Comprehensive Review Links Cannabis Use to Elevated Violence Risk for Users and Bystanders
A meta-analysis of 63 studies involving over 265,000 participants found that cannabis users faced approximately double the risk of committing violent acts compared to non-users, with rates increasing to three to four times higher for criminal convictions. The association persisted across both general and psychiatric populations, with psychiatric patients who used cannabis being 2.5 times more likely to be violent than non-using psychiatric patients. While the research cannot definitively establish causation, it highlights the broader social impact of cannabis use beyond individual health consequences.
The meta-analysis examined violence across multiple dimensions, including physical, sexual, and psychological harm. Researchers distinguished between self-reported violent behavior and acts resulting in criminal conviction, with the latter showing particularly elevated risk ratios. The study's longitudinal component—tracking participants over time—revealed weaker associations than cross-sectional studies, suggesting temporal measurement affects observed correlations.
Cannabis users themselves faced elevated victimization risks, experiencing violence at roughly 1.5 times the rate of non-users. This bidirectional relationship was more pronounced among women, indicating potential gender-based patterns in violence exposure that warrant further investigation.
As cannabis legalization expands globally, these findings could influence public health messaging, clinical practice guidelines, and policy discussions around age restrictions and potency limits. Healthcare providers may face questions about screening cannabis users for violence risk, while policymakers may consider whether packaging warnings or prevention programs should address interpersonal safety. The research could also shape conversations among users themselves about potential behavioral risks, though individual variation means population-level correlations may not apply predictively to specific persons.