Development and validation of new bioelectrical impedance equations to accurately estimate fat mass percentage in a heterogeneous Caucasian population

Abstract

Fat mass percentage (%FM) is frequently determined by nutritionists and personal trainers with bioelectrical impedance analysis (BIA) devices. The aims of the present study were: (1) to develop new regression equations using dual-energy X-ray absorptiometry (DXA) as the reference method for estimating %FM in a heterogeneous Caucasian population with a hand-to-hand device (BIA-101) and a foot-to-hand device (BIA-TELELAB) and (2) to compare the new equations with the manufacturers’ equations. We hypothesized that the new equa- tions would lead to more accurate estimations compared with DXA. A total of 218 healthy Caucasian participants aged 18 to 65 years were divided into a development group and a val- idation group. The accuracy of the different equations was assessed by mean differences, coefficient of determination, standard error of the estimate (SEE), intraclass correlation co- efficients (ICC), and Bland–Altman plots. The proposed equation for BIA-101 explained 90.0% of the variance in the DXA-derived %FM, with a low random error (SEE = 2.98%), excellent agreement (ICC = 0.94), no fixed bias, and relatively low individual variability (5.86%). For BIA-TELELAB, the proposed equation explained 88.0% of the variance in the DXA-derived %FM, with a low random error (SEE = 3.27%), excellent agreement (ICC = 0.93), no fixed bias, and relatively low individual variability (6.37%). The results obtained for the manufactur- ers’ equations confirm that these equations are not a good option for %FM assessment. As hypothesized, the new regression equations for BIA-101 and BIA-TELELAB devices can ac- curately estimate %FM in a heterogeneous Caucasian population with a broad age range.