Search

Non-communicable diseases are the global disease burden of our time, with physical inactivity identified as one major risk factor. Green spaces are associated with increased physical activity of nearby residents. But there are still gaps in understanding which proximity and what characteristics of green spaces can trigger physical activity. This study aims to unveil these differences with a rigorous sensitivity analysis. We gathered data on self-reported health and physical activity from 1365 participants in selected neighbourhoods in Porto, Nantes, Sofia, and Høje-Taastrup. Spatial data were retrieved from OpenStreetMap. We followed the PRIGSHARE guidelines to control for bias. Around the residential addresses, we generated seven different green space indicators for 15 distances (100–1500 m) using the AID-PRIGSHARE tool. We then analysed each of these 105 green space indicators together with physical activity and health in 105 adjusted structural equation models. Green space accessibility and green space uses indicators showed a pattern of significant positive associations to physical activity and indirect to health at distances of 1100 m or less, with a peak at 600 m for most indicators. Greenness in close proximity (100 m) had significant positive effects on physical activity and indirect effects on health. Surrounding greenness showed positive direct effects on health at 500–1100 m and so do green corridors in 800 m network distance. In contrast, a high quantity of green space uses, and surrounding greenness measured in a larger radius (1100–1500 m) showed a negative relationship with physical activity and indirect health effects. Our results provide insight into how green space characteristics can influence health at different scales, with important implications for urban planners on how to integrate accessible green spaces into urban structures and public health decision-makers on the ability of green spaces to combat physical inactivity. [Display omitted] • Effect size, significance and direction of effects were sensitive to the proximity. • Green space health effects operated at different proximities and may mask each other. • Physical activity mediated health effects for distinct green space characteristics. • Indirect effects were stronger with semi-public and private green spaces included. • Direct health effects for intermediate surrounding vegetation and green corridors. [ABSTRACT FROM AUTHOR]