A natural experimental study of new walking and cycling infrastructure across the United Kingdom: The Connect2 programme
Journal of transport & health 2021 ; 20: 100968.
Le Gouais A, Panter J, Cope A, Powell J, Bird EL, Woodcock J, Ogilvie D, Foley L, iConnect consortium
DOI : 10.1016/j.jth.2020.1009687
URL : https://doi.org/10.1016/j.jth.2020.100968
High quality evaluations of new walking and cycling routes are scarce and understanding contextual mechanisms influencing outcomes is limited. Using different types of data we investigate how context is associated with change in use of new and upgraded walking and cycling infrastructure, and the association between infrastructure use and overall physical activity.
We conducted repeat cross-sectional pre-post analysis of monitoring data from a variety of walking and cycling routes built in 84 locations across the United Kingdom (the Connect2 programme, 2009–2013), using four-day user counts (pre n = 189,250; post n = 319,531), next-to-pass surveys of route users (pre n = 15,641; post n = 20,253), and automatic counter data that generated estimates of total annual users. Using multivariable logistic regression, we identified contextual features associated with 50% increase and doubling of pedestrians, cyclists, and sub-groups of users. We combined insights from monitoring data with longitudinal cohort data (the iConnect study) from residents living near three Connect2 schemes. Residents were surveyed by post at baseline, one-year (n = 1853) and two-year follow-up (n = 1524) to investigate associations between use of the new infrastructure and meeting physical activity guidelines.
The routes were associated with increased use (median increase in cyclists 52%, pedestrians 38%; p < 0.001). Large relative increases were associated with low baseline levels (e.g. odds of doubling cycling were halved for each additional 10,000 annual cyclists at baseline: OR 0.52, 95% CI 0.31, 0.77). Use was associated with meeting physical activity guidelines in both repeat cross-sectional and longitudinal analyses (users vs. non-users after one year, OR 2.07, 95% CI 1.37, 3.21; after two years, OR 2.00, 95% CI 1.37, 2.96).
This examination of use, users, benefit-cost ratios, and physical activity associated with new walking and cycling infrastructure across contexts, using multiple types of data, suggests that building walking and cycling infrastructure could improve population health and reduce inequalities.