Listen to this story
In early 2018, Esquire ran this headline: “The World’s Oldest Man Drank a Glass of Red Wine Every Single Day.” Why did the editors choose this headline? Because they know people are enthralled with the idea that drinking alcohol might actually be good for us. Any “evidence” that corroborates the confirmation bias in favor of alcohol’s benefits tends to appeal to a wide audience.
According to 2015 data from the National Institute on Alcohol Abuse and Alcoholism, more than 15 million adults in the United States abuse alcohol, but less than seven percent of them receive treatment. In other words, there are a lot of heavy drinkers out there who love a reason to fool themselves into believing that alcohol is good — or at least not bad — for their health. These statistics coupled with the media’s insatiable hunger for clicks are breeding an epidemic of inappropriate alcohol-related narratives that conflate correlation with causality.
The Independent took the same story about the drinking habits of the world’s oldest man a step further. They baited their headline hook with a blatantly causal claim: “Man lives to 107 ‘by only drinking red wine.’” The Telegraph published an even more ridiculous headline: “This 107-year-old’s secret to a long life? Four bottles of red wine a day.”
Some people live a long time regardless of whether they drink, smoke, or eat six and a half Twinkies every day.
Anecdotes are not useful data in the context of public health issues like drinking. Some people live a long time regardless of whether they drink, smoke, or eat six and a half Twinkies every day. We know this, yet news stories about the drinking habits of really old people still climb to the top of the headline reels. Drinking is pleasurable. If it weren’t, we wouldn’t do so much of it. And if it’s also good for our health, even better. But while we want that to be the case, there’s really no evidence that it is.
The driving force behind headlines that say booze is healthy is correlation. Lots of studies have found that people who drink a little bit — not teetotalers and not heavy drinkers — have a little less cardiovascular disease. But it’s possible that all that means is that people who do things in moderation have healthier hearts, which seems more reasonable. Meanwhile, there’s convincing evidence available out there, too, that shows alcohol causes cancer. An epidemiological study funded by the Bill and Melinda Gates Foundation and published in 2018 found that no amount of alcohol is safe. It was part of researchers’ worldwide effort to track global burdens of health problems. The conclusion is pretty straightforward:
We found that the risk of all-cause mortality, and of cancers specifically, rises with increasing levels of [alcohol] consumption, and the level of consumption that minimises health loss is zero.
Another problem is that people who think they’re drinking in the healthy range often aren’t. A study by the U.K. Department of Health found that drinkers can underestimate their alcohol consumption by about 40%. And yet we still see headlines like this one in 2018 from Newsweek: “Drinking Alcohol Tied to Long Life in New Study.”
What this points to is our enduring struggle to grasp the fact that two things that happen together don’t necessarily cause one another. That is, correlation does not mean causation.
The tendency of even supposedly smart people, like doctors, to misconstrue correlation as causation has created lots of problems in health policy. The tumultuous history of hormone replacement therapy (HRT) is a great example. Observational findings suggested that estrogen protected against heart disease in aging women, but when randomized trials looked at what happened when women took HRT in controlled settings, heart disease actually increased.
We like to think of ourselves as intelligent, rational beings. But we regularly fail to recognize basic logical fallacies.
Despite these lessons, we still seem to get it wrong about alcohol a lot of the time, at least in the media. Here’s a headline that Maxim published in early 2018: “Drinking Alcohol Is Better For You Than Exercise When It Comes To Living Longer, Says Awesome New Study.”
Human behavior is fascinating. We like to think of ourselves as intelligent, rational beings. But we regularly fail to recognize basic logical fallacies. Our hedonistic propensities propel us to accept correlation as causation when it appears to validate behaviors we like to engage in, like drinking. Conversely, when our wants create an inclination to reject new evidence, we disparage correlational data as insufficient to establish causation. Take climate change, for instance. Changing our collective behavior in order to forestall climate change will take effort and a disruption of the status quo, so many people staunchly reject the notion that human activity causes climate change despite mountains of correlational evidence.
Have misleading headlines that assert that alcohol is healthy contributed to overconsumption? Probably in some cases. But the mounting evidence that we really shouldn’t claim any health benefits for alcohol will hopefully curtail the flood of bad journalism praising the virtues of booze.
Our challenges with wrapping our heads around correlation versus causation probably aren’t going away anytime soon. After all, philosophers have been struggling with this concept for a long time. What we know is that if you’re a person with the self-control to drink only lightly, not to excess, you’re probably a little healthier overall. But that’s probably because you’re level-headed and take a moderate approach to life. You don’t need to credit booze for your good health.