Journal of Affective Disorders 130 (2011) 138–144

Contents lists available at ScienceDirect

Journal of Affective Disorders

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j a d

Research report

Mindfulness-based cognitive therapy vs cognitive behaviour therapy as a

treatment for non-melancholic depression
Vijaya Manicavasgar ⁎, Gordon Parker, Tania Perich
Black Dog Institute, Hospital Rd, Randwick, NSW, 2031, Australia
School of Psychiatry, University of NSW, Randwick, NSW, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Aim: To examine the comparative effectiveness of Mindfulness-Based Cognitive Therapy
Received 19 February 2010 (MBCT) and Cognitive Behaviour Therapy (CBT) as treatments for non-melancholic depression.
Received in revised form 9 September 2010 Method: Participants who met criteria for a current episode of major depressive disorder were
Accepted 9 September 2010
randomly assigned to either an 8-week MBCT (n = 19) or CBT (n = 26) group therapy
Available online 20 November 2010
condition. They were assessed at pre-treatment, 8-week post-group, and 6- and 12-month
follow-ups.
Keywords: Results: There were significant improvements in pre- to post-group depression and anxiety
Mindfulness-based cognitive therapy scores in both treatment conditions and no significant differences between the two treatment
Depression
conditions. However, significant differences were found when participants in the two
Cognitive behaviour therapy
treatment conditions were dichotomized into those with a history of four or more episodes
Mindfulness
of depression vs those with less than four. In the CBT condition, participants with four or more
previous episodes of depression demonstrated greater improvements in depression than those
with less than four previous episodes. No such differences were found in the MBCT treatment
condition. No significant differences in depression or anxiety were found between the two
treatment conditions at 6- and 12-month follow-ups.
Limitations: Small sample sizes in each treatment condition, especially at follow-up.
Conclusions: MBCT appears to be as effective as CBT in the treatment of current depression.
However, CBT participants with four or more previous episodes of depression derived greater
benefits at 8-week post-treatment than those with less than four episodes.
© 2010 Published by Elsevier B.V.

1. Introduction and automatic thoughts, as well as understanding maladap-

tive assumptions and beliefs judged as reflecting enduring
Of all the psychological therapies, Cognitive Behaviour cognitive structures or schemas (Beck, 2005). CBT may be
Therapy (CBT) has become the treatment of choice for conducted with individual clients or by using a group-based
patients with depression with numerous efficacy and effec- format. Group-based CBT approaches have been found to be
tiveness studies attesting to its benefits (Beck, 2005). The cost effective and therapeutically equivalent to individual
goals of most CBT approaches for depression have focused on face-to-face approaches, thus further contributing to the
instructing patients about a cognitive approach to under- popularity of CBT as an acute treatment for depression (Oei
standing the aetiology and maintenance of mood disorders, and Dingle, 2008; Tucker and Oei, 2007).
applying specific skills to identify and modify dysfunctional Mindfulness-Based Cognitive Therapy (MBCT) was spe-
cifically developed to prevent relapse amongst those in
remission from depression and for those with a chronic,
⁎ Corresponding author. Black Dog Institute, Hospital Rd, Randwick NSW
unremitting depressive history (Segal et al., 2002a). MBCT
2031, Australia. Tel.: + 61 2 9382 4320; fax: + 61 2 9382 3712. comprises an 8-week group therapy program with sessions
E-mail address: v.manicacvasgar@unsw.edu.au (V. Manicavasgar). (each lasting approximately 2 to 2.5 h) teaching a ‘mindful’

0165-0327/$ – see front matter © 2010 Published by Elsevier B.V.

doi:10.1016/j.jad.2010.09.027
 V. Manicavasgar et al. / Journal of Affective Disorders 130 (2011) 138–144 139

