Biased Research Fueled Countless Harmful Diet Fads

(Bloomberg Opinion) -- For all its prominence in public health, there’s little real understanding of overeating, being overweight, and how these things connect to each other and to illness. Last week, a widely circulated piece in the Huffington Post stated the obvious — “Smoking is a behavior, being fat is not” — which apparently needed to be pointed out because these factors so often appear together as the primary reasons to blame people for the high cost of health care.

There’s a powerful cultural prejudice that overeating causes people to be overweight, and being overweight causes people to get sick and die. But this is an assumption, not a scientific observation. The assumption made it easy to believe the claims of America’s most celebrated overeating expert, Cornell’s Brian Wansink, whose headline-making work had propelled him into a prominent role in reshaping school lunches under the Obama administration.

Last week Wansink resigned his Cornell professorship amid a misconduct scandal. Critics such as statistics professor Andrew Gelman had been warning for years that Wansink’s conclusions were probably untrue. Wansink’s claims centered on the assumption that eating too much causes poor health, and that subtle cues such as the size of bowls and the color of plates nudge people to eat too much.

As journalist Gary Taubes has pointed out in his books, and in interviews for my column, thin people don’t necessarily count every calorie, or go to bed hungry, or show self-discipline. They eat their fill and somehow manage to step into the same size pants every day. He has proposed that putting on fat might be a symptom of a metabolic disorder rather than the result of bad behavior. His contention has been that a diet too heavy in sugar and starch is throwing people’s internal physiology out of whack.

In other cases, size might be just part of natural human variation. A few years ago, CDC epidemiologist Katherine Flegal, now at Stanford, caused a scandal with a string of papers demonstrating that overweight people actually lived slightly longer than normal weight or underweight people. When I checked in with her this week, she said that others sliced up the data differently, eliminating smokers and people who might already be sick. That yielded the status quo answer that it’s better to be “normal” weight.

There’s some merit to the argument that we shouldn’t make the same mistake we made with smoking and leaded gasoline and wait years beyond the point of sufficient data before warning people of the risks. If looked at as a risk-benefit ratio, then it’s reasonable to keep sodas and sugary foods out of school lunches, and to warn people of their likely harm. But when doctors pester their patients to just eat less, they may be advocating something that doesn’t work, and causing unhealthy stress in the process.

It’s possible that even with the best of intentions, medical providers have been practicing based on misleading research. As the Wansink scandal has unfolded, it looks like he was interested only in getting a positive result, rather than accepting whatever answer the data showed — just one bias sadly common among researchers. If he’d been careful and honest, perhaps his answers would have bolstered the view that gaining or losing weight is not a simple matter of controlling calories.

To contact the editor responsible for this story: Philip Gray at philipgray@bloomberg.net

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Faye Flam is a Bloomberg Opinion columnist. She has written for the Economist, the New York Times, the Washington Post, Psychology Today, Science and other publications. She has a degree in geophysics from the California Institute of Technology.

©2018 Bloomberg L.P.