Validity of bioelectrical impedance equations for estimating percent fat in males

Joan M. Eckerson, Jeffrey R. Stout, Terry J. Housh, Glen O. Johnson

Research output: Contribution to journalArticle

15 Scopus citations

Abstract

The present study examined the validity of selected bioelectrical impedance (BIA) equations for estimating percent fat (% fat) in males and compared their validity with that of commonly used skinfold equation (Sum3). One-hundred twenty-two Caucasian males (X̄ ± SD = 12.5 ± 5.8% fat, as determined by underwater weighing) served as subject. Selection of the BIA equations was dependent upon meeting at least one of three criteria: 1) developed from a previous interlaboratory investigation, 2) derived on a large sample size (>200), or 3) previously been shown to accurately estimate body composition when cross-validated against a criterion method. Cross- validation analyses included examination of the constant error, standard error or estimate (SEE), r, and total error (TE). The Sum3 equation which resulted in the lowest SEE and TE values (2.6% fat and 3.4% fat, respectively) and the highest validity coefficient (r = 0.90, P <0.001), most accurately estimated % fat and, therefore, was recommended over RIA equations for estimating body composition in Caucasian males with lean to average body fatness. The fat-specific interlaboratory BIA equation of Segal et al. (26) for males ≤20% fat (N = 107) which resulted in a TE value of 3.6% fat and the generalized equation of Guo et al. (10) (TE = 4.1% fat) may, however, be considered as acceptable alternatives.

Original languageEnglish (US)
Pages (from-to)523-530
Number of pages8
JournalMedicine and Science in Sports and Exercise
Volume28
Issue number4
DOIs
StatePublished - Apr 29 1996

All Science Journal Classification (ASJC) codes

  • Orthopedics and Sports Medicine
  • Physical Therapy, Sports Therapy and Rehabilitation

Fingerprint Dive into the research topics of 'Validity of bioelectrical impedance equations for estimating percent fat in males'. Together they form a unique fingerprint.

  • Cite this