- If BMI fell between 23.0 and 23.9 – on the high side of the WHO’s “overweight” category – there was no increased risk of death.
- If BMI fell between 21 and 22 – in the “normal” range, there was a 12% increase in death risk.
- If BMI fell between 20 and 20.9 – on the lower end of “normal” – there was a 19% increase in death risk.
- By contrast, if BMI fells between 33 and 33.9 – on the high side of “obese” – there was (only) an 8% increase in death risk.
Critics of the standard BMI measurements say the formula doesn't account for other important factors, like age, and muscle-to-fat ratio. An article in Medical News Today highlights several other concerns about using BMI to evaluate health:
What is the problem with BMI?
BMI is a very simple measurement which does not take into account the person's waist, chest or hip measurements. An Olympic 100 meters sprint champion may have a BMI higher than a couch potato of the same height. The couch potato may have a big belly, not much muscle and a lot of body fat on his hips, upper thighs, in his blood and other parts of his body. While the athlete will have a smaller waist, much less body fat, and most likely enjoy better health. According to a purely BMI criteria, the couch potato is healthier.
BMI does not take into account bone density (bone mass). A person with severe osteoporosis (very low bone density) may have a lower BMI than somebody else of the same height who is healthy, but the person with osteoporosis will have a larger waist, more body fat and weak bones.
Many experts criticize BMI as not generally useful in evaluation of health. It is at best a rough ballpark basic standard that may indicate population variations, but should not be used for individuals in health care.
Put simply: experts say that BMI underestimates the amount of body fat in overweight/obese people and overestimates it in lean or muscular people.
More information on BMI, together with imperial and metric BMI calculators, is available here.
Nick Trefethen, a Professor of Numerical Analysis at Oxford University's Mathematical Institute, has created what he believes to be a better, more accurate and relevant formula for the BMI one for deciding whether people are carrying too much fat. Humans do not grow equally in all three dimensions, he explains - the existing BMI formula presumes we do.