Does greenspace mitigate air pollution and motivate physical activity? A case study of four European cities

INTRODUCTION: Exposure to urban greenspace may affect health via a complex set of pathways, including lessened exposures to air pollution and enhanced opportunity for physical activity. The HEALS study included a personal monitoring component to pilot and collect environmental exposure data from individuals in four European cities (Edinburgh, UK; Utrecht, Netherlands; Athens and Thessaloniki, Greece) using novel monitors and mobile devices. The data collected over ~1-week periods thus presents an opportunity to assess these two important pathways for which greenspace may benefit health.
METHOD: We will include three different metrics of greenspace (Normalised Difference Vegetation Index [NDVI], proportion of tree canopy coverage, and proximity to green land use) and will track actual usage through GPS data. NDVI values, as an indicator of greenness, will be generated from Sentinel-2 satellite data and will be aligned to the timing of monitoring periods. We will use GPS data collected from the MOVES mobile phone application to assign the total duration engaging in active and passive transport. Fitbit units worn by study participants recorded steps per minute, which will be linked to the MOVES data. For exposure to air pollution, continuous measurements of indoor PM2.5 levels were collected from participants’ homes, and, at present, we are exploring the use of the Data Integration Model for Air Quality (DIMAQ) to estimate outdoor concentrations during the monitoring campaign. Analysis is ongoing to determine how these different types of greenspace affect exposure to air pollution and engagement in physical activity.
RESULT: Preliminary results of residential greenspace metrics did not suggest any associations between residential levels and indoor concentrations of PM2.5, noise levels, or indicators of physical activity. However, the mean (June) NDVI levels of the GPS locations (walking and running activities only) demonstrated significant positive correlations with the overall distance (adjusted for monitoring period duration) (r=0.46; p=0.02) and overall average hourly steps (r=0.41; p=0.04). No association was identified between the duration of walking and running and levels of NDVI.
CONCLUSION: Our preliminary analysis has shown positive associations between certain elements of greenspace and objective indicators of physical activity. We will expand the analysis to incorporate potentially important covariates of individual participants and also pool data from the other cities to confirm and refine these early study results.