Hostname: page-component-6766d58669-7cz98 Total loading time: 0 Render date: 2026-05-21T11:04:28.829Z Has data issue: false hasContentIssue false
  • English
  • Français

Searching for bridges between psychopathology and real-world functioning in first-episode psychosis: A network analysis from the OPTiMiSE trial

Published online by Cambridge University Press:  10 June 2022

Francesco Dal Santo Open the ORCID record for Francesco Dal Santo [Opens in a new window] ,
Eduardo Fonseca-Pedrero ,
María Paz García-Portilla Open the ORCID record for María Paz García-Portilla [Opens in a new window] ,
Leticia González-Blanco ,
Pilar A. Sáiz Open the ORCID record for Pilar A. Sáiz [Opens in a new window] ,
Silvana Galderisi Open the ORCID record for Silvana Galderisi [Opens in a new window] ,
Giulia Maria Giordano  and
Julio Bobes Open the ORCID record for Julio Bobes [Opens in a new window]
Francesco Dal Santo
Affiliation:
Área de Psiquiatría, Universidad de Oviedo, Oviedo, Spain Servicio de Salud del Principado de Asturias (SESPA), Oviedo, Spain Instituto de Investigación Sanitaria del Principado de Asturias (ISPA), Oviedo, Spain Instituto de Neurociencias del Principado de Asturias (INEUROPA), Oviedo, Spain
Eduardo Fonseca-Pedrero
Affiliation:
Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Madrid, Spain Department of Educational Sciences, University of La Rioja, Logroño, Spain
María Paz García-Portilla*
Affiliation:
Área de Psiquiatría, Universidad de Oviedo, Oviedo, Spain Servicio de Salud del Principado de Asturias (SESPA), Oviedo, Spain Instituto de Investigación Sanitaria del Principado de Asturias (ISPA), Oviedo, Spain Instituto de Neurociencias del Principado de Asturias (INEUROPA), Oviedo, Spain Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Madrid, Spain
Leticia González-Blanco
Affiliation:
Área de Psiquiatría, Universidad de Oviedo, Oviedo, Spain Servicio de Salud del Principado de Asturias (SESPA), Oviedo, Spain Instituto de Investigación Sanitaria del Principado de Asturias (ISPA), Oviedo, Spain Instituto de Neurociencias del Principado de Asturias (INEUROPA), Oviedo, Spain Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Madrid, Spain
Pilar A. Sáiz
Affiliation:
Área de Psiquiatría, Universidad de Oviedo, Oviedo, Spain Servicio de Salud del Principado de Asturias (SESPA), Oviedo, Spain Instituto de Investigación Sanitaria del Principado de Asturias (ISPA), Oviedo, Spain Instituto de Neurociencias del Principado de Asturias (INEUROPA), Oviedo, Spain Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Madrid, Spain
Silvana Galderisi
Affiliation:
Department of Psychiatry, University of Campania “Luigi Vanvitelli”, Naples, Italy
Giulia Maria Giordano
Affiliation:
Department of Psychiatry, University of Campania “Luigi Vanvitelli”, Naples, Italy
Julio Bobes
Affiliation:
Área de Psiquiatría, Universidad de Oviedo, Oviedo, Spain Servicio de Salud del Principado de Asturias (SESPA), Oviedo, Spain Instituto de Investigación Sanitaria del Principado de Asturias (ISPA), Oviedo, Spain Instituto de Neurociencias del Principado de Asturias (INEUROPA), Oviedo, Spain Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Madrid, Spain
*
*Author for correspondence: María Paz García-Portilla, E-mail: albert@uniovi.es

Abstract

Background

Network analysis has been used to explore the interplay between psychopathology and functioning in psychosis, but no study has used dedicated statistical techniques to focus on the bridge symptoms connecting these domains. The current study aims to estimate the network of depressive, negative, and positive symptoms, general psychopathology, and real-world functioning in people with first-episode schizophrenia or schizophreniform disorder, focusing on bridge nodes.

Methods

Baseline data from the OPTiMiSE trial were analyzed. The sample included 446 participants (age 40.0 ± 10.9 years, 70% males). The network was estimated with a Gaussian graphical model, using scores on individual items of the positive and negative syndrome scale (PANSS), the Calgary depression scale for schizophrenia, and the personal and social performance scale. Stability, strength centrality, expected influence (EI), predictability, and bridge centrality statistics were computed. The top 20% scoring nodes on bridge strength were selected as bridge nodes.

Results

Nodes from different rating scales assessing similar psychopathological and functioning constructs tended to cluster together in the estimated network. The most central nodes (EI) were Delusions, Emotional Withdrawal, Depression, and Depressed Mood. Bridge nodes included Depression, Conceptual Disorganization, Active Social Avoidance, Delusions, Stereotyped Thinking, Poor Impulse Control, Guilty Feelings, Unusual Thought Content, and Hostility. Most of the bridge nodes belonged to the general psychopathology subscale of the PANSS. Depression (G6) was the bridge node with the highest value.

Conclusions

The current study provides novel insights for understanding the complex phenotype of psychotic disorders and the mechanisms underlying the development and maintenance of comorbidity and functional impairment after psychosis onset.

Keywords

first-episode psychosisnetwork analysispsychopathologyreal-world functioningschizophrenia

Information

Type
Research Article
Information
European Psychiatry , Volume 65 , Issue 1 , 2022 , e33
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2022. Published by Cambridge University Press on behalf of the European Psychiatric Association

Introduction

Psychotic disorders are severe, complex, multifactorial mental disorders characterized by heterogeneous psychopathological features. These include positive, negative, cognitive, and affective symptoms, and disorganized behaviors [Reference Maj, van, De, Gaebel, Galderisi and Green1, Reference Owen, Sawa and Mortensen2]. None of these is pathognomonic of the illness, and individuals can present varying degrees of severity in the different symptomatologic areas [Reference Krynicki, Upthegrove, Deakin and Barnes3]. Several studies, for example, have highlighted the importance of mood symptoms, given their high prevalence in both prodromal and clinical psychosis [Reference Upthegrove, Marwaha and Birchwood4]. In fact, it has been estimated that up to 80% of individuals experience a clinically significant depressive episode during the early phase of the mental disorder [Reference Upthegrove, Marwaha and Birchwood4]. While previous models have conceptualized depressive and psychotic symptoms as distinct clinical features of the mental disorder, existing research has shown difficulties differentiating between these symptomatologic domains [Reference van Rooijen, Isvoranu, Kruijt, van Borkulo, Meijer and Wigman5], highlighting their close relationship, as in the case of negative symptoms and depression [Reference Krynicki, Upthegrove, Deakin and Barnes3].

Moreover, despite significant advances in pharmacological and psychological treatments, psychotic disorders still rank among the leading causes of disability worldwide, with undeniably substantial burdens for the affected individuals and their families and caregivers [Reference Galderisi, Rossi, Rocca, Bertolino, Mucci and Bucci6], as well as for the health and welfare systems, especially in younger age groups [Reference Charlson, Ferrari, Santomauro, Diminic, Stockings and Scott7]. However, functional impairment still represents a challenge for clinicians and researchers, this having also been suggested as a resistance criterion by some authors [Reference Bozzatello, Bellino and Rocca8]. A recent meta-analysis found a functional recovery rate of 38% in first-episode psychosis (FEP) and observed a tendency toward stabilization after the first 2 years of illness [Reference Lally, Ajnakina, Stubbs, Cullinane, Murphy and Gaughran9]. Different areas of functioning can be affected. For example, social functioning is often impaired, as people with this disorder tend to interact less with others or do so in a socially inappropriate way, which may reduce the willingness of others to engage with them [Reference Maj, van, De, Gaebel, Galderisi and Green1]. Thus, identifying the predictors of poor functioning should be a priority and could inform the development of new targets to better assess and manage individuals with FEP. However, existing research has shown heterogeneous results in this population, suggesting that functional outcomes could be related to diverse psychopathological characteristics [Reference Santesteban-Echarri, Paino, Rice, Gonzalez-Blanch, McGorry and Gleeson10].