approach to thoughts and feelings, characterized by non- The number of prior episodes of depression has also been
judgmental awareness of internal experience, including a found to influence CBT outcome for those currently depressed
significant meditation component (Segal et al., 2002a). It (Conradi et al., 2008). These authors conducted a 2-year
combines elements of cognitive therapy, such as psychoedu- follow-up of participants diagnosed with depression who
cation about depression, within this framework. Mindfulness were randomly allocated to CBT plus psychoeducation,
is thought to assist in distancing from automatic and self- psychoeducation only or TAU. Individuals who received CBT
perpetuating negative thoughts which characterize depres- plus psychoeducation demonstrated the greatest improve-
sogenic thinking (Segal et al., 2002a). In contrast to ments in depression. In addition, those who received CBT plus
conventional CBT—which encourages patients to examine psychoeducation and had 4 or more previous episodes of
the accuracy of their beliefs—MBCT encourages patients to depression returned significantly lower depression scores
recognize the occurrence of dysfunctional thoughts without than those with less than 4 prior episodes at 2-year follow-up
emotionally responding to them. (Conradi et al., 2008).
Three studies have evaluated the effectiveness of MBCT in No studies have, as yet, compared MBCT to another
reducing relapse and recurrence in major depression. In the psychological treatment. Given that CBT is widely researched
first study, Teasdale et al. (2000) randomly allocated 145 in treating current depression, we have compared MBCT to
patients who had recovered from recurrent depression to a CBT to address this issue. We also sought to determine
treatment as usual (TAU) or an MBCT condition. Their results whether there were any differences in treatment outcome for
indicated that, compared to TAU, MBCT significantly reduced those with fewer (less than four) and multiple (four or more)
the risk of relapse within a 60 week period in patients with a previous episodes of depression as previously reported by
history of three or more previous episodes of depression (66% Conradi et al. (2008). Less than four vs four or more prior
vs 37%). These findings were replicated in a later study by Ma episodes of depression was determined a priori in accordance
and Teasdale (2004) who found that those with two prior with the findings of Conradi et al. (2008) where an effect was
episodes had a relapse rate of 50% compared to 33% for those seen on treatment outcome for those with current depression
with three episodes, and 38% for those with four or more receiving CBT. The impact of previous episodes of depression
episodes. However, Michalak et al. (2008) reported that on MBCT treatment outcome for those currently depressed
relapse rates were reduced (to 37.5%) following treatment was also explored.
with MBCT at a 12-month follow-up, while the number of We hypothesized that both MBCT and CBT would be
previous episodes of depression did not predict relapse rates. effective in reducing depression amongst those experiencing
In addition to preventing depressive relapse, other studies a current episode. Furthermore, based on prior research, we
using MBCT have noted improvements for those experiencing hypothesized that both MBCT and CBT would be equally
current or residual depression (Eisendrath et al., 2008; effective in reducing depression in those with a history of four
Barnhofer et al., 2009; Kenny and Williams, 2007; Kingston or more previous episodes.
et al., 2007). Kenny and Williams (2007) investigated the
effects of MBCT in 46 participants with treatment resistant 2. Method
depression (categorized as those who had undertaken
previous treatment—either current medication or previous 2.1. Participants
CBT within the past 10 years, had 3 or more past episodes of
depression and who were still currently symptomatic). Those Participants were recruited from the general community
who undertook the MBCT program demonstrated significant through referral by general practitioners (GPs) and commu-
reductions in mean depression scores following treatment. nity advertisements (local newspapers and locally distributed
These authors also found that participants allocated to the flyers). Inclusion criteria were: (a) aged 18 years or over,
‘severe’ category on the Beck Depression Inventory (BDI) (b) meeting DSM-IV criteria for major depressive disorder on
demonstrated greater pre- to post-treatment change com- the computerized version of the Composite International
pared to those allocated to the ‘moderate’ category (Kenny Diagnostic Interview (CIDI-AUTO) (described below) (WHO,
and Williams, 2007). A randomized-controlled trial compar- 1997), (c) scoring 20 or more on the BDI-II (Beck Depression
ing MBCT to TAU with 14 participants in each condition found Inventory II) state depression measure (Beck et al., 1996) at
that MBCT was effective in treating current depression telephone screening in order to establish probable caseness
(Barnhofer et al., 2009). Participants who had completed an for current depression, (d) reporting low mood for at least
MBCT group showed a significant decrease in mean Beck three preceding months, (e) being proficient in English,
Depression Inventory II (BDI-II) scores compared to TAU, (f) not having engaged in CBT, mindfulness or meditation/
with scores dropping to the moderate range. These authors relaxation (operationalized as more than four sessions of
also found that fewer participants in the MBCT condition met regular meditation/relaxation) over the preceding 12 months,
diagnostic criteria for depression after treatment (Barnhofer (g) being under supervision of a case manager/clinician,
et al., 2009). Eisendrath et al. (2008) examined the effects of (h) not commencing antidepressant medication or, if medi-
MBCT on 55 participants with treatment resistant depression cated, not changing their antidepressant medication regime
and also noted reductions in depression scores following an over the preceding three months, and (i) preparedness to
MBCT program. Mean BDI-II scores at baseline in the commit to an 8-week group program.
‘moderate’ to ‘severe’ range dropped to the ‘mild’ to Exclusion criteria were: (a) a current diagnosis of melan-
‘moderate’ range post-treatment. These studies suggest that, cholic depression or bipolar disorder, (b) a history of any
in addition to reducing relapse to depression, MBCT may be psychotic illness, (c) dementia, (d) current active suicidal
used as an active treatment for current depression. ideation, (e) being hospitalized, (f) concurrent treatment using
140 V. Manicavasgar et al. / Journal of Affective Disorders 130 (2011) 138–144

