LAUSR

Simple Summary Hyperglycemia has been reported to increase the risk of pancreatic cancer, while the association between temporal glycemic changes and the risk of pancreatic cancer remains uncertain. This study used data from a prospective cohort and found a U-shaped association between annual change in fasting blood glucose and risk of pancreatic cancer in males. A higher coefficient of variance and a greater range of fasting blood glucose were also associated with the increased risk of pancreatic cancer. These findings may help identify individuals at high risk for pancreatic cancer; thus, appropriate measures can be taken to prevent pancreatic cancer. Abstract Hyperglycemia has been reported to increase the risk of pancreatic cancer (PC), while the association between glycemic change and PC risk has rarely been explored. Using data from a prospective cohort study conducted in China since 2006, 138,870 males with available fasting blood glucose (FBG) levels, including 106,632 males with at least two FBG measurements, were analyzed. The associations between FBG (level, change, and stability) and PC incidence were evaluated using Cox proportional hazard regression and restricted cubic splines. Baseline (p = 0.109) and recent (p = 0.070) FBG levels and incident PC were not significantly associated. U-shaped associations were observed between the annual FBG change and PC risk. Compared with stable FBG, participants with annual FBG change rates <−0.05 mmol/L or >0.15 mmol/L had about four-fold (HR, 4.010; 95% CI: 1.920–8.375) and six-fold (HR, 5.897; 95% CI: 2.935–11.848) higher PC risks, respectively. The PC risk increased by 2.5% (HRlinear = 1.025, 95% CI:1.009–1.042) for every 1% increase in the coefficient of variation for FBG. A subgroup analysis of males without diabetes at baseline showed stronger associations. Temporal FBG changes may be an important factor for identifying populations with high PC risks.