Let us today examine whether diabetics may consume alcohol, particularly wine, safely, and, further, whether the health effects of drinking might be different for diabetics than for the rest of us. These are no small questions, for there are at least 16 million diabetics in the United States, about six percent of the population. This number is increasing rapidly. Most are adult-onset diabetics, who, at proper weight and on appropriate diet, can have their blood sugars well controlled without the use of insulin. This huge group is the one we shall discuss herein.

First, what about the effect of moderate drinking on regulation of diabetes? We need to look at what wine contains that might influence blood sugar. The short answer is not much. Dry wine contains negligible carbohydrate (virtually no sugar), protein, or fat, the usual caloric-containing compounds of food. It provides no more than two insignificant calories per four-ounce glass from all sources other than alcohol. It is the alcohol that brings dry wine its calories, which total about 84 per glass. (The worst caloric case for calorie-conscious consumers might be a Port, containing 20 percent alcohol, at seven calories per gram, and ten percent residual sugar, at four calories per gram -a four-ounce glass would supply 178 calories.)

Dry wine in moderation does not compromise diabetic control, and is certified permissible, at the usual advised rate, by medical specialists and by the American Diabetes Association. Even an occasional glass of sweet wine can probably be accommodated. All this presupposes that there exist no other medical problems that might be adversely affected, and that the diabetes is reasonably well controlled.

One strong caveat: all diabetics, most particularly those taking medicine to lower blood sugar, must eat when they drink, else they risk dangerous hypoglycemia (abnormally low blood sugar)a good rule for everybody.

Heavy drinking by diabetics risks more than just quantitative differences from moderate consumption. It courts metabolic disaster. Dont! In addition, destruction of the pancreas by the incitement of life-threatening inflammation (pancreatitis), a complication of alcoholism, may cause diabetes in some.

Now that we have established the safety of moderate consumption by most diabetics, we should try to determine whether the well-known health benefits derived there from apply to diabetics. If so, they might be of even greater importance, for the most clearly established of such benefits are reductions of risks of death and disability resulting from heart attacks, ischemic strokes, amputations, and other complications of atherosclerosis, the disease of the arterial blood vessels that preferentially afflicts diabetics. The risks of these disorders, causing the greatest proportion of deaths in the developed world among all adults, regardless of sex or race, are increased threefold in diabetics. Coronary disease, for example, contributes to death, often premature, in 70 percent of diabetics. Women with diabetes may suffer these adversities more than men. Until recently, we have had little specific information on whether drinking beneficially had little specific information on whether drinking beneficially had little specific information on whether drinking beneficially alters these risks among diabetics. Now we have some encouraging data.

A World Health Organization study conducted by Diem et al, published in a supplement of Diabetes, volume 48, in 1999, demonstrated a reduction in coronary deaths associated with moderate alcohol consumption among diabetics. There were few women in the study group.

More impressive data came from the University of Wisconsin Madison Medical School the same year. Charles Valmadrid, et al., studied 983 older-onset diabetic individuals over a more than twelve-year period. Their mean age was 68.6; 55 percent were female. Compared with those who never drank, the risk of death from coronary heart disease was reduced by 31 percent among former drinkers, 46 percent for those who drank less than one drink per week, 56 percent for those who averaged between 1/5 and one drink daily, and an astonishing 79 percent for consumers of 1 1/4 or more drinks per day. Both the magnitude of the benefit and the benefit of very little alcohol were surprising. This work was published in the Journal of the American Medical Association on July 21, 1999.

The Nurses Health Study is a prospective study of 121.700 female nurses in eleven states that began in 1976. In the August 1, 2000, issue of Circulation, C.G. Solomon et al from the Brigham and Womens Hospital and Harvard School of Public Health, both in Boston, reported on the risk of fatal and non-fatal coronary heart disease among 5,103 women with diabetes. They had found results that paralleled those cited above. Women who usually drank less than one-half a drink daily had 26 percent less coronary disease than women who abstained. Women who usually drank more had 52 percent fewer coronary events than abstainers. Remarkably included in the first of these two groups were women who drank, on average, as little as the equivalent of less than 1/25 of an ounce of wine per day.

The companion Physicians Health Study has followed 87,938 American male physicians (2790 diabetic) since 1983, particularly with regard to factors that might influence the risks of coronary heart disease and cancer. For the particular report we review now (Ajani, et al., in Circulation of August 1, 2000), 850 deaths from coronary disease occurred during an observation period of five and a half years. Again, alcohol passes easily. Among the diabetic physicians, there was reduction in risk of death from heart attack of 33 and 58 percent, respectively, for individuals reporting weekly and daily alcohol consumption, as compared to those who never or rarely drank. Monthly drinkers showed no benefit, perhaps a statistically insignificant worse outlook. Each consumption group of diabetics appeared to benefit significantly more than non-

diabetics (33 vs. 18 percent reduction of coronary deaths for weekly drinkers, 58 vs. 39 for daily). I suppose the diabetics had more to correct.

Whats at work here? No doubt the same mechanisms of benefit are at play as for reductions of cardiovascular risks among non-diabetics: increased high-density lipoprotein (the "good cholesterol", inhibition, of oxidation of low-density lipoprotein ("bad cholesterol") to its most noxious form, reduction of excessive blood clotting, promotion of healthier blood vessel walls. Maybe some of these are of greater magnitude among diabetics. It is not yet clear whether additional benefits specific to diabetes exist. Among considered candidates are favorable action of alcohol upon insulin and sugar regulation and, surprisingly, a complicated benefit in which acetaldehyde, the ordinarily highly toxic first breakdown product of alcohol, participates.

William S. Weintraub, editorialist in the same issue of Circulation in which the last two cited articles appeared, in discussing "Alcohol Consumption, Diabetes, and Coronary Disease," cautiously concludes:

Thus, from a public policy point of view, it would seem that moderate alcohol consumption is safe and may be beneficial from the point of view of cardiovascular risk in selected populations, both diabetic and non-diabetic. However, there is probably not sufficient evidence to recommend alcohol consumption to decrease risk in any population.

The authors of the reports we reviewed conclude that moderate alcohol consumption is associated with reduced risk of cardiovascular disease in diabetics, as well as non-diabetics, and should not be routinely discouraged.

Harvey Finkel is Clinical Professor at The Boston University Medical Center, Chairman of the Committee on Health of the Society of Wine Educators , and a Member of the AIM Editorial Board