Mapping the Presence of Anxiety Symptoms in Adults With Major Depressive Disorder

Abstract

Background: Patients with major depressive disorder (MDD) often present with co-occurring anxiety symptoms. The network method provides a novel view on understanding the co-occurrence of depressive and anxiety symptoms. Thus, the purpose of our study was to explore it by applying network analysis methods.

Methods: We used electronic medical records from West China Hospital in China. In total, 3,424 patients who met the criteria for MDD were included. R-studio 3.6 was used to estimate the network structure. First, we estimated the network structure of depression and anxiety symptoms using the graphic LASSO algorithm. Then, we estimated the centrality indices of nodes to determine which symptoms are more central in the network. We then estimated the bridge centrality indices using the bridge function via the R package networktools.

Results: Some strong connections were found like “easy to wake up,” “wake up early,” and “difficulty falling asleep,” “suicidal thoughts,” and “hopelessness.” “Depressed mood,” “somatic anxiety,” “hopelessness,” “anxiety mood,” and “tension” have the higher centrality indices. Results revealed eight bridge symptoms (e.g., concentration/memory difficulty, gastrointestinal symptoms) in the co-occurrence network structure.

Conclusions: This research suggests that the described approach in mapping the presence of anxiety symptoms in individuals with major depression might potentially increase diagnostic precision and help choose more targeted interventions and potentially reduce the occurrence of treatment resistance.

Introduction

Major depressive disorder (MDD) is a debilitating psychiatric condition characterized by depressed mood, decreased energy level, and lack of interest in pleasurable activities (1). It has a significant burden on both individuals and society and is related to psychological impairments and health dysfunction. Among Chinese adults, MDD has a 12-month and lifetime prevalence rate of 3.6 and 6.9%, respectively (2). Gaspersz's study showed that 40–60% of patients with MDD also have anxiety symptoms (3). Another survey suggested that 50% of individuals with MDD meet the diagnostic criteria for anxiety disorders (4, 5).

The co-occurrence anxiety symptoms among patients with MDD have important clinical implications. Firstly, the co-occurrence anxiety symptoms predict a more chronic course and more severe disease progression (6). Secondly, compared with individuals who have MDD without anxiety symptoms, patients influenced by anxiety and depressive symptoms have shown greater functional disability, poorer quality of life, and higher risk of suicide behavior (7, 8). Thirdly, MDD with prominent coexisting anxiety symptoms is more difficult to treat than MDD without anxiety symptoms (9). Fourth, anxiety symptoms are associated with occupation of more health care resources among patients with MDD (10). Previous studies either concentrated on the count symptoms to clarify a diagnosis and calculate the prevalence rate of comorbidity or regarded the symptoms as indicators of latent dimensions, however, the between-symptom links are considered a byproduct of dimensional community (11).

Network theory provides us a novel perspective of mental disorders. Network analysis conceptualizes symptoms as constituents of mental disorders, as compared to traditional methods that assume an underlying disease in advance as the common cause of symptoms. Recently, network analysis models are rapidly growing, not only concerning methodological issues but also in offering an appealing interpretation of psychopathology (12, 13).

From the perceptive of topology, network structure consists of nodes(symptoms) and edges(association between symptoms) (14). The importance of nodes was evaluated via centrality indices. Edges represent the links between pairs of symptoms and thicker edges denote larger correlations (15). Specifically, centrality indices include strength, closeness, and betweenness that allow clinicians to discern the symptoms with the greatest importance in the network structure. Those with high centrality indices convey more clinical information (16, 17). Strength is one of the most commonly used centrality indices as it is easy to interpret and is the most stable centrality index (18). In the opinion of network theory, comorbidity is regarded as a constellation of symptom-level relationships (19). Symptoms that link two mental disorders are regarded as “bridge symptom.” Bridge symptoms indicated by bridge centrality indices mainly included bridge strength, bridge closeness, and bridge betweenness (20). Bridge symptoms may have an important role in the development and maintenance of co-occurring mental disorder (21). Thus, when one mental disorder presents, intervention on potential bridge symptoms may contribute to preventing co-occurrence (20). For example, if we suppose that sleep disturbance is a bridge symptom between depression and anxiety, then patients who suffer poor sleep quality as one of their MDD symptoms would be at greater risk for anxiety compared to those without sleep disturbance. Thus, it would be wise for psychiatrists to treat these bridge symptoms therapeutically to reduce the co-occurrence.

