A prospective population-based study of alcohol use and non-insulin-dependent diabetes mellitus.

The effect of alcohol use on risk of non-insulin-dependent diabetes mellitus was assessed in a defined population. A 24-hour recall of alcohol intake and past-week alcohol intake were obtained by separate interviewers between 1973 and 1975, and responses were coded by the Nutrition Coordinating Center, University of Minnesota. Of the 524 adults aged 30-79 years without diabetes at baseline, 31 men and 44 women were identified as diabetic by means of a glucose tolerance test (World Health Organization criteria) between 1984 and 1987. Men, but not women, who developed diabetes reported significantly more alcohol intake in the past week and in the past 24 hours. The highest rate of diabetes among alcohol users was in heavy drinkers--statistically significant only in men. Alcohol use remained a significant predictor of diabetes in men after adjustment for baseline age, body mass index (weight (kg)/height (m)2), cigarette smoking, family history of diabetes, and systolic blood pressure with a logistic regression model. The relative risk associated with past-week alcohol intake was 1.5 per 137.8 g; for past-24-hour alcohol intake, it was 1.5 per 24.5 g. Adjustment for baseline fasting plasma glucose and triglycerides did not change the results. Alcohol intake appears to be associated with risk of non-insulin-dependent diabetes mellitus in men.