Background – Bioelectrical impedance analysis (BIA) has been commonly used as a convenient, cost effective way to measure body composition in large population studies. However, the validity of this technique remains uncertain. Objective – To compare measurements of body fat mass using single frequency BIA with dual-energy X-ray absorptiometry (DEXA).
2
Design – Forty-two overweight and obese (body mass index 25 to 40 kg/m ) but otherwise healthy volunteers aged
30 to 70 y were recruited from the general population. In a cross sectional analysis, fat mass was measured using foot-to-foot BIA (Tanita Scale Model TBF-300) and whole body DEXA (GE Lunar Prodigy scanner).
Outcomes – Although fat mass measured using BIA was strongly correlated with DEXA (r2=0.89), this analysis is not useful when comparing methods of measurement. There was bias in the measurement of fat mass with BIA measuring lower than DEXA (mean±SD: -2.96±2.91 kg). In weighted least products regression there was a significant fixed bias (mean (95% CI): -6.24 (-10.75, -1.72). However, there was no significant proportional bias (1.09 (0.96, 1.23)). That is, the slope did not differ significantly from 1. This implies that the difference between the methods did not increase with increasing fat mass.
Conclusions – Although the results suggest that there may be some concern regarding accuracy of the BIA technique in overweight and obese individuals, this method may be useful for estimating of fat mass in large cross- sectional and longitudinal population studies, as well as in intervention studies.