Wine and Health Redux

This article by Tim Requarth in Slate is one of the better articles on wine and health that I have come across. It tells the story of how we came to believe that wine had health benefits and why the medical community no longer thinks it is so. And in the end he gives a thoughtful account of how we should think about this question.

Remember the French Paradox in the 1990’s? That was the genesis of the belief that moderate consumption of wine was healthy. The French used to consume a lot of wine and had a diet laden with butter and cheese yet had better health outcomes than Americans who consumed less alcohol.

A variety of studies were done comparing wine drinkers with non- drinkers and, surprisingly, the correlation between moderate wine consumption and cardiovascular health held up. Requarth summarizes:

“On balance they suggest that about one drink per day correlates with 14-25 percent less cardiovascular disease or death compared with abstaining.”

This generated lots of favorable press, some of it pushed by the wine industry, and it become widely accepted even in the medical community that wine had health benefits.

The problem was that in those studies that compared wine drinkers with non-drinkers the sample of non-drinkers included people who had given up drinking because of health problems. Thus, unhealthy non-drinkers were over-sampled. Subsequent studies that controlled for that variable have shown no health benefit for moderate wine consumption.

The good news is that moderate wine consumption doesn’t make cardiovascular health worse. However, the story about cancer is not so good. Even moderate wine drinking increases cancer risk by over 100% according to this summary by the National Cancer Institute.

But what I really like about the Slate article is that it goes on to interpret that statistic.

One of my pet peeves about health journalism is that risk assessments seldom take into consideration relative risk. It sounds scary when the headlines report that a lifetime of ingesting substance-X increases the risk of a specific cancer by 100%. But if the chances of any individual getting that form of cancer are say 1%, by ingesting substance-X you have only increased your risk to 2 %. If you really like substance-X the risk may be worth it.

The author of the Slate piece does a good job of considering relative risk for moderate wine consumption.

As bad as alcohol consumption is from a public health perspective Requarth concludes that “from the perspective of the individual drinker, it’s less dire.” Quoting a health researcher from Washington University he writes:

A middle-aged man’s baseline risk of dying from any cause over the next five years of his life is 2.9%. If he upped his drinking from a few drinks a week to a few drinks per day, this risk would rise to 3.6% or an absolute risk increase of .7%….Yes these numbers might add up to a lot of deaths and disability when we’re talking about the global population, but they aren’t reason for any individual person to panic.

Too often the headlines come at this issue from the perspective of public health, and that is an important perspective. But an individual’s risk is a much different matter. The two perspectives should not be conflated.