I ran into this today in the New York Times, but this graphic they link is much more eloquent than the article:
BMI (a function of weight and height) has been criticized for not distinguishing between fat and muscle, and generally being an unhelpful number. Healthy athletes often have a high BMI, and unhealthy elderly people with wasting muscles often have a “good” BMI.
In this 2013 article Diana M Thomas came up with a function of height and waist circumference that better measures “body roundness”, which is a decent proxy for visceral fat, which is the element of obesity that is most likely to actually hurt you.
The body roundness indicator (BRI) has the virtue of being something that’s easy to measure with cheap, available tools (no giant tank of water to calculate body density) and apparently, it’s getting traction as a better measure of “healthy weight” (or, i guess, healthy shape) than BMI. Apparently the formula is complex, but values range from 1-15. Those with a bri over 6.9 have a 50% increase in the risk of death over those with bri from 4.5-5.5, with a 25% increase in between. Those with unusually low bri (under 3.41) also have an elevated risk of death.
“The percentage of fat can vary from 10 percent to 40 percent among people of the same B.M.I., age and sex,” Dr. Heymsfield said. “The B.R.I. is a way of capturing these variations.”
Anyway, apparently, BRI is catching on. Sounds like an improvement. I wish they’d pointed to a calculator.
I had tried that calculator the other day and the Metabolic age calculation is quite suspect. Changing my weight by 1kg drops my metabolic age by 11 years.
It also shows my BRI at 3.85 which makes no sense, it’s too low,
Some athletes who have a lot of fat will have a very low amount of viceral fat. It’s an example of how being active can help to reduce the harmful effects of fat. In the show “Nova: The Truth About Fat”, they discussed how sumo wrestlers avoid the problems of obesity. They showed these MRI scans of two people of with excessive amounts of fat:
The sumo wrestler is on the right and the sedentary person is on the left. In the sumo wrestler, the fat is almost all between his skin and muscles. There is very little fat in the abdominal and chest cavities. It’s like he’s wearing an overcoat of fat. But in the sedentary person, fat permeates their abdominal and chest cavities, surrounding their vital organs. They said it’s because exercise releases a hormone called adiponectin. That hormone instructs the body to direct fat to the safer places under the skin.
I’m not sure that competitive athletes really care, but probably the BRI will be inaccurate for them in the same way that the MRI is. However, for regular people, it will probably be more accurate at indicating fat related health problems. Some people tend to store fat in their abdomen rather than evenly distributed around their body. Two people may be the same weight with the same body fat percentage, but one person may have most of their fat in a big “beer belly” while the other person has their fat distributed evenly over their whole body. In terms of health risk, the person with a beer belly is likely at more risk since that’s an indication of a high level of viceral fat.
Yes, no measure is perfect. But it looks like BRI may be a significant improvement over BMI, without adding any significant additional cost to the measurements required.
(For what it’s worth, they both say about the same for me. My BRI is maybe a little worse.)
I’ve never been sure why BMI gets such a bad rap. From the wiki:
[Ancel] Keys [, author of the paper in which the term was coined (although not the inventor)] explicitly judged BMI as appropriate for population studies and inappropriate for individual evaluation. Nevertheless, due to its simplicity, it has come to be widely used for preliminary diagnoses. Additional metrics, such as waist circumference, can be more useful.
Are there lots of cases where healthy body builders with high BMIs have suffered because someone has used it inappropriately?
The benefit of BMI is that it’s quick and easy with measurements most people know in their head. A high BMI may warrant further investigation, but whether that follow up is BRI or a volumetric analysis or just a doctor looking a patient up and down and realizing that the limitation of the measurement is obvious, I don’t see the big deal.
Many of the people who question BMI would also not fare well on any follow up measurements, which suggests that BMI is doing it’s job.
Now me, I’m skinny fat. I don’t exercise enough and what fat I do have is all visceral. It’s not enough for doctors to be concerned, but the fact that BMI let’s people like me slide is perhaps a bigger problem from a population perspective than false positives. Since my BMI is healthy, people tend not to follow up to see if my organs are actually packed in fat.
