Objective: To test the hypothesis that variations in trunk circumferences influence the accuracy of bioimpedance analysis (BIA) for assessment of percent fat mass (%FM).
Subjects and methods: %FM was predicted with BIA, and compared with air-displacement plethysmography (ADP) in a small sample of 35 overweight (OW), 21 normal weight and 8 underweight volunteers. Waist and hip circumferences were assessed, and 15 of the OW subjects were measured before and after weight reduction.
Results: BIA and ADP provided similar cross-sectional estimates of group mean %FM (28.9±10.0 and 31.3±13.0%, respectively). However, within individuals, there were large between-method differences (Diff(BIA-ADP)) ranging from -13 to +13 %FM. Furthermore, we found a systematic bias of BIA related to the degree of adiposity. Consequently, %FM and fat mass loss during weight reduction in OW were underestimated with BIA when compared with ADP. Waist and hip circumferences were inversely associated with resistance (R) and reactance (P<0.01), and with Diff(BIA-ADP) (P<0.001). In women, the variability in hip circumference explained 76%, and in men, the variability in waist circumference explained 59% of Diff(BIA-ADP).
Conclusion: Resistance changes associated with variations in trunk circumferences decrease resistance, and therefore impair the accuracy of BIA to assess %FM.