Agreement between segmental multifrequency bioelectrical impedance analysis and anthropometric methods for estimating adiposity and muscle mass in amateur adult soccer players

Abstract

Objective: To analyze the differences and agreement between segmental multifrequency bioelectrical impedance analysis (BIA) and anthropometric methods for estimating adiposity and muscle mass in amateur adult soccer players. Methods: A cross-sectional observational study was conducted in 27 soccer players with paired BIA and anthropometric measurements. For adiposity, BIA was compared with Durnin-Siri, Faulkner, Yuhasz, and adipose mass estimated using Kerr’s five-compartment model. For muscle mass, BIA was compared with Kerr, Lee, and Martin methods. Paired differences, bias, limits of agreement, intraclass correlation, mean errors, and proportional bias slope were calculated. Results: For adiposity, Durnin-Siri was the closest method to BIA, with no significant differences. Faulkner and Yuhasz provided lower values than BIA, whereas Kerr adipose mass showed higher values. For muscle mass, Kerr did not differ significantly from BIA; Lee yielded lower values and Martin higher values. Conclusion: BIA and anthropometric methods are not directly interchangeable in this population. Although Durnin-Siri and Kerr showed greater mean proximity to BIA, individual variability and proportional bias require cautious interpretation and methodological consistency during follow-up.