In this context, emerging network analysis techniques, which aim to suggest new ways of modeling and understanding psychopathological processes [Reference Fonseca-Pedrero11–Reference Borsboom13], could help disentangle the complex and dynamic interplay between symptoms and functional outcomes. This approach focuses on conceptualizing symptoms as mutually interacting and often reciprocally reinforcing elements of a complex network rather than interpreting them as a function of a latent disorder [Reference Fonseca-Pedrero11, Reference Borsboom and Cramer14, Reference Borsboom, Cramer and Kalis15]. So, from this perspective, mental disorders are presumed to arise from direct interactions between symptoms in a network architecture [Reference Borsboom, Cramer and Kalis15].

One of the main reasons for the increasing popularity of these theories is the inadequacy of the traditional categorical approach, which has led to a simplified and incomplete vision of mental problems [Reference Fonseca-Pedrero11]. Moreover, network techniques offer multiple methodological advantages over other methods, including the possible verification of simultaneous relations among variables [Reference Izquierdo, Cabello, Leal, Mellor-Marsá, Ayora and Bravo-Ortiz16] and the lack of need for a priori assumptions regarding relationships among the variables or the selection of predictors, mediators, and outcome measures [Reference Galderisi, Rucci, Kirkpatrick, Mucci, Gibertoni and Rocca17]. Furthermore, this paradigm shift may also have important clinical implications and could promote a personalized and integrated approach to the treatment of psychosis [Reference Galderisi, Rucci, Kirkpatrick, Mucci, Gibertoni and Rocca17]. For example, it could reveal novel therapeutic targets (e.g., influential nodes susceptible to deactivation) that might otherwise go unnoticed with traditional methodology and thus be neglected by clinicians in a real-life setting.

It therefore comes as no surprise that, in recent years, network analysis has been applied to the study of correlates of functioning in both established schizophrenia [Reference Galderisi, Rucci, Kirkpatrick, Mucci, Gibertoni and Rocca17, Reference Galderisi, Rucci, Mucci, Rossi, Rocca and Bertolino18] and FEP [Reference Izquierdo, Cabello, Leal, Mellor-Marsá, Ayora and Bravo-Ortiz16, Reference Chang, Wong, Or, Chu, Hui and Chan19, Reference Izquierdo, Cabello, la Torre-Luque, Ayesa-Arriola, Setien-Suero and Mayoral-van-Son20]. Using this approach, Galderisi and colleagues highlighted the critical role of real-world functioning, primarily the everyday life skills domain, in a sample of community-dwelling individuals with schizophrenia, where functioning nodes were linked to different clinical correlates such as disorganization, expressive deficits, and avolition [Reference Galderisi, Rucci, Kirkpatrick, Mucci, Gibertoni and Rocca17]. While the overall network structure of the sample was similar at the 4-year follow-up, a further analysis revealed a very sparse network, with real-life functioning disconnected from other nodes in the recovered subgroup [Reference Galderisi, Rucci, Mucci, Rossi, Rocca and Bertolino18], emphasizing the dynamic nature of these interactions and their relevance to clinical and functional recovery. Regarding FEP, Chang et al. [Reference Chang, Wong, Or, Chu, Hui and Chan19] found that psychosocial functioning was strongly associated with amotivation, moderately with positive symptoms, and only weakly with other psychopathological variables. However, the study sample was constituted of participants aged 26–55 years, making these findings less generalizable to first-episode cohorts with younger age of illness onset. On the other hand, another study recently observed several connections between functioning problems and psychopathology, including hallucinations, conceptual disorganization, and depression [Reference Izquierdo, Cabello, la Torre-Luque, Ayesa-Arriola, Setien-Suero and Mayoral-van-Son20].

Nonetheless, generalizability of previous research is problematic, mainly due to methodological differences. Some studies, for example, computed the network by introducing the total of all scores on rating scales (or all subscale scores) [Reference Galderisi, Rucci, Kirkpatrick, Mucci, Gibertoni and Rocca17–Reference Chang, Wong, Or, Chu, Hui and Chan19], limiting an examination of the role played by individual symptoms in determining functional impairment. Conversely, despite their use of individual items, other authors did not introduce the full list of PANSS items or a specific instrument for depression [Reference Izquierdo, Cabello, Leal, Mellor-Marsá, Ayora and Bravo-Ortiz16]. Also, these studies included participants from single nations. This could be a limitation because, for example, different cultural contexts may vary in the degree to which they accept particular symptoms (e.g., hallucinations and magical thinking) as normative experiences [Reference Fonseca-Pedrero, Chan, Debbané, Cicero, Zhang and Brenner21].

Moreover, previous network analysis studies focusing on functional outcomes have not made use of dedicated statistical procedures to identify so-called “bridge nodes” [Reference Jones, Ma and McNally22]. These nodes represent the key connection points between different groups of nodes in a network (e.g., groups of nodes that belong to specific psychopathological or functioning domains) and deserve special attention for their possible contribution to the onset and maintenance of comorbid conditions in mental disorders [Reference Cramer, Waldorp, Van Der Maas and Borsboom12]. From a translational perspective, bridge nodes might be used to develop targeted interventions, and deactivating symptoms based on their bridge strength rather than on other centrality measures may constitute an effective strategy to prevent comorbidity [Reference Jones, Ma and McNally22].

Thus, the purposes of the current study are (a) to use a network approach to shed light on the interplay among depressive, negative, and positive symptoms, general psychopathology, and deficits in personal, social, and occupational functioning in people with first-episode schizophrenia or schizophreniform disorder and (b) to statistically identify the bridge nodes of the estimated network.

Method

Study design and participants

This study is a secondary analysis of data from the Optimization of Treatment and Management of Schizophrenia in Europe (OPTiMiSE) trial, for which a detailed description of the rationale and methodology can be found elsewhere [Reference Leucht, Winter-van Rossum, Heres, Arango, Fleischhacker and Glenthøj23].

Individuals with FEP were recruited at the participating centers, which included 27 general hospitals and psychiatric specialty clinics in 14 European countries (Austria, Belgium, Bulgaria, Czech Republic, Denmark, France, Germany, Italy, the Netherlands, Poland, Romania, Spain, Switzerland, and the UK) and Israel. Eligible participants aged 18–40 years who met the Diagnostic and Statistical Manual of Mental Disorders (4th edition) criteria for schizophrenia, schizophreniform disorder, or schizoaffective disorder were recruited. Diagnoses were confirmed by the Mini-International Neuropsychiatric Interview-Plus [Reference Sheehan, Lecrubier, Sheehan, Amorim, Janavs and Weiller24].

They were excluded if: (a) more than 2 years had elapsed between the onset of psychosis and enrollment; (b) they had been treated with any antipsychotic medication for more than 2 weeks in the previous year or a total of 6 weeks or more lifetime; (c) they had a previous history of intolerance to one of the study drugs; (d) they met any of the contraindications for any of the study drugs; (e) they were coercively treated or represented by a legal guardian, or both, or in legal custody; or (f) they were pregnant or breastfeeding at the study time.

This study was conducted in accordance with the ethical principles of the Declaration of Helsinki, and each country obtained ethics approval. All participants received information about the purposes and protocol of the study and signed the informed consent before any procedures were performed.

For the current analysis, we employed data collected as part of phase 1 of the original study (open label amisulpride treatment at a daily dose of 200–800 mg), in which a total of 446 participants were enrolled from an initial sample of 481 subjects who were assessed for eligibility and signed an informed consent [Reference Kahn, van, Leucht, McGuire, Lewis and Leboyer25].

Measures

After completing an initial screening visit to assess eligibility, baseline data were collected, including sociodemographic variables, diagnoses, current treatments, and rating scales.

The psychopathological assessment included the positive and negative syndrome scale (PANSS) [Reference Kay, Fiszbein and Opler26] to characterize psychotic symptoms and the Calgary Depression Scale for Schizophrenia (CDSS) [Reference Addington, Addington and Schissel27] to assess depression. The clinical global impression-schizophrenia [Reference Guy28] was used to evaluate severity of illness.

