Abstract
Background
The death of a child is a highly traumatic event for parents and often leads to posttraumatic stress disorder (PTSD) and depression. PTSD and depression are frequently comorbid. However, the patterns of comorbidity at the symptom level among bereaved parents remain unclear. This study aims to identify the symptom network connectivity of PTSD comorbid with depression in bereaved Chinese parents who have lost their only child, known as Shidu parents in Chinese society.
Methods
Data were obtained from 477 bereaved individuals who had lost an only child. A regularized partial correlation network was used to construct comorbidity networks of PTSD and depression symptoms. A relative importance network was computed to determine directionality among symptoms in the network.
Results
PTSD symptoms of diminished interest, irritability/anger, reckless/self-destructive behavior, a negative emotional state, the avoidance of reminders, and flashbacks acted as important bridging nodes. The relative importance network analysis show two bidirectional pathways (hypervigilance ⇆ exaggerated startle response ⇆ difficulty concentrating ⇆ sleeping difficulties ⇆ poor sleep ; irritability/anger ⇆ reckless/self-destructive behavior ⇆ could not get going ⇆ bothered by things), and exaggerated startle response, difficulty keeping mind, hypervigilance, and diminished interest have higher outstrength values with greater influence on a multitude of other symptoms within the comorbidity network.
Conclusions
The findings highlight the important role of specific symptoms and identified bidirectional pathways underscore the interconnected nature of PTSD and depression symptoms, suggesting that targeting these core or highly influential symptoms may disrupt the broader network and improve therapeutic outcomes.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12888-025-07283-4.
Keywords: Posttraumatic stress disorder, Depression, Bereaved parents (Shidu parents), Comorbidity, Network analysis
Introduction
The death of a child poses a significant risk to the mental health of bereaved parents [1]. These parents often experience mental health issues, such as posttraumatic stress disorder (PTSD) and depression [2]. Notably, PTSD and depression are highly comorbid, with an estimated co-occurrence rate of 52.0% among trauma-exposed populations [3]. In China, this issue takes on unique dimensions due to the emergence of Shidu parents who have lost their only child and, having passed their reproductive years (with mothers typically over 49 years of age), are unable or unwilling to have or adopt another child [4, 5]. Among Shidu parents, the comorbidity rate of PTSD and depression reaches 47.6% [6], a figure comparable to broader trauma-affected groups but compounded by unique socio-cultural stressors. Critically, research suggests that comorbid PTSD and depression may exacerbate mental health burdens and impair quality of life more severely than PTSD alone [7]. Therefore, understanding this pattern of comorbidity is essential for developing targeted screening and intervention strategies [8], particularly for high-risk groups like Shidu parents.
In recent years, studies utilizing network analysis to look at the comorbidity patterns of PTSD and depression [9, 10]. In the network, symptoms are represented as nodes, and the associations between these symptoms are represented as edges. Mental disorders arise from dynamic causal interactions and feedback loops (edges), among their defining symptoms (nodes) [11]. Additionally, network analysis can elucidate the relationships between disorders at the symptom level by identifying the bridge symptoms that occur in two or more disorders and linking communities of symptoms [12].
Although numerous previous network studies have explored the comorbidity of PTSD and depression in adults [13–16], providing a foundation for understanding their comorbidity patterns, several research limitations persist. Firstly, there is still no consensus regarding the bridging roles of symptoms within the comorbidity network of PTSD and depression. For example, some studies have indicated that sleep problems and concentration difficulties serve as significant bridging roles [8, 13, 15], while other research has highlighted that flashbacks, avoidance of thoughts, emotional upset by trauma reminders, and anhedonia play crucial bridging roles [14, 16].
Secondly, previous studies have primarily focused on the comorbidity network among adult populations with various traumatic experiences, including veterans, firefighters, and victims of interpersonal violence [13, 14, 16]. However, there remains a significant gap in research specifically addressing Shidu parents. While these network outcomes have enhanced our understanding of the field, the conclusions drawn from prior studies have limited generalizability to Shidu parents for two key reasons. First, the loss of a child is distinct from other types of traumatic events, as it can fundamentally dismantle parents’ social identity, which is rooted in familial roles (e.g., caregiver, lineage bearer) [17, 18]. Unlike other trauma survivors who may retain alternative social roles, Shidu parents often face a permanent vacancy in these roles [18]. Additionally, Chinese families are typically child-centered, with the child serving as the primary source of hope and meaning in life for parents [19]. The loss of their only child represents not only a loss of spiritual sustenance but also a perceived loss of the right to happiness for bereaved parents. Consequently, these parents often feel that life has become meaningless and hopeless [20]. Second, Shidu parents experience both psychological trauma and cultural pressure, as Chinese culture views the death of a child as a sign of bad luck. This belief is often accompanied by stigmatization and social withdrawal [18], which may intensify PTSD symptoms such as diminished interest and emotional detachment [21]. Given the evidence that the type of trauma influences the structure of PTSD networks [22], further empirical research is essential to clarify the unique characteristics of comorbid PTSD and depression in Shidu parents.
