Is there any basis for the BMI? It is essentially the weight (why call it the “mass index”) divided by the square of the height. This doesn’t make sense dimensionally since the weight will scale as the cube of the height. To put this in perspective, imagine a 5 footer and a 6 footer of exactly the same proportions. If the 5 footer has a BMI of 25 (just at the boundary of overweight), then the 6 footer will have a BMI of 30 and will be called obese (again, just at the boundary).
I just read that Shaqille is 7’3" and weighs 340 lb. This gives him a BMI of around 31.5. Is he obese? Obviously this index was dreamt up by dwarfs.
Comparing weight to height is cut and dry for big cement blocks of constant density, but pretty tricky for people who are shaped funny and who have organs that weigh very different amounts. In medical practice, body surface area in square meters is often used which is thought to correlate well with weight. So as a doctor and mechanical engineer, I do not know if BMI makes dimensional sense, but I do not see why it needs to be a dimensionless number. This ain’t the Reynolds.
No one thinks the BMI is perfect. Strong people with lots of muscles have elevated weight for their heights and bigger BMIs. BMIs are used since, in addition to being easy to do in the office (no messy skin fold measurements or float tanks) they are clinically sown in studies to correlate with ideal weights, morbidity and mortality. Few of the people used in this study were seven feet tall. Conclusions from a study are often only as good as the population studied.
So no one thinks the BMI is perfect. Levels of obesity and overweight are slightly arbitrary. But for most people, including most tall people, it remains a useful index.
Other indices for obesity which can be done quickly and correlate to disease are waist measurment and hip to waist ratio.
I think the argument of the first guy is quite clear. Most people would believe that to a large degree the shape of a human is height invariant. The BMI suggests it is extremely height dependent. It would be interesting to understand why that is so. Is it because the shape changes or because the density of humans change because of bones etc. or is this just a stupid rule
BMI is a risk factor. If you’ve got two guys who are both in the same excellent shape overall but one outweighs the other by 20 lbs., the heavier one is at higher risk for heart disease and stuff like that than the lighter one, even though they’re both at lower risk for that stuff than the general population.
BMI doesn’t work well for very short people, and it may not work well for very tall people, but it works pretty well for average folks, and it’s dirt cheap to measure.
Weight may scale as the cube of dimension, but I’m less sure it scales as the cube as the height. There is considerable variation in the width and breadth of individuals. Obese people may be just as tall as a skinny person, but they also tend to be wider and thicker.
I agree the BMI doesn’t make full dimensional sense, but two out of three ain’t bad.
If you’re interested in more info, this calculator is one of the best out there. He has also calculates weight/height percentiles and has lots of fascinating links explaining the BMI basis, why it may not be accurate for you, and how weight/height averages vary by ethnicicity, age, etc.
Generally, BMI is a nice guideline. However, things gets askew when you have someone who drifts from the median range for height, or drifts from the median range for muscle mass.
Even when carrying very low body fat, I would still not look great on the BMI, because of muscle mass.
Checking body fat would be a step better than BMI, and body fat analyzers are getting to be as affordable as good digital scales.
I’m not so sure checking body fat is better than BMI. Body fat is hard to measure and requires special equipment, even so the result is often inaccurate. BMI is quick and easy and requires no special equipment. Many of the studies that show cardiac morbidity (e.g. risk of having a heart attack over the next ten years) use BMI as a measure; furthermore BMI correlates well with obesity for the average Joe. The fact this does not apply to dedicated weightlifters can be taken into account, but dedicated weightlifters do not make up a high percentage of the population.
It’s easy; and special equipment is about as special as a scale. Last time I checked, to get a BMI, you need to do the same thing you do for body fat: step onto a scale-sized object and get a reading.
Since body fat % standards apply to virutally all humans, is easy, and requires special equipment on the scale of a scale, it should reign supreme in any discussion about health.
One of the reasons that BMI is so often used, despite it’s known limitations, is that it is so often used.
At this point, we (epidemiologists) have observed increased BMI as a risk factor for adverse health outcomes in many studies, over many years, for many outcomes, and most importantly in many populations.
I’m not doing to try to defend BMI as any kind of ideal or ‘gold standard’ for measuring fitness or body composition. And, studies that want to examine the ideal measure question in detail can be done and have been done.
However, as has been mentioned, BMI is easy to measure, it’s readily available even from data that were not collect originally for research purposes (very important for some outcomes), and we do know that it is predictive of outcome in the general population.
We use it because we know it works, not because we think it’s perfect.
A family doctor can do a house call and measure a BMI. When I did some medical electives in India, I could find a scale that measures weight easier than a fancy scale which measured impedance (I don;t think I could have found this at all). When you talk about WHO guidelines (as on the cool BMI site above), one finds the world is bigger than G7. Given how inaccurate the Tanika scales are for a group of people, lots of medical specialists would rather stick with a regular scale (which has lots of other applications, e.g. pregnancy, CHF, cancer) then purchase a fancy one.