Finally, real-world functioning was assessed with the personal and social performance (PSP) scale [Reference Morosini, Magliano, Brambilla, Ugolini and Pioli29]. The PSP scale is a clinician-rated instrument that evaluates four areas of functioning: (a) Socially Useful activities, (b) Personal and Social Relationships, (c) Self-care, and d) Disturbing and Aggressive behavior. These subscores range from 0 to 6, where higher scores indicate worse functioning. A total score (ranging from 0 to 100) is also calculated, with higher scores corresponding to better personal and social functioning.

Data analyses

First, the sociodemographic characteristics of the sample and the descriptive statistics of all measures used were analyzed, expressing the results with means, standard deviations (SD), and percentages.

Second, the network of psychosis phenotype, depression symptoms, and functional outcomes was estimated. A Gaussian graphical model [Reference Epskamp and Fried30] was used for this purpose. In this network, the scores on the individual items of the instruments were used instead of the scale or subscale total scores. Therefore, the analysis was performed with a total of 43 nodes: the 30 items of the PANSS, the nine items of the CDSS, and the four items of the PSP.

The details of network analysis have been documented in previous publications [Reference Epskamp, Borsboom and Fried31, Reference Epskamp, Cramer, Waldorp, Schmittmann and Borsboom32]. A network consists of nodes (study variables, such as item scores on each measurement instrument) and edges (estimated statistical relationships among variables). We used partial correlations: if two nodes are connected in the resulting graph via an edge, they are statistically related after controlling for all other variables in the network; if they are unconnected, they are conditionally independent. The least absolute shrinkage and selection operator (LASSO) procedure was used to limit the number of spurious connections among nodes [Reference Epskamp and Fried30]. This regularization method applies a penalty to small edges, shrinking them to zero and thus dropping them out of the model, and returning a network model that is more stable and easier to interpret. For the layout, we used the Fruchterman–Reingold algorithm, placing the strongly connected nodes closer to each other and the less connected nodes far apart [Reference Epskamp, Cramer, Waldorp, Schmittmann and Borsboom32].

In keeping with previous studies examining networks [Reference Fonseca-Pedrero, Ortuño, Debbané, Chan, Cicero and Zhang33], we estimated two inference measures: strength centrality and expected influence (EI). Strength centrality is the sum of the correlations of one node to all other nodes of the network. High values reflect great centrality of the node in the network. EI identifies the most important nodes within a network graph [Reference Robinaugh, Millner and McNally34]. We used EI along with strength centrality in order to avoid possible problems associated with centrality measures [Reference Haslbeck and Fried35, Reference Opsahl, Agneessens and Skvoretz36]. It is noteworthy that strength centrality uses the sum of absolute weights (i.e., negative edges are turned into positive edges before summing), so the interpretation could be distorted if negative edges are present (as in the present article). On the other hand, EI takes into account negative associations among nodes and can assume negative values. If the EI value of a node is negative, changes in the node should produce network changes in the opposite direction (e.g., decreases in node activation should lead to increases in overall network activation) [Reference Robinaugh, Millner and McNally34].

Network stability and accuracy were estimated using the bootstrapping analysis in the R bootnet package [Reference Epskamp, Borsboom and Fried31].

Finally, further analyses were performed to identify the bridge nodes. Two bridge centrality statistics (bridge strength and bridge betweenness) were estimated with the bridge function in the networktools package [Reference Jones37]. We selected the top 20% scoring nodes on bridge strength as bridge nodes, following the methods from previous research [Reference Jones, Ma and McNally22].

We used IBM SPSS Statistics for Windows, Version 22.0 [38], JASP (https://jasp-stats.org/), and R [39] to perform data analyses.

Results

Participants and descriptive statistics

A total of 446 participants enrolled and started phase 1. Sociodemographic, clinical, and psychometric characteristics of the sample as well as the descriptive statistics of all measures are depicted in Table 1.

Table 1.

Sociodemographic data and clinical and functional assessment of the sample (n = 446).

Note: Data are expressed as mean (SD) or n (%).

Abbreviations: CDSS, Calgary Depression Scale for Schizophrenia; CGI-SCH, clinical global impression-schizophrenia; GP, general psychopathology; PSP: Personal and Social Performance Scale; PANSS, Positive and Negative Syndrome Scale.

Network structure of psychosis phenotype

The estimated psychosis network showed a high degree of interconnectedness between nodes (see Figure 1). Most of the associations between edges were positive (see Figure 1).

Figure 1.

Estimated network for psychosis phenotype, depression symptoms, and real-life functioning. CDSS: Calgary Depression Scale for schizophrenia; G: PANSS, general psychopathology dimension; N: PANSS, negative psychosis dimension; PSP: personal and social performance; P: PANSS, positive psychosis dimension. Numbers represent item numbers in the scale; blue edges represent positive associations; red edges represent negative associations. Thickness and saturation of edges indicate the strength of these associations.

Strength centrality and standardized EI values are depicted in Figure 2. The most central nodes in terms of strength were Depression (G6), Delusions (P1), Emotional Withdrawal (N2), and Depressed Mood (CDSS1). The most central nodes in terms of EI were Delusions (P1), Emotional Withdrawal (N2), Depression (G6), Depressed Mood (CDSS1), and Poor Rapport (N3). Somatic Concern (G1), Early Wakening (CDSS7), Grandiosity (P5), and Disorientation (G10) showed a negative value in this centrality index.

Figure 2.

Inference measures of the estimated psychosis network. CDSS: Calgary Depression Scale for Schizophrenia; G: PANSS, general psychopathology dimension; N: PANSS, negative psychosis dimension; PSP: personal and social performance; P: PANSS, positive psychosis dimension. Numbers represent item numbers in the scale.

The adaptive LASSO network showed that items from different rating scales but assessing similar psychopathological and functioning constructs tended to cluster together (see Figure 1).

Depression symptoms formed a cluster of nodes with a high degree of interconnectedness, with the exception of the Early Wakening node (CDSS7), which showed a direct connection with only the Morning Depression item (CDSS6) of the CDSS. In addition, the CDSS nodes were positively and closely related to the Guilty Feelings (G3) and Depression (G6) nodes of the PANSS general psychopathology subscale and not clearly related to psychosis symptom nodes. Furthermore, these nodes were separate from psychosocial functioning as measured with the PSP.

Regarding psychosis phenotype symptoms as measured with the PANSS, the items of the negative dimension tended to be consistently more strongly related compared with the items that make up the positive dimension, revealing a cluster made up of the Blunted Affect (N1), Emotional Withdrawal (N2), Poor Rapport (N3), Passive/Apathetic Withdrawal (N4), Lack of Spontaneity, and Flow of Conversation (N6) nodes. The psychotic positive cluster included Delusions (P1), Hallucinatory Behavior (P3), and Suspiciousness/Persecution (P6). Mixed results were found for the items of the PANSS general psychopathology domain, scattered throughout the network and associated with different psychopathological domains. This is exemplified by the Unusual Thought Content (G9) and Motor retardation (G7) nodes, which were closely related to the positive and negative groups of nodes, respectively. On the other hand, disorganized symptoms congregated to form a cluster of items from the three PANSS subscales, which included the Conceptual Disorganization (P2), Difficulty in Abstract Thinking (N5), Stereotyped Thinking (N7), Preoccupation (G15), Disturbance of Volition (G13), Poor Attention (G11), Disorientation (G10), and Mannerisms and Posturing (G5) nodes. Lack of Insight (G12) fluctuated between positive and disorganized domains.

With respect to real-world functioning nodes, the network depicted a clear separation between two dimensions. The first one was related to psychosocial functioning and incorporated Socially Useful Activities (PSP-A), Personal and Social Relationships (PSP-B), and Self-care (PSP-C). This group of nodes maintained connections with both the positive and disorganized clusters of the network. By contrast, the Disturbing and Aggressive Behavior item of the PSP (PSP-D) was strongly associated with the Hostility (P7) and Poor Impulse Control (G14) nodes.

