Log In / Join Now

Debate Over Alcohol's Health Effects Grows

A recent global analysis on drinking has raised alarms, but should responsible wine consumers really be worried?
An ambitious study examined alcohol's worldwide impact, but some question its conclusions.
Photo by: istockphotos
An ambitious study examined alcohol's worldwide impact, but some question its conclusions.

Lexi Williams
Posted: September 4, 2018

Is drinking any amount of alcohol dangerous? That's the question on the minds of wine lovers and public-health experts alike after the recent release of an ambitious meta-analysis of worldwide health data. The study, published in the Lancet medical journal, argues that the only safe level of alcohol intake is zero. But according to other experts, there's no reason for wine drinkers to panic.

The analysis, part of the larger 2016 Global Burden of Disease Study from the University of Washington's Institute for Health Metrics and Evaluation, looked at data from nearly 700 studies to estimate how common alcohol consumption is worldwide. They used another 592 studies—which included data on 28 million people in 195 countries accumulated between 1990 to 2016—comparing alcohol consumption in those nations to health statistics to try to gauge the health risks associated with alcohol.

The findings—including the fact that there were approximately 2.8 million alcohol-associated deaths in 2016—are sobering. But should light and moderate alcohol consumption be lumped in with heavy drinking? The authors say yes, dismissing the positive health effects of a daily drink compared to the risks. "There is no safe level of alcohol consumption," they write.

But others say no.

A global study: The pros and cons

"It's a very ambitious study. That's a good and a bad thing," said Howard Sesso, an associate professor of medicine at Brigham and Women's Hospital who was not involved in the study. "Inherently, when you're doing these types of meta-analyses or combined analyses, you gain information by combining the studies together, but you lose some of the specificity."

By examining global data, the authors were able to look for correlations between alcohol consumption and everything from cardiovascular disease to cancers to auto accidents.

But having such a big-picture view can be misleading. For example, in the study, one of the leading causes of alcohol-related deaths among 15- to 49-year-olds was tuberculosis. On a global scale, this is significant; according to the Centers for Disease Control (CDC), 10.4 million people developed the disease worldwide in 2016. But in the United States, that number was 9,272.

Tuberculosis, whether it's brought on by alcohol or not, is simply not as big of a concern here as it is elsewhere in the world—and the small chance of developing it shouldn't necessarily be a factor in determining whether to cut back on drinking in the U.S.

The study was also unable to look into many confounding factors that play a role in the reasons why people drink or don't drink, and why they may develop health issues. "The reasons for people drinking in the United States might be very different from the reasons why people drink in Japan, or in any other country for that matter," Sesso said. "It could be a function of maybe they're smoking more, maybe they're eating better or worse. It's hard to know what role these other factors were in determining the association between alcohol and overall health [with] this global measure they constructed."

One of the most notable confounding factors is whether the people in the studies consumed wine, beer, spirits, or some combination of the three. Red wine in particular has been linked to a variety of health benefits, including a lower risk of prostate cancer and healthier arteries. Had the researchers been able to subdivide the drinkers by the types of alcohol they consumed, the results may have been more edifying.

How big is the risk, anyway?

The researchers built a complex statistical model to determine the risk of developing 23 different alcohol-related health problems, including cardiovascular diseases, seven types of cancer, non-communicable diseases such as cirrhosis of the liver and diabetes, communicable diseases such as lower respiratory infections and tuberculosis, violence and self-harm, transportation-related injuries and other unintentional injuries brought on by events like poisoning and drowning.

Starting with just one drink per day, the study found a rise in estimated risk of developing one of the aforementioned 23 health issues. With each additional drink, the risk appeared to grow. Compared to non-drinkers, people who had one alcoholic beverage per day had a 0.5 percent higher risk of developing one of the problems. For those who drank two drinks per day, that risk jumped to 7 percent higher than non-drinkers. At five drinks per day, the risk was 37 percent higher.

