ABSTRACT
Obesity has been defined traditionally as a level of excess body fat associated with health risk. There is increasing evidence that central body fat distribution, independent of total body fat, is associated with cardiovascular risk factors in childhood.1,2 This suggests that, ideally, both the amount of excess body fat and its distribution should be considered in the assessment of obesity in children. However, in clinical practice, assessment of body fatness and/or fat distribution is usually impractical (see ‘Measurement of body fat’) and so simpler proxy measures are required. Subjective (clinical) diagnosis of childhood obesity is unreliable, even when carried out by experienced observers and so obesity must be defined using objective body measurements (anthropometry). Body weight alone is inadequate as an index of obesity because of its association with height, so weight must be adjusted for height in some way. Several approaches to this adjustment for height are in use, but there is now an international consensus that the body mass index (BMI), weight (kg)/height2 (m2), is the most practical and clinically meaningful.3-5
The BMI cannot be used in the same way as in adults for a number of reasons. First, BMI values of children/ adolescents are lower than those of adults. Second, BMI changes with age and differs between the sexes. This means that interpreting a BMI value requires comparison
with population reference data. Good, accessible, population reference data now exist for many countries including the UK5 and the USA.6 These are usually available as centile charts on which BMI values can be plotted to assess obesity or to monitor changes over time.
