NYT: Is BMI a scam?

[Anonymous]

https://www.nytimes.com/2021/05/18/style/is-bmi-a-scam.html?action=click&algo=bandit-all-surfaces&block=editors_picks_recirc&fellback=false&imp_id=773181498&impression_id=487210da-b8d2-11eb-9405-7145593dc1bd&index=1&pgtype=Article&pool=editors-picks-ls®ion=footer&req_id=836519829&surface=home-featured&variant=0_bandit-all-surfaces

Interesting article on how the BMI calculation was developed, what it may or may not personally signify for you, and other measures of health than solely weight.

[Anonymous]

"This explains why muscular athletes often have high B.M.I.s despite having little body fat" is only true for men. There's a literature on this. Women can't put on enough muscle naturally to have high BMIs without also having high body fat.

In general, whether this is an accurate indicator is separate from whether it makes sense to focus on. I'd argue it doesn't make tons of sense to focus on, but the reason I'd give is one that they don't even mention, which is that long-term, significant weight loss is really hard. Even if being overweight is making you unhealthy, if you're not going to change that and it's really stressful to think about, then it's better to focus on things you can change. But that doesn't mean we have to say it's a bad indicator, and that it makes people feel bad is totally separate from whether it has predictive value. (And the bit about "People who have felt discriminated against because of heavier weight are also about 2.5 times more likely to have mood or anxiety disorders" suggests a pretty obvious alternative causal explanation.)

[Anonymous]

Thank you for posting this. I've become increasingly frustrated with the focus on BMI both in the real world and on DCUM. It's fine for population studies. It's not useful for individuals. And I say this as a "thin" person for whom BMI makes me seem healthier than I am. I'm lucky my parents taught me relatively healthy eating habits, but I spent most of my life doing little to no strength training and now its catching up with me. We need to get away as a society from this idea that being thin automatically makes you healthy, and being "overweight" automatically makes you unhealthy. Especially when judging other people.

[Anonymous]

Anonymous wrote:"This explains why muscular athletes often have high B.M.I.s despite having little body fat" is only true for men. There's a literature on this. Women can't put on enough muscle naturally to have high BMIs without also having high body fat.

In general, whether this is an accurate indicator is separate from whether it makes sense to focus on. I'd argue it doesn't make tons of sense to focus on, but the reason I'd give is one that they don't even mention, which is that long-term, significant weight loss is really hard. Even if being overweight is making you unhealthy, if you're not going to change that and it's really stressful to think about, then it's better to focus on things you can change. But that doesn't mean we have to say it's a bad indicator, and that it makes people feel bad is totally separate from whether it has predictive value. (And the bit about "People who have felt discriminated against because of heavier weight are also about 2.5 times more likely to have mood or anxiety disorders" suggests a pretty obvious alternative causal explanation.)

But it IS a bad indicator at the individual level. Did you see the other thread where people are splitting hairs over whether a 5'4" woman who wants to get back to 112 has disordered thinking because 112 is either 1-2 or 3-4 pounds away from being underweight? This makes absolutely no sense at all, but people think the idea of being underweight/overweight based on BMI is some super scientific measure. It's not, and it means very little.

[Anonymous]

Anonymous wrote:

Anonymous wrote:"This explains why muscular athletes often have high B.M.I.s despite having little body fat" is only true for men. There's a literature on this. Women can't put on enough muscle naturally to have high BMIs without also having high body fat.

In general, whether this is an accurate indicator is separate from whether it makes sense to focus on. I'd argue it doesn't make tons of sense to focus on, but the reason I'd give is one that they don't even mention, which is that long-term, significant weight loss is really hard. Even if being overweight is making you unhealthy, if you're not going to change that and it's really stressful to think about, then it's better to focus on things you can change. But that doesn't mean we have to say it's a bad indicator, and that it makes people feel bad is totally separate from whether it has predictive value. (And the bit about "People who have felt discriminated against because of heavier weight are also about 2.5 times more likely to have mood or anxiety disorders" suggests a pretty obvious alternative causal explanation.)

But it IS a bad indicator at the individual level. Did you see the other thread where people are splitting hairs over whether a 5'4" woman who wants to get back to 112 has disordered thinking because 112 is either 1-2 or 3-4 pounds away from being underweight? This makes absolutely no sense at all, but people think the idea of being underweight/overweight based on BMI is some super scientific measure. It's not, and it means very little.

