Welcome Obesity Panacea! NOW EARN YOUR KEEP!

My biggest issue with the BMI is that it doesn't take the sex of the individual into account. Men and women have different ratios of muscle mass to fat mass and that's going to skew the results. I can't honestly say for sure how much it's going to do so and I would be thrilled to read a post about it by an expert.

Vene- Although women usually have a much higher body fat %, it also seems to be healthier for them.
Of course very low body fat % (e.g. 6%) can be healthy for a man and not a woman. In addition if you look at the global numbers women are more likely to be obese, yet they also tend to live longer. Incidentally, women's rate of coronary artery disease are much lower than men's if they are <60 years of age, and does not approach men's until ~80... to me, this smacks of sex steroids modulating such disease risk, which makes a lot of sense to me when I think of what I know about the molecular signaling of e.g. estrogen.
It may be the case that in obesity research, like many other fields of medicine, many of the early studies focused on men as the 'normal' patient. It seems like the links between "obesity" per se and e.g. coronary artery disease are weaker in women (although, the link between inflammatory status or lack-of-physical activity are strong for women).

well, for one, BMI confuses height/weight ratio with health, or more correctly, the users of BMI do.

The other issue is that it's not designed to be an individual measurement, but rather for comparing populations. Using the BMI of say, NYC compared to SF, sure. Using BMI for yourself? not so much.

Comment above got slightly borked. It should have read "...Incidentally, women's rates of coronary artery disease are much lower than mens if you compare people under 60 years of age, but by ~80 years, they start to equal out. This strongly hints at an effect of sex hormones on that disease risk, with it taking a number of years past menopause for women's risk to look like men's. Which makes sense, given what I know about estrogen signaling and adipose tissue. It may be the case in obesity research..."

I think how well it applies to a person varies. According to a page at the NIH, I'm nearly at the lower end of 'obese' at 194 lbs. That seems fairly reasonable, but I'd have to be at 164 in order to be at the top edge of normal, which seems off, because I'm pretty skinny at that weight. (I was there around late 2000. Then I moved back in with my folks. Thus the 30 pounds.) I can't see anyway short of cholera that I'm going to get below 160, even though BMI would have me in the 'normal' range all the way down into the 140's.

So I think for my body shape, the BMI categories skew a little low. But it's still a useful tool, as it's in the ballpark of being correct. It's clearly true that I don't want to have my current BMI or higher. And it would clearly be better if I were at 164, even if I have no prayer of getting much below that.

Of course, I don't really need another number to tell me this, either. My weight and pants size are sufficient.

BMI interpretation is a statistic. It is no more accurate to apply the category to an individual as applying a class average to an individual's test score.

Cholesterol interpretation is also a statistic. I know of people who are fit, exercise routinely, eat right, etc. but have "bad cholesterol"; and I know people who are none of those with "good cholesterol".

Why don't we wait to hear back from the experts before we all tip our hats as to how ignorant we really are?

One thing I think isn't stressed enough for folks monitoring their BMI is that for an individual changes in BMI exactly correspond to changes in weight. At my height, 5'7", there's no difference in telling me that my target BMI is 22 or that my target weight is 140 lbs.

As it happens, I'm obese, I over double that target. But a few years ago I had the opportunity to use one of those electronic scales which purport to measure body composition as well as weight. Using the it's results, I computed that if I were to lower my body fat to the recommended level and increase my hydration to the recommended level I'd still need to lose lean mass (muscle or bone) to break 200 pounds. There's no way I'll ever see myself anywhere near a "healthy" BMI without long-term hospitalization for major illness.

Emerson-- Why don't we wait to hear back from the experts before we all tip our hats as to how ignorant we really are?
*evil squint*

That is not how we do things on the internet, sir/madame.

*evil squint*

hehehe!

BMI is clearly inappropriate for anything but population statistics, because the only two variables are height and weight.

No allowance is made for musculature, skeletal mass, or even sex.

Using it as a diagnostic criterion of anything is completely inappropriate.

