotapes of all 40 MBCT
sessions. The total adherence scale scores in the trial (M 29.1;
SD 4.69) were at least comparable to those reported in the
psychometric evaluation of this scale (Segal, Teasdale, et al.,
2002) and indicate acceptable adherence to protocol. This rater
also confirmed that the MBCT was delivered competently across
therapists and groups. To assess possible differences owing to
MBCT group or MBCT therapist, we compared relapse/recurrence
rates across these variables. The relapse/recurrence rates did not
differ statistically between the five MBCT groups (46%, 49%,
68%, 44%, and 53%, respectively),
(4, N 61) 0.87, p .93,
or across the two therapists (Therapist 1: 49%, Therapist 2: 46%),
(1, N 61) 0.06, p .814. Given the homogeneity in
relapse/recurrence rates across groups and therapists, adjustment
for the group-administered nature of MBCT (Baldwin, Murray, &
Shadish, 2005) was deemed unnecessary in this case.
At each follow-up point in the m-ADM arm, ADM was checked
to ensure that type and dose were within parameters set out in the
British National Formulary (BMA–RPSGB, 2006) and that patient
adherence was acceptable. The average score on the MMAS
(Morisky et al., 1986) was 0.63 (SD 0.62), suggesting consis-
tently high levels of adherence throughout the follow-up period.
Our trial sought to ask if MBCT enabled patients to taper/
discontinue their ADM. Over the course of the follow-up period
(ca. 450 days), the mean number of days on ADM treatment was
significantly different between the m-ADM (M 411.4, SD
91.77) and MBCT (M 266.46, SD 167.74) groups, t(101)
5.40, p .0001, d 1.07. At the end of the 6-month window
allowed for tapering/discontinuation, 46 (75%) of the patients in
the MBCT arm of the trial discontinued their medication. In the
further 6 months of the follow-up period (the post–tapering/
discontinuation window), the rates of ADM usage between the two
groups continued to be highly significantly different, t(105)
4.85, p .0001, d 0.93.
Outcome Analysis: Relapse/Recurrence to
Major Depression
Figure 2 shows survival (i.e., non–relapse/recurrence) curves
over the 15-month study period for the intention to treat MBCT
and m-ADM groups. Cox regression showed borderline evidence
of a reduction in the hazard of relapse/recurrence with MBCT
compared with m-ADM in both ITT analysis, Wald (1, N
123) 1.82, p .07, hazard ratio 0.63 (95% CI: 0.39 to 1.04),
and PPT analysis, Wald (1, N 101) –1.95, p .05, hazard
ratio 0.59 (95% CI: 0.34 to 1.00). In the ITT sample over the
total 15-month follow-up period, 47% (29/61) of the MBCT pa-
tients relapsed, compared with 60% (37/62) of the m-ADM pa-
tients, log-rank
(1) 1.54, p .21. In the PPT sample, 46%
(24/52) of the MBCT patients had a relapse/recurrence, compared
Assessed for eligibility
N = 1,469)
Excluded (
= 1,346)
362 (24.6%) found not to be
it bl
449 (30.6%) did not return contact
533 (36.3%) declined
2 (0.1%) did not contact (did not
answer phone)
N = 123
Allocated to MBCT intervention
= 61)
Allocated to ADM intervention
(n = 62)
Lost to follow-up (
= 2)
Lost to follow-up (n = 6)
Prophylaxis and
Too busy / “moved on”
Anxious about group and talking about
her difficulties
Did not attend
4 MBCT sessions (n
Includes 2 participants who were
Unwilling to stay on ADM
Too busy and doesn’t like forms
Lost to contact
Not adherent to ADM protocol (n = 10)
because discontinued medication
subsequently lost to follow-up
= 61)
= 62)
Per protocol (n = 52)
Per protocol (n = 52)
Figure 1. CONSORT flow diagram. MBCT mindfulness-based cognitive therapy; ADM antidepressant
Page 6
with 60% (31/52) of the m-ADM patients, log-rank
(1) 3.32,
p .07.
