Patients with schizophrenia are at increased risk of suicide, but data from controlled studies of how pharmacotherapy is related to suicide risk is limited.

To explore suicide risk in relation to prescription of antipsychotics and antidepressants.

Of all patients with a first clinical discharge diagnosis of schizophrenia or schizoaffective disorder in Stockholm County between 1984 and 2000 (n = 4,000), patients who died by suicide within five years from diagnosis were defined as cases (n = 84; 54% male). Individual matching was performed with schizophrenia controls from the same population. Information on prescribed medication was retrieved from psychiatric records in a blinded way. Odds ratios (OR) of the association between medication and suicide risk were calculated by conditional logistic regression.

No significant association was observed between suicide and having ever been prescribed any antidepressant (33 cases and 30 controls) or any antipsychotic (83 cases and 82 controls). A lower suicide risk was found in patients who had ever been prescribed a second generation antipsychotic (risperidone, ziprazidone, olanzapine or clozapine; 12 cases and 23 controls): OR 0.2 (95% confidence interval [CI], 0.1–0.7). When the 6 cases and 8 controls who had been prescribed clozapine were excluded, the OR was 0.1 (95% CI, 0.03–0.6).

The lower suicide risk for patients who had been prescribed second generation antipsychotics may be related to a pharmacological effect of these medications, to differences in compliance, or to differences in other characteristics associated with a lower suicide risk.