At present, 10 studies utilized network analysis to explore comorbidity and co-occurrence in depression (22–31). Previous studies have explored the comorbidity between anxiety and depression (23), posttraumatic stress disorder and co-morbidity depressive symptoms (31), and comorbidity between obsessive-compulsive disorder and depression (26). While previous network researches in MDD have broadened our knowledge of the field, several limitations remain. Firstly, most studies use self-reported questionnaires (e.g., Patient Health Questionnaire-9 and General Anxiety Disorder-7) to assess symptoms among the general population (29, 30), while self-report questionnaires have a risk for response biases (32). Few studies used a sample that comprised treatment-seeking patients with MDD (23). Secondly, while sample size plays an important role in establishing a reliable network structure, most studies used a small sample size that ranges from 296 (25) to 1,029 (23, 33). Thirdly, no studies to date have explored the co-occurrence between MDD and anxiety symptoms based on real-world evidence. Real-word evidence is important for establishing the network structure of co-occurrence between MDD and anxiety symptoms. In this regard, electronic medical records (EMRs) bring new chances in clinical research, providing the potential chance for low-cost and high-volume data on clinical research (34).

The main aim of this study, which was based on the retrospective of EMRs, was to establish the co-occurring network structure between MDD and anxiety symptoms to identify the bridge symptoms.

Methods

Study Population

This was a retrospective study based on pre-existing data derived from EMRs undertaken at West China Hospital. Eligible patients were those with a diagnosis of MDD, aged between 18 and 65. In this study, we identified MDD patients through recorded primary diagnosis at discharge and based on the International Classification of Disease, Tenth Revision (Clinical Modification Codes F32, single episode major depressive disorder and F33, recurrent major depressive disorder), which has been described in another study (35). We extracted anonymous clinical-related information. Exclusion criteria are the following 3 items. Patients' follow-up discharge diagnosis code is F30 or F31. Patients with MDD did not accomplish Hamilton Depression Scale-24 (HAMD-24) and Hamilton Anxiety Scale-14 (HAMA-14) at admission. ③ Age ≤ 18 or age > 65. The details are shown in Figure 1. The data used to support the findings of this manuscript are restricted by the West China Hospital in order to protect patient privacy and avoid legal and ethical risks. Data are available from the West China Hospital for researchers who meet the criteria for access to confidential data (data.cd120.com).

Figure 1

We obtained approval from the Ethics Committee of the West China Hospital, Sichuan University (2017 N0.185). As this is a retrospective study, Institute Review Board (IRB) waived the requirement for obtaining informed consent from the individual patients. We did not use information on the identity of patients and all related information was kept confidential. All procedures were in accordance with the ethical standards of the Ethics Committee and the revised Helsinki Declaration of 2008 (36, 37).

Measures

HAMD-24 is one of the most widely used scales in clinical practical assessment of depression. It is assessed by psychiatrists and takes about 15 to 20 min to finish (38). HAMA-14 is one of the most widely used scales in the clinical practical assessment of anxiety. HAMA-14 includes 14 items with each item divided into 5 levels from 0 to 4 (39). The Chinese versions of HAMD-24 and HAMA-14 have good reliability and validity (40).

Data Analysis

All analysis was accomplished using the R-3.6 studio. In this research, the missing data belong to the category of missing at random (MAR). So, we used the method of unconditional mean imputation to handle the missing data (41). The network structure consists of two elements: nodes and edges. Every node represents a symptom, and each edge demonstrates a relationship between the two symptoms. In our study, the nodes (symptoms) were represented the scale items of HAMD-24 and HAMA-14. We used a graphical Gaussian model to estimate the networks. In addition, we used the least absolute shrinkage and selection operator (LASSO) to regularize our model and used the qgraph package to visualize the network. Next, we computed the centrality indices (i.e., betweenness, strength, and closeness) of nodes to find which symptoms are more central in the network structure. Betweenness and closeness are often not reliably estimated (17). Thus, we only reported the strengths in this article, while other node centrality indices are provided in the Supplementary Material. We assessed the accuracy stability of the centrality using the bootstrap approach in the bootnet package. To gain a stable and interpretable centrality, the CS coefficient should be >0.25 (42). We estimated the stability of edge-weights by bootstrapping the 95% confidence intervals (CIs), where fewer overlaps in the CIs indicate higher stability. Jones uses the term community to demonstrate a theoretically based group of nodes that correspond to a mental disorder according to clinical criteria, instead of according to any network analytic procedure (20). This method contributes to identifying bridge nodes, especially when networks are large, complex, or difficult to account for visually. The networktools package (43) were used to calculate the values and create plots. You can get the code from the Git hub (https://github.com/fenfenge/Network-structure).

