Monday, June 19, 2017

If you want fries with that, get 'em

Science journalist Sara Chodosh is giving you permission.
There are plenty of reasons not to eat potatoes, and only one reason to eat them: they’re freaking delicious. That’s the only reason you need. No one is eating fries because they think they’re healthy. But the next time you eat some delicious, oil-crisped taters and someone blurts out “hey, you know fries double your risk of mortality, right?” because they read a clicky headline, you can rest easy knowing that they are wrong. And superiority is the ultimate reward, right?
Here's the issue, and this applies to many "scientific: studies.
You could pretty much sum up the whole problem with the recent study on taters in one sentence: correlation doesn’t imply causation. Let’s all say it together. Correlation doesn’t imply causation. Now try to remember that next time you read literally any nutrition story. Just because a group of people who eats fried potatoes more than three times a week has a higher mortality rate than a group of people who hardly ever eat them doesn’t mean that potatoes are the source of those deaths. You can control all you want for age and socioeconomic status and every other obvious confounding factor. At the end of the day, people who eat fries three-plus times a week are almost certainly going to have other habits that make them more likely to die.
For starters, it's likely that people who choose to eat that many taters are exercising less than the people who care enough about healthy dining to avoid that starchy temptation. And they probably consume more sugar generally as well. Or maybe they take in less fiber. You can’t eliminate the confounding effect of these other habits—that’s the real problem with nutritional studies like this. It appeared in The American Journal of Clinical Nutrition recently and it's showing up all over the web. Nearly every article based on the study claims that fries doubled the mortality risk of the subjects, and media outlets aren't technically wrong about the results.
Supersize it.

No comments:

Post a Comment