Finally, most existing network studies on the comorbidity between PTSD and depression lack directional determination. Most cross-sectional studies have used regularized partial correlation network analysis to identify the comorbid network structure [14, 15, 23, 24]. However, this method calculates an undirected sparse network in which nodes represent symptoms and edges represent the partial correlation between two nodes after controlling for all other nodes [25]. A relative importance network approach can help overcome the limitations of a partial correlation network because it enables the examination of directional relationships among items within cross-sectional data [26]. In a relative importance network, each edge represents the relative importance weights of node X in predicting node Y after controlling for all the other nodes [26]. The resulting network is weighted and directed. Combining these two approaches is more helpful for clarifying bridge symptoms and the directional relationship between symptoms.
To address these gaps in the literature, the current cross-sectional study utilizes two types of network approaches to identify the comorbidity network structure between PTSD and depression among Shidu parents. First, we use the regularized partial correlation network to investigate the comorbid network patterns of PTSD and depression and to identify bridge symptoms, the strongest edges, and potential causal loops. Second, we examine a relative importance network to predict the predominant pathways of activation between PTSD and depression.
Methods
Participants and procedure
The present study utilized self-reported data from 447 bereaved individuals who had lost their only child in China (Shidu parents). A total of 385 participants from 10 cities in 5 provinces were surveyed from November 2016 to July 2017. Details of the sampling and procedure of the original data have been reported in a previous publication [21]. Sixty-two participants from Luohe city, Henan Province, China, were surveyed in July 2022. The investigation procedures and inclusion criteria for the two batches of data were the same. The current study specifically examined data on the PTSD and depression of Shidu parents.
Among the 447 participants, 188 (42.1%) were male, with a mean age of 62.3 ± 7.6 years (ranging from 44 to 84 years); 224 (50.1%) were female, with a mean age of 61.6 ± 7.2 years (ranging from 55 to 88 years); and 35 (7.8%) had missing sex information. The losses had occurred an average of 9.6 years (SD = 7.3) prior to data collection. The sociodemographic characteristics of the participants are presented in Table 1.
Table 1.
Sociodemographic characteristics of the participants (N = 447)
| Sociodemographic characteristic | Category | Frequency | % |
|---|---|---|---|
| a Sex | Male | 188 | 42.1 |
| Female | 224 | 50.1 | |
| a Age | ≤ 60 years | 191 | 42.7 |
| > 60 years | 255 | 57.0 | |
| a Educational level | Junior high school or below | 286 | 64.0 |
| High school or above | 158 | 35.3 | |
| a Marital status | Married (first/remarriage) | 339 | 75.8 |
| Single/divorced/separated/widowed | 83 | 18.6 | |
| a Religious beliefs | None | 354 | 79.2 |
| Buddhism | 15 | 3.4 | |
| Other | 19 | 4.3 | |
| a Family income | ≤ 3000 | 307 | 68.7 |
| > 3000 | 40 | 8.9 | |
| a Subjective assessment of family economic status | Poor and extremely poor | 129 | 28.9 |
| Moderate | 196 | 43.8 | |
| Wealthy and extremely wealthy | 27 | 6.0 | |
| a Cause of death | Unnatural cause | 223 | 49.9 |
| Natural cause | 158 | 35.3 | |
| a Time since child’s death | 1–3 years | 93 | 20.8 |
| 4–8 years | 105 | 27.5 | |
| 9–15 years | 113 | 30.4 | |
| ≥ 16 years | 60 | 16.1 |
a: There are missing values for sociodemographic characteristics; thus, the sum of the effective percentage is equal to 100% in these cases; unnatural causes: accident, homicide, suicide, natural disaster, or other; natural causes: illness
All the participants provided written informed consent to participate in the study. This study was approved by the ethics review committee of the Institute of Psychology, Chinese Academy of Sciences (Ethics approval number: H21044).
Measures
PTSD symptoms were assessed via the Posttraumatic Stress Disorder Checklist for the DSM-5 (PCL-5) [27]. The PCL-5 is a self-reported measure assessing the severity of 20 PTSD symptoms with four subscales: intrusion (Cluster B), avoidance (Cluster C), negative alterations in cognition and mood (Cluster D), and alterations in arousal and reactivity (Cluster E). The participants indicate the extent to which each symptom has bothered them over the past month via a five-point Likert scale ranging from 0 (not at all) to 4 (Very much). Higher scores indicate greater PTSD symptom severity.
A cutoff of 33 or above on the PCL-5 total score indicates the presence of significant PTSD symptoms [27]. In the present study, the Cronbach’s alpha value for this measure was 0.96, and the KMO value of the confirmatory factor analysis was also 0.96.
Depression symptoms were assessed via the shortened Center for Epidemiological Studies-Depression Scale (CES-D-10). The CES-D-10 consists of 10 items from the original 20-item questionnaire [28]. The participants indicate the extent to which each symptom has bothered them during the past week via a 4-point Likert scale. Two of the ten positively rated items (“I felt hopeful about the future” and “I was happy”) were reverse scored for the analysis. The total score ranges from 0 to 30, and higher scores indicate greater severity of depression. A cutoff of 10 or above on the CES-D-10 total score indicates the presence of significant depression [28]. In the present study, the Cronbach’s alpha value for this measure was 0.84, and the KMO value of the confirmatory factor analysis was 0.89.
