Uncovering the depressive symptom network in self-harming rural children: a Bayesian undirected network analysis

Self-harm among children is a growing public health concern, particularly in rural settings where unique socioeconomic and environmental factors increase vulnerability. This study explored depressive symptom networks in rural children with and without self-harm behaviors, using Bayesian undirected network analysis to identify symptom-level differences and inform targeted interventions. Data were collected from 2,009 rural elementary school students (Mage = 10.73 years, SD = 1.65) across seven Chinese provinces. Depressive symptoms were assessed using the Child Depression Inventory, and self-harm behaviors were identified through self-reported measures. Bayesian undirected network analysis compared depressive symptom networks between children with (N = 417) and without (N = 1592) self-harm behaviors. The results revealed that children with self-harm behaviors exhibited less densely connected symptom networks, with Anhedonia, Feeling Unloved, and Somatic Concerns as the most central symptoms. In contrast, children without self-harm behaviors displayed denser, strongly clustered networks, with Anhedonia, Permissive Worrying, and Somatic Concerns as central symptoms. Significant differences in network density and clustering coefficients were observed between the two groups. These findings suggest that self-harming children experience more fragmented emotional and social dynamics, while non-self-harming children demonstrate stronger symptom interconnectivity that may support better emotional regulation. These findings emphasize the need for rural-specific mental health policies and services, including school-based programs, telepsychiatry, and community-driven efforts to reduce mental health stigma and improve early detection of depressive symptoms and self-harm behaviors. By addressing the unique challenges of rural communities, this study contributes to improved mental health outcomes and supports the development of evidence-based, symptom-focused interventions for underserved populations.