This paper was coauthored by an international group of researchers—Kaye Middleton Fillmore (U.S.), William C Kerr (U.S.), Tim Stockwell (Canada), Tanya Chikritzhs (Australia), Alan Bostrom (U.S.)—whose research challenged an "established fact" accepted by most scientists and the public at large. The "established fact" was that the use of alcohol prevents the incidence of coronary heart disease. We believe that citations of our paper result from our paper re-opening the debate on this issue—and also because the extent of alcohol’s cardioprotective effect is at the core of debates around contemporary alcohol policy.

“...our paper says that the science behind alcohol’s purported health benefits is not reliable and more sophisticated methods are required to establish to what extent moderate drinking is good for heart health.”

We used meta-analysis to evaluate whether there was a systematic misclassification error in 54 published studies evaluating all-cause mortality and coronary heart disease mortality. The great majority of these suggest that abstainers were at higher risk for premature death, meaning, in turn, that moderate drinkers were "protected" against coronary heart disease. The errors we evaluated were whether former drinkers and occasional drinkers were combined with long-term abstainers.

The rationale was that these two groups contained a great many people whose drinking had declined or terminated because, with increasing age, they had become ill or frail or were using medications. Our meta-analysis suggested that, when studies were without the error evaluated, alcohol did not "protect" against premature death for coronary heart disease or death in general. As such, the paper attempts to redefine appropriate methodology in alcohol epidemiology and questions what had become a widely accepted synthesis of existing knowledge.

First and foremost, people don’t have to think they need to drink alcohol to prevent premature death. There are other alternatives to "thinning the blood" like an aspirin a day. Second, the paper has economic implications. For example, the alcohol industry has used the notion that drinking alcohol imparts specific benefits in order to help sell its product. In summary, though, our paper says that the science behind alcohol’s purported health benefits is not reliable and more sophisticated methods are required to establish to what extent moderate drinking is good for heart health.

We became interested in the research because we had performed other research (not in this area) that suggested people’s drinking declined with increasing age and the onset of ill health. We also had read a marvelous paper published in Lancet on December 3^rd, 1988 by A. Gerald Shaper, Goya Wannamethee, and Mary Walker: "Alcohol and mortality in British men: explaining the U-shaped curve," suggesting this hypothesis.

We felt it had never been properly evaluated. We had trouble finding funding for the project and publishing the paper primarily because we were testing something that was regarded as "established fact." Eventually, funding was found from an Australian source, the Alcohol Education and Rehabilitation Foundation, a charitable foundation established by in the Commonwealth government in 2001.

We hope our research stimulates other researchers to more carefully evaluate these kinds of errors in their studies or, at the least, to be very sensitive to them. We also hope to extend our research to other health effects associated with alcohol consumption, including cancer, and a host of other health conditions, which purportedly benefit from moderate alcohol consumption (e.g., the prevention of dementia and a cure for the common cold).

As mentioned, there are personal and economic implications. There are also political ones in that the notion of alcohol being a health "benefit" has stood in the way of social policies seeking to control the use of alcohol, the latter to decrease its many harmful effects.

From a political point of view, governments obtain substantial tax revenue from alcohol and, collectively, the various manufacturing and retail arms of the alcohol industry are potent forces. The extent to which alcohol is good for our health and at what dose, is a fundamental element to consider when societies decide how to regulate the availability of this most popular of recreational drugs.