Statistical analyses
Given the difference in the scale range between the PCL-5 and the CES-D-10, the scores for all the items were converted to z scores. Network analysis was conducted via R Core Software, version 4.1.3 (R Team, Vienna, Austria).
Partial correlation network analysis
The network structure of the comorbidities was estimated via the R package qgraph [29] and visualized via the Fruchterman–Reingold algorithm [30]. The network was estimated via the least absolute shrinkage and selector operator (LASSO) regularization procedure, which sets all weak partial correlations (determined by a set parameter) to exact zero. The parameter is chosen via the extended Bayesian information criterion (EBIC) and is set to 0.5 [25]. The algorithm positions nodes with more or stronger connections at the center of the network. The green and red edges represent positive and negative associations, respectively. Line thickness reflects the strength of the association.
To identify bridge nodes linking PTSD and depression, we estimated bridge centrality measures using the R package networktools [12]. Following methodological recommendations in the literature, we adopted bridge expected influence (bEI; 1-step) as the primary outcome metric. The bEI quantifies a node’s bridging role by summing the absolute weights of all edges connecting it to nodes in the other disorder, thereby capturing both positive and negative associations. Consistent with established practices [12], nodes with bEI values in the top 20% were classified as bridge nodes, as these represent the most influential connections between the two symptom networks.
Network accuracy and stability estimation
The network accuracy and stability were tested via the R package bootnet [25]. The network accuracy was tested by calculating the bootstrapped 95% confidence intervals (CIs) of the edge weights 1000 times. Smaller CIs represent greater accuracy in estimation. Network stability was assessed via the case-dropping subset bootstrap method with a correlation stability coefficient (CS coefficient). The CS coefficient should not be below 0.25; a value above 0.50 indicates strong stability and interpretability.
Relative importance network analysis
To describe the directionality of the comorbidity network, a relative importance network was constructed using the “relimp” option of the estimateNetwork function in the bootnet package in R, by using the lmg metric [31]. The relative importance networks are based on the proportion of variance (e.g., R2) that one item explains in another item after controlling for all other items [31]. In other words, the relative importance weight quantifies the amount of explained variance attributable to each predictor after controlling for multicollinearity, and it ranges between 0 and 1. This procedure was repeated for every node of the network. The graph illustrates the magnitude of the relationship And the direction of prediction, with arrows originating in the predictor symptom And terminating in the predicted symptom. In this study, the threshold was set to 0.05.
Furthermore, we calculated three indexes of centrality—outstrength, betweenness, and closeness—to qualify the importance of each node in the relative importance network. All the centrality indexes were computed via the R package qgraph [29]. Outstrength summarizes the total influence that a certain node exerts on all the other nodes. Betweenness refers to the number of times that a specific node lies on the shortest path between two other nodes, whereas closeness is computed as the inverse of the sum of the total length of all the shortest path lengths between a specific node and the remaining network [9]. However, we focused only on outstrength in this study because recent studies have indicated that betweenness and closeness centrality estimates tend to be unstable [32, 33].
Results
Descriptive analyses
The mean PCL-5 and CES-D-10 total scores for all the participants were 38.2 (SD = 20.9) and 14.1 (SD = 6.6), respectively. Among the 447 participants, 55.0% (246) reported PTSD, 74.3% (332) reported depression, and 52.3% (234) reported the comorbidity of PTSD and depression.
Partial correlation network
The partial correlation network structures of PTSD and depression symptoms are presented in Fig. 1. The results of the edge weight bootstrap (Figure S1) revealed considerable overlap between the 95% CIs of the edge weights in the network. However, some of the strongest edges exhibited nonoverlapping CIs in the network, reflecting the moderate accuracy of the estimated network. The results of the associated case-dropping subset bootstrapping of networks (Figure S2) revealed a CS coefficient of 0.75, reflecting a highly stable network.
Fig. 1.
Partial correlation network of PTSD and depression symptoms
In the comorbidity network, 198 of the 435 edges were nonzero (density of 0.46), and the majority of symptoms were positively connected. PTSD and depression exhibited strong internal connections. The strongest edge connections were symptoms E3 and E4 (hypervigilance and an exaggerated startle response), B1 and B2 (recurrent thoughts and nightmares), Dep5 and Dep8 (hopelessness about the future and unhappiness), and Dep1 and Dep10 (bothered by things and could not get going).
The network centrality bEI for PTSD and depression is presented in Fig. 2. Symptoms D5 (diminished interest), E1 (irritability/anger), E2 (reckless/self-destructive behavior), D4 (negative emotional state), C2 (avoidance of reminders), and B3 (flashbacks) played crucial bridging roles in connecting to other symptoms of PTSD and depression.
Fig. 2.
