Utility of multifrequency bioelectrical impedance compared with dual-energy x-ray absorptiometry for assessment of total and regional body composition varies between men and women Lindsey J. Anderson, David N. Erceg, E. Todd Schroeder ⁎ Division of Biokinesiology and Physical Therapy at the School of Dentistry, The Clinical Exercise Research Center, University of Southern California, ET Schroeder, 1540 E. Alcazar CHP-155, Los Angeles, CA 90033, USA ARTICLE INFO ABSTRACT Article history: Received 31 January 2012 Revised 18 May 2012 Accepted 21 May 2012 Multifrequency bioelectrical impedance analysis of body composition may be an appropriate alternative to dual-energy x-ray absorptiometry. We hypothesized that there would be no significant differences between dual-energy x-ray absorptiometry and either the Biospace (Los Angeles, CA, USA) InBody 520 or 720 multifrequency bioelectrical impedance analysis devices for total lean body mass (LBM), appendicular lean mass (ALM), trunk lean mass (TM), and total fat mass (FM) in 25 men and 25 women (including lean, healthy, and obese individuals according to body mass index), age 18 to 49 years, weight of 73.6 ± 15.4 kg. Both devices overestimated LBM in women (~ 2.5 kg, P < .001) and underestimated ALM in men (~ 3.0 kg, P < .05) and women (~ 1.0 kg, P < .05). The 720 overestimated FM in men (1.6 kg, P < .05) and underestimated TM in women (0.6 kg, P ≤ .05). Regression analyses in men revealed R 2 (0.87-0.91), standard error of the estimate (SEE; 2.3-2.8 kg), and limits of agreement (LOAs; 4.5- 5.7 kg) for LBM; R 2 (0.62-0.87), SEE (1.5-2.6 kg), and LOA (3.2-6.0 kg) for ALM; R 2 (0.52-0.71), SEE (2.4-3.0 kg), and LOA (4.6-6.1 kg) for TM; and R 2 (0.87-0.93), SEE (1.9-2.6 kg), and LOA (5.9-6.2 kg) for FM. Regression analyses in women revealed R 2 (0.87-0.88), SEE (1.8-1.9 kg), and LOA (4.1-4.2 kg) for LBM; R 2 (0.78-0.79), SEE (1.4-1.5 kg), and LOA (2.7-2.9 kg) for ALM; R 2 (0.76-0.77), SEE (1.0 kg), and LOA (2.2-2.3 kg) for TM; and R 2 (0.95), SEE (2.2 kg), and LOA (4.3-4.4 kg) for FM. The InBody 520 and 720 are valid estimators of LBM and FM in men and of LBM, ALM, and FM in women; the 720 and 520 are valid estimators of TM in men and women, respectively. © 2012 Elsevier Inc. All rights reserved. Keywords: Multifrequency bioelectrical impedance Dual-energy x-ray absorptiometry Body composition Lean mass Fat mass Segmental lean mass 1. Introduction More than one-third (34.4%) of Americans are classified as obese, having a body mass index (BMI) of 30 kg/m 2 [1], and obesity trends in the United States have continued to rise over the past 3 decades [2]. Body mass index, however, does not provide an accurate measure of an individual's specific body composition (ie, water, fat, and muscle). The increased prevalence of obesity stresses the need for safe, accurate methods of assessing body composition that are more accessible and economical than the traditional methods, which are largely exclusive to research and clinical settings. NUTRITION RESEARCH 32 (2012) 479 – 485 Abbreviations: ALM, appendicular lean mass; BMI, body mass index; BIA, bioelectrical impedance analysis; CV, coefficient of variation; DXA, dual-energy x-ray absorptiometry; FM, total fat mass; ICC, intraclass correlation coefficient; LBM, total lean body mass; LOAs, limits of agreement; MF, multifrequency; ρ C , Lin correlation coefficient; SEE, standard error of the estimate; TM, trunk lean mass. ⁎ Corresponding author. Tel.: +1 323 442 2498; fax: +1 323 442 1515. E-mail address: eschroed@usc.edu (E.T. Schroeder). 0271-5317/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.nutres.2012.05.009 Available online at www.sciencedirect.com www.nrjournal.com