Around a dozen years ago, researchers noticed that some patients with chronic conditions such as heart disease fared better than others. This should have been encouraging news, perhaps a clue to future treatments. Instead, researchers were baffled. Because the factor that seemed to be protecting these patients was fat: They were all overweight or mildly obese.
“When health-care professionals get their first nutrition books, there’s a chapter on obesity,” says Glenn Gaesser, director of the Healthy Lifestyles Research Center at Arizona State University. “And it generally says that fat people are unhealthy and thin people are healthy.”
Researchers immediately began trying to explain this “obesity paradox”—or, more often, to explain it away. Carl Lavie, a cardiologist in Jefferson, Louisiana, was one of the first clinicians to describe the paradox. It took him over a year to find a journal that would publish his findings. “People thought, ‘This can’t be true. There’s got to be something wrong with their data’,” he told Quartz.

“People thought, ‘This can’t be true. There’s got to be something wrong with their data.’”
Since then, dozens of studies have confirmed the existence of the paradox. Being overweight is now believed to help protect patients with an increasingly long list of medical problems, including pneumonia, burns, stroke, cancer, hypertension, and heart disease. Researchers who have tried to show that the paradox is based on faulty data or reasoning have largely come up short. And while scientists do not yet agree on what the paradox means for health, most accept the evidence behind it. “It’s been shown consistently enough in different disease states,” says Gregg Fonarow, a cardiology researcher at the University of California, Los Angeles.
The researcher who did most to kick off the debate, and in the process became the object of much of the pushback it generated, is an epidemiologist at the US Centers for Disease Control and Prevention named Katherine Flegal. Together with colleagues, she looked at hundreds of mortality studies that included data on body mass index (BMI), which is calculated by dividing a person’s weight in kilograms by the square of their height in meters. People with BMIs of more than 25 are classed as overweight, and those with a BMI over 30 as obese.

Researchers immediately began trying to explain this “obesity paradox”—or, more often, to explain it away.

Flegal found the lowest mortality rates among people in the overweight to mildly obese categories. It’s true that these groups are slightly more likely to suffer from heart disease and some other life-threatening conditions in the first place. But many factors influence the likelihood of a person getting heart disease. And a strong link between weight and disease only emerges among people with severe obesity. So taken at face value, the results seemed to be showing that a little extra weight is genuinely beneficial.
Flegal is a meticulous researcher: her most recent analysis incorporated data from almost 100 studies and close to three million people. It was published by the prestigious Journal of the American Medical Association. Yet Flegal’s work has made her a target for those who scoff at the paradox. Walter Willett, a researcher at the Harvard School of Public Health who has taken a high-profile stance against obesity, told NPR that one recent Flegal study was “really a pile of rubbish” and that “no one should waste their time reading it.” (He was later admonished by the editors at Nature. In recent comments to Quartz, he reiterated his view that the study was “rubbish.”)

Being overweight is now believed to help protect patients with an increasingly long list of medical problems.

Willett’s complaints are starting to look less credible, however, because no one has been able to make the paradox go away. One of the most popular explanations is that fat people get more aggressive treatment than thin people, because their weight raises red flags at the doctor’s office. This seems questionable: studies show that overweight and obese people tend to avoid doctors, get fewer preventive screenings, and receive worse treatment because they’re often misdiagnosed as “fat” rather than with a specific medical condition.

What’s more, at least one team has examined and dismissed the better-treatment explanation. The researchers, led by a French endocrinologist named Boris Hansel, analyzed data on 54,000 patients who were at risk of stroke and heart attack. The optimal treatment for these patients is well known: protective drugs like statins and beta-blockers. But mild obesity seemed to protect at-risk patients whether or not they were taking the drugs.
Another potential explanation is that the data on people of normal weight are skewed. Researchers know that people tend to lose weight toward the end of life, but don’t always realize that they are sick. Smoking also makes people thinner and sicker. So, goes the theory, maybe researchers have inadvertently lumped mortally ill people and smokers in with healthy folk of normal weight, thus making the normal weight group look less healthy than it really is.

No matter how many ways Carnethon sliced and diced the data, the obesity paradox persisted.

There’s some evidence to back up this argument, but the studies on the issue are far from clear. The argument certainly does not seem to make the obesity paradox go away, at least according to Mercedes Carnethon, an epidemiologist at Northwestern University who has analyzed data on diabetes patients. Carnethon began by excluding patients who died within two years of diagnosis, to account for people who were already sick but didn’t know it; she still found higher mortality rates among thin people. Then she ran the data separately for smokers and non-smokers; still no difference. No matter how many ways she sliced and diced the data, the obesity paradox persisted. (Flegal also ran her data with and without smokers, and found no difference.)

If the paradox is real, and extra weight can bring benefits, what constitutes a healthy life? Is there any point trying to diet to lose weight, for example?
Researchers are divided on the public health implications of the paradox, but the approach that makes most sense to me is Health at Every Size. This is based on the idea that healthy behaviors, including nutrition and physical activity, matter more than weight.

Healthy behaviors, including nutrition and physical activity, matter more than weight.

Take exercise. Paul McAuley, a health education researcher at Winston-Salem State University in North Carolina, has been studying fitness for close to 20 years. He says most studies on weight and health fail to take it into account. “Or they ask one question about it,” he says, and don’t bother to go further. When McAuley collects data on fitness, he finds that it predicts health and longevity much more strongly than fatness.

Other researchers have found that Health at Every Size, when compared with a weight-loss approach, leads to lower cholesterol, blood pressure, and other metabolic markers. “We’re so stuck on the fact that the only way to mediate health is through weight,” says Linda Bacon, a nutrition professor at University of California, San Francisco and author of a book on the approach.
If Health at Every Size is taken up more widely and continues to deliver results, we may look back and conclude that the most disturbing element of this controversy is that it was a controversy at all. We don’t know as much as we would like about the complex relationship between weight and health. We don’t know for sure what the obesity paradox means and how to interpret it. Why does it inspire so much pushback?

“We’re so stuck on the fact that the only way to mediate health is through weight.”

“People are furiously looking for some way to make this not the case,” says Deb Burgard, a clinical psychologist in Los Altos, California who treats eating disorders. “And I think that bears some comment. Theoretically we should be very happy to find out that people aren’t dying the way we thought they were going to, that there’s not going to be this terrible outcome. That people at higher weights are going to be OK.”
Even scientists whose own research has identified the paradox often seem ambivalent about the possibility that it might hold true. Carnethon has published several studies documenting the link between overweight or obese and better survival rates among people with type two diabetes. Yet like nearly every researcher I’ve interviewed on the subject, she resists the idea that fat might not always be unhealthy. “We’d never want to back away from weight-loss recommendations,” she says.
Lavie, who recently wrote a book on the paradox, also seems to buy in to the idea that everyone should aim for a BMI in the normal range. “People who are lean develop heart disease despite having a perfect body composition,” he told Quartz.
But where did this definition of “perfect body composition” come from? People of all sizes develop heart disease, and fat people with heart disease tend to do better overall than thin people with heart disease. Maybe the real paradox here lies in our assumptions about what constitutes normal weight.