2017 - Research

The Relationship Between Objectively Measured Walking and Risk of Pedestrian—Motor Vehicle Collision

This is the first study that combines GPS, accelerometer, and travel diary data to measure pedestrian exposure to collision risk.

Key takeaway:

  • This is the first study that combines GPS, accelerometer, and travel diary data to measure pedestrian exposure to collision risk. The researchers found that people typically walked in areas with low collision risk when they were walking for recreation, walking at faster paces, or going on longer walks. This suggests that pedestrians moderate collision risk by using lower-risk routes. Improving road safety for pedestrians may promote more walking for both transportation and physical activity, thereby merging transportation safety and active transportation goals.

 

Results:

  • People who walk for recreation, walk at faster paces, or walk for longer durations tend to walk in areas where the risk of traffic collision is lower.
  • Vehicle owners, residents of single-family homes, and parents of young children also typically walked in areas with low collision risk. This reflects how:
    • Safety is a key concern for people with children;
    • People who can afford to own cars and live in single-family homes may live in areas where there is a lower risk of pedestrian-vehicle collisions;
    • People who own cars also have more opportunities to drive to places where it is safer to walk (i.e., recreational trails, streets with less traffic, etc.).
    • Men walked in areas with a higher collision risk (i.e., areas with higher road speeds) compared to women, which may explain why they’re overrepresented in pedestrian injuries/fatalities.
    • The researchers found no significant differences on pedestrian-vehicle collision risk based on age or race/ethnicity, but that could be due to the study’s small sample size.
    • Black, Hispanic, American Indian/Alaskan Native people tend to have higher risks of collisions than white people, which may be explained by socioeconomic and neighborhood factors (i.e., people of color may live in neighborhoods with more high-volume and/or fast-moving vehicular traffic, which makes it less safe to walk).   

 

Implications:

  • This study suggests that people who walk more are more aware of safer walking routes and therefore moderate their risk of being hit by a car. Since people’s perceptions of safety inform their modal and route choice, future studies involving travel diaries can have people report perceptions of safety with respect to traffic as well as personal safety. After all, qualitative and subjective aspects of people’s experiences walking matter, too.  
  • Policymakers and practitioners should create supportive walking environments (i.e., improve traffic safety by lowering speeds, extending sidewalks or creating pedestrian islands on longer roads, etc.) to reduce the risk of pedestrian-vehicle collisions and encourage people to walk more for transportation and health.

 

 

Methods:

  • In this study, there were 537 participants who were all previously recruited subjects in the Travel Assessment and Community Study in King County, Washington (Seattle) in 2008–2009. They were given a portable GPS and an accelerometer, and kept travel diaries, therefore enabling data collection from quantitative and spatial methods as well as personal, qualitative accounts. The researchers superimposed people’s walking patterns onto maps of the historical probabilities of pedestrian-vehicle collisions for intersections and midblock segments in Seattle. They used linear mixed models to calculate the mean risk of pedestrian-vehicle collision in specific walking locations based on walking exposure (duration, distance, and intensity) and participant demographics (i.e., age, gender, race, etc.).

 

Quitsberg, D.A.; Howard, E.J.; Hurvitz, P.M.; Moudon, A.V.; Ebel, B.E.; Rivara, F.P.; and Saelens, B.E. (2017). The Relationship Between Objectively Measured Walking and Risk of Pedestrian—Motor Vehicle Collision. American Journal of Epidemiology, 185(9).