Two exploratory subgroup analyses were set out in the analysis
plan. First, although we restricted our sample to people with three
or more episodes, we wanted to establish that there were no
interaction effects for those with more prior episodes, Wald (1,
N 123) –1.29, p .20, hazard ratio 0.9 (95% CI: 0.76 to
1.06). Second, as this was the first trial to include patients in full
and partial remission from recurrent depression, we looked at
effects of severity of depression (i.e., asymptomatic vs. partially
Table 1
Characteristics of MBCT and m-ADM Intention to Treat Samples
Variable MBCT (n 61) m-ADM (n 62)
Demographic characteristics
Women: n (%) 47 (77) 47 (76)
White: n (%)
60 (98) 62 (100)
Age (in years)
M (SD) 48.95 (10.55) 49.37 (11.84)
Range 26–66 21–72
Marital status: n (%)
Single 4 (7) 9 (15)
Married or cohabiting 42 (69) 40 (65)
Separated, divorced, or widowed 15 (25) 13 (21)
Level of education: n (%)
No educational qualification 9 (15) 17 (27)
Some school qualification 16 (26) 16 (26)
High school and/or vocational qualification 24 (39) 15 (24)
University degree/professional qualification 12 (20) 14 (23)
Religion: n (%)
None 12 (20) 16 (26)
Christian 46 (75) 45 (73)
3 (5) 1 (2)
Social class: n (%)
Class 1 22 (36) 23 (37)
Class 2 15 (25) 12 (19)
Class 3 7 (12) 7 (11)
Class 4 6 (10) 2 (3)
Class 5 11 (18) 17 (27)
Psychiatric characteristics
HRSD score: M (SD) 5.62 (4.3) 5.76 (4.69)
BDI-II score: M (SD) 18.51 (10.91) 20.15 (12.86)
Depression diagnosis at intake: n (%)
In full remission 42 (69) 41 (66)
In partial remission 19 (31) 21 (34)
Previous episodes: M (SD) 6.43 (3.04) 6.35 (2.91)
Median 6 6
With 10 episodes: n (%) 23 (38) 19 (31)
No. of comorbid DSM–IV Axis I psychiatric diagnoses: M (SD) .83 (.96) 1.04 (1.11)
Age (in years) at first depression onset: M (SD) 26.34 (11.7) 26.11 (12.65)
Time (in months) since last depressive episode: M (SD) 24.20 (27.74) 18.68 (23.89)
Severity of last depressive episode (no. of DSM–IV symptoms
recorded): M (SD) 7.27 (1.3) 7.04 (1.35)
Attempted suicide: n (%) 20 (33) 22 (35)
No. of previous attempts: M (SD) 0.69 (1.37) 0.66 (1.05)
Range 0–7 0–4
Previous psychiatric treatment: n (%) 17 (28) 13 (21)
Quality of life
: M (SD)
Physical 22.64 (5.59) 23.0 (5.18)
Psychological 17.8 (3.82) 18.03 (3.63)
Social 9.52 (2.32) 9.27 (2.65)
Note. MBCT mindfulness-based cognitive therapy; m-ADM maintenance antidepressant medication;
HRSD Hamilton Rating Scale for Depression; BDI-II Beck Depression Inventory II; DSM–
IV Diagnostic and Statistical Manual of Mental Disorders (4th ed.).
The non-White participant was British Asian.
There was 1 Muslim in the m-ADM group and 1 Bahai, 1
Buddhist, and 1 Spiritualist in the MBCT group.
Social class was according to UK National Office of
National Statistics, and the range was from professional and managerial occupations (Class 1) to semiroutine and
routine occupations (Class 5). Data were missing for 1 case in the m-ADM arm of the trial.
Data determined
on the basis of the World Health Organization Quality of Life assessment (brief version).
Page 7
symptomatic). There was no evidence of an interaction with treat-
ment condition for severity of depression, Wald (1, N 123)
–1.50, p .13, hazard ratio 0.47 (95% CI: 0.17 to 1.26);
patients with and without residual symptoms did not respond
significantly differently to the two interventions.
To assess the possibility that differential discontinuation of
m-ADM affected the primary outcome findings, we undertook two
further sets of post hoc analyses on the primary outcome. First, we
compared participants in the MBCT arm who did and did not
taper/discontinue m-ADM on all demographic and psychiatric
characteristics (see Table 1). Only age of onset and severity of last
depressive episode were significantly different ( p .05), and
crucially this was in the direction that patients with earlier age of
onset and greater severity of last episode were more likely to taper
ADM. Second, we reran the survival analysis including only (a)
those patients who received MBCT and tapered/discontinued
m-ADM and (b) ADM patients who received an adequate dose of
m-ADM throughout the follow-up period. Cox regression showed
no difference between MBCT and m-ADM, Wald (1, N 89)
–1.27, p .21, hazard ratio 0.77 (95% CI: 0.49 to 1.16).
Outcome Analyses: Secondary Outcomes
Each of the secondary outcomes was assessed in turn, covarying
the stratification variable in all instances. The first secondary
outcome concerned the total number and qualitative nature of
depressive relapses/recurrences (duration, severity, and subjective
distress) across the MBCT and m-ADM arms. For the ITT anal-
yses, the two groups were not statistically different on any index of
relapse/recurrence: mean total number of relapses/recurrences:
m-ADM 1.57 (95% CI: 1.32 to 1.81), MBCT 1.45 (95% CI: 1.21
to 1.69), F(1, 66) 1, d 0.26; duration of relapses/recurrences
(in months): m-ADM 3.0 (95% CI: 2.1 to 3.9), MBCT 3.36 (95%
CI: 2.2 to 4.5), F(1, 66) 1, d 0.08; severity of relapses/
recurrences (DSM–IV severity specifier, 0 4): m-ADM 1.72 (95%
CI: 1.48 to 1.95), MBCT 1.79 (95% CI: 1.56 to 2.02), F(1, 66)
1, d 0.13; and subjective distress: m-ADM 62.56 (95% CI:
56.16 to 68.96), MBCT 59.65 (95% CI: 51.82 to 67.18), F(1,
66) 1, d 0.24. After rerunning these analyses for the PPT
sample, we found that the pattern of findings was unchanged, all
Fs (1, 52) 1.