Bridge nodes and bridge centrality measures

Bridge centrality statistics (bridge strength and bridge betweenness) are reported in Figure 3.

Figure 3.

Bridge centrality measures of the estimated psychosis network. CDSS: Calgary Depression Scale for Schizophrenia; G: PANSS, general psychopathology dimension; N: PANSS, negative psychosis dimension; PSP: personal and social performance; P: PANSS, positive psychosis dimension. Numbers represent item numbers in the scale.

The top 20% scoring nodes on bridge strength were Depression (G6), Conceptual Disorganization (P2), Active Social Avoidance (G16), Delusions (P1), Stereotyped Thinking (N7), Poor Impulse Control (G14), Guilty Feelings (G3), Unusual Thought Content (G9), and Hostility (P7). Most of the bridge nodes belonged to the general psychopathology subscale of the PANSS (see Figure 1).

Depression (G6) and Guilty Feelings (G3) represented the bridge between depressive symptoms and the rest of the network, especially through the shared connection with the Anxiety node (G2), which showed the highest value of bridge betweenness. Hostility (P7) and Poor Impulse Control (G14) nodes constituted the bridge between the positive symptoms cluster and the Disturbing and Aggressive Behavior item of the PSP (PSP-D). Delusions (P1), the most central node in terms of EI, emerged as a bridge between the positive dimension and psychosocial functioning. Active Social Avoidance (G16) connected the positive and negative dimensions and also showed a weaker link with Personal and Social Relationships (PSP-B), while Unusual Thought Content (G9) linked the positive dimension with the disorganized symptoms. The latter group also formed bridges with psychosocial functioning and the negative dimensions through Conceptual Disorganization (P2) and Stereotyped Thinking (N7).

Network stability and accuracy analysis

The results of the stability and accuracy analysis [Reference Epskamp, Borsboom and Fried31] indicated that the psychosis network was accurately estimated, with adequate confidence intervals around the edge weights. Details are accessible in Supplementary Figures S1 and S2.

Discussion

The present study was designed to explore the interplay between psychosis phenotype, general psychopathology, depression, and real-world functioning in people with FEP and represents one of the first attempts to assess these interactions at symptom level in this clinical population. Novel network analysis techniques, such as identification of bridge nodes, were employed to identify central nodes. The analysis yielded a stable and accurately estimated network.

Structure of the network

The topological structure of this network offers several interesting findings and highlights the complex nature of psychopathology in FEP. One of the main findings of our analysis is that the items belonging to different rating scales but assessing similar psychopathological constructs tend to cluster together in a manner not always consistent with the original structure of the scales. This is especially evident in the case of the PANSS: while the nodes from the CDSS tended to form a well-connected group, the items from the Positive, Negative, and General Psychopathology subscales sparsely mixed with each other or with the other items of the network. Taken together, these results seem to indicate that this three-dimensional solution may prove to be unsatisfactory when describing psychopathology in people with FEP. In this sense, it is interesting to observe a partial overlap of the clusters of nodes identified in our network with earlier five-factor solutions [Reference Wallwork, Fortgang, Hashimoto, Weinberger and Dickinson40]. However, those approaches produced very few models with acceptable fits, making it advisable to explore new methods [Reference Higuchi, Cogo-Moreira, Fonseca, Ortiz, Correll and Noto41]. In this regard, the emergence and refining of network analysis techniques could offer an additional tool to capture the dimensional nature of psychotic symptoms, focusing not only on identification of symptom clusters but above all on their connecting (and, by extension, activating) patterns.

It is somewhat surprising that our findings are contrary to other network analyses, which have observed that nodes that belong to the same subscale are highly interconnected [Reference van Rooijen, Isvoranu, Kruijt, van Borkulo, Meijer and Wigman5, Reference Amore, Murri, Calcagno, Rocca, Rossi and Aguglia42, Reference Griffiths, Leighton, Mallikarjun, Blake, Everard and Jones43]. However, it should be noted that a direct comparison of our findings with those of previous studies is not feasible in most of the cases due to methodological differences. First, some of the existing studies included the total score (or subscale total scores) of the rating scales instead of computing the network via individual symptoms [Reference Galderisi, Rucci, Kirkpatrick, Mucci, Gibertoni and Rocca17–Reference Chang, Wong, Or, Chu, Hui and Chan19, Reference Piao, Yun, Nguyen, Kim, Sui and Kang44]. Despite considering individual symptoms as network nodes, other authors did not include general psychopathology in their analyses [Reference van Rooijen, Isvoranu, Kruijt, van Borkulo, Meijer and Wigman5, Reference Griffiths, Leighton, Mallikarjun, Blake, Everard and Jones43] or selected a reduced number of items from the PANSS general psychopathology subscale [Reference Izquierdo, Cabello, Leal, Mellor-Marsá, Ayora and Bravo-Ortiz16, Reference Amore, Murri, Calcagno, Rocca, Rossi and Aguglia42]. Interestingly, a recent network analysis of a sample of individuals at clinical high risk for psychosis or with recent-onset psychosis highlighted the importance of general psychopathology as a potential trigger of the pathway from negative life events to the expression of psychotic symptomatology [Reference Betz, Penzel, Kambeitz-Ilankovic, Rosen, Chisholm and Stainton45], while other authors have found general psychopathology subscale score to be the node with the highest strength [Reference Piao, Yun, Nguyen, Kim, Sui and Kang44]. Therefore, based on our findings, we recommend including the PANSS general psychopathology subscale in network models, along with the positive and negative subscales. We believe that this approach could prove useful to better capture the diverse psychopathological presentation of FEP beyond the positive/negative symptom duality.

Regarding functioning, the high degree of interconnectedness between Socially Useful Activities, Personal and Social Relationships, and Self-care found in our network analysis is consistent with earlier findings [Reference Galderisi, Rucci, Kirkpatrick, Mucci, Gibertoni and Rocca17, Reference Galderisi, Rucci, Mucci, Rossi, Rocca and Bertolino18]. However, in contrast, the functioning nodes were not the most important in our network in terms of strength centrality and standardized EI values, supporting evidence from previous observations in FEP samples [Reference Izquierdo, Cabello, Leal, Mellor-Marsá, Ayora and Bravo-Ortiz16, Reference Chang, Wong, Or, Chu, Hui and Chan19]. A possible explanation for these results may lie in the differences in the samples studied. First, the former studies by Galderisi and colleagues were not conducted in FEP, suggesting that the network structure could change over the course of the illness. Moreover, they focused on stabilized community-dwelling individuals who could still exhibit significant functional impairment despite resolution of the acute symptom exacerbation, resulting in the lower relative importance of the psychotic symptoms in their network. On the other hand, the Disturbing and Aggressive Behaviors node was associated with other symptoms, namely Hostility (P7) and Poor Impulse Control (G14), and seems to be better conceptualized as a separate and specific construct.

Bridges nodes

Another initial objective of the project was to identify bridge nodes using a specific statistical analysis. If we consider the network a dynamic representation of psychopathology [Reference Borsboom13] and its impact on functioning, specific interventions targeting these bridge nodes could lower the degree of activation of other nodes and therefore lead to better overall outcomes. Given that the majority of the studies relied on visual inspection rather than more refined statistical techniques to identify these key symptoms, the current study provides novel insights for understanding the complex phenotype of psychotic disorders and the mechanisms underlying the development and maintenance of comorbidity and functional impairment after psychosis onset.

