Which estimation method to choose in network psychometrics? Deriving guidelines for applied researchers

Comparing the Clique Percolation algorithm to other overlapping community detection algorithms in psychological networks: A Monte Carlo simulation study

In psychological networks, one limitation of the most used community detection algorithms is that they can only assign each node (symptom) to a unique community, without being able to identify overlapping symptoms. The clique percolation (CP) is an algorithm that identifies overlapping symptoms but its performance has not been evaluated in psychological networks. In this study, we compare the CP with model parameters chosen based on fuzzy modularity (CPMod) with two other alternatives, the ratio of the two largest communities (CPRat), and entropy (CPEnt). We evaluate their performance to: (1) identify the correct number of latent factors (i.e., communities); and (2) identify the observed variables with substantive (and equally sized) cross-loadings (i.e., overlapping symptoms). We carried out simulations under 972 conditions (3x2x2x3x3x3x3): (1) data categories (continuous, polytomous and dichotomous); (2) number of factors (two and four); (3) number of observed variables per factor (four and eight); (4) factor correlations (0.0, 0.5, and 0.7); (5) size of primary factor loadings (0.40, 0.55, and 0.70); (6) proportion of observed variables with substantive cross-loadings (0.0%, 12.5%, and 25.0%); and (7) sample size (300, 500, and 1000). Performance was evaluated through the Omega index, Mean Bias Error (MBE), Mean Absolute Error (MAE), sensitivity, specificity, and mean number of isolated nodes. We also evaluated two other methods, Exploratory Factor Analysis and the Walktrap algorithm modified to consider overlap (EFA-Ov and Walk-Ov, respectively). The Walk-Ov displayed the best performance across most conditions and is the recommended option to identify communities with overlapping symptoms in psychological networks.

Dynamics of postnatal depressive symptoms in early parenthood

BACKGROUND

New mothers and fathers are at risk of developing postnatal depressive problems. To understand how postnatal depressive symptoms unfold over time, analyses at the within-person level are necessary. Inspecting postnatal depressive problems at the symptom level provides a novel perspective, ultimately offering insight into which symptoms contribute to the elevation of other symptoms over time.

METHODS

Panel graphical vector-autoregression (GVAR) models were applied to analyze the within-person temporal and contemporaneous relations between depressive symptoms across the postnatal period in new mothers and fathers (at T1; Nmothers = 869, Nfathers = 579). Depressive symptoms were assessed at 6-, 12-, and 18-months postpartum, using the Edinburgh Postnatal Depression Scale.

RESULTS

The results revealed that for mothers, sadness was a key symptom predicting symptom increases in multiple other depressive symptoms and itself (autoregressive effect) over time. Furthermore, anxiousness and feeling scared predicted each other across the postnatal period in mothers. For fathers, the most central predicting symptom in the overall network of symptoms was being anxious, while self-blame and being overwhelmed had strong self-maintaining roles in the fathers' symptomatology, indicating that these could be key features in fathers experiencing postnatal depressive problems. The pattern of symptoms that mothers and fathers experienced within the same time window (contemporaneous associations), shared many of the same characteristics compared to the temporal structure.

CONCLUSIONS

This study suggests that across the postnatal period, from 6- to 18-months postpartum, depressive symptoms in mothers and fathers contribute differently to the pattern of depressive problems, highlighting sadness as a key feature in maternal symptomatology and anxiousness components in paternal symptomatology.

Network analysis of pandemic fatigue symptoms in samples from five South American countries

BACKGROUND

Pandemic fatigue generates low motivation or the ability to comply with protective behaviors to mitigate the spread of COVID-19.

AIMS

This study aimed to analyze the symptoms of pandemic fatigue through network analysis in individuals from five South American countries.

METHOD

A total of 1,444 individuals from Argentina, Bolivia, Paraguay, Peru, and Uruguay participated and were evaluated using the Pandemic Fatigue Scale. The networks were estimated using the ggmModSelect estimation method and a polychoric correlation matrix was used. Stability assessment of the five networks was performed using the nonparametric resampling method based on the case bootstrap type. For the estimation of network centrality, a metric based on node strength was used, whereas network comparison was performed using a permutation-based approach.

