Why BMI is flawed – and how to redefine obesity

https://www.nature.com/articles/d41586-023-03143-x

by Brajeshwar • 3 years ago

23 74 3 years ago

nradov4• 3 years ago

Good article in general, but it's odd they didn't mention DXA scans. Those are now the gold standard for body composition measurement and are fairly cheap in most developed countries.

streptomycin• 3 years ago

Also seems to not mention that the biggest inaccuracy of BMI is that it tends to underestimate the rate of obesity (bodybuilders with the opposite problem are a rounding error in population-level statistics, many more people have very low muscle).

brightlancer2• 3 years ago

It does, though it's almost in passing:

"For instance, Asian people tend to be at a greater risk of conditions such as heart disease at lower BMIs than are white people, probably owing to differences in body-fat percentage and distribution. The WHO recommends that Asian populations use lower BMI cut-offs for overweight and obesity, which several Asia-Pacific countries now do."

johncessna• 3 years ago

Any athlete has known this for ages.

smcl• 3 years ago

Well many people know it for ages. I was thought it in school (as in … when I was like 15 doing Standard Grade Biology). But somehow it is still surprisingly commonly used as a general way to categorise people into groups for various purposes.

oldpersonintx2• 3 years ago

active athletes tend not to have chronic obesity issues, outside of shotput and competitive face slapping

OJFord2• 3 years ago

I think this overcomplicates the flaw, a very simple way of looking at it is that muscle has approx 5x the density of fat.

So if you take someone (or two people) slim, small frame, and compare them piling on muscle vs. fat, the former is obviously healthier, and yet will see them be 'overweight' per BMI far sooner.

I don't have any clinical experience, IANAD, so take with mountains of salt etc., but I'd have thought a simple change that would go a long way would be to have say three BMI ranges instead of one (well, for a given country anyway!) categorised by eye-balling or self-reporting.

avrionov• 3 years ago

It is much harder to measure muscle mass at home, or even at the doctors office. Different methods show very different results when people are measured at the same time.

We need a simple proxy method for estimating the healthy weight range.

OJFord2• 3 years ago

I didn't suggest measuring muscle mass, I said you could categorise a person (or yourself) into one of say three BMI ranges according to whether you were small/mid/big 'build'.

Obviously not as good as actually measuring muscle mass, or anything more scientific, but I think it would be a pretty trivial way of making the BMI estimate another chunk better.

kyleyeats• 3 years ago

Muscle is about 15% more dense than fat.

gus_massa• 3 years ago

I agree, but eye-balling and self reporting is too subjective. In some countries you get priority in the covid-19 vaccination line if you have a high BMI because it was correlated to deaths.

Perhaps add another variable like volume to distinguish fat and muscle (measured as Archimedes in a bathtub), but as you notice it break the reason that makes BMI so popular. From the article:

“BMI doesn’t cost anything — it’s cheap and quick,”

chrisbrandow• 3 years ago

I'm not trying to be "clever" when I say that I thought it was common knowledge among professionals or the large audience of Nature that BMI is only valid at the population level.

Symmetry• 3 years ago

Even regardless of that, we have hand calculators now so raising height to the 2.6th power instead of squaring it isn't too hard a piece of math to ask a doctor or military recruiter to be doing on their own.

Fin_Code• 3 years ago

BMI is the least worst. Tons of methods have been tried and they are all worse. We need cheap, fast and easy. Thats what BMI is. To get more accurate you need specialized equipment and or some math knowledge. Better = costs more. Thats why we have BMI and why people complain endlessly about it.

RC_ITR2• 3 years ago

But it could be a more complex formula.

For example, the max 'healthy' weight for a 6'10" person is 230lbs.

That's more or less Kevin Durant's draft stats and he looked like this:

https://twitter.com/si_vault/status/749993324195479553

confidantlake4• 3 years ago

For sure, but most people are not 6'10. For 95%+ of the population it works fine.

orangepurple2• 3 years ago

Differences in body proportion, fat-free body density, and hydration may have a larger effect on the validity of body-composition devices in minority populations than previously assumed. Based on the review of literature, DXA is a valid method in a multi-ethnic sample, if individuals are Caucasian/White and African American/Black. However, there is insufficient evidence to recommend use in Hispanic/Latinx and Asian adults, Native American males, or African-American/Black females. ADP is valid for Hispanic and African-American/Black males when utilizing race-specific equations; however, results are inconclusive in other racial/ethnic groups and sexes. For BIA, body-composition estimates may be valid in a multi-ethnic sample, but the literature demonstrates disparate results between races/ethnicities. BIA may provide valid results in Hispanic and Native American populations, as well as Asian populations utilizing race-specific equations. However, BIA is still not recommended for African-American/Black individuals based on current data. The lack of validation using a multi-compartment model criterion limits the certainty of conclusions, particularly regarding the validity of ADP and BIA.

