Prediction of body fat in adult female runners

V. C. Rogosheske, R. C. Serfass, J. M. Kelly, B. Ainsworth, G. A. Stull

Research output: Contribution to journalArticlepeer-review

Abstract

The relative body fat of 108 female Caucasian runners (X̄ age = 29.4 ± 6.7 years) was determined using densitometry and predicted using the skinfold equations of (a) Jackson, Pollock, and Ward; (b) Sloan, Burt, and Blyth; and (c) Sinning. All skinfold equations predicted mean body fat within 1.0% compared to the 18.9 ± 4.9% calculated from measures taken by hydrostatic weighing. Only the mean of 19.90 ± 3.20% fat, using the Sloan-Burt-Blythe equation, was significantly different (p < .05) from the densitometric mean. The correlation coefficients between body fat, determined by hydrostatic weighing, and the predicted values were .82 for Jackson, Pollock, and Ward; .72 for Sloan, Burt, and Blyth; and .65 for Sinning, Standard errors of estimate (SEE) and total errors for the respective equations were 2.80 and 2.82 for Jackson, Pollock, and Ward; 3.36 and 3.49 for Sloan, Burt, and Blyth; and 3.70 and 4.53 for Sinning. Plots of the regression lines along with the SEE's relative to lines of identity revealed that the Jackson-Pollock-Ward equation provided the most accurate estimation of the body fat of the runners in this study. The Sloan-Burt-Blyth equation, while overpredicting the leaner subjects, was accurate within the range of 3.1 to 3.5% prediction error commonly found for population-specific equations. The Sinning equation underpredicted body fat in leaner subjects, overpredicted fatter subjects, and yielded the highest standard and total errors of estimate.

Original languageEnglish (US)
Pages (from-to)90-96
Number of pages7
JournalClinical Kinesiology
Volume47
Issue number4
StatePublished - Jan 1 1994

ASJC Scopus subject areas

  • Physical Therapy, Sports Therapy and Rehabilitation
  • Rehabilitation

Fingerprint

Dive into the research topics of 'Prediction of body fat in adult female runners'. Together they form a unique fingerprint.

Cite this