A good example is a study published today on the correlation between alcohol consumption and diabetes. The paper, published in the journal Diabetologia, concluded that men and women who drink alcohol three to four times a week have a significantly lower chance of acquiring diabetes compared to people who drink less than one day a week on average.
Men who drank a few days a week had a 27 percent lower diabetes risk than infrequent drinkers, according to the report, and women had a 32 percent lower chance of getting the disease, which is the sixth leading cause of death worldwide, claiming 1.59 million lives each year.
Janne Tolstrup with the National Institute of Public Health at the University of Southern Denmark was lead author of the alcohol and diabetes study. In an email, she drew a clear line between drawing reasonable scientific conclusions from a single study and making broad generalizations about the health benefits or risks of drinking.
“In this study, we have a narrow focus on diabetes only,” wrote Tolstrup, “but since alcohol is related to more than 50 different diseases and conditions — reflecting that alcohol affects virtually every organ system of the body — any recommendations about how to drink and how much to drink should not be inferred from this study or any study investigating associations between alcohol and a single outcome.”
RELATED: Eating Chocolate Might Help Decrease the Risk of Irregular Heartbeat
Study methodology is also an important consideration, something that’s often glossed over by snappy headlines. For Tolstrup’s paper, she and her team analyzed data from the Danish Health Examination Survey, which provided detailed behavioral information from more than 70,000 respondents, a large and diverse sample. Also, when comparing drinkers to non-drinkers, Tolstrup was careful to distinguish between lifelong abstainers and “current” abstainers. The latter group may have quit drinking because of an underlying health problem, which would inflate the benefits of drinking.
Not all alcohol studies are so careful. In a paper published earlier this year in the Journal of Studies on Alcohol and Drugs, researchers from Penn State University pointed out a flaw in alcohol studies based on large longitudinal study in the United Kingdom called the National Childhood Development Study. One in five of the 55-year-olds in the study who said they were “abstainers” had previously reported drinking in their youth, highlighting the difficulty of relying on people’s memories for important data. The Penn State researchers found much stronger negative health effects when drinking is combined with smoking.
"Medical professionals and public health officials should be wary of drawing conclusions about the so-called 'dangers' of never drinking without more robust evidence," said Penn State researcher Jeremy Staff in a statement.
Although longitudinal health surveys and large databases are excellent ways to determine the relative health risks associated with different behaviors, it doesn’t mean that scientists understand the mechanisms behind those risks. In the diabetes study, for example, Tolstrup and her co-authors could only theorize why wine lowered diabetes risk more than beer or liquor — polyphenols in wine seem to aid the body’s management of blood sugar. She wrote that follow-up studies are required to really understand the processes at work.
Even the well-established link between red wine and a lower risk of heart disease isn’t fully understood. Researchers know that all alcohol, not just wine, can cause slight increases in levels of HDL or “good” cholesterol, but those same benefits could be obtained from physical activity or niacin pills. For that reason, the American Heart Association doesn’t recommend drinking wine or any alcohol as a way to stave off heart attacks.
RELATED: Diabetes? There's an App for That
Meanwhile, the public is barraged with seemingly contradictory information about the benefits and risks of drinking. One recent study linked moderate drinking with brain deterioration and cognitive impairment, while another one claimed that drinking boosts long-term memory. And the new study about drinking lowering the risk of diabetes needs to be weighed against other strong evidence that even moderate alcohol consumption by women increases their risk of breast cancer by as much as 9 percent.
“It’s natural that people can get a little confused about the health recommendations of alcohol,” wrote Tolstrup. “Whereas every alcohol researcher agrees that heavy drinking is very harmful, there is an ongoing debate about effects of low to moderate drinking – is it harmful, beneficial or neutral?”
The jury is still out.