Will the NIH trial on alcohol and health answer the question once and for all? Maybe not

What used to be accepted as gospel – that moderate drinkers are healthier than nondrinkers or heavy drinkers – has been challenged in recent years, and a new study to be conducted by the National Institutes of Health aims to settle the question once and for all. The study plans to enroll about 8,000 volunteers aged 50 or older from around the world, who will be assigned to avoid drinking or have one drink per day for 6 years. The lack of such large scale prospective studies is one reason why the question of alcohol’s influence on health and longevity remains subject to debate. However I am not sure the study will yield the answers it seeks to, but not for the reasons others are already finding to criticize the project.

 It’s an ambitious undertaking, with an equally ambitious price tag of US$100 million. The plan is for most of the money to come from the alcoholic beverage industry through grants, and $68 million has reportedly already been pledged. Skeptics point out that many of the study’s investigators have accepted money from alcohol industry groups in the past, and raise the issue of scientific influence peddling. But the study’s principal investigator, Dr. Kenneth Mukamal of Harvard, has no such ties and has authored numerous research articles on alcohol and health. As someone who has been a clinical investigator myself, I can vouch for the many levels of independent oversight that prevent any influence by investigators on the results. There's just no way to slip things through, and no reason to risk one's reputation by attempting to.

The challenges to the long held view that alcohol in moderation has a net positive benefit come largely from different types of studies, such as the one I critiqued in last month’s blog. These studies aggregate previously done observational studies (as opposed to interventional studies, like the NIH trial). The example I used last month essentially “cherry picked” the studies that met the criteria that would yield the desired findings. It's not new research, just old research repackaged to imply a different interpretation - one that fit the goals of an alcoholism treatment center.

But conflicts of interest are not the problem with the NIH trial, it is the complexities of human behavior. Volunteers willing to give up drinking will come from a specific subset of individuals that may not be representative of the average consumer. Allowing people to choose their beverage or to vary which one they drink could make it difficult to see if wine drinkers, for example, are different (evidence suggests that wine drinking is not the same as other types of drinking.) It may be that more than one drink per day is the optimal dose, and the effect too small to be statistically significant with only one. Another issue is the pattern of drinking; with wine, it is the relationship with meals. And how will the one drink per day rule be enforced? What of people who for whatever reason choose not to have a drink, or make up for it by skipping a couple of days then having several?

I do applaud the effort required to do this study, but I expect that whatever the results they will be challenged by those with an opposing ax to grind.