Obesity expert: A better fat measure than BMI

You have come up with an alternative to the body mass index as a measure of obesity. First off, what's wrong with BMI?

The BMI has been around since the 1840s, but it has a number of weaknesses. Firstly, it doesn't give a real estimate of percentage body fat. Secondly, the BMI can be quite different for a man and a woman with the same percentage of body fat. And thirdly, your BMI can be high even if you don't have much fat, especially if you are male and very muscular.

How did you go about searching for an alternative to BMI?

Our goal was to find a simple index of obesity, something that a practising clinician could use. To do that we looked at a population of 2000 people of Latin American descent who had had their percentage body fat measured directly using dual-energy X-ray absorption, which is an accurate way to quantify body fat. We then asked what parameters we could measure in these people that would best predict the true percentage of fat.

What did you find?

It turned out that hip circumference and height were more correlated with percentage body fat than anything else, including waist circumference and weight. So we designed an equation that could take both of these into account. We call this the Body Adiposity Index. It turns out that BAI is a good predictor of percentage adiposity, so if your BAI is 30, then your percentage body fat is around 30 per cent. It is reasonably accurate - not terribly accurate - but usable as a clinical tool.

Is BAI better than BMI?

We think it's better, but we have still got to prove it. Unlike BMI, the BAI for men and women is the same if they have the same percentage body fat. We have validated the BAI in African American populations too. Its utility has not been confirmed in Caucasian subjects, although we have tested it on a small group and it seemed to fit.

What are the downsides of the BAI?

The real challenge is to be able to predict the risk of obesity-related diseases such as cancer, diabetes, cardiovascular disease and hypertension, and then to intervene. It remains to be shown that BAI is a more useful predictor of these outcomes than other measures of body adiposity.

Were you surprised that weight isn't part of the BAI calculation?

Yes. But this means that BAI has the unexpected characteristic that it can be used where scales are unavailable or not correctly calibrated. BAI could be useful in remote locations with no reliable scales; in India, for example, where obesity is a serious problem.

Do you think BAI will one day succeed BMI as a measure of obesity?

I am agnostic on that, but I'm hopeful that BAI is better than BMI, which is misused by a lot of medical practitioners who don't realise that it is often not a good measure of percentage body fat.

فیروز نیکزاد منکل