Fat-free mass predictive equation using multifrequency bioelectrical impedance data in adolescent soccer athletes: development and cross-validation

Abstract

Objectives: This study aimed to develop and cross-validate a fat-free mass (FFM) predictive equation using multifrequency bioelectrical impedance data (MF-BIA) in adolescent soccer athletes. Methods: Male adolescent soccer athletes (n=149; 13 to 19 years old) were randomly sorted using Excel and independently selected for development group (n=100) or cross-validation group (n=49). The FFM reference values were determined using dual-energy X-ray absorptiometry (DXA). Single-frequency BIA was used to plot tolerance ellipses. Multifrequency-BIA raw data were used as independent variables in regression models. Student's independent t-test was used to compare development and cross-validation groups. Stepwise multiple regression was used to develop the FFM predictive equation. Bland-Altman plots, Lin's concordance correlation coefficient (CCC), according to McBride criteria, precision, accuracy, and standard error of estimation (SEE) were calculated to evaluate the concordance and reliability of estimates. Bioelectrical impedance vector analysis (BIVA) was plotted to evaluate hydration status. Results: No differences (P>0.05) were observed between development and validation groups in chronological age, anthropometric data, bioelectrical impedance data, and FFM values obtained using DXA. BIVA tolerance showed that all participants presented adequate hydration status when compared with the reference population. The new FFM predictive equation developed and validated: FFM (kg)= 7.064+ 0.592 * Age (years)+ 0.554* Weight(kg) + 0.365 * Height²/Resistance (cm²/Ω), presented R²= 0.95; SEE= 1.76 kg; CCC= 0.95, accuracy= 0.98, and strength of concordance= 0.99. Conclusions: The present study developed and cross-validated an FFM predictive equation based on multifrequency bioelectrical data providing substantial FFM accuracy for male adolescent soccer athletes.