Background/objectives: Estimating body composition using bioelectric impedance analysis (BIA) requires specific prediction equations. The purpose of our study was to examine the validity of published BIA equations for assessing total body water (TBW) and fat-free mass (FFM) using deuterium oxide dilution (D2O) as a reference method and to develop new FFM and TBW BIA equations if needed for Moroccan pre-pubertal children.
Subjects/methods: Data were collected from 247 schoolchildren aged 8-11 years old. Children were sorted by gender and age group and assigned in development and validation groups. D2O dilution was used as reference method for estimating TBW and FFM. Bland and Altman test, effect size, pure error, and proportional bias were used to assess the reliability of previous published equations. Cross-validation was performed by Bland and Altman test and BIA new equations were developed by linear regression.
Results: Previously published equations were tested and showed significant bias values indicating that if used they would provide biased values of TBW and FFM. The new prediction equations developed were: [Formula: see text] (l) = 0.269 + 0.292 Ht2/R (cm2/Ω) + 0.221 weight (kg) + 0.824 sex (boys = 1, girls = 0) + 0.291 age (years) (R2 = 0.91, RMSE = 1.54 kg); [Formula: see text] (kg) = -0.450 + 0.380 Ht2/R (cm2/Ω) + 0.291 weight (kg) + 1.294 sex (boys = 1, girls = 0) + 0.446 age (years) (R2 = 0.91, RMSE = 1.97 kg). These equations provide better values of proportional bias, agreement, and pure error than the other tested equations.
Conclusions: The new BIA prediction equations seem to be the most accurate for Moroccan pre-pubertal children.