Psilocybin, cognitive stability and drinking reduction in co-occurring alcohol use disorder and depression: a pilot secondary analysis of a randomized, placebo-controlled trial
Ceyda Sayalı, Eli Weisman, Frederick Streeter Barrett
medRxiv October 29, 2026 preprint DOI: 10.64898/2026.10.27.26366630 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Secondary analysis of a randomized, double-blind, placebo-controlled trial |
|---|---|
| Sample size | 25 |
| Population | Patients with co-occurring alcohol use disorder and major depressive disorder (placebo n=13, psilocybin n=12) |
| Intervention | Psilocybin |
| Duration | Cognitive tasks completed before drug administration and again about one week after (mean 6.7 days, SD 1.1) |
| Measures | PHQ-9 |
| Topics | Addiction Depression Psilocybin |
| Registration | NCT04620759 |
| Key findings | About one week after dosing, psilocybin slowed cued task switching relative to placebo (switch-trial RT d=1.12) but showed no consistent effect on feedback-driven flexibility and no detectable effect on willingness to exert control. Drinking reduction was associated with increased cognitive stability (composite rho=-.64), while depressive symptom improvement was not associated with cognitive change. |
Abstract
Background. Change in cognitive control is a candidate transdiagnostic mechanism of psychedelic therapy. One account predicts that, once acute effects resolve, psilocybin shifts control toward flexibility while sparing inhibition and working memory. A motivational account (PRISM) predicts that psilocybin increases the willingness to exert control rather than the ability to do so. We tested both in patients with co-occurring alcohol use disorder (AUD) and major depression. Methods. In this secondary analysis of a randomized, double-blind, placebo-controlled trial of psilocybin for AUD with comorbid major depressive disorder (MDD; ClinicalTrials.gov NCT04620759 ), a subset of participants (placebo n=13, psilocybin n=12) completed five cognitive tasks before drug administration and again about one week after it (mean 6.7 days, SD 1.1). Outcome measures spanned the stability-flexibility continuum. Free choice between levels of cognitive demand indexed the participant's willingness to exert control. Effect sizes and Bayes factors were used to estimate drug effects and associations with change in drinking and depressive symptoms across 34 retest measures. Results. Relative to placebo, psilocybin slowed responding in cued task switching (learning-phase RT d=1.01, BF10=7.7; switch trials d=1.12, BF10=17). Psilocybin effects on feedback-driven measures were inconsistent in direction (7 of 11 toward flexibility); the largest was on perseveration, which rose less with practice under psilocybin than under placebo (d=-0.79, BF10=1.1). Other drug effects on the 34 retest measures had inconclusive Bayes factors (BF10=0.40-2.4), and none survived false discovery rate correction. Demand avoidance, analyzed separately from the 34 measures, showed no detectable drug effect (BF01=2.0). Baseline demand avoidance showed no detectable relation to performance but tracked depressive symptoms (rho=.71, BF10=24). Drinking fell in both arms, and on 15 of 17 stability-flexibility measures change toward stability accompanied drinking reduction (stability composite rho=-.64, BF10=29; rho=-.43 when change was adjusted for baseline scores). Lower baseline working-memory span predicted larger drinking reductions (BF10=19-25). Depressive symptoms fell more after psilocybin than after placebo (PHQ-9, d=-0.88, BF10=2.7), but their change was not associated with cognitive change or predicted by working-memory span. Conclusions. About one week after dosing, psilocybin slowed cued task switching but had no consistent effect on feedback-driven flexibility, and there was no evidence of an effect on the willingness to exert effort. Drinking reduction was associated with increased, not reduced, cognitive stability, whereas improvement in depressive symptoms was not associated with cognitive change. These pilot estimates can guide adequately powered trials.
Comparable studies
Other randomized controlled trials on psilocybin for addiction, most cited first.