The study shows that alcohol seems to confer only a slight advantage in moderate drinkers.
To date, the effects of light to moderate alcohol consumption on the risk of developing adult onset diabetes remain unclear.
The researchers monitored over 5000 men between the ages of 40 and 59 for almost 17 years.
None of the men had a history of coronary heart disease, diabetes, or stroke at the start of the study.
The researchers assessed how much the men drank regularly, as well as the amount of fat (lipids), glucose, and insulin in their blood.
During this time, 198 of the men developed type II diabetes.
Heavy drinkers, who regularly consumed over 6 units a day, were at the highest risk of developing the disease, which seemed to be due to contribution of alcohol to excess weight gain.
Nonetheless, after taking into account the effects of smoking, exercise, and undiagnosed heart disease, moderate drinkers who regularly drank between 3 and 6 units daily had a significantly lower risk of becoming diabetic than occasional drinkers did.
| Alcohol showed little protective effect in men at low risk of developing diabetes.
| Journal of Epidemiology and Community Health |
Alcohol was associated with lower levels of insulin and higher levels of high-density lipoprotein (HDL or "good") cholesterol.
In heavy drinkers alcohol increased HDL by a factor of 8, and reduced insulin by 40%.
Yet this was the very group most at risk of developing diabetes, possibly because of the calorific content of alcohol adding to weight gain and its toxic effect on pancreatic cells needed to regulate insulin.
In moderate drinkers, the effects of alcohol on HDL cholesterol and insulin seemed to account for only 20% of the reduction in risk of developing diabetes.
Further analysis showed that the "protective effect" of alcohol was most evident in those at greatest risk. Those at greatest risk were heavier, current smokers, and those with higher insulin and glucose levels and lower HDL cholesterol at the start of the study.
Little protective effect was seen in men at low risk of developing the disease.
The authors conclude that the complex association between alcohol, insulin, and blood lipids needs to be explored.
It might be that other factors in a moderate drinker's lifestyle might account for some of the protective effect, they suggest.
However, they say there is no justification for encouraging light drinkers to increase their intake, or for non-drinkers to take up drinking in the hope of protecting themselves against diabetes.