Just because something is being misused doesn't mean it doesn't have explanatory value in another context. And with any continuous variable that we split into discrete categories, there's going to be some amount of arbitrariness close to the lines. Is "18" really that different from "19"? Of course not. But if someone shows up with a BMI of 17 in your office, there are questions you should be asking them that you would not ask someone at a 'normal' weight, not because there's no overlap between the answers to those questions between underweight and normal-weight people, but because it's a matter of probability.

[Anonymous]

I have mixed feelings about this. I think all these things are true:

1) Doctors rely too much on obesity being the be all end all health problem source of a fat person and this results in bad care

2) For people in the 'normal' BMI category it can mask health problems unrelated to weight because they believe they are healthy

3) For people who are 36 BMI (like I was!) it can be a useful to be a really be a sure sign that you need to do something about it and for me personally I used the 'BMI doesn't matter' to blow past a lot of numbers I should have cared about.

So like it isn't the be all and end all but I think it has its uses.

[Anonymous]

Unless you are an NFL linebacker, you’re high BMI is because you are fat not because you are well muscled. Yes NMI breaks down at the margins, but most people aren’t at the margins

[Anonymous]

Anonymous wrote:"This explains why muscular athletes often have high B.M.I.s despite having little body fat" is only true for men. There's a literature on this. Women can't put on enough muscle naturally to have high BMIs without also having high body fat.

In general, whether this is an accurate indicator is separate from whether it makes sense to focus on. I'd argue it doesn't make tons of sense to focus on, but the reason I'd give is one that they don't even mention, which is that long-term, significant weight loss is really hard. Even if being overweight is making you unhealthy, if you're not going to change that and it's really stressful to think about, then it's better to focus on things you can change. But that doesn't mean we have to say it's a bad indicator, and that it makes people feel bad is totally separate from whether it has predictive value. (And the bit about "People who have felt discriminated against because of heavier weight are also about 2.5 times more likely to have mood or anxiety disorders" suggests a pretty obvious alternative causal explanation.)

Alternately, people with mood and anxiety orders are prone to assuming discrimination that was only in their head.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:"This explains why muscular athletes often have high B.M.I.s despite having little body fat" is only true for men. There's a literature on this. Women can't put on enough muscle naturally to have high BMIs without also having high body fat.

In general, whether this is an accurate indicator is separate from whether it makes sense to focus on. I'd argue it doesn't make tons of sense to focus on, but the reason I'd give is one that they don't even mention, which is that long-term, significant weight loss is really hard. Even if being overweight is making you unhealthy, if you're not going to change that and it's really stressful to think about, then it's better to focus on things you can change. But that doesn't mean we have to say it's a bad indicator, and that it makes people feel bad is totally separate from whether it has predictive value. (And the bit about "People who have felt discriminated against because of heavier weight are also about 2.5 times more likely to have mood or anxiety disorders" suggests a pretty obvious alternative causal explanation.)

But it IS a bad indicator at the individual level. Did you see the other thread where people are splitting hairs over whether a 5'4" woman who wants to get back to 112 has disordered thinking because 112 is either 1-2 or 3-4 pounds away from being underweight? This makes absolutely no sense at all, but people think the idea of being underweight/overweight based on BMI is some super scientific measure. It's not, and it means very little.

Just because something is being misused doesn't mean it doesn't have explanatory value in another context. And with any continuous variable that we split into discrete categories, there's going to be some amount of arbitrariness close to the lines. Is "18" really that different from "19"? Of course not. But if someone shows up with a BMI of 17 in your office, there are questions you should be asking them that you would not ask someone at a 'normal' weight, not because there's no overlap between the answers to those questions between underweight and normal-weight people, but because it's a matter of probability.

There's plenty of evidence that doctors treat obese people horribly, so even in a medical context BMI isn't really being used in a productive way. Why can't doctors just look at more specific markers of health? People with high BMIs can be healthy but instead they have their medical issues ignored and they get treated as though any issue they have is their own fault for eating too much.
From the article:

. In a 2016 study of more than 40,000 adults in the United States, researchers compared people’s B.M.I.s with more specific measurements of their health, like their insulin resistance, markers of inflammation and blood pressure, triglyceride, cholesterol and glucose levels. Nearly half of those classified as overweight and about a quarter of those classified as obese were metabolically healthy by these measures. On the other hand, 31 percent of those with a “normal” body mass index were metabolically unhealthy.