meditation or CBT, (g) drug/alcohol dependence, (h) daytime from 0 to 13 for ‘minimal’, from 14 to 19 for ‘mild’, from 20
anxiolytic medication (which could potentially impair concen- to 28 for ‘moderate’ and from 29 to 63 for ‘severe’
tration) (Choy, 2007), (i) current antenatal or postnatal depression. The BDI-II has high internal reliability
depression (which could be related to hormonal factors) (α = 0.91) and good convergent validity with the BDI-I
(Bloch et al., 2000), (j) currently in receipt of antipsychotic or (0.93) (Dozois et al., 1998).
mood stabilizing medication, and (k) a history of treatment 4. Beck Anxiety Inventory (BAI) (Beck et al., 1988)—a 21-
with more than two antidepressant drugs. item measure assessing severity of anxiety symptoms.
Informed consent was obtained from all participants prior Items are scored on a 4-point scale listing common anxiety
to their involvement in treatment and they were not symptoms from 0 = ‘not at all’ to 3 = ‘severely’. Scores
reimbursed for their participation in the study. Ethics range from 0 to 21 for ‘low’, from 22 to 35 for ‘moderate’,
approval was given by the University of NSW Human while scores over 35 indicate ‘high’ anxiety. The BAI has
Research Ethics Committee (HREC 05269). high internal consistency (α = 0.94) (Fydrich et al., 1992).

2.2. Procedure 2.3.2. Interview schedules

1. Composite International Diagnostic Interview (CIDI)
Initial assessment included a telephone screen using the (WHO, 1997). The anxiety and depression modules of
BDI-II (Beck, Brown, and Steer, 1996). Participants who met the CIDI-AUTO, a computerized version of the CIDI, were
BDI-II criteria for depression caseness (a score of 20 or more) administered at interview by a research assistant who was
were provisionally accepted into the trial. trained on the CIDI to assign DSM-IV and ICD-10 diagnoses
Participants were then interviewed by a research psychol- (Peters and Andrews, 1995). The CIDI-AUTO has moderate
ogist using the Composite International Diagnostic Interview validity for the depression and anxiety disorder modules
(CIDI) (WHO, 1997) for diagnostic clarification and to exclude (kappa = 0.40) against clinician-rated diagnoses (Peters
melancholic depression and bipolar disorder, and to establish and Andrews, 1995).
the number of prior episodes of depression. The Social and 2. Social and Occupational Functioning Scale (SOFAS) (Goldman
Occupational Functioning Scale (SOFAS) (Goldman et al., 1992) et al., 1992). The SOFAS assesses DSM-IV Axis V of functioning
was administered to assess level of psychosocial impairment. and impairment across the domains of work, social function-
Participants were also required to complete a booklet of ing and relationships.
self-report measures, including re-administration of the BDI-
II within 2 weeks prior to being assigned to their treatment 2.4. Group treatments
condition. Participants were then assigned to one of two
group conditions. Eight of the 11 group programs were All group sessions were conducted by two therapists, the
randomly allocated to either the CBT or MBCT conditions. primary therapist always being an experienced clinical
Random allocation was conducted via a computer program by psychologist. Group co-facilitators were medical or psychol-
a researcher who was not involved with the research study. ogy students. One clinical psychologist was always the
The three non-random group programs were run subject to primary therapist for the MBCT groups, while two clinical
therapist availability at that time. Participants that were not psychologists alternated as the primary therapist for the CBT
randomly assigned were sequentially enrolled in one of the groups. The primary therapist for the MBCT groups had
two treatment conditions until the group program was filled. 10 years prior experience in using mindfulness in clinical
All participants were unaware of their group allocation until group settings while the therapists for the CBT conditions had
the first therapy session. previous experience and training in conducting CBT in group
On completion of the group programs, participants settings. The primary therapists for both conditions had over
completed another set of self-report questionnaires and 20 years experience each working professionally as clinical
were again rated on the SOFAS (Goldman et al., 1992). psychologists. Training in MBCT was unavailable in Australia
These assessments were completed within 2 weeks of the at the time of the commencement of this study.
completion of the group therapy programs. All group sessions were conducted at the Black Dog
At 6- and 12-month follow-ups all self-report measures Institute, Prince of Wales Hospital Campus, NSW, Australia.
were re-administered. Each group comprised of 6 to 8 participants. In total, eleven
group therapy sessions were run—five for the MBCT and six
2.3. Measures for the CBT conditions. Each group session lasted for 2 to 2.5 h
for both the MBCT and CBT conditions.
2.3.1. Self-report questionnaires
1. Demographic questions (e.g. gender, age and marital status). 2.4.1. MBCT
2. Expectancy of improvement—a single question requiring Treatment comprised a modified version of the 8-week
participants to rate their expected improvement in the MBCT course developed by Segal et al. (2002a). Modifications
first group session immediately following disclosure of to the program included the removal of the yoga component,
their actual treatment condition. This item was scored on a optional rather than compulsory purchase of the book ‘Full
5-point scale where 1 = ‘expecting mood to improve’ and Catastrophe Living’ (Kabat-Zinn, 1990) and the omission of
5 = ‘expecting mood to get worse’. the DVD-based ‘Mindfulness-Based Stress Reduction’ pro-
3. Beck Depression Inventory II (BDI-II) (Beck, Brown, and gram (Segal et al., 2002a). The yoga component was removed
Steer, 1996)—a 21-item measure assessing severity of as the therapists did not have certification for teaching yoga
depression symptoms on a scale of 0 to 3. Scores range and would have been unable to advise on issues that may
 V. Manicavasgar et al. / Journal of Affective Disorders 130 (2011) 138–144 141

