The relationships of classic psychedelic use with criminal behavior in the United States adult population

Hendricks and colleagues frame criminal behaviour as a persistent public-health and social problem in the United States, noting high rates of supervision, arrest, and recidivism and limited effectiveness of conventional interventions. The authors point to historical and emerging evidence that classic psychedelics (non-selective 5-HT2A agonists such as LSD, mescaline, and psilocybin) have produced psychological and prosocial effects in some clinical and population studies, and they summarise earlier, methodologically limited efforts that tested psychedelic-assisted psychotherapy with offending populations. Against this background, the study set out to examine whether lifetime use of classic psychedelics in the general US adult population is associated with recent criminal behaviour, and whether psilocybin use specifically shows similar relationships. The authors hypothesised that lifetime classic psychedelic use, and lifetime psilocybin use per se, would be associated with a decreased likelihood of past-year criminal behaviour and therefore aimed to test these associations in a large, nationally representative dataset to inform potential clinical research in forensic settings.

The investigators used pooled cross-sectional data from the National Survey on Drug Use and Health (NSDUH) for years 2002–2014, analysing adult respondents (≥18 years). The pooled sample comprised over 480,000 unweighted respondents; the NSDUH is designed to represent the non-institutionalised civilian US population. Analyses and the study protocol were approved by relevant institutional review boards. Lifetime classic psychedelic use was defined as any lifetime report of ayahuasca, DMT, LSD, mescaline, peyote or San Pedro, or psilocybin mushrooms. Four mutually exclusive exposure groups were created: Psilocybin Only (psilocybin but no other classic psychedelic), Psilocybin & Other Psychedelics, Non-Psilocybin Psychedelics Only, and No Psychedelics. Six criminal-behaviour outcomes were examined: three self-reported past-year behaviours—drug distribution (selling illegal drugs), larceny/theft, and assault—which were collected with ordinal response categories and collapsed into three levels (0 times, 1–2 times, >2 times); and three dichotomous past-year arrest outcomes for drug-related, property, and violent crimes constructed from specific arrest questions. A broad set of covariates was included to address confounding: survey year; age; sex; detailed ethnoracial categories; education; household income; marital and employment status; self-reported engagement in risky behaviour; degree to which religious beliefs influence decisions; and lifetime illicit use of a range of other substances (cocaine, other stimulants, sedatives, tranquilizers, heroin, pain relievers, marijuana, MDMA/ecstasy, PCP, and inhalants). Statistical analysis used generalized ordered logit regression for the three ordinal outcomes (allowing a partial proportional odds specification when proportionality was violated) and binomial logistic regression for the dichotomous arrest outcomes. All analyses accounted for the NSDUH complex survey design, sampling weights, and pooling across years. Data management was performed in SAS 9.4 and analyses in STATA 14 using the gologit2 package for generalized ordered logit models.

