BACKGROUND: Weight scales as height squared, which is an observation that forms the basis of body mass index (weight/height(2)). If, and how, circumferences, including waist circumference (WC) and hip circumference (HC), scale to height remains unclear, but this is an important consideration when developing normative ranges or applying WC/height and HC/height as risk indexes. OBJECTIVE: The study aim was to examine the scaling of weight, WC, and HC to height in NHANES (National Health and Nutrition Examination Survey) III participants. DESIGN: Subjects were adult non-Hispanic white, non-Hispanic black, and Mexican American men (n = 7422) and nonpregnant women (n = 7999) who had complete demographic and anthropometric data. In addition to height, allometric models were developed for each measure that controlled for age, race, and self-reported health status. RESULTS: After adjustment for age and race, weight scaled to height in men and women with mean (±SEE) powers of 2.29 ± 0.11 and 1.80 ± 0.07, respectively (both P < 0.001). Although univariate circumference-height models were weak or nonsignificant, when adjusted for age and race WC and HC scaled to height with powers of 0.76 ± 0.08 and 0.45 ± 0.05, respectively, in men and 0.80 ± 0.05 and 0.53 ± 0.04, respectively, in women (all P < 0.001). Age- and race-adjusted incremental increases in circumferences ranged from 0.2 to 0.5 cm per centimeter increase in height. Both WC/height and HC/height scaled negatively to height in men and women, and WC/HC scaled negatively to height in women only (all P < 0.001). Health status-adjusted models were similar. CONCLUSIONS: Circumferences and related ratios scale significantly to height, notably after adjustment for age and race, across subjects who are representative of the US population. These observations have implications for the clinical and epidemiologic use of these anthropometric measures and indexes.
BACKGROUND: Weight scales as height squared, which is an observation that forms the basis of body mass index (weight/height(2)). If, and how, circumferences, including waist circumference (WC) and hip circumference (HC), scale to height remains unclear, but this is an important consideration when developing normative ranges or applying WC/height and HC/height as risk indexes. OBJECTIVE: The study aim was to examine the scaling of weight, WC, and HC to height in NHANES (National Health and Nutrition Examination Survey) III participants. DESIGN: Subjects were adult non-Hispanic white, non-Hispanic black, and Mexican American men (n = 7422) and nonpregnant women (n = 7999) who had complete demographic and anthropometric data. In addition to height, allometric models were developed for each measure that controlled for age, race, and self-reported health status. RESULTS: After adjustment for age and race, weight scaled to height in men and women with mean (±SEE) powers of 2.29 ± 0.11 and 1.80 ± 0.07, respectively (both P < 0.001). Although univariate circumference-height models were weak or nonsignificant, when adjusted for age and race WC and HC scaled to height with powers of 0.76 ± 0.08 and 0.45 ± 0.05, respectively, in men and 0.80 ± 0.05 and 0.53 ± 0.04, respectively, in women (all P < 0.001). Age- and race-adjusted incremental increases in circumferences ranged from 0.2 to 0.5 cm per centimeter increase in height. Both WC/height and HC/height scaled negatively to height in men and women, and WC/HC scaled negatively to height in women only (all P < 0.001). Health status-adjusted models were similar. CONCLUSIONS: Circumferences and related ratios scale significantly to height, notably after adjustment for age and race, across subjects who are representative of the US population. These observations have implications for the clinical and epidemiologic use of these anthropometric measures and indexes.