Evolution of the network pattern of posttraumatic stress symptoms among children and adolescents exposed to a disaster

https://doi.org/10.1016/j.janxdis.2020.102330Get rights and content

Highlights

  • Sleep problems and nightmares exhibited high centrality at 3 months, and their centrality decreased over time.

  • The centrality of physiological cue reactivity and flashbacks increased over time and reached a high level at 27 months.

  • Temporal differences in PTSD symptoms merit more consideration in research and clinical practice.

Abstract

Growing network approach analyses of posttraumatic stress disorder (PTSD) have enhanced the knowledge of PTSD symptomatology. This study aims to explore changes in network patterns of PTSD symptoms among youth survivors following the Zhouqu debris flow through a network approach. A two-year longitudinal study was conducted to follow 1460 children and adolescents at 3, 15, and 27 months after the disaster. Symptoms with high centrality varied at different time points. Sleep problems and nightmares exhibited high centrality at 3 months, and their centrality decreased over time, while the centrality of physiological cue reactivity and flashbacks increased over time and reached a high level at 27 months. The global connectivity of the network was stronger at 27 months than at 3 and 15 months. These findings provide novel insights into youths’ PTSD symptom evolution. Temporal differences in PTSD symptoms merit more attention from researchers. Different core symptoms in acute and chronic PTSD structures should be treated as targets at different stages following trauma in clinical practice.

Introduction

Posttraumatic stress disorder (PTSD) is a maladaptive and disabling reaction caused by a traumatic event (Lowell et al., 2018). Researchers and clinical workers are very concerned about the dynamics of PTSD development because previous studies have found that the level of PTSD symptoms is closely related to time (Galatzer-Levy, Huang, & Bonanno, 2018; Liang, Cheng, Ruzek, & Liu, 2019), and the most prominent symptoms of PTSD may change over time (Ge, Yuan, Li, Zhang, & Zhang, 2019). To improve the effectiveness of clinical work, it is important to identify the core symptoms of PTSD at different stages.

The emerging symptom network analysis in recent years is a good approach to gain insight in this issue because the important advantage of network analysis is that it can fully understand the role of each symptom in mental disorders (McNally, 2016). A network approach study has important implications for identifying the most influential symptoms in a potentially interactive network, which is defined as high centrality in network analyses (Borsboom & Cramer, 2013; McNally, 2016). The network approach to psychopathology conceptualized that psychopathological symptoms are correlated in a syndrome but are not reflective of a higher-level latent construct, whose emergence advanced our understanding and complemented the traditional perspective (Borsboom & Cramer, 2013; Borsboom, 2017; Bryant et al., 2017; McNally, 2016). From a network perspective, the emergence and maintenance of a mental disorder are caused by strong causal interactions and feedback loops of symptoms (Borsboom, 2017). Symptoms with high centrality may be the main constituent symptoms of a mental disorder and should be treated as treatment targets (Borsboom & Cramer, 2013; McNally, 2016). Therefore, network research is particularly interested in identifying psychopathological symptoms with high centrality.

Recently, rapidly growing network approach studies of PTSD have provided novel insights into PTSD symptoms (Bartels et al., 2019; Birkeland, Greene, & Spiller, 2020). Although there are considerable network approach studies related to PTSD, few of them have focused on children and adolescents. Children and adolescents show some different posttraumatic responses and PTSD development patterns than adults due to the different psychological characteristics of children and adults (Braun-Lewensohn, 2015; Galatzer-Levy et al., 2018; Liang, Cheng, Zhou, & Liu, 2019). Moreover, developmental differences also exist in PTSD symptomatology. Following traumatic events, compared to adults, children and adolescents have been demonstrated to have distinct PTSD symptoms and broader associated symptoms (Bartels et al., 2019; Helpman et al., 2015; Russell, Neill, Carrion, & Weems, 2017). To date, only five studies have explored the network structure of PTSD symptoms among children and adolescents (Bartels et al., 2019; Cao et al., 2019; De Haan et al., 2020; Ge et al., 2019; Russell et al., 2017). Among these studies, flashbacks and physiological cue reactivity were demonstrated to be symptoms with high centrality in the DSM-IV criteria (Cao et al., 2019; Russell et al., 2017), and negative beliefs and persistent negative emotional state were demonstrated to be symptoms with high centrality in the DSM-5 criteria (Bartels et al., 2019). However, among most network studies focused on PTSD in adult samples, recurrent thoughts of trauma and difficulty concentrating were demonstrated to be symptoms with high centrality in the DSM-IV criteria (Bryant et al., 2017; McNally et al., 2015; Phillips et al., 2018), and persistent negative emotional state and feelings of detachment from others were demonstrated to be symptoms with high centrality in the DSM-5 criteria (Armour, Fried, Deserno, Tsai, & Pietrzak, 2017; Moshier et al., 2018; Ross, Murphy, & Armour, 2018; von Stockert, Fried, Armour, & Pietrzak, 2018). These differences indicate limited generalizability of findings from the adult population to children and adolescents.

