Beyond obesity

Interesting JAMA editorial last week (unfortunately mostly behind wall) that I thought many here would like to be aware of: “Beyond the Obesity Paradox in Diabetes. Fitness, Fatness, and Mortality”

It was inspired by this article in the same issue that noted that the roughly 12% of those who are normal BMI at time of diagnosis with diabetes had a worse mortality rate than those who were overweight or obese at time of diagnosis. To me that seems almost like a no duh: if you have diabetes with a healthy level of fatness then you both have a different make up than someone who only develops diabetes at a higher fat level (be it by a genetic cause or an epigenetic cause) and if nothing else weight loss alone is less likely to be helpful. But the editorial uses it as cause to riff on something else, which even if not the main issue for this particular finding is one who many here will be glad to see the mainstream medical rag embracing: the issue of the “metabolically obese normal-weight (MONW)” Overlapping mightily with what gets more popularly called the “skinny fat.”

Thoughts or reactions?

It sounds like another complication of an increasingly sedentary world; people who have a normal body weight but low muscle mass and consequently too much body fat for their weight due to insufficient exercise. In particular, this sentence stands out:

That may also explain the findings of some past studies, such as those that found that being slightly overweight (BMI around 25-27 or so) had a lower risk of mortality than those who were normal weight; somebody who is fat but gets regular exercise can be healthier than somebody who is thin but never exercises.

Well, we know obesity is associated with metabolic abnormalities. I can’t read the article but it appears to imply a problem with BMI in individuals who have high fat content and low lean mass. That seems like common sense…BMI is a blunt instrument and maybe better methods are necessary to identify those at risk.

I’m not clear exactly what you are getting at? Do you have some issue with this editorial?

Actually I am pleased with the editorial. I shared it because of two main recurring and intertwined discussions that occur on these boards: how blunt of an instrument is BMI; and weight being set up as the goal rather than fitness.

  1. Issues regarding BMI as a measure. For reasons of convenience overweight and obesity are defined by BMI levels and for tracking population trends it is a fairly good tool. This editorial represents the increasing understanding and greater acceptance of the need to utilize BMI at an individual level in a manner more sophisticated than that. The BMI is screening tool that will miss some who are normal by BMI but are still “metabolically obese” and will, if used as the definition of the problem (rather than identifying someone at being at higher risk of having unhealthy habits and lower fitness), have an unappreciated false positive rate.

  2. Many threads here have posters hyperfocused on weight loss as their goal or others telling them that it should be their goal and that achieving such is a simple thing to do. In a recent thread one poster advised an op that his mother with abdominal obesity needn’t bother much with exercise because exercise won’t help her lose much weight. I think the attitudes of many docs only exacerbate the focus on the weight, the scale, as the goal. Like BMI the weight is easy to measure and easy to follow. My hope is that this editorial helps them recognize the quote highlighted by Michael63129 above and the data that shows that not only can someone who is fat but fit be healthier than someone who is skinny but unfit; they are:

Thanks for the article, DSeid. I don’t have much to add on this topic, but you always seem to have up to date information and you always explain it with much clarity in a field that’s often full of noise misinformation. As someone who mostly lurks on this board, I really appreciate your posts and analysis on topics related to obesity and health. I wish more doctors were like you.

The thing that most surprises me about scientific research is how many of the findings are of the “sun rises in the East” variety. Nothing in that editorial is at all novel. I recalling seeing this book at a friend’s a couple of years ago and surely something like Top 10 Reasons Why The BMI Is Bogus from NPR should have heralded the death knell of using BMI as a measure. But I routinely see pieces lauding new research that is not new or novel at all.

From Don’t Ask’s second link:

I believe this one. I’ve always had a problem with BMI. I’m a dense sort of person and when ever I have to be weighed, they always do that double-take on the scale. When I had insurance, you can bet that I got charged more for their BMI calculation, despite the fact that I could unload and stack nearly 8 tons of hay at one go and daily carried four 100 pound bales on my back the 300 feet from the barn to the field.

I’ve been exercising a good bit lately. Most of it is aerobic exercise, and I try to mix in some weight-bearing stuff like walking to make things interesting. I’ve been able to do around 25 miles on hilly road on my bike, and my walking speed has definitely increased, from around 2.5 miles per hour to 3.5+ miles per hour. My endurance has never been this good ever, and that includes after Navy boot camp. (I’m dealing with some residual respiratory issues due to not smoking, which isn’t helping that much, but the exercise is clearing the crap out of my lungs as is the fact that I live in a humid climate. But I digress.)

That being said, my overall weight loss has stalled out, which is disappointing. Exercise increases my hunger, and even though I’m trying to make good choices (fruit and sugar-free Jello, for example), my weight loss has apparently stalled out a bit due to a combination of emotional eating and probably the exercise. And, thanks to the exercise, I look much better than I ever have, except maybe after boot camp. But Weight Watchers is myopically focused on BMI, as well, and I can’t achieve my goal weight until I lose those last pounds. It’s pretty damn frustrating.

don’t ask,

Thing is that the BMI is not bogus. And there is a big difference between sensationalized articles claiming such a thing, distorting how the modern use of BMI developed and playing to silly conspiracy theories, and the AMA mouthpiece journal promoting a more nuanced usage of the tool. That NPR list is very stupid (if you want I could go down point by point). There is also a big difference between this editorial’s points and a book that tells people to “eat what you want, when you want, enjoy pleasurable foods …” Focusing the fight on fitness and the nutritional plan rather than on the scale is not the same as advising people to “give up the fight.”

