Fulltext

A composite score combining waist circumference and body mass index more accurately predicts body fat percentage in 6- to 13-year-old children

Aeberli, I. ; Gut-Knabenhans, M. ; Kusche-Ammann, R. ; Molinari, L. ; Zimmermann, M.

In: European Journal of Nutrition, 2013, vol. 52, no. 1, p. 247-253

Add to personal list
    Title
    • A composite score combining waist circumference and body mass index more accurately predicts body fat percentage in 6- to 13-year-old children
    Author
    • Aeberli, I.. Institute of Food, Nutrition and Health, Human Nutrition Laboratory, ETH Zurich, LFV D22, Schmelzbergstrasse 7, 8092, Zürich, Switzerland
    • Gut-Knabenhans, M.. Institute of Food, Nutrition and Health, Human Nutrition Laboratory, ETH Zurich, LFV D22, Schmelzbergstrasse 7, 8092, Zürich, Switzerland
    • Kusche-Ammann, R.. Institute of Food, Nutrition and Health, Human Nutrition Laboratory, ETH Zurich, LFV D22, Schmelzbergstrasse 7, 8092, Zürich, Switzerland
    • Molinari, L.. Child Development Center, University Children's Hospital, Zurich, Switzerland
    • Zimmermann, M.. Institute of Food, Nutrition and Health, Human Nutrition Laboratory, ETH Zurich, LFV D22, Schmelzbergstrasse 7, 8092, Zürich, Switzerland
    Document Type
    • Postprint
    OAI-PMH Identifier
    • oai:doc.rero.ch:322491
    Summary
    Purpose: Body mass index (BMI) and waist circumference (WC) are widely used to predict % body fat (BF) and classify degrees of pediatric adiposity. However, both measures have limitations. The aim of this study was to evaluate whether a combination of WC and BMI would more accurately predict %BF than either alone. Methods: In a nationally representative sample of 2,303 6- to 13-year-old Swiss children, weight, height, and WC were measured, and %BF was determined from multiple skinfold thicknesses. Regression and receiver operating characteristic (ROC) curves were used to evaluate the combination of WC and BMI in predicting %BF against WC or BMI alone. An optimized composite score (CS) was generated. Results: A quadratic polynomial combination of WC and BMI led to a better prediction of %BF (r 2=0.68) compared with the two measures alone (r 2=0.58-0.62). The areas under the ROC curve for the CS [0.6*WC-SDS+0.4*BMI-SDS] ranged from 0.962±0.0053 (overweight girls) to 0.982±0.0046 (obese boys) and were somewhat greater than the AUCs for either BMI or WC alone. At a given specificity, the sensitivity of the prediction of overweight and obesity based on the CS was higher than that based on either WC or BMI alone, although the improvement was small. Conclusion: Both BMI and WC are good predictors of %BF in primary school children. However, a composite score incorporating both measures increased sensitivity at a constant specificity as compared to the individual measures. It may therefore be a useful tool for clinical and epidemiological studies of pediatric adiposity