have resulted from this aspect of the program. The DVD was baseline assessment, and 69 participants were enrolled in the
removed as it was unavailable for use in Australia at the time group therapy programs (26 participants dropped out prior to
of study commencement. randomization).
Homework requirements were the same as those de- Thirty-nine participants were allocated to the CBT and 30 to
scribed in the Segal et al. (2002a) MBCT program, with the the MBCT condition. Sixteen participants (23%) dropped out
MBSR program CD set, ‘Guided Mindfulness Meditation’, during the treatment study (4 from the MBCT group and 12
being provided to each participant. An exception to the from the CBT group program). Eight participants scored less
standard program occurred during the week where the yoga than 20 (suggesting mild rather than moderate depression) on
CD was prescribed as homework. In this week, participants the BDI-II at the pre-group assessment point and were excluded
were required to alternate between a 40-minute body scan from the analysis. Overall, 45 participants completed the group
and the 40-minute sitting meditation CDs. therapy programs (CBT = 26 and MBCT = 19).
Random allocation of participants to either the CBT or MBCT
2.4.2. CBT condition was possible for 8 of the 11 group programs as, for
An 8-week program was developed by the investigators 3 groups (CBT= 2 and MBCT = 1), therapists were available
based on standard CBT strategies outlined by Beck et al. only for one of the treatment conditions. Participants were
(1979). The program included teaching participants to considered non-completers if they failed to attend two
identify links between thoughts, feelings and behaviours, consecutive group sessions. Participant numbers ranged from
identify cognitive distortions, challenge negative thoughts 6 to 8 members per group for each group. An independent
and to use specific strategies such as ‘behavioural experi- samples t-test revealed that there was no significant difference
ments’ to address worry and avoidance behaviours. Partici- between those that were randomized to a treatment condition
pants were required to complete daily homework tasks and those that were not on baseline BDI or BAI scores.
throughout the program.
3.1. Program completers
2.5. Treatment fidelity
Demographic characteristics of those that completed the
All group therapy sessions were audio-taped in order to trial and were included in the analysis are listed in Table 1.
assess treatment fidelity. Treatment fidelity was assessed Chi-square analyses indicated that there were no significant
using the Mindfulness-Based Cognitive Therapy (MBCT-AS) differences between the two conditions on any of the baseline
and the Cognitive Behaviour Therapy Adherence Scales (CBT- demographic variables. There was no significant difference
AS) (Segal et al., 2002b). These two scales assess the extent to found between the two conditions on number of prior
which diverse psychological interventions reflect the key episodes of depression.
constructs of MBCT and CBT. Each scale has high reliability Mean age of completers for the CBT condition was 45 (SD
(both at 0.8) (Segal et al., 2002b). The item referring to the 12.94) years and for the MBCT condition 47 (SD 13.84) years.
use of video material in the MBCT program was removed from There was no significant difference in mean age or baseline
the MBCT-AS. depression and anxiety scores between participants in the
A trained clinical psychologist with Masters level training two treatment conditions. Participants in either condition did
and familiarity with mindfulness-based approaches who was not differ on the Expectancy for Improvement Questionnaire
blind to group assignment rated the audiotapes. Audiotapes (MBCT: M = 2.22, SD 0.81; CBT: M = 2.61, SD 0.99) nor on
were randomly selected, with one tape selected from each number of prior episodes of depression.
session of each group program to create a total of 16 (8 MBCT Eighteen (40%) participants were unmedicated prior to
tapes and 8 CBT) tapes that were assessed for treatment the trial, while 25 (56%) had been on stable medication for at
fidelity. Mean scores on the MBCT-AS were 19.75 (SD 3.41) least three months prior to trial commencement. Information
for the MBCT condition and 12.38 (SD 2.33) for the CBT on antidepressant use was unclear for 3 participants. In the
condition (p b 0.001). Mean scores on the CBT-AS were 28.13 CBT condition, 9 (35%) participants were unmedicated while
(SD 2.80) for the CBT condition and 9.76 (SD 2.82) for the 9 (53%) participants were unmedicated in the MBCT
MBCT condition (p b 0.001) indicating that each of the
conditions was sufficiently distinct from each other on
those relevant constructs. Table 1
Baseline Demographics.
2.6. Data analysis
CBT MBCT