RESULTS

The results showed that the relationships between pandemic fatigue symptoms were strongest in the demotivation dimension. Variability in the centrality of pandemic fatigue symptoms was observed among participating countries. Finally, symptom networks were invariant and almost identical across participating countries.

CONCLUSIONS

This study is the first to provide information on how pandemic fatigue symptoms were related during the COVID-19 pandemic.

Network analysis of depression, cognitive functions, and suicidal ideation in patients with diabetes: an epidemiological study in Iran

AIMS

The main aim of this study was to assess the prevalence of suicidal ideation and previous suicide attempts among Iranian patients diagnosed with Type-1 diabetes (T1D) and Type-2 diabetes (T2D). Additionally, the study sought to estimate the network structure of depressive symptoms and cognitive functions.

METHODS

1073 patients participated in the current study. We used Patient Health Questionnaire-9 (PHQ-9), Ask Suicide-Screening Questionnaire, diabetes-related factors, and a battery of cognitive functions tasks to estimate network structures. Also, suicidal ideations and suicide attempts prevalence have been estimated. Statistical analyses were performed using R-studio software, including mixed-graphical models (MGMs) for undirected effects and Directed Acyclic Graphs (DAGs) for directed effects.

RESULTS

The prevalence of suicidal ideation was 29.97% in T1D and 26.81% in T2D (p < 0.05). The history of suicide attempts was higher in T1D (10.78%) compared to T2D (8.36%) (p < 0.01). In the MRF networks for T1D, suicidal ideation was directly linked to 'feeling guilt (PHQ.6)', 'Suicide (PHQ.9)', HbA1c, and FBS, while the Inhibition node was directly related to suicidal ideation. The DAGs suggested connections between 'depression', HbA1c, and 'inhibition' with suicidal ideation, along with a link between the current family history of suicide attempts and the patient's history of suicide attempts. For T2D, the MRF networks indicated direct links between suicidal ideation and 'anhedonia (PHQ.1)', 'suicide (PHQ.9)', age, being female, and BMI, with inhibition also being directly related to suicidal ideation. The DAGs revealed connections between 'depression', age, and 'inhibition' with suicidal ideation, as well as links between being female or single/divorced and the patient's history of suicide attempts.

CONCLUSION

The findings suggest that suicide ideation is highly prevalent in patients with diabetes, and these symptoms should be carefully monitored in these patients.

The structure of depressive symptoms using CES-D and ZDS in outpatients in a general hospital in Lima, Peru

BACKGROUND

Depression represents one of the leading causes of disability due to illness worldwide. Previous studies have demonstrated the significant heterogeneity of the diagnosis of depression, making it necessary to develop new diagnostic approaches. Network analysis is a perspective that considers symptoms as constituents of the psychiatric disorder itself. The objective was to determine the structure of depressive symptoms using the CES-D and ZDS depression scales.

METHODS

Cross-sectional study of secondary analysis of 194 patients using the CES-D and ZDS scales. Correlation matrices and regularised partial correlation networks were constructed from the database. Centrality measures were estimated, and a network stability analysis was performed.

RESULTS

On the CES-D scale, the most central item was "Sad"; while on the ZDS scale, the most central items were "Sad" and "Live". On the CES-D scale, the connection between "Enjoy" and "Happy" was the strongest. On the ZDS scale, the strongest connection was between the items "Live" with "Useful". The item "Morning" was the least connected on the ZDS.

CONCLUSIONS

The most central symptom from the CES-D scale was sadness, while from the ZDS scale, was sadness and anhedonia.

Networks of Adversity in Childhood and Adolescence and Their Relationship to Adult Mental Health

Adverse experiences before the age of eighteen are common and include diverse events ranging from sexual abuse to parental divorce. These stressful experiences have been linked to physical and mental health issues. Previous research has focused mainly on childhood adversity, such as experiences in the family environment. Little consideration has been given to adversities that may be particularly harmful in adolescence. To understand adolescents' adverse experiences, this project used data from the Avon Longitudinal Study of Parents and Children (ALSPAC, total N = 14,901, N ≈ 1,200 - 10,000 per measure). We modelled interrelations of adversities in childhood (1-11 years) and adolescence (11-23 years) and examined adversity clusters using network analysis. We found two similar clusters in the childhood and adolescence networks: (1) direct abuse and (2) adverse family factors. We identified a third cluster of (3) educational and social adversities for adolescence. For both age groups, emotional abuse in the family environment was closely linked to mental health in early adulthood and most adversities were linked with depression in early adulthood. In adolescence, housing and academic issues and abuse by a romantic partner were particularly central to the network of adversities. Thus, we found commonalities and differences in the relevance of adverse experiences at different developmental stages. These findings highlight the need to develop age-dependent frameworks for adversity research and policymaking.

