First longitudinal study of indigenous children's growth in Brazil: the Xavante project
METHODS: This is the first prospective longitudinal study on the subject in South America. Research was carried out in two Xavante villages in Central Brazil. After the initial phase of data collection (July 2009), five follow-up waves were performed at intervals of approximately six months (January 2010 to January 2012). A generalized additive mixed model (GAMM) was used for constructing Xavante child growth curves according to sex, considering the existing correlation structure and compared to the WHO reference. Socioeconomic, demographic, and maternal variables were analyzed in relation to height (response variable). Retained in the final model were sex, place of residence (village), income per capita, and number of individuals in the household, with a smoothing spline term for age.
RESULTS: Xavante children’s height was lower than the WHO reference curve median, approaching the 3rd percentile in the first years of life. Girls were 1.49cm shorter than boys and children in the first village were 0.81cm smaller than those in the second. Height inequalities were observed according to several variables, suggesting internal socioeconomic differentiation with impacts on children’s nutritional status. Having larger households and lower per capita income were associated with lower height-for-age. The height difference between two children of the same age with disparate per capita incomes reached averages as high as 1.22cm.
CONCLUSIONS: The study’s longitudinal design demonstrated the persistence of health disparities between indigenous and non-indigenous peoples, with Xavante children growing far below what is considered optimal according to the WHO reference, as well as compared to non-indigenous children in Brazil.