Urban green space as a public health intervention: a systematic review of causal inference methods in observational studies

In the context of rapid urbanization, urban green spaces (UGS) are widely promoted as a lever for improving population health, yet a central debate remains: are reported health benefits an urban myth or a substantiated causal reality? We systematically review recent empirical studies that apply causal inference designs to estimate the health effects of UGS, including matching, synthetic control method (SCM), instrumental variable (IV), difference-in-difference (DID), and regression discontinuity design (RDD). We identified 37 records, of which 33 were empirical studies (excluding 4 Research protocols). Risk of bias was assessed using ROBINS-I, and we scrutinized whether key identification assumptions were explicitly addressed for each design. Across the 33 empirical studies, 30 report a positive causal effect of UGS on at least one health outcome. Crucially, this positive consensus remains consistent among the most methodologically rigorous papers, such as those meeting all key DID assumptions or effectively combining PSM and DID. Rather than merely echoing the limitations of data and design, this review underscores that when endogeneity and self-selection bias are strictly controlled, the health-promoting role of UGS remains highly credible. Future studies should integrate multiple approaches, leverage big data and machine learning technologies, and foster interdisciplinary collaborations to enhance causal inference accuracy and provide scientific guidance for UGS planning and public health policy.