Longitudinal association between urban blue-green space exposure and mortality: A systematic review and meta-analysis of exposure types and buffers
Xingcan Zhou, Kojiro Sho, Hongfei Qiu, Shenglin Elijah Chang, Qingya Cen
Abstract
• Twenty-one longitudinal studies of urban blue-green space and mortality were reviewed. • NDVI showed a more significant effect size than green land cover. • The impact of NDVI exposure on mortality was greatest at a 500 m buffer. • Effect size trends varied with buffer changes across mortality types. Exposure to urban blue-green space 1 1 HR: Hazard ratio, LULC: Land Use or Land Cover, NDVI; Normalized Difference Vegetation Index, NDWI: Normalized Difference Water Index, PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses, UBGS, Urban blue-green space (UBGS) affects human health, but how to integrate long-term exposure to guide the measurement and intervention of UBGS remains unclear. We aimed to synthesize the latest evidence from longitudinal cohort studies on the association between objectively measured UBGS exposure and mortality, highlighting differences in health effect sizes across exposure types and buffers. We systematically reviewed articles published through January 2024 from PubMed, Web of Science, and Scopus and conducted a meta-analysis of the longitudinal associations between exposure and mortality in 21 studies involving 28,700,187 participants from 14 countries with a median follow-up of 10.3 years. Quantitative assessment indicated that the normalized difference vegetation index from satellite data and the green proportion from land use or land cover were protective against all-cause mortality: the former had a significantly larger effect size (per 0.1-unit increase, pooled HR 95% CI: 0.97, 0.96–0.98) and showed the greatest effect at 500m compared with the ≤300m and ≥1000m buffers. UBGS exposure had a more pronounced protective effect on respiratory mortality than on all-cause, circulatory, and cancer mortality, with opposite trends across buffer sizes. The findings were primarily for green space, as studies on blue space were limited in number and included varied metrics. Although the hazard ratios were fully adjusted for sociodemographic covariates and buffered subgroup analysis was conducted, residual confounding cannot be completely excluded. Further research should focus on differences in exposure types, especially blue spaces, analyze potential mechanisms, and validate the findings across different geographical contexts.