# Why does the BMI account for height at the square rather than the cube?

The body mass index accounts for an individual’s weight by increasing the height divided at the square, rather than the cube.

Do tall and short people not roughly have the same relative dimensions as normal-sized people?

The surfare area of a shape increases at the square of the dimensions but the volume increases at the cube. Presuming that density is roughly maintained, that means that weight should increase at the cube of dimensions as well.
A 6’ man who weighs 184 pounds would be considered overweight, which just sounds wrong.

I’m 6’4" and 205. I have a significant gut and could easily lose 20 lbs. In fact I’m working on that and have a goal weight of about 185.

The reason is that while some proportions do increase with height, a lot doesn’t. At my height I am not significantly wider or deeper (I guess that is the best way to express that) than a man in similar shape who is a foot shorter than I am. My waist is similar to that of shorter men. As is my chest and shoulder width. My shoe size is pretty standard for men (11). Although I am 10% taller than the average. I’m not 10% larger in any other dimension or body part.

BMI is just a ballpark figure. It doesn’t really take into account different body shapes and especially those who have a decent amount of muscle. But it’s quick and easy to calculate. Percentage of body fat is a better number to use but it’s harder to calculate.

Bartman, aren’t you just saying that you have a slim build? Loach, aren’t you implying that it isn’t important to calculate it right?

This has bothered me for years. Little Buddha statues would have low BMIs and basketball pros fairly high ones, the way BMIs are calculated now. I take the op to imply it should be the cube and I agree. The only thing the current method should do well is predict floor loadings for crowded rooms.

I thought BMI is an empirical method (find a formula that works and use it), not one based on weight-volume ratio. So you have 2 easy measurements: weight and height. Find a simple formula that gives a range of numbers if an average person is fat, and another range if they’re thin.

I am a 6’ man and I can say for sure that anything at 185 or so pounds is definitely overweight. I’m at 175 or so right now and trying to cut down on some of my flab.

There have been some recent studies, however, that show that people in the “overweight” category (not obese though) are actually just as healthy if not healthier than “average” BMIers.

So, medical research tends to indicate that if you’re average, or on the lighter side of overweight, you’re not at higher risk for certain diseases. But if your BMI is on the higher side of overweight, and into obese, then you are at a higher risk for certain diseases.

That’s really all there is to it.

That being said, I’ve heard (though I don’t have a cite) that the whole “body types” thing is more or less a myth. There aren’t a lot of different body types that will skew the results of a typical BMI calculation for the vast, vast majority of people. People with a LOT of muscle (you really do have to have a lot) will have a high BMI, but not necessarily be at risk for certain diseases. From what I’ve heard, that’s about it.

I don’t think that it’s calculated at all. It’s measured. Right? That’s why it’s not used.

Harder to measure. As far as I know the most effective way of measuring it is an immersion test.

This. BMI is not about what “should” or “shouldn’t” be overweight. It’s about delineating a point where excess weight starts to correlate with health problems. They didn’t set the “overweight” line where it is out of some sort of moralizing. That really is the point where weight starts to be associated with health risks, and it really isn’t that high. Human are designed to be lean, active creatures.

IMHO, we have a very skewed idea of what a normal weight is. A normal weight is actually quite actively skinny, it’s not just “not full of rolls and obviously grossly obese.” People who have weights correlated with the best health outcomes are have almost visible ribs, no little belly stashes, and trim hips. Your body doesn’t care if you look nice in clothes or you are slimmer than your outright chubby coworkers. That ten pounds that you could probably stand to lose isn’t a free bonus just because it doesn’t make you look obese. If you have any real amount of extra fat beyond some very slight padding, you would probably be healthier without it. We just don’t have a lot of wiggle room for putting on weight before it starts to affect us.

BMI seems awfully meaningless, IMO. Two people could have the same height, waist, body fat percentage, muscle tone, etc. and not weigh the same. Maybe one has broader shoulders. Is there some reason people with broad shoulders should be thinner than people with more narrow frames?

That’s why they offer a range, rather than saying “everyone who is 5"6’ should weigh exactly 120 pounds.” And these ranges are not all that narrow. The “healthy” range for a 5’6" female is 117-143, which takes me from something I can only reach when I am working out every day, to my comfortable weight, to being outright chunky with a potbelly and plenty of jiggle on the high end.

