The BMI: Wrong Again

[Ironhold]

link

Yet another study has come out noting that the BMI system is badly flawed.

In this case, the study has noted that there's no direct correlation between one's BMI score and other, more consistent health measurements such as blood pressure and cholesterol.

Here's hoping that this will finally help put the BMI to rest.

 

[JerushaC]

It's been known for a long while that the BMI calculation fails to take muscle mass and family line genetics into considerations. Now it's a scale used by diet scams and fitness scams. I have never seen a licensed doctor use it.. ever.

 

[Aryeh Jay]

WOO HOO!! I am no longer morbidly obese at 4 foot 9 and 360 pounds, I am fit and trim. Take that junk science!! Now, back to my Paczkis…


I never did like BMI.

 

[jayem]

Two comments:

1) From the article:

“There are healthy people who could be penalized based on a faulty health measure, while the unhealthy people of normal weight will fly under the radar and won’t get charged more for their health insurance,” said Tomiyama.

Currently, this doesn't happen. Premiums for those with job-based group health coverage (which is about 60% of all insured people) are not based on individual medical risk. And ACA forbids basing premiums on health status for those buying policies in the individual market. Neither are premiums for Medicare Part B, or Medigap policies adjusted for personal health risk.

2) While the predictive value of BMI in isolation may be limited, there is no doubt that a high BMI in someone who already has an obesity-related illness is a major concern. An obese patient with diabetes, hypertension, arthritis, coronary artery disease, sleep apnea, and other conditions will definitely improve with weight loss. I've seen people with diabetes and high blood pressure cured--needing no medications at all--just by going on a good diet, exercising, and losing weight.

 

[Paulos23]

BMI is quick and easy, which is why it is used. But it was first used to measure the general health of a population, not an individual. It would be more accurate to do a dunk tank weighting to figure out fat percentage than BMI.

 

[Tallguy88]

It also scales weight to height incorrectly. So tall people will have a worse BMI score than short people with the same weight to height ratio.

I have had doctors tell me I am morbidly obese based on my BMI, even though I'm 6'3" and of decent proportions. I am overweight and need to work on it, but blood pressure, cholestorol, sugar, etc always are within good parameters.

 

[ThatRobGuy]

BMI has a couple of critical flaws. The obvious one being that it doesn't take into account muscle mass (as others have mentioned) and levels of physical activity. Now, it's true that not everyone who's 260lbs looks like The Rock...I get it, however, it's also true that people who are in the "overweight category", but who exercise regularly are less likely to get heart disease than a skinny person who lives sedentary lifestyle.

However, the 2nd flaw, and the one I feel to be more troublesome, is how BMI categories are used to assess health risks among people who fall in the various BMI bucket classifications set up by the CDC. The classifications are far too broad and need to be waaaayyyy more granular in order to have any sort of applicable meaning.

I'll explain what I mean, here are the classifications:

• A BMI below 18.5 (shown in white) is considered underweight.
• A BMI of 18.5 to 24.9 (green) is considered healthy.
• A BMI of 25 to 29.9 (yellow) is considered overweight.
• A BMI of 30 or higher (red) is considered obese.

And these are used to attempt to determine likelihoods of particular ailments based on your weight. For example, they report that people in the "Obese category" are 3 times more likely to suffer from heart disease than a person in the "healthy category"

Here's the issue with that, if we're talking about a person who's 6 foot tall, the obese category spans from 220lbs...to infinity lol.

So according to how these are reported, they're saying that a guy who's 220lbs is 3 times more likely to have heart disease than a guy who weighs 185lbs. Which is definitely not the case.

Now the guy who weighs 350lbs?...yes, that's a very real possibility, however, it's a fundamental flaw to operate on the premise that 220lbs and 350lbs are presenting the same risk levels. However, the CDC operates on that flawed premise.

In order to give BMI any meaning or statistical value, it needs to be split out into about 15 buckets (instead of 5) ranging from BMI (<= 18) through (>=50)

 

[iluvatar5150]

I'll explain what I mean, here are the classifications:

• A BMI below 18.5 (shown in white) is considered underweight.
• A BMI of 18.5 to 24.9 (green) is considered healthy.
• A BMI of 25 to 29.9 (yellow) is considered overweight.
• A BMI of 30 or higher (red) is considered obese.

And these are used to attempt to determine likelihoods of particular ailments based on your weight. For example, they report that people in the "Obese category" are 3 times more likely to suffer from heart disease than a person in the "healthy category"

Here's the issue with that, if we're talking about a person who's 6 foot tall, the obese category spans from 220lbs...to infinity lol.

So according to how these are reported, they're saying that a guy who's 220lbs is 3 times more likely to have heart disease than a guy who weighs 185lbs. Which is definitely not the case.

Well, to be fair, your example of 6'/185 lb is just over the line of "overweight" and 6'/220 lb is just under the limit for "obese," so what you're comparing isn't really "healthy" to "obese;" it's "overweight" to "overweight." For someone 6' tall, the middle of the "healthy" range would be about 160-165 lb. I wouldn't be surprised if there were a 3-fold difference in risk between 6'/160 and 6'/220.

Now the guy who weighs 350lbs?...yes, that's a very real possibility, however, it's a fundamental flaw to operate on the premise that 220lbs and 350lbs are presenting the same risk levels.

How do you know this? Have you seen the risk data? It could very well be that above a certain level of obesity, the risks plateau or at least flatten out to a degree. For example, is 450 really that much worse than 400 or 350? I don't know the answer to that, but it seems plausible and would contradict your expectation that the risk increases in the way you seem to think it does.

