Exploratory Analysis of Height-for-Age in Guatemala: An Examination of Potential Spatial and Environmental Components

Stephen D. McCracken, U.S. Centers for Disease Control and Prevention (CDC)
Paul Stupp, U.S. Centers for Disease Control and Prevention (CDC)

Recently demographers are exploring community effects and GIS for understanding contextual variables associated with malnutrition. In this poster presentation we begin with a geographic display of the spatial pattern of mean height-for-age z-scores (HAZ) for 373 PSUs (6,308 children) from the 2002 National Maternal and Child Health Survey of Guatemala. OLS regression on HAZ is carried out to obtain predicted HAZ and residuals based on individual characteristics. Mean residuals are geographically displayed and spatial statistics techniques applied to identify spatial patterns. Visual examination indicates the importance of altitude, geographic accessibility, and municipal level measures of poverty. These variables are extracted and assigned to the individual level data, and used in multilevel regression. Results indicate the important role of ethnicity, maternal and paternal education, work activities of father, and demographic characteristics of child at the individual level as well as the contextual role of access to local services, altitude, and poverty.