Urban Green Space and Mental Health: Mediating Roles of Physical Activity and Social Cohesion

Purpose This study investigates the relationship between urban green spaces (UGS) and residents’ mental health in Dhaka, Bangladesh, with a focus on the mediating roles of social cohesion and physical activity. Design Multilevel regression and mediation analysis with cross-sectional data were constructed. Setting 10 UGS in Dhaka were selected based on spatial variation. Sample A total of 276 adult residents (response rate: 85.2%) were surveyed using quota sampling to ensure demographic diversity. Measures The Short Warwick-Edinburgh Mental Wellbeing Scale (SWEMWBS) was used for mental health assessment, validity scales for social cohesion and IPAQ-SF for physical activity. UGS accessibility and quality were evaluated using NDVI, GIS buffers, and on-site audits. Analysis Multilevel linear regression and Sobel tests were employed to examine direct and mediated effects of UGS on mental health. Results UGS-level factors explained 18% of the variance in mental health. UGS accessibility (β = 0.285***), safety (β = 0.212***), and amenities (β = 0.220***) were positively associated with mental health. Social cohesion (Z = 2.82***) and physical activity (Z = 2.15**) partially mediated these relationships. Conclusion UGS enhances mental health in dense urban environments through both direct and mediated pathways. Subgroup analysis reveals that high and medium-quality UGS have more substantial positive effects through physical activity and social cohesion on mental health. In contrast, the impact of low-quality UGS’s is minor and insignificant. Respondents with higher education and income benefit more from UGS on their mental health. Future research using longitudinal data or experimental designs could enhance the robustness of causation.