When examining the bridge strength measure, we found that Depression (G6) was the bridge node with the highest value. This node, which was also one of the most important in terms of EI, connected the symptoms of the CDSS to the rest of the network along with Guilty Feelings (G3), and their bridge function acted through an interesting shared pathway via the Anxiety node (G2). The significance of these nodes in our network is not surprising, as several reports have stressed the high rates of depression in schizophrenia, especially during the early phase of the disorder [Reference Upthegrove, Marwaha and Birchwood4, Reference Herniman, Allott, Phillips, Wood, Uren and Mallawaarachchi46], with potential serious outcomes such as suicidal behaviors [Reference Canal-Rivero, López-Moríñigo, Setién-Suero, Ruiz-Veguilla, Ayuso-Mateos and Ayesa-Arriola47]. In keeping with the present results, a recent longitudinal study in an FEP sample found that depression was one of the most important nodes in their network [Reference Griffiths, Leighton, Mallikarjun, Blake, Everard and Jones43] and further revealed that it maintained its central role at the 12-month follow-up despite the amelioration of psychotic symptoms [Reference Herniman, Allott, Phillips, Wood, Uren and Mallawaarachchi46]. The topographic proximity of depressive to positive symptoms in our network is also consistent with previous findings [Reference van Rooijen, Isvoranu, Kruijt, van Borkulo, Meijer and Wigman5, Reference Griffiths, Leighton, Mallikarjun, Blake, Everard and Jones43, Reference Herniman, Phillips, Wood, Cotton, Liemburg and Allott48] and constitutes an alert that potential worsening of the affective domain could translate to a flare up of psychotic symptoms [Reference Griffiths, Leighton, Mallikarjun, Blake, Everard and Jones43] and vice versa. However, their networks did not include general psychopathology, so the present study raises the possibility that other nodes may underlie these relationships. Anxiety (G2), which showed the highest value of bridge betweenness in our network and is considered among the least studied features of schizophrenia [Reference Buonocore, Bosia, Baraldi, Bechi, Spangaro and Cocchi49], offers a clear example of a potential treatment target to avoid the spread of contagion between affective symptoms and the psychotic dimension.

On the other side of the network, Delusions (P1), unsurprisingly the most important node of our estimated network, showed a double connection with Personal and Social Relationships (PSP-B) as an indirect pathway through Active Social Avoidance (G16) emerged, in addition to the direct link between P1 and PSP-B. Moreover, Active Social Avoidance (G16) mediated the pathway from positive to negative symptom clusters. This combination of findings may have different clinical implications. First, due to its influence on the network and its connecting pattern, a worsening of delusions may potentially lead to an overall activation of the network, directly reinforcing other psychopathological nodes as well deteriorating functional abilities. Eradicating positive symptomatology, for example, by following evidence-based treatment schemes [Reference Kahn, van, Leucht, McGuire, Lewis and Leboyer25], seems therefore crucial to prevent inter-symptom contagion. Second, the impact of positive symptoms on functioning appears to be partly related to a specific subtype of social avoidance, which also bridges the path between positive and negative symptoms. A double reading can be derived from these findings: while early intervention on positive symptoms may prevent social isolation and, by extension, avoid the emergence of negative symptoms, if the aforementioned pathways are followed in reverse, promoting better social interactions, this could facilitate clinical recovery in FEP, as suggested by previous research [Reference Bjornestad, Hegelstad, Joa, Davidson, Larsen and Melle50]. With this in mind, clinicians should carefully distinguish active social avoidance from passive/apathetic social withdrawal, as it may require other kinds of interventions [Reference Hansen, Torgalsbøen, Melle and Bell51, Reference Galderisi, Kaiser, Bitter, Nordentoft, Mucci and Sabé52], for example, targeting theory of mind processes with metacognitively oriented psychotherapies [Reference Phalen, Dimaggio, Popolo and Lysaker53].

With respect to the bridge nodes linking psychosocial functioning with the disorganized dimension, our findings support the existing literature on applying the network approach, which stresses the impact of disorganized symptoms on functional outcomes in individuals with established schizophrenia [Reference Galderisi, Rucci, Kirkpatrick, Mucci, Gibertoni and Rocca17] and highlights the prominent role of conceptual disorganization as a connector between functioning and clinical symptoms in FEP [Reference Izquierdo, Cabello, Leal, Mellor-Marsá, Ayora and Bravo-Ortiz16]. Interestingly, another study previously found that conceptual disorganization could have an impact on community activities up to twice as high as core symptoms such as delusions and avolition [Reference Rocca, Galderisi, Rossi, Bertolino, Rucci and Gibertoni54]. Moreover, both Conceptual Disorganization (P2) and Stereotyped Thinking (N7) were recently signaled as key bridge nodes in a network from a sample of people with FEP [Reference Griffiths, Leighton, Mallikarjun, Blake, Everard and Jones43], although that study did not include a functional assessment. Hence, while formal thought disorder risk may be under-recognized during the first stages of illness, its early detection could offer a novel treatment target [Reference Griffiths, Leighton, Mallikarjun, Blake, Everard and Jones43] with potential beneficial effects on both clinical and functional recovery.

In the case of the Disturbing and Aggressive Behaviors node (PSP-D), our network model revealed several direct and indirect connections with a broad range of psychosis symptoms on the PANSS subscales. This reflects the complex interplay between psychopathology and aggressive behavior in early-stage psychosis and corroborates the findings of existing research studies, which found relationships with delusions [Reference Coid, Ullrich, Kallis, Keers, Barker and Cowden55], uncooperativeness [Reference Faay and van Os56], impulsivity [Reference Moulin, Golay, Palix, Baumann, Gholamrezaee and Azzola57], excitement [Reference Witt, van Dorn and Fazel58], lack of insight [Reference Witt, van Dorn and Fazel58], and negative symptoms [Reference Lopez-Garcia, Ashby, Patel, Pierce, Meyer and Rosenthal59], among others. According to the present results, a comprehensive assessment and management of these symptoms could facilitate prevention strategies and directly or indirectly reduce the risk of aggression due to diminished Poor Impulse Control (G14) and Hostility (P7). These findings are important not only in light of the common occurrence of disturbing and aggressive behaviors in people with FEP but also because of the drop-off in aggression rates observed in individuals with regular follow-up by mental health services after a first psychotic episode [Reference Lopez-Garcia, Ashby, Patel, Pierce, Meyer and Rosenthal59, Reference Faay, van Baal, Arango, Díaz-Caneja, Berger and Leucht60]. However, it should be mentioned that the study sample was characterized by low rates of disturbing and aggressive behavior, as discussed elsewhere [Reference Faay, van Baal, Arango, Díaz-Caneja, Berger and Leucht60].

Finally, it is worth mentioning that, to the best of our knowledge, our study provides the first cross-national network model of the inter-relationships among psychopathology, depression, and functioning in FEP. Although this collaborative effort could translate into better overall generalizability of the current findings, it should also be noted that specific social, cultural, and historical aspects may play a modulating role and result in symptom networks differing partially across place and time [Reference Borsboom, Cramer and Kalis15]. For instance, differences between countries have been observed in previous multinational network analyses of schizotypal personality traits [Reference Fonseca-Pedrero, Ortuño, Debbané, Chan, Cicero and Zhang33]. Different cultural contexts may also vary in identifying particular symptoms as anomalous or unusual [Reference Fonseca-Pedrero, Chan, Debbané, Cicero, Zhang and Brenner21], possibly determining, for example, a different impact on the level of social functioning of young people with FEP. However, the characteristics and the size of the sample limited our possibility of performing in-depth country-based analyses. Nevertheless, since the vast majority of the participating centers are located in European countries that share a similar cultural and social background, we would not expect substantial differences in this regard. More research is therefore warranted to address this issue, especially longitudinal cross-cultural studies that could provide further knowledge in the field of psychosis about the impact of cultural and social environment on the dynamic evolution of these interactions over time.

Strengths and limitations

It is worth pointing out some strengths of our study, such as the size of the collected sample, the multicenter nature of the study, and the wide-ranging psychopathological and functional variables assessed. Moreover, the strict inclusion criteria of the trial allow the control of a series of confounders. Furthermore, the network model was designed to consider individual symptoms as nodes, the stability analyses indicate that this network is accurately estimated, and novel inference measures such as EI and predictability were computed. Finally, bridge centrality statistics were calculated to identify bridge nodes, instead of relying on visual inspection as previous studies did. Also, while previous research has mainly employed bridge analysis to explore comorbidity between different mental disorders, the current study focused on various aspects of the phenotype of psychotic disorders.

