Background and aims
Substance-induced psychosis has previously been linked to an excess risk of suicide; however, the association between substance-induced psychosis and suicide attempt has hitherto not been investigated. We investigated whether substance-induced psychosis was associated with a higher risk of subsequent suicide attempt.

Design
Nation-wide prospective register-based cohort study.

Setting
Denmark.

Participants
All people living in Denmark aged 13 years or more during 1995 to 2017.

Measurements
Substance-induced psychosis and suicide attempts were identified through hospital records as ICD-10 codes.

Findings
A total of 8900 (78.8% males) individuals were diagnosed with a substance-induced psychosis, and 740 of these had a suicide attempt during follow-up. People with a substance-induced psychosis had a higher risk of a subsequent suicide attempt [hazard ratio (HR) = 13.4, 95% confidence interval (CI) = 12.4–14.4] when compared with the general population. The highest hazard ratios were found for psychosis induced by opioids (HR = 26.4, 95% CI = 18.2–38.2); alcohol (HR = 17.7, 95% CI = 15.2–20.6); sedatives (HR = 17.2, 95% CI = 8.9–33.0); and cocaine (HR = 15.6, 95% CI = 10.7–22.8), while cannabis-induced psychosis was linked to an HR of 8.9 (95% CI = 7.7–10.3). Approximately 15% of patients with substance-induced psychosis had had a suicide attempt within 20 years of their substance-induced psychosis diagnosis.

Conclusions
In Denmark, substance-induced psychosis appears to be strongly associated with subsequent suicide attempt, underscoring the importance of attention and better follow-up for this patient group.