When Formulas Fail: A Secondary Analysis of Dual-Energy X-ray Absorptiometry (DXA) Body Composition Data Comparing Ideal and Lean Body Weight Equations Against Measured Lean Body Weight in Obese Children.
Mike Wells, Lara N Goldstein
Abstract
Open AccessBackground Accurate weight-based drug dosing in obese children is essential for safe and effective therapy, yet the relationships between ideal body weight (IBW), lean body weight (LBW), and actual body composition remain poorly defined. IBW was originally developed to assess nutritional status rather than guide drug dosing, but it is often used as a surrogate for LBW despite limited validation. This study aimed to compare commonly used IBW and LBW formulas against measured LBW to evaluate their accuracy and clinical applicability in obese children. Methods A secondary analysis was conducted using data from 30 obese children (BMI-for-age Z-score ≥2.0) obtained in a hospital-based pediatric body composition study. Anthropometric data and dual-energy X-ray absorptiometry (DXA) measurements of LBW were available. Seven published equations for IBW and LBW estimation were evaluated. Accuracy was assessed using mean percentage error, root mean square percentage error, and proportions of estimates within 10% (P10) and 20% (P20) of measured LBW. Results All IBW formulas substantially overestimated measured LBW, with mean biases ranging from 18% to 30%, and none achieved acceptable accuracy (P10 <70%). Among LBW equations, the Foster formula demonstrated the highest accuracy, while the Peters and Callaghan methods performed poorly. Adjusting IBW formulas with simple correction factors improved LBW prediction accuracy, though similar corrections were not feasible for the Peters and Callaghan methods. Conclusions IBW formulas cannot reliably substitute for LBW in obese children without modification. The Foster equation offers the best accuracy for LBW estimation, while the BMI₅₀ IBW method shows potential for adaptation in clinical tools. Further validation of hybrid or adjusted models is required to optimize pediatric drug dosing safety.