Synthesizing Risk Factors for Child Health: The Case of Mosquito-Borne Diseases in Dakar

In Senegal, mosquito-borne febrile diseases constitute a major burden in terms of child morbidity and mortality. In this context, the current research examines child mosquito-borne disease in an urban environment and aims, firstly, to illustrate a new approach for synthesizing the network of its determinants. Secondly, the paper analyzes the relative role of these determinants in children's febrile disease in Dakar, focusing on protective behaviors, socio-economic characteristics and domestic environment. The data used comes from a survey conducted in 2008 in Dakar and the statistical analysis consists of two steps: first, several latent variables are estimated, using Latent Class Analysis (LCA). The yielded latent variables subsequently constitute the explanatory part of a multi-level logistic regression, in order to explain the occurrence of a recent febrile episode in children. Findings indicate that the febrile episodes are influenced primarily by health-related behaviors and ? to a lesser extent ? by factors relating to the socio-economic status of the family, while environmental factors play a minor role. The paper illustrates the use of LCA for synthesizing a wide array of indicators, in order to create a comprehensive causal image of mosquito-borne disease risks, all the while using parsimonious statistical models.