Thanks for sharing the visceral fat visuals. That makes a lot of sense - it explains why you see guys with huge hard guts with no flab and skinny limbs. They carry a lot of fat inside. There’s a doctor on youtube I’ve taken an interest in and his whole thing is visceral fat. he’s has hours and hours of videos talking about scans and how to attack this issue. his prescriptions seem a little out there (and obviously he’s selling treatments), so I’m not sure his solutions are solid, but I think he’s identified a huge issue.
I think the number of well-muscled and lean healthy individuals with a high BMI is very overblown. Yes, some healthy and not fat individuals can fall in the overweight category. But a lot of people are fatter and less healthy than they think. Guys who think they are jacked often just have “fatceps” and all competitive bodybuilders and a large chunk of fitness models and social media influencers are on steroids or other PEDs.
There’s something called the “fat free mass index” and there are generally accepted limits to natural fat free mass. A lot of numbers get thrown around in Hollywood and sports PR and whatever, but if you ever hear of like a football player who is supposedly 6’3 and 250 and 3% bodyfat or whatever, either he’s a lot fatter (or lighter) than what he’s listed at or he’s on juice. When you see a natural bodybuilder at 5’10 and 180 and lean, that’s believable. 5’6 and 220 - roids.
This caused a lot of people grief when I was in the military. You could be placed on a “fatboy” program if your ht/wt ratio was off. I recall one Chief who weighed in at about 250, most of it muscle. He marched into the med officer’s office, picked up the guy’s steel desk and slammed in back onto the floor, demanding that the doc find an ounce of fat on him. I didn’t have that kind of muscle, but I was/am pretty broad across the shoulders and was fit back then, and I was always on the cusp of having to go through the shame of being a “fatbody”.
My father was very fit, but when he lost an inch due to age his BMI became high enough that it bothered him. I feel like people who are muscular shouldn’t feel the need to make excuses, and people who are light because they have no muscle (like my mom, when she was old) shouldn’t be liked into thinking they are a “healthy weight”.
I’m of lean build, but probably 20 lbs over my former maximally scrawny weight. All of which is belly fat. It gives my metabolic age same as my actual age = 66. Altering my weight in their calculator from actual down by 10 even or 20 lbs makes no difference in metabolic age. Meanwhile all those values are in the green normal range of BMI.
Adding 10 lbs to my actual 20-over scrawny, maybe 10-over healthy, bumps BMI to the upper edge of green amd makes no change in metabolic age.
Adding 20# however costs me 13 years; suddenly I’m 79.
Clearly their idea of metabolic age has some cutoff of degree of obesity at which they start applying a pretty heavy discount.
ETA: More experimenting shows a 13 year age increase (for a 66 yo male) just as I exceed the BMI=25 top of normal range. And no further increase no matter how overweight I tell it I am.
Clearly that’s not valid statistically. But it may be consistent with what little data is out there on exactly how overweight /obese costs years. IOW some legit careful study may end with a conclusion of a 13 year loss on average, and without further detail on how that breaks down by degree of obesity. So the calculator is stuck being based on the non-detailed conclusion of the study.
However, I call Bullshit on the whole concept. Maybe there is something there but the BRI formula is idiotic. Here is the formula I got from Google: 364.2-365.5\sqrt(1-[w/2\pi]^2/[0.5h]^2)
where w and h refer to waist size, specified in cm. First, those numbers 364.2 and 365.5 simply adjust the scale and the range; what are really there for? Mathematically, it makes no sense. Second, with a trivial bit of algebra the formula reduces to 364.2 - 365.5\sqrt(1-[w/h\pi]^2)
which makes it perfectly clear that it depends only on the ratio of waist to height and you can measure them in cm, or inches, or light years, or angstroms as long as you do the same for both. What is \pi doing there anyway, except to make it spuriously “scientific” looking. Then using the well-known approximation for the square root of a number near 1 (which this will always be, you can approximate the formula by 182.65(w/h\pi)^2-1.3
For my own body, the approximation differed from the “exact” formula by an insignificant .05. But what this really shows is that the significant factor is (w/h)^2 and the rest is scaling. Probably, to avoid very small numbers you might make scale based simply on (h/w)^2. It might or might not mean much but it avoids the spurious bull.
Dang, I was about to write that post, including the square root approximation. I noticed how silly their formula was when reading the paper earlier. And they could have left out the useless 1.3 offset and multiplied by some nicer round number like 20 (leaving out the \pi). Heck, you don’t even really need to square it since it’s monotonic and we’re just looking at thresholds.
