Review of 23 meta-analyses on psychedelic-assisted therapy for mental disorders

The question

How strong is the overall evidence that psychedelic-assisted therapy works, and is safe, for conditions such as depression, post-traumatic stress disorder (PTSD) and substance use disorders?

What they did

This was an umbrella review: a study that gathers together existing reviews rather than individual trials. The authors identified 23 meta-analyses covering more than 100 primary studies of psilocybin, MDMA, LSD and ayahuasca. Results were converted to a common measure of effect size (Hedges’ g) so different studies could be compared. The quality of each review was rated with the AMSTAR2 tool, and the certainty of the evidence with the GRADE framework.

What they found

The evidence base was uneven. Psilocybin for depression (9 meta-analyses) and MDMA for PTSD (10 meta-analyses) were best covered, while only 2 meta-analyses each existed for LSD in alcohol use disorder, LSD in depression, ayahuasca in depression, and MDMA in autism spectrum disorder. Psilocybin showed large effects in major depression (Hedges’ g of about 1.05), with some evidence of benefit lasting up to six months. MDMA showed very large effects on PTSD symptoms (Hedges’ g of about 1.24), usually after 2 to 3 sessions, and LSD showed short-term benefit in alcohol use disorder (odds ratio of about 2.0) that faded over time. Reported side effects were generally mild and temporary, with no consistent signal of serious harm.

The fine print

An umbrella review can only be as good as the reviews it summarizes, and the authors describe the methodological quality as low, with small and sometimes overlapping samples, inconsistent results and risk of bias. Long-term safety and efficacy data are scarce, so most conclusions apply to the short term. The abstract does not give a total number of participants, and this summary is based on the abstract alone.

What it means

Pooled evidence points to short-term benefits from psilocybin in depression and MDMA in PTSD, with mostly mild and short-lived side effects, but the quality of the underlying research is limited and long-term data are still missing.

Evidence grade: Moderate. It pools 23 meta-analyses of randomized trials, but the authors themselves rate the underlying studies and reviews as methodologically weak, with overlapping samples and little long-term data.

This summary is based on the study’s published abstract.