Hacker News Reader

At a population level, it's great. For individuals, it can fail in all kinds of ways. For me: it is not very good because my body has very strange proportions. I'm 6'2", but my waist is barely higher than my 5'1" wife, and my torso is nearly the same length as my 6'7" brother.

In other words, for my height, I have a very long torso and very short legs. I think it should be relatively obvious that an inch of leg weighs significantly less than an inch of torso, so at a given level of body fat percentage, I'm going to weigh quite a bit more than someone my same height with more typical proportions, and thus my BMI reads me as more overweight than it otherwise would.

My point is not that BMI is bad or useless or anything else. My point is that it was designed as a population statistic and that it can be fraught when one tries to apply it to any individual with no nuance. A high BMI should cause one to consider and examine your health and weight. But it should not over-ride specific details about your physiology that point in the other direction.

It's actually still useful. It doesn't function as well as a measure of fatness for you as an individual, but that was only ever a proxy anyway.

It still functions well as a measure of load on your circulatory system, joints, you name it.

Your body is working harder than a typical 6'2" person, even if you have a great waist-to-hip ratio.

(But you'd still be better off with a smaller WTH ratio!)

Do you have a concrete example, though? Like an actual body fat percentage measurement that conflicts with what BMI explains for you?

I was also in the "overweight" category at one point while being low body fat. It was never a problem for anything, anywhere. I wasn't going to doctors who looked only at the BMI sheet and then told me to lose weight, because everyone could see that I was not overweight.

Body fat percentage can't conflict with BMI because BMI is not an estimate of body fat percentage.

And I agree that it's never caused me problems, but that's because I and, as you suggest, doctors who look at the number, have understood that it's not that big of a problem in my case. That's what I mean by not looking at the BMI number naively and ignoring other physiological states.

I don't think anything you said conflicts with anything I said, and it sounds like we mostly agree.

The serious backwoods rock climbing cohort calls that "goon coefficient". I (also) have a high goon coefficient.

Yeah, people don't want to understand that your body structure messing with BMI is quite real but only a few points. It's unquestionably "wrong" in scoring me higher than my wife despite the fact that she's the one with a few extra pounds. But it scores us both as normal, the difference doesn't matter.

Anything to pretend they aren't too heavy.

> screening tool

People tend to think in absolutes. Even perfectly rational people fall logically foul to not considering outliers.

> I’ve been in the “overweight” BMI range with low body fat before. It’s not too hard to get there

Did you do that as a natural lifter? It’s hard for me to imagine most guys being able to get into the “overweight” range at < 15% body fat without some assistance.

Yes, sorry, I meant as someone lifting weights and working on it deliberately.

For someone not working out or working a job that builds muscle, getting into the overweight category probably does mean they should take a look at their body weight.

I suspect he was not nearly as lean as he suspected. Everyone over estimates how much fat they have to lose until they actually try to get shredded. My own efforts let me to discover that despite lifting weights in a dedicated fashion for years, BMI is still relatively accurate for me. Might be off by 15-20 lbs at the most, but still somewhat useful as a diagnostic tool since it if it says I'm obese, I am at the very least pretty overweight. It's only at the margin of overweight/healthy where it becomes only a tiny bit problematic.

The vast majority of natural lifters will never get enough muscle to really seriously throw it off.

The farther from average height you are the worse it is. If you are from some part of the world where the average height distribution is near the point BMI based itself you will perceive it to be so much better.

I am only 6' and only an intermediate lifter. I'm overweight, but BMI makes me look obese. It really only takes a normal person a year or so to get to three plates on squat with no supplements other than chicken broccoli and rice if they don't skip workouts. Assuming no injuries. That level of strength easily distorts bmi wildly if you are even a little taller than average.

IIRC, the farther you are from an "ideal" 19th century Frenchman, the less BMI says about you.

What percentage of any random western country population do you think has reached three plates on squat, and therefore has to consider this?

>Getting all the way to the obese BMI range while having healthy body fat is only happening for people with an extreme dedication to body building and diet

And sumo wrestlers.

The line seems to be drawn at (effortful) activity level more than anything else. And BMI doesn't say anything about that.

I think the vexation comes from the focus on an attribute that is not directly mutable, per one's agency, because that's easier to sell things related to. Versus action that you have a lot more direct control over (social or physical circumstances notwithstanding).

> The lifestyle has a negative effect on their health, with sumo wrestlers having a much lower life expectancy than the average Japanese man.

(from https://en.wikipedia.org/w/index.php?title=Sumo&oldid=136533...), so I'm not sure this is a counterexample.

Average sumo wrestler bodyfat% is ~25–27%

It's complicated. If you actually look at the citations, they were looking at - and kind of necessarily so, if you think about it - a cohort of wrestlers who were born more than a century ago, and not later than mid-century; so, not the current generation. The article puts an emphasis on retained weight after retirement, head injuries, and extensively cites a Discovery Channel article from 2005 for many of the most damning claims (https://web.archive.org/web/20050831220303/http://discoveryc...), including the one about the sumo diet and weight causing metabolic syndrome - even though, earlier in the Wikipedia article, it's mentioned that more recent scholarship shows the remarkable (relative) metabolic health of active wrestlers.

There's an enormous amount of wear-and-tear on their bodies, but the current cohort of wrestlers have access to much better knowledge about what's happening to them during their careers and afterwards, when compared to previous ones. During their prime years, they're objectively more metabolically healthy than the general population. Once retired, they're probably better able to manage their weight than earlier generations, and know to manage incidental issues like sleep apnea and alcohol consumption. I also wonder about the effects the tight-knit stables have on them, since we know positive and abundant social connection early in adulthood has ramifications later on.

>so I'm not sure this is a counterexample.

Yeah, don't just cite Wikipedia blindly. Some of the effects described seem strong, but there are caveats.