Author(s): ZHANG, Z., YAO, X., Institution: PEKING UNIVERSITY, Country: CHINA, Abstract-ID: 2434

Background: Urban parks serve as important venues for promoting physical activity (PA). However, findings regarding the relationship between park proximity and PA have been inconsistent, and most studies have examined PA as a unitary construct without differentiating intensity levels. Additionally, discrepancies between perceived and objective proximity measures have been reported among general community residents, yet less is known about this correspondence among actual park users. This study aimed to examine the agreement between perceived and objective proximity measures among urban park users, and investigate the associations between both proximity measures and within-park PA across different intensity levels.
Methods: This cross-sectional study was conducted in October 2023 across four urban parks in Beijing, China. A total of 399 park users (52.1% female; mean age = 46.2 years) completed questionnaires assessing perceived proximity (a single-item question measuring perceived closeness to the park), demographic characteristics, and within-park physical activity using a modified International Physical Activity Questionnaire adapted for park settings. Objective proximity was measured as Euclidean distance from participants' residences to park centers using GIS and categorized into three groups: within 0.5 mile, 0.5–1 mile, and greater than 1 mile. Correspondence between proximity measures was assessed using Kendall rank correlation and ANOVA. Multilevel linear regression models were used to examine the associations between proximity measures and within-park walking, moderate-intensity PA, and vigorous-intensity PA, controlling for sociodemographic covariates, with bootstrap methods employed for non-normally distributed outcomes.
Results: Moderate correspondence was found between perceived and objective proximity measures (Kendall's τ = 0.474, p < .001). Both proximity measures demonstrated significant negative associations with moderate-intensity PA and walking within parks. Notably, a threshold effect emerged at approximately 1 mile: participants living beyond 1 mile reported significantly reduced moderate-intensity PA (β = -0.39, 95% CI [-0.67, -0.11]) and walking (β = -0.58, 95% CI [-0.84, -0.24]), while no significant differences were found between the within-0.5-mile and 0.5–1-mile groups. Neither perceived nor objective proximity was significantly associated with vigorous-intensity PA within parks.
Conclusion: These findings suggest that proximity may play a more important role in promoting moderate-intensity PA and walking within parks, while vigorous-intensity PA may be more influenced by other park characteristics beyond distance. Public health initiatives may benefit from increasing park awareness and engagement to bridge perception gaps and promote within-park PA among urban residents.