Data was analysed using Statistical Package for the Social N % N %

Sciences (SPSS) (SPSS Inc, 2009) and SOLAS (Statistical Gender
Solutions Ltd., 2001). Male 9 34 7 37
Female 17 65 12 63
Marital status
3. Results
Married 7 27 8 42
Widowed, separated/divorced 11 42 4 21
One hundred and sixty three people were screened for the Never married 8 31 7 37
study. Fifty six participants did not meet study criteria and an Employment status
additional 12 participants withdrew participation prior to the Working or full time student 15 58 12 63
Unemployed or retired 11 42 7 37
interview procedure. Ninety-five participants completed the
142 V. Manicavasgar et al. / Journal of Affective Disorders 130 (2011) 138–144

condition. Seventeen (65%) participants in the CBT condition Table 3

were medicated while 8 (47%) participants in the MBCT Pooled regression estimates of BDI and BAI at follow-up time points.

condition were medicated. There was no significant differ- Measure Mean estimate SE t df P
ence between conditions on medication use.
BDI-II
Post vs 6-month − 1.78 4.70 − 0.38 21.51 0.71
Post vs 12-month − 6.06 4.46 − 1.36 16.00 0.19
3.2. Attrition 6-month vs 12-month − 4.27 3.72 − 1.15 19.47 0.26

