Perceptions of the environment moderate the effects of objectively-measured built environment attributes on active transport. An ACTI-Cités study

International audience; Purpose: Both perceived and objectively-measured attributes of the built environment have been related to active transport. They do not refer, however, to the same construct: objective measurements provide an inventory of a neighborhood?s active-friendly attributes, while the perceptions of individuals refer to their appraisal of their neighborhood opportunities for active travel. Both aspects are complementary in explaining active transport. This study examines the independent associations of perceived and objective environmental attributes with active transport, and whether perceptions moderate the objectively-measured environment active transport relationship.Methods: Based on a sample of 2670 participants from the NutriNet-Sante? Study (France), we examined (i) the concordance between perceptions and objective measurements of residentialbased environmental attributes using kappa statistics, (ii) the associations between perceived and objectively-measured attributes and active transport by trip purpose (commuting and errand) using zero-inflated negative-binomial regressions, and (iii) the potential moderation effect of perceptions on the associations between objectively-measured environmental attributes and active transport. Information related to demographics, perceptions and transport modes was collected in 2013.Results: A poor to fair level of agreement was observed between perceived , objectively -measured environmental attributes; both were associated with active transport for commuting and errands. Environmental correlates differed between the likelihood to engage in active transport and the time spent in transport. Perceptions positively moderated the associations be-tween active transport and objectively-measured walkability and access to facilities. The differ-ences in predicted active transport time between a high versus low perception of the environmental attributes for a given neighborhood ranged from 20 to 55 min for active commuting , from 84 to 37 min for active transport for errands.Conclusion: Associations between objectively-measured environmental attributes and active transport are dependent on individuals & rsquo; perceptions of opportunities provided by their residential neighborhood. Increasing individuals & rsquo; awareness of active-friendly environmental attributes may be crucial for enhancing active transport.