[Anonymous]

There's no excuse for using a measure that was developed based on the bodies of White men on women and people of color. And this article makes me wonder - what data were used to develop BMI categories for children?

[Anonymous]

Anonymous wrote:There's no excuse for using a measure that was developed based on the bodies of White men on women and people of color. And this article makes me wonder - what data were used to develop BMI categories for children?

Is there evidence that belly fat is healthy in people of color?

[Anonymous]

Anonymous wrote:

Anonymous wrote:There's no excuse for using a measure that was developed based on the bodies of White men on women and people of color. And this article makes me wonder - what data were used to develop BMI categories for children?

Is there evidence that belly fat is healthy in people of color?

A. BMI doesn't measure belly fat
B. Yes, there is evidence that the relationship between high BMI and mortality is much weaker for people of color and especially women of color https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3844096/

[Anonymous]

Anonymous wrote:Thank you for posting this. I've become increasingly frustrated with the focus on BMI both in the real world and on DCUM. It's fine for population studies. It's not useful for individuals. And I say this as a "thin" person for whom BMI makes me seem healthier than I am. I'm lucky my parents taught me relatively healthy eating habits, but I spent most of my life doing little to no strength training and now its catching up with me. We need to get away as a society from this idea that being thin automatically makes you healthy, and being "overweight" automatically makes you unhealthy. Especially when judging other people.

+1 This is what our pediatrician told us, she doesn't refer to BMI charts. My DD is muscular and has a very heavy frame and measures high on BMI even though she doesn't look overweight and her doctor thinks she's right where she should be. Even as a baby/toddler my friends would pick her up and be shocked at how heavy she was. She's just solid. So many things go into weight that have nothing to do with fat storage and how healthy you are or not -- muscle mass, bone size, bone density, breast size (for women).

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:"This explains why muscular athletes often have high B.M.I.s despite having little body fat" is only true for men. There's a literature on this. Women can't put on enough muscle naturally to have high BMIs without also having high body fat.

In general, whether this is an accurate indicator is separate from whether it makes sense to focus on. I'd argue it doesn't make tons of sense to focus on, but the reason I'd give is one that they don't even mention, which is that long-term, significant weight loss is really hard. Even if being overweight is making you unhealthy, if you're not going to change that and it's really stressful to think about, then it's better to focus on things you can change. But that doesn't mean we have to say it's a bad indicator, and that it makes people feel bad is totally separate from whether it has predictive value. (And the bit about "People who have felt discriminated against because of heavier weight are also about 2.5 times more likely to have mood or anxiety disorders" suggests a pretty obvious alternative causal explanation.)

But it IS a bad indicator at the individual level. Did you see the other thread where people are splitting hairs over whether a 5'4" woman who wants to get back to 112 has disordered thinking because 112 is either 1-2 or 3-4 pounds away from being underweight? This makes absolutely no sense at all, but people think the idea of being underweight/overweight based on BMI is some super scientific measure. It's not, and it means very little.

Just because something is being misused doesn't mean it doesn't have explanatory value in another context. And with any continuous variable that we split into discrete categories, there's going to be some amount of arbitrariness close to the lines. Is "18" really that different from "19"? Of course not. But if someone shows up with a BMI of 17 in your office, there are questions you should be asking them that you would not ask someone at a 'normal' weight, not because there's no overlap between the answers to those questions between underweight and normal-weight people, but because it's a matter of probability.

There's plenty of evidence that doctors treat obese people horribly, so even in a medical context BMI isn't really being used in a productive way. Why can't doctors just look at more specific markers of health? People with high BMIs can be healthy but instead they have their medical issues ignored and they get treated as though any issue they have is their own fault for eating too much.
From the article:

. In a 2016 study of more than 40,000 adults in the United States, researchers compared people’s B.M.I.s with more specific measurements of their health, like their insulin resistance, markers of inflammation and blood pressure, triglyceride, cholesterol and glucose levels. Nearly half of those classified as overweight and about a quarter of those classified as obese were metabolically healthy by these measures. On the other hand, 31 percent of those with a “normal” body mass index were metabolically unhealthy.