Table 2 reports the means and inferential tests for the secondary
outcomes concerning residual depressive symptoms and quality of
life. The MBCT group reported significantly fewer residual de-
pressive symptoms across the five follow-ups compared with the
m-ADM group in both the ITT and PPT analyses. MBCT reported
better quality of life than did m-ADM in the physical and psycho-
logical domains in both the ITT and PPT analyses.
The fourth secondary outcome was psychiatric comorbidity. The
number of comorbid diagnoses at study end was significantly less
in the MBCT than m-ADM group for the ITT analysis (Mann–
Whitney U 1,332, p .05, d 0.43) and PPT analysis
(Mann–Whitney U 580, p .05, d 0.51), N 114 (ADM:
M 0.7, SD 1.01; and MBCT: M 0.34, SD 0.64).
Service Use, Productivity Losses, and Costs
There was little difference in the use of health and social
services between the two groups. Table 3 displays the total costs
per participant for the two groups. The most frequently accessed
services included UK National Health Service hospital outpatient
appointments (M 2.2 in the MBCT group vs. 3.2 in the m-ADM
group), primary care physician contacts (M 9 in the MBCT
group vs. 8 in the m-ADM group), and primary care nurse contacts
(M 3 in the MBCT group vs. 2 in the m-ADM group).
The total cost for an initial one-to-one session ($99) followed by
eight MBCT sessions and four follow-up sessions ($16.50 per
participant per session) was estimated to be $297 per participant.
In practice, some MBCT participants also received additional
one-to-one sessions, as well as support by telephone or e-mail,
giving a mean cost of $340 per participant (see Table 3). The cost
of antidepressant prescriptions was significantly lower in the
MBCT group than in the m-ADM group over the 15-month
follow-up period (M $172 vs. $275, respectively). There were
Figure 2. Survival (non–relapse/recurrence) curves comparing relapse/recurrence with major depression for
mindfulness-based cognitive therapy (MBCT) and antidepressant medication (ADM) groups over a 15-month
follow-up period.
Page 8
no significant differences between the two groups in any other cost
category. In total, the per-person cost for the MBCT group was
$457 more than that for the m-ADM group, but this difference was
not significant. Exploration of costs over time reveals that MBCT
is consistently more expensive than m-ADM over the first 12
months but that costs converge and MBCT becomes cheaper over
the final 3-month period (12 to 15 months).
Including all health and social services costs and productivity
losses, the incremental cost-effectiveness ratio was $962 per
relapse/recurrence prevented and $50 per depression-free day.
Table 2
Posttreatment and 15 Months Follow-up for MBCT and m-ADM Groups for Secondary Outcomes
Secondary outcomes
1 month posttreatment 15 months follow-up
Inferential between-groups test
(see data analysis)MBCT m-ADM MBCT m-ADM
Residual depressive symptoms
M 5.83 7.75 7.05 8.69
95% CI 4.49 to 7.3 5.86 to 9.34 5.53 to 8.74 6.64 to 10.5
ITT F(1, 116) 5.8, p .02,
PPT F(1, 98) 4.63, p .03,
M 13.12 17.47 12.61 17.02
95% CI 10.27 to 15.97 14.31 to 20.62 9.96 to 15.26 13.16 to 20.87
ITT F(1, 114) 2.52, p .12,
PPT F(1, 97) 4.0, p .04,
Quality of life
M 24.08 22.86 23.97 22.93
95% CI 22.62 to 25.53 21.34 to 24.39 22.63 to 25.30 21.18 to 24.69
ITT F(1, 114) 4.03, p .04,
PPT F(1, 98) 4.48, p .04,
M 18.88 17.47 18.61 17.36
95% CI 17.88 to 19.89 16.24 to 18.70 17.65 to 19.57 15.93 to 18.78
ITT F(1, 114) 6.23, p .01,
PPT F(1, 98) 6.37, p .01,
M 10.09 9.07 10.10 9.66
95% CI 9.55 to 10.64 8.37 to 9.77 9.53 to 10.68 8.88 to 10.44
ITT F(1, 112) 1.87, p .18,
PPT F(1, 98) 0.3, p .59,
Note. n 61 for MBCT; n 62 for m-ADM. MBCT mindfulness-based cognitive therapy; m-ADM maintenance antidepressant medication;
HRSD Hamilton Rating Scale for Depression; CI confidence interval; ITT intention to treat; PPT per protocol treatment;
the proportion
of effect and error variance that is attributed to the effect for the between-groups comparison; BDI-II Beck Depression Inventory II.
