A recent study by the National Audit Office found that obesity accounts for 18m lost working days a year and, if cases increase at their current rate, it will cost the UK economy £3.5bn a year by 2010.

ran an article on this growing, near-epidemic, problem of obesity. The piece, with which I had no gripe at all, was illustrated by a cartoon of an overweight and uncouth lout with a scowl on his face and a T-shirt bearing the slogan 'Gross National Product'. I was moved to complain to the editor. The cartoonist, with his clever play on words, had unwittingly illustrated a major barrier to the effective management of obesity in this country: the confusion of facts (ie, the medical risks of obesity) with values (ie, the aesthetics of body form).

The word 'grossly' is frequently used in both medical casenotes and by journal-ists to qualify the term 'obese'. 'Gross' does not mean that the person is very fat, or even that their obesity is severe enough to threaten their health, but that the observer finds them unattractive. Such terminology has no more place in a medical discussion than saying someone has a gross breast lump or an ugly sore throat.

Obesity as a medical condition should be described in terms of the body mass index (BMI) - the individual's weight in kilograms divided by the square of their height in metres. For example, I am 165 cm tall and weigh 58 kg, which gives me a body mass index of 21.3. In my role as a university examiner, I have a reputation for penalising medical students who describe patients as 'grossly obese' - especially if they have failed to measure the person's actual weight and calculate their BMI. A BMI below 20 kg/m2 may indicate underweight. Between 20 and 25 kg/m2 is described as normal - but note that the limits of normality are wide: I would still be 'normal' if I gained 10 kg (22 pounds) in weight. In this context, the word 'normal' means 'not proven to increase the risk of medical complications'. A 10 kg weight gain would take me from a size 12 to a size 16 - but it wouldn't, statistically, increase my risk of heart disease, arthritis of the knees, breast cancer or any of the other 33 conditions associated with more severe degrees of overweight.

A BMI between 25 and 30 kg/m2 counts as medically overweight - ie, at this level, your size is not simply a matter of aesthetics but a proven risk to your health. A third of women and 46% of men in the UK fall into this category.

With a BMI above 30 kg/m2 (which includes a further 21% of women and 17% of men), the health risks are much greater - with an average loss in life expectancy of nine years, mainly from heart disease. One third of these people (6% of the population) will die directly or indirectly from the consequences of obesity. Obesity costs the National Health Service £500m a year and the economy an additional £2bn. And if we keep getting worse at the current rate, we will be as bad as the Americans within 10 years.

The problem is usually expressed in terms of individual behaviour. We eat too much of the wrong foods, take too little exercise, and engage in a host of idle, passive leisure activities. Our children are the playstation generation - driven to school, brought up to expect one-click multi-media entertainment, and grazing almost continuously on what is rightly termed 'junk food'.

But epidemiologists Garry Egger and Boyd Swinburn disagree. The full text of their article An 'ecological' approach to the obesity pandemic is available on http://www.bmj.com/cgi/content/full/ 315/7106/477. Egger and Swinburn acknowledge that obesity in an individual is inevitably the result of an ongoing energy (calorie) intake that exceeds expenditure, and that the balance in that equation is determined by individual behaviour choices (such as taking the lift instead of walking upstairs). But, they argue, the influence of individual behaviour choices must be considered in epidemiological terms - ie, against a background of an environment that is increasingly obesity-promoting.

Compare, for example a very different epidemic - one of cholera. An individual who gets cholera has of course been infected with the cholera bacillus and has (we infer) consumed contaminated food or water. But the cholera-producing environment is one in which poverty and overcrowding continually set the scene for the contamination of water supplies with infected sewage. Just as we could not contain a cholera epidemic simply by telling everyone to be more careful with what they eat and drink, we cannot contain the obesity epidemic by putting out platitudes about gluttony and sloth in health education leaflets and TV programmes.

So what is the solution? According to Egger and Swinburn, society as a whole needs to confront and change the 'obesogenic' environment. We don't need an elite few to take absurd levels of physical exercise - we need a relatively small increase in energy expenditure in the population as a whole. Fewer of the pointless labour-saving devices (such as the TV remote) that reduce 'incidental movement' would be a good start. Let's see as many apples as chocolate bars on sale at news-stands. A jug of water should be provided as standard service on every table in restaurants. And the next time you want a document from your filing cabinet, don't send an email to your secretary - get up and pull the drawer open yourself!