Comparison of generalized and athletic bioimpedance-based predictive equations for estimating fat-free mass in resistance-trained exercisers

Objectives: This study aimed to test whether athlete-specific, bioelectrical, impedance-based equations to estimate fat-free mass (FFM) could be more accurate than generalized equations when testing resistance-trained exercisers. Methods: A total of 50 resistance-trained men (age 30.9 ± 7.4 y; body mass index: 25.3 ± 2.2 kg/m2) and 20 men from the general population (age 29.9 ± 9.1 y; body mass index: 22.8 ± 2.4 kg/m2) underwent bioelectrical impedance and dual-energy x-ray absorptiometry (DXA) evaluations. FFM was derived by one bioelectrical impedance-based equation specific for athletes and three generalized equations, all developed with foot-to-hand bioimpedance technologies at a 50 kHz frequency. DXA was the reference method for the FFM assessment. Results: Compared with DXA, when assessing the resistance-trained participants, the athletic-specific equation had neither mean (–0.89 kg; P = 0.789) or proportional bias (r = –0.104; P = 0.474) with a coefficient of determination equal to R2 = 0.91. In contrast, the three generalized predictive equations overestimated FFM (range, 4.11–5.37 kg; P < 0.05) with R2 ranging from 0.84 to 0.90. The athletic-specific equation underestimated FFM in the general population participants (–2.93 kg; P < 0.05). Conclusions: When assessing body composition in resistance-trained exercisers, specific equations for athletes should be preferred to generalized ones to avoid an overestimation in FFM. Furthermore, athlete-specific and generalized formulas cannot be used interchangeably, even when assessing body composition in the general population.