Complete data set for N 118 (i.e., 59 in each group).
Complete data set for N 117 (i.e., 59 in MBCT group and 58 in m-ADM group).
determined on the basis of the World Health Organization Quality of Life assessment (brief version; World Health Organization, 2004). Complete data set
for N 119 (i.e., 60 in MBCT group and 59 in m-ADM group).
Table 3
Total Cost Breakdown (in Dollars) per Participant
M (SD)
M (SD)
Mean difference of MBCT
vs. m-ADM (95% CI) p
MBCT 340 (58) 0 (0) 340 (325 to 355)
Antidepressants 172 (212) 275 (279) –103 (–191 to –14)
Hospital services 720 (1,245) 733 (1,983) –13 (–607 to 578)
Community health and social services 844 (1,091) 569 (580) 275 (–36 to 586)
Productivity losses 1,297 (2,707) 1,338 (4,343) –42 (–1,337 to 1,252)
Total cost over follow-up 3,370 (4,002) 2,915 (4,838) 457 (–1,130 to 2,043) .865
Total cost per year 2,767 (313) 2,340 (3,822) 427 (–852 to 1,705) .788
Note. MBCT mindfulness-based cognitive therapy; m-ADM maintenance antidepressant medication;
CI confidence interval.
Adjusted for stratification variable and baseline costs.
Page 9
For health care costs only, these ratios were $439 and $23,
The cost-effectiveness acceptability curve shown in Figure 3
demonstrates that if society’s willingness to pay for preventing an
additional relapse/recurrence is zero (society is unwilling to spend
any additional amount to prevent relapse/recurrence), then the
probability of MBCT being the more cost-effective option is 42%,
while the probability of the m-ADM being the more cost-effective
option is 58%. However, the probability of MBCT being the more
cost effective of the two options increases as willingness to pay
increases, suggesting that MBCT has a higher probability of being
more cost-effective than has m-ADM for willingness-to-pay levels
of approximately $1,000 and above.
In people with recurrent depression, MBCT produces compara-
ble outcomes to those for people using m-ADM in terms of
relapse/cost effectiveness and superior outcomes concerning resid-
ual depressive symptoms, psychiatric comorbidity, and the phys-
ical and psychological domains of quality of life. The reductions in
ADM usage in the MBCT group were substantial, and 75% of
patients in the MBCT arm completely discontinued their ADM.
MBCT provides a promising alternative approach to m-ADM, with
over 50% of people participating in MBCT staying well through
the follow-up period, compared with 40% in the m-ADM group.
Rates of adherence to MBCT were comparable to those in
previous trials (85%) and suggest the acceptability of this ap-
proach. However, in the recruitment to the study, the time com-
mitment and group aspect were noted as reasons to decline par-
ticipation. Moreover, 25% of the MBCT patients did not
discontinue their m-ADM, and those who experienced relapses/
recurrences were more likely to resume ADM. There is increasing
acknowledgement that depression is a recurrent disorder (Judd,
1997a), and management in primary care needs to develop pro-
phylactic approaches and routinely monitor for relapses. The fact
that MBCT did more to reduce residual symptoms than m-ADM
did is significant, as residual symptoms tend to predict relapse/
recurrence (Judd, 1997a) and interventions that target residual
symptoms tend to produce better outcomes over long-term follow-
ups (Fava, Rafanelli, Grandi, Canestrari, & Morphy, 1998; Paykel
et al., 1999). MBCT’s positive impact on comorbidity is likely to
have longer term benefits as patients have to contend with fewer
psychiatric symptoms. Similarly, patient quality of life is an es-
sential aspect of outcome assessment (Kuyken et al., 1995), and
the fact that MBCT produced additional gains in the physical and
psychological domains of life suggests that it may produce incre-
mental benefits in quality of life for some patients when compared
with m-ADM.
Cost-effectiveness analysis suggests that the additional cost of
MBCT may be justified in terms of improvements in the propor-
tion of patients who relapse— but only if willingness to pay for
such improvements is $1,000 or above. In terms of depression-free
days, the incremental cost effectiveness of MBCT is comparable to
that of similar studies, with a ratio of $50 per depression-free day
for total costs and $23 for health service costs. Recent estimates for
collaborative care programs include $33 in terms of total outpa-
tient costs (Liu et al., 2003) and $21 in terms of total inpatient and
outpatient costs (Simon et al., 2001). Estimates of $14–$24 have
been reported for a depression relapse prevention program (Simon
et al., 2002). Exploration of costs over time suggests that differ-
ences in cost converge and MBCT becomes cheaper than m-ADM
over the final 3-month period of the study. If this trend were to
continue, the relative cost effectiveness of MBCT may increase
over time. Future studies should consider longer follow-up periods
in order to test this hypothesis.
