Britain has an obesity problem — of that there is no question. Walk the streets of the average town or city and the evidence will be waddling around you, gasping for breath and overcompensating for spent calories at fast-food outlets.

The statistics, aired frequently, are alarming. One adult in four is dangerously overweight, up 75 per cent on a decade ago, while more than a quarter of children under the age of 11 fall into the overweight or obese categories, triple the level of 1990.

But the super-fat can take heart — if that overworked pump can still function — because they are in exalted company. Under the current one-size-fits-all obesity measurement, called the body mass index (BMI), Brad Pitt’s chiselled physique turns out to be portly, while the muscular build of Russell Crowe consigns the Gladiator star to the legions of the lardy.

Confused? You should be, and so are the scientists. A team from the Mayo Clinic in Minnesota found last week that “overweight” sufferers of heart disease had better survival rates than those classified as “normal”. While severely obese patients had a higher risk of death, overweight people had fewer heart problems than those with a normal BMI. The normal group were, in turn, less likely to die than people with a low BMI.

It has led to further questioning of the BMI scale, which has defined healthy and abnormal weight for more than 100 years. Invented by the Belgian statistician Adolphe Quetelet in 1869, it is calculated by dividing a person’s weight in kilograms by their height in metres squared. Someone with a BMI of less than 18.5 is considered underweight, between 18.5 and 24.9 is “normal”, and 25 to 29.9 is classified as “overweight”. Clinical obesity is a BMI of 30 or greater.

The scale, which is used universally in clinical trials assessing the health risks of weight, has remained largely intact since the 19th century. The categories were revised downwards in 1998, sending many “normal” people into the overweight class with a single keystroke.

The BMI’s shortcomings have been exposed repeatedly by anomalies such as Pitt, Crowe and the “dangerously overweight” All Blacks’ rugby star Jonah Lomu (BMI: 32). Yet important medical research continues to rely on it, hospitals and GPs still refer to it, and the findings and health of patients are compromised as a consequence.

The Mayo team quickly distanced last week’s study from its obvious and incongruous conclusion, that overweight people are more healthy than normal people. The researchers instead explained that the better outcomes for overweight patients were most likely on account of muscle, which weighs more than fat.

More importantly, the results showed the pressing need to replace the BMI for a new scale that better reflects fat and muscle distribution. It is this subtlety that, when ignored, can see a fit athlete with little fat but highly developed muscles classified as obese. Measuring a person’s waist circumference or waist-to-hip ratio (WHR) is widely considered a better means of detecting visceral fat, the fat that produces

the classic “middle-aged spread” of an extended abdomen and which remains a good predictor of diabetes, heart disease and heart attack risk.

Yet for reasons that can be down only to habit — and certainly not convenience — the BMI still persists. If public health experts really wish to tackle obesity, its alarming contribution to chronic illness and the huge financial tolls it exacts on the NHS, they must wean themselves off the index. Because while belt-busting fatties may not be in the right shape, nor are the numbers, and only by addressing them will we discover who among the supposed eight million British adults is really most at risk.

Sam Lister is the Times health correspondent. Mark Henderson is away