Simulating symptom-specific interventions linking depression and anxiety among civil aviation college students: a network analysis differentiated by perceived parental relationships.

Simulating symptom-specific interventions linking depression and anxiety among civil aviation college students: a network analysis differentiated by perceived parental relationships - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice BMC Psychol . 2026 Apr 20;14:805. doi: 10.1186/s40359-026-04553-y Simulating symptom-specific interventions linking depression and anxiety among civil aviation college students: a network analysis differentiated by perceived parental relationships Hongzhi Hu 1, 2 , Lang Liang 2 , Zhongfa Ouyang 3 , Yijing Song 4, ✉ Author information Article notes Copyright and License information 1 School of Marxism, Southwest Jiaotong University, Chengdu, China 2 Chengdu Aeronautic Polytechnic University, Chengdu, China 3 College Student Affairs Department, Sichuan Nursing Vocational College, Chengdu, China 4 Teacher Education Development Center, Sichuan Tourism University, Chengdu, China ✉ Corresponding author. Received 2026 Jan 15; Accepted 2026 Apr 10; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13227748  PMID: 42010725 Abstract Depression and anxiety frequently co-occur among civil aviation college students, and parental relationships have been shown to play an important role in the development of this comorbidity. However, the symptom-level mechanisms underlying depression–anxiety comorbidity remain unclear among them. Examining this comorbidity from a network perspective, while accounting for perceived parental relationships, may help predict key bridge symptoms and simulate interventions for these bridge symptoms. A total of 1,033 participants completed the Patient Health Questionnaire-9 and Generalized Anxiety Disorder-7 scale that assessed depressive and anxiety symptoms. Then, based on their self-reported perceptions of parental relationships, 503 students were classified into a group with a harmonious parental relationship and 530 into a group with a disharmonious parental relationship. Symptom network analysis, Bayesian-based directed acyclic graph analysis, and the symptom-specific intervention simulation analysis were conducted. Findings indicated that although there was no significant difference in the overall network structure, the symptom associations differed markedly between the two groups. In the harmonious group, “ guilt ”, “ motor disturbances ”, “ uncontrollable worrying ”, and “ feeling afraid ” linked the depression–anxiety comorbidity ( p < .05), whereas in the disharmonious group, the depression-anxiety comorbidity was associated with “ motor disturbances ” and “ irritability ” ( p < .05). In addition, “ fatigue ” may be the most effective alleviating intervention target for both groups. Besides, “ depressed or sad mood ” and “ motor disturbances ” could be considered separately for preventive care when dealing with those with harmonious and disharmonious parental relationships, respectively. The findings suggest potential implications for parental relationship-specific strategies to alleviate the linkage between depression and anxiety symptoms among civil aviation college students. Supplementary Information The online version contains supplementary material available at 10.1186/s40359-026-04553-y. Keywords: Depression, Anxiety, Civil aviation college students, Symptom network, Bayesian analysis, Simulation intervention Introduction The simultaneous presence of anxiety and depression is reported with considerable frequency, prompting clinical researchers to describe this phenomenon as a major comorbidity challenge [ 1 ]. Depression is widely understood as a complex psychopathological syndrome involving multiple symptoms, such as persistent low mood, pervasive guilt, and pronounced worthlessness [ 2 ]. As a frequent co-occurring mental disorder with depression, anxiety denotes anticipatory distress about forthcoming events and is associated with features like irritability, physical tension, and sleep disturbances [ 3 ]. Regardless of the diagnostic criteria or sample characteristics used, the comorbidity rate of anxiety and depression remains persistently high, ranging from 40% to as high as 68% [ 4 ]. Compared to those with a single psychiatric disorder, people with co-occurring anxiety and depression tend to demonstrate poorer cognitive performance, higher self-injury or suicidal thoughts, and poorer prognoses [ 5 ]. The transition to university life introduces numerous developmental challenges, including academic pressure, social adjustment, and identity formation, which significantly increase vulnerability to such comorbidity [ 1 ]. A recent study showed that approximately 40% of university students experienced clinically relevant depressive or anxiety symptoms, and co-occurrence is highly common [ 6 ]. It is crucial to emphasize that the pressures faced by civil aviation college students, such as quasi-military discipline, heightened awareness of occupational risks, and the burden of a lofty mission, are highly distinctive [ 7 ]. A study on civil aviation students found that their mental health is not optimistic, with prominent issues in adaptability, anxiety, and emotional dysregulation, particularly among first-year students [ 8 ]. These unique pressures may give rise to psychological response mechanisms that differ significantly from those observed in the general university student population. Consequently, research specifically targeting this group is not only necessary but constitutes the scientific foundation for implementing precise mental health services. Throughout the extensive literature, scholars have increasingly investigated the role of parental conflict in shaping their children’s mental health [ 9 – 11 ]. Prior research indicates that such interparental conflict can exacerbate the occurrence of anxiety and depression [ 12 – 14 ]. According to the family system theory [ 15 ], the family constitutes an interconnected emotional system in which dysfunction within the parental subsystem (i.e., the marital relationship) inevitably spills over into the parent-child subsystem. Children may become entangled in parental conflicts, leading them to bear excessive responsibility and chronic stress, which can contribute to emotional distress and the development of severe psychological issues. For instance, a meta-analysis revealed that compared with parent-reported parental conflict, children-reported interparental conflict has a stronger correlation with depressive symptoms, highlighting parental conflict as an important risk factor for individual depression [ 16 ]. From the perspective of social learning theory [ 17 ], children acquire strategies for coping with stress and regulating emotions by observing parental behaviors. Conflictual parents model negative interpersonal interactions, which can teach children maladaptive emotion-regulation strategies or foster a negative attributional style [ 18 ], a core cognitive factor in depression. Therefore, a deeper understanding of how perceived parental relationships influence the co-occurrence of anxiety and depression in civil aviation college students is critical for early identification of at-risk individuals and for guiding targeted interventions. According to the network-symptom theory of psychopathology [ 19 , 20 ], mental disorders arise from the causal interactions among symptoms themselves. In the corresponding symptom network analysis, each symptom is treated as a node, and the statistical associations between nodes are represented by