Greater proximity to greenspace is associated with fewer depressive symptoms among urban middle-aged adults

Understanding the factors that influence depression is crucial for both its treatment and prevention. Residential greenspace has received increasing attention for its potential role in mitigating depressive symptoms. However, traditional statistical methods are fundamentally constrained to evaluating simple bivariate associations between greenspace exposure and depression risk, overlooking the complex interactions among key health-related factors and failing to comprehensively assess the relative importance of greenspace compared to other contributing factors.

We employed machine learning (ML) methods to assess the effects of multiple factors, including greenspace, on the development of depressive symptoms among 1,542 urban middle-aged adults in the China Health and Retirement Longitudinal Study, with depressive symptoms measured using the 10-item Center for Epidemiological Studies Depression Scale. Individual greenspace exposure for each subject was quantified using vegetation indices obtained from the NASA satellite database.

Among the five ML models, the Support Vector Machines model achieved the best predictive performance, with an AUC of 0.776. Life satisfaction, EVI_250m (a measure of greenspace), social activity, pain, sleep duration, and impairment in instrumental activities of daily living were significant predictors of depressive symptoms. EVI_250m, identified by the ML model, indicated that closer proximity to greenspace was associated with a greater reduction in depressive symptoms. Furthermore, Shapley Additive Explanations values confirmed that greenspace ranked among the top influential factors.

Residential greenspace is associated with fewer depressive symptoms and ranks prominently among the most influential factors within the examined health-related variables. The potential beneficial association with greenspace appears to be stronger with closer proximity to residential areas.