Abstract P365: The Beneficial Effect of Moderate Alcohol Consumption on Type 2 Diabetes Risk is Modified by Smoking Status and BMI in The French E3n Cohort
Background - A U-shape relationship between alcohol consumption and risk of type 2 diabetes (T2D) has already been suggested. While smoking has been shown to act synergically with alcohol, with improvement of insulin sensitivity in ex-smokers, little is known about a possible interaction between alcohol and body mass index (BMI) on T2D risk.
Material & Methods - We investigated the relationship between alcohol consumption (categorized as non-consumers, 2 drinks/day, 1 drink=10g of ethanol) and T2D among 66,118 women from the French prospective E3N cohort, the French component of EPIC. Multivariate Cox models were performed to estimate hazard ratios (HR) and 95% confidence intervals (CI) and were then stratified by BMI and smoking status (non-, former and current smokers). Between 1992 and 2007, 1,369 cases of incident diabetes were identified and validated using a drug reimbursement dataset and a supplementary questionnaire.
Results - The median alcohol intake among consumers was 0.8 drink/day. Compared to non-consumers, women in the [1-2 drinks/day] category had a significantly lower rate of T2D (HR=0.81 [0.66-0.99]). Higher or lower intakes were not significantly associated with risk of T2D but the pattern of HRs was U-shaped across categories (HR=1, 0.98, 0.94, 0.81 and 0.95 respectively; P=0.04 for the quadratic term). The inverse association between moderate drinking and T2D risk was restricted to overweight or obese women (HR=0.73 [0.57-0.94]). The association was strongest among former smokers (HR=0.60 [0.41-0.87]).
Conclusion - The observed decrease in T2D risk associated with moderate alcohol consumption may be limited to overweight or obese women. Our findings also support the hypothesis of a combined beneficial effect of moderate alcohol consumption and smoking cessation on T2D risk.
- © 2012 by American Heart Association, Inc.