Other

Spatial Patterns and Determinants of Childhood Stunting in Ethiopia: A multilevel analysis of the 2024–2025 Ethiopia Demographic and Health Survey

Gebretsadik S, Tefera M, Jemal JB, Deybasso HA, Argaw R et al. · 19 Aug 2026

Abstract

Abstract Background Childhood stunting remains a major public health concern in Ethiopia, with substantial socioeconomic and geographic variation. Understanding both the determinants and spatial distribution of stunting is essential for designing targeted interventions. Therefore, this study aimed to identify factors associated with childhood stunting, spatial clustering, and geographic variation using multilevel and spatial analyses. Methods Data from 5,820 children aged 0–59 months in 787 DHS clusters in Ethiopia (2024–2025) were analyzed using multilevel mixed-effects logistic regression. Global and local spatial patterns were assessed using Moran’s I and Local Indicators of Spatial Association (LISA), respectively, and geographically weighted regression (GWR) examined spatially varying contextual determinants. Results The weighted national stunting prevalence was 39.8% (95% CI: 37.4–42.1). Child age, birth size, maternal education, maternal BMI, household wealth, and region were significant determinants in multilevel models. Spatial analysis revealed significant clustering (Moran’s I = 0.153, p  < 0.001), with distinct high- and low-burden areas across the country. GWR demonstrated geographic heterogeneity in the influence of maternal education, water and sanitation access, and household environmental conditions with stunting. Conclusion Stunting in Ethiopia remains high and unevenly distributed, shaped by both hierarchical and spatial factors. Integrating multilevel and spatial evidence highlights where and for whom interventions should be prioritized, supporting geographically tailored nutrition and health strategies.