Bridge symptoms between parenting styles and proximal psychological risk factors associated with adolescent suicidal thoughts: a network analysis

BACKGROUND

Parenting styles and the associated proximal psychological factors are suggested to increase suicidal risks in adolescents. However, how the two factors interact and confer risks on the emergence of adolescent suicidal thoughts remains unclear. Herein, we used a network approach to investigate their interrelationship and explore whether the network properties predict adolescent suicidal thoughts.

METHODS

Self-report questionnaires were completed by 1171 students aged 12-16. Network analyses were performed by Gaussian graphical models estimating the adolescent psychosocial network structure of parenting styles and psychological variables including depression, anxiety, affective lability, rumination, and resilience. Furthermore, we re-examined the network by adding a variable measuring active suicidal thoughts. Moreover, we conducted linear regressions to examine the predictive utility of bridge symptoms for adolescent suicidal thoughts.

RESULTS

Resilience, Afraid, Rumination, Concentration, and affective lability (Anger) had the highest bridge strengths in the adolescent psychosocial network. Among the identified bridge symptoms, Resilience was negatively correlated with active suicidal thoughts (regularized edge weights = -0.181, bootstrapped 95% CIs: [-0.043, -0.155]), whereas affective lability (from Anxiety to Depression, Anger), Rumination, and Afraid were positively correlated with active suicidal thoughts, with edge weights (bootstrapped 95% CIs) ranging from 0.057 (0.001, 0.112) to 0.081(0.026, 0.136). Regression analysis showed that bridge strength was significantly correlated with active suicidal thoughts (R2 = 0.432, P = 0.001).

CONCLUSION

Negative parenting styles may drive and maintain suicidal thoughts by modifying the key proximal psychological variables. Our findings highlight the important role of bridge symptoms, which may serve as vital targets for triggering adolescent suicide.

An exploration of gender and prolonged grief symptoms using network analysis

BACKGROUND

Gender has been proposed as a potentially important predictor of bereavement outcomes. The majority of research in the field has explored this issue by examining gender differences in global grief severity. Findings have been mixed. In this study, we explore potential gender differences in grief using network analysis. This approach examines how individual symptoms relate to and reinforce each other, and so offers potential to shed light on novel aspects of grief expression across genders.

METHOD

Graphical lasso networks were constructed using self-report data from 839 spousally bereaved older participants (584 female, 255 male) collected at 2- and 11- months post-bereavement. Edge strength, node strength and global network strength were compared to identify similarities and differences between gender networks across time.

RESULTS

At both time points, the strongest connection for both genders was from yearning to pangs of grief. Yearning, pangs of grief, acceptance, bitterness and shock were prominent nodes at time 1. Numbness and meaninglessness emerged as prominent nodes at time 2. Males and females differed in the relative importance of shock at time 1, and the female network had greater overall strength than the male network at time 2.

CONCLUSIONS

This study identified many similarities and few differences in the relationships between prolonged grief symptoms for males and females. Findings suggest that future studies should examine alternate sources of variation in grief outcomes. Limitations are discussed.

Partial Identification of Latent Correlations with Ordinal Data

The polychoric correlation is a popular measure of association for ordinal data. It estimates a latent correlation, i.e., the correlation of a latent vector. This vector is assumed to be bivariate normal, an assumption that cannot always be justified. When bivariate normality does not hold, the polychoric correlation will not necessarily approximate the true latent correlation, even when the observed variables have many categories. We calculate the sets of possible values of the latent correlation when latent bivariate normality is not necessarily true, but at least the latent marginals are known. The resulting sets are called partial identification sets, and are shown to shrink to the true latent correlation as the number of categories increase. Moreover, we investigate partial identification under the additional assumption that the latent copula is symmetric, and calculate the partial identification set when one variable is ordinal and another is continuous. We show that little can be said about latent correlations, unless we have impractically many categories or we know a great deal about the distribution of the latent vector. An open-source R package is available for applying our results.

