Neighborhood Walkability and Income: Associations with Walking for Leisure and Transportation: 644: May 31 8:45 AM 9:00 AM

Sallis, James F. FACSM; Saelens, Brian E.; Frank, Lawrence D.; Slymen, Donald J.; Conway, Terry L.; Cain, Kelli; Chapman, James C.

Medicine & Science in Sports & Exercise: May 2007 - Volume 39 - Issue 5 - p S31
doi: 10.1249/
C-13 Free Communication/Slide - Environmental and Community Physical Activity Considerations: MAY 31, 2007 8:00 AM - 10:00 AM ROOM: 335

1San Diego State University, San Diego, CA

2University of Washington, Seattle, WA

3 University of British Columbia, Vancouver, BC, Canada

4Lawrence Frank & Co, Atlanta, GA.


PURPOSE: The study tested the hypothesis that walking for both leisure and transportation purposes is higher in people living in high walkability neighborhoods.

Method: Thirty-two neighborhoods were selected in Seattle-King County, Washington, and Baltimore-Washington, DC regions that varied in “walkability” and income. High walkability means neighborhoods have high residential density, street connectivity, and mixed land use, consistent with the potential for active transportation. Adults aged 20–65 years were recruited from study neighborhoods (n=2199; 48% female; mean age=45 years; 26% ethnic minority). Minutes of walking for leisure and transportation purposes in the past week were reported separately using items from the long form of the International Physical Activity Questionnaire. Neighborhood walkability was objectively measured with land use and street network data using Geographic Information System software. Neighborhood income was based on census data. Multilevel mixed effect regressions were conducted, accounting for clustering within neighborhood and covarying for individual-level demographic variables including age, gender, education, ethnicity, and time at address.

RESULTS: There was a significant walkability X income interaction (p<.027) for walking for transportation, indicating a greater positive effect of walkability in higher income neighborhoods. Overall, participants living in high-walk ability neighborhoods walked significantly more for transportation than those living in low-walkability neighborhoods (adjusted means, 44.3 vs 12.8 min/day). Walking for leisure also was significantly higher (p<.012) for residents living in highwalkability versus low-walkability neighborhoods (adjusted means, 18.5 vs 14.2 min/day); neither the income effect nor the interaction was significant for leisure walking.

CONCLUSION: In most studies, walkability has been associated only with walking for transportation, so the present finding of more walking for leisure in high walkable neighborhoods indicates walkability supports physical activity for multiple purposes. Reasons for the smaller effect of neighborhood walkability on walking for transportation among low income participants need to be explored.

© 2007 American College of Sports Medicine