Active School Transportation and the Built Environment across Canadian Cities: Findings from the Child Active Transportation Safety and the Environment (CHASE) Study
Author(s): Rothman, Hagel, Nettel-Aguirre, Cloutier, McCormack, Howard, Macpherson, Fuselli, Ling, Buliung, HubkaRao, Zanotto, Rancourt, Winters
Slidedeck Presentation:
Abstract:
Background:
Walking and bicycling to school (active school transportation, AST) has been in decline for decades in North America and globally with the rise of automobility. An important factor influencing AST is the roadway built environment. This cross-sectional study from the Child Active-Transportation Safety and the Environment (CHASE) program of research assessed associations between the built environment and AST across six Canadian cities and one regional municipality and provides insights into built environmental features most support of AST.
Aims:
To examine the associations between proportions of children who use AST, and features of the built environment.
Methods:
We conducted an observational study in spring 2018 and 2019, at 552 publicly funded elementary schools (JK-grade 8) in Montreal (n=67), Laval (n=50), Toronto (n=76), Peel (n=71), Calgary (n=125), Vancouver (n=67), and Surrey (n=96), Canada. Trained observers counted children arriving at school by car or active modes during morning drop off time. The proportion of children using AST was calculated from the total number of children observed, excluding those arriving by school bus. Built environment features were included as explanatory variables related to density, land use diversity and design, based on spatial databases and site audits. The proportion of AST was modelled for all cities using random effects beta regression, and then separate models developed for each city.
Results:
Travel mode was recorded for ~118,000 students. Across all schools, the average proportion of AST was 54.3% (SD 18.9%), ranging from 39.5% (SD 22.1%) in Laval, Quebec to 69.7% (SD 18.1%) in Montreal, Quebec. Across all cities, higher odds of AST were associated with higher child population density (OR: 1.08 95% CI 1.05, 1.11), cycling infrastructure (OR: 1.05; 95% CI: 1.00, 1.11), school crossing guards (OR: 1.13 ; 95% CI: 1.01, 1.28), traffic signal density (OR: 1.44; 95% CI: 1.04, 2.00), and local road density (OR: 1.10; 95% CI: 1.04, 1.17).. Increased residential land use was associated with a lower odds of AST (OR 0.96; 95% CI: 0.94, 1.00). While the magnitude varied across the city-specific models, increased child population density was associated with more AST in each city, and more residential land use, higher Walk Score®, local road density, and school crossing guards were associated with more AST in most cities.
Discussion:
This was a large-scale study, considering a broad range of built environment correlates across a diversity of contexts. Given the ecological observational design, data on individuals were not available; however, ecological studies can point to population-based interventions with the greatest potential for a sustained public health impact.
Conclusions:
Several cities had upwards of 60% of children using AST, on average, but the high degree of variability between schools and between cities suggests opportunities to increase AST. Results underscore the substantial differences in built environments across Canadian cities, and point to the potential of local interventions, instead of universal “one size fits all” approaches.