> The fact that it takes into account just height and weight is problematic. No consideration of bone structure or body frame. That's a bug, not a feature, if

> The fact that it takes into account just height and weight is problematic. No consideration of bone structure or body frame. That's a bug, not a feature, if you allow me to grab the expression from software. It's problematic as any other indicator has "problematic" features. My point being: when you assess any indicator, you can't look only at it's ability to correctly reflect reality (validity). You'll also have to consider other stuff: is it cheap? can I trust people to use it correctly? can it be used consistently? Taking into account only height and weight will diminish validity, but give you positive answers to these three questions. It's a trade-off (and, IMHO, one that's being bashed excessively, considering the empirical evidence saying it's not that "invalid" at all). Compare it with e.g. skinfold estimation of body composition: skinfold estimation is more valid, yet science shows us that it requires a highly trained specialist for it to be reliable (i.e. to do it consistently; and even with professionals, there's not much consistency). This is also problematic, but - again - it's a bug, not a feature. > At my absolutely lightest, when you could see every rib sticking out, my BMI was still comfortably in the range considered overweight. ... and that's you being an outlier, or close. :) I'm one as well; also overweight, by BMI standard. But I have to consider that my care with nutrition and fitness should also be close to outlier in nature, regarding population in general. Knowing this, you and I (and our physicians) will take it into account. And - if the need arises - use another metric. I do both skinfold estimation of body composition and waist circumference measurement but rely mostly on the latter for tracking progress (I'm not a professional and, of course, it's me measuring myself, so I choose the method where I might mess up less).