Abstract
Background. The level of body fat mass (BFM) in childhood that is associated with weight related morbidity is unclear. Bioelectrical impedance analysis (BIA) offers an inexpensive, acceptable and portable method for measuring body composition in children. However, different equations have been derived to estimate BFM, and relationships between equations have not been explored. Objective. To compare body mass index (BMI) and BIA as tools for estimating adiposity-related health risks in children. Methods. Height, weight and BIA were measured in a population based cross-sectional survey of 341 healthy 5-year-old Australian children. Percent BFM was estimated using four published BIA-based predictive equations for pre-school children. Ranking of children according to total BFM was compared for all equations. Results. Each equation produced different estimates of percent BFM. In general, increasing BMI was associated with increasing BFM, but wide ranges of BFM estimates were produced for children of similar BMI. For all of the equations, females had a higher percent BFM compared with males of the same BMI (p <0.001). Percent BFM estimates rose rapidly in children classified as overweight/obese (1990 UK growth standard). The equations were highly correlated in their ranking of children from lowest to highest percent BFM. Conclusion. Results support concerns about the validity of BMI as an accurate measure of absolute BFM. Percent BFM estimates produced by the four BIA equations were highly correlated, indicating they rank children according to BFM in the same order. This suggests any single equation could provide a measure of relative BFM in children for population and longitudinal studies.
Acknowledgements
The PEAS Kids Growth Study was supported by grants from the Murdoch Childrens Research Institute and the National Health and Medical Research Council (NHMRC). Michele Campbell was funded by postgraduate research scholarships from the Murdoch Childrens Research Institute and the NHMRC. The Murdoch Childrens Research Institute and the NHMRC had no role in the study design, data collection, analysis, interpretation or writing of the article.