Network centrality bridge expected influence (bEI) for PTSD and depression. Note. Higher values of bridge expected influence reflect greater bridging node centrality. B1 = recurrent thoughts; B2 = nightmares; B3 = flashbacks; B4 = psychological cue reactivity; B5 = physiological cue reactivity; C1 = avoidance of thoughts; C2 = avoidance of reminders; D1 = amnesia; D2 = negative beliefs; D3 = distorted blame; D4 = negative emotional state; D5 = diminished interest; D6 = detachment; D7 = restricted affect; E1 = irritability or anger; E2 = reckless or self-destructive behavior; E3 = hypervigilance; E4 = exaggerated startle response; E5 = difficulty concentrating; E6 = sleeping difficulties. Dep1 = bothered by things; Dep2 = difficulty keeping mind; Dep3 = depressed mood; Dep4 = everything is an effort; Dep5 = hopelessness about the future; Dep6 = fear; Dep7 = poor sleep; Dep8 = unhappiness; Dep9 = loneliness; Dep10 = could not get going
Relative importance network
Figure 3 shows the directed relative importance network of PTSD and depression. In the figure, the edges specify the strength and direction of the association between two symptoms, and the arrow denotes that the predictor has a high level of relative importance for predicting the symptom that it is pointing to but does not imply causality.
Fig. 3.
Directed relative importance network of PTSD And depression symptoms. Note. Each node represents a symptom. Each edge represents the relative importance of a symptom as a predictor of Another symptom, and the thickness signifies its magnitude. Edges under 0.05 were omitted from the networks. The arrows indicate the direction of prediction
Our results show that the arrows between most symptoms are bidirectional. For example, E3 (hypervigilance) ⇆ E4 (exaggerated startle response) ⇆ E5 (difficulty concentrating) ⇆ E6 (sleeping difficulties) ⇆Dep7 (poor sleep); E1 (irritability/anger) ⇆ E2 (reckless/self-destructive behavior) ⇆Dep10 (could not get going) ⇆Dep1 (bothered by things) ⇆Dep2 (difficulty keeping mind) ⇆Dep3 (depressed mood) ⇆Dep4 (everything is an effort). In addition, E4 (exaggerated startle response), Dep2 (difficulty keeping mind), E3 (hypervigilance), and D5 (diminished interest) higher outstrength values, reflecting they exert a more potent and far-reaching influence on a multitude of other symptoms within the network (see Table 2).
Table 2.
Indexes of centrality for relative importance network
| Node | Outstrength | Closeness | Betweenness |
|---|---|---|---|
| B1 | 0.650 | 0.022 | 0.034 |
| B2 | 0.699 | 0.022 | 0.030 |
| B3 | 0.698 | 0.023 | 0.022 |
| B4 | 0.610 | 0.020 | 0.007 |
| B5 | 0.610 | 0.020 | 0.022 |
| C1 | 0.576 | 0.018 | 0.005 |
| C2 | 0.626 | 0.019 | 0.018 |
| D1 | 0.264 | 0.017 | 0.000 |
| D2 | 0.563 | 0.020 | 0.004 |
| D3 | 0.656 | 0.021 | 0.032 |
| D4 | 0.692 | 0.022 | 0.015 |
| D5 | 0.727 | 0.022 | 0.039 |
| D6 | 0.712 | 0.022 | 0.000 |
| D7 | 0.709 | 0.023 | 0.027 |
| E1 | 0.671 | 0.022 | 0.043 |
| E2 | 0.635 | 0.022 | 0.078 |
| E3 | 0.741 | 0.022 | 0.022 |
| E4 | 0.785 | 0.022 | 0.049 |
| E5 | 0.715 | 0.021 | 0.025 |
| E6 | 0.652 | 0.020 | 0.032 |
| Dep1 | 0.701 | 0.024 | 0.179 |
| Dep2 | 0.768 | 0.024 | 0.123 |
| Dep3 | 0.697 | 0.021 | 0.023 |
| Dep4 | 0.637 | 0.020 | 0.001 |
| Dep5 | 0.358 | 0.013 | 0.000 |
| Dep6 | 0.589 | 0.020 | 0.010 |
| Dep7 | 0.502 | 0.020 | 0.007 |
| Dep8 | 0.370 | 0.014 | 0.065 |
| Dep9 | 0.529 | 0.020 | 0.009 |
| Dep10 | 0.605 | 0.021 | 0.010 |
Discussion
The current study employed two network analysis approaches—a regularized partial correlation and relative importance network—to map the relationship between PTSD and depression and to identify how symptoms interact with each other in comorbidity conditions. Our findings provide valuable insights into the symptom-level interactions of these two disorders among Shidu parents.
The partial correlation network analysis revealed several important results. First, PTSD and depression symptoms emerged as two discrete subnetworks, with the strongest connections appearing among almost all the nodes within each disorder. For example, the strongest connections were observed from recurrent thoughts to nightmares (B1–B2), hypervigilance to an exaggerated startle response (E3–E4), hopelessness about the future to unhappiness (Dep5–Dep8), and depressed mood to everything is an effort (Dep3–Dep4). These findings suggest that within-disorder symptom connectivity is greater than between-disorder connectivity, which is consistent with previous studies indicating that PTSD and depression are two separate disorders with mutually influential symptom structures [13, 16, 34].
