Psychotic-like experiences (PLEs) are common in adolescence and often associated with later mental health difficulties. Although many psychosocial factors are related to PLEs, little is known about how these factors interact over time. Longitudinal network analysis allows examination of the stability of symptom associations and identification of potential intervention targets. This study investigated the structure and temporal stability of PLE networks in a large community-based adolescent cohort.

Adolescents aged 13–19 years (N = 605 with complete data across all time points) completed assessments at baseline, 12 months, and 24 months. Measures included positive and negative PLEs, cognitive biases, depression, anxiety, trauma, and interpersonal sensitivity. Networks were estimated at each time point, and permutation-based tests were used to compare network structure and overall connectivity across time. Centrality stability was assessed using bootstrapping procedures.

Network structures were stable across the 2-year period, with no significant differences in overall organization or connectivity between time points. Depression consistently showed the highest centrality, followed by anxiety and attributional bias. Positive PLEs were most strongly associated with anxiety, while negative PLEs showed their strongest associations with depression. Attributional bias remained centrally positioned and was strongly linked to trauma. All networks showed robust accuracy and high stability.

Despite considerable developmental change during adolescence, the psychosocial architecture of PLEs remained notably stable. Depression, anxiety, and attributional biases emerged as consistent key nodes, highlighting them as promising targets for prevention and early intervention in adolescents at risk for persistent PLEs.

Occupational stress triggers psychological/physical health issues, elevating the risk of burnout and depression. This study explored the interrelationships among these constructs via network analysis (undirected/directed graphs).

A total of 1363 participants from Beijing hospitals and a university completed House and Rizzo’s Work Stress Scale, Zung’s Self-Report Depression Scale, and Maslach Burnout Inventory-General Survey. Graphical Gaussian Model and directed acyclic graphs (DAG) identified core/bridge/upstream nodes and causal pathways.

Emotional exhaustion (EE) was the core node (expected influence = 2.11). The strongest edge was D11–D12 (weight = 0.46). EE, occupational stress 11, cynicism (CY), and personal accomplishment (PA) served as key bridging nodes. The network showed high stability (0.75). DAG identified upstream occupational stress 1/7/8, confirming direct occupational stress to depression pathways (emotional dysregulation model) and CY/PA mediated pathways (burnout structural theory).

Targeted interventions on core/bridge/upstream nodes may prevent depression onset and progression in occupational settings.

The network theory of mental disorders posits that associations between symptoms activate other symptoms to maintain a disorder over time. Network analytic approaches therefore may inform treatment targets. In the present study, we compared baseline OCD symptom networks among treatment responders to non-responders and examined how network structure and connectivity changed from before to after exposure and response prevention (ERP) treatment.

Community adults with OCD (n = 712) who underwent intensive outpatient treatment were assessed using the Yale-Brown Obsessive Compulsive Scale (YBOCS) at admission and discharge. Network comparison tests were used to (a) examine differences in baseline symptom network structures between treatment responders versus non-responders and (b) examine changes in network structures from pre- to post-treatment.

Pre-treatment network structures and global connectivity did not differ significantly between treatment responders and non-responders. However, post-treatment networks exhibited greater global strength (i.e., stronger associations between OCD symptoms) and significantly different network structure (i.e., different patterns of associations between OCD symptoms) relative to the pre-treatment network.

Findings showed that network structure and connectivity in OCD may be more informative as a marker of therapeutic change than in discriminating treatment responders from nonresponders using baseline symptoms. After ERP treatment, associations between obsessions and compulsions demonstrated significantly greater global network strength and altered network structure, thus underscoring the potential for network approaches to identify mechanisms of change throughout OCD treatment. Future studies incorporating session-by-session data may clarify when and how these network shifts occur over the course of therapy to help identify treatment targets.

Network analysis was employed to test whether the overall pattern of depressive–anxious symptom connections remains stable or whether specific symptom-to-symptom links shift from pregnancy to postpartum.

In a perinatal sample (n = 4,461 pregnant women, n = 5,711 postpartum women), depressive and anxiety symptoms were assessed with the Edinburgh Postnatal Depression Scale (EPDS) and Generalized Anxiety Disorder-7 (GAD-7). Phase-specific polychoric Gaussian graphical models were estimated with EBICglass. We examined strength and bridge centrality, community structure, and nodewise predictability, and compared networks using the network comparison test.

Depression and anxiety formed four reproducible communities (one GAD-7 worry/arousal and three EPDS affective/anhedonic, anxious–cognitive distress, and depressed affect/sleep–suicidality modules) with identical partitions across phases. Global strength was similar, but postpartum networks showed higher edge density and more negative partial correlations, suggesting localized changes in which symptom pairs were directly linked—and how strongly—across phases. Across phases, Sadness, Crying, Uncontrollable worrying, and Trouble relaxing were most central and predictable. Worry-, arousal-, and sleep-related symptoms (e.g., hard to sleep) showed the strongest bridge centrality postpartum, and Self-harm was a prominent bridge during pregnancy; several edges shifted between phases, including stronger Enjoyment–Self-harm and weaker Hard to sleep–Self-harm postpartum.

Perinatal depression and anxiety organize into cohesive yet partially distinct symptom networks that remain globally stable but show localized shifts in direct symptom-to-symptom connections from pregnancy to postpartum. Central affective and arousal nodes, particularly sadness, pathological worry, and sleep disturbance, may be high-yield targets for phase-tailored screening and intervention.

