Why I am not surprised that the NIH cancelled the alcohol-health study

Not long after enrolling the first patients in the much hyped prospective study on alcohol and health, the National Institutes of Health recently announced that they were pulling the plug. I am actually more surprised that they ever got it off the ground in the first place. As I wrote a year ago when the study was still in its planning stages, there were too many competing interests, criticisms of the study design, and concerns about funding to expect that whatever results came out would be universally accepted. Nevertheless, I am disappointed.

The study, called Moderate Alcohol and Cardiovascular Health Trial (MACH) was intended to provide hard evidence about the health effects of moderate alcohol consumption by prospectively assigning subjects with heart disease to one drink per day or not drinking, which they were to follow for up to 10 years. Most existing data on the question is retrospective, or simply tracks a subject population according to their drinking preferences, which is known to be inaccurate and variable. There are many issues with these types of studies so the results have been subject to debate. Prospective randomized studies like the MACH trial with objectively measurable behaviors and outcomes are designed to provide the highest level of evidence.

Follow the money: Where the trouble started

Large scale prospective clinical trials are expensive to do however, and that is where some of the trouble started: most of the projected $100 million in funding was to come from alcohol companies including Anheuser-Busch InBev, Carlsberg Breweries A/S, Diageo (maker of Tanqueray gin and Captain Morgan rum) and Pernod Ricard USA (maker of Absolut Vodka and Glenlivet Scotch whisky). The funds were to come through donations to a nonprofit entity called the Foundation for the National Institutes of Health (FNIH), which was created by congress as a sort of firewall between donors and the studies they subsidize. While it is usually reasonable to assume that the FDA-registered study design had adequate safeguards about sponsors influencing the results, concerns were raised about communications between the organizers of the study and its industry underwriters, bypassing the FNIH. Whether real or not, the appearance of conflict of interest was enough for the NIH advisory committee to recommend cancellation.

Was the whole mess truly a scandal or a misunderstanding? One report lambasted the study organizers, calling the trial’s demise a “case study” in how to conduct science (and how not to.) They accused the researchers of designing the trial to get a favorable result for moderate drinking, by not following the subjects long enough to find a potential increase in cancer (10 years not long enough?) But that was not the hypothesis being tested; clinical trials are not supposed to be random stabs in the dark, but attempts to validate a concept that already has support. 

I think there is a good chance that they would have found support for the role of moderate consumption in prevention of diabetes and heart disease, but I am more interested in the specific role of wine. As I expounded on last month, it is convergence of drinking patterns - essentially treating all types of alcohol consumption as equal – that may be responsible for diluting the French Paradox effect. What I would like to see is a prospective trial on just wine. Wine with dinner, wine the traditional French way; just don’t ask me to be in the control group.