Second, we identified several important bridge symptoms, including diminished interest, irritability/anger, reckless/self-destructive behavior, a negative emotional state, the avoidance of reminders, and flashbacks, which reaffirmed previous findings [8, 14, 16]. However, we did not find a bridging role of sleep problems and concentration difficulties, which is inconsistent with some previous research [13, 15]. These inconsistent findings suggest that the two symptoms are not the primary intermediate factors that facilitate the interaction between PTSD and depression. Instead, they may be associated with other core symptoms, such as exaggerated startle response and hypervigilance among Shidu parents. One possible explanation for this phenomenon is the cultural tendency in China to avoid discussing topics related to bereavement, leading relatives and friends to mention such topics less frequently [35]. As a result, Shidu parents may suppress many of their somatic symptoms, while sleep and attention problems are gradually rationalized during their adaptation to trauma. Consequently, comorbidity is manifested through cognitive-emotional symptoms. Future research should consider adopting longitudinal designs and exploring the neural mechanisms involved to gain a deeper understanding of the underlying processes.
Diminished interest was identified as an important bridge symptom in the PTSD and depression network, which was in line with the findings of a previous study [8]. This is also the most central symptom in the PTSD network among Shidu parents [21]. These results indicate that diminished interest plays an important role in the development of PTSD symptoms, as well as in the comorbidity of PTSD and depression. This may be related to the collapse of the meaning system and the lack of social support for Shidu parents. Children are the main source of parents’ hope and meaning in life [19]. Losing their only child means losing spiritual sustenance, and parents who have experienced this often feel that their lives have become meaningless and hopeless, which may lead to diminished interest in things and people around them [20, 35]. Moreover, diminished interest is associated with detachment in Shidu parents [21], and these parents have fewer social contacts. Limited social resources (e.g., social support) may weaken an individual’s ability to cope with trauma, increasing the comorbidity of PTSD and depression.
Symptoms of irritability/anger and reckless/self-destructive behavior were also important bridging nodes. Moreover, reckless/self-destructive behavior was strongly associated with irritability/anger and could not get going (depression symptoms) in the current network. In Chinese culture, children play an important economic and social support role for parents, and this support is also vital for generational continuity [36]. Suddenly losing an only child is more devastating for Shidu parents and makes them prone to irritability/anger and reckless/self-destructive behavior, which may make parents feel could not get going. These findings may be linked to dysregulation in neural circuits involved in emotional processing and impulse control [37, 38]. Heightened irritability/anger and reckless/self-destructive behavior could reflect impaired prefrontal control over amygdala-driven emotional responses [37], while “could not get going” (depressive symptoms) might relate to reduced dopaminergic activity in the mesolimbic reward system [38]. Such neural dysfunctions, triggered by the traumatic loss of an only child, disrupt emotional regulation and motivational pathways, aligning with the observed symptom interconnections in Shidu parents. In addition, a negative emotional state, the avoidance of reminders, and flashbacks were identified as bridging symptoms. These results are consistent with those of previous studies among different populations, such as veterans and firefighters [8, 14, 16], emphasizing their significance in the comorbidity network of PTSD and depression among Shidu parents.
The relative importance network analysis revealed two strong bidirectional pathways, highlighting the mechanisms underlying the comorbidity of PTSD and depression. The first pathway (hypervigilance (E3) ⇆ exaggerated startle response (E4) ⇆ difficulty concentrating (E5) ⇆ sleeping difficulties (E6) ⇆ poor sleep (Dep7)) illustrates a physiological-cognitive-emotional cascade: physiological hyperarousal triggers cognitive impairments and sleep disturbances, ultimately culminating in depressive symptoms. This supports the hypothesis that the physiological symptoms of PTSD (e.g., hypervigilance, startle reactivity) may act as risk factors for depression by disrupting cognitive and sleep processes [39]. In the context of Chinese culture, for Shidu parents, the traditional emphasis on family lineage continuation and the central role of children in family life [18] intensifies the physiological stress response. The sense of losing the “successor” can lead to persistent hypervigilance, as they constantly relive the trauma of loss, which further disrupts cognitive and sleep functions and increases the likelihood of developing depressive symptoms.
The second pathway (irritability/anger (E1) ⇆ reckless/self-destructive behavior (E2) ⇆ could not get going (Dep10) ⇆ bothered by things (Dep1)) illustrates a vicious cycle of emotional and behavioral dysregulation that contributes to core depressive symptoms. Notably, this pathway’s eventual connection to the re-experiencing symptoms of PTSD suggests that emotional dysregulation may intensify intrusive traumatic memories [40], further entrenching the comorbidity. Together, these two pathways demonstrate that PTSD and depression mutually reinforce each other through interconnected biological and psychological mechanisms: physiological hyperreactivity impairs cognition and sleep, exacerbating emotional distress, while emotional dysregulation heightens physiological arousal, creating a feedback loop. Neurologically, these interactions may arise from dysregulation between the amygdala (which drives hypervigilance and startle responses) [37] and the prefrontal cortex (which is responsible for attention and emotion regulation) [38]. This dysfunction may perpetuate symptom connections, as impaired top-down control fails to modulate amygdala-driven stress responses [41].
