The body mass index is known to be a 'blunt' instrument. It does distinguish fat from muscle, says nothing about where fat is distributed, and cannot begin to assess metabolic risk.

But it is nonetheless very useful for providing the 'view from altitude,' or in other words, the epidemiologic view. As the average BMI in our population goes up, it is a very reliable indication that obesity rates are rising. It would be nice if we had to worry about a sudden outbreak of muscularity, but alas- we don't!

And, as more people get heavy, more people get heavy in harmful ways- excess fat around the middle, for example. So tracking BMI at the population level is also a reliable gauge of the overall health toll of obesity.

But that's where the true utility of the BMI ends. It is of limited use at the individual level, where waist circumference is better. Body composition may be better still, but adds cost. Until we can measure body composition as reliably and inexpensively as BMI or waist circumference, it is unlikely to be 'transformative.'

I do find it very useful, however. For one thing, patients losing weight the best way of all- by both eating better, and exercising- may lose fat, but not pounds. They may gain muscle as they lose fat, and get leaner without getting lighter. Body composition measures are required to convince them of this 'success' to which the scale is oblivious.

BMI will be around for the foreseeable future as a useful tool in epidemiology. Waist circumference is an important, cheap, and easy addition for assessing individual risk. Body composition measures add value, but at a cost- which is justified thus far in select cases.