Nope, "moderate" alcohol consumption isn't good for you: Study

In a landmark systematic review of existing medical literature, researchers publishing in the Journal of Studies on Alcohol and Drugs have found that previous research finding that moderate alcohol consumption had health benefits was badly flawed.

Conventional wisdom for the last several decades, based on earlier research, has held that drinking moderately — on the order of four or five alcoholic drinks a week for adults — has some health benefits and may even be associated with longer lifespans. But that view has increasingly come into doubt, and scientists who examined 107 published studies on alcohol use and health now report that many of those included a key methodological error, one that seriously undermined their conclusions.

These studies focused on older adults in an effort to demonstrate a link between alcohol and longevity, the researchers report, and many such studies did not distinguish between adults who now drink moderately, but previously drank little or not at all, and those who now drink moderately after a lifetime of potentially damaging drinking habits. This difference, researchers claim, is crucial: "Lower quality" studies that did not ask self-described moderate drinkers about their previous drinking habits appeared to show that drinking was connected to a longer lifespan, whereas "higher quality" studies that controlled for subjects' past drinking histories did not show any such correlation.

Those higher-quality studies also had a mean cohort age of 55 years or younger and followed up with their subjects past age 55, adding to their methodological rigor. The 107 studies reviewed by the authors covered the experiences of 4,838,825 participants, including 425,564 recorded deaths. 

“If you look at the weakest studies, that’s where you see health benefits.” lead researcher Tim Stockwell, a scientist with the Canadian Institute for Substance Use Research at the University of Victoria, said in a statement.

Widespread misconceptions that limited or moderate alcohol use is healthy could be harmful, the researchers argue.

"Studies with life-time selection biases may create misleading positive health associations," the authors write. "These biases pervade the field of alcohol epidemiology and can confuse communications about health risks."