Body mass vs. body fat: How often is BMI misleading?

[bradley]

http://nyti.ms/1Jqysdn

[Chad]

BMI is a rough calculation and needs to be looked at in the context of the overall picture. For instance, when I'm at 225 Lbs I'm basically on the line between overweight and obese, but my body fat is below 20%. Just getting to the top of the BMI normal range would probably put me at an unhealthy fat level. Probably below 5%. So, for a sample of one (me), it's definitely not a good measure.

[7Wannabe5]

http://www.ncbi.nlm.nih.gov/pubmed/21521449

http://www.ncbi.nlm.nih.gov/pubmed/24966306

IOW, it's not the overall percentage of fat either. Abdominal fat is the strong correlation. Fat on the hips in females stores estrogen and may be correlated with additional muscle tissue which is protective of hip fracture. Reduction in overall body fat in females to a level that inhibits the production or storage of estrogen or development of bone tissue may therefore prove a greater risk. The tendency to equate health and fitness with measures most appropriate for a man in late adolescence is sort of ridiculous and, as the studies above prove, simply not based in reality.

At 69 inches tall and 172 lbs. weight (slightly overweight BMI) my measurements are approximately 40/30/42, so although I'm certain my body fat percentage is waaaay higher than Chad's, it is also the case that BMI is not the most accurate indication of my relative health risks (or appearance!-especially if I go to the trouble of throwing on a pair of heels-lol)

[Riggerjack]

I'm fat, no question about it. My fault, and I'll fix it. However, at just over 6', BMI sets 220 as the point where I went from overweight to obese.

When I weighed 220, I was still in the army, running 20 miles each week and working out 5 days a week.

Since then, I have built a lot more muscle in my thighs and shoulders (climbing ladders and working with my arms over my head.) And added a right proper beer belly.

I can lose weight, but getting down to 240 would still be "obese", but a low body fat %.

Look at football players. Corners and wide receivers have a different body type than linebackers, than tackles. Yeah, if a tackle worked out differently, he could have a running back's body type, but there is no way for a corner to get a nose tackle's body type, or vice versa.

BMI is way too simplistic a model to encompass the range of body types.

This has caused me to be turned down by insurance companies for decades. I'm OK with that. If it were important to me, I could get my annual physical results to them to resolve it. BMI is probably a good enough measure for insurance purposes, so long as there is a workable work around.

[enigmaT120]

It's only off 18% of the time? That's not bad for such a simple metric. I'm 5' 11" and weigh 175. If I gain about one more pound I'll pass from normal to overweight, but I certainly don't want to lose any weight. I don't know my body fat percentage but I think it's probably fine. Like the rest of you, I work out.

[7Wannabe5]

I think that really what all the research indicates is that you don't need a tank of water, electrodes, scale or even a measuring tape. All you have to do as an individual is look at yourself and answer the question "Do I have excess fat on my abdomen?" and then figure out what is the most likely cause of that for you. Most likely excess calories and lack of exercise, but it might be alcohol intake, stress, or hormonal issues. If I was a man over 50, I would seriously consider testosterone supplementation as a possible course of action. I am definitely going to look into bio-identical supplementation for myself if/when my waistline starts thickening relative to the rest of me.

[jacob]

It's only off 18% of the time? That's not bad for such a simple metric.

+1000!!

Indeed, for something that's only off in 18% of the cases, a predictive model that only involves TWO WIDELY AVAILABLE variables---weight and height which almost everybody has access to---, and TWO INSTANCES OF LONG DIVISION after conversion to metric which just about every other adult should be capable of, it's a rather impressive piece of modelling.

In addition, it doesn't require deep medical expertise to realize when the measure is off, i.e. those who are mostly muscle and bone (athletes) and those who are mostly fat and bone (runway models).

And from a public health perspective, it's better to "lose" the fraction of the 18% who don't understand the limitation of the model as having no model would mean losing all those who are overweight but remain convinced that they just got some meat on the bones/no more than average/other excuses/etc.

TL;DR - The BMI is an impressive piece of modelling judging by explanatory power / simplicity.

[7Wannabe5]

It is only 18% off in predicting body fat percentage BUT, as is indicated in the studies I posted, body-fat-percentage itself is not the best way to judge heart disease/stroke risk because, for instance, fat on your inner thighs is inert in this regard. Do pear-shaped women live longer on average than moderately fit men? I think the answer is "Yes." IOW, Burpees suck as measure of longevity!!!

[jennypenny]

Yeah, I'm with 7W5. I think measurements/ratios are more indicative of health in women assuming their BMI is in the normal or overweight range, and not obese. Hip/thigh fat is a good thing post-menopause. Abdominal fat is never good, but worse for women than men.

