Self-criticism statistically mediates symptom change in MDMA-assisted therapy for PTSD

This is a post hoc, exploratory secondary analysis of the phase 3 trial of MDMA-assisted therapy for severe PTSD (NCT03537014), run at 15 sites in the United States, Canada and Israel. Of 90 randomized adults, 82 had both baseline and 18-week data: 42 in the MDMA-assisted therapy group and 40 who received placebo plus the same therapy. A team from Yale School of Medicine asked whether self-compassion, measured with the 26-item Self-Compassion Scale, carries part of the treatment effect.

One scale, two halves

The authors split the scale into compassionate self-responding (self-kindness, common humanity, mindfulness) and uncompassionate self-responding (self-judgment, isolation, over-identification). Uncompassionate responding fell from 3.78 (SD 0.80) to 2.63 (SD 0.93) in the MDMA group and from 3.95 (SD 0.78) to 3.67 (SD 1.05) in the placebo group, a between-group difference of -0.88 (p<.0001). Compassionate responding rose in the MDMA group as well. The largest single change was in self-judgment, at -1.10.

Mediation, and what the design allows

In the mediation models, the change in uncompassionate responding carried the treatment effect on PTSD severity (CAPS-5, indirect effect B = 8.63, 95% CI 5.14–12.78) and on depressive symptoms (BDI-II, B = 8.51, 95% CI 5.07–12.28); the direct paths were not significant. Indirect effects on alcohol and drug use scores (AUDIT, DUDIT) were not significant. Self-compassion and PTSD severity were measured at the same two time points, so the analysis cannot show which changed first.

A target outside the protocol

The largest shift fell on self-judgment, and a clinician can ask about self-judgment, score it and follow it in any trauma treatment, with or without a study drug in the room. This study did not test self-compassion work outside the MDMA protocol. The authors themselves propose adding formal self-compassion components to MDMA-assisted therapy.