Network Analysis Approach Research Articles

Interpersonal symptoms in adolescence depression across Asian and European regions: a network approach

BackgroundMajor Depressive Disorder (MDD) poses a significant global health challenge, with symptom presentation potentially varying between adolescents and adults. Adolescence is a critical period marked by heightened vulnerability to interpersonal stresses, yet the impact of these stresses on the structure of depressive symptoms is not well understood. Recognizing the cultural nuances in how depression manifests among adolescents is crucial. To this end, this paper employs a network analysis approach, utilizing a comprehensive symptom checklist from the Multidimensional Depression Assessment Scale (MDAS). Our study investigates the role of interpersonal symptoms within the broader cluster of emotional, cognitive, and somatic symptoms and explores variations in adolescent groups in four Asian and European regions.MethodsWe recruited a diverse sample of 6,348 adolescents aged 12 to 18 from Hong Kong, Taiwan, the UK, China, and the Netherlands using the Qualtrics platform. Employing the Gaussian Graphical Model, we established a network model of depressive symptoms as measured by the MDAS, segregating the sample into Asian and European regions to examine the interconnections between them. The study focused on identifying central symptom nodes and comparing the network structures between the two groups.ResultsThe analysis identified feeling worthless, low energy, being a burden to others, and low mood as central symptoms of depression. Notably, there were significant differences in the connections between depressive symptoms among Asian (Hong Kong, China and Taiwan) and European (UK and the Netherlands) adolescents, highlighting cultural variations in how interpersonal symptoms interact with emotional, cognitive, and somatic symptoms.ConclusionThis study is pioneering in applying network analysis to include interpersonal symptoms in examining depression among a diverse adolescent population. It demonstrates that interpersonal symptoms are integral to the central features of depressive symptoms. Furthermore, our findings suggest that, compared to their UK and Dutch peers, interpersonal symptoms in Asian adolescents are uniquely connected to other symptom clusters, reflecting distinct cultural patterns. Limitations: The study engaged a broad community sample; however, future research could benefit from including a larger sample size to allow for a more detailed analysis of a greater number of symptom nodes.

The spatial correlation of economic institutional change in China and its impact on economic growth: A social network analysis approach.

Economic institutional change is a vital driving force behind the rapid rise of China's economy. However, the incremental approach to economic institutional change has caused unbalanced transformation and economic growth. To this end, we adopted the entropy method to measure the economic institutional change index, and employed social network analysis to reveal its spatial correlation characteristics. We then applied QAP analysis to empirically demonstrate the impact of China's economic institutional change on regional disparities in economic growth. The findings indicated a gradual increase in the level of economic institutions over time and a spatial gradient between the eastern, central, and western regions. Moreover, the spatial correlation network of China's economic institutional change is stable and gradually improving. Nevertheless, the role of provinces in the process of economic institutional change varies: the eastern coastal provinces play a dominant role, the central and western provinces benefit to a lesser extent, and some provinces in northeastern China play a "bridging" and "intermediary" role. Regional differences in China's economic institutional change have widened the regional disparities in China's economic growth, and the impact of each dimension of economic institutions on regional disparities in economic growth is characterized by phases.

Changes in core depressive symptoms over time in peripartum women: A network analysis

Perinatal depression (PND) is a prevalent mental health condition that affects women during pregnancy and after childbirth. Distinct clinical subtypes of PND exist, with the timing of symptom onset being a pivotal element in the classification of these subtypes. However, the specific manifestations of PND across the various stages of pregnancy and the postpartum period remain to be elucidated. This study aimed to explore the changes in depression symptoms with the stage of pregnancy. Women in their second (n = 161) and third trimesters (n = 248) of pregnancy as well as those in their first 6 weeks of the postpartum period (n = 110) were recruited. Each patient was evaluated using the Edinburgh Postnatal Depression Scale. A network analysis approach was used to explore the interconnections among depressive symptoms across the different time periods. Women in the postpartum period exhibited the most pronounced prevalence of PND and severity of depressive symptoms. In the second and third trimesters, “sadness and misery” was the most central symptom. However, its prominence diminished after childbirth. “Fear and panic” was the predominant symptom in the postpartum network. The structural integrity of PND symptom networks was maintained across all three periods, with consistent strength and closeness. This study demonstrates the temporal evolution of PND’s central symptoms in women during and after pregnancy, transitioning from depression-centric to anxiety-centric manifestations. These findings advocate for symptom-specific interventions to enhance the mental well-being of mothers and their offspring. Furthermore, this study offers clinical insights into the biological underpinnings of PND subtypes, facilitating precise diagnosis and targeted treatment strategies.

