THE RELATIONSHIP between schizophrenia and cancer morbidity and mortality has been a sort of epidemiological puzzle for decades. Several studies have reported conspicuous reductions in the occurrence of malignancies, in particular lung cancer, in schizophrenic patients. Authors have typically commented on the difficulty of explaining such findings considering the widespread tobacco addiction in this population. However, the majority of these earlier studies have been of small size, based on proportionate mortality statistics rather than on incidence, and generally lacking in methodological rigor. Baldwin%1 estimated that at least 100 000 person-years of observation would be required to reliably identify any significant association, positive or negative, between schizophrenia and cancer. Among the few studies meeting this requirement, the World Health Organization (WHO) multicenter investigation, based on record linkages across case registers, found significant and consistent decreases in the incidence and mortality of cancer in nearly all sites (standardized incidence ratio, 0.38; 95% confidence interval, 0.26-0.53 for lung cancer in men) in the national data from Denmark, but not in the regional data for Nagasaki (Japan) and Honolulu (Hawaii).%2,3 A subsequent Danish study%4 by and large replicated these findings, although the magnitude of the reduction in this more recent cohort was attenuated.
Jablensky A, Lawrence D. Schizophrenia and CancerIs There a Need to Invoke a Protective Gene?. Arch Gen Psychiatry. 2001;58(6):579–580. doi:10.1001/archpsyc.58.6.579