Conceptualizing anxiety and depression in children and adolescents: a latent factor and network analysis

The objective of this study is to gain insight into the inherent structure of anxiety and depressive symptoms by combining the strengths of latent factor analysis and network analysis. The sample comprised 743 children and adolescents aged 4–18 years (M = 11.64, SD = 3.66, 61% males) who sought routine care outpatient psychotherapy. Parents or primary caregivers rated anxiety and depressive symptoms of their children on a DSM-5-/ICD-10-based symptom checklist. First, we analyzed the factor structure of the internalizing symptoms using exploratory factor analysis (EFA). Next, we conducted a network analysis and identified central and bridge symptoms that may explain comorbidity between anxiety disorders and depression. We then employed exploratory graph analysis (EGA) as an alternative tool within the framework of network psychometrics to estimate the number of dimensions (i.e., communities within a network). Finally, we tested a model based on these results using confirmatory factor analysis. The results demonstrate a complex interplay between anxiety and depressive symptom domains. Four factors/communities were identified by EFA and EGA, but the item-community allocation differed, and the interpretation of factors/communities was unclear. A clear distinction between these domains could not be supported. However, associations within a domain were stronger than associations between the two domains. We identified pain, suicidal, irritable, and afraid of adults as bridge items between the symptom domains. In conclusion, our findings further advance the general understanding of the frequently reported co-occurrence of anxiety and depressive symptoms and diagnoses in clinical practice. Identifying bridge symptoms may inform intervention practices by targeting specific symptoms that contribute to the maintenance of anxious and depressive behaviors.

Longitudinal analysis of the Non-Motor Symptoms Scale in Parkinson's Disease (NMSS): An exploratory network analysis approach

Introduction

Parkinson's disease (PD) is a multisystem neurodegenerative disorder characterized by motor and non-motor symptoms. In particular, non-motor symptoms have become increasingly relevant to disease progression. This study aimed to reveal which non-motor symptoms have the highest impact on the complex interacting system of various non-motor symptoms and to determine the progression of these interactions over time.

Methods

We performed exploratory network analyses of 499 patients with PD from the Cohort of Patients with Parkinson's Disease in Spain study, who had Non-Motor Symptoms Scale in Parkinson's Disease ratings obtained at baseline and a 2-year follow-up. Patients were aged between 30 and 75 years and had no dementia. The strength centrality measures were determined using the extended Bayesian information criterion and the least absolute shrinkage and selection operator. A network comparison test was conducted for the longitudinal analyses.

Results

Our study revealed that the depressive symptoms anhedonia and feeling sad had the strongest impact on the overall pattern of non-motor symptoms in PD. Although several non-motor symptoms increase in intensity over time, their complex interacting networks remain stable.

Conclusion

Our results suggest that anhedonia and feeling sad are influential non-motor symptoms in the network and, thus, are promising targets for interventions as they are closely linked to other non-motor symptoms.

Presenting wicked problems in a science museum: A methodology to study interest from a dynamic perspective

Science centers and science museums have an important social role in engaging people with science and technology relevant for complex societal problems-so called wicked problems. We used the case of personalized medicine to illustrate a methodology that can be used to inform the development of exhibitions on such wicked problems. The methodology that is presented is grounded in dynamic theories of interest development that define interest as a multidimensional construct involving knowledge, behavior (personal and general) value, self-efficacy, and emotion. The methodology uses a mixed method design that is able to (1) study the predictive effects of background variables on interest, (2) study the interest dimensions predicting individual interest, and (3) identify the most influential interest dimensions. We set up focus groups (N = 16, age = 20-74, low SES) to design a survey study (N = 341, age 19-89 years olds with a broad range of SES) about people's interest in personalized medicine. Results of a network analysis of the survey data show that despite the variety in emotions and knowledge about subtopics, these dimensions do not play a central role in the multidimensional interest construct. In contrast, general value and behavior (related to understanding scientific research) seem to be interesting candidates for eliciting situational interest that could have an effect on the more long term individual interest. These results are specific for the case of personalized medicine. We discuss ways in which results of studies with the presented methodology might be useful for exhibition development.

