Walking and cycling for commuting, leisure and errands: relations with individual characteristics and leisure-time physical activity in a cross-sectional survey (the ACTI-Cités project)

International audienceBackground: Increasing active transport behavior (walking, cycling) throughout the life-course is a key element of physical activity promotion for health. There is, however, a need to better understand the correlates of specific domains of walking and cycling to identify more precisely at-risk populations for public health interventions. In addition, current knowledge of interactions between domains of walking and cycling remains limited. Methods: We assessed past-month self-reported time spent walking and cycling in three specific domains (commuting, leisure and errands) in 39,295 French adult participants (76.5 % women) of the ongoing NutriNet Santé web-cohort. Multivariate logistic regression models were used to investigate the associations with socio-demographic and physical activity correlates. Results: Having a transit pass was strongly positively associated with walking for commuting and for errands but was unrelated to walking for leisure or to all domains of cycling. Having a parking space at work was strongly negatively associated with walking for commuting and cycling for commuting. BMI was negatively associated with both walking for leisure and errands, and with the three domains of cycling. Leisure-time physical activity was negatively associated with walking for commuting but was positively associated with the two other domains of walking and with cycling (three domains). Walking for commuting was positively associated with the other domains of walking; cycling for commuting was also positively associated with the other domains of cycling. Walking for commuting was not associated with cycling for commuting.Conclusions : In adults walking and cycling socio-demographic and physical activity correlates differ by domain (commuting, leisure and errands). Better knowledge of relationships between domains should help to develop interventions focusing not only the right population, but also the right behavior