Spatial analysis of the relationship between early childhood mortality and malaria endemicity in Malawi

Spatial differences in mortality have been reported in Africa amongst children under-five years of age. Risk factors contributing to this geographical variation include bio-demographic and socio-economic factors, the prevalence of infectious diseases and the variability in the quality of child health care. This paper is concerned with investigating the link between early childhood mortality and malaria risk. We used data from the Mapping Malaria Risk in Africa (MARA) and Demographic and Health Survey (DHS) databases to explore this relationship. The DHS survey included questions on bio-demographic and socio-economic status, complete birth histories and survival time of each child within the five years preceding the survey. Survival times were computed in months until death or until the survey was done. The malaria risk was based on prevalence data estimated at the precise DHS sampling location. A spatial Cox regression model was applied to analyze child survival, assessing the influence of both individual-specific factors, malaria endemicity and group-specific environmental factors, approximated by geographical location. Geographical location was considered at subdistrict level. Our analysis shows that although malaria endemicity is not associated with the risk of infant mortality, it is an important risk factor for child mortality. The results confirm the effects of bio-demographic and socio-economic variables (maternal education, maternal age, birth order and place of residence) on infant and child mortality. The subdistrict-specific variation of infant and child mortality shows a rural-urban distinction with a relatively lower risk of mortality in main urban areas.