A recent NYTimes article asks whether you can be overweight and still healthy. I want to quickly point out an inherent problem in the use of “overweight” as a measure and a problem with using BMI.
“Overweight” is determined by having a BMI over 25, while “obese” is having a BMI over 30. BMI is, however, a flawed measure of healthy fat distribution for certain population groups. A 6′ tall individual has to be 183 lbs. or under to be considered “normal weight”. If you actually see a 6′ tall person who is 184 lbs, there’s no way you would consider him/her anyone close to overweight. Conversely, overweight shorter people benefit from this quirk in BMI.
The reason for this is the way BMI is calculated as mass divided by height squared; it is inversely proportional to the square of the height. So after a certain point, at a constant weight, as height increases the BMI decreases at an exponentially slower rate (imagine the 1/x² graph). Thus, taller people have a higher BMI than their body fat would suggest. Additionally, since muscle weighs more than fat, a healthy individual with a lot of muscle mass can be overweight while an individual with fat instead could have a healthy/healthier BMI.
So the conclusions from studies based on BMI should be taken with those facts in mind. I think ideally a better measure for fat content should be created that will be more complicated, but more accurate, than the oversimplified BMI. Some people advocate for waist circumference, but it’s a bit naive to create a measure that can be manipulated at will by a patient. Normal ranges of height/weight could be calculated demographically, like a the growth curves we have in pediatrics clinics, instead of just using a formula that attempts to capture a trend. This might not solve for muscle mass, but would solve for the graphical irregularities you get from BMI.