However, the reader should be aware of some limitations when interpreting the current findings. Selection bias is perhaps the main potential concern, as participants were recruited to take part in a randomized controlled trial. For example, the exclusion of individuals coercively treated or represented by a legal guardian, who could represent a group with worse overall symptom severity and functioning, could limit the generalizability of the current findings. Also, the cross-sectional analysis performed suffers from well-known drawbacks. Furthermore, the clinical evaluation lacked an intelligence quotient evaluation, as well as a dedicated cognitive assessment, although the direct impact of cognition on real-world functioning is at present still a matter of debate.

Conclusions

The findings of this investigation indicate that it does not seem appropriate to simplify and compartmentalize the psychopathological presentation of FEP using the original structure of the three PANSS subscales and that innovative techniques such as network analysis, focusing on the interactions between individual symptoms, could prove useful to better capture the complex interplay between symptoms and functional outcomes.

The current research also statistically identified several bridge nodes, whose deactivation could inhibit the cascade of self-reinforcing interactions between symptoms and functioning nodes in the estimated network. Although interventions on positive symptoms are crucial in FEP, we stress the importance of other symptoms such as anxiety, depression, disorganization, impulsivity, active social avoidance, and social relationships. Too often neglected by clinicians, these bridge nodes need to be carefully assessed during the early stages of the illness and should constitute preferential intensive treatment targets to reduce comorbidity and preserve real-world functioning. In accordance with the network theory [Reference Borsboom, Cramer and Kalis15], tailored interventions targeting these nodes should involve multidisciplinary approaches, including pharmacological and psychosocial strategies.

Acknowledgment

The authors wish to thank Sharon Grevet for her English assistance.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author, M.P.G-P., upon reasonable request. Restrictions in relation to potentially person identifiable information apply.

Author Contributions

Conceptualization: F.D.S., E.F.-P., M.P.G.-P., S.G., J.B.; Data curation: F.D.S., E.F.-P., G.M.G.; Formal analysis: F.D.S., E.F.-P.; Data visualization: E.F.-P., L.G.-B., P.A.S.; Writing—original draft: F.D.S., E.F.-P., L.G.-B., P.A.S.; Writing—review & editing: M.P.G.-P., P.A.S., S.G., G.M.G., J.B.; Study supervision: M.P.G.-P., P.A.S., S.G., J.B. All authors gave the final approval.

Financial Support

This study was funded by the European Commission Seventh Framework Programme (HEALTH-F2-2010-242114).

Conflicts of Interest

F.D.S. has received grants from the Spanish Foundation of Psychiatry and Mental Health and the European Psychiatric Association. M.P.G.-P. has been a consultant to and/or has received honoraria/grants from Angelini, Alianza Otsuka-Lundbeck, Instituto de Salud Carlos III, Janssen-Cilag, Lundbeck, Otsuka, and Pfizer. L.G.-B. has been a consultant to and/or has received honoraria/grants from the Spanish Foundation of Psychiatry and Mental Health, European Psychiatric Association, Otsuka, Lundbeck, Janssen-Cilag and Pfizer. P.A.S. has been a consultant to and/or has received honoraria or grants from Adamed, CIBERSAM, European Commission, GlaxoSmithKline, Instituto de Salud Carlos III, Janssen-Cilag, Lundbeck, Otsuka, Pfizer, Plan Nacional sobre Drogas and Servier. S.G. has been a consultant to Angelini, Gedeon Richter-Recordati, Innova Pharma-Recordati Group within the last 2 years and has received honoraria/grants from Angelini, Gedeon Richter-Recordati, Janssen Pharmaceuticals, Janssen Cilag, Lundbeck, Recordati Pharmaceuticals. J.B. has received research grants and served as consultant, advisor or speaker within the last 5 years for: AB-Biotics, Acadia Pharmaceuticals, Angelini, Casen Recordati, D&A Pharma, Exeltis, Gilead, GSK, Ferrer, Indivior, Janssen-Cilag, Lundbeck, Mundipharma, Otsuka, Pfizer, Reckitt-Benckiser, Roche, Sage Therapeutics, Servier, Shire, Schwabe Farma Ibérica, research funding from the Spanish Ministry of Economy and Competiveness—Centro de Investigación Biomedica en Red area de Salud Mental (CIBERSAM) and Instituto de Salud Carlos III, Spanish Ministry of Health, Social Services and Equality, Plan Nacional Sobre Drogas and the 7th Framework Programme of the European Union. All other authors declare that they have no conflicts of interest.

