Validation of body composition estimates in male and female distance runners using estimates from a four-component model

This study used estimates of body composition from a four-component model (%Fat(d,w,m)) to determine whether the assumed density (D(ffm)) and the composition of the fat-free mass (FFM), and estimates of body composition from methods based on two- and three-component models are valid in distance runners. Measures of body density (D(b)) by underwater weighing, total body water by deuterium dilution, and bone mineral by dual-energy X-ray absorptiometry (DXA) were obtained in 10 female and 12 male runners and an equal number of controls matched for age, height, weight, gender, and ethnicity. D(ffm) of the runners did not differ from 1.1 g.cm(-3) or from the controls even though the composition of the FFM differed from that assumed. Therefore, percentage of body fat (%Fat) from densitometry did not differ from %Fat(d,w,m) in the runners, although individual variation was substantial (-5.3 to 3.0% body mass, SD(diff) = 2.1% body mass). Three-component model estimates of %Fat from D(b) and body water agreed quite closely with %Fat(d,w,m) (x </= 1.6% body mass, SD(diff) = 0.5% body mass), whereas those from D(b) and body mineral (x(diff) = 1.6-2.9% body mass, SD(diff) = 2.3% body mass) and from DXA (x(diff) = 2.2-4.0% body mass, SD(diff) = 2.1% body mass) were less accurate and underestimated %Fat(d,w,m). We conclude that D(ffm) does not differ from 1.1 g.cm(-3) and that densitometry is a valid method to estimate group mean body composition, but that measurement of D(b) and total body water are necessary for accurate individual estimates of body composition in male and female distance runners. Am. J. Hum. Biol. 12:301-314, 2000. Copyright 2000 Wiley-Liss, Inc.