Catarina N. Matias, Diana A. Santos, Pedro B. Júdice, Joao P. Magalhaes,Claudia S. Minderico, David A. Fields, Henry C. Lukaski, Luís B. Sardinha,Analiza M. Silva.
Clinical Nutrition 35 (2016) 468-474
Background: This study aimed: a) to develop BIA-based models for TBW and ECW prediction based on dilution methods in a sample of national level athletes; and b) to validate the new models with a cross-validation approach in a separate cohort using dilution methods as criterion.
Subjects and Methods:Two hundred and eight highly trained athletes (21.3 ± 5.0 years) were evaluated during their respective competitive seasons. Athletes were randomly split into development (n ¼ 139) and validation groups (n ¼ 69). The criterion method for TBW was deuterium dilution and for ECW was bromide dilution, where ICW was the respective difference between both. Resistance (R) and reactance (Xc) were obtained with a phase-sensitive 50 kHz BIA device and used for the estimation of TBW and ECW.
Result and Conclusion:Athletic BIA-based models were developed for TBW and ECW [TBW ¼ 0.286 þ 0.195*S2/ R þ 0.385*Wt þ 5.086*Sex; ECW ¼ 1.579 þ 0.055*S2/R þ 0.127*Wt þ 0.006*S2/Xc þ 0.932*Sex, where sex is 0 if female or 1 if male, Wt is weight (kg), S is stature (cm), and R and Xc are in ohm (U)]. Cross validation revealed R2 of 0.91 for TBW and R2 0.70 for ECW and no mean bias.The new equations can be considered valid, with no observed bias, thus affording practical means to quantify TBW and ECW in national level athletes.