edges. The comorbidity between distinct disorders can also be represented by edges connecting bridge symptoms, and identifying these bridge symptoms may help clarify specific and targeted treatment strategies [ 21 ]. In short, the pathways connecting different symptoms can be visualized in the network structure, thereby facilitating the identification of the most influential bridge symptoms that drive the emergence of this whole comorbidity. For instance, Cheung et al. [ 22 ] found that guilt , sad mood , and energy symptoms served as key bridge symptoms in the depressive symptom network. In another network analysis study on depression-anxiety comorbidity, sad mood and worry were identified as bridging symptoms [ 3 ]. However, pathological symptoms in individuals exhibit dynamic characteristics and may fluctuate or even recur over the course of intervention. Although traditional cross-sectional symptom network analysis is capable of identifying key symptoms through centrality measures, it still has limitations in helping to predict potential changes in symptom networks during the course of treatment. To address this limitation, the NodeIdentifyR algorithm (NIRA) was recently developed as a simulation-based network approach that models the impact of specific symptom-level interventions [ 23 ]. NIRA generates projected networks by manipulating selected symptoms to observe downstream effects, thereby identifying optimal prevention and treatment targets based on the magnitude of network change. Lunansky et al. [ 23 ] applied NIRA to PTSD symptoms among survivors of the 2008 Wenchuan earthquake and identified intrusive thoughts as the most effective treatment target, a finding subsequently supported by empirical evidence. Based on existing literature, parental relationships are significantly associated with depression and anxiety symptoms among civil aviation college students. However, the underlying symptomatic linkage mechanisms between parental relationships and the comorbidity of depression and anxiety remain unclear, let al.one utilizing simulated interventions to better identify which symptoms are most critical from a clinical perspective. It is crucial to note that mental health problems, particularly depressive disorders, have become the leading medical cause for grounding military pilots [ 24 ]. Given that civil aviation university students serve as the sole pipeline and the pre-service reserve for pilots, their mental health status directly impacts the future stability and operational effectiveness of the flight corps. However, there is a lack of systematic investigation into the mental health of this critical source population during their key formative period. To ensure the precision of intervention strategies, it is essential to clarify the symptom-level connections between these conditions. Therefore, this study aims to further elucidate the intrinsic transmission mechanisms between depression and anxiety by employing symptom network analysis and simulated interventions within network models, while accounting for parental relationship factors. Guided by family systems theory and social learning theory, which emphasize the impact of family relational patterns on individual emotional processes, this study aims to clarify: 1) how the symptom network structure of depression-anxiety comorbidity differs between individuals with harmonious versus discordant perceptions of their parental marital relationship, and which symptoms form the connections within each network; 2) Which symptoms exert the greatest predicted influence on reducing or increasing the activation of depression and anxiety comorbidity across varying parental marital relationships. Method Procedures and participants Between October and December 2024, 1,788 first-year students in Chengdu were recruited to participate in the current study and completed the Patient Health Questionnaire-9 (PHQ-9; [ 25 ]) and the Generalized Anxiety Disorder-7 (GAD-7; [ 26 ]) using Wenjuanxing, a commonly used online survey platform in China ( https://www.wjx.cn ). After quality control, 144 participants who failed the instructed-response item (i.e., “ please select the ‘Completely true.‘ ”) and 444 who showed too fast responses (less than two seconds for any item; [ 27 ]) were excluded. Among the remaining 1,200 participants (valid response rate = 67.1%), 167 reported parental divorce, separation, or the death of one parent and were therefore excluded as well. The final sample comprised 1033 participants (valid response rate = 67.1%; 520 males and 513 females; M age = 18.63, SD age = 0.85). Based on their responses to the item “How is your parents’ relationship?“ , 503 participants were categorized into the harmonious parental relationship group (i.e., “ Harmonious relationship , never quarrel ”), while the remaining 530 were classified into the disharmony group (i.e., “ Occasionally quarrel , about 1–2 times per month ,” “ Often quarrel , several days each month ,” or “ Always quarrel , almost every day ”). Before participation, all individuals provided electronic informed consent and were informed of their ability to withdraw from the study at any time. This study was reviewed and approved by the Ethics Committee of the first author. Measures Patient Health Questionnaire-9 (PHQ-9) Depressive symptoms were measured using the Patient Health Questionnaire-9 (PHQ-9; [ 25 ]). Participants reported the frequency of each symptom, such as “ depressed mood ,” “ fatigue ,” and “ guilt ,” over the previous two weeks on a four-point Likert scale, ranging from 0 (not at all) to 3 (n early every day ). Higher aggregate scores indicated more severe depressive symptoms. In the present sample, the PHQ-9 demonstrated excellent internal consistency, with a Cronbach’s α of 0.89. Generalized Anxiety Disorder-7 scale (GAD-7) Anxiety symptoms were assessed using the Generalized Anxiety Disorder-7 scale (GAD-7; [ 26 ]). Participants indicated how often they had experienced symptoms, such as “ nervousness ” and “ trouble relaxing ,” over the past two weeks, using a four-point Likert scale from 0 ( Not at all ) to 3 ( Nearly every day ). Higher total scores reflected greater severity of anxiety. In this study, the GAD-7 demonstrated strong internal consistency, with a Cronbach’s α of 0.93. Statistical analysis All statistical analyses were conducted using R software (version 4.5.2, [ 28 ]). Descriptive statistics were computed for participants’ demographic characteristics (e.g., age, gender, and only-child status) for the total sample, the harmonious parental relationship group, and the disharmonious parental relationship group, and group differences were also examined. Related to the purpose of the current study, including depression and anxiety, descriptive statistics and group comparisons were performed on both the original continuous symptom scores and the total scores. To estimate the Ising network and perform corresponding simulated interventions, each symptom score was dichotomized: an original score of 0 (“ not at all ”) was recoded as 0, while scores of 1–3 (“several days” to “nearly every day”) were recoded as 1. The group comparisons on the dichotomized symptom scores were also performed. Then, we conducted the following network analyses. Estimating network structure and centrality The depression-anxiety Ising network was estimated for both the harmonious and the disharmonious parental relationship groups. Network estimation and visualization were conducted using the R packages bootnet (version 1.4.3) and qgraph (version 1.6.9) [ 29 , 30 ]. The Ising network was constructed using the IsingFit method in the