Abstract
The purpose of this study was to compare estimates of human body composition determined from single-frequency bioelectrical impedance methods (S-BIM) and multi-frequency bioelectrical impedance methods (M-BIM) . The human body composition was assessed by dual energy X-ray absorptiometry (DEXA), 5 brands of S-BIM, and 2 brands of M-BIM. Forty-five women, aged 26-58 years, served as subjects. The S-BIM and M-BIM fat-free mass (FFM) estimates were highly correlated with the FFM measured by DEXA (r=0.82-0.93) . The standard errors of estimate (SEE) for FFM were approximately 2 kg. With the exception of the MLT-100 (which slightly underestimated FFM), all brands of BIM slightly overestimated FFM. The absolute mean differences between FFMDEXA and each of the 7 BIM estimates ranged from -3.02 kg to 3.46 kg. Although the 7 brands of BIM provided slightly different estimates, the results of this study suggest that 5-BIM and M-BIM are relatively valid in human body composition.
