Revista Facultad de Medicina de la Universidad Nacional de Colombia
versão impressa ISSN 0120-0011
CUBILLOS-VASQUEZ, Sandra M et al. A methodological proposal for comparing anthropometric measurements between Colombian and US populations. Rev.fac.med.unal [online]. 2011, vol.59, suppl.1, pp. 31-31. ISSN 0120-0011.
Background. Using international growth patterns has been described as being inappropriate for describing a Latin-American population's characteristics and there is limited support for guaranteeing their applicability. Objective. Proposing a statistical methodology for comparing and predicting a target population's anthropometric values from a base population's values. Materials and methods. The Kolmogorov-Smirnov test was used for establishing whether there were significant differences in a local target population's anthropometric measurements and those for a base population and identifying data fit by using a simple linear regression model. Results. There were significant differences between Colombian and US population growth patterns; the anthropometric percentiles studied for the Colombian population were lower than US ones. A greater than 90% coefficient of determination was obtained once the linear ratio had been determined; this led to stating that a linear regression model for all ages from both genders with measurements for the Colombian population could be estimated from US population information. Conclusions. The statistical methodology proposed here fit a target population's nutritional classification when compared to foreign reference patterns where stateorganism accepted reference tables were not available. The results of the present study showed a tendency towards underestimating the Colombian population's nutritional state because it had been compared to populations such as the US population which has different growth trends; such situation may be presented in other populations.
Palavras-chave : measurements; methods and theories; statistics; honparametric; linear models; body height.