Wine-specific Mortality Benefits Disappear When Studied Appropriately
Some have suggested that wine might be more beneficial to health than other types of alcoholic beverages because of compounds in wine besides alcohol. To test this hypothesis, investigators studied 802 past-month abstainers and drinkers of low-risk (“moderate”) amounts (1 to <3 [14 g] drinks per day) age 55–65 years at enrollment. They assessed mortality over 20 years.
- The mortality rate was 69% for abstainers, 50% for low-wine-consumption drinkers (<one-third of alcohol from wine) (adjusted* odds ratio [AOR], 0.67), and 32% for high-wine-consumption drinkers (AOR, 0.59).
- Although an unadjusted analysis showed a large reduction in mortality associated with high-wine-consumption drinking compared with low-wine-consumption drinking, the difference disappeared after controlling for overall alcohol consumption, socioeconomic and health factors, and health behaviors.
If there is a mortality benefit from drinking, these results suggest beverage type doesn’t matter. But, as is very common in studies examining the potential benefits of drinking, this study tested associations between typical alcohol consumption during 1 month and mortality over 20 years without updating consumption. It also compared drinkers with those who had stopped drinking—a group well known to be less healthy (“sick quitters”). “Moderate” drinkers are well known to be healthier than others—not because of drinking, but because health-conscious people often drink low-risk amounts. The fact that the large benefit of wine consumption disappeared with appropriate statistical adjustment should serve as yet another reminder to use caution concluding that observed associations between drinking and health benefits are causal. It would not be the first time consistent results from numerous observational studies were consistently wrong.
Richard Saitz MD, MPH
Holahan CJ, Schutte KK, Brennan PL, et al. Wine consumption and 20-year mortality among late-life moderate drinkers. J Stud Alcohol Drugs.