Social network analysis approach to identify agricultural key communicators

Aim: This paper employed Social Network Analysis approach to visualize and calculate different social network metrics. It identifies key communicators who play pivotal roles in information flow. Methodology: Research was conducted at Jammulapalem and Perali villages of Bapatla district, Andhra Pradesh, India, during 2022-23 using an exploratory research design. A total of 120 farmers Eigen vector were selected using simple random technique, and data was collected using a well-developed interview schedule. Network metrics such as Degree centrality, betweenness centrality, and eigen vector centrality were computed to evaluate the network structure using R software (version 4.3.1). R packages, namely igraph, statnet and network D3, were used for network creation, analysis and visualization to identify the influential nodes. Results: The study revealed a complex web of relationships among various stakeholders within the agricultural network through a network graph, identifying key communicators with the highest Degree centrality. Interpretation: The focal points identified through Social Network Analysis represent a specific demographic and socio-economic group, typically aged between 35 and 55, primarily medium-scale farmers,with landholdings spanning 10 to 25 acres and high annual income, with educational backgrounds ranging from high school to pre-university college, and wield significant influence within their local communities. They require sensitization, training, practical demonstrations, and personalized support to effectively disseminate agricultural information. Key words: Agricultural Information System Network, Centrality measures, Information Sources, Key Communicators, Network visualization, Social Network Analysis

Blood transcriptome profiling reveals distinct gene networks induced by mRNA vaccination against COVID-19.

Messenger RNA (mRNA) vaccines represent a new class of vaccines that has been shown to be highly effective during the COVID-19 pandemic and that holds great potential for other preventative and therapeutic applications. While it is known that the transcriptional activity of various genes is altered following mRNA vaccination, identifying and studying gene networks could reveal important scientific insights that might inform future vaccine designs. In this study, we conducted an in-depth weighted gene correlation network analysis of the blood transcriptome before and 24h after the second and third vaccination with licensed mRNA vaccines against COVID-19 in humans, following a prime vaccination with either mRNA or ChAdOx1 vaccines. Utilizing this unsupervised gene network analysis approach, we identified distinct modular networks of co-varying genes characterized by either an expressional up- or downregulation in response to vaccination. Downregulated networks were associated with cell metabolic processes and regulation of transcription factors, while upregulated networks were associated with myeloid differentiation, antigen presentation, and antiviral, interferon-driven pathways. Within this interferon-associated network, we identified highly connected hub genes such as STAT2 and RIGI and associated upstream transcription factors, potentially playing important regulatory roles in the vaccine-induced immune response. The expression profile of this network significantly correlated with S1-specific IgG levels at the follow-up visit in vaccinated individuals. Those findings could be corroborated in a second, independent cohort of mRNA vaccine recipients. Collectively, results from this modular gene network analysis enhance the understanding of mRNA vaccines from a systems immunology perspective. Influencing specific gene networks could lead to optimized vaccines that elicit augmented vaccine responses.

Who has the greatest influence on adolescent gaming disorder: parents, teachers, or peers? An interpersonal relationships network model of gaming disorder.

Gaming disorder (GD) in adolescents is associated with impaired interpersonal relationships, including those with parents, teachers and peers. However, the interpersonal relationships most strongly associated with GD-related maladaptive behaviors are not well established. This study aimed to investigate the associations between these three types of relationships and the manifestation of GD in adolescents. In this cross-sectional study, 1920 Chinese adolescents participated in a survey that assessed interpersonal relationships (parent-child, teacher-student, and peer relationships) and demographic variables (e.g., gender, grade, duration of gaming), and 1414 participants were ultimately included. A network analysis approach was utilized to evaluate the key network metrics of edge weight and node centrality. The findings revealed that peer fear and inferiority (r = 0.12) and teacher-student conflict were most strongly correlated with GD, followed by parent-child conflict (r = 0.09). Peer fear and inferiority exhibited the highest strength centrality (0.84), followed by teacher-student conflict (0.83) and parent-child conflict (0.35). Moreover, the duration of gaming was significantly and positively correlated with GD (r = 0.19). The present study underscores the significant role of conflict and rejection within interpersonal relationships, particularly among peers, in the manifestation of GD-related behaviors in Chinese adolescents.

