ACCURACY OF ESTIMATING HUMAN BODY COMPOSITION CHANGES ON BI METHODS -WITH THE STUDY OF SERIAL MEASUREMENTS DURING THE WEIGHT / 体力科学

The purpose of this study was to investigate the accuracy of estimating human body composition changes using bioelectrical impedance (BI) methods during a weight-loss intervention. Subjects were forty-three obese men (age : 49.2±10.5 yr, BMI : 27.8±1.7 kg/m²) who completed a 14-week weight-loss intervention. In all subjects, fat mass (FM) and fat-free mass (FFM) were assessed by dual energy x-ray absorptiometry (DXA) as well as single- and multi-frequency BI methods (SBIM, MBIM) before and after the intervention. Resistance parameters were measured by SBIM and MBIM (SBIM : R₅₀ ; MBIM : R∞, R0, and Rfc). In nine subjects these variables were also measured at weeks 1 and 4. Weight decreased (P<0.05) by -8.0±3.2 kg during the intervention while FFM changes averaged -0.4±1.6 kg (DXA), -2.0±1.5 kg (SBIM), and -1.6±1.7 kg (MBIM). BI methods overestimated FFM before the intervention (before ; DXA : 54.4±4.8 kg, SBIM : 56.5±4.3 kg, MBIM : 55.9±4.5 kg). In nine subjects, FFM measured by SBIM (FFM_SBIM) and MBIM (FFM_MBIM) was similar to FFM measured by DXA(FFM_DXA)(after ; DXA : 54.6±5.4 kg, SBIM : 54.6±3.8 kg, MBIM : 54.6±4.1 kg), although BI methods overestimated the FFM before the intervention (before ; DXA : 54.9±5.1 kg, SBIM : 56.9±3.8 kg, MBIM : 56.3±4.4 kg). The ΔFM_SBIM and ΔFM_MBIM were highly correlated with the ΔFM_DXA(SBIM : r=0.87, MBIM : r=0.88). The ΔFFM_SBIM andΔFFM_MBIM were significantly correlated with the ΔFFM_DXA(SBIM : r=0.54, MBIM : r=0.49). The ΔR₅₀ and ΔRfc were also significantly correlated with the ΔFFM_DXA(R₅₀ : r=-0.63, Rfc : r=-0.48). These results suggest that during a weight-loss intervention, 1) BI methods and DXA provide similar estimates of human body composition change, although they overestimate FFM in obese men, and 2) changes of resistance parameters observed with BI methods may estimate human body composition change more accurately.