Subsequent cancer risk in the incidence cohort of Rochester, Minnesota, residents with diabetes mellitus.

Subsequent cancer incidence was determined in the population-based incidence cohort of Rochester, Minnesota, residents diagnosed with diabetes mellitus between 1945 and 1969. The relative risk of having cancer, excluding cervical and non-melanoma skin cancers, was not significantly increased following the diagnosis of diabetes mellitus. The potential biases of increased medical surveillance among diabetics and exacerbation of subclinical diabetes by occult malignancy did not appear to be important except in the case of subsequent pancreatic cancer.