In addition, the relative importance network analysis identified an exaggerated startle response (E4), difficulty keeping mind (Dep2), hypervigilance (E3), and diminished interest (D5) as the most influential nodes among all PTSD and depression symptoms. This suggests their potential role in driving the maintenance and exacerbation of this comorbid condition. For Shidu parents, the cultural significance attached to family and the social stigma associated with losing an only child [18] can intensify these symptoms. The constant social pressure and self-blame within this cultural context [20] make it more challenging for them to recover from the effects of these symptoms, thereby reinforcing the role of these nodes in the comorbid condition.
This study should be interpreted in light of several limitations. First, we utilized cross-sectional data, and it is not possible to infer direct causation between symptoms. Although relative importance network analysis can represent the relationship directions between pairs of nodes from cross-sectional data, the analysis is constrained by some strict assumptions that limit potential inferences. Future longitudinal studies are needed to evaluate the comorbidity patterns of PTSD and depression. Second, PTSD and depression symptoms were assessed via a self-report questionnaire instead of a clinical assessment. Future studies should assess comorbidity network structures on the basis of clinical observations to replicate and validate our findings. Finally, the participants in the current study were limited to Chinese Shidu parents, and the sample size was small because of difficulties accessing this group. These considerations may limit the generalizability of our conclusions. Therefore, the findings of this study must be interpreted with caution, and the conclusions should be generalized carefully with awareness of potential stability issues in the data.
Conclusions
The current study employed regularized partial correlation network and relative importance network analyses to enhance our understanding of the comorbidity of PTSD and the onset and progression of depression. Our findings reinforce the notion that PTSD and depression are independent yet mutually influential, with key bridge symptoms and bidirectional pathways highlighting the mechanisms of comorbidity among Shidu parents. These findings underscore the necessity for targeted interventions that address critical symptoms in this vulnerable population.
Supplementary Information
Acknowledgements
We extend our sincere thanks to all participants and organizations that supported this research.
Abbreviations
- PTSD
Posttraumatic stress disorder
- PCL-5
Posttraumatic Stress Disorder Checklist for the DSM-5
- CES-D-10
Center for Epidemiological Studies-Depression Scale
- KMO value
Kaiser-Meyer-Olkin value
- LASSO
Least absolute shrinkage and selector operator
- EBIC
Extended Bayesian information criterion
- bEI
Bridge expected influence
- lmg
Lindeman-Merenda-Gold
Authors’ contributions
Buzohre Eli: Data curation, Formal analysis, Methodology, Writing–original draft, Writing–review & editing. Xuanang Liu: Methodology, Writing–review & editing. Zhengqing Zhu: Methodology, Writing–review & editing. Fei Xiao: Writing–review & editing. Zhengkui Liu: Conceptualization, investigation, project administration, resources, supervision.
Funding
This work was supported by the National Key R&D Program of China (2023YFC3605304), the Shihezi University High-level Talent Research Launch Project (RCSK202435), the Tianchi Talent Program of Xinjiang Uygur Autonomous Region (2024, Buzohre Eli), and Shihezi University self-funded support for university-level research projects (ZZZC2023076).
Data availability
Due to considerations of participants’ privacy, the data that support the findings of this study are available from the corresponding author, upon reasonable request.
Declarations
Ethics approval and consent to participate
This study was conducted in accordance with the Declaration of Helsinki guidelines. It was approved by the ethics review committee of the Institute of Psychology, Chinese Academy of Sciences (Ethics approval number: H21044). All the surveys were conducted after informed consent was obtained from the participants.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Youngblut JM, Brooten D, Cantwell GP, del Moral T, Totapally B. Parent health and functioning 13 months after infant or child NICU/PICU death. Pediatrics. 2013;132:e1295–301. 10.1542/peds.2013-1194. [DOI] [PMC free article] [PubMed]
- 2.Aho AL, Malmisuo J, Kaunonen M. The effects of peer support on post-traumatic stress reactions in bereaved parents. Scand J Caring Sci. 2017;32:326–34. 10.1111/scs.12465. [DOI] [PubMed]
- 3.Rytwinski NK, Scur MD, Feeny NC, Youngstrom EA. The co-occurrence of major depressive disorder among individuals with posttraumatic stress disorder: a meta-analysis. J Trauma Stress. 2013;26:299–309. 10.1002/jts.21814. [DOI] [PubMed] [Google Scholar]
- 4.Chen E. The number of the only-child-died-families in China. Popul Dev. 2013;19(6):100–3. https://kns.cnki.net.psych.remotexs.cn.
