To the Editor:
In their letter, Moreault et al. suggest using a constant body mass index (BMI) to compute ideal body weight.1 Although I agree that this would simplify a commonly used calculation, and the magnitude of any error is small and likely not clinically consequential in this context, this letter provides an opportunity to point out a little appreciated concept: namely, that the normal BMI for a taller patient should be larger.
BMI is calculated as weight (a measure proportional to length [L] cubed, via density) divided by height (length) squared. If as one grows in height, much of the body also grows in width and depth, then the mathematics dictate BMI increases as the height increases (i.e., L3/L2 increases as L increases for L greater than one). This relationship is very apparent if one inspects the pediatric charts of BMI versus age with the understanding that height increases with age. In adults, the increase in BMI with height relationship can be demonstrated by calculating the BMI as a function of height from any of the ideal body weight formulas (e.g., Acute Respiratory Distress Syndrome Network, also known as ARDSnet).2 The Corpulence Index (Ponderal Index, Rohrer’s Index) is computed as weight divided by height cubed, and seems to fluctuate less with height than BMI, but in medical journals, it is not commonly used for adults.3
This concept may be more significant when stratifying patients into obesity-related categories. For those on the border, this may inappropriately place a taller patient into the more obese category than dictated by physiology.
The author is six feet six inches tall.