This blog has moved to Dr. Sharma's home site,

- If you subscribe by email, easily update your email subscribtion here!
- If you subscribe by RSS, your feed should redirect automatically.
- Please change your bookmarks (and any web links) to point to!
- All posts before the move will remain here as an archive and new posts will only be found at the new location.

Monday, October 15, 2007

Why I don't like BMI

I often get asked to explain or define the term "obesity". This is when, as a clinician, I am reminded that the conventional BMI-based definition of obesity is problematic.

To be fair, the concept of BMI has been most useful for population studies and there is no doubt that it reasonably reflects average body fat in a given group of people.

Yes, on average someone with a BMI of 30 will probably have more body fat than someone with a BMI of 25, but does this mean that everyone with a BMI of 30 needs obesity treatment and everyone with a BMI of 25 is safe? The diplomatic answer of course is "it depends"!

"Depends on what?" you may ask. Well, it depends on whether or not a) the higher BMI actually reflects more body fat in that individual and b) the person with the BMI of 30 actually needs treatment.

So the question really comes down to - does a given BMI level help me decide who needs obesity treatment? Well, most clinicians will probably agree that taken alone it doesn't. You probably also need to know the age, gender, ethnic background, waist circumference, family history, current complaints (if any) and risk factor profile to decide who needs obesity treatment.

For example, a young pre-menopausal Caucasian woman, physically active, healthy diet, no risk factors with a BMI of 30 may be safe, whereas a 50 year old South Asian male with elevated triglycerides, hypertension, waist circumference of 95 cms, family history of premature heart disease and BMI of 25 may in fact significantly benefit from losing a few pounds (and keeping them off!).

Well, that is not what the current guidelines or regulators tell me - according to them, our BMI 30 lady has "obesity" and would thus qualify for obesity treatment; our BMI 25 male is not obese and would not qualify - nonsense!

So what is obesity? My rather simple clinical definition is the following:

Obesity is that level of excess fat that threatens or affects someones socioeconomic, mental or physical health - obviously, the level of excess fat that does that will vary from individual to individual depending on their "global risk".

In fact, even with other risk factors such as dyslipidemia, diabetes or hypertension, we have now moved towards "global risk" where we factor in age, gender, co-existing disease, past history, etc. If this makes sense for dysplipidemia, diabetes or hypertension, why not adopt the same strategy for excess fat? - too complex for the busy practitioner?

Well, who said medical decision making has to be easy?


No comments: