Assessment of interrelationships between cognitive performance, symptomatic manifestation and social functioning in the acute and clinical stability phase of schizophrenia: insights from a network analysis

Schizophrenia is a neurodevelopmental disorder that often leads to lasting impairments of social functioning [1]. It is ranked among the most important causes of disability worldwide [2]. Nevertheless, there is evidence that social functioning of people with psychotic disorders tends to improve over time, especially in those with lower baseline severity of negative symptoms and greater baseline quality of life [3]. This effect is more pronounced in people with shorter illness duration. At this point, it is important to note that recovery in schizophrenia should not only include remission of clinical symptoms but also improvement of social functioning that manifests in sustained employment, independent living, overall involvement in the society, and social interactions [4]. However, functional recovery rates are still low and reach about 58% in people with first-episode psychosis (FEP) [5] and about 38% in those with multiple-episode schizophrenia [6]. In this regard, the recognition of factors contributing to functional recovery holds important implications for clinical practice.

It is important to note that impairments of social functioning in schizophrenia are largely attributable to all aspects of psychopathological manifestation, including positive and negative symptoms, formal thought disorder as well as affective symptoms [7, 8]. Furthermore, it has been shown that impairments of neurocognition serve as an important factor contributing to decreased social functioning of people with schizophrenia [9]. Neurocognitive impairments observed in people with schizophrenia affect several domains, including speed of processing, attention/vigilance, working memory, verbal learning, visual learning as well as reasoning and problem solving [10‐12]. Moreover, there is evidence that individuals with schizophrenia show deficits of social cognition defined by a range of processes that enable to interpret social contexts and respond appropriately [13]. A recent meta-analysis of studies performed in people with early psychosis demonstrated that impairments across all domains of neurocognitive performance and social cognition are associated with lower psychosocial functioning both cross-sectionally and longitudinally [14]. However, there is evidence that neurocognitive deficits might affect social functioning through the effects on social cognition [15‐17]. Other factors that have been associated with lower social functioning include earlier age of psychosis onset [18] and longer duration of untreated psychosis [19].

Neurocognitive deficits in people with schizophrenia are closely related to negative symptoms. The National Institute of Mental Health (NIMH) developed the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) initiative that aimed to develop the consensus about the definition and methods of measuring negative symptoms [20]. According to the NIMH-MATRICS consensus statement, negative symptoms include avolition, anhedonia, asociality, blunted affect, and alogia [20‐22]. Negative symptoms are now perceived as the core component of schizophrenia that largely contributes to long-term disability and poor functional outcomes [23].

Our understanding of psychopathological manifestation of schizophrenia, its underlying mechanisms, and clinical relevance has largely improved with the development of network approaches to data analysis. A network analysis assumes that mental disorders represent systems of causally associated symptoms as opposed to the effects of a single latent factor [24]. This approach to analysis of psychopathology allows to address its complexity without indicating a specific model [25]. Findings are represented by networks, where specific variables (nodes) are connected with edges. Due to the possibility to indicate central nodes of psychopathology, i.e., those that have the highest number of strong connections with other nodes, a network analysis holds the usefulness to indicate the most important targets for therapeutic interventions.

To date, some studies have aimed to address the complexity of connections between various aspects of clinical manifestation of schizophrenia that are of importance for shaping social functioning. For instance, Galderisi et al. [26] found that improving functional capacity and daily life skills might be critical for planning any therapeutic interventions. Another study investigated differential effects of interrelationships between psychopathology, metacognition, and neurocognition on functional capacity in people with early- and late-stage of schizophrenia [27]. The authors found that the global network structure and strength do not differ significantly between both stages of schizophrenia. Moreover, this study demonstrated that visual learning and disorganization symptoms serve as the most central determinants of social functioning in schizophrenia. Also, disorganization and negative symptoms as well as metacognition appeared to be directly and strongly associated with real-life functioning. Investigating determinants of social functioning at various stages of schizophrenia might be of importance for potential personalization of therapeutic approaches. However; these aspects have not been thoroughly addressed by the prior studies. In other words, it remains unknown as to whether there are the same symptomatic and neurocognitive determinants of social functioning during the acute relapse and clinical stability phase of schizophrenia. To address this research gap, the present study aimed to perform and compare separate network analyses of psychopathology, neurocognition, and social functioning in individuals with acute relapse of schizophrenia spectrum disorders and those during clinical stability phase.