BAI
Overall, 14 participants (23%) dropped out of the study,
Post vs 6-month − 3.90 3.31 − 1.18 21.32 0.25
with a trend for more drop-outs from the CBT than for the Post vs 12-month − 4.82 3.37 − 1.43 14.60 0.17
MBCT condition (i.e. 10 vs 4) although this difference was not 6-month vs 12-month − 0.91 3.02 − 0.30 12.10 0.77
statistically significant. Of the 14 that dropped out, 7 (50%)
were female and 7 (50%) male; five (36%) were married, 3
(21%) separated or divorced, 6 (43%) were never married; ten
(71%) were employed or full time students, 4 (29%) were A post-hoc ANOVA was conducted on BDI and BAI scores
unemployed. for both CBT and MBCT conditions. Participants in the
CBT group who had four or more previous episodes of
depression demonstrated greater improvements in depres-
3.3. Primary treatment comparisons sion compared to those with less than four episodes of
depression (F(1,23) = 6.02, p b 0.05). There were no signifi-
A MANOVA was conducted on pre- and post-group BDI, cant differences in participants with less than or greater than
BAI and SOFAS scores. Overall, mean BDI (Pillai's trace = 0.41, four previous episodes of depression for the MBCT condition.
F = 8.26, p b 0.001) and BAI (Pillai's trace = 0.24, F = 3.72, The relationship of SOFAS scores to previous was not
p b 0.05) scores significantly decreased from baseline to post- explored as there was no significant difference between
treatment for both treatment conditions. Recommended cut- pre- and post-treatment scores on this measure.
off scores for the BDI and BAI indicated that group scores on
both measures decreased from the ‘severe’ to the ‘mild to
moderate’ range (see Table 2). There was no significant 3.4. Follow-up analyses
difference between the SOFAS scores before and after group
treatment for either condition and no significant difference Six-month and twelve-month data were analysed using a
between those with more than four or less than four prior repeated measures MANOVA, with pre-group scores entered as
episodes of depression. a co-variate. Data substitution for missing time points was
There were no significant differences between pre-treat- conducted using multiple imputation via Rubin's (1987)
ment BDI and BAI scores for participants with four or more or Bayesian Least Squares approach, which has been found by
those with less than four previous episodes of depression. Barnes et al. (2006) to perform well in a small sample clinical
However, there was a significant interaction found between trial simulation. Small sample size correction for multiple
post-treatment mean scores on the BAI, BDI, treatment imputation was conducted in accordance with Barnard and
condition and number of previous episodes of depression Rubin's (1999) recommendations. Analyses were not con-
(Pillai's trace = 0.22, F = 3.35, p b 0.05). Tests of between- ducted on prior episodes of depression due to the small sample
subjects effects indicated that the interaction between treat- size. Results of the mean estimates are presented in Table 3.
ment condition and no. of prior episodes was significant for There was no significant difference found between the
the BDI (F(1,38) = 6.37, p b 0.05), BAI (F(1,38) = 7.05, p b 0.05) post-group, 6-month and 12-month time points on the BDI-II
and SOFAS (F(1,38) = 5.06, p b 0.05). or the BAI. There were no significant differences between

Table 2
Depression and anxiety mean scores pre- and post-group treatment.

N CBT N MBCT

Pre Post Pre Post

Mean SD Mean SD Mean SD Mean SD

BDI-II 26 36.23 11.11 23.62 16.83 19 32.42 9.01 21.21 13.83

Episodes of depression
Less than 4 16 36.19 10.35 28.50 16.42 11 30.45 10.17 17.55 13.64
4 or more 10 36.30 12.82 15.80 15.06 8 35.13 6.83 26.25 13.26
BAI 26 23.96 16.09 16.35 13.58 26 18.89 11.77 13.68 16.19
Episodes of depression
Less than 4 16 24.50 14.77 19.38 14.02 11 16.45 11.32 7.64 9.20
4 or more 10 23.10 18.82 11.50 11.93 8 22.25 12.29 22.00 20.41
SOFAS 26 65.39 10.05 72.81 11.90 19 70.58 10.77 73.26 13.33
Episodes of depression
Less than 4 16 63.19 9.16 70.88 12.07 11 70.00 10.42 78.72 13.99
4 or more 10 68.90 10.88 75.90 11.55 8 71.38 11.93 65.75 8.12
 V. Manicavasgar et al. / Journal of Affective Disorders 130 (2011) 138–144 143

Table 4 sodes of depression and relapse rates following treatment

Unadjusted mean scores at 6-month and 12-month follow-up. with MBCT. As this study was conducted with currently
N CBT N MBCT depressed individuals, while previous studies investigated
the impact of MBCT on those in remission, contrasting
Mean SD Mean SD
findings may reflect these differences. Preventing relapse to
BDI-II depression in high risk samples and treating depression may
6-month 14 15.94 14.31 13 18.39 12.68
both require separate analyses. As we did not use a diagnostic
12-month 14 18.93 16.51 9 18.56 11.25
interview at follow-up points to assess relapse rates, it is
BAI difficult to compare our findings to those of Teasdale et al.
6-month 14 14.79 15.06 14 10.93 12.42 (2000) and Ma and Teasdale (2004). Further research is
12-month 14 11.86 10.76 9 10.44 10.58 required to investigate relapse rates amongst those who were
either depressed or in remission while receiving treatment
and using a longer follow-up period.

