UNASSIGNED: The audit-based school walkability index was developed based on all street segments (n=841) within a 0.4km network buffer from each study school (n=18). The GIS-based school walkability, a combination of road connectivity, vehicular traffic exposure, and residential density, was also measured in a 2km network buffer around each school. The participants were individuals aged 8-11 years (n=315) who participated in the Walking School Bus randomized controlled trial project. Mixed-effects logistic and linear models were used to examine the association of the index\'s representations of the built environment with children\'s school travel mode (walking or biking to school 1+ times per week) and with objectively measured moderate-to-vigorous physical activity (MVPA, average weekday minutes during the 90-min before-school period). These associations were tested with the total sample as well as the subsample of children living within 1.5km from their schools.
UNASSIGNED: The audit-based school walkability index (WI) was positively associated with both active commuting to school among the subsample living within 1.5km from their schools and with children\'s before-school MVPA among the subsample and the total sample. The GIS-based school WI showed significant associations with children\'s before-school MVPA but no relationships with active school travel among the subsample and the total sample.
UNASSIGNED: The audit-based school walkability index can be used as a complementary tool for measuring walkability near low-income elementary schools along with existing GIS-based school walkability index.