Yes information about MONW has been being published for years and BMI was always known in the literature to be a convenient to use rough correlate of adiposity, particularly good at identifying the obese. Yes, information about the importance what sort of adiposity has been accumulating for years. Nothing “novel”, but still something significant.

As a screening tool BMI is pretty dang good. It was the step of using BMI as the definitions of overweight and obesity at an individual level that was regrettable and more importantly the focus on fatness over fitness.


The fact that you look so much better and have so much better endurance are proof that you’ve lost fat and become more fit. The point of this editorial is an endorsement of the POV that the scale is a poorer correlate for your health outcomes than those two, especially the latter one. It is the case that fitness is harder to measure in a Weight Watcher meeting or at a doctor’s visit than weight or BMI and thus those numbers will continue to be used. But you have to accept that your goal is not the weight number. It is the better health in both the short and long terms and that is being achieved by what you do, not what you weigh.

Please do. There’s way too much ignorance on the topic of BMI, and it would be nice to have something authoritative.

Anyway, the discussion here reminds me of something else I’ve seen recently, which is the notion that lack of exercise should be considered a medical condition in and of itself. I can buy their arguments pretty easily.

You didn’t ask, but here’s my opinion. Also, I didn’t comment on all ten.

  1. BMI “ignores waist size, which is a clear indicator of obesity level.” I was under the impression that BMI is a quick and dirty way to guess someone’s waist line when that information isn’t available. (See So, obviously, BMI is going to have all the flaws of the waist line measurement, with a few more on top of it.

  2. I can’t believe the vast majority of people in about 1850 lived sedentary lives. There is simply no way that is true.

  3. The BMI is not a number between 1 and 100. I have no idea why the article would assume it is. A BMI of 100 for a 6’0” man would mean a weight of 738 pounds. Clearly that is very high, but there’s no reason why weight couldn’t keep climbing past that.

  4. The idea of sharp boundaries is of course nonsense except insofar as you have to draw lines somewhere. Stuff like this happens in almost everything humans do and likely happens as well in whatever might replace the BMI.

The BMI is far from perfect but sometimes the critics really do let the perfect be the enemy of the good. It is a screening test with limitations which works well to identify those at high risk of high adiposity.

Some resources:

About Quetelet and the development of the Index He set out to develop a tool that spread weight to height in a normal distribution and did so. In recent decades it was found that it was still valid and correlated well with obesity related outcomes.

The AHA statement on Assessment of Obesity.

That statement goes into detail on other possible methods as well and the only one that they think might also meet the requirements of reliability, reproducibility, and ease of use across all populations, may be waist circumference, but at this point it is best viewed as a complement to BMI, not a replacement for.

This guideline is more what I need as a pediatrician. And the information gained by following the BMI curve, rather than focusing on a single measurement, is invaluable, even in preschool years.

So let’s go through the points:

The tool was validated in modern populations. Who cares if a tool is old if it works? The microscope was invented a long time ago; does it not work now because of that?

It is an observed correlation. Period. Someone with a BMI over 30 is highly likely to have high adiposity and be at risk of obesity related health problems. Few with BMI over 30 are “athletes and fit, health-conscious movie stars who work out a lot”. Maybe a few bodybuilders, but a very few. Saying that is scientific nonsense evinces a serious lack of understanding of science.

Waist circumference is a measure of central obesity which may be the most important thing to know but as of now we cannot say it is the only important thing to know.

Nope. Very logical. If someone has a high BMI and I should be suspicious that they have high adiposity and are at high risk of having behaviors that should be changed. It won’t catch all of those with behaviors that should be changed and it will be wrong for many too. But it catches a good many.

Huh? So because populations are similar then and now it is not valid as a screen? Again, it should not be considered the whole story.

Agreed that the cut-offs are arbitrary and that 25 would have always over-identified too many as overweight. It was a silly choice to make it easy for people to use…

And as said above, does this guy even know what the BMI is? 1 to 100? Maybe he’s thinking of percentiles.

An insurance company wants to screen out populations at higher risk and those with BMI over 30 are, as a population, at higher risk. If they are allowed to cherry pick the lowest risk group then of course they will.

He wants us to MRI or hydrostatically weigh every person? The only contender for screening purposes (quick, cheap to use, reproducible, decent enough predictive value for obesity complications …) is waist circumference. Yes, that should be used more often, but for now at least as an adjunct to BMI. It is too soon to say it is good enough alone.

What embarrasses the U.S. is the level of morbid obesity. Those with BMI over 35 are fat, not muscled, and those numbers have sky rocketed.

BMI is a limited tool and its limits need to be better appreciated by both the public and by clinicians. But “bogus”? Hogwash.

I think this supports my personal distaste for BMI being used as a health measure. ‘Fatness’ is something far more complicated than weight relative to height. I think newer, more reliable methods of determining body fat percentage should be emphasized.

Another study. (And the article it references.)

Another one.

And the take on it. I think this may go farther than the data warrants but is worth sharing.