Auditing the research practices and statistical analyses of the group-level temporal network approach to psychological constructs: A systematic scoping review

Network analyses have become increasingly common within the field of psychology, and temporal network analyses in particular are quickly gaining traction, with many of the initial articles earning substantial interest. However, substantial heterogeneity exists within the study designs and methodology, rendering it difficult to form a comprehensive view of its application in psychology research. Since the field is quickly growing and since there have been many study-to-study variations in terms of choices made by researchers when collecting, processing, and analyzing data, we saw the need to audit this field and formulate a comprehensive view of current temporal network analyses. To systematically chart researchers' practices when conducting temporal network analyses, we reviewed articles conducting temporal network analyses on psychological variables (published until March 2021) in the framework of a scoping review. We identified 43 articles and present the detailed results of how researchers are currently conducting temporal network analyses. A commonality across results concerns the wide variety of data collection and analytical practices, along with a lack of consistency between articles about what is reported. We use these results, along with relevant literature from the fields of ecological momentary assessment and network analysis, to formulate recommendations on what type of data is suited for temporal network analyses as well as optimal methods to preprocess and analyze data. As the field is new, we also discuss key future steps to help usher the field's progress forward and offer a reporting checklist to help researchers navigate conducting and reporting temporal network analyses.

Risk factors of suicidal behaviors in a high-risk longitudinal veteran sample: A network analysis

INTRODUCTION

Suicide is a substantial public health burden, particularly among veterans. Risk factors have been delineated for suicide; however, the dynamic interrelations between risk factors have not been fully examined. Such research has the potential to elucidate processes that contribute to suicide risk between individuals with a past suicide attempt (attempters) and those without a past suicide attempt (nonattempters).

METHODS

In the current study, network analysis was used to compare networks between attempters and nonattempters in a high-risk veteran sample (N = 770; M_age  = 32.3 years, SD = 6.8; 326 with a past suicide attempt) who were followed over 1 year. Networks were estimated to examine (1) concurrent relations of suicide risk factors at baseline and (2) predictability of prospective suicidal behavior (SB).

RESULTS

There were no differences in the overall connectivity of attempter and nonattempter networks. Perceived burdensomeness and posttraumatic stress disorder (PTSD) symptoms were most central in the attempters' network, whereas PTSD symptoms and insomnia were most central in the nonattempters' network. The risk factors prospective SB in either network. However, attempters were more likely to engage in SB over the course of the study.

CONCLUSION

These findings highlight the difficulty in predicting who will attempt suicide.

Escaping negative moods and concentration problems play bridge roles in the symptom network of problematic smartphone use and depression

Background

Problematic smartphone use (PSU) has become an issue of great concern in the age of smartphones. PSU is associated with emotional problems, one of which is depression, as shown by empirical studies. However, previous studies have been limited in that they have focused solely on the total score for symptoms of PSU and depression while ignoring the symptomatic heterogeneity of these two concepts.

Method

This study filled this gap by exploring links between symptoms of PSU and depression among 1,849 university students (59.17% female participants, 17-23 years old). Network analysis was utilized to reveal symptom connections, central symptoms, and bridge symptoms between PSU and depression.

Results

(1) A total of 17 of 81 symptom connections (about 21%) between PSU and depression existed in the symptom network. For example, "self-control failure" for PSU was positively correlated with "concentration problems" for depression; (2) "recklessly continuing" for PSU and "fatigue" for depression were central symptoms within the PSU symptom network and depression symptom network, respectively; (3) "escaping negative moods" for PSU and "concentration problems" for depression were bridge symptoms. The former was maximumly connected with the depression symptoms and the latter was maximumly connected with the PSU symptoms; and (4) gender had very minimal influence on the network characteristics.

Conclusion

The results are in keeping with the central idea of the compensatory internet use theory that excessive smartphone use may be a coping strategy for depressed emotions derived from escaping motivation. Moreover, concentration problems may be a mediator explaining how negative emotions (e.g., depression) cause PSU, which is undefined in current internet use theories. Finally, symptom connections, central symptoms, and bridge symptoms could be potential targets for the prevention and intervention of PSU and depression in young adults.

