Pancreatic cancer is one of the most fatal malignancies, with a 5-year survival of <8% and is the 4th leading cause of cancer death. It is also one of the cancers most tightly linked to diabetes even when recent diabetes diagnosis and reverse causation are ruled out. The highest rates of pancreatic cancer are found among blacks in the U.S., yet despite nearly twice the prevalence of diabetes among blacks compared to whites, as well as blacks’ increased prevalence of obesity, another key risk factor, little is known about the interrelationship between diabetes and pancreatic cancer among blacks and whether this accounts for the observed racial disparity. We tested the relationship of diabetes with the incidence of pancreatic cancer and whether this relationship varied between blacks and whites. Our null hypothesis was that diabetes would not differentially increase risk among blacks compared to whites. Using the Southern Community Cohort Study, we conducted a cancer follow-up (2002-2015) of a cohort of mostly low income black and white participants aged 40-79 with (n=17,644) and without diabetes (n=64,870) at cohort entry. Mean age and diabetes duration of those with diabetes was 55.0 and 9.5 years, respectively. Mean age of those without diabetes was 51.6 years. Logistic regression was used. A total of 307 incident cases of pancreatic cancer were observed. In analysis controlling for age, sex, race, current and former smoking, and BMI, diabetes was a significant risk factor for pancreatic cancer (OR=1.48, 95% CI= 1.13-2.94). In race-stratified analyses, the increases associated with diabetes were seen both among blacks (OR= 1.46, 95% CI=1.06-2.02) and whites (OR=1.60, 95% CI= 0.95-2.67). After control for diabetes, the black excess in risk of pancreatic cancer was attenuated (OR=1.25, 95% CI=0.96-1.64). The findings suggest that diabetes affects pancreas cancer risk regardless of race, but that the higher prevalence of diabetes among blacks contributes to their elevated risk of pancreatic cancer.


B. Conway: None. W.J. Blot: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at