This study does not speak to the mechanism whereby MBCT is
efficacious. It is possible that MBCT cultivates greater awareness
that empowers people to step out of automatic modes of reacting
and respond more skillfully at times of potential relapse (Segal,
Williams, & Teasdale, 2002). However, it is also possible that the
Figure 3. Probability that mindfulness-based cognitive therapy (MBCT) would be more cost effective than
maintenance antidepressant medication (m-ADM) over the 15-month follow-up period for a range of levels of
willingness to pay for a unit reduction in the proportion of patients who relapse. K thousand.
Page 10
behavioral (action at times of potential relapse) and cognitive
(noting and responding to negative thinking) components of
MBCT are the active ingredients. This would be consistent with
recent evidence showing that cognitive and behavioral approaches
that target these mechanisms produce prophylactic effects compa-
rable to those for m-ADM (Bockting et al., 2005; Hollon et al.,
2005). It is also possible that nonspecific effects of attention and
group support are responsible for MBCT’s efficacy. Research
explicitly focused on studying mechanisms of change is needed to
examine these outstanding questions (Coelho et al., 2007; Lau-
renceau, Hayes, & Feldman, 2007). Observed differences between
MBCT and m-ADM in rates of comorbidity, residual depressive
symptoms, and quality of life provide promising clues to potential
mechanisms of action.
This study has several limitations. First, knowing that rates of
adherence to m-ADM tend to be poor (Cooper et al., 2007), in our
protocol we sought to maximize adherence in the m-ADM group
to ensure that MBCT was indeed being compared with an active
treatment. As such, our m-ADM group should be characterized as
an enhanced care group in that the study team actively encouraged
high levels of adherence at each follow-up. The findings suggest
that the enhanced care package was successful in ensuring high
rates of adherence to m-ADM. As such, the m-ADM management
in this trial was better than would be experienced in routine
primary care settings (cf. Katon et al., 2001). However, our checks
on MBCT therapists’ competence and adherence suggest that this
arm is probably also of a higher quality than would be expected in
routine health care settings. Second, our recruitment approach was
to screen all patients on ADM in primary care and then through
several steps establish if these individuals met the study’s inclu-
sion criteria and were willing to participate (White et al., 2007).
This enabled us to recruit our sample relatively rapidly but meant
that a large number of people proved not to be eligible or declined.
However, we argue that the sample represents people identified in
primary care as suffering depression and who at their physician’s
invitation were interested in pursuing a psychological approach to
relapse prevention. Nonetheless, the next phase of effectiveness
treatment outcome research should examine the question of
MBCT’s generalizability for different subpopulations and different
settings (e.g., referral source, primary vs. secondary care, patients
with multiple comorbidities). The time commitment of participa-
tion (in MBCT and the associated research) and the group aspect
of MBCT were given as reasons for declining, so research into
MBCT’s accessibility is required. Third, different relapse preven-
tion interventions with different populations produce different
absolute rates of depressive relapse (Geddes et al., 2003; Hollon et
al., 2002). The rates of relapse across both arms of the trial were
relatively high, most likely reflecting (a) our selection of a group
at particularly high risk (3 episodes) and (b) self-selection
among patients recognizing their risk and wanting help. We would
predict that relapse rates in this highly vulnerable group might be
reduced further through augmentation of ADM with MBCT, which
would be an obvious focus for a future trial. Fourth, this study does
not speak to the mechanisms through which MBCT operates. Fifth,
the economic evaluation excluded patient and family expenses and
the cost of informal care. For assessment of the true societal costs
of depression, future studies should consider including these costs.
Finally, the fact that a subset of MBCT participants chose not to
discontinue their m-ADM suggests that a future additive design
(MBCT plus m-ADM) may produce better outcomes, at least
among those patients not yet ready to discontinue m-ADM. Alter-
natively, a future design may need to provide greater or more
specialist support in tapering/discontinuing m-ADM.
MBCT has only recently been developed, and this is the first
evaluation of its efficacy against another active treatment and the
first evaluation of MBCT’s cost effectiveness. This study differs
most significantly from previous trials (Ma & Teasdale, 2004;
Teasdale et al., 2000) in that it recruited only people being treated
with the currently most commonly used prophylactic approach
(m-ADM) and supported people participating in MBCT to taper/
discontinue their medication. Future trials will be able to base
themselves on extant effect sizes and examine outstanding questions
such as external validity in increasingly real-world settings, the costs
of training MBCT therapists, the efficacy of augmenting m-ADM
treatment with MBCT, and comparisons with other psychosocial
approaches to relapse prevention, such as behavioral activation and
cognitive therapy (Dimidjian et al., 2006; Hollon et al., 2005).
American Psychiatric Association. (1994). Diagnostic and statistical man-
ual of mental disorders (4th ed.). Washington, DC: Author.
