Mortality in schizophrenia and bipolar disorder: Clinical and serological predictors.

Persons with schizophrenia and with bipolar disorder have a reduced life expectancy due largely to death from natural causes. The reasons for this increased mortality have not been completely defined. We prospectively assessed a cohort of persons with schizophrenia and one with bipolar disorder with a clinical evaluation and a blood sample from which immune and infectious disease markers were measured. Mortality was determined with data from the National Death Index following a period of up to 14years. We examined the role of demographic, clinical, and serological factors on mortality in bivariate and multivariate models. A total of 43/710 (6.1%) persons with schizophrenia and 12/406 (3.0%) with bipolar disorder died of natural causes. In the schizophrenia group, mortality was predicted by the following variables in a multivariate model: cigarette smoking (RR=6.93, 95% CI 1.59, 30.1, p=0.0099); autoimmune disorder (RR=8.08, 95% CI 2.50, 26.1, p=0.00047); gastrointestinal disorder (GI) (RR=3.53, 95% CI 1.43, 8.69 p=0.0061); and reduced maternal education (RR=0.84, 95% CI 0.72, 0.97), p=0.018. The combination of smoking and an autoimmune disorder yielded an unadjusted relative risk of 18.1 for mortality, and the combination of smoking and a GI disorder an unadjusted relative risk of 9.45, compared with individuals with neither risk factor. In the bipolar disorder group, significant bivariate predictors of mortality included lower cognitive score (RR=0.95, p=.0085) and the presence of type 1 or 2 diabetes (RR=3.90, p=.026). Given the extraordinary high risk of death due to smoking in schizophrenia, smoking cessation remains an urgent priority.