Where are we now?
Getting to grips with obesity BMI or waist?
Continuing our series on key clinical controversies, Professor Mike Lean tackles the growing problem of obesity
The potential for GPs to improve health and prevent disease by tackling obesity is a no-brainer. Being overweight or obese is responsible for virtually all type 2 diabetes (a very rare disease in thin people), about half of all hypertension (our most expensive disease), and thereby around 30% of cardiovascular disease. Add to this perhaps 10% of all cancers, 20% of all depression, 20% of all arthritis and you begin to see the fuller picture.
Patients are screaming for help – especially to avoid weight regain – and yet GPs have resisted involvement.
It was until quite recently argued that GPs should not bother to measure anything as first, they could spot obesity by looking, and second, it was not worth doing anything about it, so why bother1?
These two points are both wrong.
• It is surprisingly difficult to identify a person as having a body mass index (BMI) of 25kg/m2 or 30kg/m2 accurately, say to within 2 units, on a casual inspection – even by experienced experts. • We have copious evidence that there are surprisingly large medical benefits from modest, achievable weight loss (5-10kg) achieved by a variety of means.
Reflecting on the evidence
There are many moderate or high-intensity interventions shown to be efficacious in randomised clinical trials, but they can never be effective under the practical or cost constraints of routine practice. Minimal intervention has been shown to be ineffective2.
The Counterweight programme
The Counterweight programme has been developed as a cost-effective evidence-based intervention programme for UK primary care. About 30-40% of obese patients can achieve 0.5% of weight loss, which DPP3 and DPS4 show will reduce diabetes by almost 60% as well as improving all cardiovascular risk factors. The key to success lay in using well-researched methods and materials (evidence based where possible) and a carefully designed training programme for practice nurses and GPs.
The potential savings from reduced drug prescriptions alone are greater than the cost of running the Counterweight intervention.
BMI – a poor predictor
GPs are correct in saying that BMI is a relatively weak predictor of ill-health. At least within the BMI range 18.5-30kg/m2, variation in muscle mass (for example in athletes) can overwhelm differences in body fat. BMI is a crude index using widely available measures (height and weight) for broad categorisation in epidemiology. It is not intended as a sole diagnostic criterion for obesity. But above a BMI of 30kg/m2, body fat is nearly always high.
Waist circumference – a better predictor
Waist circumference is now being used worldwide as a better predictor of total body fat than BMI. It is not a perfect predictor, but it is not affected by variation in height or muscle mass, so it gives better prediction of body fat and health within the normal and overweight BMI range 18.5-30kg/m2.
Some of the inaccuracy of waist measurements in predicting total body fat arises from the additional influence of intra-abdominal fat. Because intra-abdominal fat is pathophysiologically linked with metabolic syndrome, this source of error in predicting total body fat gives the waist circumference extra power in predicting ill-health.Furthermore, most ordinary people can understand and monitor waist circumference, but have difficulty conceptualising BMI. Some confusion has arisen from studies that used non-standard methods of measuring waist circumference (such as Interheart) and by persistent use of weight to height ratio, but waist circumference is one of the most robust, reproducible anthropometric measures. Each 1kg weight loss is matched by about 1cm reduction in waist.
Practical implications for GPs
From the foregoing discussion it emerges that waist circumference is a better and easier measure than BMI for any purpose:
• for epidemiology to monitor the epidemic• for statistical prediction of body fat and of health outcomes• for clinical decision-making.
Waist circumference cut-offs, first established from research in Glasgow and the Netherlands5, have now been adopted internationally, by the World Health Organisation, International Diabetes Federation, National Institutes of Health and by the British Heart Foundation and Diabetes UK. They provide the key diagnostic criterion for metabolic syndrome.NICE has reproduced a complicated scheme for BMI and waist circumference6. This arose from a misapprehension that BMI measures total body fat, while waist measures intra-abdominal fat. A simplified and more logical approach was published recently in the BMJ. Given that we are seeking ways to identify patients at high risk earlier, to encourage improvements for disease prevention, it looks as if waist circumference alone may be a sufficient criterion to initiate active weight management. The GP contract currently advises measurement of height and weight (both individually valuable) and calculating BMI. To have full practice and population-level registers of BMI would certainly help us monitor the epidemic as prevalence rises steadily. It would be better to record waist circumference7. To get BMI, patients need to take shoes off, proxy measures of height are needed for the chair-bound, and some cannot get on to scales, or exceed the upper weight limit. Waist circumference can be measured in everyone, but it does require the waist (between the iliac crest and the lowest rib) to be revealed.
Why the reluctance to record waist circumference?
It is an extra burden on primary care, and confusion about its value will probably remain until today's new graduates reach senior positions as NHS policymakers. Measuring waist circumference should not replace measuring height or weight – and weight alone remains the easiest way to monitor obesity.
After 120 years the days of BMI (Adolphe Quetelet 1796-1874) are numbered. The age of the tape measure is with us and it will remain until we have ways to measure body fat more accurately (bio-impedance and other black-box methods are highly promoted but not as good as waist circumference in estimating body fat content).
Professor Mike Lean is professor of human nutrition, Division of Developmental Medicine, University of Glasgow, and was a founder of Counterweight, the national primary care weight management programmeCompeting interests None declared