Provincial inequalities in child nutritional risk: a public health multivariate approach using Z-score matrices for...

Child malnutrition remains a persistent public health challenge in Ecuador, characterized by significant territorial inequalities that disproportionately affect children under 5 years of age. These disparities are closely linked to social, environmental, and structural determinants, requiring analytical approaches that go beyond national averages to capture spatial heterogeneity.

This study employed a quantitative, ecological, and cross-sectional design using secondary data from the National Survey on Child Undernutrition (ENDI 2022-2023). A provincial-level Z-score standardized matrix was constructed to ensure comparability across heterogeneous indicators. Multivariate techniques were applied, including Principal Component Analysis (PCA), HJ-Biplot representation, hierarchical clustering (Ward.D2), and k-means classification, to identify territorial patterns of nutritional vulnerability.

The first two principal components explained 58.47% of total variance, representing structural vulnerability and socio-environmental inequality. Cluster analysis identified three distinct territorial groups (high, medium, and low risk), confirming non-random spatial patterns. Provinces with higher malnutrition prevalence were consistently associated with limited access to drinking water, inadequate sanitation, precarious housing conditions, and lower caregiver education levels. Multivariate analyses revealed strong correlations among environmental and social determinants, highlighting the multidimensional nature of child malnutrition.

Findings suggest that child malnutrition in Ecuador is a multi-causal and territorially conditioned phenomenon shaped by persistent structural inequalities. The integration of Z-score standardization with multivariate techniques provides a robust and replicable framework for identifying priority areas and supporting targeted, evidence-based public health interventions.