Mental health problems among university students are increasingly severe, with symptoms such as depression, anxiety, loneliness, and suicidal ideation frequently co-occurring to form complex symptom networks. This study systematically analyzed association patterns of mental health symptoms among UK university students through multilayer network analysis. A cross-sectional survey was conducted with 1, 285 students from five UK universities, who are assessed using eight validated psychometric instruments evaluating depression, anxiety, mania, sleep quality, stress, suicidal ideation, psychotic experiences, and loneliness. A dual-level network analysis approach was employed, constructing both a scale-level network with 8 nodes to identify macro-association patterns and an item-level network with 33 nodes for in-depth analysis of depression, loneliness, and suicidal ideation connections. The EBICglasso algorithm estimated network structure, and key symptoms were identified through centrality indices. The scale-level network revealed depressive symptoms as most prominent across all centrality indices, establishing their core position. The strongest connections existed between anxiety-depression (edge weight = 0. 37) and anxiety-stress (edge weight = 0. 35), while loneliness connected with psychotic experiences (edge weight = 0. 23) and suicidal ideation (edge weight = 0. 144). In item-level analysis, thoughts of death (PHQ₉), lack of companionship (UCLA3₄), and frequency of suicidal thoughts (SBQ2) demonstrated strongest bridge centrality. Network stability analysis showed CS coefficients reached the good standard of 0. 5. These findings demonstrate that depressive symptoms occupy a core network position, loneliness plays a unique bridging role, and suicidal ideation closely associates with depression and loneliness, providing evidence for network-based precision intervention strategies.
Zhang et al. (Fri,) studied this question.