Symptom network for psychological distress in college freshmen: a large sample bayesian network analysis

College students show a high prevalence of psychological distress, particularly during the freshman year. Investigating the network structure of psychological distress symptoms can reveal both comorbid patterns and potential directional relationships among symptoms. Using data from a large sample of Chinese university students (N = 14,372), this study estimated a Gaussian graphical model (GGM) and a Bayesian network to explore undirected and directed connections among the nine core dimensions of the Symptom Checklist-90 (SCL-90). Anxiety exhibited the highest expected influence (EI) in the GGM, followed by interpersonal sensitivity and depression. In the DAG, depression and anxiety were located upstream in the network, with depression identified as the primary parent node, suggesting that emotional distress may serve as the “trigger” for the entire symptom network. Furthermore, interpersonal sensitivity demonstrated the highest betweenness and occupied a midstream position in the DAG, potentially functioning as a key hub linking other symptoms. Phobic anxiety emerged as a stable downstream node. These findings emphasize the importance of depression, anxiety, and interpersonal sensitivity in the network of psychological distress symptoms and highlight the value of early cognitive-emotional interventions and transdiagnostic strategies targeting symptom interactions. This study did not collect detailed demographic information from participants or include other mental health scales for comparison, which limits the generalizability of the results.