Classic psychedelics and alcohol use disorders: A systematic review of human and animal studies

Classic psychedelics — including LSD, psilocybin, ayahuasca (DMT) and mescaline — act primarily at 5-HT2A receptors and can induce profound alterations of consciousness. Historical clinical use from the 1950s to the 1970s reported reductions in craving and alcohol consumption following psychedelic treatment, but many early studies lacked modern methodological safeguards such as control groups and blinding. Renewed interest since the 1990s has focused attention on these compounds as potential treatments for psychiatric disorders, including alcohol use disorder (AUD), and on underlying neurobiological mechanisms such as serotonin receptor-mediated plasticity. Calleja-Conde and colleagues set out to systematically review the human and preclinical literature published between 2000 and 2021 on classic psychedelics and alcohol-related outcomes. The stated aim was to synthesise recent evidence on whether classic psychedelics are associated with reductions in alcohol consumption or other alcohol-related behaviours, and to summarise proposed biological and psychological mechanisms that might explain any observed effects.

The investigators conducted a systematic search of PubMed and PsycInfo for articles published from 01/01/2000 through December 2021. Searches were restricted to English and Spanish and used the string "(classic psychedelics OR lysergic acid diethylamide OR psilocybin OR ayahuasca OR mescaline) AND alcohol." Journal articles only were considered and PRISMA principles guided the review process. Inclusion criteria allowed experimental and observational studies in humans and preclinical (animal) studies that examined the relationship between classic psychedelics and alcohol consumption or related phenomena. Exclusions comprised reviews, commentaries, conference materials, qualitative-only studies, papers without full text, and studies of non-classic psychedelics (for example MDMA or ketamine). No limits were placed on age, ethnicity, geographic setting, or whether psychedelic use was recreational or therapeutic. Screening and data extraction were carried out independently by two authors with a third author resolving disagreements. The reviewers used National Institutes of Health quality assessment tools to appraise study bias. Extracted human-study data included substance, sample size, study design and main results; for preclinical studies they recorded species, substance and dose, experimental model, and main outcomes. The authors grouped results by psychedelic substance for synthesis and reported study counts and characteristics in summary tables.

The systematic search ultimately yielded 27 eligible articles. From 585 non-duplicate records initially considered, 24 were selected through database searching and three additional manuscripts were identified from references, giving a final set of 27. Of these, 20 were human studies and 7 were preclinical. Human research comprised mainly observational designs (n = 17) and a smaller number of psychedelic-assisted intervention studies (n = 3). Among the animal studies, 57% evaluated voluntary alcohol consumption and the remainder addressed related phenomena such as alcohol-induced conditioned place preference or behavioural sensitisation. Human observational studies generally reported associations or self-reported improvements in alcohol-related outcomes following psychedelic use, but these studies had important methodological limitations. Common problems included lack of dose control, reliance on subjective self-report, frequent concurrent use of alcohol and psychedelics, and the cross-sectional or descriptive nature of many datasets, limiting causal inference. Some surveys indicated frequent co-use of alcohol and psychedelics; for example, more than half of respondents in one study who had used LSD or psilocybin reported often using them concurrently with alcohol. Among naturalistic reports of mescaline use, 76% of respondents with prior alcohol misuse or AUD reported improvements after mescaline, although these data are self-reported and uncontrolled. Preclinical findings were mixed but identified psilocybin as the compound with the most consistent data supporting reductions in relapse-like drinking. Meinhardt and colleagues reported that a subchronic psilocybin treatment (1 mg/kg) produced a 20% reduction in alcohol consumption on the first day of relapse in a rat model; subsequent work from the same group found that psilocybin (1 and 2.5 mg/kg) reduced relapse by 40–50% relative to vehicle. These studies propose a mechanistic interaction between 5-HT2A and metabotropic glutamate receptor 2 (mGluR2) within medial prefrontal cortex–nucleus accumbens circuits, noting that alcohol dependence is associated with reduced mGluR2 expression and that psychedelics may restore related behavioural deficits. Other mechanistic work cited by the authors implicates 5-HT2A effects in the ventral tegmental area (VTA), restoration of chloride homeostasis via KCC2, and modulation of brain-derived neurotrophic factor (BDNF) and GABAergic signalling as potential pathways by which 5-HT2A agonists reduce alcohol consumption in rodents. Ayahuasca-related animal studies suggested that pretreatment can block development of alcohol-induced conditioned place preference (a measure of reward) and attenuate alcohol-induced behavioural sensitisation; oral ayahuasca produced anxiolytic effects during ethanol withdrawal and prevented ethanol-induced changes in hippocampal 5-HT1A receptor and prodynorphin levels, and in the dynorphin/prodynorphin ratio in the striatum. However, one intermittent two-bottle choice study in Wistar rats using doses approximating ritual consumption found no effect on voluntary alcohol intake. Evidence regarding mescaline and peyote was sparse and inconsistent. An epidemiological analysis in young American Indians (n = 3,861) reported a positive association between 30-day alcohol use and peyote use, with larger odds ratios for recreational than for spiritual use. Another survey found high rates of simultaneous use of mescaline and alcohol among students. Naturalistic self-report data also indicated perceived improvements in alcohol problems after mescaline in many users, but these are uncontrolled observations. Across studies, no clear evidence emerged that classic psychedelics increase alcohol consumption. Safety data reported by the authors emphasise that classic psychedelics are generally considered physiologically non-toxic, carry low risk of dependence or compulsive use, and only rarely precipitate serious mental health events such as psychosis, a risk mitigated by psychiatric screening.

Calleja-Conde and colleagues conclude that recent research over the last two decades provides promising but inconclusive evidence that classic psychedelics could help reduce alcohol consumption. Observational human studies largely point toward beneficial associations, and several animal experiments — particularly those with psilocybin — show reductions in relapse-like drinking. Nevertheless, the investigators stress that the human literature is frequently limited by small samples, lack of control groups, variable dosing, subjective outcome measures, and confounding due to concurrent substance use, all of which impede causal inference. The authors discuss several plausible neurobiological mechanisms drawn from preclinical work. These include interactions between 5-HT2A and mGluR2 receptors in medial prefrontal circuits connected to the nucleus accumbens, modulation of VTA function and chloride homeostasis (KCC2) with downstream effects on GABAergic control and BDNF expression, and dynorphin/kappa-opioid alterations observed with ayahuasca components. The paper presents these mechanisms as hypotheses that may explain observed reductions in craving, relapse and other alcohol-related behaviours, but notes that replication is needed and that species, dose-regimen and paradigm differences may account for inconsistent findings across animal studies. Limitations acknowledged by the authors include the paucity of rigorous clinical trials, heterogeneity of study designs, and remaining uncertainty about the psychological processes that might mediate therapeutic effects (for example, whether so-called "personal transformation" or changes in self-perception and affect contribute to reduced drinking). Regulatory stigma and historical concerns about adverse effects have also slowed research progress. Regarding safety, the review reiterates that classic psychedelics are generally less harmful than alcohol and other controlled substances, but that clinical protocols should exclude individuals with predisposition to psychosis and should include psychiatric screening. Overall, the authors call for more high-quality research — controlled clinical trials, mechanistic preclinical studies and careful observational work with better dose and outcome measurement — to determine whether classic psychedelics, particularly psilocybin, can be effective and safe treatments for problematic alcohol use. They emphasise that current evidence is encouraging but insufficient to draw firm conclusions and that further study is required before these compounds can be recommended as standard treatments.