These headlines are all referencing a study that was recently published in the August British Medical Journal. Researchers looked back at the medical records of two million people and compared different levels of drinking (non-drinker, former drinker, occasional drinker, moderate drinker, heavy drinker) with the risk that someone would see their doctor for one of 12 heart problems.
They found that the lowest risk group was moderate drinkers—people who had a few drinks each week. They were significantly less likely to see their doctor for a heart issue than every other group, and were the least likely to get treatment for things like angina, coronary heart disease and heart attacks.
Now, these researchers were really smart. They were doing a study on heart disease, and they know that a lot of things can cause heart disease—smoking, diabetes and the like. So they got a range of demographic information on the participants of the study and did some cool statistical tricks to see if there was any difference once you accounted for these factors.
The results stayed the same.
It seems like the headlines might not be so wrong after all.
Except for one thing.
This isn’t the first time that moderate drinking has been revealed to be protective for heart conditions. For the last 20 years studies have told us exactly what we want to hear: people who drink a moderate amount of alcohol are the healthiest.
So why have the public health recommendations remained exactly the same; drink less, ideally not at all?
The answer is simple; rich people screw everything up.
It turns out that people who drink in moderation aren’t like the rest of us. Specifically, they tend to be wealthy, educated, and in many ways better off. This is a big deal because people who have a higher socio-economic status have generally better health outcomes, particularly for heart disease.
In fact, the recent study showed exactly this. They found that moderate drinkers were the most privileged of any group*. And there’s very strong evidence that once you control for socio-economic status, the protective effects of moderate drinking disappear completely.
This makes perfect sense. Socio-economic status is one of the biggest drivers of poor health; there have been whole books written about how your postcode predicts how healthy you will be.
There’s also a well-known problem with comparing moderate drinkers to non-drinkers, which is that non-drinkers tend to be weird in their own ways. Many people who don’t drink at all are either sick (and so have cut out booze) or are recovering alcoholics (and so have many of the same risks as current drinkers). The authors tried to take this into account, but it remains an issue with their results.
I could go on, because this study was observational. This means that even if it demonstrated that moderate drinkers were better off (which it doesn’t), you still couldn’t say that moderate drinking leads to better health. This type of study just doesn’t work that way.
The researchers did everything they could to account for known variables, but there are just too many things that could’ve made a difference to people’s hearts to say that moderate drinking causes better health.
Same Old Story
If it sounds like this is another boring story about journalistic articles blowing fairly uneventful scientific findings way out of proportion, well, that’s because it absolutely is.
As I’ve said again, and again, and again; this could all have been avoided if the journalists writing about it had read the actual study. To quote the study authors:
It could be argued that it would be unwise to encourage individuals to take up drinking as a means of lowering their risk (although it must be noted that the findings from this study do not directly support this as we did not consider transitions from non-drinking to drinking)
Not only was this study open-access (free), easily Google-able, and reasonably easy to read, it was just eight pages long. If that wasn’t enough, there is a plain-language summary that precisely explains why you can’t say that moderate alcohol intake is good for you.
Journalists do hard work. They have crazy deadlines, and have to pump out interesting stories that we will all read. But this is just another case where 10 minutes of internet searching disproves almost every media article out there.
*In fact this is a pretty big issue that I have with this study. While the researchers did collect socio-economic indicators (you can see the data in table 1), there was no effort to control for this in their study. Since it’s never addressed or reported on, I can only assume that either the data was not good enough quality to do this analysis (very likely) or that they intentionally left it out (less likely). Either way, it is a significant weakness of the study that is never mentioned and strikes me as very strange.
Gideon is a health nerd and epidemiologist (public health person) working in chronic disease. He writes about how simple health science really is, how we get it so wrong and why being terrified of that New Scary Study is usually a bad idea. If you want to get in contact, he is shamefully addicted to Twitter and would love to hear from you!