Fasting but not casual blood glucose is associated with pancreatic cancer mortality in Japanese: EPOCH-JAPAN.

Abstract

The dose-response relationship between fasting blood glucose levels and risk of pancreatic cancer has been investigated, but the association between casual blood glucose levels and pancreatic cancer death has not been examined. We examined the association between casual and fasting blood glucose levels and death due to pancreatic cancer in Japanese. We performed a pooled analysis of the individual Japanese including 46,387 participants aged 40-79years from ten cohorts. Participants were classified into five groups: low normal, middle normal, high normal, prediabetes (casual blood glucose 140-199mg/dl, or fasting blood glucose 110-125mg/dl), and diabetes (casual blood glucose ≥200mg/dl, fasting blood glucose ≥126mg/dl, or anti-diabetic drug use). Low normal, middle normal, and high normal were defined according to tertiles of casual or fasting normal blood glucose levels. Hazard ratios (HRs) and 95% confidence intervals (CIs) for pancreatic cancer mortality were estimated stratifying casual and fasting blood glucose by cohort-stratified Cox proportional hazards regression analysis, with low normal (casual blood glucose <94mg/dl, or fasting blood glucose <90mg/dl) as a reference. Fasting blood glucose showed a dose-response relationship with pancreatic cancer mortality (p for trend = 0.005). After adjusting for covariates, HRs (95% CIs) were 2.83 (1.18-6.76) for prediabetes and 3.96 (1.56-10.08) for diabetes. However, there were no significant associations with casual blood glucose. These tendencies were observed after the exclusion of participants who were censored for the first 5years of follow-up. Fasting blood glucose is a better predictor of pancreatic cancer death than casual blood glucose.