Abstract
One potential strategy to promote physical activity (PA) involves changing the neighborhood environment. The PA-friendliness of neighborhood environments is often calculated using geospatial data. However, the association with perceived PA-friendliness seems to be low. Therefore, we examined the relationship between two Dutch geospatial measures and residents' perceptions regarding the PA-friendliness of their neighborhoods. Data from 3438 respondents aged 40-80 years from the Doetinchem Cohort Study were linked to individual geospatial data. In addition to respondents' self-reports, we used the following two geospatial measures: the Dutch walkability index and the Dutch indicator for PA-friendly environments (KBO-indicator). We performed logistic regression analyses to assess associations between perceived PA-friendliness and two objective geospatial measures, including interactions for age, sex, education, work status, and physical functioning. The majority of respondents (83%) evaluated their neighborhood as PA-friendly. The logistic analyses revealed no associations between the geospatial measures of walkability and PA-friendliness and the individuals' perception measures. Similarly, a comparison of residents from high versus low walkable or PA-friendly neighborhoods demonstrated no significant differences in their perception of PA-friendliness. Additionally, no significant interaction effects were observed with sex, age, education, employment status, or physical functioning, indicating that even among subgroups there was no correspondence between objective and subjective measures. The lack of correspondence between both objective and subjective measures for PA-friendliness in the neighborhood suggests that these are distinct constructs. Future research should focus on qualitative methods to bridge the gap between objective and subjective measures and test whether the perceived PA-friendliness is similar to the objective measures.