2022 Volume 77 Issue 5 Pages I_395-I_405
In this study, the authors focus on street characteristics around residences where actual walking activities take place, and analyze them together with the geographic environmental variables of the existing Walkability Index (WI). The authors aim to clarify the environmental factors that contribute to walking for specific purposes and the optimal analytical frameworks for the purposes. Thorough the classification of streets by focusing on cross-sectional configuration and roadside features, and the regression analysis for the amount of activity of daily-errands walking and leisure walking, the authors demonstrate that: (1) For daily-errands walking, street characteristics had a greater effect than existing WI. For leisure walking, parts of street characteristics were significant. Thus, combining street characteristics with each other or synthesizing WI with street characteristics may allow us to construct environmental variables with larger odds ratios. (2) For daily-errands walking, a single model using the median binarized walking activity can be used for analysis. On the other hand, for leisure walking, it is necessary to set up models separately for the presence or absence of walking and the amount of walking.
