AUTHOR=Mazza Orit , Gluck Chemda , Haim Amir , Bornstein Robyn Jacob 

TITLE=Spatial patterns of childhood obesity clusters linked to socioeconomic inequalities

JOURNAL=Frontiers in Public Health

VOLUME=Volume 13 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2025.1497090

DOI=10.3389/fpubh.2025.1497090

ISSN=2296-2565

ABSTRACT=IntroductionThe childhood obesity epidemic continues to be a challenge worldwide despite advances in prevention and treatment. Multifactorial causes are responsible for this epidemiology, and unequivocally, environmental factors play a key role. Studying the connection between socioeconomic factors and prevalence of childhood obesity is key to instituting change on a public health level.ObjectiveTo identify geographical areas (clusters) with high and low prevalence of childhood obesity and examine their spatial association with socio-economic and demographic factors.MethodsCluster analysis of geographic and population data for localities and regional councils was performed using growth data for children in grade 1. Analysis of childhood obesity prevalence utilized spatial autocorrelation (Moran’s I) and Getis-Ord Gi∗statistic (hotspot analysis). Local Geographically Weighted Regression (GWR) and Multiscale GWR (MGWR) were performed to examine socio-economic and demographic predictors of the z-score values from the hotspot analysis.ResultsThe cluster analysis identified several significant spatial clusters of localities with high and low z-scores from the Getis-Ord Gi∗hotspot analysis. Both the GWR and MGWR models demonstrated notable spatial variation, achieving high adjusted R-squared values (95.2 and 96.4%, respectively) and low residual variances (0.05 and 0.03, respectively). The analysis indicated that the variables exhibited a significant localized effect (p < 0.05), underscoring spatial heterogeneity. Among these variables, three showed a significant influence across the entire geographic area: average years of education for individuals aged 25–54, the percentage of families with four or more children, and the socioeconomic index. These findings emphasize the spatial variability of these factors and the ability of the model to generate a range of coefficients tailored to each locality.ConclusionThe application of geographic techniques enables examination of spatial patterns of childhood obesity. The current analysis is the first study that demonstrate a direct association between cluster areas of childhood obesity prevalence and socio-economic and demographic factors in the Middle East. It highlights that spatial dependence and heterogeneity are key factors in analyzing patterns of childhood obesity. Correlations between sociodemographic parameters are consistent with patterns observed in high-income countries (a negative socioeconomic index association) and in middle-and low-income countries (a positive association with average years of schooling). The results suggest that the immediate geographic environment plays a substantial role in childhood obesity. Therefore, it may reflect different patterns related to macroeconomic factors, such as the country’s income level.