Enhancing Understanding of Eating Psychopathology: A Network Analysis Study

Compassion Focused Therapy for eating disorders (CFT-E) has been shown to be effective in treating eating psychopathology. Exploring which processes identified by the CFT-E model are most linked with eating disorder symptoms could help develop targeted treatments. However, the relationship between these variables requires further investigation. This study explores the interconnections between eating disorder symptoms and processes identified by the CFT-E using a network analysis approach. Participants were 497 individuals (77.3% female) from a community sample who completed measures to assess body dissatisfaction, shape and weight overvaluation, cognitive restraint, bingeing, emotional eating, purging, self-compassion, social comparison through appearance, self-criticism, appearance-related shame, guilt and pride, and positive and negative affect. A regularized partial correlation network was estimated. Appearance-related shame, inadequate self and body dissatisfaction emerged as the central nodes in the network. These findings support the use of interventions targeting shame and self-criticism for the treatment of eating disorder symptoms. Studies with diverse samples are needed to further examine the interconnections between eating psychopathology and the key processes identified by the CFT-E theoretical model.

A calibrated scale to measure heritage learning in digital environments. A network analysis approach

A calibrated scale to measure heritage learning in digital environments. A network analysis approach

The dimensional structure of the Positive and Negative Syndrome Scale in first-episode schizophrenia spectrum disorders: an Exploratory Graph Analysis from the OPTiMiSE trial

The Positive and Negative Syndrome Scale (PANSS) is the most widely used rating scale to assess psychotic symptoms in patients with schizophrenia and other primary psychoses. However, a definitive consensus regarding its dimensional structure remains elusive. The present work aims to determine the number of dimensions of the scale through a network analysis approach in a sample of individuals experiencing first-episode schizophrenia spectrum disorder (FE-SSD) with minimal or no prior exposure to antipsychotic treatment. Baseline data of 446 participants (age 25.96 ± 5.99 years, 70% males) enrolled in the OPTiMiSE trial were analysed. Exploratory Graph Analysis (EGA) was conducted to evaluate the dimensionality of the PANSS, and a bootstrap approach (bootEGA) was employed to assess model stability. The analysis was replicated, excluding unstable items with stability values below 0.75, until a stable model was achieved. The analysis of the 30 items of the PANSS revealed inadequate structural consistency, resulting in the exclusion of 9 unstable items. The final model comprised 21 symptoms distributed across four communities (Positive, Cognitive/Disorganised, Excited/Aggressive and Negative) but lacked a depressive domain. In conclusion, we propose a concise version of the PANSS, incorporating 21 items, to better assess the core symptoms of the first episode of SSD. This revised version provides clinicians with a robust psychometric tool with reduced administration time, but the complementary administration of a dedicated instrument for evaluating affective symptoms is advisable.

The association between aspects of expressive suppression emotion regulation strategy and rumination traits: a network analysis approach

Expressive suppression is an abnormal emotion regulation strategy, and its relationship with rumination traits is unclear. In this study with 395 participants in China (33.9% female, Mean age = 21.22, SD = 2.11), we estimated the association between expressive suppression and rumination traits, using the Rumination Response Scale (RRS) and the Emotion Regulation Questionnaire (ERQ) respectively. Considering there may be complex correlations between different behavioral symptoms of expressive suppression (“Keeping emotions to myself”, “Inhibiting positive emotion responses”, “Controlling emotions by not expressing them”, “Inhibiting negative emotion responses”) and different subtypes of rumination traits, this study employed a symptom-based network analysis method to uncover the differential association between rumination traits and expressive suppression, and the key symptoms linking the two. The study found the S3 node (Controlling emotions by not expressing them) had significant positive correlations with symptom rumination, brooding, and reflective pondering. Among the network, the S3 node acts as a bridge between two variables. This suggests that interventions targeting the S3 symptom may improve rumination traits. The present study was a cross-sectional study with limitations in revealing the causal relationships between expression suppression and rumination traits. Future studies could employ longitudinal tracing methods to explore the relationship between them.

