Article Text

Download PDFPDF
41 Walking route safety and objectively measured walking: a multilevel analysis
  1. D Alex Quistberg1,
  2. Beth E. Ebel1,
  3. Philip M. Hurvitz2,
  4. Eric J. Howard2,
  5. Anne V. Moudon2,
  6. Brian E. Saelens3,
  7. Fred P. Rivara1
  1. 1Harborview Injury Prevention & Research Center, University of Washington, USA
  2. 2Urban Form Lab, University of Washington, USA
  3. 3Center for Child Health, Behavior and Development, Seattle Children's Hospital, USA


Statement of purpose A safe walking environment may improve public health by facilitating environmental health, physical activity and injury prevention. We studied the walking behaviour of adult pedestrians to examine the relationship between walking and walking route safety.

Methods A cross-sectional study of adults in King County, WA in 2007–2008 assessed walking using accelerometry with spatial location via Global Positioning System (GPS) devices and travel diaries. GPS and accelerometer readings were logged every 30 s for up to 7 days. Participants were included if they had at least one walking “bout” in Seattle, WA city boundaries (N = 548). Bouts were spatially matched to Seattle intersection and mid-block locations. The historical probability of a pedestrian-motor vehicle collision during 2007–2013 at each intersection and mid-block was predicted in a random-effects Poisson model which included road characteristics (e.g., traffic volume), built environment features (e.g., sidewalks) and neighbourhood characteristics (e.g., residential density). Multilevel mixed-effects linear regression was used to assess the average safety of walking routes (mean exposure to pedestrian collision risk) and individual walking (accelerometer counts, bout length, and bout time) and demographic characteristics.

Results Pedestrians who walked longer had safer walking routes (−2.65% historical probability of pedestrian collision along the route per 10 min of walking, 95% CI −3.54%, −1.76%). Female pedestrians more commonly had safer walking routes relative to males (−1.10%, 95% CI −0.17%, −2.02%), as did people who owned at least one motor vehicle (−3.98%, 95% CI −5.07%, −2.90%).

Conclusions These findings may indicate that people who walk more are significantly less likely to walk in areas with a higher risk of pedestrian collisions. Pedestrians who walk more may avoid areas that are perceived or experienced as higher risk to pedestrians.

Significance and contributions Our findings suggest that environmental changes to improve safety for walkers may encourage pedestrian activity.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.