Determinants of anemia among children aged 6-23 months in Nepal: an alternative Bayesian modeling approach

BACKGROUND: Anemia remains a major public health concern among children under two years of age in low- and middle-income countries. Childhood anemia is associated with several adverse health outcomes, including delayed growth and impaired cognitive abilities. Although several studies in Nepal have examined the determinants of anemia among children aged 6-23 months using nationally representative data, alternative modeling approaches remain underutilized. This study applies a Bayesian analytical framework to identify key determinants of anemia among children aged 6-23 months in Nepal. METHODS: This cross-sectional study analyzed data from the 2022 Nepal Demographic and Health Survey (NDHS). The dependent variable was anemia in children (coded as 0 for non-anemic and 1 for anemic), while independent variables included characteristics of the child, mother, and household. Descriptive statistics including frequency, percentage and Chi-squared test of associations between the dependent variable and independent variables were presented. Bayesian binary logistic regression model was used to identify significant determinants of anemia. A noninformative normal prior with mean 0 and variance 1000 was assumed to derive the posterior distribution for the model coefficients. The Gibbs sampling method was executed to obtain posterior samples, and the results were summarized using posterior mean, odds ratio, standard error, Monte Carlo error, and 95% credible interval (CrI). RESULTS: The analysis of data from 685 children aged 6-23 months revealed several determinants of anemia. Compared to children aged 19-23 months, children aged 6-12 months had 3.15 times higher odds of anemia (95% CrI: 2.02 - 4.91) and those aged 13-18 months had 2.35 times higher odds (95% CrI: 1.53 - 3.61). Second-born children had 1.49 times higher odds (95% CrI: 1.001 - 2.23) than first-born children. Mothers who did not take deworming medication during pregnancy had 1.57 times higher odds of having an anemic child (95% CrI: 1.14 - 2.51) compared to those who took such medication. Children living in the Terai region had 2.45 times higher odds (95% CrI: 1.53 - 3.98) than those living in the Hill region. Ethnicity also plays a role, as Dalit children had 2.29 times higher odds of anemia (95% CrI: 1.03 - 5.04) and Janajati children had 2.73 times higher odds (95% CrI: 1.24 - 5.92), compared to Muslim and other ethnic groups. CONCLUSION: Key determinants of anemia among children aged 6-23 months in Nepal include child's age, birth order, maternal intake of deworming medication during pregnancy, ecological region, and ethnicity. These findings highlight the necessity for targeted interventions that address demographic factors, maternal health, and geographic disparities to reduce childhood anemia in Nepal.