"A beer every day could keep the doctor away," chirped one headline that made the rounds this week. Another boldly proclaimed that a pint-a-day habit could flat-out prevent stroke and heart disease. Yet another promised to explain "Why drinking beer could be good for your heart," and then backpedaled. "Positive effects were negated by heavy drinking," added the subhead. That's better, but let's back up even more. These articles reported on a preliminary study, presented as a poster at the American Heart Association conference in New Orleans this week. The study found an association between moderate drinking and slower declines in high-density lipoprotein (HDL), or "good cholesterol," in a population of around 80,000 Chinese adults followed over a six-year period. Researchers found that HDL levels declined over time in all participants, as they typically do with age, but that those declines were slower in moderate drinkers than in those who abstained or drank heavily. Good cholesterol helps clear out the arteries and takes other forms of cholesterol—including low-density lipoprotein (LDL), or "bad cholesterol," which clogs the arteries and increases the risk of heart attack and stroke—with it. Therefore, retaining more HDL might be expected to promote heart health. But the links between alcohol and heart health are not as strong as these headlines may lead you to believe. For one thing, the study in question has not yet been published in a scientific journal. That means that it has not yet gone through a vetting process called peer review. During peer review, experts in a given field review a study manuscript, point out potential problems, and give the journal an opinion about whether or not to publish it. They can also recommend that the researchers do additional experiments before they publish their results. In theory, then, this new research about alcohol consumption and good cholesterol could have problems that, had the study been published in a journal, the peer review process might have caught.
Another reason not to take these results as conclusive and definitive is that the study was "preliminary"; that is, when the researchers submitted their work to the conference, they had not entirely finished analyzing their results, says Shue Huang, the Penn State Ph.D. student who presented the poster. The headlines also suggest that drinking improves heart health in a cause-and-effect relationship. But that's not actually what the study reports. Instead, it describes the correlations between self-reported alcohol consumption and rates of HDL decline. But these experiments do not prove that the moderate drinking led to slower HDL declines any more than they show that slowly declining HDL levels lead to moderate drinking.
Further, there's always the possibility that another factor tightly associated with drinking could explain the results. What if, for example, moderate drinkers had significantly different diets (another factor that affects LDL and HDL levels) than people who abstained from alcohol or drank heavily? In that purely hypothetical case, it could be possible that diet, not drinking, explained the results. To avoid such confounding variables (multiple differences between study groups that could affect the results), the groups should be as similar to each other as possible in every aspect but one—in this case, drinking. Neither the study press release nor news articles mentioned whether or not that was the case. However, study author Shue Huang says that the study had taken possible confounding variables, including age, sex, physical activity, body mass index, and more, and the association still held. News reports claiming that drinking prevents heart disease and stroke make an unsupported leap between slower HDL declines and better heart health. Yes, higher levels of good cholesterol are associated with lower incidence of heart disease, so one might predict that the retention of good cholesterol benefits heart health. But the study did not examine heart health; instead, it measured HDL levels. To truly draw a connection between drinking and heart health through a study like this, it would be important to track the incidence of heart problems among the participants. Huang did note that moderate alcohol consumption has already been linked with a lower risk of cardiovascular disease. More than 100 studies support this association. So this new study is less about confirming the alcohol-heart health link and more about providing a possible mechanism for that association, though more work would be needed to confirm this idea. "Our study might provide some possible explanation for such observed benefit," Huang adds. So what does the study actually say? For one thing, participants' HDL levels decreased across the board. Those declines, however, were slower in people who drank moderately (one to two drinks for men and half to one drink for women) than in those who abstained altogether or drank heavily, according to the release. The association between moderate drinking and slower HDL declines was strongest for beer, but also held up for moderate consumption of hard liquor. Unfortunately that's a little wordy for a headline. Photo: Cary Conover/Getty Images