Alcohol intake, insulin resistance, and abdominal obesity in elderly men

Moderate and high alcohol intake have been associated with decreased and increased risk of type 2 diabetes, respectively. Insulin resistance, insulin secretion, and abdominal obesity are major predictors of diabetes, but the links with alcohol intake remain contradictory because of limited data.

In a population-based cohort of 807 men (age, 70 years), we studied whether alcohol intake was related to insulin sensitivity, measured with the gold standard technique (euglycemic clamp), insulin secretion (early insulin response), or adiposity [BMI, waist circumference (WC), waist-to-hip ratio]. Alcohol intake was self-reported (questionnaire) and was assessed from a validated 7-day dietary record. The cross-sectional associations were evaluated using multivariable linear regression, adjusting for smoking, education level, physical activity, dietary total energy intake, hypertension, diabetes, triglycerides, and cholesterol.

In multivariable models, self-estimated alcohol intake was not related to insulin sensitivity, early insulin response, or BMI, but was positively related to WC (beta-coefficient, 0.77; 95% confidence interval, 0.15 to 1.39; p=0.02) and waist-to-hip ratio (0.006 [0.002-0.009], p=0.003). The association with WC and waist-to-hip ratio was most pronounced in men in the lowest tertile of BMI. The results using dietary records were similar.

Evaluated in a large sample in elderly men, neither insulin sensitivity measured by clamp technique nor insulin secretion was significantly associated with alcohol intake. However, high alcohol intake was associated with abdominal obesity, which might explain the higher diabetes risk previously observed in high alcohol consumers.