Biol Psychiatry. 2026 Aug 18:S0006-3223(26)01385-5. doi: 10.1016/j.biopsych.2026.06.025. Online ahead of print.
ABSTRACT
BACKGROUND: Increases in cannabis use, potency, and legal liberalization have raised concerns about psychiatric consequences. Previous studies suggest an association between cannabis use and earlier age at onset (AAO) of psychosis, but an updated meta-analysis is needed.
METHODS: MEDLINE, Embase, and PsycInfo were searched from inception to April 2025. Peer-reviewed cohort and case-control studies reporting mean and standard deviation of AAO of psychosis in cannabis users and nonusers or other effect size data were included. Pooled standardized mean difference was calculated using random-effects meta-analysis. Subgroup analyses used mixed effects models, and between-study heterogeneity was examined with random-effects meta-regression.
RESULTS: The main meta-analysis included 181 samples from 149 studies, comprising 18,272 cannabis users and 52,801 nonusers. Cannabis use was associated with a mean earlier AAO of 2.50 years. Subgroup analysis by sample mean onset age found no difference in AAO of psychosis between cannabis users and nonusers in the <20-year-old group, but there was a progressive age-related difference; cannabis users in the 20 to <25, 25 to <30, and ≥30 groups had an earlier AAO of 1.60, 4.43, and 6.30 years, respectively. Multiple meta-regression indicated that sample mean onset age was a strong and statistically independent moderator of earlier AAO effect size (z = -7.59, p < .0001).
CONCLUSIONS: The association between cannabis use and earlier psychosis onset is robust, with larger effect sizes in samples with later mean age of psychosis onset. These findings support consideration of public health warnings of the potential psychiatric risks associated with cannabis use across the lifespan.
PMID:42610992 | DOI:10.1016/j.biopsych.2026.06.025