Results

Descriptive Statistics

We included 3,424 patients with MDD comprising 2,349 females and 1,075 males, with ages ranging from 18 to 65 (M = 42.5, SD = 13.25). Table 1 shows the items, item content, sample means, standard deviation, and missing items/percentage.

Network Estimation

The graphical LASSO network is shown in Figure 2. A thicker edge indicates a stronger association between the symptoms. Green edges represent positive regularized partial correlations, and red edges represent negative regularized partial correlations. Network analysis demonstrated that five strong connections edges were among the HAMD-24 items. The top edge was between the items “easy to wake up,” “wake up early,” and “difficulty falling asleep” (D4: D5: D6). Additionally, HAMD-24 items “suicidal thoughts” and “hopelessness” (D3:D23) and “hopelessness” and “inferiority” (D23:D24) were among the strongest. There were two strongest edges between the HAMA-14 items. The strongest edge was between items “tension” and “afraid” (A2:A3), followed by “cardiovascular symptoms” and “respiratory symptoms” (A9:A10).

Figure 2

Network Inference and Stability

Strength centrality is shown in Figure 3. Firstly, nodes demonstrating “depressed or sad mood” (D1), “somatic anxiety” (D11), and “hopelessness” (D23) were among the MDD symptoms exhibiting higher levels of strength. Secondly, “anxiety mood” (A1) and “tension” (A2) were among the anxiety symptoms exhibiting higher levels of strength. In terms of stability of network analysis, bootstrap 95% CI demonstrated moderate stability for the strength index (Supplementary Materials).

Figure 3

Bridge strength is shown in Figure 4. Nodes demonstrating “anxiety mood” (A1), “insomnia” (A4), “concentration/memory difficulty” (A5), “pessimism” (A6), and “gastrointestinal symptoms” (A11) were anxiety symptoms displaying higher levels of bridge strength on all MDD symptoms. The MDD symptoms “nervousness or anxiety” (D10), “somatic anxiety” (D11), and “gastrointestinal symptoms” (D12) exert a strong bridging effect on anxiety symptoms.

Figure 4

Discussion

Many patients with MDD also experience anxiety symptoms. As far as we know, this is the first research to explore network connectivity among treat-seeking patients with MDD and co-occurrence anxiety symptoms based on real-world evidence. Perhaps the most prominent result was that not all nodes(symptoms) were equally important in the network with co-occurrence anxiety symptoms among patients with MDD (44, 45). Besides, we also found some strong linkage between symptoms, such as “easy to wake up” (D4), “wake up early” (D5), and “difficulty falling asleep” (D6), which were closely associated with each other.

Moreover, “depressed or sad mood” (D1) and “hopelessness” (D23) have the highest centrality in the network structure. Such results do not come as a surprise; previous studies that used self-report questionnaires and composite international diagnostic interviews also suggested that sad mood and hopelessness carry more weight than other symptoms of MDD (14, 46–48). “Somatic anxiety” (D11) had one of the highest centrality indices, making it a hallmark symptom of MDD among Chinese. “Somatic anxiety” is involved via overt signs of excessive autonomic activity and/or skeletal/muscle-motor tension. Due to the following three reasons that including (1) emotional symptoms are stigmatized, (2) thinking deviated far from internal experiences, (3)a special perspective of the self is emphasized, (49) somatization has been a general symptom in the Chinese population (50, 51). It is worthy that higher somatization was related to poorer performance in the cognitive and poorer antidepressant treatment clinical outcome (52). Anxiety symptoms (e.g., anxiety mood and tension) also had higher centrality. Goghari's 2-year longitudinal study found that patients with MDD reported higher levels of anxiety than those with other mental disorders. To some extent, easing the patient's anxiety level helps improve the outcome and daily function of depression (53).

Network theory helps to understand the co-occurrence of psychopathology. We could reduce co-occurrence by effectively “burning the bridge symptoms” between disorders (45). In this study, the MDD nodes that displayed the strongest association with anxiety symptoms were “nervousness” (D10), “somatic anxiety” (D11), and “gastrointestinal symptoms” (D12). Conversely, the anxiety nodes that displayed the strongest connection with MDD symptoms were “anxiety mood” (A1), “insomnia” (A4), “concentration/memory difficulty” (A5), “pessimism” (A6), and “gastrointestinal symptoms” (A11). Gastrointestinal symptoms were bridge symptoms that link MDD to anxiety symptoms and similarly linked anxiety symptoms to MDD. Gastrointestinal symptoms are common features for both MDD and anxiety disorders. They are also prominent features in posttraumatic stress disorder, schizophrenia, and autism spectrum disorder (54–56), demonstrating that gastrointestinal symptoms may be a common feature associated with a multitude of mental disorders. Simpson's review indicated that the role of gut microbiota in mood regulation and emotional processing may be of particular relevance to depression and anxiety etiology (57). In this study, concentration/memory difficulty was a bridge symptom that linked anxiety to MDD. Pettit's research found that computer-based attention training can reduce anxiety symptoms among the youth (58). These results suggested that attention training could be used to ease anxiety symptoms among MDD patients.