Nor is body fat percentage especially useful, if you're talking about health, rather than aesthetics. Someone who eats a ton of quality food but still gets lot of exercise will be at a stable weight (calories in = calories out), have a lot of excess fat, but be quite healthy.

It's not being overweight or obese that makes people unhealthy. What makes such people unhealthy are the poor dietary and exercise habits that lead them to be overweight in the first place. Even many of the professionals in the health field seem to forget the difference between correlation and causation. Perhaps mostly because causation is too complex to figure out much of the time when it comes to these issues.

"have a lot of excess fat, but be quite healthy."

Well, there's still they physics of carrying the extra weight, which puts more stress on other body systems, and can cause things like obstructive sleep apnea. The person's cardiac health might be fine, but the weight can still take a toll in other ways.

Abbie, just wait until you get out of grad school and can afford real food - we'll see what you BMI does then ;-)

The real reason BMI is being abandoned as a medical measure is it does not do a very good job of predicting ones cardiovascular/diabetic risk (which is supposedly what it does). Unlike more specific measures (like % total body fat, or not. It all comes down to the fact we're all built differently.

A far better measure (in terms of guaging risk) is your hip-to-waist ratio, which is essentially a simple measure of visceral fat.

I am 6'6" and weigh 250 pounds. I am overweight and BMI says I am overweight. To be obese I would have to gain almost 20 pounds. Of course to be in normal range I would have to lose around 30 pounds. I need to exercise more and I know it but I am okay right now.

NPR had an article (Top 10 Reasons Why the BMI is Bogus) that breaks it down pretty well. It's a misapplied population statistic.

Certainly, height/weight charts can get you *somewhere* as a broad, I-know-nothing-else-about-you health barometer. But how useful is it? Not very.

Simply put, it's mathematically wrong. If we were to run you through the Magical Enlarger and make you 20% larger all around (same proportions everywhere) your BMI would also go up 20%.

My favorite example is Shaquille O'Neal: at 7' 1" and 325 pounds, he has a BMI of 31.6. Obese, bordering on seriously so. For comparison, a 5' 2" tall individual would have to weigh 173 pounds to hit the same BMI -- and would have a waist size about the same as hir height. Which one is more appropriate to BMI? Which one should be working harder on losing weight?

The curious thing is that BMI is, on balance, a good predictor of health. The problem is that a third of it is due to the fact that other things equal, taller people aren't as healthy as shorter ones -- and height is not really something we can change so counseling O'Neal to get his weight down to 245 (about a BMI of 24) to be in good shape is ... well, can you imagine Shaq after losing eighty pounds?

So, based on BMI, Shaq may well have the overall life expectancy of someone two feet shorter who weighs 173. The latter can profitably lose weight. Shaq, not so much.

My favorite example [because of its simplicity and clarity] of misapplying statistics is this: 50% of the country is female, therefore everyone is 50% female.

"So, based on BMI, Shaq may well have the overall life expectancy of someone two feet shorter who weighs 173. The latter can profitably lose weight. Shaq, not so much."

You're making the error of treating an athlete as an average person. Whatever else its faults, the BMI simply goes wonky for athletes in general.

But the same is true for suggested calorie intake levels. An olympic swimmer takes in vast amounts of food, but that isn't an argument against an estimate for how much the average semi-sedentary person should eat.

Athletes, especially 7'1" ones, are outliers, and are really ought to be ignored when you're trying to come up with a metric that works for the majority of people.

If you try to come up with a simple rule of thumb for fitness that works for a sedentary middle aged woman or man, and also for the outliers like Manute Bol, Shaq, or a roided-out wrestler, you're going to fail or drive yourself nuts.

Thanks for letting us ease into things, ERV ;)

It's interesting that the "muscle weighs more than fat" issue seems to come up so frequently, as there are several other fundamental issues with BMI that are at least as important, if not more so.

We'll take a crack at this ASAP, and I look forward to reading more comments here in the meantime.

