Cancer Mortality Among Diabetics2

Cancer mortality was ascertained among 21,447 diabetic patients who registered at a large diabetes clinic in Boston during a 26-year period ending in 1956. Observations were made through 1959. Analysis was by the calculation of standardized mortality ratios in which observed numbers of deaths were compared with numbers expected if the age- and sex-specific rates in the general Massachusetts population observed over the same period of time applied. Total cancer mortality among the female diabetics occurred at approximately the expected frequency. By contrast, the male diabetics experienced a significantly reduced risk of cancer death that persisted even after the exclusion of deaths from lung cancer. This decreased cancer risk was believed attributable to two factors: a) an especially excessive risk of deaths from other causes, excluding cancer and b) an over-representation of Jews among the patients. The positive association between cancer and diabetes, reported in many clinical studies, is believed to be explained by the well-documented metabolic differences between normal and malignant tissues and by the failure of these studies to identify the antecedent disease. That diabetics have a significantly increased risk of death from pancreatic cancer was confirmed by this study. It is suggested that the carcinogenic potentialities of exogenous animal insulins be investigated, particularly in view of their known antigenic and teratogenic properties in laboratory animals. This study offers evidence that diabetic females do not have increased risk of death from cancer of the uterine corpus. Earlier clinical reports of a high prevalence of diabetes among endometrial cancer patients almost invariably were based on retrospective observations which made it difficult to identify with certainty the antecedent disease. Mortality from lung cancer was reduced significantly among the male diabetics, but not among the females. An over-representation among the Jewish male patients (with their known diminished lung cancer risk) is believed responsible, at least in part. However, reduced lung cancer risk among diabetics independent of religion cannot be excluded. The possibility of differences in the cigarette smoking habits of diabetics and nondiabetics could not be examined.