Around 10–30 % of children and adolescents worldwide go through childhood adversities such as sexual abuse, emotional abuse, and neglect (Stoltenborgh et al., 2015). These experiences have been shown to affect mental health (Cook et al., 2005) and can even increase the risk of adult mental illness based on the variety of traumas experienced (Briere et al., 2008; Cloitre et al., 2009). As they transition into adulthood, college students deal with a lot of changes, both academic and personal, which can make them vulnerable to mental health issues (Chi et al., 2020). Recent research among Chinese college students highlights a strong connection between their psychological challenges and childhood trauma (Li et al., 2022).

The 11th revision to the World Health Organization's International Classification of Diseases (ICD-11) (World Health Organization, 2018) includes two typical psychological responses after trauma, post-traumatic stress disorder (PTSD) and complex PTSD (CPTSD), under a category of ‘disorders specifically associated with stress’. PTSD is comprised of three symptom clusters including (1) re-experiencing of the trauma in the here and now, (2) avoidance of traumatic reminders and (3) a persistent sense of current threat that is manifested by exaggerated startle and hypervigilance. CPTSD of ICD-11 includes the three PTSD clusters and three additional clusters that reflect ‘disturbances in self-organization’ (DSO); (1) affect dysregulation, (2) negative self-concept and (3) disturbances in relationships (Maercker et al., 2013). These disturbances are proposed to be typically associated with sustained, repeated or multiple forms of traumatic exposure (Karatzias et al., 2019), reflecting loss of emotional, psychological and social resources under conditions of prolonged adversity (Cloitre et al., 2014). Previous research suggests that while individuals who have experienced various trauma types can develop CPTSD (P. Hyland et al., 2017; Møller et al., 2020), enduring adversities like childhood abuse, domestic violence may represent a common trigger for CPTSD symptoms (Ho et al., 2019; Tian et al., 2020; Karatzias et al., 2022). Therefore, exploring the symptom manifestations of CPTSD within the context of childhood trauma emerges as a significant area of inquiry.

The two-factor model for CPTSD have gained validation through previous support from factor analytic methods and cluster analysis studies (Brewin et al., 2017). In general, PTSD and DSO symptoms exhibit multidimensionality (separate but correlated). The PTSD clusters was primarily associated with symptoms related to fear and anxiety, while the DSO clusters reflect the pervasive psychological disturbances that can occur following traumatic exposure, even in the absence of traumatic reminders (Brewin et al., 2009; Herman, 1992). The theoretical framework of resource loss theory (Hobfoll, 1989) fundamentally links the PTSD with DSO. Within this framework, potential traumatic events negatively impact an individual's physical health and diminish their capacity to cope by adversely affecting essential psychological resources such as positive self-awareness, emotional regulation abilities, and interpersonal relationships. When the harm of an event exceeds the resources for effective coping, the sense of threat associated with fear is amplified. Repeated traumatic events continuously deplete resources, leading to an increased risk of harm and an augmented sense of threat (Hobfoll et al., 2011).

Network analysis is employed to examine the interrelationships among observable variables, including symptoms of psychopathology (Borsboom and Cramer, 2013). The foundational concept of network analysis posits that symptoms of mental health disorders are causally interdependent, thereby exerting mutual influence on one another (McNally, 2016). This paradigm holds significant clinical and scientific relevance, shifting emphasis from identifying the root cause of a psychopathological syndrome to exploring the intricate web of interdependent symptom relationships. In addition, symptoms identified as central through network analysis are hypothesized to be integral in maintaining the coherence and stability of the overall symptom network. These central symptoms could potentially be key targets in therapeutic interventions. The rationale is that if a clinician could effectively reduce the intensity or occurrence of a central symptom, it might consequently diminish the activation or persistence of other symptoms in the network (McNally, 2016; Robinaugh et al., 2016). A deep understanding of the symptom network of CPTSD and the identification of its central symptoms in various samples and research backgrounds may aid in the development of targeted psychological interventions (Knefel et al., 2019). However, the translation of these findings into a psychotherapy program requires careful consideration of the variability and context-specific nature of these symptoms.

The current network analyses of CPTSD primarily focus on contemporaneous networks using cross-sectional data (Knefel et al., 2019; Karatzias et al., 2020; McElroy et al., 2019; Levin et al., 2021). This approach, while informative, inherently limits the exploration of potential causal relationships between symptoms. It becomes challenging to ascertain direct influences among nodes that may span across various disorders. Cross-lagged panel network (CLPN) is an innovative methodology that integrates latent variable models and network analysis to unveil intricate longitudinal processes that transpire within and across various constructs over time (Rhemtulla et al., 2022). It harnesses node-wise regression models, as pioneered by Funkhouser et al. (2021), to compute both autoregressive and cross-lagged effects. By regressing the variables at the second measurement point (T2) against those at the initial measurement point (T1), CLPN provides valuable insights into the cross-lagged effects between these variables (Rhemtulla et al., 2022). Furthermore, CLPN can rigorously assess and confirm a fundamental tenet of network theory: the influential role of central nodes in impacting other nodes (Borsboom and Cramer, 2013). If we assume that these central nodes act as catalysts, sustaining the network's dynamics, CLPN holds the potential to predict the evolution of the other nodes over time. This predictive capacity marks a significant strength of CLPN, augmenting its utility in analyzing complex relationships and dynamic interactions within networks.

In this study, we first examined and compared the differences between contemporaneous networks of CPTSD among College students who have experienced traumatic childhood events at two time points. The aim was to elucidate the longitudinal development of CPTSD network structure and deepen understanding of the relationship between PTSD and DSO. Here, we introduced a CLPN as a valuable approach to assessing the causal relationships in symptom-symptom interaction in the various symptoms of CPTSD. We used it to assess the causal relationships between PTSD and DSO symptoms and thus elucidate the causality between the two latent disorders.