Harriet Brown

Brave Girl Eating

Dementia and the "Obesity Paradox"

Why are we still surprised when obesity is linked to better health outcomes?

Posted Apr 23, 2015

Source: Used as part of the Creative Commons

Two weeks ago, a new study published in The Lancet made headlines in the U.S. and elsewhere by describing a correlation between obesity and lower levels of dementia. In other words, people who are considered overweight or obese on the BMI chart showed lower levels of dementia later in life than those considered "normal" or underweight.

Any news about our risk of dementia is of interest, given the fact that our population is aging and more of us will be developing dementias in the next 20 years. But what’s really fascinating here is the way news outlets have been spinning the study. The BBC’s coverage, for instance, started with a headline reading “Being overweight ‘reduces dementia risk.’” Note the extra quotes around “reduces dementia risk,” which suggest  not just skepticism but suspicion. The article itself leads with the line “The researchers admit they were surprised by the findings, which run contrary to current health advice.” The subtext here (not that it’s all that hidden) is that since overweight and obesity are bad news health-wise, this link between obesity and lower dementia rates is either a huge surprise or a mistake. Or both.

Every time a study like this makes the news, scientists and journalists seem to feel obliged to register shock, suspicion, and disbelief. Researchers call such findings evidence of the “obesity paradox,” meaning a) they run counter to what they expected, and b) they can’t explain them in the usual ways.

But such findings aren’t new and shouldn’t be shocking. A robust body of research has documented such “paradoxes” in all sorts of diseases and conditions, including heart disease, heart failure, type 2 diabetes, stroke, high blood pressure, kidney disease, burn recovery, surgical recovery, and, now, dementia. (I wrote about it for the New York Times several years ago.) The idea that fat is associated with better outcomes is only paradoxical if you assume that the relationship between weight and health is a straight-up linear one—in other words, the higher the BMI, the greater the risk for disease. If you actually look at the research, though (and I’d hope both scientists and journalists do that), you’d come across lots of findings like this. Over the last 15 years, for instance, epidemiologists like Katherine Flegal of the Centers for Disease Control and Prevention have repeatedly and meticulously documented the fact that the lowest mortality rates fall in the overweight to mildly obese categories on the BMI chart; the highest mortality risks affect those on either end—underweight and severely obese. Dozens of researchers have discovered that some people with chronic diseases who are considered overweight or obese live longer and fare better than those of normal weight with those same diseases.

It's disingenuous to act shocked about this study or any other one that uncovers a health benefit of having a BMI over 25. It should be crystal clear by now that the relationship between weight and health is far more complex than the simplistic simple fat = bad, thin = good equation, and that there’s an awful lot of it we don’t yet understand.

Having spent several years immersed in the science on weight and health, I can testify to the fact that there is no linear equation here, no simple cause and effect. The literature is full of conflict, complexity, and nuance. We’re far too smart to buy in to the simplistic and often inaccurate view of weight and health being presented to us in stories like this one.

So enough with the surprise already. And anyway, would it be so terrible if there were some health benefits to being fat?

Harriet Brown’s latest book is Body of Truth: How Science, History, and Culture Drive Our Obsession with Weight—and What We Can Do About It. She teaches magazine journalism at the S.I. Newhouse School of Public Communications in Syracuse, New York.

More Posts