Glossary:

ADP: air displacement plethysmography

BIA: bioelectrical impedance devices / bioelectrical impedance analysis

Reference: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8528114/

Gwypaas2• 3 years ago

Arnold Schwarzenegger hade an obese BMI in competition trim. Outliers does not invalidate a generic measure, they are outliers.

https://upload.wikimedia.org/wikipedia/commons/0/05/Arnold_S...

orangepurple2• 3 years ago

If you are white use this formula to calculate body fat percentage with a standard error of the estimate (SEE) of 4.74%:

We studied a sample of 6,510 white subjects (2,154 men, 4,356 women), aged 18–80 years, including patients visiting our department. The study was performed to valuate the usefulness of a new equation: BF% = –44.988 + (0.503 x age) + (10.689 x sex) + (3.172 x BMI) – (0.026 x BMI^2) + (0.181 x BMI x sex) – (0.02 x BMI x age) – (0.005 x BMI^2 x sex) + (0.00021 x BMI^2 x age) where male = 0 and female = 1 for sex, and age in years, developed by multiple regression to predict BF%

This online calculator claims to use this algorithm:

https://www.omnicalculator.com/health/body-fat

Reference: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3263863/pdf/383...

nradov4• 3 years ago

Waist to hip ratio is the least worst. It's just as cheap, fast, and easy. And it's a more accurate predictor of mortality.

https://www.medscape.com/viewarticle/981431

tomjakubowski• 3 years ago

this makes sense - carrying excess abdominal fat is a sign of poor metabolic health. and taking a ratio of circumferences instead of weight/height accounts for muscle being denser than fat

swatcoder4• 3 years ago

It's almost impossible for people to consistently measure, which throws self-assessment out the window, and remains a challenge in clinical settings.

Pot belly guy measures his waist at his ribs, pear shaped lady rests the tape above her saddle bags, tape isn't level this time, tape is slack that time, etc etc

It might only require tape and might have useful health correlates given good values, but getting those good values is just not something epidemiologists or even clinicians can rely on in the field.

wil421• 3 years ago

Exactly, BMI is an OK gauge for general use.

When I was younger and in shape bulked up with single digit body fat it was not useful. I was obese/overweight in BMI only.

Now that I am more sedentary due to office job and less time because I have kids it’s a useful gauge to help find my ideal weight.

jerrycruncher• 3 years ago

And the people who are more muscular _know_ that BMI isn't going to be accurate for them. The complaints seem to be somewhat bad faith arguments from the population for whom it is accurate, but wishes it weren't.

chrisfosterelli2• 3 years ago

Yeah I've always felt that it's unrealistic to expect a single number to effectively represent a person's health. Of course it's useful and of course there's caveats and of course a proper health evaluation should have a more holistic view than just BMI.

bagels• 3 years ago

I thought it was meant more to assess how fat someone is, not health?

Gwypaas2• 3 years ago

Is not this common knowledge? It works for population measures and mostly everyone not exercising at a very high level. In other words, 9x% percent of the population can not excuse their BMI by "muscle mass", it is simply due to a sedentary lifestyle and the BMI is a correct estimate.

DoneWithAllThat• 3 years ago

These omissions are why the AMA adopted a policy in June to downplay the clinical use of BMI, calling it “an imperfect measure” that has been used for “racist exclusion” and that has caused “historical harm”. The policy suggests combining other metrics, such as waist circumference or body composition, with BMI.

Without even reading the article I already knew there would be some identity politics nonsense underpinning it. Because it’s Nature, and they have no credibility anymore.

KittenInABox• 3 years ago

I don't think this is identity politics. It's well known there's ethnicity differences in health. East Asians are more at risk for diabetes at lower BMIs than Whites, for example. There's nothing political about saying BMI categories for overweight/obese/etc are skewed towards what's normal for the average White person specifically, which affects everyone else.

confidantlake4• 3 years ago

It is incredibly racist of them to imply other races somehow can not maintain normal body weight. This directly harms the very people they claim to be helping.

sxp2• 3 years ago

BMI is imperfect, but it's better than the other fast & cheap measurements of obesity. A DEXA scan & full bloodwork would be better, but that takes a lot of time and money. A scale with body fat measurements is nice if you pay for it and use it at home. But basic BMI is good since it's so easy to calculate for *average* people. It's not useful if you're an athlete, but people with that much muscle probably have access to better measurements.

