Calibration of bioelectrical impedance analysis against deuterium dilution for body composition assessment in stunted Ugandan children

Background: BIA represents an important tool in body composition (BC) assessment, especially in low-income settings in which simple and affordable options are preferred. There is a particular need to measure BC in stunted children, in which cases population-specific BIA estimating equations are lacking. Objective: We calibrated an equation to estimate body composition from BIA using deuterium dilution (2H) as the criterion method in stunted children. Methods: We measured BC with 2H and performed BIA in stunted Ugandan children (n = 50). Multiple linear regression models were constructed to predict 2H-derived FFM from BIA-derived whole-body impedance and other relevant predictors. Model performance was expressed as adjusted R2 and RMSE. Prediction errors were also calculated. Results: Participants were aged 16-59 mo, at whom 46% were girls, and their median [IQR] height-for-age z-score (HAZ) was -2.58 [-2.92, -2.37] according to WHO growth standards. Impedance index (height2/impendance measured at 50 kHz) alone explained 89.2% variation in FFM and had an RMSE of 583 g (precision error 6.5%). The final model contained age, sex, impedance index and height-for-age z-score as predictors, and explained 94.5% variation in FFM with an RMSE of 402 g (precision error 4.5%). Conclusion: We present a BIA calibration equation for a group of stunted children with relatively low prediction error. This may help evaluate the efficacy of "nutritional supplementation in large-scale trials in the same population". J Nutr 20xx;xxx:xx.