Psilocybin in alcohol use disorder and comorbid depressive symptoms: Results from a feasibility randomized clinical trial

Luquiens and colleagues frame alcohol use disorder (AUD) as a chronic, relapsing condition often accompanied by depressive symptoms, a combination that substantially raises the risk of early relapse after detoxification. The introduction notes that fewer than 7% of people with co-occurring addiction and another mental disorder receive treatment for both conditions, and that depression-related changes in reward processing and persistent craving after detoxification are important mechanistic drivers of relapse. Given these overlaps, treatments that simultaneously target alcohol use and depression are of interest. Psilocybin is presented as a candidate intervention because of emerging evidence for efficacy in both depression and addictive disorders and its neurobiological effects on connectivity, plasticity and subjective social processes. The authors identify inconsistent early evidence for psilocybin in AUD and point out that no published data address dual disorders specifically. They therefore designed a pilot randomised, double-blind trial to assess whether two oral administrations of 25 mg psilocybin, given three weeks apart as an adjunct to standard inpatient relapse-prevention care, would be feasible, acceptable and show preliminary efficacy for recently detoxified patients with severe AUD and persistent depressive symptoms compared with a very low 1 mg dose control.

The study was a prospective, single-centre, double-blind, parallel-group randomised pilot trial with a 2:1 allocation favouring the experimental arm (25 mg psilocybin) versus a very low dose control (1 mg). Eligible adults had severe AUD by DSM-5 criteria, a Beck Depression Inventory-II (BDI-II) score ≥14, and had completed detoxification 14 to 60 days prior to inclusion. Randomisation was stratified by gender and implemented via software after sequence generation by an independent statistician. Each participant received two dosing sessions spaced 21 days (±3 days) apart, with a preparation session the day before and an integration session the day after each dosing. Therapists (psychiatrists or psychologists) followed a standard therapeutic manual combining mindfulness, acceptance and commitment therapy (ACT) elements and non-directive supportive psychotherapy during dosing. Dosing occurred in patients' rooms with continuous monitoring of vital signs and repeated 0–10 self-ratings for intensity, anxiety and perceived threat during the session. The 5-Dimensional Altered States of Consciousness Scale (5D-ASC) was administered at hour 6. Primary outcome was feasibility, operationalised as completion of both dosing sessions and screening/inclusion rates; acceptability was measured by premature withdrawals and reasons. Secondary and exploratory outcomes included alcohol consumption assessed by the 4-week Alcohol Timeline Followback (TLFB) at 6 and 12 weeks (heavy drinking days [HDDs], drinking days, total intake, relapse defined as at least one HDD, and abstinence defined as no drinking in the past month), craving (Craving Experience Questionnaire), depression (BDI-II with post hoc responder definitions of ≥50% and ≥80% reduction), anxiety (Beck Anxiety Inventory), emotion regulation (DERS), rejection sensitivity (A-RSQ), and other baseline measures (ACE, RSQ, MoCA, PCL-5). Blinding integrity was assessed by participant and investigator guesses. Safety monitoring captured adverse events and suicidal risk. Sample size followed pilot recommendations (20 experimental, 10 control) and analyses used a per-protocol approach. Statistical methods included Fisher's exact test for the primary endpoint, χ2 or Fisher tests for categorical secondary outcomes, t-tests or Mann–Whitney tests for continuous measures as appropriate, Cox regression for time-to-relapse (hazard ratio), and an exploratory linear mixed model for DERS. Significance was set at α = 0.05. A last-observation-carried-forward sensitivity analysis was reported in supplementary material. The extracted text notes missing screening-delay data due to administrative processes.