We found a strong connection between “easy to wake up,” “wake up early,” and “hard to fall asleep” (D4: D5: D6). Interestingly, the three symptoms above belong to the diagnostic criteria for insomnia (59) and MDD patients with sleep complaints are prone to more severe symptoms (60). Thus, insomnia is a valuable therapeutic target in MDD patients. The theoretical promise of network analysis is that psychiatric symptoms are not isolated and may accentuate each other (61). Symptoms may reinforce one another via creating a feedback loop, such as, “inferior helplessness” (D24) could lead to “hopelessness” (D23), and at the same time prompt “suicidal thoughts” (D3). These links may contribute to the build of a self-sustained symptom constellation. Thus, it is necessary to find potential self-sustaining mechanisms and timely interventions for feedback loops (13). Somatic symptoms (e.g., cardiovascular symptoms and respiratory symptoms) were closely connected. This can be explained by the mind-body interaction model, which states that information flows not only from the body to the brain but also from the brain to the body. In the brain, “prediction” is derived from metacognition (conceptual knowledge), namely higher-order thoughts and cognition, which helps estimate the generation of behavioral commands (62).

Clinical Implication

The network model provides a novel view to investigate the potential mechanisms underlying the etiology and maintenance of mental disorders. We can understand how the symptoms are interrelated via networks and how to intervene on comorbidities. Firstly, when depressive patients are admitted to the hospital, psychiatrists should prioritize the evaluation of identified bridging MDD symptoms to screen patients with a higher risk of suffering anxiety symptoms. Secondly, from the viewpoint of network structure, we can intervene mental disorders from the following aspects: (1) symptoms (nodes) interventions: direct intervention of one or more symptoms; (2) network interventions: intervention symptom-symptom connections (61); (3) bridge symptoms: intervention of bridge symptoms to prevent co-occurrence. For example, we can intervene on closely connected sleep symptoms to avoid using certain antidepressants that may cause or even worsen sleep problems (63), since improving sleep contributes to improving the outcomes of MDD (64, 65). Our findings may be used to demonstrate which symptoms (e.g., gastrointestinal symptoms, insomnia) drive the association, and which should be handled first to reduce MDD and anxiety co-occurrence. These findings emphasize the importance of assessing anxiety symptoms among patients with MDD. Owing to the limitations of earlier versions of the Diagnostic and Statistical Manual of Mental Disorder (DSM) on grading targeting rules, anxiety in mental disorders has been underestimated, underdiagnosed, and undertreated. CBT is very effective for the treatment of anxiety symptoms and sleep disturbance (66). Thus, patients with MDD can be treated with suitable medicines and CBT.

Strengths and Limitations

This research has several key strengths. First, we employed a sample comprised of treatment-seeking patients diagnosed with MDD according to the ICD-10. Moreover, HAMD and HAMA are evaluated by psychiatrists; thus, ambiguous/vague answers to the self-assessment scale were avoided. Second, we used a relatively large sample that contributed to establishing a reliable network structure with robust edge weights and centrality (42). To date, only one study has used large psychiatric samples of over 1,000 participants (23).

Despite the strengths, some limitations should be considered in this study. Firstly, this is a cross-sectional study, and we cannot explore changes in the co-occurrence network structure over time. Specifically, we are not sure whether the connections between symptoms appear temporarily or continuously, and what will happen to those strong connections under the intervention of strategy. Considering that network modes require the estimation of many parameters and these models need power to reliably detect small coefficients, exploration of larger data sets is necessary (67, 68). Thus, in this study, we do not compare the network structure in different severity of subgroup. Secondly, in our study, the target population was patients with MDD diagnosis so that the results could be generalized to the whole population. Specifically, the results do not suit individuals who suffer from certain depression symptoms, while failing to meet the diagnostic criteria for MDD. Third, Gureje's study showed that the presentation of mental disorders is influenced by culture and social milieu (69), while we only focused on Chinese patients. Future studies should consider the cross-cultural variation and explore how culture influences the presentation of mental disorders. Finally, we included patients with ICD-10 F32 and F33 diagnoses only. We did not check whether the patients had a comorbidity diagnosis with an anxiety disorder or any other co-morbidity. However, Wise's study (70) may support the notion that comorbidities might not be so important, at least concerning functional connectivity. Specifically, Wise's research supports the notion that biological abnormalities in functional connectivity in major depression across independent samples might overlap irrespective of the presence of anxiety comorbidities.