Findings from the present study indicate that associations of clinical manifestation with functional capacity are similar at various stages of schizophrenia with important roles of positive, negative and depressive symptoms as well as neurocognitive impairments. This observation is in agreement with results of the prior study showing no significant differences in the network of factors related to social functioning between early- and late-stage schizophrenia [27]. Also, the network comparison of remitted and non-remitted individuals with schizophrenia has demonstrated that depressive, positive, and negative symptoms are linked in a stable way [40]. In our study, impairments of neurocognition appeared to be the most central predictor of social functioning. This observation was demonstrated for the RBANS score of immediate memory (only in remitted patients with schizophrenia) and attention (in both samples) that is composed of scores from two tasks, i.e., digit span and symbol coding. Apart from attention, these tasks measure other aspects of cognitive performance including processing speed and working memory, respectively. Finally, our study confirmed that all domains of psychopathology are associated with social functioning in people with schizophrenia regardless of illness stage [7]. Although in our study both samples were similar in terms of sociodemographic characteristics (except for age) and the level of cognitive performance, some differences need to be pointed out. Indeed, acutely relapsed patients had higher levels of positive and negative symptoms as well as lower levels of social functioning. Moreover, the dosage of antipsychotics was significantly higher in remitted patients with schizophrenia.

Our observations about the central role of neurocognition in shaping functional capacity are in agreement with the prior studies. For instance, a prospective study of individuals with FEP, based on a network analysis, revealed that working memory appears to have the highest centrality in predicting social functioning [41]. Another cross-sectional study of stable outpatients with schizophrenia demonstrated that working memory deficits measured using the letter number sequencing task were among the most central nodes predicting social functioning [42]. Also, the study based on 81 individuals with schizophrenia spectrum disorders in a non-acute phase revealed the greatest centrality of the PANSS cognitive factor score and working memory. However, the authors did not assess social functioning [43]. In turn, the study by Karyakina and Shmukler [44] compared the networks of neurocognitive performance in people with schizophrenia spectrum disorders and healthy controls. The authors found that processing speed impairments, measured using the symbol coding task, were among the most central neurocognitive impairments in people with schizophrenia spectrum disorders. Among healthy controls, working memory appeared to have the greatest centrality among all neurocognitive domains. Processing speed and working memory impairments represent the most important aspects of cognitive deficits observed in psychotic disorders that can already be detected in people at clinical high risk of psychosis (especially those who transit to overt psychosis over time) [45], first-degree relatives of people with schizophrenia [46], and individuals with FEP [47].

The central role of attention, processing speed, immediate and working memory indicates that these domains of neurocognition should be the target for interventions that also aim to improve social functioning in schizophrenia. Among these interventions, several studies have focused on cognitive remediation. The most recent meta-analyses revealed that cognitive remediation strategies are effective in improving cognitive performance with low-to-moderate effect size estimates [48, 49]. Interestingly, one of these meta-analyses, that synthesized evidence from studies performed among inpatients with schizophrenia, suggested efficacy in improving processing speed, memory, and working memory [48]. However, evidence with respect to improvement of social and global functioning was less reliable. At this point, it is important to note that cognitive remediation strategies use various protocols and target various domains of neurocognitive functioning. In this regard, our findings might provide additional support for selecting targets of cognitive remediation.

There are some limitations of our study. First, our sample size was not large. However, a network analysis achieved sufficient stability and accuracy, likely due to relatively low number of nodes included in the network. Second, insights into social functioning might be limited due to the use of SOFAS that does not provide insights into various aspects of social functioning. Also, the PANSS does not provide comprehensive insights into the current conceptualization of negative symptoms and fails to fails to assess subject’s internal experience. Third, a network analysis did not analyze other important predictors of social functioning, e.g., those related to social cognition and metacognition. Nevertheless, it is important to note that impairments of social cognition and metacognition might be the consequence of neurocognitive deficits [15‐17]. Therefore, it is likely that inclusion of these variables would not change considerably the centrality measures. Fourth, our sample of individuals during acute phase of psychosis was not homogenous in terms of specific diagnostic categories. Finally, although we used a network analysis approach, a lack of prospective design does not allow to provide solid insights into causality.

In sum, findings from the present study indicate that the effects of clinical manifestation on social functioning in schizophrenia spectrum disorders do not differ considerably between acute and clinical stability phases. Importantly, our observations indicate that impairments of attention, processing speed as well as working and immediate memory might be the most central predictors of social functioning in schizophrenia spectrum disorders. This gives a translational perspective indicating that interventions aimed at improving functional capacity of individuals with schizophrenia should focus on these domains of neurocognitive performance.