treatment groups found at any time point. Unadjusted mean

scores are presented in Table 4. 4.1. Limitations

4. Discussion The generalisability of this study is limited by the small

sample sizes in both conditions and replication studies are
Our results suggest that both MBCT and CBT are equally clearly needed. The finding that participants in the CBT
effective for treating current depression. Both mean depres- condition attained the same level of improvement as those in
sion and anxiety scores significantly decreased in the CBT and the MBCT condition supports the efficacy of both treatments for
MBCT groups following the 8-week group program with no current depression however the small sample size, especially at
significant differences between the two treatment conditions. 6- and 12-month follow-ups, warrants caution when inter-
On average, participants shifted from the ‘severe’ to the ‘mild preting our results. The absence of a diagnostic interview at the
to moderate’ range on the BDI-II, indicating that although 6- and 12-month follow-up points makes it difficult to draw
both interventions were associated with improvement in definitive conclusions about the impact of the therapies on
depression, participants still experienced residual symptoms relapse prevention. Not all of our groups were randomly
post-intervention. This is consistent with our finding that the assigned, which may have impacted on the results and
level of psychosocial functioning did not change despite although participants were unaware of their group assignment
participants' improvements in depression and anxiety symp- until the first session and it is unlikely that this significantly
toms following group treatment and consistent with results compromised our findings, it would be helpful to include a non-
from the study by Eisendrath et al. (2008) which found that intervention control group in any further studies to control for
BDI-II scores also shifted to the mild-moderate range this possible confound. Prior experience with other psycholog-
following treatment with MBCT. ical therapies was not assessed in this study, and would be
Twelve-month and 6-month follow-up assessments indi- worthwhile exploring in future studies examining treatment
cated that gains were maintained over time. Again there were outcome.
no significant differences in depression and anxiety scores
between the treatment conditions, suggesting that both were
equally effective in this regard. 5. Conclusion
Importantly, we demonstrated a relationship between a
history of current depressive episodes and outcome depend- Despite some methodological limitations, we demonstrat-
ing on treatment condition. Participants in the CBT condition ed that MBCT was as effective as CBT in the treatment of
who had four or more previous episodes of depression current depression. When examined in relation to number of
returned significantly lower post-treatment BDI scores previous depressive episodes, CBT appeared to be more
compared to those with less than four episodes. In this effective at treating those with four or more episodes of
study, CBT was more effective for those with four or more depression while MBCT appeared effective irrespective of
previous episodes of depression while MBCT was equally number of prior depression episodes. Further investigation
effective for those with less than four or more than four may need to focus on the number of previous episodes of
previous episodes. depression and their effects on treatment outcome for both
A possible explanation for this finding is that the CBT CBT and MBCT and possible mechanisms of change.
condition specifically targeted and modulated trait vulnera-
bility and predisposition to depression by addressing dys-
Role of funding source
functional cognitive schemas which have been implicated in Funding for the study was provided by the National Medical Health and
depression recurrence (Conradi et al., 2008), while MBCT may Medical Research Council (NHMRC) Program Grant no. 222708 and Program
have had state benefits but did not modulate any predisposi- Grant no. 510135. The NHMRC had no further role in study design; in the
collection, analysis and interpretation of data; in the writing of the report;
tional cognitive factors.
and in the decision to submit the paper for publication.
Our findings about the effectiveness of MBCT in relation to
number of previous episodes contrast with those of Teasdale
et al. (2000) and Ma and Teasdale (2004) who found an Conflict of interest
inverse relationship between the number of previous epi- The authors have no conflicts of interest to declare.
144 V. Manicavasgar et al. / Journal of Affective Disorders 130 (2011) 138–144