Baer, R. A. (2003). Mindfulness training as a clinical intervention: A
conceptual and empirical review. Clinical Psychology: Science and
Practice, 10, 125–143.
Baldwin, S. A., Murray, D. M., & Shadish, W. R. (2005). Empirically
supported treatments or Type I errors? Problems with the analysis of
data from group-administered treatments. Journal of Consulting and
Clinical Psychology, 73, 924 –935.
Barrett, B., Byford, S., & Knapp, M. (2005). Evidence of cost-effective
treatments for depression: A systematic review. Journal of Affective
Disorders, 84, 1–13.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive
therapy of depression. New York: Guilford Press.
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). The Beck Depression
Inventory: Second Edition. San Antonio, TX: Psychological Corporation.
Bockting, C. L. H., Schene, A. H., Spinhoven, P., Koeter, M. W. J.,
Wouters, L. F., Huyser, J., et al. (2005). Preventing relapse/recurrence in
recurrent depression with cognitive therapy: A randomized controlled
trial. Journal of Consulting and Clinical Psychology, 73, 647– 657.
Bower, P., Byford, S., Sibbald, B., Ward, E., King, M., Lloyd, R., et al.
(2000). Randomised controlled trial of non-directive counselling,
cognitive-behaviour therapy, and usual general practitioner care for
patients with depression: II. Cost effectiveness. British Medical Journal,
321, 1389 –1392.
British Medical Association & Royal Pharmaceutical Society of Great
Britain. (2006). British national formulary 48. London: British Medical
Journal Books/Pharmaceutical Press.
Byford, S., Knapp, M., Greenshields, J., Ukoumunne, O. C., Jones, V.,
Thompson, S., et al. (2003). Cost-effectiveness of brief cognitive behav-
iour therapy versus treatment as usual in recurrent deliberate self-harm: A
decision-making approach. Psychological Medicine, 33, 977–986.
Claxton, K. (1999). The irrelevance of inference: A decision-making
approach to the stochastic evaluation of health care technologies. Jour-
nal of Health Economics, 18, 341–364.
Coelho, H. F., Canter, P. H., & Ernst, E. (2007). Mindfulness-based cognitive
therapy: Evaluating current evidence and informing future research. Journal
of Consulting and Clinical Psychology, 75, 1000 –1005.
Cooper, C., Bebbington, P., King, M., Brugha, T., Meltzer, H., Bhugra, D.,
et al. (2007). Why people do not take their psychotropic drugs as
prescribed: Results of the 2000 National Psychiatric Morbidity Survey.
Acta Psychiatrica Scandinavica, 116, 47–53.
Page 11
Curtis, L., & Netten, A. (2006). Unit costs of health and social care.
Canterbury, England: Personal Social Services Research Unit.
Department of Health. (2006). National health service reference costs.
London: Author.
Dimidjian, S., Hollon, S. D., Dobson, K. S., Schmaling, K. B., Kohlenberg,
R. J., Addis, M. E., et al. (2006). Randomized trial of behavioral
activation, cognitive therapy, and antidepressant medication in the acute
treatment of adults with major depression. Journal of Consulting and
Clinical Psychology, 74, 658 670.
Efron, B., & Tibshirani, R. J. (1993). An introduction to the bootstrap. New
York: Chapman & Hall.
Fava, G. A., Rafanelli, C., Grandi, S., Canestrari, R., & Morphy, M. A.
(1998). Six-year outcome for cognitive behavioral treatment of residual
symptoms in major depression. American Journal of Psychiatry, 155,
Fenwick, E., Claxton, K., & Sculpher, M. (2001). Representing uncer-
tainty: The role of cost-effectiveness acceptability curves. Health Eco-
nomics, 10, 779–787.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1995).
Structured Clinical Interview for DSM-IV Axis I disorder with psychotic
screen. New York: New York Psychiatric Institute.
Frank, E., Kupfer, D. J., Perel, J. M., Cornes, C., Jarrett, D. B., Mallinger,
A. G., et al. (1990). 3-year outcomes for maintenance therapies in
recurrent depression. Archives of General Psychiatry, 47, 1093–1099.
Geddes, J. R., Carney, S. M., Davies, C., Furukawa, T. A., Kupfer, D. J.,
Frank, E., et al. (2003). Relapse prevention with antidepressant drug
treatment in depressive disorders: A systematic review. Lancet, 361,
653– 661.
Harper, A., & Power, M. (1998). Development of the World Health
Organization WHOQOL-BREF quality of life assessment. Psychologi-
cal Medicine, 28, 551–558.
Hirschfield, R. M. A., Keller, M. B., Panico, S., Arons, B. S., Barlow, D.,
Davidoff, F., et al. (1997). The National Depressive and Manic-
Depressive Association consensus statement on the undertreatment of
depression. Journal of the American Medical Association, 277, 333–340.
