<![CDATA[Child malnutrition is a persistent public health issue in East Java, Indonesia, characterized by uneven spatial distribution across its 38 regencies and cities. This study aims to model the prevalence of malnutrition among children under five using Geographically Weighted Regression (GWR) to identify locally significant determinants. Secondary data used in this study is prevalence of child nutritional status by regencies/cities in East java, taken from the 2023 Indonesian Health Survey, incorporating seven predictor variables: low birth weight prevalence, complete immunization coverage, exclusive breastfeeding, access to improved sanitation, number of community health posts (Posyandu), access to clean water, and poverty rate. Spatial dependence and heterogeneity were confirmed through Moran’s I (p = 0.009) and Breusch-Pagan tests (p = 0.024), validating the application of GWR. Spatial dependence and heterogeneity were confirmed through Moran’s I (p = 0.009) and the Breusch-Pagan test (p = 0.024), indicating the relevance of a spatial modelling approach. The best-performing model used an adaptive bi-square kernel (CV = 0.133; R² = 94.15%). All predictors exhibited spatial variability with statistically significant effects in specific regions (p < 0.05). In Tuban Regency, for instance, five variables including low birth weight, breastfeeding practices, and sanitation were significantly associated with malnutrition rates. These findings suggest that the relationship between predictors and malnutrition is not uniform across regions. GWR enables the identification of local patterns often overlooked by global models, offering a more accurate understanding of spatial disparities. The results provide strong evidence for developing targeted, region-specific public health strategies to address child malnutrition more effectively in East Java ]]>
Copyrights © 2025
