Objective: To refine the diagnosis of childhood obesity by creating new sex-specific centile curves for body fat and to base these references on a simple and affordable method that could be widely adopted in clinical practice and surveys. Design: Body fat was measured by bio-impedance in 1985 Caucasian children aged 5-18 years from schools in Southern England. Smoothed centile charts were derived using the LMS method. Results: The new body fat curves reflect the known differences in the development of adiposity between boys and girls. The curves are similar by sex until puberty but then diverge markedly, with males proportionately decreasing body fat and females continuing to gain. These sex differences are not revealed by existing curves based on body mass index. We present charts in which cutoffs to define regions of 'underfat', 'normal', 'overfat' and 'obese' are set at the 2nd, 85th and 95th centiles. These have been designed to yield similar proportions of overweight/overfat and obese children to the IOTF body mass index cutoffs.Conclusions: Direct assessment of adiposity, the component of overweight that leads to pathology, represents a significant advance over body mass index. Our new charts will be published by the Child Growth Foundation for clinical monitoring of body fat, along with the software to convert individual measurements to Z-scores.

Objective: To develop waist circumference percentile curves for British children and to compare these curves with those from other countries. Design: Cross-sectional study. Setting: School-aged population. Subjects: A representative sample of school children from the geographical regions of Great Britain, approximately in proportion to their age distribution. The sample population consisted of 8355 children (3585 males, 4770 females) with ages ranging between 5.0 and 16.9 y. Interventions: Waist circumferences were measured with a flexible non-elastic tape and waist circumference percentiles were constructed and smoothed using the LMS method. Main outcome measures: Smoothed waist circumference percentile curves. Results: Mean waist circumference increased with age in both boys and girls. For girls, curves began to plateau after the age of 13 y whereas, for boys, waist percentile curves continued to increase more sharply after this age. However, these curves mainly reflect the patterns of waist circumference in Caucasian children. Conclusions: These curves represent the first waist circumference percentiles for British children and could be used provisionally for both clinical and possibly epidemiological use, although they should be validated against equivalent longitudinal data. Sponsorship: This project has been sponsored by UNL Diversity and Development Fund.

A study of central fatness using waist-to-height ratios in UK children and adolescents over two decades supports the simple message -'keep your waist circumference to less than half your height' Objective: To examine the influence of age and gender on the waist:height ratio (WHTR) in children and to compare changes over time in WHTR, a measure of central fatness in British children.

Objective To compare changes over time in waist circumference (a measure of central fatness) and body mass index (a measure of overall obesity) in British youth. Design Representative cross sectional surveys in 1977, 1987, and 1997. Setting Great Britain. Participants Young people aged 11-16 years surveyed in 1977 (boys) and 1987 (girls) for the British Standards Institute (n=3784) and in 1997 (both sexes) for the national diet and nutrition survey (n=776). Main outcome measures Waist circumference, expressed as a standard deviation score using the first survey as reference, and body mass index (weight(kg)/height(m)2 ), expressed as a standard deviation score against the British 1990 revised reference. Overweight and obesity were defined as the measurement exceeding the 91st and 98th centile, respectively. Results Waist circumference increased sharply over the period between surveys (mean increases for boys and girls, 6.9 and 6.2 cm, or 0.84 and 1.02 SD score units, P < 0.0001). In centile terms, waist circumference increased more in girls than in boys. Increases in body mass index were smaller and similar by sex (means 1.5 and 1.6, or 0.47 and 0.53 SD score units, P < 0.0001). Waist circumference in 1997 exceeded the 91st centile in 28% (n=110) of boys and 38% (n=147) of girls (against 9% for both sexes in 1977-87, P < 0.0001), whereas 14% (n=54) and 17% (n=68), respectively, exceeded the 98th centile (3% in 1977-87, P < 0.0001). The corresponding rates for body mass index in 1997 were 21% (n=80) of boys and 17% (n=67) of girls exceeding the 91st centile (8% and 6% in 1977-87) and 10% (n=39) and 8% (n=32) exceeding the 98th centile (3% and 2% in 1977-87). Conclusions Trends in waist circumference during the past 10-20 years have greatly exceeded those in body mass index, particularly in girls, showing that body mass index is a poor proxy for central fatness. Body mass index has therefore systematically underestimated the prevalence of obesity in young people.

BIA assessment of %FFM and SMMa represents a significant advance in nutritional assessment since these body composition components are associated with metabolic health. Muscle-to-fat ratio has the potential to provide a better index of future metabolic health.