Purpose: There is great interest in whether type 2 diabetes and its treatments alter breast cancer risk and prognosis, but previous studies are inconclusive. We conducted a cohort study within the UK General Practice Research Database to investigate associations of type 2 diabetes and patterns of diabetes treatment with breast cancer risk and all-cause mortality.
Methods: We identified 52,657 women with type 2 diabetes, diagnosed between 1987 and 2007, and 30,210 randomly selected women without diabetes. We performed a time-dependent analysis using Cox proportional hazards models.
Results: Diabetes was associated with a 29 % increased overall breast cancer risk (95 % CI: 1.16-1.44), but the association markedly attenuated when adjusted for age, period of cohort entry, region, and body mass index (BMI) (HR: 1.12; 95 % CI: 0.98-1.29). Women with breast cancer and pre-existing diabetes had a 49 % (95 % CI: 1.17-1.88) increased all-cause mortality risk compared with women with breast cancer but without diabetes, after controlling for age, period, region, BMI, smoking, alcohol, and deprivation. Compared with sulfonylurea, we found weak evidence that metformin monotherapy (HR: 1.04; 95 % CI: 0.79-1.37) and insulin (HR: 1.33; 95 % CI: 0.63-2.83) modified breast cancer risk among women with diabetes.
Conclusions: We found weak evidence that diabetes is associated with a small increased risk of breast cancer. Among treated women, there is no evidence that anti-diabetes treatments modify the risk of developing breast cancer, with wide confidence intervals indicating imprecise effect estimates. Women with breast cancer and diabetes, however, had an increased all-cause mortality risk highlighting the potential importance of maintaining adequate glycemic control alongside anti-cancer treatments and subsequent follow-up.