Conclusion

This research is the first to explore the association between depressive and anxiety symptoms among MDD patients based on EMRs, thus offering an essential basis on how the two disorders co-vary. We found that some high central symptoms (e.g., hopelessness, somatic anxiety, and tension) and some bridge symptoms (e.g., concentration/memory difficulty, pessimism, and gastrointestinal symptoms). We summarize the evidence from the current research that treatment for co-occurring anxiety symptoms among MDD patients at symptom level may be efficacious. Multiple interventions, such as improving sleep, CBT, or attention training, could be applied to address these co-occurring symptoms.

References

• 1.

  Organization WH. ICD-10 Classification of Mental and Behavioural Disorders (The). (1992).

  • Google Scholar

• 2.

  HuangYWangYWangHLiuZYuXYanJet al. Prevalence of mental disorders in China: a cross-sectional epidemiological study. Lancet Psychiatry. (2019) 6:211–24. 10.1016/s2215-0366(18)30511-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 3.

  GasperszRLamersFKentJMBeekmanATSmitJHvan HemertAMet al. Longitudinal predictive validity of the DSM-5 anxious distress specifier for clinical outcomes in a large cohort of patients with major depressive disorder. J Clin Psychiatry. (2017) 78:207–13. 10.4088/JCP.15m10221

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4.

  NuttD. Management of patients with depression associated with anxiety symptoms. J Clin Psychiatry. (1997) 58(Suppl. 8):11–6.

  • Pubmed Abstract
  • Google Scholar

• 5.

  KesslerRCBerglundPDemlerOJinRKoretzDMerikangasKRet al. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). JAMA. (2003) 289:3095–105. 10.1001/jama.289.23.3095

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 6.

  GoldbergDFawcettJ. The importance of anxiety in both major depression and bipolar disorder. Depress Anxiety. (2012) 29:471–8. 10.1002/da.21939

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 7.

  PfeifferPNGanoczyDIlgenMZivinKValensteinM. Comorbid anxiety as a suicide risk factor among depressed veterans. Depress Anxiety. (2009) 26:752–7. 10.1002/da.20583

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 8.

  ZimmermanMChelminskiIYoungDDalrympleKWalshERosensteinL. A clinically useful self-report measure of the DSM-5 anxious distress specifier for major depressive disorder. J Clin Psychiatry. (2014) 75:601–7. 10.4088/JCP.13m08961

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 9.

  GasperszRLamersFKentJMBeekmanATFSmitJHvan HemertAMet al. Anxious distress predicts subsequent treatment outcome and side effects in depressed patients starting antidepressant treatment. J Psychiatr Res. (2017) 84:41–8. 10.1016/j.jpsychires.2016.09.018

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10.

  McLaughlinTPKhandkerRKKruzikasDTTummalaR. Overlap of anxiety and depression in a managed care population: prevalence and association with resource utilization. J Clin Psychiatry. (2006) 67:1187–93. 10.4088/jcp.v67n0803

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 11.

  BorsboomD. Psychometric perspectives on diagnostic systems. J Clin Psychiatry. (2008) 64:1089–108. 10.1002/jclp.20503

  • CrossRef
  • Google Scholar

• 12.

  CramerAOWaldorpLJvan der MaasHLBorsboomD. Comorbidity: a network perspective. Behav Brain Sci. (2010) 33:137–50. 10.1017/s0140525x09991567

  • CrossRef
  • Google Scholar

• 13.

  FriedEIvan BorkuloCDCramerAOBoschlooLSchoeversRABorsboomD. 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

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 14.

  BoschlooLvan BorkuloCDBorsboomDSchoeversRA. A prospective study on how symptoms in a network predict the onset of depression. Psychother Psychosom. (2016) 85:183–4. 10.1159/000442001

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 15.

  McNallyRJRobinaughDJWuGWYWangLDesernoMKBorsboomD. Mental disorders as causal systems: a network approach to posttraumatic stress disorder. Clin Psychol Sci. (2014) 3:836–49. 10.1177/2167702614553230

  • CrossRef
  • Google Scholar

• 16.

