This meta-analysis of nine placebo-controlled trials (n=211) showed a very large effect size (g=1.21) of treatment on four mental health conditions (PTSD, end-of-life anxiety, depression, social anxiety among autistic adults).

Abstract

“After a two-decade hiatus in which research on psychedelics was essentially halted, placebo-controlled clinical trials of psychedelic-assisted therapy for mental health conditions have begun to be published. We identified nine randomized, placebo-controlled clinical trials of psychedelic-assisted therapy published since 1994. Studies examined psilocybin, LSD (lysergic acid diethylamide), ayahuasca (which contains a combination of N,N-dimethyltryptamine and harmala monoamine oxidase inhibitor alkaloids), and MDMA (3,4-methylenedioxymethamphetamine). We compared the standardized mean difference between the experimental and placebo control group at the primary endpoint. Results indicated a significant mean between-groups effect size of 1.21 (Hedges g), which is larger than the typical effect size found in trials of psychopharmacological or psychotherapy interventions. For the three studies that maintained a placebo control through a follow-up assessment, effects were generally maintained at follow-up. Overall, analyses support the efficacy of psychedelic-assisted therapy across four mental health conditions – post-traumatic stress disorder, anxiety/depression associated with a life-threatening illness, unipolar depression, and social anxiety among autistic adults. While study quality was high, we identify several areas for improvement regarding the conduct and reporting of trials. Larger trials with more diverse samples are needed to examine possible moderators and mediators of effects, and to establish whether effects are maintained over time.”

Authors: Jason B. Luoma, Christina Chwyl, Geoff J. Bathje, Alan K. Davis & Rafael Lancelotta

Notes

This paper is included in our ‘Top 10 Articles for Psychedelic Novices‘

The studies selected were all from between 2011 and 2018 and all were double-blind, randomized, placebo-controlled, (open-label) crossover or parallel arm. The studies were the following:

The three studied that reported long-term (seven weeks to six months) follow-up showed a decrease of 7.5% effect size, Hedges g = 1.36, “indicating that effects were largely sustained over the follow-up period.”

The authors are very positive about the studies they included, the critique they did have might be interesting for future studies:

An analysis could compare psychedelic therapy (so the combination that was used in these studies) with other pharma and therapy interventions (vs only looking at those separately)
Another meta-analysis with a larger sample size has historically shown to diminish the effect size
Future studies could focus on the type of placebo given
The long-term follow-up data is very tentative (as only three studies reported on it)
Reviewing the blinding procedures and implementing other ways of blinding than normally done and evaluated should also be implemented
Randomizing the amount of psychotherapy (whilst keeping the drug dosing the same) would also shed interesting light into how effective/necessary that part is

The total number of participants in the nine studies was 211. Of these, 165 received an active/high dose at least once, 125 received a placebo (and some received a high dose at a later date).

The participants received, on average, 1.9 active drug doses, if averaged over the 9 studies. If averaged over all non-placebo participants, they received 1.5 active doses. Two of the five MDMA studies had 3 dosing sessions.

The Hedges g is a measure of effect size. A score of 1 indicates a difference of 1 standard deviation between the groups. Everything above 0.8 can be interpreted as a high result. The mean between-groups effect size of 1.21 can thus be interpreted as a good validation of the effect size of psychedelics in combination with psychotherapy.

“The overall between-group effect size at the primary endpoint for psychedelic-assisted therapy compared to placebo was very large (Hedges g = 1.21). This effect size reflects an 80% probability that a randomly selected patient undergoing psychedelic-assisted therapy will have a better outcome than a randomly selected patient receiving a placebo (McGraw and Wong 1992)… Overall, results suggest that psychedelic-assisted therapy is effective with minimal adverse effects.”

“Although we compared the effect size in our study to those reported in prior reviews and meta-analyses, future research would benefit from directly comparing the effectiveness of psychedelic-assisted therapy to more traditional pharmacological, psychotherapeutic, or combined interventions. Future studies would also benefit from examining cost-effectiveness, given that psychedelic-assisted therapy, as currently delivered, may be relatively expensive compared to single-modality treatments.”

What was missing from many of the studies was the total number of psychotherapy hours (when reported it was around 10 hours). This may mean that the costs of such treatment (for the populations studied) might still be quite high. But, if compared to a multi-year psychotherapy treatment protocol without much effect, this may be much more effective.

Summary

A meta-analysis of placebo-controlled trials of psychedelic-assisted therapy is published in the Journal of Psychoactive Drugs.

After a two-decade hiatus, randomized, placebo-controlled clinical trials of psychedelic-assisted therapy have begun to be published. These studies have shown efficacy across four mental health conditions, including post-traumatic stress disorder, anxiety/depression associated with a life-threatening illness, unipolar depression, and social anxiety among autistic adults.

