Identifying the Confounders of Marginalization and Mortality in Mexico, 2003–2007

Abstract

Confounders alter any predicted outcome by modifying the relationship between cause and consequence. Confounders can appear in different proportions inside a population and their effect might not be the same for different geographic regions. This paper focuses on establishing the impact of the marginalization index and confounders like migration and percentage of indigenous population on a municipality’s mortality rate. Data come from the 2003–2007 Instituto Nacional de Estadistica y Geografia death records and the 2005 Marginalization Index from the Consejo Nacional de Poblacion. WLS regressions of age-adjusted mortality rates on the index as a whole and on seven of its nine components are calculated to measure the impact of these confounders. Overall, results were mixed for municipalities with a higher level of marginalization. Percentage of indigenous population is the most consistent confounder throughout the analysis. The results show the difficulty of quantifying confounders accurately but open the door for further refinement.