An 18-month longitudinal study of suicidality in patients diagnosed with schizophrenia.

Abstract Background Suicide kills over 30,000 people annually in the US. Schizophrenia increases the risk, even in psychiatric populations. Identifying high-risk groups within this patient population is central to suicide prevention. We tested the effects of known and putative risk factors for suicide in a clinically recognizable population with symptoms of schizophrenia and unstable illness. Methods We studied 1439 subjects with schizophrenia participating in the NIMH-funded Clinical Antipsychotic Treatment for Intervention Effectiveness (CATIE) trial and followed for 18 months. The dependent variable was moderate to severe self-reported suicidal thinking and self-reported suicide attempts. The independent variables comprised potential risk factors for suicidality including measures of clinical change during follow-up. Proportional hazards models of time to first suicidality generated bivariate and multivariate hazard ratios (HRs). Results Altogether 96 patients (7.6%) experienced moderate to severe suicidality in the course of 18 months. Multivariate analyses showed that baseline moderate or severe suicidality was the strongest correlate of subsequent moderate or severe suicidality (HR 5.1). An increase in a subject's depression score during follow-up was also independently and strongly associated with suicidality (HR 3.5). A change in psychotic symptoms was not. Conclusions Depression and despair may be more important risk factors for suicidality among people with schizophrenia than psychotic decompensation. Strategies for prevention that seek to identify groups at high risk of suicide should focus on these variables, as well as the effect of prior suicidality.