Predictive accuracy of Omron((R)) body logic analyzer in estimating relative body fat of adults

This study assessed the predictive accuracy of a new hand-held, segmental, bioimpedance (BI) analyzer in estimating the relative body fat (%BF) of a s ample of 25 men and 23 women (18 - 55 years, 7.0 to 42.8%BFHW). The referen ce method was hydrostatic weighing (HW) at residual lung volume. The %BF es timates obtained from manufacturer's (Omron) gender-specific equations were cross-validated. There were high validity coefficients (r(y,y') = .91 and .83, for men and women, respectively), moderate prediction errors (SEE = 3. 46%, E = 3.64%BF for men; SEE = 4.04%, E = 3.87%BF for women), and no signi ficant difference (p > .05) between the average %BFHW and %BFOmron for wome n (21.8% vs. 21.6%BF, respectively). For men, there was a small but signifi cant (p < .05) difference in %BFHW (18.7%) and %BFOmron (20.1%). For both m en and women, the line of identity did not differ significantly (p > .05) f rom the line of best fit. The Omron(R) BI equations accurately estimated th e %BF of 72% of the men and 65% of the women in this sample within +/- 3.5% BF. Therefore, use of the Omron(R) BI analyzer is suitable for assessing th e %BF of adults having demographic characteristics similar to those of this sample.