From Hedley Brown
How refreshing to see Richard Bergman suggesting an alternative to the body mass index (BMI) of obesity, which makes muscular men anxious (12 March, p 31). Back in the 1970s my colleagues and I discovered that there was a weirdly accurate correlation between the incidence of coronary disease in the UK and the ratio of waist to inside-leg length of trousers sold by a well-known drapery store.
I could not persuade the big trouser company to part with the raw sales data, so we could not publish in a journal. But we had seen the heart and trouser maps, and we spread the news at lectures in the hope of un-spreading waistlines.
From Frank Hollis
When I saw this item was about somebody questioning BMI as a measure of obesity I thought “at last!” But I was surprised to find that its most obvious flaw was not mentioned. Indeed, I’ve never seen it mentioned. Why does the BMI assume that one’s weight should be proportional to the square of one’s height, when it should obviously be proportional to the cube of height, that is to one’s volume, for an optimum body density?
Advertisement
Steyning, West Sussex, UK
The editor writes:
• A tall person who was a scaled-up short person could be very big-boned indeed: so the mass divided by the cube of height (the “Ponderal index”, apparently) is wrong too. BMI is a workable approximation.
Nunthorpe, North Yorkshire, UK
