Incidence of Diabetes Mellitus in Male Moderate Alcohol Drinkers: A Community-Based Prospective Cohort Study

Background/Aim. Although alcohol consumption is known to affect the incidence of diabetes mellitus (DM), reports on the effects of moderate alcohol consumption on DM incidence have been inconsistent. This community-based prospective cohort study was performed to investigate the incidence of DM in male Korean moderate alcohol drinkers. Methods. The Ansan and Ansung cohort was used for the analysis. The study population included a total of 3,492 men with no history of DM. The subjects were classified as mild (1-14 g/d), moderate (15-29 g/d), and heavy (>= 30 g/d) drinkers based on their amount of alcohol consumption. The incidence rates of DM in the three groups were compared and analyzed over a 10 year follow-up period. Results. The hazard ratios (HRs) for DM incidence were 25.12 (95% confidence interval [CI], 21.73-28.90) per 1,000 person years (PY) in mild drinkers, 31.13 (26.11-36.83) per 1,000 PY in moderate drinkers, and 31.68 (26.81-37.18) per 1,000 PY in heavy drinkers (p for trend, p = 0.043). Multivariate regression analysis showed that the HRs (95% CI) for DM were 1.25 (0.97-1.61, p = 0.086) in moderate drinkers and 1.30 (1.01-1.68, p = 0.045) in heavy drinkers compared to mild drinkers. The changes in pancreatic insulin secretion were more remarkable than those in insulin resistance in all three groups. Conclusions. The incidence of DM in male Korean moderate drinkers did not increase significantly over the observation period. However, the incidence of DM tended to increase with increasing alcohol consumption. Pancreatic insulin secretion may play a more important role than insulin resistance in the relationship between alcohol and incidence of DM. (C) 2019 IMSS. Published by Elsevier Inc.