A very lean bodybuilder could still have a high BMI. I don't have a reference offhand, but the US Army has a lot of data on military-service-aged men and their

A very lean bodybuilder could still have a high BMI. I don't have a reference offhand, but the US Army has a lot of data on military-service-aged men and their cardiovascular health risks. They found BMI isn't a great predictor of cardiovascular health. Some formula involving height and the circumference of the belly was a much better predictor of cardiovascular risks. A thin but tall westerner colleague in Japan told me that doctors in Japan are required to counsel anyone with a waist circumference over a certain limit about weight loss. There's no correction for height (or race), only sex. His doctor realized that my colleague isn't overweight, but still had to lecture him. BMI at least accounts for height, but can be similarly misleading.