Validity of Hand-to-Foot and Foot-to-Foot Consumer Bioimpedance Analyzers: A Four-Compartment Model Comparison

Authors

Sarah J. White, Texas Tech UniversityFollow
Madelin R. Siedler, Texas Tech UniversityFollow
Christian Rodriguez, Texas Tech UniversityFollow
Abegale D. Williams, Texas Tech UniversityFollow
Gabrielle DeHaven, Texas Tech UniversityFollow
Patrick S. Harty, Texas Tech UniversityFollow
Matthew T. Stratton, Texas Tech UniversityFollow
Dale Keith, Texas Tech UniversityFollow
Jacob Green, Texas Tech UniversityFollow
Jake Boykin, Texas Tech UniversityFollow
Grant M. Tinsley, Texas Tech UniversityFollow

Abstract

Body fat percentage (BF%) is a useful variable for predicting disease risk and determining overall fitness. Consumer-grade bioimpedance analyzers seek to provide accurate body composition data while remaining affordable and accessible. PURPOSE: The purpose of this study was to compare body fat percentages obtained from hand-to-foot and foot-to-foot consumer bioimpedance analyzers to a gold standard 4-compartment (4C) model. METHODS: Seventy-five adults (40 F, 35 M; age: 27.2 ± 7.3 y; height: 168.1 ± 8.8 cm; BM: 72.1 ± 16.6 kg; 4C model BF%: 25.0 ± 9.2%) were evaluated by a 4C model, a consumer-grade hand-to-foot bioimpedance analyzer (BIA-HF; Tanita BC568) and two consumer-grade foot-to-foot bioimpedance analyzers (BIA-FF; Tanita BC554 and Tanita UM081). The 4C model comprised dual-energy X-ray absorptiometry, air displacement plethysmography, and bioimpedance spectroscopy. BF% estimates obtained by each bioimpedance analyzer were compared to the criterion 4C using the coefficient of determination (R²), standard error of the estimate (SEE), and Bland-Altman analysis. RESULTS: BIA-HF underestimated BF% by 1.4 ± 4.1%, and both BIA-FF overestimated BF% by 0.5 to 0.6 ± 5.7%. The R² value was higher for BIA-HF as compared to both BIA-FF analyzers (0.81 vs. 0.64). The SEE and 95% limits of agreement (LOA) were lower for BIA-HF (SEE: 4.0%; LOA: 8.1%) as compared to both BIA-FF (SEE: 5.6%; LOA: 11.2%). No method demonstrated proportional bias based on Bland-Altman analysis. CONCLUSION: While both hand-to-foot and foot-to-foot consumer-grade bioimpedance analyzers demonstrated potentially meaningful errors when compared to a gold standard method, the hand-to-foot device exhibited better overall performance. Specifically, a stronger linear agreement with the 4C model and lower individual-level errors were observed with the hand-to-foot model as compared to both foot-to-foot models from the same manufacturer. The superior performance of the hand-to-foot analyzer could be due to its direct testing of both the upper and lower body, which is more similar to the methods used in the 4C model and a better representation of an individual’s overall body composition.

Recommended Citation

White, Sarah J.; Siedler, Madelin R.; Rodriguez, Christian; Williams, Abegale D.; DeHaven, Gabrielle; Harty, Patrick S.; Stratton, Matthew T.; Keith, Dale; Green, Jacob; Boykin, Jake; and Tinsley, Grant M. (2022) "Validity of Hand-to-Foot and Foot-to-Foot Consumer Bioimpedance Analyzers: A Four-Compartment Model Comparison," International Journal of Exercise Science: Conference Proceedings: Vol. 2: Iss. 14, Article 3.
Available at: https://digitalcommons.wku.edu/ijesab/vol2/iss14/3