Association of diabetes with survival among cohorts of Indigenous and non-Indigenous Australians with cancer.

BACKGROUND: The association between diabetes and cancer incidence has been well documented, but relatively little research has been undertaken on the potential influence of diabetes on cancer survival and the research that is available has produced inconsistent results. Because Indigenous Australians have a high prevalence of diabetes, we assessed survival, stratified by diabetes, among Indigenous Australian cancer patients. We also assessed survival, stratified by diabetes, amongst a cohort of non-Indigenous Australian cancer patients.
METHODS: All-cause survival and cancer-specific survival in diabetic versus non-diabetic cancer patients were assessed in Indigenous and non-Indigenous cohorts separately, using proportional hazards models.
FINDINGS: Indigenous cancer patients with diabetes (n = 140) had an overall survival disadvantage compared to Indigenous cancer patients without diabetes (n = 675) with all-cause Hazard Ratio (HR) = 1.4 (95% CI 1.1-1.8) adjusted for age, sex, and cancer site. After further adjustment to take into account the greater number of non-cancer deaths and co-morbidities in Indigenous cancer patients with diabetes, and their later stage at cancer diagnosis with less cancer treatment, there was no residual difference in cancer-specific survival compared to Indigenous cancer patients without diabetes (cancer-specific HR = 1.0, 95% CI 0.8, 1.3). Fewer non-Indigenous cancer patients had diabetes (n = 52) and they showed no differences in survival compared to their counterparts without diabetes.
INTERPRETATION: The poorer survival of Indigenous Australian cancer patients with diabetes was due to more non-cancer deaths, later stage at cancer diagnosis, less cancer treatment, and more co-morbidities than Indigenous Australian cancer patients without diabetes. In contrast, diabetes did not appear to affect survival in non-Indigenous Australians with cancer, either because there were too few to detect a moderate deleterious effect or because there was no association. Understanding the relation between diabetes and cancer treatment and survival is important because both diabetes and cancer are relatively common diseases, increasingly likely to co-exist.