Impact of Measured vs. Predicted Residual Lung Volume on Body Fat Percentage Using Underwater Weighing and 4-Compartment Model. Issue 9 (September 2017)

• Record Type: Journal Article
• Authors: Nickerson, Brett S.; Esco, Michael R.; Bishop, Phillip A.; Schumacker, Randall E.; Richardson, Mark T.; Fedewa, Michael V.; Wingo, Jonathan E.; Welborn, Bailey A.
• Is Part Of: Journal of strength and conditioning research. Volume 31:Issue 9(2017)
• Publication Date: 2017-09
• Subjects: multicompartment model -- hydrostatic weighing -- percent fat -- body density -- adiposity; Physical education and training -- Periodicals; Weight training -- Physiological aspects -- Periodicals; Physical fitness -- Periodicals; 613.7
• DOI: 10.1519/JSC.0000000000001698
• Languages: English
• ISSNs: 1064-8011

Abstract

Abstract : Abstract: Nickerson, BS, Esco, MR, Bishop, PA, Schumacker, RE, Richardson, MT, Fedewa, MV, Wingo, JE, and Welborn, BA. Impact of measured vs. predicted residual lung volume on body fat percentage using underwater weighing and 4-compartment model. J Strength Cond Res 31(9): 2519–2527, 2017—The purpose of this study was to compare underwater weighing (UWW) and 4-compartment (4C) model body fat percentage (BF%) for predicted vs. simultaneously measured residual lung volume (RLV). Forty-seven women and 33 men (age = 22 ± 5 years) had UWW and 4C model BF% determined using Boren et al. (RLVBOREN ), Goldman and Becklake (RLVGB ), and Miller et al. (RLVMILLER ) RLV prediction equations. Criterion UWW BF% included body density (BD) values with simultaneous RLV. Criterion 4C model BF% included BD through UWW with simultaneous RLV, total body water through bioimpedance spectroscopy, and bone mineral content through dual-energy x-ray absorptiometry. The standard error of estimate (SEE) for UWW and 4C model BF% determined through RLV prediction equations varied from 2.0 to 2.6% and from 1.3 to 1.5%, respectively. The constant error (CE) was significantly different for UWW BF% when using RLVBOREN, RLVGB, and RLVMILLER (all p < 0.016; CE = 0.7, −2.0, 1.0%, respectively). However, the CEs for RLVBOREN and RLVMILLER were not significant in the 4C model ( p = 0.73 and 0.11; CE = 0.1 and 0.2%, respectively), whereas RLVGB remained significantly different ( p < 0.001; CE = −1.5%). The 95% limits of agreement were less than ±5.2% for UWW BF% and less than ±3.1% for the 4C model when using the 3 RLV equations. When used in a 4C model, the RLV equations yielded a smaller CE, SEE, and 95% limits of agreement than UWW BF% results. However, because of the range of individual error shown in the current study, caution should be employed when using the 4C model as a criterion method with predicted RLV.