Agreement of Bioelectrical Impedance Analysis Prediction Equations with Dual-Energy X-Ray Absorptiometry for Tracking Resistance Training-Induced Changes in Fat-Free Mass in Older Women

This study examined the agreement between commonly used bioelectrical impedance analysis (BIA) prediction equations and dual-energy X-ray absorptiometry (DXA) for tracking resistance training-induced changes in fat-free mass (FFM) in older women. One hundred and five older women (68.4 ± 5.8 years) completed a 24-week supervised resistance training program. FFM was assessed before and after training using DXA and estimated using eight BIA prediction equations. DXA-derived FFM increased significantly after training (Δ = 0.9 ± 2.3 kg). All BIA equations also detected significant increases in FFM, with mean changes similar to those observed with DXA after multiplicity adjustment (p > 0.05). However, this similarity at the group level did not translate into agreement at the individual level. Correlations between BIA equation-derived and DXA-derived changes were weak to moderate (r = 0.26–0.42), and despite generally small mean biases (−0.25 to 0.59 kg), the 95% limits of agreement were wide across all equations. Moreover, significant negative proportional bias was observed for all equations, indicating progressively greater underestimation relative to DXA as the magnitude of FFM change increased. Collectively, these findings demonstrate a clear distinction between the ability of BIA prediction equations to reproduce mean group-level changes and their ability to track individual changes in FFM. Thus, although BIA prediction equations may be useful for describing average group-level adaptations to resistance training in older women, they should not be considered interchangeable with DXA for monitoring individual changes in FFM.