How a persuasive scientist with poor evidence ruined our diets
Posted by Richard on May 6, 2014
ICYMI: Put away that box of breakfast cereal and have some ham and eggs. That’s the take-away from a fascinating essay by Nina Teicholz in last Saturday’s Wall Street Journal:
“Saturated fat does not cause heart disease”—or so concluded a big study published in March in the journal Annals of Internal Medicine. How could this be? The very cornerstone of dietary advice for generations has been that the saturated fats in butter, cheese and red meat should be avoided because they clog our arteries. For many diet-conscious Americans, it is simply second nature to opt for chicken over sirloin, canola oil over butter.
The new study’s conclusion shouldn’t surprise anyone familiar with modern nutritional science, however. The fact is, there has never been solid evidence for the idea that these fats cause disease. We only believe this to be the case because nutrition policy has been derailed over the past half-century by a mixture of personal ambition, bad science, politics and bias.
The persuasive scientist who derailed nutrition policy was Ancel Benjamin Keys, and it’s an interesting story. The unintended consequences of the adoption of Dr. Keys’ diet recommendations, chiefly the increased consumption of carbohydrates and vegetable oils, have not been good. For one thing, eating less fat and more carbs, ironically, makes us fatter:
One consequence is that in cutting back on fats, we are now eating a lot more carbohydrates—at least 25% more since the early 1970s. Consumption of saturated fat, meanwhile, has dropped by 11%, according to the best available government data. …
The problem is that carbohydrates break down into glucose, which causes the body to release insulin—a hormone that is fantastically efficient at storing fat. Meanwhile, fructose, the main sugar in fruit, causes the liver to generate triglycerides and other lipids in the blood that are altogether bad news. Excessive carbohydrates lead not only to obesity but also, over time, to Type 2 diabetes and, very likely, heart disease.
Read the whole thing. You may want to change your breakfast routine.
David Bryant said
Hi, Richard. I read this article with some interest. I particularly liked the fact that Dr. Keys conducted a study of people’s dietary habits on the island of Crete during Lent. Talk about unbiased!
On the general subject of statistics, you might be interested in How Not to Be Wrong, by Jordan Ellenberg. I bought this book just a couple of days after it was published; I’ve read it, and I’d recommend it to anyone who wants to polish up his mathematical intuition.
Ellenberg’s a math geek, but he’s also a pretty good writer, and he spins a good yarn. For instance, he tells the story of Abraham Wald, a statistician who was employed by the War Department during WWII. The brass gave him a table showing where they found bullet holes in returning aircraft, saying “We think we need more armor on the places the Jerrys are hitting most often.” Wald said, “No — you need more armor on the engines. Hardly any of the returning planes have bullet holes in the engines, because when a bullet hits an engine, the plane does not return.” The Army Air Corps took his advice, and the fatality rate for bomber crews declined.
I suspect that Ellenberg is a bit of a statist — probably a Democrat (he’s currently a professor at the University of Wisconsin in Madison). Anyway, he has a blog called “Quomodocumque” — if you’re interested in nerdy stuff, check out this post and the comment I left in response.
Richard said
I try to remember Hanlon’s Razor, but it certainly sounds like Keys was deliberately cherry-picking evidence to confirm what he wanted to “prove.” Thanks for the tip on Ellenberg and the Wald story — I really like that one.