Street-View Greenspace Exposure and Childhood Asthma-Related Outcomes: Analyses in Project Viva

Abstract

Background and Aim: While higher levels of greenspace can mitigate air pollution, reduce stress, and increase physical activity, its role in childhood asthma is unclear. The aim of this study is to examine the associations of neighborhood greenspace based on Google street view (GS) images with asthma-related measures among children in Project Viva, a cohort in eastern Massachusetts participating in the Environmental influences on Child Health Outcomes (ECHO) consortium. Methods: We leveraged GSV images from 2007 to 2011 within 250m of participants’ home addresses at mid-childhood and applied deep learning algorithms to derive unique metrics of ground-level greenspace exposure. We calculated total greenspace exposure, which combined %grass, %trees, %plants, %fields, and %flowers within each image. Mothers reported the child’s asthma status in mid-childhood (median 7.7 yr) and early adolescence (median 12.7 yr). Trained research assistants performed lung function tests (FEV1 and FVC) using spirometry and collected blood for total serum IgE measurement. We constructed linear and logistic regression models to examine associations of neighborhood greenspace with these asthma-related outcomes adjusting for child’s sex, race/ethnicity, age at outcome, maternal and paternal education, marital status and household income (at enrollment), and census tract level household income and urbanicity (at delivery). Results: In adjusted cross-sectional analyses (N=486), each interquartile range increase in total greenspace was associated with lower FEV1 (-44.7 mL [95% CI: -82.7, -6.64]) and higher total IgE (56.4 kU/L [95% CI: 0.86, 112]). Similar associations were observed for %tree but not for %grass. Greenspace was not associated with asthma status cross-sectionally in mid-childhood. There were no longitudinal associations between any greenspace measures in mid-childhood and any asthma-related outcomes in early adolescence (N=414). Conclusions: We observed cross-sectional but not longitudinal associations between GSV-based greenspace with FEV1 and total serum IgE in adverse directions and found no association between greenspace and asthma status.