03, 95% CI = 1.01, 1.05), the presence of a school inhibitors crossing guard (IRR = 1.14, 95% CI = 1.07, 1.21) and primary language other than English (IRR = 1.20, 95% CI = 1.05, 1.36) were associated with more walking. Child population density, traffic lights and school crossing guards exhibited the most significant associations. Effect modification was evident only for school crossing guard (Table 4). With no crossing guard present, walking proportions BLU9931 research buy were positively associated with environmental variables and negatively associated with poor weather. Lower IRRs were evident when crossing guards were present, except for child population density. This is the first large

study to correlate direct observational counts of walking to school with objective built environment data. The mean proportion of observed walking was high at 67%; with large variability between schools. The mean proportion of other active modes (i.e. cycling and scootering) was 1.7%. On average, 31% of children arrived by car. Previous population-based national and local Canadian surveys reported 50–55% of children walking to school (Buliung et al., 2009 and Cragg et al., 2006). The higher

proportions in this study were likely due to sampling children within 1.6 km of schools, whereas previous estimates were not restricted to children living within walking distance. Observed proportions were also higher than those in Australia and the Quizartinib U.S., where approximately 48% of children living within walking distance reported walking to school (Martin et al., 2007 and Salmon et al., 2007). Strong associations with walking were found for child population density and traffic lights, which validated previous findings (Braza et al., 2004, Bringolf-Isler et al., 2008, Mitra et al., 2010b, Salmon et al., 2007 and Timperio et al., 2006). In addition to the strong positive association found between walking and school crossing guards, there was evidence of crossing guards acting as an effect modifier between the environment and

walking which has not been previously reported. With a school Mephenoxalone crossing guard present, other built and social environmental factors had less impact on walking which has important implications for potential interventions. Although road design features may be more easily modified in existing neighborhoods than those related to population density and land use, roadway modification can be a highly contested, politicized process. The process to install crossing guards is much simpler in Toronto, and involves a reported need by the community to the Toronto Police, followed by an assessment of the location. If the presence of school crossing guards overrides other negative effects of the built and social environments on walking, adding crossing guards may a feasible and effective method to increase walking proportions.