Air Displacement Plethysmography Accurately Determines Body Composition in Overweight Adults Compared to Criterion 4C Model

Abstract

PURPOSE: In comparison to body composition assessment methods such as hydrostatic weighing (HW), air displacement plethysmography (ADP) requires less technical expertise of investigators and provides more convenience and comfort for subjects. In this regard, ADP could be of significant benefit for body composition assessment among children, elderly, overweight, or unfit individuals for whom HW may pose a challenge. Although the reliability and validity of ADP has been compared to other two-compartment models in various populations, it has not been validated against the criterion four-compartment (4C) model of body composition among overweight adults. Among normal weight subjects, two previous studies comparing ADP to the 4C model concluded that percent body fat (BF%) by ADP was significantly lower than 4C estimates. Therefore, the purpose of this study was to assess the validity of ADP using the criterion 4C model of body composition among overweight adults. A secondary aim was to compare the measures of body density (Db) derived from HW and ADP in order to determine if the methods are interchangeable. METHODS: Body composition was assessed among 35 overweight males and females (BMI 27.9 ± 2.2; Age 32.7 ± 7.2 y) using ADP and Selinger's 4C equation (total body water from 18O isotope dilution; body mineral from dual energy x-ray absorptiometry; and body density from HW). Mean differences in BF% and in Db were compared with t-tests. Regression analysis was used to describe the relationship between BF% determined by the criterion 4C and ADP, and between 4C models with Db from HW and Db from ADP. Bland Altman plots were used to examine bias in the comparisons. RESULTS: BF% was not significantly different between ADP and 4C (0.1 ± 2.6, P = 0.9) and regression analysis showed that 88% of the between-subject variance in the 4C BF% was explained by ADP (R2 = 0.88, SEE 2.6, P <.0001). The regression line was not different from 1 (P = 0.3) and the intercept was not different from 0 (P = 0.3). There was no significant difference between ADP and HW in the determination of Db (−0.002 ± 0.006 kg/L, P = 0.6) or in BF% when Db from ADP and HW was utilized in the 4C equation (0.5 ± 1.6, P = 0.7). Regression analysis showed a strong relationship between the two estimates of BF% using the 4C equation with Db from ADP and HW (R2 = 0.95, SEE 1.6, P < .0001). The regression line was not different from 1 (P=0.1) and the intercept was not different from 0 (P = 0.3). CONCLUSIONS: ADP provides a valid estimate of body composition in overweight adults when compared to the criterion 4C model. Further, there is no difference in body fat percentage estimates when Db from ADP replaces Db from HW in the 4C equation.