Background: This study examines the relation between alcohol and type 11 diabetes and the possible mediating effects of HDL-cholesterol and serum insulin.
Methods: Prospective study of 5221 men aged 40-59 years with no history of coronary heart disease, diabetes, or stroke drawn from general practices in 18 British towns,
Results: During the mean follow up of 16.8 years there were 198 incident cases of type 11 diabetes. Occasional drinkers were the reference group. A non-linear relation was seen between alcohol intake and age adjusted risk of diabetes, with risk lowest in light and moderate drinkers and highest in heavy drinkers (quadratic trend p=0.03). Further adjustment for body mass index decreased risk in heavy drinkers. After additional adjustment for physical activity, smoking, and (undiagnosed) pre-existing coronary heart disease, only moderate drinkers showed significantly lower risk than occasional drinkers (RR=0.66 95% CI 0.44 to 0.99). Alcohol intake was inversely associated with serum insulin and positively associated with HDL-cholesterol. Adjustment for these factors reduced the "protective" effect in moderate drinkers (adjusted RR=0.73 95% CI 0.48 to 1.10) but the quadratic trend remained significant (p=0.02).
Conclusion: There is a non-linear relation between alcohol intake and the risk of type 11 diabetes. Serum insulin and HDL-cholesterol explained a small amount (20%) of the reduction in risk of type 11 diabetes associated with moderate drinking. The adverse effect of heavy drinking seemed to be partially mediated through its effect on body weight.