These omissions are why the AMA adopted a policy in June to downplay the clinical use of BMI, calling it “an imperfect measure” that has been used for “racist exclusion” and that has caused “historical harm”.

That sounds like the root justification for avoiding BMI.

nradov4• 3 years ago

The scales with electrical body fat measurements are crap, little better than random number generators. I have compared results on a few of those with actual DXA scans and the scales are way off. The readings sometimes also vary by 2% from day to day, which isn't physiologically possible. Don't waste your money.

RandallBrown2• 3 years ago

I don't really care if my scale is accurate, I care if it's consistent.

Mine seems to correctly report my body fat going up or down as I gain or lose weight.

nekoashide• 3 years ago

I have an app with my scale and I weigh every morning, the numbers never make sense but the trends do. Like over a whole month I see not only weight loss but other metrics that may be deviating.

sxp2• 3 years ago

The scales are noisy, but the trendlines they provide are useful and match the visual appearance of my body. The biggest source of noise that I see is when I have a lot of carbs or sugar in a day and then retain water. Then I eat fewer carbs the next day and get rid of the excess water.

antisthenes2• 3 years ago

It's not flawed for 99.5% of people.

“BMI is a somewhat crude metric for determining health risks,” says Susan Yanovski

Of course it's crude, it's a single number, as opposed to thousands of potential risk factors unrelated to BMI.

Adult BMI charts don’t reflect this variability. And the cut-offs between BMI categories — which are the same in most countries that use the index to diagnose obesity — are “something between science-based and arbitrary”

Again, not the point of BMI. You can have 40% fat or 42% fat in your body, you're still obese, and that means increased risk and a call to take action.

And because BMI was developed using measurements from white people, people of colour “don’t quite fit into these narrow parameters”

...Commenting this just means stooping down to their level, so I won't.

I understand that personal weight is an emotional topic for people, but holy crap is this article an awful strawman-fighting pile of garbage.

Nature, you can do better than that.

jmye3• 3 years ago

Do you actually have a counter argument to the fact that a measurement based on one race or sex may not be applicable to all races/sexes? This is common knowledge in health care, so I’m quite interested in hearing your evidence-based reason for discounting it, sorry, calling it a “pile of garbage”.

You do have one, right?

antisthenes2• 3 years ago

Yes.

While variations may be present, the fact is that someone who is race X and position Y on the BMI scale in the obesity category still has the same risk factors from increased weight as someone who is race Z and position Y.

You do have hard numbers about what the differences are in terms of percentages for the risk factors.

right?

brightlancer2• 3 years ago

1/3 of Americans are "obese". Not "overweight", _obese_.

Sure, there are and will be variations among populations (and individuals), but that variation is less meaningful when 99.999% of obese people are at unhealthy weights.

Instead, it appears to be a con to excuse unhealthy behaviors, particularly by playing the trump card of "that's racist!" as the Nature article does.

confidantlake4• 3 years ago

Because as a rough estimate it is fine. If you are a 40 bmi it doesn't matter what your race or sex is, it is a problem.

fsckboy• 3 years ago

what happened to waist-to-hip ratio, in the sense that nobody is mentioning it wrt this study and discussion? It's simple as hell, and I thought it was already considered better? Could they be combined, or do they share the same weaknesses?

hospitalJail4• 3 years ago

That's pretty interesting.

I imagine the reason it isnt used is because the surface is complex. Its not as simple as standing on something writing a number, standing next to something and writing a number. You need to keep the angle correct and make sure the starting location is correct.

Wonder if you could use 2 side photos to calculate it, or 1 overhead.

swatcoder4• 3 years ago

It's just extremely hard to measure well. It's good in a research setting where diligence can be applied, but not at home and not when you're turning over dozens of patients a day in a clinic.

oldpersonintx2• 3 years ago

people will do anything these days to keep from reality checking themselves on obesity

you can't give any advice - everything is "correlation does not equal causation" or "unless you have a peer reviewed paper you can't tell me anything"

eat vegetables and exercise. if you can't even make that limited leap of obvious common sense, then just accept your reduced expected lifespan and overall poor quality of life

photochemsyn• 3 years ago

This seems to be the practical issue: if more precise tests like hydrodensitometry were needed before prescriptions were issued, fewer drugs would be sold:

"That’s especially concerning with the exploding demand for new anti-obesity medications. In the United States, a person can be prescribed semaglutide for weight loss simply if their BMI is 30 (the lower bound of the obesity category) or more. With a BMI of between 25 and 30 — the overweight category — a person needs just one weight-related condition, such as high blood pressure, to get a prescription. “If you have 15 minutes in a visit,” Stanford says, “you’re going to follow the BMI … instead of really taking the time to see ‘Do they need the medication?’”