Hollon, S. D., DeRubeis, R. J., Shelton, R. C., Amsterdam, J. D., Salomon,
R. M., O’Reardon, J. P., et al. (2005). Prevention of relapse following
cognitive therapy vs medications in moderate to severe depression.
Archives of General Psychiatry, 62, 417– 422.
Hollon, S. D., Munoz, R. F., Barlow, D. H., Beardslee, W. R., Bell, C. C.,
Bernal, G., et al. (2002). Psychosocial intervention development for the
prevention and treatment of depression: Promoting innovation and in-
creasing access. Biological Psychiatry, 52, 610 630.
Judd, L. L. (1997a). Prevalence, correlates, and course of minor depression
and major depression in the National Comorbidity Survey: Discussion.
Journal of Affective Disorders, 45, 28 –29.
Judd, L. L. (1997b). The clinical course of unipolar major depressive
disorders. Archives of General Psychiatry, 54, 989 –991.
Judd, L. L., Paulus, M. P., Zeller, P., Fava, G. A., Rafanelli, C., Grandi, S.,
et al. (1999). The role of residual subthreshold depressive symptoms in
early episode relapse in unipolar major depressive disorder. Archives of
General Psychiatry, 56, 764 –765.
Kabat-Zinn, J. (1990). Full catastrophe living: How to cope with stress,
pain and illness using mindfulness meditation. New York: Delacorte.
Katon, W., Rutter, C., Ludman, E. J., Von Korff, M., Lin, E., Simon, G.,
et al. (2001). A randomized trial of relapse prevention of depression in
primary care. Archives of General Psychiatry, 58, 241–247.
Katon, W., & Schulberg, H. (1992). Epidemiology of depression in pri-
mary care. General Hospital Psychiatry, 14, 237–247.
Keller, M. B., Klerman, G. L., Lavori, P. W., Coryell, W., Endicott, J., &
Taylor, J. (1984). Long-term outcome of episodes of major depression:
Clinical and public-health significance. Journal of the American Medical
Association, 252, 788–792.
Koopmanschap, M. A., & Rutten, F. F. H. (1996). A practical guide for
calculating indirect costs of disease. Pharmacoeconomics, 10, 460 466.
Kupfer, D. J., Frank, E., Perel, J. M., Cornes, C., Mallinger, A. G., Thase,
M. E., et al. (1992). 5-year outcome for maintenance therapies in
recurrent depression. Archives of General Psychiatry, 49, 769–773.
Kuyken, W., Orley, J., Power, M., Herrman, H., Schofield, H., Murphy, B.,
et al. (1995). The World Health Organization Quality of Life Assess-
ment (WHOQOL): Position paper from the World Health Organization.
Social Science & Medicine, 41, 1403–1409.
Lau, M. A., Segal, Z. V., & Williams, J. M. (2004). Teasdale’s differential
activation hypothesis: Implications for mechanisms of depressive re-
lapse and suicidal behaviour. Behaviour Research and Therapy, 42,
Laurenceau, J. P., Hayes, A. M., & Feldman, G. C. (2007). Some meth-
odological and statistical issues in the study of change processes in
psychotherapy. Clinical Psychology Review, 27, 682– 695.
Liu, C. F., Hedrick, S. C., Caheny, E. F., Heagerty, P., Felker, B.,
Hasenberg, N., et al. (2003). Cost-effectiveness of collaborative care for
depression in a primary care veteran population. Psychiatric Services,
54, 698 –704.
Ma, S. H., & Teasdale, J. D. (2004). Mindfulness-based cognitive therapy
for depression: Replication and exploration of differential relapse pre-
vention effects. Journal of Consulting and Clinical Psychology, 72,
31– 40.
Morisky, D. E., Green, L. W., & Levine, D. M. (1986). Concurrent and
predictive validity of a self-reported measure of medication adherence.
Medical Care, 24, 67–74.
National Institute for Clinical Excellence. (2004). Depression: Manage-
ment of depression in primary and secondary care (Clinical Guideline
No. 23). www.nice.org.uk/CG023NICEguideline
Olfson, M., Marcus, S. C., Tedeschi, M., & Wan, G. J. (2006). Continuity
of antidepressant treatment for adults with depression in the United
States. American Journal of Psychiatry, 163, 101–108.
Paykel, E. S., Scott, J., Teasdale, J. D., Johnson, A. L., Garland, A., Moore,
R., et al. (1999). Prevention of relapse in residual depression by cogni-
tive therapy: A controlled trial. Archives of General Psychiatry, 56,
829 835.
Prien, R. F., & Kupfer, D. J. (1986). Continuation drug therapy for major
depressive episodes: How long should it be maintained? American
Journal of Psychiatry, 143, 18–23.