A Network Analytic Approach to Dissociation: New Insights from Clinical Data

ABSTRACT The transdiagnostic construct of dissociation, characterized by a disintegration of specific psychological functions such as consciousness, memory, identity, perception, body representation, and behavior, remains elusive to a unified conceptualization. Specifically, its dimensionality is a matter of ongoing controversy. Empirical approaches applying factor analyses to the Dissociative Experiences Scale (DES) have yielded inconsistent findings. This study adopts a novel methodological approach, utilizing Exploratory Graph Analysis (EGA) to address this issue. In a sample of 668 day-hospital patients undergoing psychotherapy for a variety of mental disorders, a Gaussian graphical model was estimated for the 28 items of the DES. Additionally, the stability of the results was ensured by bootstrap procedures. While both the original EGA and the bootstrap EGA suggested four dimensions, the structural consistency of this solution was low due to an instability of 12 items. After excluding 10 of these unstable items, re-analyses again revealed a four-factor structure, but boot EGA indicated that one factor had unsatisfactory structural consistency due to the multidimensionality of its two items. Upon removing these, our final network consisted of 16 items mapping onto 3 dimensions. Our study, using data from a diagnostically heterogeneous sample, replicates and extends previous findings on the dimensionality of dissociation as captured by the DES. The three dimensions identified correspond to segregated processes, derealization/depersonalization, and absorption. This solution aligns with a bipartite model of dissociation with two broader categories referring to either altered states of consciousness (often named detachment) or to non-integrated mental modules (labeled as compartmentalization).

Testing the validity of the broaden-and build theory of positive emotions: a network analytic approach.

The Broaden-and-Build Theory of positive emotions is one of the best known and applied theories in Positive Psychology. It argues that positive emotions initiate an upward movement by opening up the mind and broadening thoughts and thus represents a counter model to the vicious circle-models of clinical psychology. The number of studies directly testing this theory in all components is scarce, ambiguities in the model impede clear inference. To draw a conclusive picture on within-model processes, we applied network modeling on the components of the theory across two studies (N 1 = 312; N 2 = 302). In both studies, the positive relationship between positive emotions, resources and life outcomes is well-supported, yet the role of broadening, as an intermediary component within these is questioned. As the broadening component consistently deviated from the model's predictions and thus did not contribute to the model as expected in either study, the validity of the Broaden-and-Build Theory in its current conceptualization is challenged, and our results point to the need to reassess the role of broadening.

The emergence and evolution of ambiguous ideas: an innovative application of social network analysis to support systematic literature reviews

AbstractSystematic literature reviews are attempts to understand conversations between researchers working to develop solutions to common problems. These conversations often stretch back decades and can involve the participation of dozens of authors. Traditional approaches to systematic reviews are ill-equipped to make sense of the sheer volume of relevant literature when exploring the emergence and evolution of ambiguous ideas across large knowledge communities. This article presents three innovative applications of Social Network Analysis (SNA) methods to explore the emergence and evolution of accountability in higher education across a collection of 450 peer-reviewed articles published from 1974-2017 and their corresponding 12,270 references. First, qualitative data from articles and references were integrated into new interactive joint displays called Narrated Network Diagrams, creating opportunities to more accurately assess themes and meanings in literature by connecting structures in co-citation networks with relevant relational stories. Second, time was elevated in the analysis procedure to capture the dynamism of knowledge formation. Third, underutilized descriptive network statistics were applied to the co-citation network analysis to generate new insights such as different mechanisms for authors gaining influence in a knowledge community. Ultimately, this article presents an innovative longitudinal Mixed Methods Social Network Analysis (MMSNA) approach to systematic literature reviews, significantly advancing previous SNA methods integration in this critical research practice.

Classification of Simulated Fake Bandwidth Data Using LSTM

Hardly will someone acknowledge that the bandwidth we use every day is as authentic as most ISPs advertise, even those offering dedicated services. There are usually shortcomings, especially on upload and download bandwidth speeds. This paper presents the classification of simulated fake bandwidth data using the Long Short-Term Memory model, which though seldom found, is a very effective approach in network analysis. There were 1400 bandwidth data points collected from the MikroTik RB 1100 AHx device in a month, then further processed with normalization, and divided to have 80% training and 20% testing. The LSTM model applied had an accuracy rate of 98.93%, proving that it is capable of classifying either genuine or fake bandwidth instances accordingly. Of 1,400 test data points, the model managed to classify 723 as fake bandwidth and another 677 as genuine, resulting in a classification error rate of only 1.07%. The results clearly prove that LSTM has huge potential for real-time bandwidth manipulation detection, key to enhancing trust and efficiency in network management. In this respect, this research shows that bandwidth analysis combined with LSTM can be an original solution for network monitoring

Understanding the Emergence of Comorbidity between Problematic Online Gaming and Gambling: A Network Analysis Approach.

