Murrie · Schizophrenia bulletin 2020 · systematic review and meta-analysis · n=50 studies (40,783 participants)

Transition of Substance-Induced, Brief, and Atypical Psychoses to Schizophrenia: A Systematic Review and Meta-analysis.

Cited 170 times in the scientific literature.

Level 1 - systematic review of randomized trials

Systematic review and meta-analysis of observational inception cohorts evaluating prognostic risk/proportions

PubMed 31618428 · doi:10.1093/schbul/sbz102 · record verified 2026-08-26

What was done

Systematic review and meta-analysis of MEDLINE, PsycINFO, and Embase to evaluate the transition rate from substance-induced, brief, and atypical psychoses to schizophrenia. The search yielded 50 studies with 79 transition estimates among 40,783 individuals, including 25 studies reporting 43 substance-specific estimates across 34,244 individuals.

What was found

The pooled transition rate to schizophrenia was 25% (95% CI 18%–35%) for substance-induced psychoses, compared to 36% (95% CI 30%–43%) for brief, atypical, and not otherwise specified psychoses. By substance class, transition rates were: cannabis 34% (6 studies, 95% CI 25%–46%), hallucinogens 26% (3 studies, 95% CI 14%–43%), amphetamines 22% (5 studies, 95% CI 14%–34%), opioids 12%, alcohol 10%, and sedatives 9%. Rates were slightly lower in older cohorts, but did not differ by sex, country, setting (hospital vs community; urban vs rural), diagnostic methods, or follow-up duration.

Why it matters

Quantifies the substantial long-term risk of schizophrenia following substance-induced psychosis, identifying cannabis, hallucinogens, and amphetamines as the highest-risk exposures requiring targeted clinical follow-up.

Limits

Based entirely on observational cohort data, preventing causal inference regarding whether substances trigger schizophrenia or unmask preexisting vulnerability. Confidence intervals and study counts for alcohol, opioid, and sedative sub-analyses are not reported in the abstract. Substance-specific subgroup analyses for hallucinogens (n=3) and amphetamines (n=5) relied on very small numbers of studies.

Cited by