Faith lost in complaint system
By Nerys Hairon
Body mass index is the 'wrong measure' of obesity and fails to predict risk of myocardial infarction, a major international study concludes.
The Interheart study of over 27,000 patients from 52 countries found using waist-to-hip ratio instead of BMI would triple the numbers classified at high MI risk.
The study, published in this week's Lancet, was accompanied by a commentary pronouncing the 'final nail in the casket' for BMI as an independent cardiovascular risk factor.
'Current practice with BMI as the measure of obesity is obsolete and considerably underestimates the grave consequences of the overweight epidemic,' said the commentary by researchers at Akershus University Hospital in Oslo, Norway.
The study of 12,461 cases with a first MI and 14,637 controls found BMI only showed a 'modest and graded association' with MI, which disappeared after adjusting for other risk factors.
But the Interheart researchers found waist-to-hip ratio and waist and hip circumferences were still closely linked with risk of MI even after adjusting for other risk factors.
Patients in the top fifth for waist-to-hip ratio were at 75 per cent greater risk than those in the lowest fifth, and those in the highest fifth for waist circumference were at a 33 per cent increased risk.
Some 24 per cent of patients in the top two fifths for waist-to-hip ratio were at risk of MI but just 8 per cent of those in the top two fifths for BMI.
Study researcher Dr Sonia Anand, associate professor of medicine at McMaster University in Ontario, Canada, said BMI was 'the wrong measure of BMI and never remained a strong factor to predict CV risk'.
'So now we should use waist-to-hip ratio as a marker of risk probably much more informative than blood tests to predict risk of diabetes and CVD.'
But GP obesity experts reacted cautiously to the suggestion of dumping BMI.
Dr David Haslam, the chair of the National Obesity Forum, suggested that BMI should not be 'consigned to the waste bin yet' and was important at the population level and in historical data.