References

Maj, M, van, OJ, De, HM, Gaebel, W, Galderisi, S, Green, MF, et al. The clinical characterization of the patient with primary psychosis aimed at personalization of management. World Psychiatry. 2021;20:4–33. doi:10.1002/wps.20809.CrossRefGoogle ScholarPubMed
Owen, MJ, Sawa, A, Mortensen, PB. Schizophrenia. Lancet. 2016;388: 86–97. doi:10.1016/S0140-6736(15)01121-6.CrossRefGoogle ScholarPubMed
Krynicki, CR, Upthegrove, R, Deakin, JFW, Barnes, TRE. The relationship between negative symptoms and depression in schizophrenia: a systematic review. Acta Psychiatr Scand. 2018;137:380–90. doi:10.1111/acps.12873.CrossRefGoogle ScholarPubMed
Upthegrove, R, Marwaha, S, Birchwood, M. Depression and schizophrenia: cause, consequence, or trans-diagnostic issue? Schizophr Bull. 2017;43:240–4. doi:10.1093/schbul/sbw097.Google ScholarPubMed
van Rooijen, G, Isvoranu, AM, Kruijt, OH, van Borkulo, CD, Meijer, CJ, Wigman, JTW, et al. A state-independent network of depressive, negative and positive symptoms in male patients with schizophrenia spectrum disorders. Schizophr Res. 2018;193:232–9. doi:10.1016/j.schres.2017.07.035.CrossRefGoogle ScholarPubMed
Galderisi, S, Rossi, A, Rocca, P, Bertolino, A, Mucci, A, Bucci, P, et al. The influence of illness-related variables, personal resources and context-related factors on real-life functioning of people with schizophrenia. World Psychiatry. 2014;13:275–87. doi:10.1002/wps.20167.CrossRefGoogle ScholarPubMed
Charlson, FJ, Ferrari, AJ, Santomauro, DF, Diminic, S, Stockings, E, Scott, JG, et al. Global epidemiology and burden of schizophrenia: findings from the global burden of disease study 2016. Schizophr Bull. 2018;44:1195–203. doi:10.1093/schbul/sby058.CrossRefGoogle ScholarPubMed
Bozzatello, P, Bellino, S, Rocca, P. Predictive factors of treatment resistance in first episode of psychosis: a systematic review. Front Psych. 2019;10:67. doi:10.3389/fpsyt.2019.00067.CrossRefGoogle ScholarPubMed
Lally, J, Ajnakina, O, Stubbs, B, Cullinane, M, Murphy, KC, Gaughran, F, et al. Remission and recovery from first-episode psychosis in adults: systematic review and meta-analysis of long-term outcome studies. Br J Psychiatry. 2017;211:350–8. doi:10.1192/bjp.bp.117.201475.CrossRefGoogle ScholarPubMed
Santesteban-Echarri, O, Paino, M, Rice, S, Gonzalez-Blanch, C, McGorry, P, Gleeson, J, et al. Predictors of functional recovery in first-episode psychosis: a systematic review and meta-analysis of longitudinal studies. Clin Psychol Rev. 2017;58:59–75. doi:10.1016/j.cpr.2017.09.007.CrossRefGoogle ScholarPubMed
Fonseca-Pedrero, E. Network analysis: a new way of understanding psychopathology? Revista de Psiquiatria y Salud Mental. 2017;10:206–15. doi:10.1016/j.rpsm.2017.06.004.CrossRefGoogle ScholarPubMed
Cramer, AOJ, Waldorp, LJ, Van Der Maas, HLJ, Borsboom, D. Comorbidity: a network perspective. Behav Brain Sci. 2010;33: 137–50. doi:10.1017/S0140525X09991567.CrossRefGoogle ScholarPubMed
Borsboom, D. A network theory of mental disorders. World Psychiatry. 2017;16: 5–13. doi:10.1002/wps.20375.CrossRefGoogle ScholarPubMed
Borsboom, D, Cramer, AOJ. Network analysis: an integrative approach to the structure of psychopathology. Annu Rev Clin Psychol. 2013;9:91–121. doi:10.1146/annurev-clinpsy-050212-185608.CrossRefGoogle Scholar
Borsboom, D, Cramer, AOJ, Kalis, A. Brain disorders? Not really: Why network structures block reductionism in psychopathology research. Behav Brain Sci. 2018;42: e2. doi:10.1017/S0140525X17002266.CrossRefGoogle Scholar
Izquierdo, A, Cabello, M, Leal, I, Mellor-Marsá, B, Ayora, M, Bravo-Ortiz, MF, et al. The interplay between functioning problems and symptoms in first episode of psychosis: an approach from network analysis. J Psychiatr Res. 2021;136:265–73. doi:10.1016/j.jpsychires.2021.02.024.CrossRefGoogle ScholarPubMed
Galderisi, S, Rucci, P, Kirkpatrick, B, Mucci, A, Gibertoni, D, Rocca, P, et al. Interplay among psychopathologic variables, personal resources, context-related factors, and real-life functioning in individuals with schizophrenia: a network analysis. JAMA Psychiatry. 2018;75:396–404. doi:10.1001/jamapsychiatry.2017.4607.CrossRefGoogle ScholarPubMed
Galderisi, S, Rucci, P, Mucci, A, Rossi, A, Rocca, P, Bertolino, A, et al. The interplay among psychopathology, personal resources, context-related factors and real-life functioning in schizophrenia: stability in relationships after 4 years and differences in network structure between recovered and non-recovered patients. World Psychiatry. 2020;19:81–91. doi:10.1002/wps.20700.CrossRefGoogle ScholarPubMed
Chang, WC, Wong, CSM, Or, PCF, Chu, AOK, Hui, CLM, Chan, SKW, et al. Inter-relationships among psychopathology, premorbid adjustment, cognition and psychosocial functioning in first-episode psychosis: a network analysis approach. Psychol Med. 2020;50:2019–27. doi:10.1017/S0033291719002113.CrossRefGoogle ScholarPubMed
Izquierdo, A, Cabello, M, de la Torre-Luque, A, Ayesa-Arriola, R, Setien-Suero, E, Mayoral-van-Son, J, et al. A network analysis approach to functioning problems in first psychotic episodes and their relationship with duration of untreated illness: findings from the PAFIP cohort. J Psychiatr Res. 2021;136: 483–491. doi:10.1016/j.jpsychires.2020.10.019.CrossRefGoogle Scholar
Fonseca-Pedrero, E, Chan, RCK, Debbané, M, Cicero, D, Zhang, LC, Brenner, C, et al. Comparisons of schizotypal traits across 12 countries: results from the international consortium for Schizotypy research. Schizophr Res. 2018;199: 128–34. doi:10.1016/j.schres.2018.03.021.CrossRefGoogle Scholar
Jones, PJ, Ma, R, McNally, RJ. Bridge centrality: a network approach to understanding comorbidity. Multivariate Behav Res. 2021;56:353–67. doi:10.1080/00273171.2019.1614898.CrossRefGoogle ScholarPubMed
Leucht, S, Winter-van Rossum, I, Heres, S, Arango, C, Fleischhacker, WW, Glenthøj, B, et al. The optimization of treatment and management of schizophrenia in Europe (OPTiMiSE) trial: rationale for its methodology and a review of the effectiveness of switching antipsychotics. Schizophr Bull. 2015;41: 549–558. doi:10.1093/SCHBUL/SBV019.CrossRefGoogle Scholar
Sheehan, DV, Lecrubier, Y, Sheehan, KH, Amorim, P, Janavs, J, Weiller, E, et al. The mini-international neuropsychiatric interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry. 1998;59:22–33.Google ScholarPubMed
Kahn, RS, van, RIW, Leucht, S, McGuire, P, Lewis, SW, Leboyer, M, et al. Amisulpride and olanzapine followed by open-label treatment with clozapine in first-episode schizophrenia and schizophreniform disorder (OPTiMiSE): a three-phase switching study. Lancet Psychiatry. 2018;5:797–807. doi:10.1016/S2215-0366(18)30252-9.CrossRefGoogle ScholarPubMed
Kay, SR, Fiszbein, A, Opler, LA. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr Bull. 1987;13:261–76. doi:10.1093/schbul/13.2.261.CrossRefGoogle Scholar
Addington, D, Addington, J, Schissel, B. A depression rating scale for schizophrenics. Schizophr Res. 1990;3: 247–51. doi:10.1016/0920-9964(90)90005-R.CrossRefGoogle ScholarPubMed
Guy, W. Clinical global impressions. In: ECDEU assessment manual for psychopharmacology, Rockville: National Institute for Mental Health; 1976, p. 218–22.Google Scholar
Morosini, PL, Magliano, L, Brambilla, L, Ugolini, S, Pioli, R. Development, reliability and acceptability of a new version of the DSM- IV social occupational functioning assessment scale (SOFAS) to assess routine social functioning. Acta Psychiatr Scand. 2000;101:323–9. doi:10.1034/j.1600-0447.2000.101004323.x.CrossRefGoogle ScholarPubMed
Epskamp, S, Fried, EI. A tutorial on regularized partial correlation networks. Psychol Methods. 2018;23:617–34. doi:10.1037/met0000167.CrossRefGoogle ScholarPubMed
Epskamp, S, Borsboom, D, Fried, EI. Estimating psychological networks and their stability: a tutorial paper. Behav Res Methods. 2018;50:195–212. doi:10.3758/s13428-017-0862-1.CrossRefGoogle ScholarPubMed
Epskamp, S, Cramer, AOJ, Waldorp, LJ, Schmittmann, VD, Borsboom, D. Qgraph: network visualizations of relationships in psychometric data. J Stat Softw. 2012;48:1–18. doi:10.18637/jss.v048.i04.CrossRefGoogle Scholar
Fonseca-Pedrero, E, Ortuño, J, Debbané, M, Chan, RCK, Cicero, D, Zhang, LC, et al. The network structure of schizotypal personality traits. Schizophr Bull. 2018;44:468–79. 10.1093/schbul/sby044.CrossRefGoogle Scholar