That's incredibly informative for an indicator that you can take in under a minute, non-invasively, and that you don't need someone with medical training to administer. I'd use that as evidence for the usefulness of BMI, not against it -- and I'd bet that, combined with age and sex, which we also collect, it's even more informative. We don't always look at specific markers of health because they are more invasive, more expensive, and more time-consuming to collect, but one of the ways we can and should use BMI is when to decide to collect that additional information. But even BMI by itself is useful - for instance, there are medications that are less effective if you're overweight or obese, and you're going to get different recommendations about weight gain during pregnancy.

If we were going to replace BMI with something else that was better, that would be one thing. But that seems unlikely to me. And it's not like if you remove this as a category that doctors are going to stop stereotyping overweight and obese people and blaming their medical problems on their weight. Not measuring BMI isn't going to remove their ability to notice which of their patients are heavier.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:"This explains why muscular athletes often have high B.M.I.s despite having little body fat" is only true for men. There's a literature on this. Women can't put on enough muscle naturally to have high BMIs without also having high body fat.

In general, whether this is an accurate indicator is separate from whether it makes sense to focus on. I'd argue it doesn't make tons of sense to focus on, but the reason I'd give is one that they don't even mention, which is that long-term, significant weight loss is really hard. Even if being overweight is making you unhealthy, if you're not going to change that and it's really stressful to think about, then it's better to focus on things you can change. But that doesn't mean we have to say it's a bad indicator, and that it makes people feel bad is totally separate from whether it has predictive value. (And the bit about "People who have felt discriminated against because of heavier weight are also about 2.5 times more likely to have mood or anxiety disorders" suggests a pretty obvious alternative causal explanation.)

But it IS a bad indicator at the individual level. Did you see the other thread where people are splitting hairs over whether a 5'4" woman who wants to get back to 112 has disordered thinking because 112 is either 1-2 or 3-4 pounds away from being underweight? This makes absolutely no sense at all, but people think the idea of being underweight/overweight based on BMI is some super scientific measure. It's not, and it means very little.

Just because something is being misused doesn't mean it doesn't have explanatory value in another context. And with any continuous variable that we split into discrete categories, there's going to be some amount of arbitrariness close to the lines. Is "18" really that different from "19"? Of course not. But if someone shows up with a BMI of 17 in your office, there are questions you should be asking them that you would not ask someone at a 'normal' weight, not because there's no overlap between the answers to those questions between underweight and normal-weight people, but because it's a matter of probability.

There's plenty of evidence that doctors treat obese people horribly, so even in a medical context BMI isn't really being used in a productive way. Why can't doctors just look at more specific markers of health? People with high BMIs can be healthy but instead they have their medical issues ignored and they get treated as though any issue they have is their own fault for eating too much.
From the article:

. In a 2016 study of more than 40,000 adults in the United States, researchers compared people’s B.M.I.s with more specific measurements of their health, like their insulin resistance, markers of inflammation and blood pressure, triglyceride, cholesterol and glucose levels. Nearly half of those classified as overweight and about a quarter of those classified as obese were metabolically healthy by these measures. On the other hand, 31 percent of those with a “normal” body mass index were metabolically unhealthy.

That's incredibly informative for an indicator that you can take in under a minute, non-invasively, and that you don't need someone with medical training to administer. I'd use that as evidence for the usefulness of BMI, not against it -- and I'd bet that, combined with age and sex, which we also collect, it's even more informative. We don't always look at specific markers of health because they are more invasive, more expensive, and more time-consuming to collect, but one of the ways we can and should use BMI is when to decide to collect that additional information. But even BMI by itself is useful - for instance, there are medications that are less effective if you're overweight or obese, and you're going to get different recommendations about weight gain during pregnancy.

If we were going to replace BMI with something else that was better, that would be one thing. But that seems unlikely to me. And it's not like if you remove this as a category that doctors are going to stop stereotyping overweight and obese people and blaming their medical problems on their weight. Not measuring BMI isn't going to remove their ability to notice which of their patients are heavier.

Yes actually we should tell doctors that BMI is not a great indicator. Just because the medical profession is terrible at treating obesity and obese people doesn't mean they can't be better.