A comparison between composition and density of the fat-free mass of young adult Singaporean Chinese and Dutch Caucasians

Background: The body composition was measured in 30 female Singapore Chinese and 34 female Dutch Caucasians aged (mean +/- SD) 22.7 +/- 3.6 years and in 39 male Singapore Chinese and 39 male Dutch Caucasians aged 23.3 +/- 3.2 years. Methods: The body fat percentage (BF%) was measured using densitometry and deuterium oxide dilution as well as a chemical four-compartment model, including fat mass, water, mineral, and protein. The chemical composition of the fat-free mass (FFM) and its density were calculated. Results: Mean body mass index (kg/m2) and BF% based on the four-compartment model in the four subgroups were 20.3 +/- 1.5 and 29.7 +/- 5.3 for the Singapore Chinese females, 22.1 +/- 1.2 and 28.9 +/- 5.2 for the Dutch females, 21.5 +/- 2.5 and 19.4 +/- 6.1 for the Singapore males, and 22.2 +/- 1.8 and 15.8 +/- 4.9 for the Dutch males. The contributions to the FFM of water, protein, and mineral were, respectively, 72.8 +/- 1.5, 19.2 +/- 1.8, and 7.9 +/- 0.8% for the Singaporean females and 74.2 +/- 1.4, 18.3 +/- 1.5, and 7.6 +/- 0.5% for the Dutch females. In males these figures were 73.2 +/- 1.7, 19.6 +/- 1.7, and 7.3 +/- 0.5% for the Singaporeans and 72.9 +/- 1.4, 20.5 +/- 1.4, and 6.6 +/- 0.5% for the Dutch. In females, these figures were for all components significantly different between the ethnic groups; in males the differences were significant for protein and mineral only. The calculated densities (kg/l) of the FFM were 1.1074 +/- 0.0057, 1.1012 +/- 0.0051, 1.1027 +/- 0.0065, and 1.1004 +/- 0.0055 for the Chinese females, Caucasian females, Chinese males, and Caucasian males, respectively. Conclusions: Although it cannot be excluded that the differences in density and composition of the FFM are due to methodological differences between the study sites, the findings may have consequences for the validity of single (reference) methods due to violation of normally used assumptions. The findings explain the observed underestimation of BF% from density using Siri's formula in the Chinese subjects, especially in females. The use of deuterium oxide as a single method to determine the BF% resulted in a smaller bias than densitometry (overall 1.6 +/- 2.0% for densitometry, 0.3 +/- 1.7% for deuterium dilution), suggesting that the deuterium oxide method may be more suitable as a single method in comparative body composition studies.