A cohort study of association between greenbelt around schools and non-suicidal self-injury behaviors in children and adolescents

AI Summary

Higher school-area NDVI was associated with reduced risk of adolescent NSSI (HR=0.75; 95% CI 0.67-0.85; P<0.001).
Association significant across genders and urban-rural residence, with protective effects in boys, girls, urban and rural students.
Non-linear negative dose-response between NDVI and NSSI risk; enhancing school greenbelt environments suggested as a potential prevention strategy.

Zhonghua Liu Xing Bing Xue Za Zhi. 2026 May 10;47(5):823-830. doi: 10.3760/cma.j.cn112338-20251211-00892.

ABSTRACT

Objective: To investigate the association between exposure to greenbelt in the vicinity of schools and non-suicidal self-injury (NSSI) in children and adolescents, and provide evidence for the improvement of adolescent mental health via environmental interventions. Methods: A prospective cohort study was conducted in 11 092 children and adolescents without NSSI at baseline surveys from 367 public schools in 13 prefectures in Jiangsu Province, with three follow-up surveys conducted between 2022 and 2024. Greenbelt exposure was assessed by using annual average normalized difference vegetation index (NDVI) within a 1 000-meter buffer zone around each school based on the satellite data from Moderate-Resolution Imaging Spectroradiometer of National Aeronautics and Space Administration, USA. Inverse probability weighting based on propensity scores was used to balance confounding factors, and Cox proportional hazards regression model was constructed to evaluate the association between NDVI and the risk of NSSI in adolescents. Following the Schoenfeld residual test, “duration of outdoor activities”, a variable against the proportional hazards assumption, was treated as a stratification variable for stratified Cox proportional hazards regression analysis. Additionally, restricted cubic splines (RCS) were used to explore the non-linear dose-response relationship between NDVI and NSSI risk, while subgroup analysis and interaction test were conducted. Results: After weighting, baseline characteristics were similar between the exposure groups with high and low NDVI. The Cox proportional hazards regression analysis revealed that NDVI was a protective factor against NSSI (HR=0.75, 95%CI: 0.67-0.85, P<0.001), with robust findings from sensitivity analyses. Subgroup analyses demonstrated that this association was significant in boy students (HR=0.77,95%CI: 0.64-0.92, P=0.004), girl students (HR=0.74, 95%CI: 0.63-0.87, P<0.001), urban students (HR=0.73, 95%CI: 0.63-0.85,P<0.001), and rural students (HR=0.79, 95%CI: 0.65-0.97, P=0.026). RCS analysis indicated a significant non-linear negative correlation between NDVI and NSSI risk (non-linearity test P<0.001). Conclusions: Higher greenbelt exposure around schools was a protective factor for NSSI in adolescents, and this association was significant across subgroups defined by gender and urban-rural residence. Optimizing greenbelt environments in and around schools might serve as a potential public health strategy for prevention of NSSI in adolescents.