Psilocybin-assisted treatment for alcohol dependence: a proof-of-concept study

Classic (5HT2A agonist) hallucinogens were widely researched from the 1950s to the early 1970s for conditions including addiction. A meta-analysis of randomized trials of LSD for alcoholism reported an odds ratio favouring LSD at first post-treatment follow-up, and more recent work has renewed clinical interest in psilocybin. Small modern trials have examined psilocybin for anxiety in advanced cancer and as an adjunct for smoking cessation, and clinical research has generally supported the safety of classic hallucinogens when subjects are carefully screened and supervised. Despite this background and mechanistic rationale, there were no prior studies specifically testing psilocybin in people with alcohol dependence. Bogenschutz and colleagues therefore conducted a single-group, proof-of-concept study to quantify the acute psychoactive effects and tolerability of oral psilocybin in alcohol-dependent participants, and to provide preliminary outcome and safety data across a 36-week follow-up. The study aimed to evaluate whether administration of psilocybin, given alongside a manualised psychosocial intervention, was associated with changes in drinking behaviour and related psychological measures and whether acute subjective effects predicted clinical outcomes.

The investigators used a single-group, within-subjects design. Participants received a 12-week, 14-session manualised psychosocial intervention that included seven Motivational Enhancement Therapy (MET) sessions, three preparation sessions, and two debriefing sessions, with two open-label psilocybin administration sessions: the first after 4 weeks of psychosocial treatment and the second after 8 weeks. Outcome data were collected up to 36 weeks after baseline. Participants were recruited from the community and were aged 25–65 with DSM-IV alcohol dependence, at least two heavy drinking days in the prior 30 days, concern about their drinking, and not currently in treatment. Exclusion criteria included medical or psychiatric conditions judged exclusionary, family history of schizophrenia, bipolar disorder or suicide, dependence on cocaine/psychostimulants/opioids, or substantial recent hallucinogen use. Participants had to be abstinent and not in alcohol withdrawal at the time of psilocybin sessions. All provided written informed consent and the protocol was IRB-approved. Psilocybin dosing was weight-adjusted. For the first session all participants received 0.3 mg/kg orally; for the second session the planned dose was 0.4 mg/kg unless the participant declined, had adverse effects on the first session, or had experienced a "complete" mystical experience at 0.3 mg/kg. Sessions took place in a living-room-like room with two therapists present, the participant wearing eyeshades and headphones and instructed to focus inward. Vital signs were monitored frequently and medications for hypertension, anxiety, or acute psychosis were available if needed. Participants remained under observation for at least 8 hours and completed questionnaires beginning 7 hours after dosing. Assessments included medical screening, SCID for psychiatric diagnoses, CIWA-Ar for withdrawal screening, and multiple measures of acute hallucinogen effects (Hallucinogen Rating Scale intensity subscale, 5-Dimensional Altered States of Consciousness Scale, States of Consciousness Scale including the Mystical Experience Questionnaire items, and the ARCI). Drinking behaviour was measured using the Time-Line Follow-Back (TLFB) with heavy drinking days defined by standard sex-specific thresholds; drinking consequences were measured with the Short Inventory of Problems (SIP). Additional psychological measures included the Alcohol Abstinence Self-Efficacy Scale (AASE), Penn Alcohol Craving Scale (PACS), the SOCRATES motivation scale, and the Profile of Mood States (POMS). Safety monitoring captured adverse events and frequent vital signs. Statistical analysis was primarily descriptive with two a priori hypotheses formally tested. Paired t-tests compared follow-up scores with baseline and with week 4 (the pre-psilocybin time point); effect sizes (Cohen's d) were computed with correction for correlation between time points. The primary drinking outcome was percent heavy drinking days, with the main contrast being baseline versus weeks 5–12. With n = 10 the study was powered (power = 0.803) to detect a large pre–post effect (d = 1.0) at α = 0.05. The Benjamini–Hochberg procedure was used to control the false discovery rate for drinking outcomes.