Gaussian graphical models with applications to omics analyses

Gaussian graphical models (GGMs) provide a framework for modeling conditional dependencies in multivariate data. In this tutorial, we provide an overview of GGM theory and a demonstration of various GGM tools in R. The mathematical foundations of GGMs are introduced with the goal of enabling the researcher to draw practical conclusions by interpreting model results. Background literature is presented, emphasizing methods recently developed for high-dimensional applications such as genomics, proteomics, or metabolomics. The application of these methods is illustrated using a publicly available dataset of gene expression profiles from 578 participants with ovarian cancer in The Cancer Genome Atlas. Stand-alone code for the demonstration is available as an RMarkdown file at https://github.com/katehoffshutta/ggmTutorial.

The tangled webs we wreak: Examining the structure of aggressive personality using psychometric networks

OBJECTIVE

Trait aggression is a prominent construct in the psychological literature, yet little work has sought to situate trait aggression among broader frameworks of personality. Initial evidence suggests that trait aggression may be best couched within the nomological network of the Five-Factor Model (FFM). The current work sought to locate the most appropriate home for trait aggression among the FFM.

METHOD

We applied a preregistered regimen of psychometric network analyses to three datasets (combined N = 2927) that contained self-reports of trait aggression and the FFM traits.

RESULTS

Trait aggression was highly central in the factor-level networks, which contained associations consistent with the conceptualization of this construct as a lower-order component of low agreeableness. The facet-level networks revealed that the behavioral facets of trait aggression reflected low agreeableness, but that the anger and hostility facets reflected high neuroticism. The item-level network suggested that the intent to initiate aggressive encounters was the primary bridge that empirically linked trait aggression to agreeableness.

CONCLUSIONS

Our results indicate that trait aggression is primarily a lower-order facet of agreeableness, advance our understanding of trait aggression, integrate it with broader frameworks of personality, and suggest future directions to refine this complex dispositional tendency.

Anxiety, depression, stress, worry about COVID-19 and fear of loneliness during COVID-19 lockdown in Peru: A network analysis approach

This study aims to examine the relationships between symptoms of anxiety, depression, stress, worry about COVID-19 and fear of loneliness during COVID-19 lockdown in Peru using network analysis. There were 854 participants aged 18 to 50 years (Mean = 36.54; SD = 9.23); 634 females (74.20%) and 220 males (25.80%), who completed the Generalized Anxiety Disorder Scale (GAD-7), Patient Health Questionnaire (PHQ-9), Perceived Stress Scale (PSS-10), Preoccupation with COVID-19 Contagion (PRE-COVID-19), Brief Scale of Fear of Loneliness (BSFL). A partial unregularized network was estimated through the ggmModSelect function. Expected influence (EI) and bridging EI values were calculated to identify central symptoms and bridging symptoms respectively. The results reveal those two symptoms of depression-stress and anxiety-were the most central symptoms in the network. Depressive symptoms are at the same time the most comorbid and it is shown that there are no differences in the network when compared between those who left home and those who did not leave home during lockdown. Depressive symptoms are concluded to be central and bridging in the network and interconnected with some symptoms of stress and anxiety. These findings may be important to understand the experience of COVID-19 lockdown in Peru.

Depressive and anxiety symptoms in patients with COPD: A network analysis

BACKGROUND

Individuals with Chronic Obstructive Pulmonary Disease (COPD) often develop anxiety and depression, which worsen illness management and prognosis. Physical and psychological symptoms, contextual and illness-related factors display complex reciprocal interactions, which give rise to heterogeneous presentations. Examining the patterns of association between specific physical and psychological symptoms in patients with COPD may help to focus on the precision of the patient-centred care.

RESEARCH QUESTION

We used network analyses to examine the links between symptoms of COPD, depression and anxiety.

METHODS

Data from 1587 individuals with COPD from the COPDGene study were included. We estimated a Bayesian Gaussian Graphical Model to highlight the unique associations between symptoms of COPD (assessed with the COPD Assessment Test), depression and anxiety (assessed with the Hospital Anxiety and Depression Scale (HADS), while examining the role of sociodemographic characteristics, lung function tests, and health status.

RESULTS

Unique Variable Analysis reduced 14 HADS items to Tension/worry (chronic anxiety), Fear/panic (acute anxiety), Restlessness, Anhedonia, Sadness and Slowing. In network analyses, chest-tightness was related to acute anxiety, while cough and weakness were connected with core depressive symptoms (sadness and lack of pleasure). Chronic anxiety was linked with acute anxiety and depressive symptoms. Findings were confirmed accounting for the role of confounders, including lung function, sex, ethnicity and lifestyle factors. A simulation based on our model yielded distinct predictions about anxiety and depression in two participants with similar COPD severity, but different symptom profiles.