estimateNetwork function, which is specifically designed for binary data [ 31 ]. Logistic regression was employed to estimate the network parameters by regressing each dichotomized symptom on all other symptoms. The two main parameters of interest were edge weights and thresholds. Edge weights, obtained from the regression coefficients, reflected the strength of associations between nodes within the network. Thresholds, derived from the regression intercepts, indicated each symptom’s baseline tendency to be present. Positive thresholds suggest a higher likelihood of symptom activation when all other symptoms are absent, whereas negative thresholds indicate a tendency toward non-activation. In the visualized network, each node corresponded to a specific symptom of depression or anxiety, while edges represented the connections between symptoms, quantified by edge weights. Positive relationships were depicted in green and negative relationships in red, with the thickness of an edge reflecting the strength of the association [ 19 ]. We then calculated each node’s Expected Influence (EI) and bridge EI using the centralityPlot function [ 32 ]. EI represents the sum of all edge weights connected to a node, indicating its overall importance within the network. Bridge EI measures the total strength of one node’s connections within a mental community to other nodes in a different mental community, reflecting its role in linking distinct mental communities [ 33 ]. Following previous studies [ 34 ], nodes with standardized EI and bridge EI scores greater than 1 were considered core symptoms and bridge symptoms, respectively. Due to differences in gender distribution between the harmonious and disharmonious groups, we additionally estimated Ising networks including gender as a covariate. The similarity between these networks and those estimated without the covariate was assessed using a Mantel test based on Pearson correlations between the adjacency edge-weight matrices [ 35 ]. A high correlation ( r ≥ .700) indicates a high degree of similarity, suggesting that gender did not substantially influence the network estimation. In addition, because dichotomizing PHQ-9 and GAD-7 symptom items may lead to a loss of information, we estimated partial correlation networks for both groups using the original continuous item scores. Specifically, two Gaussian graphical models (GGMs) were estimated with the estimateNetwork function using the “ EBICglasso ” algorithm [ 36 ]. The resulting adjacency matrices were then compared with those derived from the corresponding dichotomized Ising networks using the same Mantel test procedure [ 37 – 39 ]. A high correlation between the two matrices would indicate that the binarization of the data did not substantially distort the underlying network structure. Estimating network stability and accuracy The accuracy and stability of the networks were examined using the bootnet R package (version 1.4.3 [ 29 ]). First, nonparametric bootstrapping was performed to generate 95% confidence intervals ( CI s) for edge weights, with narrower CI s indicating more reliable edge estimates. Next, centrality stability was evaluated using a case-dropping bootstrap procedure, producing the correlation stability coefficient (CS-C). The CS-C represents the maximum proportion of the sample that can be removed while still retaining, with 95% confidence, a correlation of at least 0.7 between centrality estimates from the full and reduced datasets. CS-C values above 0.25, 0.50, and 0.75 were interpreted as acceptable, good, and excellent stability, respectively [ 40 ]. Finally, the nonparametric bootstrapped difference tests were conducted to determine whether differences between edge weights and node centrality measures were statistically significant. Network comparison test Group differences in the depression-anxiety networks were explored using the R package NetworkComparisonTest (NCT; [ 41 ]). NCT provides a network structure invariance test and a global strength invariance test to assess differences in overall network structure and total connectivity (i.e., the sum of all edge weights), respectively. When a significant difference in network structure is detected, follow-up analyses may be conducted to identify specific edge differences [ 41 ]. In addition, the centrality invariance test in NCT was used to evaluate group differences in node centrality indices. To further assess the robustness of the findings, we also performed NCT on networks estimated from the original continuous data. This supplementary analysis aimed to examine whether the dichotomization of continuous variables influenced the observed network differences. Directed Acyclic Graph (DAG) analysis Using the R package bnlearn [ 42 , 43 ] (Scutari, 2010), a Bayesian-based DAG analysis was conducted for both groups to explore the potential causal relationships among symptoms of depression and anxiety. DAG analysis provides the strength and direction probability of each edge in the network, enabling the identification of potential causal triggers and downstream symptoms. The DAG structure was estimated using a hill-climbing algorithm, which identifies the most likely network structures by evaluating conditional independencies and optimizing model fit indices, such as the Bayesian Information Criterion (BIC; [ 44 ]). Here, BIC was calculated using bootstrapping with 1000 iterations, while the edges were retained using the optimal cut-point method [ 45 ]. The direction of an edge was determined according to the following principle: if it appeared pointing from node A to node B in at least 50% of the networks, the edge was assigned the direction from A to B [ 42 , 43 ]. The DAG network was visualized as a directed network, with upper-layer symptoms pointing to lower-layer ones. Trigger nodes, located upstream, can activate or affect other symptoms, while outcome nodes, found downstream, are influenced by these upstream nodes. Directed edges between nodes reflected the BIC values, with edge thickness indicating the extent of change in model fit when the corresponding connection was removed. Solid arrows represented negative BIC values, indicating that removing the edge worsened the model fit, while dotted arrows represented positive BIC values, suggesting that removing the edge improved the network structure [ 42 , 43 ]. Simulation of alleviating and aggravating interventions Simulation-based intervention analyses were conducted using the NodeIdentifyR algorithm (NIRA) from the IsingSampler R package [ 23 ] on the previously estimated Ising networks. NIRA computes sum scores to capture the overall symptom burden of the network and evaluates the potential effects of interventions targeting individual nodes. By systematically manipulating the threshold parameters, NIRA simulates symptom alleviation or aggravation and identifies the symptoms that most strongly influence the total network score. The impact of each intervention was quantified as the absolute difference between the network sum scores before and after the manipulation. In alleviating simulations, symptoms that most effectively reduced the total score were considered key intervention targets, whereas in aggravating simulations, those that increased the total score the most were highlighted [ 6 , 23 ]. Results Descriptive statistics and group differences Table 1 presents the detailed comparisons between participants from the harmonious and disharmonious parental relationship groups. 