[7Wannabe5]

Actually, it is bad for both, but worse for men. Here's another article that spells it out in plain language. Pear good. Apple bad. Of course, a man who is shaped like a pear likely has some other sort of health issue. Maybe, carrot good. Apple bad?

http://www.dailymail.co.uk/news/article ... apple.html

[Ego]

I once attended a training for my fitness job where the instructor mentioned the usefulness of BMI to give people a rough idea of where they fall on the health/weight spectrum. One of my fellow attendees, a recent hire with a Masters in Exercise Physiology and also the fittest person in the room, blew a gasket. He indignantly pointed out that he, at 5’7”, 210 lbs. and 5% body fat, was obese according to BMI.

What he failed to notice was that those nodding enthusiastically in his support where the staff from the head office, many of whom were obviously obese by any measure. While he is undoubtedly the outlier that proves the rule imperfect, those most in need of change latched firmly to his argument and believed that because it was imperfect that meant it was useless.

It is important to study outliers. They can tell us how they became exceptional and how they maintain it. But those are two different things. When an outlier tells us she can eat lots of saturated fat, ignore cholesterol, scoff at BMI or drink a bottle of wine on cheat day, she is speaking as someone who is maintaining outlier status. People who are trying to become an outlier must do very different things to get to where she is.

[Scott 2]

Muscular obesity still means extra load on the system. 210 at 5' 7" taxes the body more than 150.

The metric is good.

[jacob]

Of course more detailed measurements are better. This is not disputed.

The point of the BMI is that it's a simple measure that a lot of people (even those who otherwise don't own a measuring tape) are able to apply or even probably know.

If I tell Joe Sixpack that he can gauge his "shape" by taking his weight in kg and dividing his height in meters twice and see if the result is larger than 25 ... then it just might work. Or his doctor might just be able to explain it to him.

Conversely, ... if I say the optimal equations is something like k1*mass/h/waist*exp(k2*heart rate-k3*waist/hip ratio)+....+ various functions of 200 hundred other variables ... which is more accurate and is only misleading in 4% of the cases .... the pedagogical value would completely break down because Jox Sixpack wouldn''t be able to understand it and the doc probably wouldn't be able to explain it anyway.

The physical model is that of putting mass onto a skeleton which surface area scales roughly as height squared. Hence the simple fit has physical meaning too. It's mass per surface area of an "unweighted human".

That's really all there is to it.

Because of this simplicity you could use the BMI to gauge the "size" of elephants or mice as well ... or course with different numbers.

In data science, parsimony is valued highly. It is almost always preferable to have a simple model whose behavior is easy to understand so that one may recognize and explain outliers easily. The idea is for this kind of modelling to serve as a tool to help one understand. Not to be a perfect ab initio model that is 100% descriptive the relation between a given phenotype and its health.

[7Wannabe5]

I agree that BMI is fairly simple and reasonably useful, but it is not the case that the math, or much more importantly, the rationale behind the math, is simple enough for Joe 6-Pack to understand. He has to look at a chart to see his BMI, and,therefore, it falls into the realm of "magic." Then the magical mystery is further perpetuated by, really, quite arbitrary lines that assign the labels of "normal", "overweight" and "obese" that often do not necessarily correspond to the previously learned and well-ingrained meanings of these words which have cultural-status and psycho-social associations and, therefore, cause tendency towards denial. OTOH, we could choose to adopt waist-to-height-ratio which has been proven to be an even more accurate simple metric and only requires one division or ratio or comprehension of the mathematical concept of "twice as big" which most second-graders can understand. Beyond a chart, BMI requires a scale which is a much more complicated device than a measuring tape. Of course, you don't even need a measuring tape to determine if your waist-to-height ratio is out of bounds. You just need some string. Step on the string, put one hand on your head and hold it out straight past your forehead, and pull the string up with one hand to meet the other. Then put the two ends of the string together, cut it in half and try to fit it around your waist. I could instruct kids to do this in a kindergarten gym class and they would get it. Since waist-to-height ratio does also generally correspond to what human beings sense to be a healthy, attractive shape (as opposed to the cylindrical form assumed by the BMI), I would think the results would intuitively seem more acceptable and believable to most people.

OTOH, I shall at this juncture, bow to Ego's suggestion that outliers should extend "noblesse oblige" and withdraw my objection to potentially unfair insurance rates, on behalf of Beyonce and myself and all other members of the Amazon Solidarity Society. (bwah-ha-ha.)

[Chad]

....eat lots of saturated fat, ignore cholesterol, ...