For a list of more accurate measurement approaches:

https://www.medicalnewstoday.com/articles/326331

efitz• 3 years ago

When I was in the best shape of my life, in the military, my BMI was "too high" even though I was lean and of healthy proportions. I've always taken it with a grain of salt since then.

barbazoo• 3 years ago

I think BMI isn't relevant when one's already in the best shape of their lives. It's an ok metric when you're not.

hospitalJail4• 3 years ago

Similarly I was at the end of bulk season and 212lbs. I was loaded with food from yesterday, creatine water weight, and packed with muscle. I took some work health screening thing, instead of the perfect score 20/20, I got a 17/20 because I was 2lbs too heavy according to BMI. This ended up costing me some trivial about of money like $6/mo.

I cut down to 173lbs, and did the cycle again. The following year, I timed my cut better so I was only 185lbs at the time of the screening, got 20/20.

swatcoder4• 3 years ago

You were an exception and (obviously) could see that for yourself. Your doctor probably agreed.

The value of BMI in epidemiology is fantastic because it applies well for almost all normal people and is easy to capture all over the world.

The clinical value is weaker, but mostly because people like to rationalize that they too are also in exceptional athletic form. By definition, most are not.

jchw2• 3 years ago

Here is my biggest problem:

I'm fat. It's that simple, I'm fat. Why am I fat? Because I ate too much and exercised too little.

Why did I eat too much and exercise too little? Well, there's not really one reason; portion size norms are kind of bad here, I ate a lot at night, I stress-eat sometimes, and my upbringing didn't instill good exercise or eating habits for the most part. Doesn't really matter that much, at least not in my case.

How do I know that I am fat? By looking down. I have fat all over my body. And I've lost a lot of weight recently, but I can look down and see that there is more to go before I am rid of the remainder of clearly excess fat on my body. I understand that some people do have a tendency to eating disorders and a distorted view of themselves, but that's not a problem I've ever had, so I'm fine relying on my own senses.

So needless to say, I don't really give a shit about BMI. I think most people ought not put that much stock into it, because most people can, if they're being honest with themselves, tell if they're overweight by looking in a mirror.

But... I think that if anything, the big problem with most people is not that people are too concerned with BMI, it's that people are not concerned enough about the side-effects of their terrible diets and lack of exercise. I do not think that our standards are overall too high, I think our standards are too low, at least here in America. Granted, I'm not exactly a health nut, but seriously: when I was talking to my mother, and I said I have a goal to get down to at least 180 lbs, she thought I would look anorexic at 180 lbs. I had to show her pictures of people who weigh 180 lbs to get her to believe that that's a completely reasonable target weight.

Some segment of the population definitely are too obsessed with BMI and weight loss, and probably the same for some health specialists. But for all of the hand-wringing about that issue, and movements like HAES, somehow people seem to miss the part that your horribly unhealthy habits are not more healthy just because BMI is an estimate and not some sort of absolute fact. Just taking a shot in the dark, if your BMI is >35, I would guess that regardless of the population-level inaccuracies of BMI, it's very likely a good indicator of an unhealthy lifestyle, or at least, a dramatically unusual physique. If I am wrong, I'd love to know.

But what's more important is not eating absurd portions of junk food every day, and getting up and walking once in a while. The whole culture here of "dieting" as a thing you do to lose weight rather than as a lifestyle adjustment is crazy. I know it's kind of cliche by now, but as someone who mostly ignored this sort of thing for years, even I must admit that my view of what a healthy diet and lifestyle looked like was very distorted by weird American cultural things. Amusingly, the only thing that actually opened my eyes to a lot of it was realizing how much I didn't agree with some of the things that HAES advocates on the Internet said. In a roundabout way, it helped me stop rationalizing my weight problems internally and realizing that my weight problems were lifestyle problems, by virtue of almost spite.

I'm not sure how much of a hot or cold take this rant was, because I generally do my best to avoid reading too many online flame wars about lifestyle things, so I apologize ahead of time if this rant is just full of either flamebait or cliches, I genuinely do. But, this article really annoys me, because wow, I don't understand how BMI continues to be such a massive source of concern as obesity continues killing the shit out of us and ruining our quality of life. I'm starting to feel like there's got to be some crazy conspiracy going on.

hospitalJail4• 3 years ago

The whole culture here of "dieting" as a thing you do to lose weight rather than as a lifestyle adjustment is crazy.

