Roni Rabin has written an important article in the New York Times (June 16, 2009) about the purported relationship between moderate alcohol consumption and a person’s overall health. She wonders if moderate drinking is really good for you. Given the countless studies that have demonstrated its benefits, how can anyone seriously entertain this question anymore? Well, Rabin does. She says what if it’s all a big mistake? “For some scientists, the question will not go away. No study, these critics say, has ever proved a causal relationship between moderate drinking and lower risk of death—only that the two often go together.”
This is the reason I find her article so important. It gets to the nub of the many studies reported in the media about the association between a behavior, drug, or food product and another physiological or behavioral measure. For example, it is sometimes claimed that daily dose of aspirin decreases the risk of a heart attack. Or that excessive coffee consumption decreases bone mineral density and increases the risk of osteoporosis. And that eating a goodly number of fruits and vegetables decreases the risk of lung disease. The list of such claims seems limitless.
But it is essential to remember the nature of the evidence upon which they are made. Almost all are derived from correlational data in which one variable is found to go together with another with varying degrees of probability. However, it is impossible to claim a causal relationship with such data. We have no idea, for example, if the relationship is due to any number of third variables can account for the relationship.
Consider, for example, the claim that moderate alcohol consumption may promote cardiovascular health and reduce the risk of diabetes and dementia. Could it be that the moderate drinkers differ from those who abstain? Rabin suggests, “It may be that moderate drinking is just something healthy people tend to do, not something that makes people healthy.”
She quotes Kaye Middleton Fillmore, a sociologist from the University of California Medical School in San Francisco: “The moderate drinkers tend to do everything right—they exercise, they don’t smoke, they eat right and they drink moderately. It’s very hard to disentangle all of that, and that’s a real problem.”
The fact that abstainers and moderate drinkers have widely different life styles is crucial in understanding why one must be cautious in claiming a causal relationship in this case or, by extrapolation, to any statistical correlation, regardless of its size.
Rabin adds to the plausibility of this interpretation by citing the data of Dr. Naimi of the C.D.C. who reports that moderate drinkers and abstainers are “so different that they simply cannot be compared. Moderate drinkers are healthier, wealthier and more educated, and they get better health care, even thought they are more likely to smoke. Moderate drinkers tend to be socially advantaged in ways that have nothing to do with their drinking. These two groups are apples and oranges.”
Rabin reminds us that many researchers are haunted by the recently discovered mistakes in studies of hormone replacement therapy, “which was widely prescribed for years on the basis of observation studies similar to the kind done on alcohol.”
The fact remains that researchers have yet to carry out the kind of randomized, controlled experiment in which a group of abstainers are randomly assigned either to get a daily dose of alcohol or not and then closely monitored for several years to determine its effects on indices of health. Comparable designs can be imagined for other subgroups of individuals, such as those who are at risk for cardiovascular disease or diabetes. Such designs are uniformly required before the introduction of any new pharmaceutical agent in this country.
She concludes her article by quoting Dr. Sei Lee, again from the University of California Medical School in San Francisco who recently proposed a large randomized clinical trial on alcohol and health. “But this is a really important question, because here we have a readily available and widely used substance that may actually have a significant health benefit—but we just don’t know enough to make recommendations”