77.6% of the participants were not only children, and this proportion was not different between the two groups. Notably, participants in the disharmonious parental relationship group showed higher overall depression and anxiety levels ( p s ≤ 0.001, Cohen’s d s ≥ 0.54), as well as higher scores on each individual symptom ( p s ≤ 0.001, Cohen’s d s ≥ 0.27), relative to the harmonious group. Table 1. Descriptive statistics and group difference test Variable Total sample ( N = 1033) Parental relationship Group differences (a) Harmonious ( N = 503) Disharmonious ( N = 530) p Cohen’s d Demographic variables Age (in years) 18.63 (0.85) 18.65 (0.85) 18.61 (0.85) 0.489 -0.04 Sex Male 520 (50.3%) 281(54.0%) 239(43.3%) < 0.001 —— Female 513 (49.7%) 222 (46.0%) 291 (56.7%) Only child Yes 231 (22.4%) 133 (22.5%) 134 (22.3%) 1.00 —— No 802 (77.6%) 463 (77.5%) 470 (77.7%) Study variables (b) Depression (PHQ-9) 4.42 (4.47) 3.24 (3.69) 5.55 (4.85) < 0.001 0.54 PHQ1: Anhedonia 0.60 (0.67) 0.40 (0.49) 0.63 (0.48) < 0.001 0.47 PHQ2: Depressed or sad mood 0.45 (0.61) 0.30 (0.46) 0.50 (0.50) < 0.001 0.41 PHQ3: Sleep difficulties 0.64 (0.82) 0.36 (0.48) 0.55 (0.50) < 0.001 0.40 PHQ4: Fatigue 0.78 (0.80) 0.49 (0.50) 0.69 (0.46) < 0.001 0.42 PHQ5: Appetite changes 0.55 (0.72) 0.35 (0.48) 0.52 (0.50) < 0.001 0.33 PHQ6: Guilt 0.42 (0.66) 0.25 (0.43) 0.44 (0.50) < 0.001 0.40 PHQ7: Concentration difficulties 0.54 (0.72) 0.33 (0.47) 0.52 (0.50) < 0.001 0.39 PHQ8: Motor disturbances 0.31 (0.60) 0.17 (0.38) 0.32 (0.47) < 0.001 0.35 PHQ9: Suicide ideation 0.12 (0.38) 0.07 (0.25) 0.15 (0.36) < 0.001 0.27 Anxiety (GAD-7) 2.76 (3.65) 1.76 (2.78) 3.71 (4.09) < 0.001 0.56 GAD1: Nervousness 0.48 (0.63) 0.30 (0.46) 0.53 (0.50) < 0.001 0.48 GAD2: Uncontrollable worrying 0.38 (0.63) 0.20 (0.40) 0.42 (0.49) < 0.001 0.48 GAD3: Excessive worry 0.49 (0.68) 0.27 (0.45) 0.52 (0.50) < 0.001 0.51 GAD4: Trouble relaxing 0.43 (0.65) 0.25 (0.43) 0.46 (0.50) < 0.001 0.45 GAD5: Restlessness 0.29 (0.55) 0.16 (0.37) 0.33 (0.47) < 0.001 0.40 GAD6: Irritability 0.42 (0.62) 0.24 (0.43) 0.47 (0.50) < 0.001 0.48 GAD7: Feeling afraid 0.28 (0.55) 0.15 (0.35) 0.32 (0.47) < 0.001 0.42 Open in a new tab Detailed dichotomized scores for each individual symptom (0 = absent , 1 = present ) are presented in Table S1 (a) The chi-square test or Independent-Samples t-test was used to compare the groups (b) Original continuous item scores (0 = not at all , 3 = nearly every day ) and computed total scores are presented Network structures and centrality for groups Figure 1 A and C exhibited the separate network of the harmonious and disharmonious parental relationship groups (see the full edge-weight matrices in Table S2). All nodes in both networks showed negative thresholds, indicating that symptoms were most likely to be inactive when others were not activated. “ Fatigue ” (PHQ4) had the threshold value closest to zero in both the harmonious (threshold = -1.914) and disharmonious (threshold = -1.211) networks, making it the most likely symptom to be activated. Fig. 1. Open in a new tab Network structures for the harmonious ( A ) and disharmonious ( B ) parental relationship groups, and node EI centrality ( C ) and bridge EI centrality ( D ). Note. All edges in the network were green, indicating positive associations; thicker lines represent stronger connections In the harmonious network, all 71 edges were positive. “ Uncontrollable worrying ” (GAD2; EI = 1.142; Fig. 1 B) and “ fatigue ” (PHQ4; EI = 1.020) showed the highest EI centrality, emerging as the most central symptoms. GAD2 was most strongly correlated with “ nervousness ” (GAD1), “ excessive worry ” (GAD3), and “ restlessness ” (GAD5; edge weights = 1.018–1.642), while PHQ4 was most strongly correlated with “ sleep difficulties ” (PHQ3), “ appetite changes ” (PHQ5), “ concentration difficulties ” (PHQ7), and “ motor disturbances ” (PHQ8; edge weights = 0.936–1.558). Furthermore, four symptoms showed the highest bridge EI (Fig. 1 D) and were identified as key bridging symptoms: “ guilt ” (PHQ6; bridge EI = 1.778), PHQ8 (bridge EI = 1.514), GAD2 (bridge EI = 1.049), and “ feeling afraid ” (GAD7; bridge EI = 1.001). PHQ6 bridged anxiety and depression primarily through its connections with GAD3, “ trouble relaxing ” (GAD4; edge weight = 0.629), and “ feeling afraid ” (GAD7; edge weights: 0.711-989), while PHQ8 was also strongly connected to GAD3 and GAD7 (edge weights = 1.034–1.049). GAD2 bridged the two communities primarily through its connections with “ depressed or sad mood ” (PHQ2) and “ suicide ideation ” (PHQ9; edge weights = 0.748–0.804), whereas GAD7 showed a particularly strong connection with the depression symptom PHQ8 (edge weight = 1.034). In the disharmonious network, all 77 edges were positive. “ Motor disturbances ” (PHQ8) had the highest EI (EI = 1.669; Fig. 1 B) and bridge EI (EI = 1.561; Fig. 1 D) values, indicating it was the central symptom and bridging symptom in the network. PHQ8 was most strongly linked to depression symptoms such as “ suicide ideation ” (PHQ9), “ appetite changes ” (PHQ5), and “ concentration difficulties ” (PHQ7; edge weights = 0.832–1.556), as well as anxiety symptoms such as “ irritability ” (GAD6), “ restlessness ” (GAD5), and “ feeling afraid ” (GAD7; edge weights = 0.645–1.028). Moreover, GAD6 also emerged as a bridging symptom, showing the second-highest bridge EI (Bridge EI = 1.195). GAD6 bridged anxiety and depression mainly through its connections with PHQ8, “ depressed or sad mood ” (PHQ2), and “ fatigue ” (PHQ4; edge weights = 0.517–1.028). For both the harmonious and disharmonious groups, the Ising networks estimated with gender as a covariate indicated that gender was not significantly associated with other depression and anxiety symptoms (edge weights = 0). Furthermore, Mantel test results indicated that the edge-weight matrices of depression and anxiety symptoms were highly consistent between the networks estimated with and without controlling for gender (harmonious group: r = .985, p = .001; disharmonious group: r = .992, p = .001), indicating that gender had a negligible impact on the network structure. Additionally, the GGM networks estimated from the continuous data for groups are presented in Table S3 and are highly correlated with the Ising network (harmonious group: r = .746, p = .001; disharmonious group: r = .723, p = .001). These findings suggest that the binarization procedure adequately preserved the pattern of symptom associations observed in the original continuous data and did not substantially compromise the accuracy or sensitivity of the network estimation. Group differences in network The NCT results revealed no significant differences in the overall network structure ( p = .472) or the global strength (S = 0.08, p = .981). See the full results of specific edge differences in Table S4. Regarding node centrality, the disharmonious group exhibited higher EI for “ suicide ideation ” (PHQ9; diff = 3.146, p = .023), while the harmonious group exhibited marginally higher bridge EI for “ guilt ” (PHQ6; diff = 1.163, p = .093; see Table 2 and S5). Table 2. Significant network differences between groups Parental relationship Diff p Harmonious Disharmonious Node EI centrality PHQ9: Suicide ideation 3.224 6.370 3.146 0.023 * Node bridge EI centrality PHQ6: Guilt 3.121 1.958 -1.163 0.093 † Open in a new tab The higher Node EI, and Node bridge EI were in bold † p < .1 * p < .05. See the full results in Table S5 In addition, the NCT on GGM networks estimated from the continuous data indicated no significant differences in the overall network structure ( p = .827), the global strength (S = 0.07, p = .909), or node EI indices ( p s ≥ 0.134). The full results of specific edge differences were presented in Table S6. In addition, the harmonious