Clinical research on classical psychedelics was once widespread in the 1950s and 1960s, but was largely halted in the 1970s after the drugs were reclassified as Schedule I substances and banned in most countries. MDMA was widely used for therapeutic purposes before it was banned in 1985. The first systematic study of MDMA-assisted psychotherapy was not published until 2008, and it is the first meta-analysis of modern placebo-controlled clinical trials of psychedelic-assisted therapy. MDMA, psilocybin, ayahuasca, and LSD are used in psychedelic-assisted therapies. The outcomes of these therapies vary, but they all share several similarities, including similar study designs, concurrent psychotherapeutic support, and a focus on behaviour change post-drug administration. A meta-analysis of the effects of psychedelic-assisted therapy is timely because several organizations are progressing toward clinical trials that could result in MDMA and psilocybin being approved for medical use in the USA and Europe within the next several years.

We searched for all published randomized, placebo-controlled trials on psychedelic-assisted therapy published after 1993. We compared the pooled effect size between the experimental and placebo (control) group at the primary endpoint and examined the durability of effects through a long-term follow-up assessment.

We searched PsycInfo, ERIC, Medline, Academic Search Premiere and CINHAL for randomized placebo-controlled clinical trials that assessed the effect of MDMA, psilocybin, ayahuasca, N,N-dimethyltryptamine (DMT), or LSD on symptoms of a diagnosed psychiatric condition listed in either the DSM-IV or DSM-V. The process of selecting studies is outlined in the PRISMA diagram (Figure 1). The search yielded 931 records; 4 were duplicates within each database; 733 articles were excluded; 131 articles were reviewed; 9 studies were included; and 1 additional article was included.

During data extraction, an additional paper was excluded because it did not present data in a manner which allowed for meta-analysis. We calculated effect sizes for studies comparing active treatment groups to control (placebo) groups, and extracted data regarding sample, primary outcome measure, sample size, study design, drug and dose administered, number of therapy only sessions, number of dosing sessions before the primary endpoint, type of control group, and risk of bias.

Effect sizes were calculated using Cohen’s d, transformed to Hedges g, and averaged across studies. Meta-analyses were conducted using Meta-Essentials, and effect sizes were classified as small, medium, large, and very large. We tested six continuous and two categorical moderating variables using meta-regressions, including gender, race/ethnicity, mean age, number of dosing sessions, number of psychotherapy sessions, and sample size.

Nine studies were reviewed, and the average number of dosing sessions before the primary endpoint was 1.9 (SD = 0.8). Five studies examined MDMA, two psilocybin, one ayahuasca, and one LSD, and three studies included placebo-controlled long-term follow-up assessments.

The pooled effect of psychedelic-assisted therapy was significantly different from zero, and the heterogeneity was low. A total of 193 additional studies would be needed to bring the effect size down to a significance level of .05.

The observed effects of psychedelic-assisted therapy are tolerant to null results, and there is no significant difference in the overall effect size between classic psychedelic-assisted therapy and MDMA-assisted therapy.

Moderator analyses were performed on the data and found that none of the variables were statistically significant moderators.

Three studies reported long-term follow-up data with placebo control beyond the primary endpoint. The average effect size was 1.47 at the primary endpoint and 1.36 at the long-term follow-up point.

The overall between-group effect size for psychedelic-assisted therapy compared to placebo was very large (Hedges g = 1.21). This suggests a novel mechanism of action wherein effects extend beyond the direct biological effects of the psychedelic substance.

Meta-analyses of randomized clinical trials of psychedelic-assisted therapy for mental health conditions have found moderate effect sizes, with small to moderate effect sizes for pharmacological and psychological interventions for major depressive disorder, generalized anxiety disorder, social anxiety disorder, and mental health problems overall. The effect size observed in our review was larger partially because psychedelic-assisted therapy is a combined treatment. Future research would benefit from directly comparing the effectiveness of psychedelic-assisted therapy to more traditional pharmacological, psychotherapeutic, or combined interventions.

Psychedelic-assisted therapy may be cost-effective, but further research is needed to examine the reliability of the effect sizes observed. Additionally, future trials may have to include chronic or treatment-resistant populations, and use different types of placebos. We found no significant moderators of treatment effects on planned moderation analysis, suggesting comparable efficacy across the two classes of treatments (MDMA versus classical psychedelics). However, larger sample sizes may have yielded significant results.

The long-term follow-up data from three studies suggest that treatment effects were maintained, but this should be interpreted with caution because additional psychotherapy occurred during the follow-up period in one of the studies. Psychoactive compounds may have positive effects on people with end-of-life distress, those with PTSD, and among White volunteers. However, more research is needed into how to effectively achieve blinding when investigating compounds that frequently have obvious acute subjective effects.

Methodological challenges in pharmacological research may require study designs that depart from standard clinical trial methods, and researchers recommend that future investigators directly assess whether the blind was broken by asking assessors, treatment staff, and patients to guess whether they received the active treatment. Several reports did not describe or quantify the amount of concurrent psychotherapy and supportive contact.

We did not find any statistical indication of reporting bias, but a recent report of six Phase 2 randomized trials of MDMA-assisted psychotherapy for the treatment of PTSD demonstrated a combined between-groups effect size of Cohen’s d = .8. This could reflect publication bias.

A meta-analytic review of randomized placebo-controlled trials found that psychedelic-assisted therapy had large effect sizes and high quality, but that improvements were needed regarding the conduct and reporting of trials in the future.

The authors thank Brandon Strickland for his help during the literature search. The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

AKD and RL are board members at Source Research Foundation, which had no input into the present study.