  FavaGAFauRCTombaETombaE. The clinical process in psychiatry: a clinimetric approach. J Clin Psychiatry. (2012) 73:177–84. 10.4088/JCP.10r06444

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 17.

  BringmannLA-OElmerTEpskampSKrauseRWSchochDA-OWichersMet al. What do centrality measures measure in psychological networks?J Abnorm Psychol. (2019) 128:892–903. 10.1037/abn0000446

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 18.

  PhillipsRDWilsonSMSunDMoreyR. Posttraumatic stress disorder symptom network analysis in U.S. military veterans: examining the impact of combat exposure. Front Psychiatry. (2018) 9:608. 10.3389/fpsyt.2018.00608

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 19.

  BorsboomDEpskampSKievitRACramerAOSchmittmannVD. Transdiagnostic networks: commentary on Nolen-Hoeksema and Watkins (2011). Perspect Psychol Sci. (2011) 6:610–4. 10.1177/1745691611425012

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 20.

  JonesPJMaRMcNallyRJ. Bridge centrality: a network approach to understanding comorbidity. Multivariate Behav Res. (2019) 1–15. 10.1080/00273171.2019.1614898. [Epub ahead of print].

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 21.

  CramerAOJWaldorpLJvan der MaasHLJBorsboomD. Complex realities require complex theories: refining and extending the network approach to mental disorders. Behav Brain Sci. (2010) 33:178–93. 10.1017/S0140525X10000920

  • CrossRef
  • Google Scholar

• 22.

  KozelDMarusićA. Individuals with diabetes mellitus with and without depressive symptoms: could social network explain the comorbidity?Psychiatr Danub. (2006) 18:12–8.

  • Pubmed Abstract
  • Google Scholar

• 23.

  BeardCMillnerAJForgeardMJFriedEIHsuKJTreadwayMTet al. Network analysis of depression and anxiety symptom relationships in a psychiatric sample. Psychol Med. (2016) 46:3359–69. 10.1017/s0033291716002300

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 24.

  AfzaliMHSunderlandMTeessonMCarragherNMillsKSladeT. 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

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 25.

  ChoiKWBatchelderAWEhlingerPPSafrenSAO'CleirighC. Applying network analysis to psychological comorbidity and health behavior: depression, PTSD, and sexual risk in sexual minority men with trauma histories. J Consult Clin Psychol. (2017) 85:1158–70. 10.1037/ccp0000241

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 26.

  McNallyRJMairPMugnoBLRiemannBC. Co-morbid obsessive-compulsive disorder and depression: a Bayesian network approach. Psychol Med. (2017) 47:1204–14. 10.1017/s0033291716003287

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27.

  BoschlooL. Zooming in and zooming out: a network perspective on the comorbidity of depression and anxiety. J Am Acad Child Adolesc Psychiatry. (2018) 57:914–5. 10.1016/j.jaac.2018.07.880

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 28.

  ZaniratiGAzevedoPNVenturinGTGreggioSAlcaráAMZimmerERet al. Depression comorbidity in epileptic rats is related to brain glucose hypometabolism and hypersynchronicity in the metabolic network architecture. Epilepsia. (2018) 59:923–34. 10.1111/epi.14057

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29.

  GarabilesMRLaoCKXiongYHallBJ. Exploring comorbidity between anxiety and depression among migrant Filipino domestic workers: a network approach. J Affect Disord. (2019) 250:85–93. 10.1016/j.jad.2019.02.062

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 30.

  PriceMLegrandACBrierZMFHébert-DufresneL. The symptoms at the center: examining the comorbidity of posttraumatic stress disorder, generalized anxiety disorder, and depression with network analysis. J Psychiatr Res. (2019) 109:52–8. 10.1016/j.jpsychires.2018.11.016

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31.

  LazarovASuarez-JimenezBLeviOCoppersmithDDLLubinGPineDSet al. 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

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32.

  ArtinoARJrPhillipsAWUtrankarATaAQDurningSJ. “The Questions Shape the Answers”: assessing the quality of published survey instruments in health professions education research. Acad Med. (2018) 93:456–63. 10.1097/acm.0000000000002002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 33.

  EpskampSFriedEI. A tutorial on regularized partial correlation networks. Psychol Methods. (2018) 23:617–34. 10.1037/met0000167

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34.

  BergerMLSoxHWillkeRJBrixnerDLEichlerHGGoettschWet al. Good practices for real-world data studies of treatment and/or comparative effectiveness: recommendations from the Joint ISPOR-ISPE Special Task Force on real-world evidence in health care decision making. Value Health. (2017) 20:1003–8. 10.1016/j.jval.2017.08.3019

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 35.