group exhibited higher bridge EI for “ suicide ideation ” (PHQ9; diff = 0.355, p = .008) and “ uncontrollable worrying ” (GAD2; diff = 0.301, p = .090; marginally; see Table S6). Network accuracy and stability Bootstrapped analyses of edge weights (Figure S1) revealed that both the harmonious and disharmonious parental relationship networks displayed relatively narrow 95% confidence intervals, supporting satisfactory estimation accuracy. In addition, case-dropping bootstrap analyses (Figure S2) demonstrated that bridge EI remained stable across the two networks, with correlation stability coefficients (CS-Cs) of 0.25 and 0.35, respectively. These coefficients reflect the proportion of the sample that can be omitted while maintaining reliable bridge EI estimates. In contrast, the EI indices were less robust (CS-Cs < 0.25) in both groups, suggesting that these results should be interpreted cautiously. Finally, Figures S3 and S4 presented the results of the bootstrapped difference tests for edge weights and centrality measures (EI and bridge EI), respectively. In both networks, some edges were stronger, and some nodes were more central than others, suggesting heterogeneity in network connections and symptom importance. Directed Acyclic Graph (DAG) analysis Figure 2 illustrates the DAG networks for the harmonious and disharmonious parental relationship groups, while Table S7 presents the detailed edge strengths (measured by BIC) and directional probabilities. Overall, both DAG networks exhibit a hierarchical structure in which anxiety symptoms act as upstream drivers of depression; however, the specific entry points and downstream pathways differ between groups. Fig. 2. Open in a new tab DAG networks for the harmonious ( A ) and disharmonious ( B ) parental relationship groups. Note. Arrow thickness indicates the strength of the edge, measured by BIC. The detailed BIC values and directional probability of each arrow were shown in Table S7 (see Supplementary Materials) In the DAG network for the harmonious parental relationship group (Fig. 2 A), “ uncontrollable worrying ” (GAD2) was positioned at the top, initiating the activation of the entire network. GAD2 activated four anxiety symptoms, including “ trouble relaxing ” (GAD4), “ restlessness ” (GAD5), “ nervousness ” (GAD1), “ excessive worry ” (GAD3), and “ feeling afraid ” (GAD7). Moreover, GAD2 also triggered several depression symptoms such as “ depressed or sad mood ” (PHQ2), “ guilt ” (PHQ6), “ motor disturbances ” (PHQ8), and “ suicide ideation ” (PHQ9). PHQ9 functioned as a secondary downstream node additionally triggered by PHQ6 and PHQ8. “ Appetite changes ” (PHQ5) was located at the bottom of the network, serving as the ultimate downstream node. PHQ5 was primarily activated by “ fatigue ” (PHQ4), “ guilt ” (PHQ6), “ concentration difficulties ” (PHQ7), and GAD1. In the DAG network for the disharmonious parental relationship group (Fig. 2 B), “ trouble relaxing ” (GAD4) played as a triggering node positioned at the top. GAD4 triggered all other six anxiety symptoms, especially “ excessive worry ” (GAD3), “ uncontrollable worrying ” (GAD2), and “ irritability ” (GAD6), as well as a depression symptom, “ depressed or sad mood ” (PHQ2). PHQ9 also functioned as a secondary downstream node that was triggered by “ guilt ” (PHQ6), “ motor disturbances ” (PHQ8), and “ feeling afraid ” (GAD7). “ Sleep difficulties ” (PHQ3) and “ appetite changes ” (PHQ5), located at the bottom, serve as the two ultimate downstream nodes of the network, which were both activated by PHQ8, “ fatigue ” (PHQ4), and “ anhedonia ” (PHQ1). In addition, PHQ3 was also activated by PHQ2 and “ concentration difficulties ” (PHQ7). Simulation of alleviating and aggravating interventions The results of the simulation-based interventions (Fig. 3 ; see full results in Table S8) revealed distinct key targets for networks of harmonious and disharmonious parental relationship groups. In both groups, alleviating “ fatigue ” (PHQ4; Fig. 3 A and C) led to the greatest decreases in overall symptom levels, suggesting that PHQ4 may serve as a promising target for mitigating general distress. Conversely, aggravating “ depressed or sad mood ” (PHQ2; Fig. 3 B) in the harmonious network and “ motor disturbances ” (PHQ8; Fig. 3 D) in the disharmonious network resulted in the greatest increases in overall symptom levels. These findings suggest that PHQ2 and PHQ8 may serve as critical indicators of preventive symptoms within their respective groups. Fig. 3. Open in a new tab The results of the simulated intervention for the harmonious ( A & B ) and disharmonious ( C & D ) parental relationship groups. Note. A & C . Simulated alleviating interventions for groups. C & D . Simulated aggravating intervention for groups. Dots represent the network sum scores, while the lines represent the 95% confidence intervals. The symptoms are listed based on the size of the intervention’s effects, along with the original total score before the intervention. See the detailed NIRA of each symptom in Table S8 Discussion This study advances the existing literature by offering a nuanced, symptom-level examination of how college students’ perceptions of parental relationships are linked to depression and anxiety. Using network analysis, it maps the specific connections between the three variables. Also, the analysis identifies potential targets for intervention and prevention through simulations that alleviate or exacerbate symptoms. Several key findings emerge from this investigation. First, it is worth noting that although there was no significant difference between the two groups in terms of overall network structure, individuals who perceive their parental relationship as harmonious differ from those who perceive it as disharmonious in both the types and number of bridge symptoms. In the harmonious network, “ guilt ” (PHQ6, a symptom reflecting feelings of self-blame, worthlessness and inferiority), “ motor disturbances ” (PHQ8, a symptom reflecting slowness of movement or speech, or increased restlessness), “ uncontrollable worrying ” (GAD2, a symptom reflecting excessive worry about various matters.), and “ feeling afraid ” (GAD7, a symptom reflecting a fear that something terrible is about to happen) served as bridge symptoms connecting depression and anxiety. This finding aligns to some extent with previous research, with “ guilty ” being one of the critical bridging nodes of connection within the depression and anxiety symptom network [ 46 ]. It is also consistent with the study conducted by Dobson et al. [ 47 ], indicating that the physical symptoms of depression bridged the two communities. To be specific, these four bridge symptoms are predominantly internalizing mood symptoms (with the exception of “ motor disturbances ”), indicating that for individuals with positive parental relationships, the comorbidity of depression and anxiety is primarily directed inward. From an attachment theory perspective [ 48 ], individuals with secure attachments are susceptible to threats of relationship disruption. Based on this, students from harmonious families may be more likely to internalize a sense of responsibility to maintain family well-being and are reluctant to disappoint or worry their parents. As a result, their own negative emotions may be perceived as a threat to familial harmony, triggering intense guilt or worry (e.g., fear that their emotional state may disturb the family equilibrium). This, in turn, leads to a comorbid structure characterized by more internalizing symptoms as key connectors. In terms of the disharmonious network structure, however, only “ motor disturbances ” and “ irritability ” (GAD6, a symptom reflecting a tendency toward irritability or impatience) emerged as bridge symptoms. These two bridge symptoms are primarily externalizing and behavioral in nature. This finding can be readily understood through social learning theory [ 17 ], which posits that children exposed to hostile parental interactions learn maladaptive behaviors via observation. Consequently, their psychological distress is often expressed through externalizing negative reactions toward the environment, such as irritability and restlessness. Such a result, which shows that the overall network structure shows no difference but bridge symptoms differ significantly, may suggest that the fundamental pattern of connections among depression and anxiety is stable across different family backgrounds, whereas the pathways of symptom transmission vary according to perceived parental relationship quality. More specifically, these findings reveal that a harmonious parental relationship may facilitate the development of depression-anxiety comorbidity through a more internalizing and moderate pathway, whereas a discordant parental relationship may predispose individuals to develop comorbid disorders via a more externalizing and intense route. Moreover, analysis of the DAG results revealed significant structural differences in the anxiety-depression symptom networks between the harmonious and disharmonious parental relationship groups. In the harmonious group, “ uncontrollable worrying ” (GAD2) served as the triggering node, primarily driving downstream anxiety and depression symptoms, with “a ppetite changes ” (PHQ5) as the ultimate downstream node. In contrast, in the disharmonious group, “ trouble relaxing ” (GAD4) emerged as the initiating node, directly activating anxiety and depression symptoms, with “ sleep difficulties ” (PHQ3) and “ appetite changes ” (PHQ5) jointly serving as the endpoints of the network. In both groups, “suicidal ideation” (PHQ9) functioned as a secondary downstream node; however, in the disharmonious group, it was additionally triggered by “feeling afraid” (GAD7). These findings further suggest that the quality of parental relationships may alter the pathways of symptom propagation and the clustering patterns of somatic symptoms. As suggested by attachment theory [ 48 ], harmonious parental relationships create a family environment that fosters a sense of security, which may, to some extent, lead individuals to experience anxiety primarily at the cognitive level in the form of uncontrollable worrying. In contrast, in the context of disharmonious parental relationships, the family environment involves more conflict or unpredictability, which may more readily activate the individual’s physiological stress system. Consequently, for individuals in the disharmonious group, anxiety is more likely to be initially activated as difficulty relaxing, thereby triggering a broader range of anxiety and depression symptoms. Regarding the endpoints of the symptom chains, the harmonious group showed that the symptom chain terminated mainly in eating behavior, whereas the disharmonious group exhibited both eating and sleep disturbances. Clinically, this may indicate that when patients present with both sleep and appetite problems, it often reflects a higher stress burden and more severe functional impairment. Besides, the simulation intervention results indicated that for both groups, a reduction in “ fatigue ” (PHQ4, a symptom reflecting tiredness or listlessness) is likely to lead to effective alleviation of comorbid anxiety-depression symptoms in individuals. This alleviation effect suggests that, regardless of familial protective factors, “ fatigue ” could be a promising intervention target for depressive and anxiety symptoms. As Lopresti et al. [ 49 ] have found through a retrospective study, both physical exercise and sleep regulation yield significant positive effects in improving depressive symptoms. The discovery of group-specific preventive targets revealed that “ depressed or sad mood ” (PHQ2, a symptom reflecting feelings of frustration, depression, or despair) and “ motor disturbances ” may be the effective prevention targets for the harmonious and the disharmonious group, respectively. We postulate that individuals perceiving low parental conflict exhibit stronger emotional self-awareness, allowing for more proactive regulation of sadness and thus attenuating the risk of comorbid depression and anxiety. This view is supported by existing evidence suggesting that the overall symptom severity would likely be reduced if they target the depressive symptom of “ sad mood ” [ 3 ]. Conversely, those who perceive poor parental relationships are more likely to develop noticeable externalizing somatic behaviors, and thus, strategies targeting “ motor disturbances ” may serve as a more effective preventive focus for this group. Above all, this study preliminarily suggests that the quality of parental relationships may play a moderating role in the structure of anxiety-depression symptom networks, and offers some differentiated implications for clinical practice. In mental health services for aviation university students, the family environment may be preliminarily considered as a reference factor in differentiated intervention. “ Fatigue ”, as a potential intervention target common to both groups, aligns with the concept of fatigue management in aviation safety and may warrant priority attention. On this basis, for students from harmonious family backgrounds, greater attention might be paid to the impact of cognitive anxiety symptoms on flight performance; for students from disharmonious family backgrounds, more attention could be given to the potential effects of irritability and behavioral changes on teamwork and operational capacity. Nevertheless, these implications require further validation using a real-world sample [ 50 ], and practical application should be flexibly adjusted to account for individual differences and the specific training context of aviation colleges. Limitations This study also has several notable limitations. First, while the cross-sectional design allows for examining the impact of perceived parent-child relationships on the network structure of depression-anxiety comorbidity, it cannot track the persistence of this influence. Therefore, future research should consider adopting longitudinal designs to enhance the robustness of findings. Second, participants were classified into groups based on a single-item indicator of parental relationship quality. Although this approach allowed for network comparisons, it may limit the precision of the moderation assessment. Future studies could employ moderated network analysis to examine these effects using continuous measures. Third, the online implementation of this study, relying solely on self-reported data from a single province, suggests that participant bias and regional disparities may have compromised the accuracy of the findings. For instance, participants might have underreported the quality of their parental relationships or the severity of anxiety/depression symptoms to maintain a positive image, potentially weakening symptom correlations in network analysis and obscuring the true structure of the symptom network. Future research should consider employing more objective measurement methods and expanding sample sizes. This would facilitate the identification of more subtle and stable differences within network structures, thereby validating and extending the findings of this study. Fourth, a relatively large number of