Impact of Cannabis Use on Time to Psychiatric Rehospitalization in Young Adults With Bipolar and Psychotic Disorders: A Survival Analysis.

Perino J.; Daveluy A.; Truong-Bastos A.; Miremont F.; Fatseas M.; Tournier M. · The Journal of clinical psychiatry · 2026

**Background & Methods** This historical cohort study followed 200 young adults (ages 18-29 years) hospitalized for psychotic or bipolar disorders in 2017 through May 2023. Cox regression models analyzed the association between cannabis use at index admission and psychiatric readmission rates during the year following discharge, with data extracted from medical records. **Key Findings** - Cannabis was the most frequently used substance (58% lifetime use, 43% current use at admission), followed by cocaine and ecstasy; 37% of the cohort met criteria for substance use disorder. - Cannabis use at admission significantly increased readmission risk within one month postdischarge (HR=2.8; 95% CI 1.3-5.9), with risk persisting but diminishing at three months (HR=1.9; 95% CI 1.1-3.3) and becoming statistically nonsignificant by six months. - The first three months postdischarge represented a critical vulnerability window, suggesting this period is optimal for targeted clinical interventions in cannabis-using patients with these disorders. **Dosage & Administration** Not reported. **Safety & Adverse Effects** Not reported. The study focused on psychiatric readmission as the primary outcome rather than specific adverse effects or safety profiles. **Evidence Quality** This study demonstrates moderate evidence quality with several limitations. The historical cohort design is susceptible to selection bias and unmeasured confounding—cannabis use may correlate with disease severity, medication adherence, or other factors affecting readmission independent of cannabis itself. The population was limited to young adults in a single healthcare system, potentially reducing generalizability. Medical record documentation of substance use may be incomplete or inaccurate, introducing information bias. The analysis does not distinguish between cannabis frequency, quantity, potency (THC concentration), or route of administration, limiting mechanistic understanding. Additionally, the study does not control for potential confounders such as medication compliance, psychosocial support, or concurrent psychiatric treatment intensity. While the findings are clinically relevant and suggest an association between cannabis use and early psychiatric relapse, causality cannot be established. The evidence supports cannabis use as a risk marker for early readmission but requires confirmation through prospective studies with more rigorous methodology.

More recent cannabis research