 

[ThatRobGuy]

How do you know this? Have you seen the risk data? It could very well be that above a certain level of obesity, the risks plateau or at least flatten out to a degree. For example, is 450 really that much worse than 400 or 350? I don't know the answer to that, but it seems plausible and would contradict your expectation that the risk increases in the way you seem to think it does.

Actually, here's the study I was making reference to:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2223029/

The bands for overweight and obesity in the US, for example, are the product of the 1997 National Institutes for Health task force report on the prevention and treatment of obesity that supposedly links these bands to increased risk of death. However, the study on which the report is based does not support these linkages.5 It found that the death risks for men with a body mass index of 19-21 were the same as those for men who were overweight and obese (29-31).

So in evaluating the link between obesity and mortality, it's pretty clear that the magic number we're looking for is something that's higher than 31.

Here's a visual view of the data

It would appear that the sharpest upturn would happen at around BMI 35, which if we're going with our 6' example, is around the 255-260 range. ...and then grows exponentially as we move up to BMI 40.

 

[iluvatar5150]

Actually, here's the study I was making reference to:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2223029/

The bands for overweight and obesity in the US, for example, are the product of the 1997 National Institutes for Health task force report on the prevention and treatment of obesity that supposedly links these bands to increased risk of death. However, the study on which the report is based does not support these linkages.5 It found that the death risks for men with a body mass index of 19-21 were the same as those for men who were overweight and obese (29-31).

So in evaluating the link between obesity and mortality, it's pretty clear that the magic number we're looking for is something that's higher than 31.

Here's a visual view of the data

It would appear that the sharpest upturn would happen at around BMI 35, which if we're going with our 6' example, is around the 255-260 range. ...and then grows exponentially as we move up to BMI 40.

Interesting.

 

[keith99]

It also scales weight to height incorrectly. So tall people will have a worse BMI score than short people with the same weight to height ratio.

I have had doctors tell me I am morbidly obese based on my BMI, even though I'm 6'3" and of decent proportions. I am overweight and need to work on it, but blood pressure, cholestorol, sugar, etc always are within good parameters.

Bingo. My understanding is it was originally a screening tool for hypertension. In that context the miscalling is offset because just being bigger does mean additional work for the heart.

I once ran the numbers for Wilt Chamberlain in his early years when he got the nickname Wilt the Stilt because he was thin as a rail. He came out as obese according to BMI!

 

[Ada Lovelace]

I think people have also been incorrectly labeled as unhealthy because according to the BMI they are underweight. The number gives a glimpse about our health but definitely not a complete portrayal of it.

 

[trunks2k]

Meh, I've never trusted the BMI. I'm a tall guy, and it seems that its accuracy is way off for tall people. I've calculated my BMI when you could look at me and think there's no way I'm close to overweight and my BMI had me as being overweight, and to be in the healthy range, I'd have to go back to what I was as a teen and looked like I was being starved.

 

[BrianJK]

In the military, we still use BMI as part of the determination of whether a soldier is fit, but there is an additional test taking body measurements if a soldier is over the allowable weight for his/her height. So for us, it's a quick and easy indicator, but it isn't the final word. I kind of like that view. Old tools which are imperfect can still be used if that grain of salt is there.

 

[Sistrin]

My BMI broke down and I had to have it towed to the shop. The estimate is $350.oo, not including tires.

 

[ThatRobGuy]

I think people have also been incorrectly labeled as unhealthy because according to the BMI they are underweight. The number gives a glimpse about our health but definitely not a complete portrayal of it.

I think the BMI mortality chart reflects that way for very low BMIs because in many cases, if a person is very very underweight, it often times can mean there is some sort of disease (like cancer for example) that could be the cause. If an adult is under 90 pounds, there's a good chance that there's a serious health issue there.

 

[iluvatar5150]

Meh, I've never trusted the BMI. I'm a tall guy, and it seems that its accuracy is way off for tall people. I've calculated my BMI when you could look at me and think there's no way I'm close to overweight and my BMI had me as being overweight, and to be in the healthy range, I'd have to go back to what I was as a teen and looked like I was being starved.

The formula for BMI is:
BMI = (weight in kg) / ((height in m)^2)

This formula assumes a constant ratio between height, width, and depth for a healthy person. But as any tall person who's every shopped for clothes can tell you, proportions often don't hold constant through all height ranges. If the real-life ratios change, then the results are going to be skewed.

 

[Ada Lovelace]

I think the BMI mortality chart reflects that way for very low BMIs because in many cases, if a person is very very underweight, it often times can mean there is some sort of disease (like cancer for example) that could be the cause. If an adult is under 90 pounds, there's a good chance that there's a serious health issue there.

Underlying health issues can be significant contributing factors in people who are underweight as well as those who are overweight. A friend who struggles with being overweight has PCOS and hypothyroidism. I have endocrine issues that have caused me to be underweight, but consider myself to still be healthy. To echo what has been written about the necessity of evaluating the whole person to determine the quality of health rather than whether their weight falls within the BMI parameter of healthy, there are are factors that need to be considered when someone is underweight, too. My BMI is currently 15.9, which would lead some to construe me as being sickly or feeble because I'm in the underweight category, but I'm a very fit dancer and athlete. I have no nutritional deficiencies according to recent blood work. I'm running the LA Marathon this weekend and am feeling optimistic I can BQ.