Arguments Against Obesity As A Disease #1: BMI Is Not A Good Measure Of Health

scaleOver the past months, I have been involved in countless discussions and debates about whether or not obesity should be considered a chronic disease (as it has by the American and Canadian Medical Associations and a number of other organisations).

I therefore thought it perhaps helpful if I discussed each of the common pros and cons on this issue in a series of blog posts.

To begin this short series, I’d like to discuss perhaps the most common argument against calling obesity a disease, namely, the well-known shortcomings of BMI.

As regular readers will know, I have long railed against the use of BMI as a clinical definition of obesity as it is neither a direct measure of body fatness nor does it directly measure health. In fact, its specificity and sensitivity to pick up health problems commonly associated with obesity (such as type 2 diabetes or hypertension) is so limited, that it would not even remotely meet the criteria commonly applied to other diseases for diagnostic testing.

Thus, especially around the BMI cut off of 30 (widely used to “define” obesity in Caucasians), anywhere from 5-25% of individuals would be considered pretty healthy by almost any clinical measure. Even at higher BMI cut offs, it is not all that difficult to find individuals with very mild to non-existent health problems related to their size (as in EOSS 0-1).

While some of these individuals may well go on to develop health problems over time, “risk for” a disease is generally not considered a “diagnosis” of that disease. Thus, even if an elevated BMI may indicate increased risk of obesity, it cannot be used to “define” an individual as having the “disease of obesity”.

This shortcoming of BMI has been widely (albeit perhaps not widely enough) recognised, which is exactly why, for e.g. the Canadian Medical Association, in their declaration of obesity specifically states that,

“BMI is a useful operational definition for obesity but should not be used as the defining characteristic of the disease….in the case of individuals who are very obese, issues of definition and measurement are not relevant.” (emphasis mine)

Similarly the WHO in its definition of obesity states that,

“BMI provides the most useful population-level measure of overweight and obesity as it is the same for both sexes and for all ages of adults. However, it should be considered a rough guide because it may not correspond to the same degree of fatness in different individuals.” (emphasis mine)

So, if not BMI, what exactly should be used as the defining characteristic of obesity?

This brings us to the actual WHO definition of obesity, which states that obesity is defined as

“…abnormal or excessive fat accumulation that may impair health.”

It is as simple as that!

If your body fat affects your health you have “obesity” – if it doesn’t you don’t.

So what is the BMI cutoff for obesity, you may ask – the simple answer is – there is none!

Whether excess or abnormal body fat affects your health or not is not something you can measure by simply stepping on a scale or pulling out a measuring tape.

Answering the question of whether or not someone has obesity always takes a medical exam and tests, at the end of which your health professional should be able to determine whether or not you have “obesity” or just “adiposity” (the medical term for “fatness”).

So if BMI cannot be used to define “obesity” – how can obesity be a disease?

Because, whether or not your body fat is affecting your health (the actual definition of obesity as a disease) has nothing to do with BMI.

Obesity is a “clinical diagnosis” based solely on the clinical assessment of whether or not the quantity (e.g. body fat%) or quality (e.g. visceral fat) of your body fat is affecting your health.

Thus, the argument against the use of BMI to define obesity is not really an argument against obesity as a disease at all. It is just an argument (and a very valid one at that) in favour of finding (or rather applying) a better definition of obesity in clinical practice.

So while doctors should by all means examine patients with a higher BMI for the presence of “obesity” – they should NOT use BMI alone to define it.

Thus, diagnosing obesity in clinical practice is far more like diagnosing “depression” or “ADHD” (where you do not have a numeric cut off but rather a clinical symptom score) than diagnosing diabetes or hypertension (where you do have a numeric cut off).

Even the most vocal critic of considering obesity a disease will likely agree that when obesity affects your health it deserves to be treated (how, is an entirely different discussion).

On the other hand, even the most enthusiastic proponents of obesity as a disease will hopefully agree that when excess body weight does not affect your health, there is no documented benefit of “treatment”.

That is really all that this discussion is about.

Berlin, Germany