Travis

Athletes, especially 7'1" ones, are outliers, and are really ought to be ignored when you're trying to come up with a metric that works for the majority of people.

The math is just as bad for people closer to the middle, but it's not as obvious. People don't do 10% numbers well; they're not clearly wrong. However, if you compare someone who's 5' 2" tall and someone 6' 2" tall with the same BMI, it's pretty damned obvious that the shorter one is plumper than the taller one -- to the tune of about 20%.

That's just extremely basic mathematics, and that's why from the very beginning even the creator of the BMI said it wasn't applicable to individuals.

Travis-- Thanks for letting us ease into things, ERV ;)
BWAHAHAHAHA!

;)

Try some boundary problems with the BMI calculation. At 6', how light would I have to be to qualify as "underweight," and how heavy to qualify as "obese"?

At 6', I would have to be 135 lbs to be considered underweight. This seems ridiculous to me. I'm about 170 lbs currently and routinely described as "skinny" -- my mother and grandmother won't shut up about it, in fact. I know people about my height who look absolutely emaciated at 160 lbs.

At 6', a friend of mine who weighs 230 lbs is obese. Never mind the fact that he's actually just a broad-shouldered dude with a relatively small beer gut.

To me, 6' 230 lbs seems maybe a LITTLE overweight, but not so far outside reasonable variation as to qualify as "obese." On the other hand, 160 lbs at 6' seems like it's unhealthy enough to qualify as "underweight" -- even though such a person would have to lose 25 lbs before their BMI indicated as such.

Maybe it works fine for average women -- 6', 230 lbs is much farther from the average woman than it is from the average man, and 6' 135 lbs is much closer to the average woman than to the average man. But for lower-middle to lower-high weights for a 6' man, BMI seems like a little bit of bullshit.

A 6' 135 lb lady would be a very skinny lady indeed. Scarily so.

I have never been able to reconcile my size with my BMI. When I was in my "fighting trim" at 5'3", my lean body mass was 135. Add 20% body fat onto that, and I'm in the "overweight" range.

I just ignore it and go by how my clothes fit.

Not to be too much of a jackass, but as bad as BMI can be, "I'm skinner than my fatass friends" does not a good metric make. Yet, according to the psychological studies on the issue, it *is* a very commonly used one. Having fatter friends leads you to think fatter is normal.

There also appears to be a weird "maybe BMI works for somebody else" attitude. Which, if Shaq is posting here, is legit.

A large part of it boils down to the communication strategies available for physicians. And to be perfectly blunt in the current medical school curriculum we are trained to practically whore ourselves out so our patients will loose weight. I am okay with that. Obesity is a serious concern. A very very serious concern. How much weight is clinically relevant? Every pound. It has been pretty well documented that there is a strong clinical difference between gaining a pound in fat a year vs loosing a pound of fat a year. To answer the medically related questions about physiology. Among other things a positive energy balance leads to an increase in pro-growth hormones which directly influence and speed the growth of cancers. Additionally individual body cells compete less for glucose and down-regulate insulin receptors leading to insulin resistance the precursor of diabetes. Insulin resistance and frank diabetes lead to elevated blood glucose. Glucose is actually quite reactive and tends to randomly find and attach to various proteins leading to vasculature destruction particularly in the kidneys though also harming the vulnerable vasculature of the heart. Elevated glucose also tends to feed infections as well as growing tumors. Furthermore, simply needing to pump the extra blood through the fine vasculature of adipose tissue perpetually strains the heart, but sadly not in a good way. As mentioned cholesterol is technically independent of weight. However, often the diets that lead to a high weight are rich in cholesterol which builds up and also leads to more vasculature damage and inflammation independent of obesity. These effects are all well documented and sadly work together to cause vascular disease and death.

To try to answer some other questions it is true that fat stored in breasts has a different clinical significance than visceral fat, and as it stands we seem to be uncertain as to why that is. I will let the PhDs try to answer that one, and obviously BMI being just about weight and height ignores that fact. If you are personally interested in your own health there are other much more intricate and involved health calculators available online that take into account many items including gender, age, weight, height, smoking, fasting glucose, blood pressure, blood lipid profiles, past medical history and family history. While yes BMI is pretty sloppy these combined statistics tend to work out pretty well and have been verified. The catch is that most people don't like hearing about the risk factors they can't change such as sex, age, and family history while interestingly the other factors all tend to improve along with weight loss.

As a budding physician I am very much troubled by one message that is often lost in the dialogue about weight. It seems that people think of themselves as skinny/healthy or fat/unhealthy when in reality it is almost a perfect spectrum. If you are anywhere in the middle there are very specific benefits of only loosing 5-10 pounds and then keeping that weight off. So it all comes down to talking to people and making that happen. Yes, the problems with BMI are well documented, and to be frank basic waist circumference is even better than BMI. Except in a clinical practice most patients who are already sensitive from society that considers obesity as a moral failing don't want their physicians pulling out a tape measurer. However, bringing up BMI is a much more empathetic way to start that discussion, and with any luck can make those 5-10 pounds happen.

@D.C.Sessions: A 5'2" person of 170ish lb would have the same waist circumference as hir height? Are you for real? I'm a bit heavier at that height and my waist is 32". You probably wouldn't even think "there goes an obese person" if you saw me on the street. "There goes a chick with truly unfortunate thighs" maybe.

I think this does illustrate one of the problems when talking about obesity: people look at these rather extreme statistical outliers -- like the person who really is 62" tall and 62" around (and probably, what, 250+lb?) and think *that* is who has the problem. But it's really more about the average person who is heavier than they would have been 20 years ago. It's more of a population-wide problem than a that-fat-person-over-there problem.

As far as the BMI relates to average folks, I think it's misleading and not in the "OMG I'm not obese the BMI lies" direction. A woman of my height with a fine-boned build could easily have the same body fat percentage as I do yet weigh within the "normal" range according to the BMI. Is that woman getting an accurate assessment from a doctor who depends on BMI? I'm going to guess not. I know there are special guidelines for some populations (e.g. Asians) attempting to address this sort of thing, but it really seems to make more sense to just use a better tool.

I think the problem here is that it's not well known that BMI is for the *average.* People just need to understand that and not take it so seriously. BMI is, at best, a guide that you have to adjust for you body type.

Unless you're an outlier like Shaq or "The Rock," it can still be a semi-useful "quick" reference. Just meaning that if you're a pretty average person all around, you might want to take heed if it tells you that you're obese or morbidly obese.

Still, chances are even better that you know it already or not if you're fit (though I guess Dunning-Kruger might come into effect here). As a measure of how fit/unfit... you'll want to find something else.

I was the same height in High School as I am now (6'6") but in HS I weighed 195 pounds. That would be within the range of normal according to BMI. If you looked at me then, I was skinny, very skinny. I had to special order jeans that fit me (31, 36). As I said earlier, today I weigh 250. I feel that is overweight. I would rather be around 225 to 230, which is still considered overweight by BMI, but like other said there is limitations to BMI. I still need to exercise more.

No one I know considers me to be overweight. I will happily drop and give you 50 anytime you want. My BMI is 27.8, smack in the middle of "Overweight" My wife on the other hand is visually overweight as even she will tell you. Her BMI is just inside 25, making her "Normal." The difference is height: Im a foot taller than her.

As any mathemetician will tell you, if you scale up a 3 dimensional object the volume and therefore mass (or weight), assuming constant density will go up by the cube of the increase in one dimension (height). But the BMI uses the square of height. Now the human body isnt a perfect sphere, so a cube-law is probably not perfectly accurate either. I did some reading on the history of the BMI a while back and it seems the population data used to come up with it actually indicated the appropriate exponent should be about 2.5. However, the exponent of 2 was used for ease of calculation, which results in shorter than average people generally having lower BMI's than taller than average people.

