Association between alcohol consumption and pancreatic cancer risk differs by glycaemic status: A nationwide cohort study

Background: The dose–response association between alcohol consumption and the subsequent pancreatic cancer risk by individuals’ glycaemic status is unclear. Research design and method: This large-scale nationwide cohort study included 9,514,171 adults without cancer who underwent health examinations under the Korean National Health Insurance Service in 2009 and were followed-up until December 2017 for pancreatic cancer development. Multivariable Cox proportional hazards regression analysis was performed. Results: During a median follow-up period of 7.3 years, 12,818 patients were newly-diagnosed with pancreatic cancer. Among individuals with normoglycemia, a J-shaped association was observed between the frequency of alcohol consumption (1–2 and ≥5 days/week: hazards ratio [HR]; 95% CI, 0.91; 0.85–0.97 and 1.13; 1.002–1.27, respectively) and pancreatic cancer risk, after adjusting for potential confounders. However, in patients with impaired fasting glucose (IFG), pancreatic cancer risk increased with increased frequency and average daily amount of alcohol consumption (all P for trend <0.01). IFG combined with heavy alcohol consumption (30 g/day) was associated with 38% increased pancreatic cancer risk (HR, 1.38; 95% CI, 1.23–1.54). Diabetes was associated with an increased pancreatic cancer risk regardless of alcohol consumption and 70% increased risk even in non-drinkers (HR, 1.70; 95% CI, 1.61–1.80). Conclusions: The J-shaped dose–response association between alcohol consumption and pancreatic cancer risk was observed only in individuals with normoglycemia, not in patients with IFG and diabetes. Complete alcohol abstinence may help reduce pancreatic cancer risk in patients with IFG and diabetes.