Location or design? Associations between neighbourhood location, built environment and walking

In examining the association between environmental exposures and walking, conducting research on a neighbourhood scale has been the dominant approach whereas the association of the regional-scale environment with behaviours has rarely been explored. Because regional location and neighbourhood built environment attributes are likely to be correlated, the findings in neighbourhood-scale studies may be biased. In contrast to existing literature, this study is based on the assumption that a neighbourhood’s location may be associated with walking or physical activity and that this association may be separately identifiable from the influence of the neighbourhood built environment on behaviours. The findings indicated that residing in a highly urban location had a consistently positive association with walking and transportation-purpose physical activity when the neighbourhood built environment and individuals’ socio-demographic factors were controlled. Meanwhile the inclusion of the neighbourhood location variable did not result in significant changes to the models for recreation-purpose activity.