However, the way the analysis represented these risks may make them seem more dramatic than they actually are. "The paper includes relative risks, such as people consuming two drinks a day having a 7 percent increased risk of experiencing a serious health problem each year. But 7 percent of what?" David Spiegelhalter, a professor for the public understanding of risk at the University of Cambridge, said in an email. "Psychology research shows that to properly interpret the importance of risks, we need to know absolute risks. For example, what does this mean for 1,000 people?"

A separate press release on the findings break it down to the absolute risks: Over the course of one year, an estimated 914 out of 100,000 people who don't drink could be expected to develop one of the 23 health problems. Comparatively, if 100,000 people consumed one drink (in this study, 10 grams of alcohol, or a bit less than one standard glass of wine) per day, 918 of them could be expected to develop one of those health issues—four more than the one-drink-per-day group.

Is this very small difference in risk between one drink and zero actually meaningful? Many experts think not. They do agree, however, that the risk increases with each additional drink, and that exceeding moderate levels of consumption is unsafe.

"I would tell someone that drinking within [the U.K.'s] current guidelines [of 14 units per week or less] is very low risk, but drinking more than two drinks a day on average increases the health risks " Spiegelhalter said.

Sesso concurs: "The study's value is in the fact that it's reinforcing what we've already known, that excessive levels of alcohol intake do more harm than good."

So what about wine's benefits?

The findings in the Lancet paper don't invalidate all the work that's been done to understand the health benefits of wine. In fact, it even highlights some of them. Though the researchers determined that the risks, such as cancer, ultimately outweighed any of the health benefits, they did find protective effects, specifically against heart disease and diabetes, and more notably among women.

These findings back up previous studies that have found alcohol to be beneficial in these health areas. And for those who are more concerned with heart disease than they are with tuberculosis, for example, it's certainly a noteworthy observation. (Individuals should consult with their doctor about what's right for their particular health situation.)

Considering the extremely broad scope of the analysis and the very small estimated risk of low amounts of drinking that it found, many experts are considering its recommendations for abstention to be extreme. Future studies that focus on specific populations, especially randomized clinical trials, are needed in order to further understand the risks and benefits of alcohol as they relate to the many confounding factors that were not explored in this analysis.

There is also overall quality of life to consider. Since abstaining from alcohol appears to only slightly reduce the risk of developing a health problem, is cutting out a beloved beverage really worth it?

As Spiegelhalter wrote in an online analysis of the study, "Claiming there is no 'safe' level does not seem [to be] an argument for abstention. There is no safe level of driving, but [the] government [does] not recommend that people avoid driving. Come to think of it, there is no safe level of living, but nobody would recommend abstention."


Want to learn more about how wine can be part of a healthy lifestyle? Sign up for Wine Spectator's free Wine & Healthy Living e-mail newsletter and get the latest health news, feel-good recipes, wellness tips and more delivered straight to your inbox every other week!

Chas Paddock
West Boylston, MA. USA —  September 4, 2018 6:02pm ET
First, correlation is not cause and effect.

Second, in a “complex statistical model”, factor confounding and correlation must be analyzed and unconfounded to obtain the actual statistical significance vs. random occurrence, of consuming JUST the alcohol.

Third, the reason people drink has no statistical validity in a REAL study intended to conclude cause and effect of the factor being studied: Alcohol

Lastly, this was not a study. These were people playing with statistical software and showing it to medical people who took a Hippocratic oath. A real study would at least attempted to confirm the effects and the findings. It appears there wasn’t. This “study” is hogwash.

Would you like to comment? Want to join or start a discussion?

Become a WineSpectator.com member and you can!
To protect the quality of our conversations, only members may submit comments. Member benefits include access to more than 315,000 reviews in our Wine Ratings Search; a first look at ratings in our Insider, Advance and Tasting Highlights; Value Wines; the Personal Wine List/My Cellar tool, hundreds of wine-friendly recipes and more.

WineRatings+ app: Download now for 365,000+ ratings.