Another limitation to note in existing network studies for PTSD symptoms is the lack of a longitudinal design. Only a minority of studies have explored the changes in the network structure of PTSD symptoms over time among adults (Bryant et al., 2017; von Stockert et al., 2018), and only one study investigated the changes among youth (Ge et al., 2019). After exposure to traumatic events, PTSD symptoms change over time, particularly in the early stage following trauma (Cheng, Liang, Zhou, Eli, & Liu, 2019; Galatzer-Levy et al., 2018; Lai, Lewis, Livings, La Greca, & Esnard, 2017). Ge et al. (2019) explored the changes in the network of PTSD symptoms at 2 weeks, 3 months and 6 months following the Lushan earthquake among youth survivors. They found that flashback and upset by reminders were symptoms with high centrality at all time points, and avoiding thoughts increased in centrality over time. However, the evolution of the long-term network of PTSD symptoms in youth is still unknown. Previous studies have found that flashbacks and physiological cue reactivity had high centrality in 2.5 years (Cao et al., 2019) and 3–4 years after natural disasters (Russell et al., 2017) among children and adolescents, which were inconsistent with Ge et al.'s (2019) findings. Thus, high central symptoms may be varied at early and later stages after a disaster. Many PTSD symptoms often resolve spontaneously in the acute phase of traumatic events, and symptoms often decline substantially in the first two years among youth survivors following a natural disaster; therefore, most youth survivors recover without treatment (Cheng et al., 2019; Fan, Long, Zhou, Zheng, & Liu, 2015; Lai et al., 2017). However, approximately 10–20 % of youth survivors who are initially symptomatic continue to experience PTSD symptoms over a course of two years following a natural disaster (Fan et al., 2015; Lai et al., 2017; Osofsky, Osofsky, Weems, King, & Hansel, 2015). Therefore, identifying the long-term pattern of PTSD symptoms reflecting chronic PTSD symptoms has important implications for PTSD diagnosis and long-term intervention.

To identify changes in the network of PTSD symptoms among children and adolescents, we conducted a longitudinal study after the Zhouqu debris flow, which occurred in Gansu Province, China. At 11:00 p.m. on August 7, 2010, a sudden downpour in the mountainous area in the northeast of Zhouqu County triggered landslides. The mudslides passed through the densely populated areas of the county and blocked the Bailong River. More than half of the Zhouqu county urban area was flooded, causing significant losses of personnel and property. A total of 1557 people were killed, 284 people were missing and 2315 outpatients were treated in the disaster. Additionally, 4496 households and 20,227 people were affected by the disaster, 1417 acres of farmland were destroyed by water, and 5508 houses were destroyed by the disaster. Four schools (2 primary and 2 secondary schools) located in Zhouqu County that were seriously destroyed by the disaster were chosen to be investigated. Among these participants, 15.21 % were trapped and 4.52 % were injured in the debris flow; 28.42 % of them witnessed a building collapse, 30.34 % witnessed a dead body and 39.38 % lost one (or more) relative(s) in the debris flow.

To increase our understanding of the dynamics of PTSD, the current study was aimed to explore youths’ network structures of PTSD at different stages during two years following the Zhouqu debris flow. We were particularly interested in identifying changes in the network of PTSD and the most central symptoms at different times, which may provide inspiration for formulating time-specific and effective intervention programs (McNally, 2016). We hypothesize that (i) flashback and emotional cue reactivity were symptoms with high centrality at 3 months after the disaster; and (ii) flashbacks and physiological cue reactivity were symptoms with high centrality at 27 months after the disaster.