- 5.Eli B, Chen Y, Zhang J, Huang X, Wang Q, Ma Z, Yv Y, Liu Z. Time course of attentional bias and its relationship with PTSD symptoms in bereaved Chinese parents who have lost their only child. Eur J Psychotraumatol. 2023;14: 2235980. 10.1080/20008066.2023.2235980. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Zhang Y, Jia X. Mental health status of the shiduers: based on latent profile analysis. Chin J Clin Psychol. 2019;27:362–6. 10.16128/j.cnki.1005-3611.2019.02.031.
- 7.Flory JD, Yehuda R. Comorbidity between post-traumatic stress disorder and major depressive disorder: alternative explanations and treatment considerations. Dialogues Clin Neurosci. 2015;17:141–50. 10.31887/DCNS.2015.17.2/jflory. [DOI] [PMC free article] [PubMed]
- 8.Afzali MH, Sunderland M, Teesson M, Mills NC, Slade K. A network approach to the comorbidity between posttraumatic stress disorder and major depressive disorder: the role of overlapping symptoms. J Affect Disord. 2017;208:490–6. 10.1016/j.jad.2016.10.037. [DOI] [PubMed]
- 9.Borsboom D, Cramer AO. Network analysis: an integrative approach to the structure of psychopathology. Annu Rev Clin Psychol. 2013;9:91–121. 10.1146/annurev-clinpsy-050212-185608. [DOI] [PubMed] [Google Scholar]
- 10.Armour C, Fried EI, Deserno MK, Tsai J, Pietrzak RH. A network analysis of DSM-5 posttraumatic stress disorder symptoms and correlates in U.S. military veterans. J Anxiety Disord. 2017;45:49–59. 10.1016/j.janxdis.2016.11.008. [DOI] [PubMed] [Google Scholar]
- 11.Fried EI, van Borkulo CD, Cramer AO, Boschloo L, Schoevers RA, Borsboom D. Mental disorders as networks of problems: a review of recent insights. Soc Psychiatry Psychiatr Epidemiol. 2017;52:1–10. 10.1007/s00127-016-1319-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Jones PJ, Ma R, McNally RJ. Bridge centrality: a network approach to understanding comorbidity. Multivar Behav Res. 2021;56:353–67. 10.1080/00273171.2019.1614898. [DOI] [PubMed]
- 13.An Y, Shi J, Chuan-Peng H, Wu X. The symptom structure of posttraumatic stress disorder and co-morbid depression among college students with childhood abuse experience: a network analysis. J Affect Disord. 2021;293:466–75. 10.1016/j.jad.2021.06.065. [DOI] [PubMed]
- 14.Cheng P, Wang L, Zhou Y, Ma W, Zhao G, Zhang L, Li W. Post-traumatic stress disorder and depressive symptoms among firefighters: a network analysis. Front Public Health. 2023;11: 1096771. 10.3389/fpubh.2023.1096771. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Duek O, Spiller TR, Pietrzak RH, Fried EI, Harpaz-Rotem I. Network analysis of PTSD and depressive symptoms in 158,139 treatment-seeking veterans with PTSD. Depress Anxiety. 2021;38:554–62. 10.1002/da.23112. [DOI] [PubMed]
- 16.Lazarov A, Suarez-Jimenez B, Levy O, Coppersmith DDL, Neria Y. Symptom structure of PTSD and co-morbid depressive symptoms – a network analysis of combat veteran patients. Psychol Med. 2020;50:2154–70. 10.1017/S0033291719002034. [DOI] [PMC free article] [PubMed]
- 17.Thoits PA. Identity-relevant events and psychological symptoms: a cautionary tale. J Health Soc Behav. 1995;36:72–82. 10.2307/2137288. [PubMed] [Google Scholar]
- 18.Zheng Y, Lawson TR. Identity reconstruction as shiduers: narratives from Chinese older adults who lost their only child. Int J Soc Welfare. 2015;24:399–406. 10.1111/IJSW.12139. [Google Scholar]
- 19.Feng Z. Childlessness and vulnerability of older people in China. Age Ageing. 2018;47:275–81. 10.1093/ageing/afx137. [DOI] [PMC free article] [PubMed]
- 20.Wang N, Hu Q. It is not simply the loss of a child: the challenges facing parents who have lost their only child in post-reproductive age in China. Death Stud. 2019;45:209–18. 10.1080/07481187.2019.1626941. [DOI] [PubMed]
- 21.Eli B, Liang Y, Chen Y, Huang X, Liu Z. Symptom structure of posttraumatic stress disorder after parental bereavement – a network analysis of Chinese parents who have lost their only child. J Affect Disord. 2021;295:673–80. 10.1016/j.jad.2021.08.123. [DOI] [PubMed]
- 22.Benfer N, Bardeen JR, Cero I, Kramer LB, Whiteman SE, Rogers TA, Silverstein MW, Weathers FW. Network models of posttraumatic stress symptoms across trauma types. J Anxiety Disord. 2018;58:70–7. 10.1016/j.janxdis.2018.07.004. [DOI] [PubMed]
- 23.Garabiles MR, Lao CK, Siyuan W, Hall BJ. The network structure of posttraumatic stress disorder among filipina migrant domestic workers: comorbidity with depression. Eur J Psychotraumatol. 2020;11:1765544. 10.1080/20008198.2020.1765544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Gilbar O. Examining the boundaries between ICD-11 PTSD/CPTSD and depression and anxiety symptoms: a network analysis perspective. J Affect Disord. 2020;262:429–39. 10.1016/j.jad.2019.11.060. [DOI] [PubMed]
- 25.Epskamp S, Borsboom D, Fried EI. Estimating psychological networks and their accuracy: a tutorial paper. Behav Res Methods. 2018;50:195–212. 10.3758/s13428-017-0862-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Brusco M, Watts AL, Steinley D. A modified approach to fitting relative importance networks. Psychol Methods. 2022;29(1):1–20. 10.1037/met0000496. [DOI] [PubMed] [Google Scholar]
- 27.Weathers FW, Litz B, Keane TM, Palmieri PA, Marx BP, Schnurr PP.The PTSDchecklist for DSM-5 (PCL-5). Scale available from the National Center for PTSD. 2013. https://www.ptsd.va.gov.