  GeFJiangJWangYYuanCZhangW. Identifying suicidal ideation among Chinese patients with major depressive disorder: evidence from a real-world hospital-based study in China. Neuropsychiatr Dis Treat. (2020) 16:665–72. 10.2147/ndt.s238286

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 36.

  World Medical Association declaration of Helsinki. Recommendations guiding physicians in biomedical research involving human subjects. JAMA. (1997) 277:925–6.

  • Google Scholar

• 37.

  World Medical Association declaration of Helsinki. Recommendations guiding physicians in biomedical research involving human subjects. J Med Liban. (1994) 42:88–9.

  • Google Scholar

• 38.

  HamiltonM. A rating scale for depression. J Neurol Neurosurg Psychiatry. (1960) 23:56–62. 10.1136/jnnp.23.1.56

  • CrossRef
  • Google Scholar

• 39.

  HamiltonM. The assessment of anxiety states by rating. Br J Med Psychol. (1959) 32:50–5.

  • Google Scholar

• 40.

  YanM. XS, Hu M. The Chinese translation, reliability and validity of some depression scales in China. Chinese Mental Health J. (2016) 30:501–5. 10.3969/j.issn.1000-6729.2016.07.005

  • CrossRef
  • Google Scholar

• 41.

  DondersARvan der HeijdenGJStijnenTMoonsKG. Review: a gentle introduction to imputation of missing values. J Clin Epidemiol. (2006) 59:1087–91. 10.1016/j.jclinepi.2006.01.014

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 42.

  EpskampSBorsboomDFriedEI. Estimating psychological networks and their accuracy: a tutorial paper. Behav Res Methods. (2018) 50:195–212. 10.3758/s13428-017-0862-1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 43.

  JonesPJ. Networktools: Tools for Identifying Important Nodes in Networks. R Package Version 1.1.0 (2017). Available online at: https://CRAN.R-project.org/package=networktools

  • Google Scholar

• 44.

  KnefelMTranUSLueger-SchusterB. The association of posttraumatic stress disorder, complex posttraumatic stress disorder, and borderline personality disorder from a network analytical perspective. J Anxiety Disord. (2016) 43:70–8. 10.1016/j.janxdis.2016.09.002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 45.

  JonesPJMairPRiemannBCMugnoBLMcNallyRJ. A network perspective on comorbid depression in adolescents with obsessive-compulsive disorder. J Anxiety Disord. (2018) 53:1–8. 10.1016/j.janxdis.2017.09.008

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 46.

  FriedEIEpskampSNesseRMTuerlinckxFBorsboomD. What are 'good' depression symptoms? Comparing the centrality of DSM and non-DSM symptoms of depression in a network analysis. J Affect Disord. (2016) 189:314–20. 10.1016/j.jad.2015.09.005

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 47.

  BoschlooLSchoeversRAvan BorkuloCDBorsboomDOldehinkelAJ. The network structure of psychopathology in a community sample of preadolescents. J Abnorm Psychol. (2016) 125:599–606. 10.1037/abn0000150

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 48.

  CramerAOJvan der SluisSNoordhofAWichersMGeschwindNAggenSHet al. Dimensions of normal personality as networks in search of equilibrium: you can't like parties if you don't like people. Eur J Pers. (2012) 26:414–31. 10.1002/per.1866

  • CrossRef
  • Google Scholar

• 49.

  RyderAG. Cross-cultural Differences in the Presentation of Depression: Chinese Somatization and Western Psychologization. Vancouver, BC: University of British Columbia (2004).

  • Google Scholar

• 50.

  ZhouXDereJZhuXYaoSChentsova-DuttonYERyderAG. Anxiety symptom presentations in Han Chinese and Euro-Canadian outpatients: is distress always somatized in China?J Affect Disord. (2011) 135:111–4. 10.1016/j.jad.2011.06.049

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 51.

  DereJSunJZhaoYPerssonTJZhuXYaoSet al. Beyond “somatization” and “psychologization”: symptom-level variation in depressed Han Chinese and Euro-Canadian outpatients. Front Psychol. (2013) 4:377. 10.3389/fpsyg.2013.00377

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 52.

  BraundTAPalmerDMWilliamsLMHarrisAWF. Dimensions of anxiety in Major depressive disorder and their use in predicting antidepressant treatment outcome: an iSPOT-D report. Psychol Med. (2020) 50:1032–42. 10.1017/s0033291719000941

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 53.