I, for one, am not going to let a couple of Canadian pretty boys tell me how to throw the medicine ball around or what size Indian clubs I need.

Hell, the closest those chaps come to a fine waxed dollar coil mustache is, if my magnifying glass is to be trusted, a "soul patch".

Great Caesar's ghost! I'm off to the natatorium for wand drills and my Jahn turnplatz aquatics....harumph.

Hi ERV,

Thanks for the kind welcome and the thoughtful blogwarming present. I tried my best to collect my thoughts on this issue, and they now appear on our site - I had been intending to discuss this for months. Thanks for the friendly nudge;)

"Canadian pretty boy"

Peter

Notice your presupposition that body fat percentage is the ideal, the most important number. Is it?

@Dave- "Women tend to imagine their ideal weight is unrealistically low, so they diet unnecessarily. Men tend to allow their ideal weight to be higher than medically recommended. Men and Women should learn from each other." (see the link)
The BMI formula takes into account height, and (for reasons I've described) is if anything *medically* MORE applicable to men.
In other words, you and your wife can think whatever you like about aesthetic aspects, and of course you may be an athlete like Shaq who BMI was not meant for, but without any further info than what you've supplied I'd much have your wife factored into my pool for insurance premiums than you.
If it'll make you feel better, you can think of dogs. An Irish Wolfhound will tend to live just ~7 years, a toy poodle ~15. The Wolfhound doesn't have to be a fatass to have more health troubles.

I consider myself an Average Jane. To be considered BMI 'overweight', I would have to gain 35 pounds. To be that weight but keep my current body fat percentage, only 5 of those 35 lbs could be fat.

There is no way I could do that without steroids. No friggen way.

Is Ms. Smith insinuating that Bev Francis, who tipped the scales at 180 pounds and could bench press more then twice her body weight, was on steroids?

Re Becca @ #2

A 6% body fat ratio would be very dangerous for a woman. It is my understanding that a woman stops menstruating if her body fat ratio falls below 10%.

@SLC- exactly. I was using 6% body fat as an example of a number that could be healthy for a male (athlete) but not for a female.

Becca - Although you are addressing me, Im not sure I understand what your point is relative to mine. Yes, issues with women's body image can result in inaccurate self-assessments, but frankly, after enough years of marriage, you get pretty darn familiar with the other's psyche and I can assure you that is not affecting this right now. But you dont know my wife and have no reason to trust my assessment, so lets leave her out of this. That will also eliminate any male/female issues with BMI.

As mentioned before, my BMI is 27.8. Neither I nor any of my acquaintances would consider me to be overweight. Nevertheless, I am not particularly athletic, Im a middle-aged swivel-chair jockey. I keep reasonably active, but I havent seen the inside of a gym in over a decade. So I have no reason not to believe I am an "Average Joe." To reach "Normal" by the BMI I would have to lose over 20 lbs, to reach the middle of "Normal" I would have to lose 40 lbs. When I was that weight 20 years ago, I was considered rail-thin, not normal.

There are numerous reasons for having a standard, mathematical definition of obesity. There are also valid reasons for using an integral power in that definition, not the least of which is ease of calculation. That said, there several studies that suggest that 2 is not the optimal exponent to be used, if non-integral values are allowable. (It is generally the only integral value within the error bars.) My point is that while generally a valid calculation, by reducing the exponent to 2, BMI over-represents obesity in those of above average stature and under-represents it in those of below average stature. If you have a BMI of 35, unless you are a professional athelete, you are probably obese, I dont care what your height is. But if you are 6-6 and your BMI is 31, you would probably be better classified as "Overweight" than "Obese" Simillarly if its 26.

Dave- You appear to be chiefly define obesity as "when people tell you you're fat" as opposed to, say, "increased risk for cardiac disease". Like it or not, between protective estrogens and the large-organism = more strain effect (as by analogy to the dog breeds), on average a woman at a BMI of ~25 is significantly more healthy than a man at a BMI of 27.8.
Which says nothing about individuals, of course- you might be much healthier than your wife, I'm just telling you why your input does not convince me that you are healthier or that BMI is extra wrong for tall folks.