Problematic online gaming and gambling tend to co-occur. The exact mechanisms underlying this phenomenon and the potential effects of gender differences remain unknown. This study aimed to identify the early clustering patterns of problematic online gaming and gambling in a community sample of young adults without a lifetime history of psychiatric treatment. Data were collected through an online survey and analyzed using partial correlations and Bayesian networks. Altogether, 1441 individuals (aged 18-40 years, 51.4% females) participated in the survey. Both problematic online behaviors were weakly interrelated, suggesting that they serve as distinct constructs. Men's networks appeared to be more complex and had significantly higher global connectivity. Moreover, men and women differed with respect to the specific nodes that bridged both constructs. In men, the bridge nodes were "being criticized because of betting or being told about gambling problems", "loss of previous interests due to gaming", "deceiving other people because of gaming", and "health consequences of gambling". Among women, the bridge nodes were "feeling guilty because of gambling", "loss of previous interests because of gaming", "social consequences of gaming", and "continued gaming problems with other people". In men, the strongest edge was found between "borrowing money/selling anything to gamble" and "financial problems because of gambling", while in women, the strongest edge appeared between "betting more than afforded to be lost" and "tolerance symptoms of gambling". The findings indicate that problematic online gaming and gambling tend to emerge in different ways among men and women. Therapeutic interventions should be planned considering gender differences.

Elucidating the molecular markers and biological pathways associated with extrahepatic cholangiocarcinoma: a transcriptome sequencing study.

Cholangiocarcinoma is a malignancy with high aggressiveness, and extrahepatic cholangiocarcinoma (ECCA) represents the predominant subtype. However, the molecular architecture and underlying pathogenic mechanisms of ECCA remain poorly understood. The objective of this study is to elucidate the molecular markers and biological pathways associated with ECCA. In order to identify the factors influencing ECCA, we conducted transcriptome sequencing on a cohort of 8 surgically resected ECCA specimens. To validate our findings, we integrated data from The Cancer Genome Atlas and Gene Expression Omnibus (GEO) databases using batch integration analysis. Finally, we confirmed our results using clinical samples. The findings of this study reveal that through the analysis of sequencing data, we have successfully identified the genes that are differentially expressed and have a significant role in the development of ECCA. Utilizing the Weighted Gene Co-expression Network Analysis approach, we have integrated these identified gene modules with the GEO dataset, leading to the identification of four key genes (PTGDS, ITIH2, LSAMP, HBB) that are strongly associated with the progression-free survival of ECCA. We screened a key gene LSAMP from four genes using immunohistochemistry. The gene primarily participate in crucial biological processes such as the ECCA cell cycle and DNA replication. The qRT-PCR reaction and Western Blot conducted on the tissues provided confirmation of the expression levels of the gene, which exhibited consistency with the outcomes of our analysis. Our study has successfully identified potential biomarkers LSAMP for ECCA, which can serve as valuable tools for early detection and targeted therapeutic interventions in clinical settings.

Emotion regulation and psychopathology: Investigating differential associations between emotion regulation skills and psychological symptoms using a network approach.

Emotion regulation plays an important role in the development and maintenance of psychopathology. However, the question whether specific ER skills are related to specific psychological symptoms has rarely been studied, but has important implications for targeted interventions. This analysis aims to explore potential differential associations between various ER skills and psychological symptoms using a network analysis approach. Routine data from a transdiagnostic clinical sample of 716 patients (460 females, 256 males) from an outpatient clinic for psychotherapy were analysed. Nine ER skills were assessed with the Emotion Regulation Skills Questionnaire, and nine symptom dimension scores were obtained using the Brief Symptom Inventory. A regularized partial correlation network models including ER skills and individual symptom domains were calculated. Bridge expected influence was calculated to estimate the strength of association of each ER skill with psychological symptoms. Only the following ER skills were most strongly related to psychological symptoms (bridge expected influence): Tolerance, Confrontation and Modification. All other ER skills were indirectly connected to symptom severity through these four skills. The strongest direct edges between ER skills and symptoms were Modification-Depression, Confrontation-Obsession-Compulsion and Tolerance-Interpersonal Sensitivity, which were significantly stronger than the vast majority of other associations between ER skills and psychological symptoms. These exploratory findings provide valuable targets for future studies to investigate specific associations between ER skills and psychological symptoms which could help to improve outcome monitoring and efficacy of interventions targeting ER.