  GoghariVMHarrowM. Anxiety symptoms across twenty-years in schizoaffective disorder, bipolar disorder, and major depressive disorder. Psychiatry Res. (2019) 275:310–4. 10.1016/j.psychres.2019.03.050

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 54.

  VirtanenTEskelinenSSailasESuvisaariJ. Dyspepsia and constipation in patients with schizophrenia spectrum disorders. Nordic J Psychiatry. (2017) 71:48–54. 10.1080/08039488.2016.1217044

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 55.

  WernerssonRCarlssonJ. Posttraumatic stress disorder is correlated to irritable bowel syndrome. Ugeskrift Laeger. (2015) 177:1248–52.

  • Pubmed Abstract
  • Google Scholar

• 56.

  HolingueCNewillCLeeLCPasrichaPJDaniele FallinM. Gastrointestinal symptoms in autism spectrum disorder: a review of the literature on ascertainment and prevalence. Autism Res. (2018) 11:24–36. 10.1002/aur.1854

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 57.

  SimpsonCADiaz-ArtecheCElibyDSchwartzOSSimmonsJGCowanCSM. The gut microbiota in anxiety and depression - a systematic review. Clin Psychol Rev. (2020) 83:101943. 10.1016/j.cpr.2020.101943

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 58.

  PettitJWBechorMReyYVaseyMWAbendRPineDSet al. A randomized controlled trial of attention bias modification treatment in youth with treatment-resistant anxiety disorders. J Am Acad Child Adolesc Psychiatry. (2020) 59:157–65. 10.1016/j.jaac.2019.02.018

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 59.

  MedicineAAoS. International Classification of Sleep Disorders, 3rd ed. Darien: American Academy of Sleep Medicine (2014).

  • Google Scholar

• 60.

  HinkelmannKMoritzSBotzenhardtJMuhtzCWiedemannKKellnerMet al. Changes in cortisol secretion during antidepressive treatment and cognitive improvement in patients with major depression: a longitudinal study. Psychoneuroendocrinology. (2012) 37:685–92. 10.1016/j.psyneuen.2011.08.012

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 61.

  BorsboomD. A network theory of mental disorders. World Psychiatry. (2017) 16:5–13. 10.1002/wps.20375

  • CrossRef
  • Google Scholar

• 62.

  Mallorquí-BaguéNBulbenaAPailhezGGarfinkelSNCritchleyHD. Mind-body interactions in anxiety and somatic symptoms. Harvard Rev Psychiatry. (2016) 24:53–60. 10.1097/hrp.0000000000000085

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 63.

  FangHTuSShengJShaoA. Depression in sleep disturbance: a review on a bidirectional relationship, mechanisms and treatment. J Cell Mol Med. (2019) 23:2324–32. 10.1111/jcmm.14170

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64.

  McCallWVBlockerJND'AgostinoRJr.KimballJBoggsNet al. Treatment of insomnia in depressed insomniacs: effects on health-related quality of life, objective and self-reported sleep, and depression. J Clin Sleep Med. (2010) 6:322–9. 10.5664/jcsm.27872

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 65.

  ManberREdingerJDGressJLSanPedro-Salcedo MGKuoTFKalistaT. Cognitive behavioral therapy for insomnia enhances depression outcome in patients with comorbid major depressive disorder and insomnia. Sleep. (2008) 31:489–95. 10.1093/sleep/31.4.489

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 66.

  FruehBCGrubaughALCusackKJKimbleMOElhaiJDKnappRG. Exposure-based cognitive-behavioral treatment of PTSD in adults with schizophrenia or schizoaffective disorder: a pilot study. J Anxiety Disord. (2009) 23:665–75. 10.1016/j.janxdis.2009.02.005

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 67.

  EpskampSRhemtullaMBorsboomD. Generalized network psychometrics: combining network and latent variable models. Psychometrika. (2017) 82:904–27. 10.1007/s11336-017-9557-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 68.

  EpskampSKruisJMarsmanM. Estimating psychopathological networks: be careful what you wish for. PLoS ONE. (2017) 12:e0179891. 10.1371/journal.pone.0179891

  • CrossRef
  • Google Scholar

• 69.

  GurejeOLewis-FernandezRHallBJReedGM. Systematic inclusion of culture-related information in ICD-11. World Psychiatry. (2019) 18:357–8. 10.1002/wps.20676

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 70.

  Wise MarwoodMAPerkins Herane-Vives Jouleset al. Instability of default mode network connectivity in major depression: a two-sample confirmation study. Transl Psychiatry. (2017) 7:e1105. 10.1038/tp.2017.40

  • Pubmed Abstract
  • CrossRef
  • Google Scholar