Lj. Brandon, COMPARISON OF EXISTING SKINFOLD EQUATIONS FOR ESTIMATING BODY-FAT IN AFRICAN-AMERICAN AND WHITE WOMEN, The American journal of clinical nutrition, 67(6), 1998, pp. 1155-1161
The purpose of this study was to compare seven skinfold equations with
underwater weighing (UWW) for estimating body fat in 39 African Ameri
can [age: 22.8 +/- 3.6 y ((x) over bar +/- SD); weight: 59.6 +/- 8.3 k
g) and 39 white (age: 22.1 +/- 2.9 y; weight: 61.7 +/- 7.3 kg) women.
The hypothesis examined was that the equations would produce more accu
rate body fat estimates in white women, but would be appropriate for u
se in African American women. Body fat estimated from two quadratic, t
hree linear, and two logarithmic skinfold equations was compared with
body fat estimated from UWW; the same procedures were used to evaluate
the results in both African Americans and whites. The data were analy
zed by using t tests, analysis of variance, Scheffe's honestly signifi
cant difference tests, correlations, error assessments, and agreement.
The results showed that total error, SEE, and SD values were larger i
n the African American women than in the white women and were not with
in acceptable limits listed in the literature. The correlation coeffic
ients were lower in the African American women than in the white women
. Agreement between the skinfold equations and UWW, based on deviation
s from mean differences, was better in the white women. In conclusion,
the skinfold prediction equations evaluated in this study were more v
ariable and produced more error when used in African American women. T
herefore, population-specific equations for African American women sho
uld be used to estimate body fat because they will probably yield more
accurate estimates.