Segal, Z. V., Teasdale, J. D., Williams, J. M., & Gemar, M. C. (2002). The
Mindfulness-Based Cognitive Therapy Adherence Scale: Inter-rater re-
liability, adherence to protocol and treatment distinctiveness. Clinical
Psychology & Psychotherapy, 9, 131–138.
Segal, Z. V., Williams, J. M. G., & Teasdale, J. D. (2002). Mindfulness-
based cognitive therapy for depression: A new approach to preventing
relapse. New York: Guilford Press.
Simon, G. E., Katon, W. J., Von Korff, M., Unutzer, J., Lin, E. H. B.,
Walker, E. A., et al. (2001). Cost-effectiveness of a collaborative care
program for primary care patients with persistent depression. American
Journal of Psychiatry, 158, 1638–1644.
Simon, G. E., Von Korff, M., Ludman, E. J., Katon, W. J., Rutter, C.,
Unutzer, J., et al. (2002). Cost-effectiveness of a program to prevent
depression relapse in primary care. Medical Care, 40, 941–950.
Teasdale, J. D., Segal, Z. V., Williams, J. M. G., Ridgeway, V. A., Soulsby,
J. M., & Lau, M. A. (2000). Prevention of relapse/recurrence in major
depression by mindfulness-based cognitive therapy. Journal of Consult-
ing and Clinical Psychology, 68, 615– 623.
Thompson, S. G., & Barber, J. A. (2000). How should cost data in
pragmatic randomised trials be analysed? British Medical Journal, 320,
Vanhout, B. A., Al, M. J., Gordon, G. S., & Ruten, F. F. H. (1994). Costs,
effects and C/E ratios alongside a clinical trial. Health Economics, 3,
309 –319.
Page 12
van Schaik, D. J. F., Klein, A. F. J., van Hout, H. P. J., van Marwijk,
H. W. J., Beekman, A. T. F., de Haan, M., et al. (2004). Patients’
preferences in the treatment of depressive disorder in primary care.
Psychiatry and Primary Care, 26, 184 –189.
White, K., Holden, E. R., Byng, R., Mullan, E., & Kuyken, W. (2007).
Under/over-recruitment to mental health trials. Acta Psychiatrica Scan-
dinavica, 116, 158.
Williams, J. B. (1988). A structured interview guide for the Hamilton
Depression Rating Scale. Archives of General Psychiatry, 45, 742–747.
Williams, J. M. G., Teasdale, J. D., Segal, Z. V., & Kabat-Zinn, J. (2007).
The mindful way through depression: Freeing yourself from chronic
unhappiness. New York: Guilford Press.
World Health Organization. (2004). The World Health Organization
Quality of Life (WHOQOL)-BREF [Questionnaire]. Retrieved May 7,
2008, from http://www.who.int/substance_abuse/research_tools/en/
Zimmerman, M., McGlinchey, J. B., Posternak, M. A., Friedman, M.,
Attiullah, N., & Boerescu, D. (2006). How should remission from
depression be defined? The depressed patient’s perspective. American
Journal of Psychiatry, 163, 148–150.
Received January 8, 2008
Revision received May 7, 2008
Accepted June 16, 2008
New Editors Appointed, 2010 –2015
The Publications and Communications Board of the American Psychological Association an-
nounces the appointment of 4 new editors for 6-year terms beginning in 2010. As of January 1,
2009, manuscripts should be directed as follows:
Psychological Assessment (http://www.apa.org/journals/pas), Cecil R. Reynolds, PhD, De-
partment of Educational Psychology, Texas A&M University, 704 Harrington Education
Center, College Station, TX 77843.
Journal of Family Psychology (http://www.apa.org/journals/fam), Nadine Kaslow, PhD,
Department of Psychiatry and Behavioral Sciences, Grady Health System, 80 Jesse Hill Jr.
Drive, SE, Atlanta, GA 30303.
Journal of Experimental Psychology: Animal Behavior Processes (http://www.apa.org/
journals/xan), Anthony Dickinson, PhD, Department of Experimental Psychology, University
of Cambridge, Downing Street, Cambridge CB2 3EB, United Kingdom
Journal of Personality and Social Psychology: Personality Processes and Individual Differ-
ences (http://www.apa.org/journals/psp), Laura A. King, PhD, Department of Psychological
Sciences, University of Missouri, McAlester Hall, Columbia, MO 65211.
Electronic manuscript submission: As of January 1, 2009, manuscripts should be submitted
electronically via the journal’s Manuscript Submission Portal (see the website listed above with
each journal title).
Manuscript submission patterns make the precise date of completion of the 2009 volumes
uncertain. Current editors, Milton E. Strauss, PhD, Anne E. Kazak, PhD, Nicholas Mackintosh,
PhD, and Charles S. Carver, PhD, will receive and consider manuscripts through December 31,
2008. Should 2009 volumes be completed before that date, manuscripts will be redirected to the new
editors for consideration in 2010 volumes.
Page 13