Robinaugh, DJ, Millner, AJ, McNally, RJ. Identifying highly influential nodes in the complicated grief network. J Abnorm Psychol. 2016;125:747–57. doi:10.1037/abn0000181.CrossRefGoogle ScholarPubMed
Haslbeck, JMB, Fried, EI. How predictable are symptoms in psychopathological networks? A reanalysis of 17 published datasets. 2017;47:2767–76. doi:10.1017/S0033291717001258.CrossRefGoogle Scholar
Opsahl, T, Agneessens, F, Skvoretz, J. Node centrality in weighted networks: generalizing degree and shortest paths. Soc Networks. 2010;32:245–51. doi:10.1016/j.socnet.2010.03.006.CrossRefGoogle Scholar
Jones, PJ. Networktools: tools for identifying important nodes in networks, https://cran.r-project.org/package=networktools; 2021 (Accessed 27 November 2021).Google Scholar
IBM Corp Released. IBM SPSS statistics for windows, version 22.0. Armonk, NY: IBM Corp; 2013.Google Scholar
R Core Team. R: a Language and environment for statistical computing, https://cran.r-project.org/; 2018 (Accessed 27 November 2021).Google Scholar
Wallwork, RS, Fortgang, R, Hashimoto, R, Weinberger, DR, Dickinson, D. Searching for a consensus five-factor model of the positive and negative syndrome scale for schizophrenia. Schizophr Res. 2012;137: 246–50. doi:10.1016/J.SCHRES.2012.01.031.CrossRefGoogle Scholar
Higuchi, CH, Cogo-Moreira, H, Fonseca, L, Ortiz, BB, Correll, CU, Noto, C, et al. Identifying strategies to improve PANSS based dimensional models in schizophrenia: accounting for multilevel structure, Bayesian model and clinical staging. Schizophr Res. 2022;243:424–30. doi:10.1016/J.SCHRES.2021.06.034.CrossRefGoogle ScholarPubMed
Amore, M, Murri, MB, Calcagno, P, Rocca, P, Rossi, A, Aguglia, E, et al. The association between insight and depressive symptoms in schizophrenia: undirected and Bayesian network analyses. Eur Psychiatry. 2020;63:1–21. doi:10.1192/j.eurpsy.2020.45.CrossRefGoogle ScholarPubMed
Griffiths, SL, Leighton, SP, Mallikarjun, PK, Blake, G, Everard, L, Jones, PB, et al. Structure and stability of symptoms in first episode psychosis: a longitudinal network approach. Transl Psychiatry. 2021;11:567 doi:10.1038/S41398-021-01687-Y.CrossRefGoogle ScholarPubMed
Piao, YH, Yun, JY, Nguyen, TB, Kim, WS, Sui, J, Kang, NI, et al. Longitudinal symptom network structure in first-episode psychosis: a possible marker for remission. Psychol Med. 2021:1–9. doi:10.1017/S0033291720005280.CrossRefGoogle ScholarPubMed
Betz, LT, Penzel, N, Kambeitz-Ilankovic, L, Rosen, M, Chisholm, K, Stainton, A, et al. General psychopathology links burden of recent life events and psychotic symptoms in a network approach. NPJ Schizophr. 2020;6:40 doi:10.1038/S41537-020-00129-W.CrossRefGoogle Scholar
Herniman, SE, Allott, K, Phillips, LJ, Wood, SJ, Uren, J, Mallawaarachchi, SR, et al. Depressive psychopathology in first-episode schizophrenia spectrum disorders: a systematic review, meta-analysis and meta-regression. Psychol Med. 2019;49:2463–74. doi:10.1017/S0033291719002344.CrossRefGoogle ScholarPubMed
Canal-Rivero, M, López-Moríñigo, JD, Setién-Suero, E, Ruiz-Veguilla, M, Ayuso-Mateos, JL, Ayesa-Arriola, R, et al. Predicting suicidal behaviour after first episode of non-affective psychosis: the role of neurocognitive functioning. Eur Psychiatry. 2018;53:52–7. doi:10.1016/j.eurpsy.2018.06.001.CrossRefGoogle ScholarPubMed
Herniman, SE, Phillips, LJ, Wood, SJ, Cotton, SM, Liemburg, EJ, Allott, KA. Interrelationships between depressive symptoms and positive and negative symptoms of recent onset schizophrenia spectrum disorders: a network analytical approach. J Psychiatr Res 2021;140:373–80. doi:10.1016/J.JPSYCHIRES.2021.05.038.CrossRefGoogle ScholarPubMed
Buonocore, M, Bosia, M, Baraldi, MA, Bechi, M, Spangaro, M, Cocchi, F, et al. Exploring anxiety in schizophrenia: new light on a hidden figure. Psychiatry Res. 2018;268:312–6. doi:10.1016/J.PSYCHRES.2018.07.039.CrossRefGoogle ScholarPubMed
Bjornestad, J, Hegelstad, W ten V, Joa, I, Davidson, L, Larsen, TK, Melle, I, et al. With a little help from my friends” social predictors of clinical recovery in first-episode psychosis. Psychiatry Res. 2017;255:209–14. doi:10.1016/j.psychres.2017.05.041.CrossRefGoogle ScholarPubMed
Hansen, CF, Torgalsbøen, AK, Melle, I, Bell, MD. Passive/apathetic social withdrawal and active social avoidance in schizophrenia: difference in underlying psychological processes. J Nerv Ment Dis. 2009;197:274–7. doi:10.1097/NMD.0B013E31819DBD36.CrossRefGoogle ScholarPubMed
Galderisi, S, Kaiser, S, Bitter, I, Nordentoft, M, Mucci, A, Sabé, M, et al. EPA guidance on treatment of negative symptoms in schizophrenia. Eur Psychiatry. 2021;64(1):e23. doi:10.1192/J.EURPSY.2021.13.CrossRefGoogle Scholar
Phalen, PL, Dimaggio, G, Popolo, R, Lysaker, PH. Aspects of theory of mind that attenuate the relationship between persecutory delusions and social functioning in schizophrenia spectrum disorders. J Behav Ther Exp Psychiatry. 2017;56:65–70. doi:10.1016/J.JBTEP.2016.07.008.CrossRefGoogle ScholarPubMed
Rocca, P, Galderisi, S, Rossi, A, Bertolino, A, Rucci, P, Gibertoni, D, et al. Disorganization and real-world functioning in schizophrenia: results from the multicenter study of the Italian network for research on psychoses. Schizophr Res. 2018;201:105–12. doi:10.1016/j.schres.2018.06.003.CrossRefGoogle ScholarPubMed
Coid, JW, Ullrich, S, Kallis, C, Keers, R, Barker, D, Cowden, F, et al. The relationship between delusions and violence: findings from the East London first episode psychosis study. JAMA Psychiat. 2013;70:465–71. doi:10.1001/jamapsychiatry.2013.12.CrossRefGoogle ScholarPubMed
Faay, MDM, van Os, J. Aggressive behavior, hostility, and associated care needs in patients with psychotic disorders: a 6-year follow-up study. Front Psychiatry. 2020;10:934. doi:10.3389/fpsyt.2019.00934.CrossRefGoogle ScholarPubMed
Moulin, V, Golay, P, Palix, J, Baumann, PS, Gholamrezaee, MM, Azzola, A, et al. Impulsivity in early psychosis: a complex link with violent behaviour and a target for intervention. Eur Psychiatry. 2018;49:30–6. doi:10.1016/j.eurpsy.2017.12.003.CrossRefGoogle Scholar
Witt, K, van Dorn, R, Fazel, S. Risk factors for violence in psychosis: systematic review and meta-regression analysis of 110 studies. PLoS One. 2013;8:e55942. doi:10.1371/journal.pone.0055942.CrossRefGoogle ScholarPubMed
Lopez-Garcia, P, Ashby, S, Patel, P, Pierce, KM, Meyer, M, Rosenthal, A, et al. Clinical and neurodevelopmental correlates of aggression in early psychosis. Schizophr Res. 2019;212:171–6. doi:10.1016/j.schres.2019.07.045.CrossRefGoogle ScholarPubMed
Faay, MDM, van Baal, GCM, Arango, C, Díaz-Caneja, CM, Berger, G, Leucht, S, et al. Hostility and aggressive behaviour in first episode psychosis: results from the OPTiMiSE trial. Schizophr Res. 2020;223:271–8. doi:10.1016/j.schres.2020.08.021.CrossRefGoogle ScholarPubMed
Figure 0

Table 1. Sociodemographic data and clinical and functional assessment of the sample (n = 446).

Figure 1

Figure 1. Estimated network for psychosis phenotype, depression symptoms, and real-life functioning. CDSS: Calgary Depression Scale for schizophrenia; G: PANSS, general psychopathology dimension; N: PANSS, negative psychosis dimension; PSP: personal and social performance; P: PANSS, positive psychosis dimension. Numbers represent item numbers in the scale; blue edges represent positive associations; red edges represent negative associations. Thickness and saturation of edges indicate the strength of these associations.

Figure 2

Figure 2. Inference measures of the estimated psychosis network. CDSS: Calgary Depression Scale for Schizophrenia; G: PANSS, general psychopathology dimension; N: PANSS, negative psychosis dimension; PSP: personal and social performance; P: PANSS, positive psychosis dimension. Numbers represent item numbers in the scale.

Figure 3

Figure 3. Bridge centrality measures of the estimated psychosis network. CDSS: Calgary Depression Scale for Schizophrenia; G: PANSS, general psychopathology dimension; N: PANSS, negative psychosis dimension; PSP: personal and social performance; P: PANSS, positive psychosis dimension. Numbers represent item numbers in the scale.

Submit a response

Comments

No Comments have been published for this article.