Bayesian Geo-Additive Modelling of State-Level Heterogeneity in Anaemia Prevalence and Risk Factors among Under-Five Children in Nigeria

Abstract

Anemia is a very serious global health issue, particularly in sub-Saharan African countries, where it has significant implications for both human health and economic development. This study aims to identify the socioeconomic, demographic and climatic factors as well as geographical variations linked to the prevalence of anaemia in young children in Nigeria, based on the results of the 2018 Nigeria demographic and health survey. The association between various types of covariates and possible spatial variations was explored using a hierarchical Bayesian geo-additive modelling approach. In particular, the study focused on a binary response variable. Out of the four formulated semi-parametric models, geo-additive model with both structured and unstructured random effects was found to be the best fitted. The findings of the study reveal significant spatial variation in anemia risk, with the highest risk observed in the states of Sokoto, Niger, Akwa Ibom, Ebonyi and Bauchi. Besides, educational level of a child’s mother, wealth status of household, a child’s area of residence, prevalence of malaria and land surface temperature were all associated with childhood anemia. The prevalence of infant anemia decreased with increasing child's age. Mother’s Body Mass Index also has inverse relationship with the risk of childhood anaemia. Given the observed state-level patterns of anemia risk, it is important to implement targeted programs that address the specific needs of vulnerable children in each state. Th is could help reduce the prevalence of childhood anemia in the Nigeria.