Here is an interesting podcast by two researchers on saturated fat and cholesterol. Ron Krauss on Re-thinking Saturated Fat & LDL Cholesterol (it's the newest one on the list at this time):

https://itunes.apple.com/us/podcast/fou ... 98322?mt=2

The research he did has shown that saturated fat isn't the problem. It actually helps create the the larger or fluffy LDL cholesterol instead of the small particle cholesterol, which is not as cell receptor friendly. As a result, the small LDL particles stay in the blood longer and cause problems, while the fluffy/large LDL particles get picked up by cell receptors and used properly. This appears to prevent damage by the LDL cholesterol.

He also theorized the real health issue was combining saturated fat with highly processed carbs (the average western diet). Though, he stated he doesn't have the research yet to show this, as he doesn't have the money to do the research.

Sorry, about hijacking the thread a little.

[tommytebco]

BMI is only misleading If you are mired in self-adulation and denial. Death camp skinny and weight lifter muscle bound will give out of bounds numbers and an alert which may be correct.

If you are apt to "suck it in" and turn to the most favorable angle, smile and nod "YEP! Muscle Bound!!" then you might be misled.

[banker22]

As Scott said, are the heart / organs able to distinguish between carrying around 220lbs of muscle vs. 220lbs of fat? Have there been any studies on this. I understand that hormonally there are massive differences to having muscle vs. fat, but for most aspects of health it seems BMI is a pretty good measure.

[jacob]

@banker22 - Yes it/they are. The heart for physical reasons e.g. increased requirement to pump blood (in the muscle case, athlete's hearts are physically different, larger left chamber, thicker muscle ... if overdone it can be damaged (see old endurance athlete thread)... if not overdone you're more likely to survive light heart attacks). Most other organs for chemical reasons as the existence of [a lot of] fat [think of it as yet another organ] alters blood chemistry (basic food metabolism) if/when it starts overriding normal functioning (think diabetes).

It's easier to grok this when fat is not considered an inert substance that just sits there (beer belly) like some kind of unused fuel tank but rather an active organ that interacts with all the other organs. Once the fat-organ becomes super-large (e.g. weighs 40+ pounds), it starts to interfere with the other organs' functionality.

Here's a paper weight example:
http://www.amazon.com/One-Pound-Fat-Rep ... B00B5YQY6O

There are many studies. BMI is just trying to summarize the leading indicator in a way that most of the public can understand and get right, e.g. height, weight, long division, ...

[banker22]

@jacob that's pretty terrifying...considering I probably have about 50lbs of in on my frame!

What is your recommended weight loss regime? Warrior diet + exercise every other day or so? I understand at the end of the day its calories in vs. calories out...but wanted to hear your specific recommendation.

[jacob]

@banker22 - I get the impression that with few exceptions (medical, etc.) overweight/obese folks have reached such a stage because they don't see the excess weight as a problem or see it as normal. The primary difference between this group and the normal weight group is excuses and/or low priority.

It's similar to how people feel differently about "financial health". To some it's normal to owe a few annual salaries to other people. Others go into near panic mode if they manage to take on just a little bit of debt in the first place. And yet others will only be satisfied once they have FU money. Actions will be taken accordingly.

I see two high BMI groups:
1) Former athletes who used to eat a lot and exercise a lot and then at some point (often shortly after leaving school) stopped exercising but kept eating and then they blew up. These guys know how to eat and how to move. They should have the discipline and the skills based on history. They just need to get back to where they were. This can be a fast process.

Only thing holding this group back is the decision to rearrange priorities. E.g. health is more important than putting in 100 hour weeks to further one's career prospects.

2) People who have been nonathletic all their life. They've never really moved but unlike the skinny-fat group they ate slightly more than they burned for a long long time and just slowly added a couple of pounds of fat every year, year after year. In my experience, these guys have no idea what to do---but they often think they do; but the gap between what they think is good enough and what's actually good enough is often positively huge. They'd need to learn both what's healthy to eat (e.g. that if it comes in a box, it's probably no good even if it says "healthy" on the box). They'd need to learn how to move whether that's training the muscles to do a single push-up or recognizing the difference between joint injury and simple fatigue. Likely they also need to acquire physical discipline to push themselves hard.

What's holding this group back is that there's a lot of things they've never learned as well as the decision to learn them. This is a much steeper hill to climb and it gets harder and harder the older one gets. Even people who've do everything right need to work harder after age 40 when the body starts changing.

The cheapo solution is to buy some workout+diet program and follow it to the letter. The expensive solution is to hire a trainer who will do the same. Unless you're supermotivated, I wouldn't recommend asking friends/family. Nothing is more frustrating and a greater strain on relations than hearing endless excuses. A trainer gets paid for that. The DVD doesn't care.