The anthropometric estimation of body density and lean body weight of male athletes

Abstract

The purpose of this investigation was to evaluate the validity of using selected equations which had been derived from anthropometric measures of general populations for estimating the body density (D_B) and lean body weight (LBW) of college athletes. New regression equations were also derived on the basis of measurements taken on the sample of athletes studied. True D_B (underwater weighing), true percent fat, skinfolds (SF), and skeletal diameters were measured in 50 male, spring sports participants (17 baseball, 15 track and field-11 spring football and 7 tennis). The correlation for true D_B and estimates from the equation of Pascale, et al. were significant (.76) as was that for Sloan's equation (.74). Correlations between Behnke's technique using skeletal diameters and different k values were significant (all .85). Multiple correlations between true D_B and D_B estimated from newly derived equations ranged from .43 for equations using only skeletal diameters to .90 for equations using SF's and diameters. The highest multiple correlation (.90) was obtained from the equation D_B = 1.02967 - .00131 subscapular SF + .00196 bitrochanteric diameter - .00126 abdominal SF - .00096 triceps SF + .00260 sum of knee diameters - .00114 bi-iliac diameter. When the established and newly derived equations were used to estimate D_B values of sub-groups formed according to the sports represented, it was found that groups at the high or low extremes of body density were not as accurately evaluated as those groups nearer the total sample mean.