Age-Dependent Heterogeneity Trends in the Associations between Urban Green Space and Cardiovascular Health in a Humid-Subtropical Megacity

Abstract Urban green space (UGS) is a promising nature-based solution for cardiovascular health, yet its associative pathways in pollution-saturated, high-density environments remain contested. We applied a two-stage framework combining gradient boosting decision trees and structural equation modeling to a clinical cohort of 23,312 patients in Nanjing, China. Air-quality-related pathways accounted for more than 75% of the relative statistical contribution to model variance, exceeding psychological restoration proxies. At the macroscale, the aggregated air-quality pathway showed a marginal positive coefficient; although this may conceptually resemble aerodynamic interference, the interpretation remains speculative because no morphological or fluid-dynamics data were available. Stage 2 models identified distinct protective associative networks. For chronic ischemic heart disease, vegetation vitality was inversely associated with NO2 and SO2 and was linked to more favorable metabolic markers. For heart failure, the vegetation condition was associated with mixed pollutants and hemodynamic-metabolic indicators. Age-stratified analyses suggested heterogeneity in canopy density-PM2.5 associations, but subgroup slopes were nonsignificant, and effect sizes were extremely small, indicating fragile relative fluctuations rather than structural reversal. These findings characterize UGS primarily as a biophysical modulator and support targeted ecological strategies in traffic corridors and aging communities.