Descriptively, approximately one in seven respondents reported lifetime classic psychedelic use; the largest group was No Psychedelics, followed by Psilocybin & Other Psychedelics, Non-Psilocybin Psychedelics Only, and Psilocybin Only. Past-year criminal behaviours and arrests were infrequent in the weighted data (weighted percentages ranged about 0.2% to 1.2%). Lifetime classic psychedelic use was concentrated in younger adult strata (26–49 years), more common among men, non-Hispanic Whites and Native Americans/Alaska Natives, those with higher education and income, those reporting greater engagement in risky behaviour, employed individuals, and those reporting less influence of religious beliefs on decisions. The Psilocybin Only group in particular tended to be younger, more educated, higher income, more likely male and non-Hispanic White, and differed from other groups on lifetime use of other illicit substances. In adjusted models testing lifetime classic psychedelic use (versus no use) the investigators report specific associations: lifetime classic psychedelic use was associated with lower adjusted odds of past-year larceny/theft (adjusted odds ratio, aOR = 0.73; 95% CI 0.65–0.83) and lower odds of past-year assault (aOR = 0.88; 95% CI 0.80–0.97). For arrest outcomes, lifetime classic psychedelic use was associated with reduced odds of past-year arrest for a property crime (aOR = 0.78; 95% CI 0.65–0.95) and reduced odds of past-year arrest for a violent crime (aOR = 0.82; 95% CI 0.70–0.97). In contrast, lifetime classic psychedelic use was associated with increased odds of past-year drug distribution; authors report a 47% to 68% increased odds for this outcome and indicate the association was stronger among individuals who distributed drugs more frequently. Lifetime classic psychedelic use was not associated with past-year arrest for a drug-related crime. Analyses of the four mutually exclusive groups found that, relative to the Psilocybin Only group, the No Psychedelics and Non-Psilocybin Psychedelics Only groups were less likely to report past-year drug distribution, while Psilocybin Only participants were less likely than No Psychedelics to report past-year larceny/theft and assault. The Psilocybin Only group, however, showed higher likelihood of past-year drug distribution compared with No Psychedelics and Non-Psilocybin Only groups, and higher larceny/theft than the Psilocybin & Other Psychedelics group in some comparisons. For most outcomes, lifetime illicit use of other substances was associated with increased odds of criminal behaviour or arrest; exceptions varied by substance and outcome (for example, lifetime sedative and PCP use were not associated with drug distribution). Where proportional odds assumptions were violated for certain covariates or exposures, partial proportional odds models allowed coefficients to vary across outcome levels, and Wald tests were used to identify such instances.

Hendricks and colleagues interpret their results as broadly consistent with a protective relationship between lifetime classic psychedelic use and certain antisocial criminal behaviours in the general US adult population. They note that lifetime classic psychedelic use was associated with 27% lower odds of past-year larceny/theft, 12% lower odds of past-year assault, 22% lower odds of past-year arrest for a property crime, and 18% lower odds of past-year arrest for a violent crime, while lifetime use of most other illicit substances tended to be associated with increased odds of criminal outcomes. The authors highlight that the findings replicate and extend earlier, smaller or forensic-sample studies and that associations were observed both for self-reported offending and for arrest outcomes, which the authors argue reduces the likelihood that results reflect differential detection or evasion of arrest. The investigators also discuss an important exception: lifetime classic psychedelic use was associated with substantially increased odds of past-year drug distribution. They consider possible interpretations without asserting causality — for example, that distribution may relate to distributing psychedelics (which comprise a small share of drug arrests) or reflect political and anti-authoritarian attitudes among psychedelic users — and they note that classic psychedelic use was not associated with drug-related arrests, possibly because most drug arrests involve substances such as heroin, cocaine, or marijuana. Several limitations are emphasised. The cross-sectional, self-report NSDUH data prohibit causal inference and cannot resolve temporality or dose–response relationships. Important potential confounders (for example, personality traits such as openness, spirituality, or political orientation) might account for observed associations despite adjustment for many covariates. The authors acknowledge response biases, inability to identify which substances were distributed, and the possibility that psychedelics can cause adverse acute or longer-term mental-health effects in some individuals. Mechanisms linking psychedelic experiences to prosocial or reduced antisocial behaviour remain speculative; mystical-type experiences and emotions such as awe are proposed hypotheses but were not measured in these data. Finally, the paper stresses ethical considerations for any translation into forensic settings. Past coercive or non-consensual administration of psychedelics in correctional contexts is noted as a cautionary precedent. The authors argue for individual autonomy, rigorous ethical safeguards, and avoidance of any treatment context that could be perceived as institutionally mandated. They conclude that the pattern of findings supports further clinical investigation of classic psychedelics, particularly psilocybin, in populations at elevated risk of criminal behaviour, while urging careful attention to ethics and potential harms.

The authors conclude that lifetime classic psychedelic use, and to some extent lifetime psilocybin use specifically, is associated with lower likelihood of larceny/theft and other property crimes, and lower likelihood of assault and other violent crimes in the US adult population. They suggest these findings, together with prior evidence, provide a rationale for clinical research on classic psychedelics in forensic settings, while underscoring the need for ethical safeguards and voluntary participation.