Furthermore, what really matters is not so much the comparison with other men or women but specifically whether one has an addictive tendency. Correspondingly, anaemia is always individually important even though it is generally more common in some countries than in others. Men who suffer from the "nurturant of self" group of addictive outlets illustrated here, most commonly channel their addictive urges into work and exercise whereas women overeat or starve or become bulimic and will also tend to shop and spend compulsively. Incidentally, when women shop compulsively, they tend to buy food or clothes whereas men tend to buy books, computers, sports or gardening equipment or something to do with motor vehicles; but these sexual stereotypes are not always true. We do see men who binge or starve or shop compulsively and we frequently see anorexic women who exercise excessively and overeaters who take no exercise at all. Their average state is meaningless.

Consequently, I am not really interested in how one woman overeater compares with women in general or with men because there may be more women bingers in general than men and more women bingers and starvers than men. Nor am I particularly interested in comparing the numbers of male and female workaholics even though the proportion in each sex may be the same. My interest is primarily in identifying them. I think it better to keep things simple and look for the addictive characteristics upon which the questions were based rather than complicating matters in a maze of numbers. Thus a man scoring eleven points on the food starving questionnaire may be comparable to only twenty per cent of men but still not concern me from an addictive perspective because anorexia in men is rare and this skews the cut-off point downwards inappropriately if one is examining the general population.

Correspondingly, an overeating woman would definitely concern me on a score of twenty-two points even though thirty per cent of other women are like her. Not only are eating disorders common in women but there tend to be more overeating problems in women than in men because in our culture women tend to be more involved in food preparation. This does not mean that compulsive overeating is environmentally acquired rather than genetically linked. It means that those women who have the "nurturant of self" addictive tendency (which I believe is probably genetically inherited) will be more likely to express it in bingeing rather than, say, in work. Overeating women will therefore tend to skew the cut-off point upwards, just as in England the fact that ten per cent of the adult population drink fifty per cent of the total volume of alcohol consumed skews the average per capita consumption upwards.

The decision on which form of treatment is required (simple abstinence on its own or coupled with outpatient or inpatient professional care) will depend upon assessment of the whole range of addictive outlets rather than solely upon the answers to the bingeing questionnaire. That surely is what should really matter to me or to the patients.

The purpose of making this broad assessment of addictive behaviour is to show patients themselves the range and intensity of their addictive outlets. Patients commonly say "everyone I know does that", and it is helpful to show them that, while