CONCLUSION

Network analyses highlighted specific associations between symptoms of COPD, depression and anxiety. Accounting for symptom-level interactions may help to promote personalized treatment approaches.

Meta-analytic Gaussian Network Aggregation

A growing number of publications focus on estimating Gaussian graphical models (GGM, networks of partial correlation coefficients). At the same time, generalizibility and replicability of these highly parameterized models are debated, and sample sizes typically found in datasets may not be sufficient for estimating the underlying network structure. In addition, while recent work emerged that aims to compare networks based on different samples, these studies do not take potential cross-study heterogeneity into account. To this end, this paper introduces methods for estimating GGMs by aggregating over multiple datasets. We first introduce a general maximum likelihood estimation modeling framework in which all discussed models are embedded. This modeling framework is subsequently used to introduce meta-analytic Gaussian network aggregation (MAGNA). We discuss two variants: fixed-effects MAGNA, in which heterogeneity across studies is not taken into account, and random-effects MAGNA, which models sample correlations and takes heterogeneity into account. We assess the performance of MAGNA in large-scale simulation studies. Finally, we exemplify the method using four datasets of post-traumatic stress disorder (PTSD) symptoms, and summarize findings from a larger meta-analysis of PTSD symptom.

Depression, COVID-19 Anxiety, Subjective Well-being, and Academic Performance in University Students With COVID-19-Infected Relatives: A Network Analysis

This study aimed to examine the relationship between anxiety, depression, subjective well-being, and academic performance in Peruvian university health science students with COVID-19-infected relatives. Eight hundred two university students aged 17-54 years (Mean 21.83; SD = 5.31); 658 females (82%) and 144 males (18%); who completed the Patient Health Questionnaire-2, Coronavirus Anxiety Scale, Subjective Well-being Scale (SWB), and Self-reporting of Academic Performance participated. A partial unregularized network was estimated using the ggmModSelect function. Expected influence (EI) values were calculated to identify the central nodes and a two-tailed permutation test for the difference between the two groups (COVID-19 infected and uninfected). The results reveal that a depression and well-being node (PHQ1-SWB3) presents the highest relationship. The most central nodes belonged to COVID-19 anxiety, and there are no global differences between the comparison networks; but at the local level, there are connections in the network of COVID-19-infected students that are not in the group that did not present this diagnosis. It is concluded that anxious-depressive symptomatology and its relationship with well-being and evaluation of academic performance should be considered in order to understand the impact that COVID-19 had on health sciences students.

Multifarious Linkages Between Personality Traits and Psychological Distress During and After COVID-19 Campus Lockdown: A Psychological Network Analysis

BACKGROUND

The novel coronavirus disease pandemic is still proliferating and is not expected to end any time soon. Several lockdowns and social distancing measures might be implemented in the future. A growing body of research has explored the effect of personality on individuals' psychological wellbeing during the pandemic. However, most prior studies have not discussed the dynamic and reciprocal transactions between personality and psychological distress in various situations. Therefore, this study aims to explore the internal mechanisms of the ways in which certain personality traits triggered specific symptoms during and after college lockdown, by using network analysis.

METHODS

Based on survey data from 525 university students in China, the study detected the connection between individual personality and psychological distress through network analysis. Of the participants, 70.1% were female, and 20.9% were male. The mean age of the participants was 19.701 (SD = 1.319) years. We estimated networks via two steps: First, two networks that only contain the Big Five personality traits and the six symptoms of psychological distress during and after the lockdown measure were estimated. Second, we add control variables and re-estimated the networks to check whether the linkages among the Big Five personality traits and the six symptoms of psychological distress observed in the first step were stable. Moreover, we employed strength centrality as the key indicator to present the potential significance of diverse variables within a network.

