Abstract P167: Residential Exposure to Greenness and Mortality Following Percutaneous Coronary Intervention

Abstract

Background: Evidence regarding environmental exposure to green spaces and outcomes in coronary disease patients is lacking. We evaluated the association between residential exposure to greenness and mortality in patients undergoing percutaneous coronary interventions (PCI). Methods: Consecutive patients undergoing PCI at the Rabin Medical Center in Israel between 2004-2014 (n = 12,104) were studied. Clinical data at the time of hospitalization were extracted from medical records. Mortality data (through 2017) were obtained from the Ministry of Health. Patients with incomplete information on residential addresses were excluded. Exposure to greenness was estimated using normalized difference vegetation index (NDVI), a satellite-based index derived from Landsat 30 m spatial resolution imagery, with larger values indicating higher levels of vegetative density. NDVI was estimated within a buffer of 300 m around each patient’s home and as the point value of each 30 m pixel (immediate living environment). Additional residential-based environmental measures were obtained. Cox models assessed the hazard ratios (HRs) for mortality associated with greenness measures. Results: Among 11,262 patients analyzed [median age, 69 (IQR 61-78) y, 24% women], median NDVI-300 was 0.15 (IQR 0.13-0.17) and median NDVI-30 was 0.14 (IQR 0.11-0.17). Patients with higher NDVI-300 were slightly older; NDVI-30 was inversely correlated to ambient air pollution. During a median follow-up of 8.1 (IQR 5.1-10.6) years, 3,217 participants died. After adjustment for sociodemographic and clinical factors, NDVI-30 -but not NDVI-300- was associated with lower mortality (Figure), with an HR of 0.96 (95% CI: 0.92-0.99) per 1 SD increase. Conclusions: In this PCI registry, residential exposure to green spaces in the immediate living environment was associated with lower mortality. Results for the extended area of the living environment were inconclusive. This inconsistency between different spatial resolutions warrants further investigation