Glucose tolerance status and 20 year cancer incidence.

BACKGROUND Type 2 diabetes, an extreme state of glucose intolerance, has been found to be associated with cancer mortality; less is known about impaired glucose tolerance and cancer incidence. OBJECTIVES To examine the association between fasting and post-load plasma glucose and insulin, and the 20 year incidence of cancer. METHODS We followed a sample of the Jewish Israeli population (n = 2780), free of cancer at baseline, from 1977-1980 to 1999 for cancer incidence and mortality. Baseline fasting and 1 and 2 hour post-load plasma glucose levels were recorded, as was insulin, in 1797 of them. RESULTS During 20 years, 329 individuals (11.8%) developed cancer. Cancer incidence for all sites differed between men and women (13.0% and 10.7%, P = 0.03), and among different glucose tolerance status groups (P = 0.01). Cancer incidence hazard ratio, by glucose status adjusted for gender, age, ethnicity, smoking and body mass index, was 1.24 (95% CI 0.96-1.62, P = 0.10) for impaired fasting glucose or impaired glucose tolerance, and 1.32 (95% CI 0.96-1.81, P = 0.09) for type 2 diabetes mellitus, compared with those who were normoglycemic at baseline. Fasting insulin and cancer incidence were not associated. CONCLUSIONS An increased long-term cancer risk for individuals with impaired fasting glucose or glucose tolerance, or diabetes, is suggested. Even this modest association could have substantial public health consequences.