Predictors of death from natural causes in schizophrenia: 10-year follow-up of a community cohort

Purpose People with schizophrenia have increased natural mortality. There is much speculation but little evidence about the reasons behind this. This paper describes a study designed to measure the impact of pre-selected clinical, demographic and lifestyle variables on the natural mortality of a cohort with schizophrenia. Methods Ten-year Cox proportional hazards regression analysis of a community cohort of 95 people with schizophrenia. Results Death from natural causes was significantly associated with psychosis (HR 2.62, 95% CI 1.13–6.07), age (HR 1.08, 95% CI 1.02–1.13) and cigarette smoking (HR 2.53, 95% CI 1.01–6.34) at outset. There was a trend to association with low dietary unsaturated fat (P = 0.06). Conclusions Active psychosis appears to predict natural mortality in people with schizophrenia. Mental health services should prioritise the effective treatment of psychosis. Further research is needed to clarify other risk factors and evaluate health promotion interventions.