Section snippets

Participants and procedure

A two-year longitudinal study was conducted beginning 3 months following the Zhouqu debris flow. In the first investigation (T1: early November 2010), 3957 participants were surveyed in grades four to nine from 2 primary and 2 secondary schools located in Zhouqu County. These 4 schools were the schools most affected by the disaster and covered the vast majority of children in this age range in Zhouqu County. In the second investigation (T2: early November 2011), 5344 participants were surveyed

Descriptive results

The average PTSD scores of the participants at the three time points (3, 15, 27 months) were 22.69 (SD = 11.16), 20.01 (SD = 11.03), and 18.76 (SD = 11.10). According to the criterion (cut-off scores of 38), the prevalence rates of PTSD at T1, T2, T3 were 9.32 %, 7.26 % and 5.82 %, respectively.

Networks and centrality estimation

The networks of PTSD symptoms at 3, 15, and 27 months are shown in Fig. 1A–C, respectively. Many similar connection patterns appeared in the networks across different time points, such as a continuous

Discussion

The present study represents a network analysis investigation of long-term changes in PTSD symptoms among children and adolescents. Generally, different symptoms had high centrality in different periods. Inconsistent with hypothesis (i) and a previous study (Ge et al., 2019), sleep problems and nightmares had high centrality in 3 months. The differences between our study and Ge et al.'s (2019) might be caused by the different measurements of PTSD in the two studies. Consistent with hypothesis

Conclusion

The current study is a network analysis investigation of long-term changes in PTSD symptoms among children and adolescents. Sleep problems and nightmares exhibited high centrality at 3 months, and their centrality decreased over time, while the centrality of physiological cue reactivity and flashbacks increased over time and reached a high level at 27 months. The global connectivity of the network was stronger at 27 months than at 3 and 15 months, which might reflect the hysteresis of PTSD.

Funding

This work was supported by the National Key R&D Program of China (2020YFC2003000).

Declaration of Competing Interest

None.

Acknowledgments

We thank the schools and the children for their participation.

References (48)

  • American Psychiatric Association

    Diagnostic and statistical manual of mental disorders

    (2013)
  • L. Bartels et al.

    The importance of the DSM‐5 posttraumatic stress disorder symptoms of cognitions and mood in traumatized children and adolescents: Two network approaches

    Journal of Child Psychology and Psychiatry

    (2019)
  • M.S. Birkeland et al.

    The network approach to posttraumatic stress disorder: A systematic review

    European Journal of Psychotraumatology

    (2020)
  • D. Borsboom

    A network theory of mental disorders

    World Psychiatry : Official Journal of the World Psychiatric Association (WPA)

    (2017)
  • D. Borsboom et al.

    Network analysis: An integrative approach to the structure of psychopathology

    Annual Review of Clinical Psychology

    (2013)
  • O. Braun-Lewensohn

    Coping and social support in children exposed to mass trauma

    Current Psychiatry Reports

    (2015)
  • R.A. Bryant et al.

    Acute and chronic posttraumatic stress symptoms in the emergence of posttraumatic stress disorder: A network analysis

    JAMA Psychiatry

    (2017)
  • X. Cao et al.

    Sex differences in global and local connectivity of adolescent posttraumatic stress disorder symptoms

    Journal of Child Psychology and Psychiatry

    (2019)
  • Z. Chen et al.

    Structure of the Children’s Revised Impact of Event Scale (CRIES) with children and adolescents exposed to debris flood

    PloS one

    (2012)
  • J. Cohen

    Statistical power analysis for the behavioral sciences

    (1988)
  • A.O. Cramer et al.

    Complex realities require complex theories: Refining and extending the network approach to mental disorders

    The Behavioral and Brain Sciences

    (2010)
  • A. De Haan et al.

    Dysfunctional posttraumatic cognitions, posttraumatic stress and depression in children and adolescents exposed to trauma: A network analysis

    Journal of Child Psychology and Psychiatry

    (2020)
  • S. Epskamp et al.

    A tutorial on regularized partial correlation networks

    Psychological Methods

    (2018)
  • S. Epskamp et al.

    Estimating psychological networks and their accuracy: A tutorial paper

    Behavior Research Methods

    (2018)
  • Cited by (0)

    View full text