Compared to the general population, people with schizophrenia are at risk of dying prematurely due to suicide and due to different somatic illnesses. The potential role of antipsychotic treatment in affecting suicide rates and in explaining the increased mortality due to somatic disorders is highly debated. A recent study of death registers in Finland compared the cause-specific mortality in 66,881 patients versus the total population (5.2 million) between 1996 and 2006, suggesting that antipsychotic use decreased all-cause mortality compared to no antipsychotic use in patients with schizophrenia, and that clozapine had the most beneficial profile in this regard (Tiihonen et al., 2009). The benefits of clozapine were conferred by significant protective effects for suicide compared to perphenazine, whereas, a mixed group of 'other' antipsychotics, haloperidol, quetiapine and risperidone were reported to be associated with significantly higher all-cause mortality than perphenazine. By contrast, despite known differences in effects on cardiovascular risk factors, there were no significant differences between any of the examined antipsychotics regarding death due to ischemic heart disease. A number of methodological and conceptual issues make the interpretation of these findings problematic, including incomplete reporting of data, questionable selection of drug groups and comparisons, important unmeasured risk factors, inadequate control for potentially confounding variables, exclusion of deaths occurring during hospitalization leading to exclusion of 64% of deaths on current antipsychotics from the analysis, and survivorship bias due to strong and systematic differences in illness duration across the treatment groups. Well designed, prospective mortality studies, with direct measurement of and adjustment for all known relevant risk factors for premature mortality, are needed to identify risk and protective medication and patient factors and to, ultimately, inform clinical practice.
(c) 2009 Elsevier B.V. All rights reserved.