Seventy individuals were screened and 10 participants were enrolled (four women, six men). Demographically the sample included two Native American/Alaska Native, one African American, four Hispanic and three white non-Hispanic participants; mean age was 40.1 years and mean duration of alcohol dependence 15.1 years. Ten participants completed the first psilocybin session. Seven began the second session and six received 0.4 mg/kg (one received 0.3 mg/kg due to a complete mystical experience in the first session). Nine participants completed all follow-up assessments; one discontinued shortly after the first psilocybin session and did not provide usable outcome data. Fidelity coding of MET sessions indicated adherence above the proficiency benchmark. Acute physiological and subjective effects were recorded. Systolic and/or diastolic blood pressure showed modest but significant increases between 30 and 180 minutes post-dose; heart rate did not change significantly. Monitor ratings of global drug effect and "distance from ordinary reality" peaked between 120 and 180 minutes. Self-report measures obtained 7 hours after dosing showed marked inter-individual variability; on average MEQ and HRS scores were numerically lower than reported in normal volunteers at comparable doses. Subjective ratings were strongly correlated between sessions for most hallucinogen-effect scales, and no statistically significant differences between the 0.3 mg/kg and 0.4 mg/kg sessions were observed for the six participants who completed both at the higher dose. Clinical drinking outcomes showed significant improvement following psilocybin administration. Percent heavy drinking days decreased during weeks 5–12 relative to baseline: mean difference 26.0 (SD 22.4), 95% CI 8.7–43.2, t(8) = 3.477, p = 0.008. The same measure also decreased relative to weeks 1–4 (the pre-psilocybin psychosocial period): mean difference 18.2 (SD 20.0), 95% CI 2.8–33.5, t(8) = 2.723, p = 0.026. Percent drinking days decreased during weeks 5–12 relative to baseline (mean difference 27.2 (SD 23.7), 95% CI 9.0–45.4, t(8) = 3.449, p = 0.009) and relative to weeks 1–4 (mean difference 21.9 (SD 21.8), 95% CI 5.1–38.6, t(8) = 3.010, p = 0.017). Improvements were maintained at subsequent follow-up points and most contrasts remained significant after controlling the false discovery rate; reported Cohen's d effect sizes were generally large (greater than 0.8) with one exception (range reported 0.75–1.38). Secondary outcomes including drinking consequences, craving, self-efficacy, and motivation improved at multiple time points relative to baseline and/or week 4. Changes in mood (POMS) were not significant except for increased Vigor at week 24. Because of substantial inter-individual variability in acute subjective response, the investigators examined correlations between intensity of the first session's acute effects and short-term clinical outcomes: large correlations were observed between measures of acute effect intensity and changes in drinking, and with craving and self-efficacy in some cases. Adverse events related to treatment were generally mild and transient: five participants reported headaches resolving within 24 hours, one had nausea with an episode of vomiting, one experienced diarrhoea (with pre-existing irritable bowel syndrome), and one reported transient insomnia. No participant required medication or intervention for blood pressure, anxiety, or psychiatric symptoms during sessions, and there were no reports of illicit hallucinogen use during study participation.

Bogenschutz and colleagues interpret the findings as demonstrating feasibility and preliminary signals of clinical benefit when psilocybin is administered within a structured psychosocial treatment for alcohol dependence. Participants' acute reactions were qualitatively similar to those seen in other populations, adverse events were transient and mild, and substantial reductions in drinking occurred after psilocybin administration rather than during the initial weeks of counselling alone. Strong correlations between the intensity of acute drug effects and subsequent reductions in drinking were identified, although the authors note that those correlations were not specific to measures of "mystical" experience and were also associated with other acute effect ratings. The investigators acknowledge important limitations that constrain causal interpretation. Prominent among these are the small sample size, the open-label single-group design with no control condition or blinding, and lack of biological verification of alcohol use; these factors make it impossible within this study to disentangle the specific pharmacological effects of psilocybin from expectancy, attention, psychosocial treatment, or time. The authors also note the high inter-individual variability in subjective response and that, on average, participants' acute subjective scores were lower than those reported in normal volunteers at comparable doses—consistent with historical observations that some people with alcohol dependence may require higher doses to achieve strong subjective effects. Finally, the extracted text of the Discussion appears truncated and the final remarks are not present in the provided material. The authors conclude that these preliminary results support conducting controlled trials with larger samples to test efficacy and to investigate mechanisms and predictors of therapeutic response.