We studied the correlations between body mass index (BMI) and percent body fat (fat%) measured by dual-photon absorptiometry (DPA) in 445 white and 242 Asian adults aged 18-94 y. In addition, comparisons in six circumferences and eight skinfold-thickness measurements between whites and Asians were made to explain the discrepancies. Although Asians had lower BMI, they were fatter than whites of both sexes. The correlations between fat% and BMI varied by BMI and sex and race. Comparisons in anthropometry show that Asians had more subcutaneous fat than did whites and had different fat distributions from whites. Asians had more upper-body subcutaneous fat than did whites. The magnitude of differences between the two races was greater in females than in males. Prediction equations developed for each sex and race, based on BMI alone, gave SEEs ranging from 4.4% to 5.7%. All were significantly improved to the range of 3.5-4.4% when age and several skinfold-thickness measurements were added.

WC values at the 4 commonly used anatomic sites differ in magnitude depending on sex, are highly reproducible, and are correlated with total body and trunk adiposity in a sex-dependent manner. These observations have implications for the use of WC measurements in clinical practice and patient-oriented research.

OBJECTIVE:Although the body mass index (BMI, kg/m 2 ) is widely used as a surrogate measure of adiposity, it is a measure of excess weight, rather than excess body fat, relative to height. We examined the relation of BMI to levels of fat mass and fat-free mass among healthy 5-to 18-y-olds. METHODS AND PROCEDURES: Dual-energy X-ray absorptiometry was used to measure fat and fat-free mass among 1196 subjects. These measures were standardized for height by calculating the fat mass index (FMI, fat mass/ht 2 ) and the fat-free mass index (FFMI, fat-free mass/ht 2 ). RESULTS: The variability in FFMI was about 50% of that in FMI, and the accuracy of BMI as a measure of adiposity varied greatly according to the degree of fatness. Among children with a BMI-for-age Z85th P, BMI levels were strongly associated with FMI (r ¼ 0.85-0.96 across sex-age categories). In contrast, among children with a BMI-for-age o50th P, levels of BMI were more strongly associated with FFMI (r ¼ 0.56-0.83) than with FMI (r ¼ 0.22-0.65). The relation of BMI to fat mass was markedly nonlinear, and substantial differences in fat mass were seen only at BMI levels Z85th P. DISCUSSION: BMI levels among children should be interpreted with caution. Although a high BMI-for-age is a good indicator of excess fat mass, BMI differences among thinner children can be largely due to fat-free mass.