responses were excluded in our work due to failed data quality checks. This procedure may introduce potential selection bias and should therefore be considered when interpreting the findings. Future studies could further improve response quality through optimized survey design (e.g., ecological momentary assessment [ 51 ]), additional quality control procedures, and recruitment strategies that enhance participant engagement. Finally, this study examined only broadly perceived parental relationships without focusing on specific parental conflicts, while potential intergenerational dynamics also warrant further consideration [ 52 ]. This choice may limit deeper insights into the conflict-specific dynamics. Future work is recommended to select a particular conflict as a dedicated research focus. Conclusions To sum up, this cross-sectional study reveals different connectivity patterns between depression and anxiety symptoms among civil aviation college students reporting harmonious versus disharmonious parental relationships. Specifically, in the group with harmonious parental relationships, “ guilt ”, “ motor disturbances ”, “ uncontrollable worrying ”, and “ feeling afraid ” were identified as key bridge symptoms connecting these two disorders. Conversely, in the group with disharmonious parental relationships, “ motor disturbances ” and “ irritability ” served as bridge symptoms. Furthermore, “ fatigue ” may be a potential intervention target for both groups. However, preventive priorities may carefully distinguish “ depressed or sad mood ” for the good-relationship group and “ motor disturbances ” for the poor-relationship group. These findings offer preliminary insights for the treatment of anxiety and depression among university students from differing family environments. Supplementary Information Supplementary Material 1. (20.9MB, docx) Acknowledgements None. Authors’ contributions Contributed to conception and design: Hongzhi Hu, Yijing Song; Contributed to acquisition of data: ZhongfaOuyang, Lang Liang, Hongzhi Hu, Yijing Song. Contributed to analysis and interpretation of data: Yijing Song, Hongzhi Hu; Drafted the article: Hongzhi Hu, Yijing Song. Revised the article: YijingSong, Zhongfa Ouyang, and Lang Liang; Approved the submitted version for publication: ZhongfaOuyang, Lang Liang, Hongzhi Hu, Yijing Song. Funding The current study was supported by the Project of the Key Research Base of Philosophy and Social Sciences in Chengdu (Chengdu Aviation Industry Development and Cultural Construction Research Center) (Project No. CAIACDRCXM2026-03). Data availability The dataset generated and analyzed in the current study is not publicly available due to our Departmental policy, but is available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate The research was examined and approved by the ethics committee of Chengdu Aeronautic Polytechnic University. Informed consent was obtained from all individual participants included in the study. The study was conducted in accordance with the principles of the Declaration of Helsinki. 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. Cummings CM, Caporino NE, Kendall PC. Comorbidity of anxiety and depression in children and adolescents: 20 years after. Psychol Bull. 2014;140(3):816–45. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Association AP. Diagnostic and statistical manual of mental disorders: DSM-5™. 5th ed. Arlington, VA, US: American Psychiatric Publishing, Inc.; 2013. [ Google Scholar ] 3. Beard C, Millner AJ, Forgeard MJ, Fried EI, Hsu KJ, Treadway MT, Leonard CV, Kertz SJ, Björgvinsson T. Network analysis of depression and anxiety symptom relationships in a psychiatric sample. Psychol Med. 2016;46(16):3359–69. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Gaspersz R, Nawijn L, Lamers F, Penninx B. Patients with anxious depression: overview of prevalence, pathophysiology and impact on course and treatment outcome. Curr Opin Psychiatry. 2018;31(1):17–25. [ DOI ] [ PubMed ] [ Google Scholar ] 5. Kircanski K, LeMoult J, Ordaz S, Gotlib IH. Investigating the nature of co-occurring depression and anxiety: Comparing diagnostic and dimensional research approaches. J Affect Disord. 2017;216:123–35. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Li P, Huang N, Yang X, Fang Y, Chen Z. A simulation-based network analysis of intervention targets for adolescent depressive and anxiety symptoms. Asian J Psychiatr. 2024;99:104152. [ DOI ] [ PubMed ] [ Google Scholar ] 7. Chen W, Zhao B, Tu H, Liao M, Zhang X, Su Y. Perceived social support, insomnia, and psychological resilience: a chain mediation model of internet addiction in civil aviation cadets. Front Psychiatry. 2025;16:1698824. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Gao Cuicui L, Jiawei, Chunxiao L. Research in the mental health status of students of civil aviation characteristic majors. J Xinjiang Vocat Univ. 2009;17(4):70–3. [ Google Scholar ] 9. Baethge A, Junker NM, Garthus-Niegel S. Too much to handle? Trajectories of work-home conflict as the family grows and its impact on parents’ mental health. J Occup Health Psychol. 2025;30(1):16–33. [ DOI ] [ PubMed ] [ Google Scholar ] 10. Niu H, Wang S, Tao Y, Tang Q, Zhang L, Liu X. The association between online learning, parents’ marital status, and internet addiction among adolescents during the COVID-19 pandemic period: A cross-lagged panel network approach. J Affect Disord. 2023;333:553–61. [ DOI ] [ PubMed ] [ Google Scholar ] 11. Tao Y, Hou W, Niu H, Ma Z, Zhang S, Zhang L, Liu X. Centrality and bridge symptoms of anxiety, depression, and sleep disturbance among college students during the COVID-19 pandemic—a network analysis. Curr Psychol. 2024;43(15):13897–908. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Yap MB, Pilkington PD, Ryan SM, Jorm AF. Parental factors associated with depression and anxiety in young people: a systematic review and meta-analysis. J Affect Disord. 2014;156:8–23. [ DOI ] [ PubMed ] [ Google Scholar ] 13. Yap MB, Jorm AF. Parental factors associated with childhood anxiety, depression, and internalizing problems: a systematic review and meta-analysis. J Affect Disord. 2015;175:424–40. [ DOI ] [ PubMed ] [ Google Scholar ] 14. van Eldik WM, de Haan AD, Parry LQ, Davies PT, Luijk M, Arends LR, Prinzie P. The interparental relationship: Meta-analytic associations with children’s maladjustment and responses to interparental conflict. Psychol Bull. 2020;146(7):553–94. [ DOI ] [ PubMed ] [ Google Scholar ] 15. Bowen M. The use of family theory in clinical practice. Compr Psychiatry. 1966;7(5):345–74. [ DOI ] [ PubMed ] [ Google Scholar ] 16. Jiang H, Yan Y, Zhang Y. The Relationship between Interparental Conflict and Depressive symptoms Spanning from Childhood to Youth: A Systematic Review and Meta-Analysis. J Youth Adolesc. 2025;54(10):2419–40. [ DOI ] [ PubMed ] [ Google Scholar ] 17. Bandura A, Ross D, Ross SA. Transmission of aggression through imitation of aggressive models. J Abnorm Soc Psychol. 1961;63:575–82. [ DOI ] [ PubMed ] [ Google Scholar ] 18. Azevedo MS, Meira L, Câmara DP, Ferreira TB, Martins EC. Parental Emotion Socialization and Difficulties in Emotion Regulation in Adolescents: A Network Analysis. J Youth Adolesc. 2025;54(11):2778–93. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. Borsboom D, Cramer AO. Network analysis: an integrative approach to the structure of psychopathology. Annu Rev Clin Psychol. 2013;9:91–121. [ DOI ] [ PubMed ] [ Google Scholar ] 20. Borsboom D. A network theory of mental disorders. World Psychiatry. 2017;16(1):5–13. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Cramer AO, Waldorp LJ, van der Maas HL, Borsboom D. Comorbidity: a network perspective. Behav Brain Sci. 2010;33(2–3):137–50. discussion 150–193. [ DOI ] [ PubMed ] [ Google Scholar ] 22. Cheung T, Jin Y, Lam S, Su Z, Hall BJ, Xiang YT. Network analysis of depressive symptoms in Hong Kong residents during the COVID-19 pandemic. Transl Psychiatry. 2021;11(1):460. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Lunansky G, Naberman J, van Borkulo CD, Chen C, Wang L, Borsboom D. Intervening on psychopathology networks: Evaluating intervention targets through simulations. Methods. 2022;204:29–37. [ DOI ] [ PubMed ] [ Google Scholar ] 24. Blakeley K, Jansen DJ. Post-Traumatic Stress Disorder and Other Mental Health Problems in the Military: Oversight Issues for Congress . In: 2013 (No. CRSR43175). Washington D.C. 2013. 25. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606–13. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Spitzer RL, Kroenke K, Williams JB, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006;166(10):1092–7. [ DOI ] [ PubMed ] [ Google Scholar ] 27. Huang JL, Curran PG, Keeney J, Poposki EM, Deshon RP. Detecting and Deterring Insufficient Effort Responding to Surveys. J Bus Psychol. 2012;27(1):99–114. [ Google Scholar ] 28. R. A language and environment for statistical computing. R foundation for statistical computing. [ https://www.R-project.org/ ]. 29. Epskamp S, Borsboom D, Fried EI. Estimating psychological networks and their accuracy: A tutorial paper. Behav Res Methods. 2018;50(1):195–212. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Epskamp S, Cramer AOJ, Waldorp LJ, Schmittmann VD, Borsboom D. qgraph: Network Visualizations of Relationships in Psychometric Data. J Stat Softw. 2012;48(4):1–18. [ Google Scholar ] 31. van Borkulo CD, Borsboom D, Epskamp S, Blanken TF, Boschloo L, Schoevers RA, Waldorp LJ. A new method for constructing networks from binary data. Sci Rep. 2014;4(1):5918. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Opsahl T, Agneessens F, Skvoretz J. Node centrality in weighted networks: Generalizing degree and shortest paths. Social Networks. 2010;32(3):245–51. [ Google Scholar ] 33. Jones PJ, Ma R, McNally RJ. Bridge Centrality: A Network Approach to Understanding Comorbidity. Multivar Behav Res. 2021;56(2):353–67. [ DOI ] [ PubMed ] [ Google Scholar ] 34. Tang Q, Zou X, Gui J, Wang S, Liu X, Liu G, Tao Y. Effects of childhood trauma on the symptom-level relation between depression, anxiety, stress, and problematic smartphone use: A network analysis. J Affect Disord. 2024;358:1–11. [ DOI ] [ PubMed ] [ Google Scholar ] 35. Scheiner SM, Gurevitch J. Mantel tests: spatial structure in field experiments. Design and analysis of ecological experiments (pp. 342-359). edn.: Chapman and Hall/CRC. 2020:342–59. 36. Epskamp S, Waldorp LJ, Mõttus R, Borsboom D. The Gaussian Graphical Model in Cross-Sectional and Time-Series Data. Multivar Behav Res. 2018;53(4):453–80. [ DOI ] [ PubMed ] [ Google Scholar ] 37. Zhang C, Li R, Zhang W, Tao Y, Liu X, Lv Y. A simulation-based network analysis of intervention targets for comorbid symptoms of depression and anxiety in Chinese healthcare workers in the post-dynamic zero-COVID policy era. BMC Psychiatry. 2025;25(1):457. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Niu H, Lu X, Lv Y, Gui J, Wang S, Tao Y, Zhou J. Simulating Interventions for Symptoms Linking Problematic Social Networking Sites Use to Online Aggressive Behavior Among Chinese College Students: A Gender-Differentiated Network Analysis. Psych J. 2026;15(2):e70085. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. Xu H, Wu Y, Cong X, Ma Z, Wang H, Bian R, Zhang T, Ge W, Li Y, Liu X et al. Identifying and simulating interventions on central depressive symptoms: a network analysis combining clinical and sub-clinical samples. J Affect Disord. 2026;394(Pt B):120602. [ DOI ] [ PubMed ] 40. Bringmann LF, Vissers N, Wichers M, Geschwind N, Kuppens P, Peeters F, Borsboom D, Tuerlinckx F. A Network Approach to Psychopathology: New Insights into Clinical Longitudinal Data. PLoS ONE. 2013;8(4):e60188. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. van Borkulo CD, van Bork R, Boschloo L, Kossakowski JJ, Tio P, Schoevers RA, Borsboom D, Waldorp LJ. Comparing network structures on three aspects: A permutation test. Psychol Methods. 2023;28(6):1273–85. [ DOI ] [ PubMed ] [ Google Scholar ] 42. Scutari M. Learning Bayesian Networks with the bnlearn R Package. J Stat Softw. 2010;35(3):1–22.21603108 [ Google Scholar ] 43. Scutari M. Bayesian Network Constraint-Based Structure Learning Algorithms: Parallel and Optimized Implementations in the bnlearn R Package. J Stat Softw. 2017;77(2):1–20. [ Google Scholar ] 44. McNally RJ. Can network analysis transform psychopathology? Behav Res Ther. 2016;86:95–104. [ DOI ] [ PubMed ] [ Google Scholar ] 45. Nagarajan R, Scutari M, Lèbre S. Bayesian Networks in R with Applications in Systems Biology. Volume 122. New York: Springer; 2013. [ Google Scholar ] 46. Zou H, Gao J, Wu W, Huo L, Zhang W. Which comes first? Comorbidity of depression and anxiety symptoms: A cross-lagged network analysis. Soc Sci Med. 2024;360:117339. [ DOI ] [ PubMed ] [ Google Scholar ] 47. Dobson ET, Croarkin PE, Schroeder HK, Varney ST, Mossman SA, Cecil K, Strawn JR. Bridging Anxiety and Depression: A Network Approach in Anxious Adolescents. J Affect Disord. 2021;280(Pt A):305–14. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 48. Bowlby J. Attachment and loss: retrospect and prospect. Am J Orthopsychiatry. 1982;52(4):664–78. [ DOI ] [ PubMed ] [ Google Scholar ] 49. Lopresti AL, Hood SD, Drummond PD. A review of lifestyle factors that contribute to important pathways associated with major depression: diet, sleep and exercise. J Affect Disord. 2013;148(1):12–27. [ DOI ] [ PubMed ] [ Google Scholar ] 50. Bernstein EE, Phillips KA, Greenberg JL, Curtiss J, Hoeppner SS, Wilhelm S. Mechanisms of cognitive-behavioral therapy effects on symptoms of body dysmorphic disorder: a network intervention analysis. Psychol Med. 2023;53(6):2531–9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 51. Curtiss JE, Pinaire M, Fulford D, McNally RJ, Hofmann SG. Temporal and contemporaneous network structures of affect and physical activity in emotional disorders. J Affect Disord. 2022;315:139–47. [ DOI ] [ PubMed ] [ Google Scholar ] 52. Tao Y, Guo S, Lu X, Niu H, Liu X. Intergenerational symptom level transmission of comorbid anxiety and major depressive disorder: A dyadic network approach with simulated intervention modeling. J Anxiety Disord. 2026;120:103159. [ DOI ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1. (20.9MB, docx) Data Availability Statement The dataset generated and analyzed in the current study is not publicly available due to our Departmental policy, but is available from the corresponding author upon reasonable request. Articles from BMC Psychology are provided here courtesy of BMC ACTIONS View on publisher site PDF (1.9 MB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Biased and inflexible interpretations of social situations predict depressive symptoms and relational outcomes in parent-adolescent dyads.#301580
• Mental Health in Relation to Parental Absence: A Systematic Review.#269404
• Mothers With a History of Depression and Their 10-14 Year Old Daughters#849606
• Parent-Adolescent Relationships, Emotional Regulation, and HRV in Adolescents With Depression#890489
• Mediational Role of Executive Functions on Parental Style and Anxiety and Depression Symptoms in Adolescents From Different Vulnerability Environments: A Cross-Sectional Study.#476766