Checking BMI's a fat lot of good (by Andy Ho in the Straits Times dated 10 October 2009)

By Andy Ho, Senior Writer

IN 2006, my ticker gave me a health scare. So I invested in a motorised treadmill and a stack machine with pulleys and weights. Helped along by some obsessive calorie-counting, I dropped 15kg within months.

Alas, I have since regained half of that weight despite one hour on the treadmill every day - and watching what I eat. True, the stack machine has collected dust but that is because my doctor repeatedly reminds me to focus on aerobic exercise, not strength training.

However, if you just walk or run and diet without strength training, your body fat can remain high. You could even appear slim and yet have lots of body fat. And your body mass index (BMI) - weight in kilos divided by the square of height in metres - could appear normal even if you have lots of body fat but too little lean body mass.

Lean body mass is the body mass minus the fat - or basically, muscles and bones. It is only weight training that can build up one's lean body mass.

However, the BMI - a measurement that is widely cited - does not distinguish between fat and lean mass. While the former is related to significant morbidity, the latter lowers health risks.
Moreover, the BMI glosses over fat distribution, which matters. While fat at the waist could be harmful, fat at the hips and thighs may not be so.

Deciding how best to distinguish between the types of fat matters because it will lead to a more accurate delineation of higher- and lower-risk sub-groups. These groups can then receive different intensity of treatment, including specifying what levels of good and bad cholesterol they should attain.

But what the public health authorities care about most is the overall relationship between BMI and the risk of death from all causes. Alas, the BMI misses the smaller picture - the individual: You!

For example, patients with a normal BMI may also have Syndrome X - the combination of high levels of blood lipids, especially triglycerides, diabetes and hypertension - which raises their risk of cardiovascular problems.

So a normal BMI may not mean that your risk of ischemic heart disease is low. In fact, it may even lull you into a false sense of security. You might even dismiss early warning signs - like feeling slightly breathless at a weekend game of soccer.

Despite these shortcomings, the World Health Organisation uses the BMI to define the various grades of obesity. Large studies do consistently show that the most obese groups are two to four times more likely to die earlier than normal-weight groups. But notice the operative word: 'groups'.

In large populations, almost any reasonable surrogate measure of fatness will correlate well with obesity-related health conditions. Survey measures like waist circumference or waist-to-height ratio are probably as good as BMI.

In March this year, The Lancet reported a huge Oxford University analysis that pooled together 57 studies involving 900,000 participants from four continents. It confirmed 'once and for all... obesity shortens lifespan' - obesity here defined by BMI.

But some people are heavy because they are also tall. A muscular man may weigh a lot and yet be healthy. Statistically, errors introduced into the BMI by such facts all wash out when there are many participants in a public health study. Thus, in a large population survey, a high BMI can mean obesity - overall.

However, populations are made up of individuals, whom doctors have to treat one at a time. As an adult's height does not change dramatically, the BMI reduces to really his body weight. Why repeatedly calculate a patient's BMI?

In exercise studies, body weights - and thus BMI - often remain the same because one may lose some fat but put on muscle and become healthier.

Measuring body fat, however, used to be a very cumbersome task. It basically involved being submerged in a tank of water - if you could find a facility with the right equipment to do that.
Enter the Bod Pod, a human-size, computerised egg-shaped pod that can accommodate even a sumo wrestler. You enter and sit inside while the gizmo measures your body fat. Like the tank method, it relies on the Archimedean 'Eureka' principle - or in this instance, how much air you displace from the pod. If one has too much body fat - even if one's weight and BMI are normal - the Bod Pod will be able to tell.

Normal-weight patients who nevertheless have too much body fat are said to be prone to be 'metabolically obese'. Such non-obese looking people who actually have lots of body fat are often found to be also afflicted with Syndrome X. These not-fat-lookers thus have a high cardiovascular risk. To alert people to this risk, they are now given the newly minted diagnosis of 'normal weight obesity'.

A Mayo study last year showed that 55 per cent of adults with normal BMI have enough body fat to be classified as 'obese'. According to WHO criteria, men with 25 per cent body fat and women with 35 per cent body fat are 'obese'. It turns out that even people with normal body weights may carry such amounts of body fat.

In the Mayo study, women with 'normal weight obesity' were four times more likely to have heart disease than controls. (There were too few males in the study for statistical analysis.)

The upshot of all this is: First, a normal weight or a normal BMI need not mean your health is good. Second, the BMI must be seen for what it is - a poor tool to use to diagnose obesity in the individual. And third, we should screen for body fat - including those people of normal weight.

If getting our body fat measured periodically could save lives, here is a business idea: Set up an island-wide chain of Bod Pod outlets.

Any takers?

Note: Now read BMI has its uses, says health board [Ref: add-i09J30-2352]