Abstract Emerging research highlights the correlations between green spaces and physical activity, with increasing attention being paid to the role of campus green spaces. Despite the significance of greenways in campus exercise environments, the correlations between their green spaces and built environmental indicators with physical activity remain poorly understood. Previous research has mainly consisted of small‐scale surveys, with a notable absence of large‐scale physical activity tracking studies. This study bridges the gaps by conducting a large‐scale analysis of 121 greenways across 38 campuses in Shanghai, utilizing 700,742 trajectory data points to delineate jogging, walking and cycling activities in around 7 years. We employed multi‐dimensional green exposure assessments, encompassing Normalized Difference Vegetation Index (NDVI), Green View Index (GVI) of trees, shrubs and grass, and canopy height, derived from street view and remote sensing data. Utilizing negative binomial regression within a generalized linear model and incorporating quadratic terms, this study identified both linear and non‐linear exposure relationships. Furthermore, the study investigated the moderating effect of the urban–suburban variable in the green spaces and physical activity exposure‐response model. The study findings indicate that: (1) The GVI significantly correlates with jogging and walking in an inverted U‐shaped exposure‐response relationship, reaching a peak at a threshold around 0.27; (2) Cycling behaviour is significantly associated with canopy height; (3) The visibility of trees and shrubs positively correlates with physical activity, while grass exhibits no significant correlation; (4) The urban–suburban variable moderates the relationship between NDVI and jogging activity. This research clarifies the nexus between greenway environment and various physical activities, explores the dynamic interplay between people and nature and offers valuable insights for campus greenway development and policy‐making. Read the free Plain Language Summary for this article on the Journal blog.
Mao et al. (Mon,) studied this question.