Incidence of cancer among persons with schizophrenia and their relatives

Abstract

Background: It has repeatedly been reported that the risk for cancer in patients with schizophrenia is different from that of the general population, specifically a lower risk for lung cancer despite increased smoking. Confirmation of these associations could lead to hypotheses on shared risk or protective factors, either genetic or environmental.

Methods: From Finland's National Hospital Discharge and Disability Pension registers, Helsinki, we identified a cohort of 26 996 individuals born between 1940 and 1969 and treated for schizophrenia between 1969 and 1991. They were followed up for cancer from 1971 to 1996 by record linkage with the Finnish Cancer Registry, yielding 446 653 person-years at risk, and standardized incidence ratios (SIRs) were calculated. Likewise, 39 131 parents and 52 976 siblings of the patients with schizophrenia were followed up to explore familial genetic hypotheses on deviations in cancer risk.

Results: In patients with schizophrenia, an increased overall cancer risk was found (724 cases observed vs 619 expected; SIR, 1.17; 95% confidence interval [CI], 1.09-1.25). Half of the excess cases were attributable to lung cancer (SIR, 2.17; 95% CI, 1.78-2.60), and the strongest relative increase in risk was in pharyngeal cancer (SIR, 2.60; 95% CI, 1.25-4.77). Cancer incidence in siblings (SIR, 0.89; 95% CI, 0.83-0.94) and parents (SIR, 0.91; 95% CI, 0.89-0.93) was consistently lower than that in the general population.

Conclusion: Although specific lifestyle factors, particularly tobacco smoking and alcohol consumption, probably account for the increased cancer risk in patients with schizophrenia, the decreased risk in relatives would be compatible with a postulated genetic risk factor for schizophrenia offering selective advantage to unaffected relatives.