Psychedelic use is associated with reduced daily opioid use among people who use illicit drugs in a Canadian setting

Opioid use disorder (OUD) causes major morbidity and mortality and remains difficult to treat in many settings, in part because of toxic contamination of the illicit supply and barriers to sustained engagement with conventional treatments such as opioid agonist therapy. Recent renewed scientific interest has examined whether classic psychedelics — primarily 5-HT2A agonists such as LSD, psilocybin, DMT/ayahuasca and mescaline — might have therapeutic potential for mental health and substance use problems. Proposed mechanisms include enhanced neural plasticity and profound subjective experiences (often described as mystical experiences) that may mediate enduring changes in mood, behaviour and motivation to use substances. Argento and colleagues set out to examine whether recent naturalistic use of classic psychedelics is associated with subsequent daily illicit opioid use among people who use drugs (PWUD) in Vancouver, Canada. The study aimed to address a gap in prospective, individual-level observational evidence by analysing harmonised cohort data with repeated measures of substance use and a broad set of potential confounders, using lagged exposure modelling to strengthen temporal inference.

The investigators analysed longitudinal data pooled from three ongoing community-recruited prospective cohorts of PWUD in Vancouver: VIDUS (injection drug users, HIV-negative), ACCESS (people living with HIV who use drugs) and ARYS (street-involved youth). Recruitment used community outreach, snowball sampling and self-referral; psychedelic-use questions were introduced in 2006. Participants completed interviewer-administered questionnaires and bi-annual follow-up visits, providing self-reported information on sociodemographics, detailed substance-use patterns (including psychedelic use), healthcare engagement and social-environmental factors. Biosamples for HIV and HCV testing were collected where appropriate. The primary outcome was at least daily use of any illicit opioids (heroin, fentanyl, nonmedical pharmaceutical opioids) in the prior six months, used as a pragmatic proxy for OUD when diagnostic data were not available. The primary exposure was any use of classic psychedelics in the prior six months, derived from an ‘other drug use’ questionnaire item and including LSD/acid, magic mushrooms/psilocybin, mescaline (San Pedro, peyote), salvia and DMT/ayahuasca. Most covariates were time-updated to the past six months; baseline, time-fixed covariates included age, gender, ethnicity and childhood trauma measured by the Childhood Trauma Questionnaire. Time-updated potential confounders comprised HIV serostatus, homelessness, incarceration, sex work, recent physical/sexual violence, barriers to accessing addiction treatment, current enrolment in drug treatment (including OAT), and frequent (≥ daily) use of other substances (cannabis, benzodiazepines, crack, cocaine, methamphetamine), plus a defined measure of heavy alcohol use (>14 drinks/week or >4 drinks on one occasion for men; >7/week or >3 on one occasion for women). For analysis, the sample comprised participants with at least one visit between January 2006 and December 2018. Descriptive statistics compared baseline characteristics by psychedelic use. Associations were estimated using generalized linear mixed-effects models (GLMM) with a logit link and random intercepts to account for repeated measures. To ensure exposure preceded outcome, psychedelic use was lagged to the value recorded at the prior visit; this lagging required at least two visits and led to exclusion of single-visit observations from models requiring the lagged variable. Bivariable associations informed a conservative stepwise confounding selection for the multivariable GLMM: starting from a full model, covariates were removed iteratively based on their impact on the primary exposure coefficient until remaining changes exceeded 5%. Observations with missing values on outcome or explanatory variables were excluded via listwise deletion. Results are presented as unadjusted and adjusted odds ratios with 95% confidence intervals; analyses were conducted in R version 4.0.3.

