OP89 A natural experimental study of new walking and cycling infrastructure across the United Kingdom: quantitative analysis of use and physical activity

Abstract

Background Physical inactivity is a growing problem, increasing risks of non-communicable disease. Changing the built environment, including the provision of walking and cycling infrastructure, can influence population levels of physical activity, but high quality evaluations of these interventions are scarce. We demonstrate how different types of data can be used to investigate how context is associated with change in use of new walking and cycling infrastructure, and the association between use of infrastructure and overall physical activity. Methods We combined insights from multiple ‘routine’ and academic data sources. We conducted repeat cross-sectional pre-post analysis of pragmatic monitoring data from 84 new walking and cycling routes in the United Kingdom (the ‘Connect2’ programme), using four-day manual counts (pre n=189,250; post n=319,531), route intercept surveys (next-to-pass) (pre n=15,641; post n=20,253), and automatic count data which generated total estimated annual users, alongside longitudinal data from the iConnect cohort which involved postal questionnaires with local residents near three of the Connect2 schemes at baseline, 1-yr (n=1853) and 2-yr follow-up (n=1524). Multivariable binary logistic regression analysis was conducted using R to analyse: Contextual features associated with increases of at least 50%, and doubling, of pedestrians, cyclists and particular user sub–groups. Models were adjusted for each independent contextual variable and time from infrastructure completion to post–monitoring. Association between use of new walking and cycling infrastructure and meeting physical activity guidelines. Models were adjusted for demographic variables. Cohort data were additionally adjusted for baseline physical activity and scheme. Results The new routes were associated with increased use (median increase in cyclists=52%, pedestrians=38% (p Conclusion Creating new walking and cycling infrastructure may help to increase levels of population physical activity and places with existing low levels of walking and cycling could see the largest relative increases. ‘Routine’ and academic research evaluations involve trade-offs between scale, representativeness of sample and ability to capture within-participant change. Combining findings across data sources can help to understand the impacts of complex interventions on health.