What is already known on this topic

Several meta-analyses of about 20 year’s work on randomised controlled trials (RCTs) with mindfulness-based cognitive therapy for depression (MBCT)1 have concluded that MBCT is efficacious in reducing relapse/recurrence where people have had at least three major depressive episodes (MDEs).

Methods of the study

In an update to a previous meta-analysis trial identification process,2 2555 potentially relevant studies published or accepted for publication between November 2010-November 2014 were identified through a systematic literature search using EMBASE, PsycINFO, PubMed/MEDLINE, Scopus, Web of Science and the Cochrane Controlled Trials Register. Of these, 766 studies were screened against study inclusion criteria; 8 full-text publications assessed for eligibility. Inclusion criteria were: (1) randomised trials of MBCT for prevention of depressive relapse reported in English in peer-reviewed journals; (2) adult participants in full or partial remission from MDE; (3) MBCT delivered according to treatment manual;1 (4) at least 1 non-MBCT treatment arm; and (5) outcome measurement of relapse/recurrence. After excluding five studies, the remaining three were pooled with six studies identified from the earlier meta-analysis. Meta-analytic approaches were applied to individual patient data (IPD) on available time to postintervention MDE combined across the nine studies (n=1258).

What this paper adds

• IPD analysis, better powered than RCTs separately, shows that MBCT reduces the risk of depressive relapse within a 60-week follow-up compared to controls (HR, 0.69, 95% CI 0.64 to 0.97) and is efficacious across broad demographic groups.

• MBCT is beneficial in comparison to antidepressant medication (HR, 0.77, 95% CI 0.60 to 0.98).

• Greater severity of baseline depression symptoms was associated with greater effect for MBCT (HR, 0.80, 95% CI 0.66 to 0.97).

• Adverse events occurred in three of the six studies where these were recorded: none were attributed to MBCT.

Limitations

As noted by the authors, asymmetry in the funnel plot suggests the possibility of some publication bias, to which we would add the comment that the timespan of the constituent studies means that clinical trial registration has not necessarily been consistently applied through the sampling period. Not all identified studies could be included in IPD analyses because of institutional ethical constraints. While no adverse events were attributed to MBCT, the paper highlights a lack of attention to their recording and reporting. The authors note that studies varied in data collection processes for time-course of MDEs and we would add that time to MDE may usefully be complemented by inclusion in analyses of variables such as depression-free days or differential quality of life.3 Earlier relapse may not necessarily be a poorer outcome in overall quality of life if that relapse is relatively brief and its severity mild. The authors note a lack of power for moderation analyses and use of some of these other outcomes might help address this limitation.

What next in research

Although adverse effects from MBCT were not identified in this meta-analysis, there is increasing literature and public discussion about the side effects of meditation and mindfulness, for example, increased anxiety, stress and depression.4 We lack validated and accepted measures of side-effects of MBCT, so furthering development of such measures could be a useful enabling priority along with systematic adverse event reporting. Beyond this, investigators could build on this work and that of others to refine our understanding of the range of applicability of MBCT including more detailed examination of mechanisms by which individual MBCT may prevent depressive relapse.5 For MBCT research to have real-world impact, there also needs to be continued attention to the advancement of the translational research agenda.6 With the availability of MBCT limited or absent in many areas, the research in this area may otherwise remain practically irrelevant for many clinicians.

Do these results change your practices and why?

Yes. The association of increased severity of symptoms with better outcome may guide intervention or referral. This relationship is presently unexplained; perhaps symptom flurries provide opportunities for practice of the cognitive and behavioural routines on which the effectiveness of MBCT is based, or motivation may be greater where depressive symptoms are experienced as an active threat. Where MBCT is unavailable, it would be helpful to advocate for the importance of preventative interventions and MBCT as mainstream. When advocating for MBCT, clinicians need to engage with accounts of negative experiences and critiques of secular mindfulness to present a balanced and credible picture.