Network modeling of post-concussion symptoms following mild traumatic brain injury (mTBI) has emerged as a promising tool for understanding how cognitive, emotional, and somatic symptoms co-occur and interact. However, the generalizability of networks developed in individual studies remains unclear. This study aimed to develop the first-ever meta-analytic pooled between-persons network structure of post-concussion symptoms and systematically examine the between-study heterogeneity of these symptom networks.

Using the Meta-Analytic Gaussian Network Aggregation (MAGNA) framework, a single pooled network model was developed by aggregating data from 6 distinct samples, comprising a total of 5,776 participants. Additionally, this study quantitatively assessed the degree of heterogeneity across these studies.

Strong symptom clusters between cognitive, emotional, and somatic symptoms were identified. Concentration difficulty and slowed thinking were the most central symptoms in the pooled MAGNA network. Large between-study heterogeneity was observed.

Findings from this meta-analysis highlight cognitive symptoms as most important for defining the network structure after mTBI at a group level, potentially perpetuating and/or being perpetuated by symptoms in other domains. The large heterogeneity observed between studies underscores the need for an idiographic (person-specific) approach to studying post-concussion symptom networks to inform precision rehabilitation.

Coroners’ Prevention of Future Death reports (PFDRs, also known as Regulation 28 reports) provide an opportunity to understand factors contributing to mental health-related deaths.

To examine available mental health-related PFDRs, addressing three core questions: (a) What is the overall profile of these reports? (b) What relational patterns emerge from these reports? and (c) What concerns and preventive actions do coroners highlight in these reports, and how they evolved over time?

We collected all mental-health related public PFDRs available up to June 2025 (N = 586). Data extraction combined automated web scraping, optical character reading and large language model (LLM)-assisted (GPT-4o) parsing to capture demographics, settings, coroner areas, co-occurring categories, concerns and recommended actions. Descriptive statistics, category and recipient co-occurrence network analysis and thematic analysis were used to provide a comprehensive landscape of these reports.

Report numbers increased steadily from 2013, peaking in 2021 and then declined. Some jurisdictions, including Manchester South, East Sussex and East London, consistently had more PFDRs issued. The deceased were typically young, male and had died mainly outwith hospital, most often at home; 78.0% of reports included at least one formal response from recipients, whereas 22.0% had no corresponding response available. The network analyses suggested that PFDRs seldom identified isolated issues. Coroners’ concerns changed over time, from service access and resources to inter-agency coordination and then, more recently, to risk assessment and management.

Mental health-related deaths examined by coroners arise within complex, evolving multi-sector contexts and do not frequently identify single errors. Minimising such deaths may require coordinated strategies across healthcare, social care and justice systems. Analysis of PFDRs allows identification of patterns that may inform such actions. PFDRs should be analysed routinely and patterns followed over time.

Language impairments are common in affective and psychotic disorders, yet their patterns and underlying pathomechanisms remain insufficiently understood. A transdiagnostic perspective provides a framework for identifying shared and disorder-specific language alterations across diagnostic boundaries. Combining natural language processing (NLP) with network analysis enables the investigation of complex associations between linguistic, cognitive, and psychopathological features.

Spontaneous speech from N = 372 participants (119 MDD, 27 BD, 48 SSD and 178 HC) was elicited using four Thematic Apperception Test pictures (~12 min per participant). NLP models were applied to extract latent linguistic variables across various levels, including lexical diversity, syntactic complexity, semantic coherence, and disfluencies. Network analysis was used to relate linguistic variables, psychopathology (SAPS, SANS, HAM-A, HAM-D, YMRS, TLI, GAF), and cognitive performance (attention, verbal memory, recognition, and verbal fluency).

Linguistic variables formed the densest network cluster, with type–token ratio, mean length of utterance, and syntactic complexity emerging as central nodes. Psychopathology variables were less cohesive, while TLI “Impoverishment”, coherence mean, and executive functioning bridged linguistic, cognitive, and psychopathological domains. Network comparison tests revealed no significant differences in linguistic–cognitive network structure across HC, MDD, BD, and SSD.

Linguistic networks show high structural consistency across healthy individuals and patients, whereas psychopathological symptom networks reflect transdiagnostic profiles. These findings support a dimensional and transdiagnostic framework underscore shared language–cognition mechanisms, and highlight executive functioning as key cross-domain connection, which opens up new avenues for dimensional research into the pathophysiological and etiological mechanisms underlying language dysfunctions.

Compared with well-studied internal adaptive systems, there remains a lack of comprehensive exploration of external correlated factors of resilience, as well as the way in which each ingredient of resilience is influenced.

This study aims to explore the dimensional associations among resilience and several factors, including parenting rearing style, childhood trauma and negative life events.

A series of social demographic variables, parental rearing patterns, childhood trauma, negative life events and resilience were assessed. Multiple linear regression analysis was used to explore correlated factors of resilience, with all the above factors included in the model. Network analysis was conducted to identify the central factor and key associations, and to visualise complex interactions among resilience, parenting rearing style, childhood trauma and negative life events.

This cross-sectional study was conducted among 4302 freshmen (2388 females, 55.5%; mean 18.59; s.d. = 0.95) from three colleges between October and December 2020. Three key associations were discovered: ‘learning pressure and emotional control’ (r = −0.195, P < 0.05), ‘emotional neglect and family support’ (r = −0.129, P < 0.05) and ‘maternal care family support’ (r = 0.193, P < 0.05). ‘Emotional abuse’ (bridge expected influence, −0.588) was the core node of the estimated network.

This study found that learning pressure, emotional neglect and maternal care emerged as the most critical external correlates of resilience. Emotional abuse occupies the most central position in the external correlated network of resilience. Future longitudinal research should clarify the temporal impacts of these associations, and the key factors, in the dynamic resilience system.