Content area
Abstract
This thesis investigates the extent and the determinants of poor child health and nutrition as reflected by attained height, in rural Guatemala. We analyze anthropometric data on height-forage because, when calculated and interpreted correctly, these offer the most appropriate measures of children's long-term health and nutritional status for empirical research.
The ultimate goals of this research are threefold. First, we estimate the role played by exogenous individual, household and community characteristics in shaping children's height differentials in the context of a socially, economically, ethnically, and geographically diverse developing country such as Guatemala. Secondly, we investigate which proximate determinants of child health and nutrition are responsible for the poor growth outcomes of Guatemalan children. thirdly, we address empirical questions that have been ignored in previous anthropometric research, such as the distribution of child stunting across communities and the magnitude of intra-family correlation of height-for-age outcomes, before and after controlling for observed covariates.
We use a conceptual and empirical framework that models children's growth as a cumulative process produced within the social, economic and ecological environment in which children grow up. The economic model of the family and the proximate determinants framework guide the selection of the covariates and the interpretation of the parameter estimates. We fit multilevel models to hierarchically clustered data in order to control for both family and community heterogeneity.
Our results confirm findings from previous research suggesting that poor child growth outcomes in Guatemala are the consequence of widespread poverty. We find that ethnic differentials in height-for-age outcomes between children of ladino mothers and children of indigenous mothers who do not speak Spanish are larger among children of more educated parents and among children living in communities with better health care facilities. The beneficial effects of mother's education, husband's education and household resource availability seem to operate through better environmental conditions found in the household, and longer birth intervals among siblings.
Estimates derived from multilevel models reveal a large amount of clustering of child height-for-age outcomes within communities and families. Our models account for most of the community-level variation in child growth patterns but explain only half of the overall intra-family correlation.
