Association between anthropometric indices and body fat for identifying excess body fat in elementary school children: a population-based cross-sectional study

TL;DR

This study finds that BMI and obesity measures are effective in identifying excess body fat in children.

Contribution

The study evaluates multiple anthropometric indices for identifying excess body fat in children using comprehensive statistical methods.

Findings

BMI and degree of obesity showed strong association with body fat percentage across key percentiles.

ROC and PR curve analyses indicated high discriminatory ability of BMI and obesity measures.

Classification metrics like precision and recall exceeded 70% for most indices.

Abstract

Identifying and managing obesity in children is essential to prevent obesity-related diseases in adulthood. This study aimed to evaluate the association between body mass index (BMI), degree of obesity, waist circumference, waist-to-height ratio, and body fat—particularly excess body fat.

Participants included 660 children aged 9–12 years (349 boys and 311 girls). Fat mass, fat-free mass, and body fat percentage were assessed using bioelectrical impedance analysis. The discriminatory ability of BMI, degree of obesity, waist circumference, and waist-to-height ratio to identify excess body fat—defined as body fat percentage exceeding the 85th, 90th, or 95th percentile—was evaluated using receiver operating characteristic (ROC) curve and precision-recall (PR) curve analyses. Classification performance was further evaluated using a confusion matrix, accuracy, precision, recall, F1 score, Cohen’s kappa coefficient, and Matthews correlation coefficient (MCC).

The areas under the ROC curve (AUCs) and 95% confidence intervals (CIs) for BMI, degree of obesity, waist circumference, and waist-to-height ratio in identifying obesity based on body fat percentage were > 0.9 in both sexes in most cases. PR AUCs and 95% CIs for BMI and degree of obesity were ≥ 0.8 in most cases. Precision, recall, and F1 scores for BMI and degree of obesity in identifying obesity at the 85th or 95th percentiles were > 70% in nearly all cases. Kappa coefficients indicated substantial agreement between BMI and the 85th or 90th percentiles of body fat percentage, and moderate agreement for the degree of obesity. The MCC index showed a pattern similar to that of the kappa coefficients.

These findings suggest that BMI and degree of obesity are strongly associated with body fat percentage across the 85th, 90th, and 95th percentiles and that obesity classifications based on BMI as well as degree of obesity align closely with those based on body fat percentage.

The online version contains supplementary material available at 10.1186/s40101-025-00410-w.

Figures