Meta-analysis found cannabis use associated with 53% higher all-cause mortality, but evidence quality was low

Across 14 cohort studies with 17,545,076 participants (3,000,667 cannabis users), cannabis use was associated with a risk ratio of 1.53 (95% CI: 1.09-2.14) for all-cause mortality. The effect was significantly larger in prospective designs (RR 2.07) versus retrospective (RR 1.11), and in general population cohorts (RR 2.53) versus patient populations (RR 1.03). Larger sample sizes were associated with smaller effects. GRADE assessment rated overall evidence certainty as low.

Systematic review and meta-analysis registered with PROSPERO (CRD42023396915). Searched Scopus, PubMed, Web of Science, and ProQuest through October 2023. Included cohort studies comparing cannabis users to non-users for all-cause mortality. Random-effects meta-analysis with sensitivity analysis, meta-regression, and Newcastle Ottawa Scale quality assessment.

This is the first meta-analysis to synthesize evidence on cannabis use and all-cause mortality. The finding of a 53% increased risk is notable but comes with substantial caveats: high heterogeneity, potential confounding, and the association disappearing in patient populations suggest the relationship is more complex than a simple causal link.

Whether cannabis use itself increases mortality or whether cannabis users have other risk factors that drive mortality remains one of the most consequential unanswered questions in cannabis epidemiology. This meta-analysis finds an association but cannot resolve the causation question due to the limitations of the underlying studies.

Very high heterogeneity (I²=98%) reduces confidence in the pooled estimate. Unadjusted estimates included in many studies mean confounding factors (tobacco, alcohol, socioeconomic status) likely inflate the association. GRADE assessment rated evidence as low certainty. Larger studies showed smaller effects, suggesting potential publication bias or confounding.