Weight a minute - just how accurate is BMI?

No matter how much we would like to believe it's not true, being overweight is bad for your health. Along with the increased risk of type 2 diabetesheart attack, stroke and osteoarthritis, being overweight also increases the risk of getting several cancers. But now the newspapers tell us that BMI may not be an accurate predictor of health-related weight risk after all. They point to stories that rugby star Jonny Wilkinson was classified as overweight when he was captaining the England team, and a bodybuilder told her BMI was bordering on the obese range as she prepared for international competition.

The traditional measure for ideal weight has been based for more years than I've been a doctor (which is more years than I care to remember!) on a formula called BMI, or Body Mass Index. This takes a formula of your weight and height to give a single number. Under 18.5 is underweight; between 18.5 and 25 is 'ideal'; 25-30 is overweight and over 30 is obese.

There's no doubt that muscle, being denser than fat, weighs more, and that for top rank athletes using a BMI may not give an accurate indication of their health risks - or their waist size. With one in four Britons tipping the scales in the obese range (up from 12.5% just 20 years ago) and almost two thirds of Britons now classified as overweight or obese, it's hardly surprising that many people have a vested interest in disproving the traditional measures.

A decade ago, doctors started sitting up and taking notice of alternative measures of size with the publication of a major international study of heart attack risks, which showed that abdominal obesity was a more accurate predictor of risk than BMI. (1) It has become increasingly clear that visceral fat - the kind laid down deep inside your stomach cavity, rather than the kind that accumulates under your skin - is actively sabotaging your body's attempts to stay healthy. It predisposes you to insulin resistance, which in turn drives type 2 diabetes, to unhealthy ratios of heart-protective HDL cholesterol compared to unhealthy LDL cholesterol and to high blood pressure. The World Health Organisation defines a truly high-risk waist circumference as over 102 cm (40 inches) in men and over 90 cm (36 inches) in women - but risk starts to rise with figures over 35 inches for men and 30 inches for women. Waist-hip ratio may be a more accurate indicator still.

But that certainly doesn't mean BMI is useless. If you're superfit, or if you're a 'pear' (whose excess weight is evenly distributed over your body) rather than an 'apple') where excess always seems to accumulate around your midriff) and are in the overweight (BMI 25-30) rather than obese (BMI over 30) range, you might be able to shrug off BMI. But if your idea of exercise is running for a bus once a week, or your BMI is over 30, there's no doubt abdominal obesity and raised BMI go hand in hand.

The risk of developing type 2 diabetes rises steadily with every stage of BMI over the middle of the ideal range (243); with a BMI of 30 the risk is almost 10 times higher in men and 30 times higher in women; with a BMI over 35 there's a 60-fold (for men) and 90-fold (for women) higher chance of being diagnosed. (2,3)

The sad truth is that most of us aren't international athletes. Most of the people I see who struggle with their weight also struggle to do anything like the half an hour on five days a week recommended by the Department of Health. So most of us can't afford to ignore what the scales say, or what they tell us about our BMI. Nor can you afford to ignore practical solutions like the MyHealth tool, which allows you to look at which areas of your lifestyle you can pat yourself on the back for, and which ones you need to take a long hard look at. Packed with practical tips to help you cut your risk of heart disease and other illnesses, it may not turn you into Jonny Wilkinson but it could be the start of a whole new you.


1. Salim Yusuf et al . Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study). Lancet 364; 9437 11 Sept 2004

2. Colditz GA et al. Ann Intern Med 1995;122:481-486

3. Chan J et al. Diabetes Care 1994;17:961-969

Disclaimer: This article is for information only and should not be used for the diagnosis or treatment of medical conditions. Patient Platform Limited has used all reasonable care in compiling the information but make no warranty as to its accuracy. Consult a doctor or other health care professional for diagnosis and treatment of medical conditions. For details see our conditions.