Acknowledgements Kabat-Zinn, J., 1990. Full Catastrophe Living: Using the Wisdom of Your Body
and Mind to Face Stress, Pain, and Illness. Delta Books, New York.
Kenny, M.A., Williams, J.M., 2007. Treatment-resistant depressed patients
The authors thank Dusan Hadzi-Pavlovic, Sava Tsolis, show a good response to Mindfulness-based Cognitive Therapy. Behav.
Aimee Gayed, Peter Walker and David Gilfillan for their Res. Ther. 45, 617–625.
Kingston, T., Dooley, B., Bates, A., Lawlor, E., Malone, K., 2007. Mindfulness-
contributions to this study. based cognitive therapy for residual depressive symptoms. Psychol.
Psychother. 80, 193–203.
References Ma, S.H., Teasdale, J.D., 2004. Mindfulness-based cognitive therapy for
depression: replication and exploration of differential relapse preven-
tion effects. J. Consult. Clin. Psychol. 72, 31–40.
Barnard, J., Rubin, D.B., 1999. Small-sample degrees of freedom with multiple
Michalak, J., Heidenreich, T., Meibert, P., Schulte, D., 2008. Mindfulness predicts
imputation. Biometrika 86, 948–955.
relapse/recurrence in major depressive disorder after mindfulness-based
Barnes, S.A., Lindborg, S.R., Seaman Jr., J.W., 2006. Multiple imputation
cognitive therapy. J. Nerv. Ment. Dis. 196, 630–633.
techniques in small sample clinical trials. Stat. Med. 25, 233–245.
Oei, T.P.S., Dingle, G., 2008. The effectiveness of group cognitive behaviour
Barnhofer, T., Crane, C., Hargus, E., Amarasinghe, M., Winder, R., Williams, J.M.G.,
therapy for unipolar depressive disorders. J. Affect. Disord. 107, 5–21.
2009. Mindfulness-based cognitive therapy as a treatment for chronic
Peters, L., Andrews, G., 1995. Procedural validity of the computerized version
depression: a preliminary study. Behav. Res. Ther. 47, 366–373.
of the Composite International Diagnostic Interview (CIDI-Auto) in the
Beck, A.T., Brown, G.K., Steer, R.A., 1996. BDI-II Manual. The Psychological
anxiety disorders. Psychol. Med. 25, 1269–1280.
Corporation, San Antonio.
Rubin, D.B., 1987. Multiple Imputation for Nonresponse in Surveys. Wiley,
Beck, A.T., Rush, A.J., Shaw, B.F., Emery, G., 1979. Cognitive Therapy of
New York.
Depression. The Guilford Press, New York.
Segal, Z.V., Williams, M., Teasdale, J.D., 2002a. Mindfulness-based Cognitive
Beck, A.T., 2005. The current state of cognitive therapy: a 40-year
Therapy for Depression: A New Approach to Preventing Relapse.
retrospective. Arch. Gen. Psychiatry 62, 953–959.
Guildford Publications, New York.
Beck, A.T., Epstein, N., Brown, G., Steer, R.A., 1988. An inventory for
Segal, Z.V., Teasdale, J.D., Williams, M., Gemar, M.C., 2002b. The mindfulness-
measuring clinical anxiety: psychometric properties. J. Consult. Clin.
based cognitive therapy adherence scale: inter-rater reliability, adherence
Psychol. 56, 893–897.
to protocol and treatment distinctiveness. Clin. Psychol. Psychother. 9,
Bloch, M., Schmidt, P.J., Danaceau, M., Murphy, J., Nieman, L., Rubinow, D.R.,
131–138.
2000. Effects of gonadal steroids in women with a history of postpartum
Statistical Solutions Ltd., 2001. SOLAS for Missing Data Analysis. Statistical
depression. Am. J. Psychiatry 157, 924–930.
Solutions, Cork, Ireland.
Choy, Y., 2007. Managing side effects of anxiolytics. Prim psychiatry 14, 68–76.
SPSS Inc, 2009. SPSS for Windows, Rel. 18., Chicago.
Conradi, H.J., de Jonge, P., Ormel, J., 2008. Cognitive behavioral therapy v.
Teasdale, J.D., Segal, Z.V., Williams, J.M.G., Ridgeway, V.A., Soulsby, J.M., Lau, M.
usual care in recurrent depression. Br. J. Psychiatry 193, 505–506.
A., 2000. Prevention of relapse/recurrence in major depression by
Dozois, D.J.A., Dobson, K.S., Ahnberg, J.L., 1998. A psychometric evaluation of
mindfulness-based cognitive therapy. J. Consult. Clin. Psychol. 68, 615–623.
the Beck Depression Inventory-II. Psychol. Assess. 10, 83–89.
Tucker, M., Oei, T.P.S., 2007. Is group more cost effective than individual
Eisendrath, S.J., Delucchi, K., Bitner, R., Fenimore, P., Smit, M., McLane, M.,
cognitive behaviour therapy? The evidence is not solid yet. Behav. Cogn.
2008. Mindfulness-based cognitive therapy for treatment-resistant
Psychother. 35, 77–91.
depression: a pilot study. Psychother. Psychosom. 77, 319–320.
World Health Organisation, 1997. CIDI-Auto Version 2.1: Administrator's
Fydrich, T., Dowdall, D., Chambless, D.L., 1992. Reliability and validity of the
Guide Training and Reference Centre for WHO CIDI, Sydney.
Beck Anxiety Inventory. J. Anxiety Disord. 6, 55–61.
Goldman, H.H., Skodol, A.E., Lave, T.R., 1992. Revising axis V for DSM-IV: a
review of measures of social functioning. Am. J. Psychiatry 149, 1148–1156.