Summary and Comments – Mindfulness interventions for psychosis: A meta-analysis

Mindfulness interventions for psychosis: A meta-analysis

by Khoury, B., Lecomte, T., Gaudiano, B.A., & Paquin, K. (2013)

(Schizophrenia Research, 150(1). 176–184. )

Background & Research

Among the so-called Third Wave of cognitive-behavioural therapies, mindfulness-based approaches have garnered great interest in recent years, within clinical psychology as well as among the more general public. Definitions of “mindfulness” vary, but it is generally understood to refer to non-judgmental and non-reactive awareness, observation and acceptance of all inner experiences, the centering of attention and experience to the here and now, and in most cases, cultivating an attitude of kindness or (loving) compassion. In terms of interventions based on mindfulness, their unifying goal is to learn to willingly embrace and accept present experiences in the moment, both pleasant and unpleasant, without avoiding, suppressing or clinging on to them.

As regards psychosis, developing mindfulness skills could be helpful for alleviating the distress and suffering related to psychotic symptoms, instead of attempting to control them. Individuals might be able to change their ways of responding and ascribing meaning to the symptoms as they appear, regarding them as transient experiences that do not define one as a person or necessarily reflect reality. The symptoms or sensations are likely to remain unpleasant, but by promoting acceptance of them and their transient nature, individuals may reclaim power over themselves better than by attempting to fight, correct or counteract the symptoms. 

Traditional CBT approaches have been found to be effacious for treating residual positive and negative symptoms in psychotic disorders, but their effectiveness for treating affective problems (anxiety, depression etc.) is less clear. It is the affective symptoms as well as symptom-related distress and internalized stigma that mindfulness-based approaches might best impact. A number of mindfulness-based interventions for psychosis have already been developed, with different specific focuses including mindfulness meditation, acceptance based protocols and compassion based approaches. The APA already considers one mindfulness approach, Acceptance and Commitment Therapy, to be an empirically supported treatment for psychosis (as an aside, the Finnish Current Care Guidelines for schizophrenia do not currently mention any mindfulness-related interventions).

No meta-analyses on the effectiveness of mindfulness methods for treatment of psychosis have been carried out so far, nor is much known about what roles different components may play in their effectiveness. The aim in this article was thus to conduct “a meta-analysis of the currently existing studies that form the emerging evidence-base for mindfulness treatments for psychosis”. The result is an “effect-size analysis (1) to quantify the size of the effect of mindfulness interventions for psychotic disorders; and (2) to investigate and quantify the moderators of the effectiveness of mindfulness interventions for psychosis.”

Out of 378 potential articles initially identified, 14 were included in the final meta-analysis, representing 13 studies on 468 individuals with different psychotic spectrum disorders. The studies varied in the interventions used, the quality and stringency of methodology, the measures used to study clinical effectiveness, as well as the participants, but all represented mindfulness-based approaches for psychosis. Six studies featured pre-post analyses only, while the rest had control groups of different kinds.

Results & Discussion

The meta-analysis was performed by computing stardardized differences in means. Mean effect sizes for groups of studies were calculated by pooling individual effect sizes using a random effect model. Overall, the results show significant effects for both pre-post analyses (as measured by Hedge’s g, g = 0.52, p < 0.001), and controlled analyses (g = 0.41, p < 0.001). The effects were thus higher in pre-post analyses, but the authors suggest caution in interpreting this, as heterogeneity was moderate to high. Higher effects were found for negative rather than positive symptoms in both types of analyses, and no differences were found between individual or group treatments. Modest evidence was found for maintenance of the effects at follow-up (not many studies included proper follow-up).

Half the studies reported mindfulness-related outcome variables. Based on these studies, the average pre-post effect size of clinical outcomes was moderated by effects on mindfulness outcomes (n = 5; β = .33, SE = .11, p < .005, indicating a medium effect) and by effects on mindfulness, acceptance and compassion strategies combined (n = 6; β = .52, SE = .13, p < .0005, large effect). Study quality or length of treatment did not moderate effects, though this may be due to insufficient power.

Studying risk of bias across studies (publication bias),  analyses suggested that effect-size estimates for pre-post analyses were unbiased and robust. For controlled analyses, the effect-size estimates were somewhat less robust and might vary from small to moderate, depending on the strength of the control group used.

In sum, in this meta-analysis of 13 studies, mindfulness interventions for psychotic disorders were found to be moderately effective in pre-post studies. When compared with a control group, the effect sizes were small to moderate. Though the focus of these interventions was not on symptom reduction, they were found to be moderately effective in reducing negative and affective symptoms in individuals with psychotic disorders and in increasing functioning and quality of life. For positive symptoms, the effects appeared smaller.

The limitations of this meta-analysis include the relatively small number of studies and thus low power, high heterogeneity among some study groups, dissimilar outcome variables and varied quality of the studies. Despite these limitations, the meta-analysis does demonstrate the effectiveness of mindfulness-based interventions for individuals with psychotic disorders, especially in treating negative symptoms. Further, mindfulness and related elements appear to be active components in such interventions.

For future research, the authors stress the importance of clarifying the similarities and differences between traditional CBT versus mindfulness interventions for psychosis, especially as their effectiveness appears to be at a similar level (though perhaps targeting different symptoms?). Effectiveness studies should also include at least one validated measure of distress, as well as specific measures for changes in mindfulness, acceptance or compassion. They also call for reaching a better consensus regarding definitions and conceptualizations of mindfulness itself.


The initial, admittedly rather limited, studies and this methodologically impressive meta-analysis together begin to prove that mindfulness-based interventions can be helpful for psychotic disorders, and that their effects are in fact at least partially based on the specific mindfulness elements included. They may not reduce (at least positive) symptoms that significantly, but they seem to be at least moderately effective where it matters: improving the individuals’ overall functioning and quality of life. These findings have real clinical significance. Too often the treatment of psychotic disorders and schizophrenia in particular is only pharmacological, or pharmacological and psychosocial at best. In my view, the inclusion of (cognitive-behavioural and/or mindfulness-based) therapy and interventions should at this point become the standard.

The (quite tentative) finding that individual and group settings work equally well is also significant. In cash-stripped public healthcare systems, group interventions may be the more realistic option for the time being. However, individual psychotherapy should also be a real option for more individuals with psychotic disorders. Too often people with such severe problems are excluded from these services, seen as somehow unable to benefit from them. At least for CBT/mindfulness based therapies, this really does not seem to be the case.

The interventions studied here had been modified to suit individuals/patients prone to psychotic symptoms. Indeed, it would appear prudent not to include long-lasting, deep meditative exercises in interventions designed for such individuals, as the possibility of triggering symptoms or causing iatrogenic harm seems quite real to me, based on (limited) clinical experience. Emphasizing acceptance of and detachment from positive symptoms, as well as compassion and living in the present moment, would seem the safest and most likely useful approaches as regards mindfulness.

Lastly, I agree with the authors that the term “mindfulness” is in danger of becoming too all-encompassing and losing some of its specific meaning. In addition to its roots in meditative and spiritual practices, in layman usage, all sorts of approaches and thinking seem to be promoted under the title. Thus, when used in scientific studies, it should be defined as clearly and unambiguously as possible.