This meta-analysis examined whether cannabis use is associated with an earlier age at onset (AAO) of psychosis, pooling 181 samples from 149 cohort and case-control studies (18,272 cannabis users vs. 52,801 nonusers) identified from MEDLINE, Embase, and PsycInfo through April 2025.
Cannabis use was associated with a mean earlier psychosis onset of **2.50 years** overall. The effect grew with later typical onset age: no difference in the <20-year-old group, but 1.60, 4.43, and 6.30 years earlier in the 20–<25, 25–<30, and ≥30 age groups, respectively. Sample mean onset age was a strong independent moderator (z = −7.59, p < .0001).
- Relies on cohort and case-control study designs; cannot establish causation. - Between-study heterogeneity is expected given the broad range of included samples and methodologies. - Cannabis potency and frequency of use data were not uniformly reported across studies, limiting dose-response analysis.
Clinicians should counsel patients—especially adults over 20—that cannabis use is linked to meaningfully earlier psychosis onset, with the gap widening substantially in older age groups. Public health messaging about cannabis and psychiatric risk should not be limited to adolescents.
