2 ms·

IMHO, it's still bullshit. >This means that if you heroically exercised so much that you converted 10% of your body volume from fat to muscle (wow!), your BMI

IMHO, it's still bullshit. >This means that if you heroically exercised so much that you converted 10% of your body volume from fat to muscle (wow!), your BMI reading would go up just 1.8%. That's much less than the corrections just mentioned for short or tall people. This is wordplay. You don't turn fat into muscle. A correct measurement of someone who gains 10 lbs of muscle would be the same as someone who weight 10 lbs less, not someone who carried it as fat.

kazinator 11y ago
The BMI concept is volume based. If people grew equally in all dimensions, and this was proportional across all tissue types, then height would be cubed. What Trefethen is doing there is arguing that although muscle is more dense (and so a muscular person of the same weight and height has less volume than a fat person), it doesn't make that much difference. In other words, I think he's arguing that the height ^ 2.5 is still a good normalizer if we restrict ourselves to the population of muscular people. Like say, if we focus on males who have 4% body fat, of different heights, the height ^ 2.5 based BMI could tell you how built they are, corrected for height, still in a better way than using an exponent of 2. On a related topic, there is a "fat-free mass index" (FFMI) which applies the BMI formula to just an individual's lean mass. This could work well with the same correction.