Effect of mode selection when using contact-electrode bioelectrical impedance analyzers to estimate percent body fat...

When using bioelectrical impedance analysis (BIA) analyzers to assess body composition, a computer-programmed prediction equation referred to as a mode (i.e., standard STD or athletic ATH) must be selected prior to the assessment. This study examined the effect of mode selection on percent body fat (%BF) determined by leg-to-leg and segmental bioelectrical impedance analysis (LBIA; SBIA) in young adults.

Ninety-two young adults (52 women; 40 men) had %BF estimated using LBIA (Tanita: TBF-300A) and SBIA (Tanita: BC-418) which was then compared to air displacement plethysmography (ADP). During the BIA assessment, %BF was determined using both the STD and ATH pre-programmed modes.

In the women, %BF (mean±SD) was significantly (P<0.001) underestimated by the ATH modes of LBIA (23.0±6.9%) and SBIA (23.4±6.2%) when compared to ADP (25.5±7.3%). In the men, the STD mode of LBIA (18.3±5%) overestimated %BF when compared to ADP (14.1±7%). All standard error of estimate and pure error values (range=3.8% to 6.3%) exceeded the recommended range for accuracy (<3.5%).

The STD mode produced mean %BF values most similar to ADP in the women, whereas the ATH mode did so in the men; a finding that was consistent for both analyzers. However, large individual prediction errors and wide confidence intervals were observed regardless of the mode selected or analyzer used for the assessment. As such, when precision is critical, the prediction equations programmed in the LBIA and SBIA analyzers examined in this study cannot be recommended to measure %BF on an individual basis.