TY - JOUR
T1 - Urban greenspace and the indoor environment
T2 - Pathways to health via indoor particulate matter, noise, and road noise annoyance
AU - Mueller, William
AU - Steinle, Susanne
AU - Pärkkä, Juha
AU - Parmes, Eija
AU - Liedes, Hilkka
AU - Kuijpers, Eelco
AU - Pronk, Anjoeka
AU - Sarigiannis, Denis
AU - Karakitsios, Spyros
AU - Chapizanis, Dimitris
AU - Maggos, Thomas
AU - Stamatelopoulou, Asimina
AU - Wilkinson, Paul
AU - Milner, James
AU - Vardoulakis, Sotiris
AU - Loh, Miranda
N1 - Funding Information:
We thank Hilary Cowie for her input to statistical methods and manuscript review. We are grateful to Reinier Sterkenburg and Meredith Franklin for providing expertise on GIS and spatial analysis. Remy Franken and John Cherrie are acknowledged for their work on the HEALS project. Luc Cluitmans is thanked for providing data management assistance. This study was part of the HEALS project, which was funded from the European Union's Seventh Programme for research, technological development and demonstration under grant agreement No 603946 .
Publisher Copyright:
© 2019
PY - 2020/1
Y1 - 2020/1
N2 - Background/Aim: The exposome includes urban greenspace, which may affect health via a complex set of pathways, including reducing exposure to particulate matter (PM) and noise. We assessed these pathways using indoor exposure monitoring data from the HEALS study in four European urban areas (Edinburgh, UK; Utrecht, Netherlands; Athens and Thessaloniki, Greece). Methods: We quantified three metrics of residential greenspace at 50 m and 100 m buffers: Normalised Difference Vegetation Index (NDVI), annual tree cover density, and surrounding green land use. NDVI values were generated for both summer and the season during which the monitoring took place. Indoor PM2.5 and noise levels were measured by Dylos and Netatmo sensors, respectively, and subjective noise annoyance was collected by questionnaire on an 11-point scale. We used random-effects generalised least squares regression models to assess associations between greenspace and indoor PM2.5 and noise, and an ordinal logistic regression to model the relationship between greenspace and road noise annoyance. Results: We identified a significant inverse relationship between summer NDVI and indoor PM2.5 (−1.27 μg/m3 per 0.1 unit increase [95% CI -2.38 to −0.15]) using a 100 m residential buffer. Reduced (i.e., <1.0) odds ratios (OR) of road noise annoyance were associated with increasing summer (OR = 0.55 [0.31 to 0.98]) and season-specific (OR = 0.55 [0.32 to 0.94]) NDVI levels, and tree cover density (OR = 0.54 [0.31 to 0.93] per 10 percentage point increase), also at a 100 m buffer. In contrast to these findings, we did not identify any significant associations between greenspace and indoor noise in fully adjusted models. Conclusions: We identified reduced indoor levels of PM2.5 and noise annoyance, but not overall noise, with increasing outdoor levels of certain greenspace indicators. To corroborate our findings, future research should examine the effect of enhanced temporal resolution of greenspace metrics during different seasons, characterise the configuration and composition of green areas, and explore mechanisms through mediation modelling.
AB - Background/Aim: The exposome includes urban greenspace, which may affect health via a complex set of pathways, including reducing exposure to particulate matter (PM) and noise. We assessed these pathways using indoor exposure monitoring data from the HEALS study in four European urban areas (Edinburgh, UK; Utrecht, Netherlands; Athens and Thessaloniki, Greece). Methods: We quantified three metrics of residential greenspace at 50 m and 100 m buffers: Normalised Difference Vegetation Index (NDVI), annual tree cover density, and surrounding green land use. NDVI values were generated for both summer and the season during which the monitoring took place. Indoor PM2.5 and noise levels were measured by Dylos and Netatmo sensors, respectively, and subjective noise annoyance was collected by questionnaire on an 11-point scale. We used random-effects generalised least squares regression models to assess associations between greenspace and indoor PM2.5 and noise, and an ordinal logistic regression to model the relationship between greenspace and road noise annoyance. Results: We identified a significant inverse relationship between summer NDVI and indoor PM2.5 (−1.27 μg/m3 per 0.1 unit increase [95% CI -2.38 to −0.15]) using a 100 m residential buffer. Reduced (i.e., <1.0) odds ratios (OR) of road noise annoyance were associated with increasing summer (OR = 0.55 [0.31 to 0.98]) and season-specific (OR = 0.55 [0.32 to 0.94]) NDVI levels, and tree cover density (OR = 0.54 [0.31 to 0.93] per 10 percentage point increase), also at a 100 m buffer. In contrast to these findings, we did not identify any significant associations between greenspace and indoor noise in fully adjusted models. Conclusions: We identified reduced indoor levels of PM2.5 and noise annoyance, but not overall noise, with increasing outdoor levels of certain greenspace indicators. To corroborate our findings, future research should examine the effect of enhanced temporal resolution of greenspace metrics during different seasons, characterise the configuration and composition of green areas, and explore mechanisms through mediation modelling.
KW - Air pollution
KW - Exposome
KW - Greenspace
KW - Noise annoyance
KW - Particulate matter
UR - http://www.scopus.com/inward/record.url?scp=85073825430&partnerID=8YFLogxK
U2 - 10.1016/j.envres.2019.108850
DO - 10.1016/j.envres.2019.108850
M3 - Article
C2 - 31670081
AN - SCOPUS:85073825430
SN - 0013-9351
VL - 180
SP - 1
EP - 13
JO - Environmental Research
JF - Environmental Research
M1 - 108850
ER -