I think it depends on the person. I diet once a year and that diet keeps me at 6pack abs for almost the entire year until around November/December. A 3-4 month diet allows me to eat whatever I want the rest of the year. This gives me a chance to grow muscle and not put brainpower into food.

When you are far overweight and you can't quite see the finish line, maybe it is a lifestyle thing.

jchw2• 3 years ago

Just to me personally, that seems reasonable enough. It seems reasonable to have a diet that you adjust on and off. I plan to do the same thing once I hit my target, and for me, the thing that seems to work well is something between OMAD and IF where I just do one major meal and barely eat outside of it. I don't seem to get any of the negative side-effects, so it works great for me.

But the culture of fad diets and "90 day weight loss" routines as a cure for their unhealthy lifestyles is ridiculous. The lifestyle a lot of people are living is not healthy to do for an entire month, less most of the year. And I'm not sure I'd say that's "most people" in America, but I'd venture to guess it's a whole lot of them.

Vt71fcAqt7• 3 years ago

What is your non dieting diet? What is your TDEE? Do you exercise? You generally (always?) cannot maintain visible abs if you are eating above your TDEE.

bazmattaz2• 3 years ago

Loved this post. A lot of what you said resonated with my story. You’re so right about diet culture- it exists to make money from diet related products.

I can imagine it’s hard in the states (although here in the UK we’re on the same track) with huge portion sizes, lots of fast food and lots of processed food. It’s hard, and expensive to cook and eat healthy these days.

I lost a bunch of weight recently. What really worked for me was, - fasting between 6:30pm and 10:30am. No late eating was a game changer for me - exercise. I do HIIT classes 4 times a week - trying to cook healthier - cut down alcohol

zingababba• 3 years ago

Pretty funny, at my most muscular I was considered obese on BMI charts but could still (barely) see my abs. I do feel waaaaaaaaaay better at a lower BMI though, currently at a 22. I'm now of the opinion that tissue is tissue and any amount of mass isn't great for the heart.

drbig3• 3 years ago

I'd like to share a different approach, however limited it is I find it still surprisingly useful:

For an adult male the test is to:

- Bench press 80% of your weight

- Squat with 100% of your weight

- Dead lift 150% of your weight

Yes, this is muscle/power specific and not good at all for children and the elderly (and women...), but being a lower-bound and "auto-scaling" (because % of body mass) and not requiring any medical equipment...

Perhaps give it a try next time you're at the gym :-)

goosinmouse3• 3 years ago

Im going to begin with saying that nobody should go to the gym and test what you said. An untrained male that is 200lbs will absolutely get hurt if they try to attempt a 160lb bench press, 200lb squat, or 300lb deadlift. In fact i'd argue that they would be more likely than not to get a serious injury if they were untrained and actually attempted any of those. Untrained numbers are going to be half of what you said to test.

What is this supposed to be an approach for? Overall fitness level?

Although im going to add a disclaimer for anyone reading this that wants to give this a try.

According to StrengthLevel, which is community sourced stats, for an average male on their site which skews exclusively towards people who do strength training here are the bodyweight ratio's for beginners, who are usually classified as AT LEAST a month of training to 6 months

- Bench press .50x of weight - Squat .75x of weight - Deadlift 1.00x of weight

Moving up to novice which is 6 months to 2 years the numbers are in line with what you shared. So people who haven't lifted should definitely not try to test their strength by going straight to novice numbers on the bar. They WILL hurt themselves if they put anywhere close to that weight and have no spotter or spotter bars.

confidantlake4• 3 years ago

Definitely this. Also even if you can say bench press 160lbs you should be doing several warm up sets starting from an unweighted bar. Just throwing on 160lbs cold and going for it is madness.

drbig3• 3 years ago

Thanks for discussing! I overall agree.

First lemme be explicit that I didn't imagine anyone sane who has not been to a gym ever - or even in a long time - to plop in ~100% of their weight for anything. Warm up and going there in increments was presumed by me as the standard procedure.

- Bench press .50x of weight - Squat .75x of weight - Deadlift 1.00x of weight

Yeah, I wasn't sure about the percentages, just the general relation between them. My fault here most likely! I don't remember how many lives ago I heard about this "test". Your numbers sound reasonable too.

But it was clearly after I done my work (out, over 3 years) of getting from "nothing" to "something".

What is this supposed to be an approach for? Overall fitness level?

I'd say more of the power (strength) level. There is a lot more to fitness.

bazmattaz2• 3 years ago

This is a terrible idea. You can be fit and healthy but not be able to lift really heavy things. As someone who lifts heavy things a few days a week I know it’s no joke and not for everyone

Why BMI is flawed – and how to redefine obesity

Related Stories