Between 5 February and 24 October 2024, 350 individuals were screened and 30 were randomised into the per-protocol sample: 20 to 25 mg and 10 to 1 mg. Mean age was 49 ± 10 years and 43% were female. Baseline drinking in the last drinking period included a mean of 20 ± 9 HDDs and 21 ± 9 drinking days; the extracted text did not clearly report total daily alcohol consumption at baseline. Feasibility and acceptability: all participants received the first dosing session and integration visit. One participant in the 25 mg arm did not receive the second dose because of a myocardial infarction attributed to undiagnosed coronary artery disease occurring 3 days before the scheduled second session. Four control-group participants refused the second session after suspecting their allocation; three of those left the facility early and three declared intentions to self-administer psychedelics after discharge (one self-administered MDMA during follow-up). Completion of the second dosing session was higher in the 25 mg group (95% vs 60%; P = 0.031). Acute subjective effects: peak self-rated experience intensity reached 10/10 in the 25 mg group and 6/10 in controls. Two 25 mg recipients reported transient anxiety rated 10 that resolved to 0 within 30 minutes; the highest anxiety in controls was 4. Perceived threat peaked at 5 in one 25 mg participant and decreased to 0 within 30 minutes. No rescue medications were required. 5D-ASC profiles showed similar overall shapes across sessions in the 25 mg arm, with lower anxiety in the second session. Efficacy (12-week outcomes): at week 12 the abstinence rate was 11/20 (55%) in the 25 mg group versus 1/9 (11%) in the control group (one control lost to follow-up), a between-group difference of −44% (95% CI −82% to −5.9%), P = 0.043. Median percentage change in drinking days at 12 weeks was −100 (IQR −100 to −49) in the 25 mg group versus −93 (−96 to 0) in controls, P = 0.038. Craving frequency median change was −8 (−23 to −1) in the 25 mg arm versus +7 (−2 to 11) in controls, P = 0.045. Forty-five percent of participants in the 25 mg group achieved an ≥80% reduction in BDI-II at week 12 compared with none in the control group (P = 0.027); there was no significant difference for the ≥50% BDI-II reduction threshold. Time-to-first HDD (relapse) produced relapse rates of 35% in the 25 mg group and 50% in controls, with a hazard ratio of 0.52 (95% CI 0.16 to 1.65), a non-significant finding. Subgroup and exploratory analyses: no statistically significant differences in drinking outcomes were detected according to baseline antidepressant (SSRI/SNRI) use within the 25 mg group. Baseline PCL-5 score was reported as associated with abstinence status, but the extracted text truncates the full comparison and does not permit a complete statement of effect sizes. A sensitivity analysis using LOCF is reported in supplementary material but specific results are not fully included in the extracted text. Safety and adverse events: 19 adverse events occurred in 10 patients (50%) in the 25 mg group and six events in six patients (60%) in the control group; some participants experienced multiple events. Four events in the 25 mg group and one in the control group were judged attributable to treatment. Attributable events in the experimental arm included two brief episodes of dizziness in a single participant during a session, one episode of nightmares beginning 2 days after dosing and lasting 9 days, and a hypoglycaemic episode in a patient with comorbid anorexia that resolved with food. Suicidal risk events included one instance of suicidal ideation in the control group related to suspected allocation, one suicide attempt in a control participant about 1 month after the first session coinciding with relapse and life stress (who then refused the second session), and two episodes of suicidal ideation in the 25 mg group occurring during heavy drinking after discharge. Blinding: correct blinding-guess rates were high: 93.33% among patients and 86.67% among investigators. Patients in the control group more often guessed correctly than investigators, with the reverse pattern in the experimental group.

Luquiens and colleagues interpret their pilot data as supporting the feasibility and acceptability of two 25 mg psilocybin-assisted psychotherapy (PAP) sessions as an adjunct to intensive relapse-prevention care for recently detoxified patients with severe AUD and persistent depressive symptoms. Recruitment of 30 participants at a single centre in under 9 months and high completion of procedures are cited as indicators of demand and feasibility. The authors suggest the trial intervention may have functioned as a motivational lever for some patients to initiate detoxification. Although efficacy was not the primary outcome, the investigators highlight several statistically significant secondary findings favouring the 25 mg arm: a substantially higher abstinence rate at 12 weeks, greater reductions in percent drinking days and craving frequency, and a higher proportion achieving an ≥80% BDI-II reduction. They note that traditional depression responder thresholds (≥50% reduction) did not differ between groups and suggest that psychedelic responders sometimes experience near-complete symptom resolution, implying that conventional outcome measures may not fully capture the therapeutic profile of psychedelics. The discussion addresses the continuation of antidepressants in many participants; the authors observe that SSRI/SNRI use appeared to attenuate some anxiety-related dimensions of the acute experience but did not demonstrably reduce overall treatment response, while acknowledging limited power to detect subgroup differences. The relatively high proportion of female participants (43%) is presented as a study strength given the typical under-representation of women in severe-AUD trials. Blinding integrity is discussed, with the authors contending that using a very low dose comparator aligns with regulatory guidance and may minimise induced side effects; they also report that blinding-guess rates are consistent with prior AUD trials. Key limitations acknowledged include the single-centre pilot design, small sample size limiting representativeness and statistical power, incomplete feasibility metrics (screening delays), and some contamination risk from participant self-administration of other psychoactive substances. The authors recommend methodological refinements for future trials: refining drinking outcome measures, considering a higher (e.g. 3 mg) control dose, offering a later guaranteed high-dose session to control-arm participants to reduce refusals and contamination, and exploring mechanisms and optimal dosing schedules including the potential utility of an additional maintenance high-dose session.

The authors conclude that two 25 mg psilocybin-assisted psychotherapy sessions as an adjunct to intensive relapse-prevention care yielded promising feasibility and preliminary efficacy signals in recently detoxified patients with severe AUD and depressive symptoms. They call for mechanistic investigations and a larger efficacy trial, and recommend, for future studies, increasing the control dose to 3 mg and offering a third high-dose session to all participants to assess maintenance effects and reduce contamination bias.