Linking urban green infrastructure and maternal and child health using spatial autocorrelation analysis of perinatal outcomes in India

Urban Green Infrastructure (UGI), including urban forests and vegetated spaces, has been associated with improved maternal and child health (MCH) outcomes, particularly in high-income countries. However, empirical evidence remains limited in rapidly urbanising and climate-vulnerable regions of the Global South, where UGIs may simultaneously buffer environmental and health risks. This study addresses this gap by developing a spatially explicit, data-driven framework to examine associations between green cover and perinatal outcomes in Karnataka, India. Remote sensing–derived Normalized Difference Vegetation Index (NDVI) data were integrated with district-level health indicators, including antenatal care (ANC), caesarean deliveries, and stillbirth rates. Multiple linear regression and spatial autocorrelation analyses were applied to assess UGI–health linkages. The regression model (adjusted R2= 0.691) indicates a statistically significant negative association between green cover and stillbirths. Spatial autocorrelation analysis using Moran’s I (0.44, p < 0.01) and Local Indicators of Spatial Association (LISA) reveals distinct geographic clusters, with low-UGI sub-districts overlapping hotspots of adverse birth outcomes. K-Nearest Neighbour (KNN) spatial weights enhanced detection of local clustering patterns. While the predictive model reproduced key aspects of the observed spatial structure, this does not constitute formal model validation. Overall, the findings provide evidence that UGI is associated with perinatal health outcomes in India, highlighting its potential role in climate adaptation and health equity. The study offers actionable insights for spatially informed urban planning and public health strategies in data-constrained, high-growth Global South contexts.