Between 2006 and 2018, 3,813 participants contributed 34,831 observations and were included in analyses after removing five participants with missing psychedelic-use data. The median number of follow-up visits per participant was 6 (IQR 2–15) and the median interval between visits was 6.2 months (IQR 5.6–7.0). At baseline the median age was 32 years, 65% identified as white (n = 2464), and 34% (n = 1307) were cis/trans women or other gender minorities. Psychedelic use in the prior six months was reported by 6% (n = 229) at baseline and by 35% (n = 1333) at any point over follow-up. Among baseline psychedelic users, LSD was the most frequently reported agent (61%), followed by psilocybin mushrooms (51%); smaller numbers reported DMT, salvia or mescaline. At baseline, 29% (n = 1093) reported daily illicit opioid use in the prior six months; across follow-up, 22% (n = 7539) of observations recorded daily opioid use. In bivariable GLMM analyses, psychedelic use was associated with lower odds of subsequent daily opioid use (OR 0.59; 95% CI 0.40–0.87; p = 0.007). Several covariates were positively associated with daily opioid use at p < 0.001, including younger age, identifying as a woman, recent homelessness, incarceration, involvement in sex work, recent violence, barriers to treatment, and daily use of benzodiazepines, crack, cocaine or methamphetamine; the latter group had approximately 2.5 to over 3 times increased odds of daily opioid use. Conversely, living with HIV, daily cannabis use, heavy alcohol use, and recent enrolment in drug treatment were associated with reduced odds of daily opioid use (all p < 0.001). In the final multivariable confounding model that adjusted for age, HIV status, childhood trauma, homelessness, daily cannabis use, daily stimulant use, and enrolment in opioid/other drug treatment, lagged psychedelic use remained independently associated with a 55% reduction in the odds of subsequent daily illicit opioid use (Adjusted Odds Ratio 0.45; 95% CI 0.29–0.70; p < 0.001). A prespecified sensitivity analysis restricted to participants who had ever used illicit opioids produced a similar statistically significant association (AOR 0.58; 95% CI 0.38–0.86). The paper reports the full multivariable model in supplementary material.

Argento and colleagues interpret their findings as evidence that recent naturalistic use of classic psychedelics is associated with substantially lower odds of subsequently reporting daily illicit opioid use among a large, prospectively followed community sample of PWUD. They note this is, to their knowledge, the first prospective cohort analysis with semi-annual individual-level data demonstrating this association after adjustment for a wide range of measured confounders and after lagging the exposure to preserve temporal ordering. The authors situate the results alongside prior observational and experimental literature that has reported links between psychedelic use and reduced rates of various substance use disorders, and they discuss plausible mechanisms including neurobiological effects on plasticity and psychological processes such as mystical or meaning-enhancing experiences that may facilitate behavioural change. They also reference clinical trial and historical RCT findings suggesting potential therapeutic effects for alcohol, tobacco and opioid dependence, and acknowledge related findings from small clinical pilot studies. Several limitations are emphasised. The observational design precludes causal inference and residual confounding is possible; for example, unmeasured personality traits or spiritual orientations that influence both likelihood of trying psychedelics and propensity to use opioids could bias results. Reliance on self-reported substance-use data may introduce recall or response bias, and psychedelic use may have been underreported because it was captured under an ‘other drug use’ item. The investigators used daily opioid use as a proxy measure for OUD owing to the absence of DSM diagnoses, and they acknowledge that their lifetime illicit opioid measure may underestimate true prevalence. To mitigate temporality concerns they lagged exposure and conducted a sensitivity analysis among those who had ever used opioids, which yielded consistent results. Finally, while the sample included some gender and ethnic diversity (34% women/gender minorities and 35% non-white participants, including 31% Indigenous), the authors call for greater inclusion of diverse populations in future research. Implications raised by the authors include the need to consider psychedelic-assisted psychotherapy as a potential adjunct to existing treatments while continuing to address upstream socio-structural drivers of OUD (for example, stigma, housing instability and incarceration). They state that confirmation in other settings, controlled trials, longer-term follow-up and qualitative work to understand psychosocial mechanisms are required before translating these observational findings into clinical or policy recommendations.

The study found that recent naturalistic use of classic psychedelics was independently associated with significantly lower odds of subsequent daily illicit opioid use among a community-based sample of people who use drugs in Vancouver. The authors conclude that further research — including controlled trials, longer-term follow-up, qualitative investigations, and studies in more diverse populations — is needed to clarify the therapeutic potential and mechanisms by which psychedelics might augment interventions for substance use disorders.