- 28.Andresen E, Malmgren JA, Carter WB, Patrick DL. Screening for depression in well older adults: evaluation of a short form of the CES-D (center for epidemiologic studies depression scale). Am J Prev Med. 1994;10:77–84. 10.1016/S0749-3797(18)30622. [PubMed] [Google Scholar]
- 29.Epskamp S, Cramer AOJ, Waldorp LJ, Schmittmann VD, Borsboom D. Qgraph: network visualizations of relationships in psychometric data. J Stat Softw. 2012;48:1–18. 10.18637/jss.v048.i04. [Google Scholar]
- 30.Fruchterman TMJ, Reingold M. Graph drawing by force-directed placement. Softw Pract Experience. 1991;21:1129–64. 10.1002/spe.4380211102.
- 31.Grömping U. Relative importance for linear regression in R: the package relaimpo. J Stat Softw. 2006;17:1–27. 10.1080/10629360600569147. [Google Scholar]
- 32.Bringmann LF, Elmer T, Epskamp S, Krause RW, Schoch D, Wichers M, Wigman JTW, Snippe E. What do centrality measures measure in psychological networks? J Abnorm Psychol. 2019;128:892–903. 10.1037/abn0000446. [DOI] [PubMed] [Google Scholar]
- 33.Yang F, Huang N, Zhang B, Fu M, Guo J. Network analysis of COVID-19-related PTSD and depressive symptoms among Chinese general population and comorbidity subgroup. Psychol Trauma Theory Res Pract Policy. 2023;15:431–42. 10.1037/tra0001434. [DOI] [PubMed]
- 34.Qi J, Ye Y, Sun R, Zhen R, Zhou X. Comorbidity of posttraumatic stress disorder and depression among adolescents following an earthquake: a longitudinal study based on network analysis. J Affect Disord. 2023;324:354–63. 10.1016/j.jad.2022.12.119. [DOI] [PubMed]
- 35.Eli B. Latent profiles and stage characteristics of posttraumatic stress disorder symptoms among parents who lost their only child[Master’s thesis]. University of Chinese Academy of Sciences, Beijing. 2019.
- 36.Wei Y, Jiang Q, Gietel-Basten S. The well-being of bereaved parents in an only-child society. Death Stud. 2016;40:22–31. 10.1080/07481187.2015.1056563. [DOI] [PubMed] [Google Scholar]
- 37.Kleshchova O, Rieder JK, Grinband J, Weierich MR. Resting amygdala connectivity and basal sympathetic tone as markers of chronic hypervigilance. Psychoneuroendocrinology. 2019;102:68–78. 10.1016/j.psyneuen.2018.11.036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Fitzgerald JM, Digangi JA, Phan KL. Functional neuroanatomy of emotion and its regulation in PTSD. Harv Rev Psychiatry. 2018;26(3):116–28. 10.1097/HRP.0000000000000185. [DOI] [PMC free article] [PubMed]
- 39.Adenauer H, Pinösch S, Catani C, Gola H, Keil J, Kissler J, Neuner F. Early processing of threat cues in posttraumatic stress disorder-evidence for a cortical vigilance-avoidance reaction. Biol Psychiatry. 2010;68(5):451–8. 10.1016/j.biopsych.2010.05.015. [DOI] [PubMed]
- 40.Battaglia S, Fazio CD, Borgomaneri S, Avenanti A. Cortisol imbalance and fear learning in PTSD: therapeutic approaches to control abnormal fear responses. Curr Neuropharmacol. 2025;23(7):835–46. 10.2174/1570159X23666250123142526. [DOI] [PMC free article] [PubMed]
- 41.Jhaveri DJ, Mcgonigal A, Becker C, Benoliel JJ, Nandam LS, Soncin L, Kotwas L, Bernard C, Bartolomei F. Stress and epilepsy: towards understanding of neurobiological mechanisms for better management. eNeuro. 2023;10(11):0200232023. 10.1523/ENEURO.0200-23.2023. [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Due to considerations of participants’ privacy, the data that support the findings of this study are available from the corresponding author, upon reasonable request.