March of Dimes says: "No level of alcohol use during pregnancy has been proven safe."

ACSH Newsletter "Priorities"
Volume 8 Number 1 1997
American Council on Science and Health

Should Physicians Recommend Alcohol to Their Patients? - No
by Dr. Albert B. Lowenfels

If physicians were to encourage their patients to drink alcohol, what patients would be the target group? Certainly not heavy drinkers, whose health, job and family may already suffer from alcohol abuse or addiction; the advice for these unfortunate individuals should be to reduce alcohol consumption or, preferably, to abstain entirely from alcohol.

Light and moderate drinkers need no encouragement to drink alcohol; instead, they need advice about safe levels for drinking, the dangers of drinking while driving or operating motorized equipment and, for females, the necessity for abstinence from alcohol prior to conception and during pregnancy.

The only target group, therefore, would be those patients who are nonconsumers of alcohol. Physicians would never advocate alcohol consumption for children, so our advice would be limited to nondrinking adults. The size of this group can be estimated as follows: There are currently about 200 million adults in the United States. Although the exact number of nondrinkers in that population is unknown, a good estimate is about 25 to 30 percent, or at least 50 million persons.

We know that this large group of nondrinkers includes many different subgroups. Some nonconsumers avoid alcohol because they already suffer from an alcohol-related disease. Others abstain because they have a chronic disease and have been advised to avoid alcohol. A third group may have an alcoholic parent and intuitively know they must avoid alcohol. A final group abstains from alcohol because of religious convictions. Clearly, it would be unwise to recommend light or moderate drinking to patients in any of these categories.

What about the residual group of nondrinkers who have no definite reason to avoid alcohol? Would their health improve if they began drinking? To give a thoughtful answer to this important question, we must first review the complex relationship between alcohol and health. What are the detrimental effects of alcohol consumption and what, if any, are its health benefits? This problem has attracted an enormous amount of interest: In the past few years thousands of articles have been published on alcohol and health.

Alcohol consumers are known to have increased risks of many diseases. These include cirrhosis of the liver; digestive-tract diseases such as ulcers or pancreatitis; several painful and often lethal cancers such as throat cancer, esophageal cancer and liver cancer; and certain neurologic disorders such as blackouts and seizures. In addition, all types of accidents, including fatal car crashes, are more frequent in drinkers than in nondrinkers. Finally, the fetal alcohol syndrome, now thought to be the most common cause of mental retardation, occurs only in the children of alcohol consumers. While it is true that some of these health problems occur primarily in heavy drinkers, any amount of alcohol may be hazardous for other diseases such as the fetal alcohol syndrome, for which a safe, lower limit is unknown.

There is only one well-recognized health "benefit" of alcohol consumption: Health professionals now agree that drinking small amounts of alcohol seems to reduce the risk of coronary heart disease. But is this single gain enough to balance the long list of alcohol-associated health problems?

We could find a convincing answer to the overall impact of alcohol on health if we were able to conduct the following experiment, a prospective randomized trial. Nondrinking adults would be randomly assigned to one of two groups: an alcohol-consuming group in which all the participants would be required to drink a daily glass of fruit juice spiked with about an ounce or two of alcohol, and a second, "control" group who would drink only fruit juice without alcohol. The two groups would be followed for 10 or 20 years so that we could compare the death rates in alcohol consumers to the rates in nonconsumers.

For various ethical and practical reasons, this experiment - which would give us badly needed information about the potential health benefits of light or moderate drinking - will never be performed. Therefore, to answer the "to drink or not to drink" question, we're forced to rely upon indirect, weaker evidence from nonrandomized trials - retrospective studies that look back at past alcohol exposure and cross-cultural studies that compare drinking levels and health status among different groups. These types of studies can be plagued by confounding and bias.

If we accept the premise that alcohol protects against certain types of heart disease, will we gain or lose by telling our nondrinking patients they should drink? We know that there are already at least 100,000 alcohol-related deaths each year in the United States. It is difficult to predict the number of heart disease deaths caused by alcohol abstinence, but the number has been estimated to be approximately equal to the number of alcohol-related deaths. Thus, a health policy of advocating light or moderate drinking for our abstinent patients would be unlikely to save many lives.

According to a report prepared for the Robert Wood Johnson Foundation, the cost of alcohol addiction for the year 1990 in the United States amounted to almost 100 billion dollars - higher than the estimated 67 billion dollars we spend for illicit drugs and the 72 billion dollars we spend for tobacco addiction. An unpredictable number of new alcohol consumers would eventually turn into heavy drinkers or become addicted to alcohol, requiring additional funds to cover the costs of their alcohol-related problems.

for moderate drinking often speak of the "French paradox." In the southwest of France - despite high consumption of foods rich in cholesterol, such as buttery sauces, various cheeses and goose liver - the risk of heart disease, particularly in men, appears to be lower than expected. According to moderate-drinking advocates, this "paradox" of a high-cholesterol diet and a low risk of heart disease can be explained by the beneficial, protective effect of copious amounts of alcohol - particularly red wine.

But men in France actually die about two years earlier than do men in Sweden or Norway, even though per capita alcohol consumption in Scandinavia is only about one third the consumption in France. Frenchmen, although they may not be dying of heart disease, are dying of other causes. Drinking alcohol does not guarantee longevity - and it certainly does not provide immunity against death!

And what has been the health experience of groups of individuals who have been lifelong abstainers? Do they die prematurely? Do they suffer from excess heart disease or other illnesses? Fortunately, such information is available from many reports reviewing the health of Seventh Day Adventists and Mormons - groups that abstain from alcohol on religious grounds. Their survival rates are generally higher than the American average. Avoiding alcohol does not interfere with an active, prolonged, healthy life.

From available statistics we know that there are more female than male nondrinkers. We also know that women are more likely to develop complications of alcohol, such as liver cirrhosis, at lower levels of alcohol intake than men. We therefore can predict that a policy of telling our nondrinking patients to begin drinking would be likely to yield more alcohol-related complications in women than in men.

There are many readily available nonaddictive drugs that effectively reduce the risk of coronary-artery occlusion. Why, then, should we recommend a drug that we know leads to loss of control or alcohol addiction in about 10 percent of users? It makes little sense to recommend alcohol as a safeguard against coronary heart disease when there are so many much safer drugs already at hand.

As we focus on the problem of alcohol and public health, we can learn a great deal by reviewing recommendations from organizations with recognized expertise in this area.

In 1991 the World Health Organization assembled a special review group to formulate worldwide alcohol policy. The group's conclusion on drinking and heart disease was this: "Any attempt to put across a message which encourages drinking on the basis of hoped-for gains in coronary heart disease prevention would be likely to result in more harm than benefit to the population."

In the United States, the National Institute on Alcohol Abuse and Alcoholism warns us that vulnerability to alcoholism and alcohol-related pathologies varies among individuals and cannot always be predicted before a patient begins to drink.

Finally, the Christopher D. Smithers Foundation, the largest private philanthropic organization devoted to research on alcoholism in America, does not advocate light or moderate drinking as a public health measure.

Over 2,000 years ago Hippocrates, one of our wisest physicians, reminded us, "Above all, do no harm." Let us remember this prudent advice as we decide what we should tell our patients about alcohol and health.

Albert B. Lowenfels, M.D., is a Professor of Surgery at New York Medical College.

Return to FAS: Fetal Alcohol Syndrome

Return to FAS Community Resource Center