Nonetheless, what really triggered my comment was that IF you were to say: "EVERYBODY says I look great at BMI 27.8 and everybody, including my wife, says she is a lardass at BMI 25"
...this indicates to me that you are:
1) a douche
and
2) in denial about your actual health/life expectancy.
So, although you did *not* say that, and may in fact not be a douche or in denial, it is definitely true that:
*You may be able to do pushups, but you're still gonna die
*I'm glad I'm not your wife

Abbie,

BMI is actually a decent measurement to look at population health stats/comparisons etc.

I'm sure most people who land in a normal or healthy BMI range think the measurement is just fine for them.

My guess is that people who fall outside of what is deemed a normal or healthy BMI take issue with it because they've been put in a category they don't like or don't agree with...and it might just force them to do some introspection and wonder if they actually have some work to do in the gym or on their diet. It's much easier emotionally to dismiss the credibility and merits of the BMI (or whatever other measurement) than it is to look at yourself and decide there might be some work that needs to be done.

John

In one paragraph, NBC Sports sums up a disgraceful state of affairs: "And for those who persist in thinking Shaquille OâNeal is obese, he has a simple message: 'You think that, stick to science. Top 50, three rings, lot of money, two mansions.'"

Of course, NBC Sports should have followed up this obscenity by mentioning that people who bounce a ball and toss it through a hoop on average draw no salary for their efforts, which is as it should be.

http://nbcsports.msnbc.com/id/7129586/page/2/

My wife got referred to a consultant because the midwife we're seeing was told to refer everybody with a BMI above 30 and she came in at 30.5 . I think it was the shortest consultation ever, because anybody with eyes and a brain can see that she's fine.

My BMI is off da hook, nawwhatimsayin'?

I got all them conic sections filled up in dis here bidness.

My insurance premiums are EPIC.

Wow, John Barban, you're right: no one has every had any cultural cues urging "introspection" about body size until the minute someone calculates a BMI score. *eyeroll* Have you SEEN the mainstream culture?

I don't pay much attention to measurments like this and like for and such as, but here's a personal anecdote that I've been wanting to relay-

In junior high, I lifted weights and my goal was to get up to 155. I'm 5'8. When I finally got to 155, I was really happy. Then in highschool I went veg, quit lifting weights so much and dropped down to 145. Then I got sick after highschool and dropped down to 140. Was at 140 for many years. Now I'm up to 160 and basically haven't exercised(still sick) in at least half a decade.

The point being, right now I'm probably a little 'overweight', but if I was to get better and start working out I would imagine that I would lose about 10-20lbs. of fat and then gain most of it right back as muscle if I were to keep lifting. So even a difference of +/-20lbs on a relatively small frame on the scale doesn't really tell jack shit about overall health. 155lbs=really fit vs. 145lbs=still pretty fit vs. 160lbs=quite out of shape.

Plus for girls you've got the whole boob issue. Even moderately big boobs seem like they would throw everything off balance in terms of scales. +1 for the male-centric body charts! :p

I should start jogging (again-ish) now that the sun is returning and I'm in the habit(ish) of getting up early.

But what to wear? I look in bafflement at gear in the supermarket, and I'm afraid to ask.

My sophmore year of college, I was on the crew team. I was 5'4" and 160 lbs. I trained 11 times a week, some aerobic training and some weights; I was in the best shape I've ever been in my life.

A year later, I was still 5'4" and 160 lbs. I was off the team, and had taken a summer job that required me to be sedentary 5 days a week. I was in some of the worst shape I've ever been in my life.

Same BMI in both cases.

Another problem with clinicians using BMI is that it is a better predictor of problems for white populations, which it was based on.

http://www.nature.com/oby/journal/v16/n4/abs/oby200811a.html
http://www.ncbi.nlm.nih.gov/pubmed/17639654
http://www.ncbi.nlm.nih.gov/pubmed/19108808