Asymmetrical Information - Megan McArdle

01.10.13

Rethinking Thin

Last week, a study came out showing that overweight people had lower mortality than "normal" ones. There's been a lot of pushback--but the arguments are pretty weak.

Last week, a study came out indicating having a Body Mass Index in the overweight range was actually associated with better mortality outcomes than being thin.  This has actually been known about for a long time by public health experts--Paul Campos and I discussed it years ago in a blog interview--but it came as a great surprise to most readers, and apparently, not a few health journalists.  

The pushback began immediately.  Lindsay Abrams of The Atlantic sums up the counterarguments:  

Aside from the obvious limitations -- people who pass away after a lengthy period of disease, for example, will likely be thinner than they might have been had they died unexpectedly -- the study fails to take into account any of the various other measures used to assess health. It ignores blood pressure, blood sugar, and cholesterol -- high levels of all are directly associated with a variety of chronic conditions and diseases -- not to mention mental health and life satisfaction scores. As another large-scale study recently pointed out, longevity isn't everything. The population as a whole is living longer than it was twenty years ago, but the number of those years spent in poor health are increasing as well. 

That BMI is an imperfect measure of body size is emphasized here as well. The simple calculation of height and weight ignores gender, age, and muscle mass -- I remember being hopelessly confused the first time a guy told me his goal was to gain weight. Where on the body fat is located is important as well (belly fat, for example, poses a greater health risk than excess weight that's more evenly distributed). A BMI in the "overweight" range, from 25 to just below 30, encompasses a broad sweep of body diversity: A frequently cited argument is that Michael Jordan, at his prime, would have been classified as overweight. By almost any other measure but BMI, we would almost certainly put him in the range of ideal health.  

Okay, yes, but . . . no.  It's possible that most of the thin people who die are meth addicts or have cancer, but even a study which threw out the folks who died within three years of entry into the study found that once you accounted for physical activity*, "underweight" BMIs were correlated with excess mortality risk, while "overweight" BMIs were not.  And arguing that the study fails to control for things like blood pressure, blood sugar, and cholesterol seems like fairly weak sauce; those are the very mechanisms by which obesity is supposed to kill us.  

I do agree that mortality isn't everything; morbidity (poor health) matters too.  But it is not true that we are living longer, more miserable lives.  On the contrary, active life expectancy is increasing faster than life expectancy, according to the experts I interviewed for an article in the Atlantic.  You can see this in the proportion of people on Medicare who have limitations in their ability to perform the tasks of daily living, which has been falling, not rising.  (Note: this chart is from 2008, before the bulge of baby boomers started skewing Medicare demographics younger).  

To be sure, it's possible that these statistics would be even more awesome if folks were thinner--being heavy is indisputably hell on the joints.  But it is simply not the case that we are collectively living longer, unhealthier lives than we used to.  And being thin can also be associated with morbidity problems--you're more likely to hurt yourself when you fall, for example, because there's no padding.

As for the last point, it seems a bit rich for public health experts and health journalists to suddenly be arguing that, well, of course, BMI doesn't really tell you how fat people are.  For twenty years, we've been reading about America's obesity epidemic, which was going to kill us all in our beds, bankrupt the health care system, and Jesus Christ, have you seen how these people look in sweatpants?  All of those articles were based on . . . rising BMI's.  That's because we have no idea how much body fat has increased.  It's absolutely true that visceral fat predicts mortality better than many other measures, but visceral fat is really difficult and expensive to measure, so we mostly haven't.  

It's even more absurd to suddenly resort to the Michael Jordan argument that Fat Acceptance folks have been using for years.  For starters, there's a reason that people specifically use NBA players, and not, say, soccer players, in this example: part of the reason that NBA players have such outsized BMIs is that the usefulness of the calculation tends to fall apart at the tails of the height distribution, not because they're so gosh-darned muscular. Star Real Madrid player Cristiano Ronaldo has an entirely "normal" BMI of 21.9  

I think it's unlikely that we've had an undetected epidemic of seven-footers.  And while a few epic body-builders can also have "misleading" BMIs, it seems unlikely that there are enough of those to be skewing studies that include millions of people.  Check out these professional body builders; a few of the men (and most of the women) are in the normal range, despite having a considerably higher muscle-to-fat ratio than 99% of the population. Several more are within a point of the cutoff.   I find it impossible to believe that there are enough technically "obese" bodybuilders in the general population to be somehow messing up our mortality statistics.

No, mostly BMI is telling us what it's always told us: the US population has gained some weight.  And this new study is telling us what studies have been telling us for a while: that this doesn't necessarily mean we're all going to die young.


* But Megan, I hear you cry, why are you controlling for physical activity? Doesn't exercise make you thin?  You'd think so, but there's not much evidence that exercise affects BMI enough to take you from "overweight" to "underweight".  It may not effect it at all; appetite tends to increase to compensate for increased activity.  


This article has been corrected to rectify an arithmetic mistake.