Limitations and shortcomings

Some argue that the error in the BMI[16] is significant and so pervasive that it is not generally useful in evaluation of health.[17][18] University of Chicago political science professor Eric Oliver says BMI is a convenient but inaccurate measure of weight, forced onto the populace, and should be revised.[19]

The medical establishment has generally acknowledged some shortcomings of BMI.[20] Because the BMI is dependent only upon weight and height, it makes simplistic assumptions about distribution of muscle and bone mass, and thus may overestimate adiposity on those with more lean body mass (e.g. athletes) while underestimating adiposity on those with less lean body mass (e.g. the elderly).

A 2005 study in America showed that overweight people had a similar relative risk of mortality to normal weight people as defined by BMI, while obese people had a higher death rate.[21]
In an analysis of 40 studies involving 250,000 people, patients with coronary artery disease with normal BMIs were at higher risk of death from cardiovascular disease than people whose BMIs put them in the "overweight" range (BMI 25–29.9).[22] In the intermediate range of BMI (25–29.9), BMI failed to discriminate between bodyfat percentage and lean mass. The study concluded that "the accuracy of BMI in diagnosing obesity is limited, particularly for individuals in the intermediate BMI ranges, in men and in the elderly... These results may help to explain the unexpected better survival in overweight/mild obese patients."[23]

Body composition for athletes is often better calculated using measures of body fat, as determined by such techniques as skinfold measurements or underwater weighing and the limitations of manual measurement have also led to new, alternative methods to measure obesity, such as the body volume index. However, recent studies of American football linemen who undergo intensive weight training to increase their muscle mass show that they frequently suffer many of the same problems as people ordinarily considered obese, notably sleep apnea.[24][25]
A further limitation relates to loss of height through aging. In this situation, BMI will increase without any corresponding increase in weight.

A study by Romero-Corral et al., using data representing noninstitutionalized civilians in the United States, found that BMI-defined obesity was present in 19.1% of men and 24.7% of women, but that obesity as measured by bodyfat percentage was present in 43.9% of men and 52.3% of women.[26]

The exponent of 2 in the denominator of the formula for BMI is arbitrary. It is meant to reduce variability in the BMI associated only with a difference in size, rather than with differences in weight relative to one's ideal weight. If taller people were simply scaled-up versions of shorter people, the appropriate exponent would be 3, as weight would increase with the cube of height. However, on average, taller people have a slimmer build relative to their height than do shorter people, and the exponent which matches the variation best is between 2 and 3. An analysis based on data gathered in the USA suggested an exponent of 2.6 would yield the best fit for children aged 2 to 19 years old.[27] The exponent 2 is used instead by convention and for simplicity.
As a possible alternative to BMI, the concepts fat-free mass index (FFMI) and fat mass index (FMI) were introduced in the early 1990s.[28]