RESULTS

The findings demonstrate that, first, "depress" was the central symptom in the network during the college lockdown, while "efforts" was the central symptom after the lockdown. Second, the symptoms of "restless" and "worthless" significantly declined after the lockdown. Third, we found that there is an internal mechanism through which personality affected certain psychological symptoms during and after lockdowns. Specifically, neuroticism triggered certain symptoms during and after the lockdown, while extraversion and conscientiousness suppressed certain symptoms. Substantial evidence on internal linkages is imperative to develop effective interventions.

CONCLUSION

This study explores the internal mechanisms of the ways in which certain personality traits trigger specific symptoms. Overall, our results provide empirical evidence that personality traits play a key role in how individuals with certain traits respond to college lockdown during a pandemic. The study makes a significant contribution to the literature because it is among the first few studies which explores the effects of personality traits on individual psychological distress using network analysis during the pandemic.

The Interplay Between Quality of Life and Resilience Factors in Later Life: A Network Analysis

Age-related challenges and transitions can have considerable social, psychological, and physical consequences that may lead to significant changes in quality of life (QoL). As such, maintaining high levels of QoL in later life may crucially depend on the ability to demonstrate resilience (i.e., successful adaptation to late-life challenges). The current study set out to explore the interplay between several resilience factors, and how these contribute to the realization and maintenance of (different facets of) QoL. Based on the previous work, we identified behavioral coping, positive appraisal, self-management ability, and physical activity as key resilience factors. Their interplay with (various facets of) QoL, as measured with the WHOQOL-OLD, was established through network analysis. In a sample of community-dwelling older adults (55+; N=1,392), we found that QoL was most strongly (and directly) related to positive appraisal style and self-management ability. Among those, taking care of multifunctional resources (i.e., yielding various benefits at the same time) seemed to be crucial. It connected directly to "satisfaction with past, present, and future activities," a key facet of QoL with strong interconnections to other QoL facets. Our analysis also identified resilience factor(s) with the potential to promote QoL when targeted by training, intervention, or other experimental manipulation. The appropriate set of resilience factors to manipulate may depend on the goal and/or facet of QoL that one aims to improve.

Network analysis of physical and psychiatric symptoms of hospital discharged patients infected with COVID-19

In the current study, we aimed to investigate the network structure of COVID-19 symptoms and its related psychiatric symptoms, using a network approach. Specifically, we examined how COVID-19 symptoms relate to psychiatric symptoms and highlighted potential pathways between COVID-19 severity and psychiatric symptoms. With a sample of six hundred seventy-five recovered COVID-19 patients recruited 1 month after hospital discharge, we respectively integrated COVID-19 symptoms with PTSD, depression, and anxiety symptoms and analyzed the three network structures. In all three networks, COVID-19 severity and ICU admission are not linked directly to COVID-19 symptoms after hospitalization, while COVID-19 severity (but not ICU admission) is linked directly to one or more psychiatric symptoms. Specific pathways between COVID-19 symptoms and psychiatric symptoms were discussed. Finally, we used directed acyclic graph estimation to show potential causal effects between COVID-19 related variables and demographic characteristics.

Feelings of worthlessness links depressive symptoms and parental stress: A network analysis during the COVID-19 pandemic

BACKGROUND

The prevalent co-occurrence between parental stress and depression has been established prior to and during the COVID-19 pandemic outbreak. However, no studies to date have identified the connections through which these symptom domains interact with each other to emerge into a complex and detrimental mental health state, along with the plausible mechanistic variables that may play key roles in maintaining parental stress and depression. The aim of this research is to uncover these interactions in a period where parents experience heightened demands and stress because of the strict social distancing protocols.

METHODS

Network analysis is utilized to examine parental stress and depressive symptoms during the COVID-19 pandemic in a large cross-sectional study (N = 2,868) of parents. Two graphical Gaussian graphical network models were estimated, one in which only parental stress and depression symptoms were included, and another in which several mechanistic variables were added.

RESULTS

Expected influence and bridge expected influence revealed that feeling worthless was the most influential node in the symptoms network and bridged the two psychological states. Among the mechanistic variables, worry and rumination was specifically relevant in the depressive cluster of symptoms, and self-criticism was connected to both constructs.

CONCLUSION

The study displays that the co-occurrence of parental stress and depression has specific pathways, was manifested through feelings of worthlessness, and has specific patterns of connection to important mechanisms of psychopathology. The results are of utility when aiming to avoid the constellation of co-occurring parental stress and depressive symptoms during the pandemic.