LAUSR

Background The number of patients with diabetes and impaired fasting blood glucose in Korea is rapidly increasing compared to the past, and other metabolic indicators of population are also changed in recent years. To clarify the mechanism more clearly, we investigated the association between fasting blood glucose and incidence of pancreatic cancer in this retrospective cohort study. Methods In Korea National Health Information Database, 19,050 participants without pancreatic cancer in 2009 were enrolled, and followed up until 2013. We assessed the risk of incident pancreatic cancer according to the quartile groups of fasting blood glucose level (quartile 1: <88 mg/dL, quartile 2: 88–97 mg/dL, quartile 3: 97–109 mg/dL and quartile 4: ≥109 mg/dL). Multivariate Cox-proportional hazard model was used in calculating hazard ratios (HRs) and 95% confidence interval (CI) for incident pancreatic cancer. Results Compared with quartile1 (reference), unadjusted HRs and 95% CI for incident pancreatic cancer significantly increased in order of quartile2 (1.39 [1.01–1.92]), quartile3 (1.50 [1.09–2.07]) and quartile4 (2.18 [1.62–2.95]), and fully adjusted HRs and 95% CI significantly increased from quartile2 (1.47 [1.05–2.04]), quartile3 (1.61 [1.05–2.04]) to quartile4 (2.31 [1.68–3.17]). Conclusion Fasting blood glucose even with pre-diabetic range was significantly associated with the incident pancreatic cancer in Korean.