Cannabis Use Linked to Earlier Psychosis Onset in Meta-Analysis

Researchers combined results from 149 peer-reviewed cohort and case-control studies. These studies compared the age when psychosis first began in people who used cannabis and those who did not.

Psychosis is a term for experiences such as hallucinations, delusions, or severely disorganized thinking and behavior. It can occur in several mental health conditions. The research team searched three medical research databases from their start dates through April 2025.

The main analysis included 181 samples from the 149 studies. In total, there were 18,272 cannabis users and 52,801 nonusers. The researchers used a random-effects meta-analysis, a method that combines findings while allowing for differences between studies.

Overall, cannabis use was associated with psychosis beginning 2.50 years earlier on average. That is an average difference between groups in the studies, not a prediction of what will happen to any one person.

The researchers also looked at whether the association differed depending on the average age when psychosis began in each study sample. They found no difference between cannabis users and nonusers in samples where psychosis began before age 20.

In samples with an average onset age from 20 to under 25, cannabis users developed psychosis 1.60 years earlier on average. The difference was 4.43 years in samples with average onset from 25 to under 30. In samples with an average onset age of 30 or older, the difference was 6.30 years.

A further analysis found that the sample’s average age of psychosis onset was strongly linked to the size of the earlier-onset difference. In plain terms, the association was larger in study groups where psychosis tended to start later in life.

The authors describe the overall association as robust, meaning it appeared consistently across the studies they combined. They said the findings support considering public-health warnings about possible psychiatric risks of cannabis use across the lifespan, rather than focusing only on adolescence.

Still, this research does not establish that cannabis caused psychosis or directly moved its timing earlier. The included studies were cohort and case-control studies, not experiments. People who use cannabis may differ from nonusers in other ways that could also be related to when psychosis begins.

The abstract does not provide details about how each original study defined cannabis use, how often people used it, the products or potency involved, or how fully individual studies accounted for other possible influences. Those differences can affect how confidently a combined result can be interpreted.

More research would need to better separate cannabis use from other factors linked with psychosis risk and timing. Studies that measure patterns of use carefully over time could help clarify why the association was stronger in samples with later average onset ages.