Dynamics of symptom network in patients with first-episode schizophrenia: Insight from the CNFEST project

BackgroundSchizophrenia is a heterogeneous psychotic disorder. Recent theories have emphasized the importance of interactions among psychiatric symptoms in understanding the pathological mechanisms of schizophrenia. In the current study, we examined the symptom network in patients with first-episode schizophrenia (FES) at four time points during a six-month follow-up period. MethodsIn total, 565 patients with FES were recruited from the Chinese First-Episode Schizophrenia Trial (CNFEST) project. Clinical symptoms were measured using the Positive and Negative Syndrome Scale (PANSS) at baseline and follow-up (514 patients at one month, 429 at three months, and 392 at six months). We used a network analysis approach to estimate symptom networks with individual symptoms as nodes and partial correlation coefficients between symptoms as edges. A cross-lagged panel network (CLPN) model was used to identify predictive pathways for clinical symptoms. ResultsWe found stable and strongly connected edges in patients across the time points, such as links between delusions and suspiciousness/persecution (P1:P6), and emotional withdrawal and passive/apathetic social withdrawal (N2:N4). Emotional withdrawal (N2), poor rapport (N3), and passive/apathetic social withdrawal (N4) had high centrality estimates across all four time points. CLPN analysis showed that negative symptoms, including emotional withdrawal (N2), poor rapport (N3), and passive/apathetic social withdrawal (N4), and stereotyped thinking (N7) may have predictive effects for negative and general symptoms at follow-ups. ConclusionsThe symptom network of schizophrenia may be dynamic as treatment progresses. Negative symptoms remain the central and stable symptoms of schizophrenia. Negative symptoms may be potential therapeutic targets that predict other symptoms.

Long-term psychological implication of the COVID-19 pandemic: a network analysis among the Italian population

The enduring psychological impact of the COVID-19 pandemic is a subject of growing concern. Despite existing research highlighting immediate challenges, there is a lack of comprehensive studies elucidating specific pathways through which psychological variables contribute to the pandemic’s long-lasting implications. Through a network analysis approach, we aimed to investigate the mutual relationships and associations between eating disorder symptoms, guilt- and shame-proneness, psychopathological symptoms, self-esteem, defence mechanisms, and mentalization and understand how they might interact two years after the pandemic's outbreak. From March to August 2022, 651 Italian young adults (females = 593, 91.1%) aged between 18 and 38 (Mage = 23.84, SD = 3.74) were recruited online and completed a digital booklet of self-report questionnaires. A weighted, undirected graphical LASSO network was performed to investigate the structure of the nodes. A stable and accurate network structure showed that the most interconnected nodes were psychoticism, interpersonal sensitivity, obsessive-compulsive symptoms, depression, anxiety, paranoid ideation, and immature defences, suggesting their central role in the network functioning. Findings emphasized the enduring psychological impact of the pandemic, elucidating the complex interplay and dynamic impact of key psychopathological symptoms and defence mechanisms. The network tested also provides valuable insights for developing targeted interventions and effective strategies to address post-pandemic mental health challenges.

Shared and distinct cortical morphometric alterations in five neuropsychiatric symptoms of Parkinson’s disease

Neuropsychiatric symptoms (including anxiety, depression, apathy, impulse–compulsive behaviors and hallucinations) are among the most common non-motor features of Parkinson’s disease. Whether these symptoms should be considered as a direct consequence of the pathophysiologic mechanisms of Parkinson’s disease is controversial. Morphometric similarity network analysis and epicenter mapping approach were performed on T1-weighted images of 505 patients with Parkinson’s disease and 167 age- and sex-matched healthy participants from Parkinson’s Progression Markers Initiative database to reveal the commonalities and specificities of distinct neuropsychiatric symptoms. Abnormal cortical co-alteration pattern in patients with neuropsychiatric symptoms was in somatomotor, vision and frontoparietal regions, with epicenters in somatomotor regions. Apathy, impulse–compulsive behaviors and hallucinations shares structural abnormalities in somatomotor and vision regions, with epicenters in somatomotor regions. In contrast, the cortical abnormalities and epicenters of anxiety and depression were prominent in the default mode network regions. By embedding each symptom within their co-alteration space, we observed a cluster composed of apathy, impulse–compulsive behaviors and hallucinations, while anxiety and depression remained separate. Our findings indicate different structural mechanisms underlie the occurrence and progression of different neuropsychiatric symptoms. Based upon these results, we propose that apathy, impulse-compulsive behaviors and hallucinations are directly related to damage of motor circuit, while anxiety and depression may be the combination effects of primary pathophysiology of Parkinson’s disease and psychosocial causes.