Godoy R, Goodman E, Levins R, Leonard W R
Heller School for Social Policy and Management, Brandeis University, Waltham, MA 02454-9110, USA.
Ann Hum Biol. 2005 Jul-Aug;32(4):469-86. doi: 10.1080/03014460500154384.
The study analysed variability in physical stature, weight, and body mass index (BMI) in the USA during 1971-2002.
Subjects were non-Hispanic Blacks and Whites, 2-74 years of age from the National Health and Nutrition Examination Surveys (NHANES I-III and 1999-2002).
The coefficient of variation and the standard deviation of the logarithm of stature, weight, and BMI were used to assess anthropometric variability for groups defined by age, race, sex, income, and survey year. Weighted ordinary least squares regressions were used to estimate the effect of socio-economic variables on anthropometric variability.
(a) The relation between age and variability in weight or BMI resembles an inverted U, (b) men have lower variability in BMI than women, (c) Blacks and the poor have greater variability in weight and BMI than Whites or than the non-poor, and (d) variability in anthropometric indices increased during 1971-2002. Results were robust to the measure of variability used and to the use of the mean and mean square of the anthropometric indicators as explanatory variables.
Since anthropometric indices correlate reliably with canonical indicators of well-being (e.g. income), growing variability in anthropometric indices, particularly among the Blacks and the poor, signals growing inequality in quality of life--a worrisome trend.