Different body types will probably have different ideals within the ranges, but it’s safe to say that even a husky, big-boned women is going to be medically overweight if she is 5’6" and clocking in at 150. If you shoot for somewhere in the “healthy” range, you are probably not facing increased health risks due to weight. If you are outside of the healthy range, it might be a good idea to think about getting closer to the healthy range.

Not what I was trying to say. But it is on the right track. I am decidedly overweight, but my basic build is average for a man of my height. But I suppose you caught something I wasn’t aware that I was implying, as men increase in height, builds tend toward slimmer. And that is true. So the BMI reflects that.

The OP’s question “Do tall and short people not roughly have the same relative dimensions as normal-sized people?” has a pretty clear answer… at least to me. No they don’t. Taller people tend to have slimmer builds and shorter people tend to have stockier builds. To the degree that BMI is a helpful measurement, it does properly map this trend. But as mentioned BMI is only a useful approximation anyway.

Really? I thought people tended to be scaled copies of the same proportions. Do you know of any research on this?

Humans are not built in the shape of a cube with sides the measuring the height, therefore it does not follow “that weight should increase at the cube of dimensions.”

The origins of the BMI go to Adolphe Quetelet in 1832, based on empirical cross-sectional data in an attempt to fit body builds into a normal distribution. Later it was found to be a reasonable and easily applied correlate of weight related health risks and popularized

It doesn’t matter if an object is cube-shaped. As long as the proportions stay the same, if it’s scaled up, the volume will increase with the cube of the length; whether it’s cube-shaped, spherical, donut-shaped, or shaped like a pixie riding a unicorn.
What matters is that as humans get taller, their proportions don’t tend to stay the same. Taller people are typically proportionally skinnier.

Sure. Hereis one that looks at several metrics (leg bone size, stability, heat loss, and actual measurements from Singapore and the UK) to estimate that the body scales between 2.3 and 2.8. Hereis one that suggests 2.6 is an appropriate number for children 2-19 in the US. So BMI understates scale as you go up in height because it only squares. But cubing would overstate it.

BMI does not take into account age, gender, bone mass, or muscle mass. It can’t distinguish between lean body mass and fat mass. BMI could best be considered a poor estimation of someone’s overall health and further investigation IS REQUIRED before someone could be considered over or under weight.

Heavily muscled athletes will have a high BMI and a low percentage of body fat. Old folks may appear “BMI normal” even though they have lost muscle as they age. During his prime playing days, Michael Jordan’s BMI was 27-29, which classifying him as overweight when in fact he simply carried a lot of muscle.

As others have pointed out though, you have to be very muscular for that to be an issue in using your BMI. For most of us, it’s a useful rough measure. Like the fuel gauge in your car.

I think a lot of the hate about BMI is just not liking the result. More to the point, I think there are a lot of people who are approximately average weight and therefore think of themselves as healthy, and then are horrified to be classified as overweight. But the thing is, the average american is on the hefty side by many measures.

Well here’s a real world example that I think helps illustrate why people find BMI frustrating. About a year ago was the last time I met with my bariatrician (MD specializing in weight loss)

According to BMI, my healthy weight range as a 5’10" man is 149 - 183 lbs.

At the time I was 200 pounds with about 160 pounds of lean mass and 40 pounds of fat. That’s a reasonably healthy 20% body fat. I was in no way muscular, in fact I lost about 20 pounds of lean mass while dieting under the doctor’s supervision.

According to BMI I needed to lose 17 pounds to break into the very top of the healthy weight range.

If I lost 17 pounds of fat to get down to 183 (the absolute minimum to be healthy under BMI), I would be down to 12.5% body fat. 12.5% body fat is not at all realistic for most adults given their fitness level. And that’s the absolute top of the healthy weight range.

Obviously the other option is to lose more lean mass as well as fat. I could’ve lost 17 pounds as long as I didn’t have to lose 17 pounds of fat only. But why would i want to? Why would anyone deliberately lose lean mass just to fit into some calculated guideline when they’re already at a healthy body fat percent?

So the bottom line is it’s absolutely false that only extremely muscular people are foiled by misleading BMI judgments.