Urban forest characteristics and childhood asthma onset

Abstract

The influence of the urban forest on asthma onset is unclear as trees reduce air pollution but also emit pollens and biogenic organic compounds. Additionally, rudimentary vegetation data have so far been used to assess risks. We aimed to assess associations between residential urban forest characteristics and the onset of childhood asthma. We used an open cohort of all children born in Montreal (Canada) between 2000 and 2015 and followed up to 12 years, created from linked medico-administrative databases. New cases of asthma were defined with a validated algorithm. Coniferous and deciduous tree crown volumes were estimated with Lidar point cloud data and random forest models. Total volumes of coniferous and deciduous tree canopies for pollen/summer seasons were calculated for buffers around participants’ residential postal codes.Cox models (age as the time axis, stratified for sex) for the residential volumes of coniferous and deciduous canopies were developed; splines were used for non-linearity. Models were adjusted for calendar year, and neighborhood social and material deprivation. The cohort included 352,966 children with 30,825 new cases of asthma, and an overall incidence rate of 18 new cases per 1,000 person-years. Volumes of deciduous and coniferous tree canopies in 100 m buffers around postal codes ranged 0 - 465,686 m3 and 0 - 64,198 m3 respectively. Volumes of coniferous and deciduous tree canopies had non-linear relationships with asthma onset with null or negative slopes at low levels, becoming positive at higher volumes. Preliminary results suggest complex relationships between urban forest characteristics and childhood asthma onset that deserve investigation to enhance understanding of ecosystem services (i.e. benefits to people) of urban trees.