ABSTRACT
We examined relationships between skinfold (SKF) and optical density (delta OD) measurements across age and levels of body fatness (%BF) for 151 women, 20 to 72 years. There were significant (p < .05) relationships between delta ODs and SKFs at all sites, except the thigh. The interaction (SKF x Age) was significant (p < .05) for pectoral and biceps delta ODs. Slope comparisons indicated the relationships for younger (29 years) and older (59 years) women differed significantly from zero and each other (p < .05). Analysis of SKF x %BF interactions revealed that relationships between SKFs and delta ODs at the pectoral and biceps sites for leaner (22% BF) women differed significantly from zero (p < .05) and were larger than those for obese (39% BF) women (p < or = .05). Thus, the relationship between SKFs and delta ODs is stronger for younger and leaner women compared to older and fatter women. These findings may reflect differences in fat layering due to age or body fatness and provide insight as to why the manufacturer's near-infrared (NIR) equation significantly underestimates the %BF of obese women.
Subject(s)
Adipose Tissue/anatomy & histology , Aging/pathology , Body Composition , Body Mass Index , Adult , Aged , Body Height , Body Weight , Factor Analysis, Statistical , Female , Humans , Infrared Rays , Middle Aged , Muscles/anatomy & histology , Optics and Photonics , Regression Analysis , Reproducibility of Results , Skinfold ThicknessABSTRACT
Three methods of body composition assessment were used to estimate percent body fat (%BF) in nonobese (n = 77) and obese (n = 71) women, 20-72 yrs of age. Skinfolds (SKF), bioelectrical impedance (BIA), and near-infrared interactance (NIR) methods were compared to criterion-derived %BF from hydrostatic weighing (%BFHW). Nonobese subjects had < 30% BFHW and obese subjects had > or = 30% BFHW. The Jackson, Pollock, and Ward SKF equation and the manufacturer's equations for BIA (Valhalla) and NIR (Futrex-5000) were used. For nonobese women there were no significant differences between mean %BFHW and %BFSKF, %BFBIA, and %BFNIR. The rs and SEEs were 0.65 and 3.4% BF for SKF, 0.61 and 3.6% BF for BIA, and 0.58 and 3.7% BF for NIR for nonobese subjects. For obese women, mean %BFHW was significantly underestimated by the SKF, BIA, and NIR methods. The rs and SEEs for the obese group were 0.59 and 3.4% BF for SKF, 0.56 and 3.5% BF for BIA, and 0.36 and 3.9% BF for NIR. The total errors of the equations ranged from 5.6 to 8.0% BF in the obese group. It is concluded that all three field methods accurately estimate %BF for nonobese women; however, none of the methods is suitable for estimating %BF for obese women.
