Psychological Medicine Influence of social cognition as a mediator cambridge.org/psm between cognitive reserve and psychosocial functioning in patients with first episode psychosis Original Article Cite this article: González-Ortega I et al I González-Ortega1,2,3,4 , A González-Pinto1,2,3, S Alberich1,2,3,4, E Echeburúa1,5, (2020). Influence of social cognition as a M Bernardo1,6, B Cabrera1,6, S Amoretti1,6, A Lobo1,7, C Arango1,8, I Corripio1,9, mediator between cognitive reserve and psychosocial functioning in patients with first E Vieta1,10, E de la Serna1,11, R Rodriguez-Jimenez1,12, R Segarra1,13, episode psychosis. Psychological Medicine 50, 2702–2710. https://doi.org/10.1017/ JM López-Ilundain14, AM Sánchez-Torres14, MJ Cuesta14, PEPs Group: S0033291719002794 I Zorrilla1,2,3, P López1,2,3, M Bioque1,6, G Mezquida1,6, F Barcones7,15, Received: 23 January 2019 Revised: 9 August 2019 C De-la-Cámara1,7, M Parellada1,8, A Espliego1,8, A Alonso-Solís1,9, EM Grasa1,9, Accepted: 13 September 2019 C Varo10, L Montejo1,10, J Castro-Fornieles1,11, I Baeza1,11, M Dompablo1,12, First published online: 22 October 2019 I Torio1,12, A Zabala1,13, JI Eguiluz1,13, L Moreno-Izco14, J Sanjuan1,16, R Guirado17, Key words: Cognitive reserve; first episode psychosis; I Cáceres18, P Garnier18, F Contreras1,19,20, J Bobes1,21, S Al-Halabí1,21,22, psychosocial functioning; social cognition J Usall1,23, A Butjosa1,23, S Sarró1,24, R Landin-Romero24, A Ibáñez1,25 Author for correspondence: I González-Ortega, and G Selva1,26 E-mail: itxaso.gonzalezortega@osakidetza.eus 1 Centre for Biomedical Research in the Mental Health Network (CIBERSAM), Madrid, Spain; 2Department of Psychiatry, Araba University Hospital, Bioaraba Research Institute, Vitoria, Spain; 3Department of Neurosciences, University of the Basque Country, Bizkaia, Spain; 4The National Distance Education University (UNED), Vitoria, Spain; 5Department of Personality, Assessment and Psychological Treatment, University of the Basque Country, San Sebastián, Spain; 6Barcelona Clinic Schizophrenia Unit, Neuroscience Institute, Hospital Clinic of Barcelona, Barcelona, Spain; 7Department of Medicine and Psychiatry, University of Zaragoza, Aragon Institute for Health Sciences (IIS Aragón), Zaragoza, Spain; 8Child and Adolescent Psychiatry Department, Gregorio Marañón General University Hospital, School of Medicine, Universidad Complutense, IiSGM, Madrid, Spain; 9Department of Psychiatry, Institut d’Investigació Biomèdica-Sant Pau (IIB-SANT PAU), Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona (UAB), Barcelona, Spain; 10Bipolar Disorders Unit, Hospital Clinic, University of Barcelona, IDIBAPS, Barcelona, Spain; 11Child and Adolescent Psychiatry Service, Hospital Clinic of Barcelona, Barcelona, Spain; 1212 de Octubre Hospital Research Institute (i+12), Madrid, Spain; 13Department of Neurosciences, University of the Basque Country, Cruces University Hospital, Biocruces Bizkaia Health Research Institute, Vizcaya, Spain; 14Department of Psychiatry, Navarre Hospital Complex, IdiSNA, Navarre Institute for Health Research, Pamplona, Spain; 15Department of Family Medicine, Hospital Universitario Miguel Servet, Zaragoza, Spain; 16INCLIVA, University of Valencia, Hospital Clinico Universitario of Valencia, Spain; 17Neurobiology Unit, Department of Cell Biology, Interdisciplinary Research Structure for Biotechnology and Biomedicine (BIOTECMED), University of Valencia, Valencia, Spain; 18Department of Psychiatry, Hospital del Mar Medical Research Institute (IMIM), Barcelona, Spain; 19Psychiatry Department, Bellvitge University Hospital-IDIBELL, Barcelona, Spain; 20Department of Clinical Sciences, School of Medicine, University of Barcelona, Barcelona, Spain; 21Psychiatry Department, University of Oviedo, Oviedo, Spain; 22Institute of Neurosciences of the Principality of Asturias, INEUROPA, Oviedo, Spain; 23Research Unit, Parc Sanitari Sant Joan de Déu, Universitat de Barcelona (UB), Sant Boi de Llobregat, Barcelona, Spain; 24FIDMAG Hermanas Hospitalarias Research Foundation, Barcelona, Spain; 25Psychiatry Department, Ramón y Cajal University Hospital, Ramón y Cajal Health Research Institute (IRyCIS), University of Alcalá, Madrid, Spain and 26Teaching Unit of Psychiatry and Psychological Medicine, Department of Medicine, University of Valencia, INCLIVA Health Research Institute, Valencia, Spain Abstract Background. Social cognition has been associated with functional outcome in patients with first episode psychosis (FEP). Social cognition has also been associated with neurocognition and cognitive reserve. Although cognitive reserve, neurocognitive functioning, social cogni- ©The Author(s) 2019. This is an Open Access tion, and functional outcome are related, the direction of their associations is not clear. article, distributed under the terms of the Therefore, the main aim of this study was to analyze the influence of social cognition as a Creative Commons Attribution licence (http:// mediator between cognitive reserve and cognitive domains on functioning in FEP both at creativecommons.org/licenses/by/4.0/), which baseline and at 2 years. permits unrestricted reuse, distribution, and reproduction in any medium, provided the Methods. The sample of the study was composed of 282 FEP patients followed up for 2 years. original work is properly cited. To analyze whether social cognition mediates the influence of cognitive reserve and cognitive domains on functioning, a path analysis was performed. The statistical significance of any mediation effects was evaluated by bootstrap analysis. Results. At baseline, as neither cognitive reserve nor the cognitive domains studied were related to functioning, the conditions for mediation were not satisfied. Nevertheless, at 2 https://doi.org/10.1017/S0033291719002794 Published online by Cambridge University Press Psychological Medicine 2703 years of follow-up, social cognition acted as a mediator between cognitive reserve and func- tioning. Likewise, social cognition was a mediator between verbal memory and functional out- come. The results of the bootstrap analysis confirmed these significant mediations (95% bootstrapped CI (−10.215 to −0.337) and (−4.731 to −0.605) respectively). Conclusions. Cognitive reserve and neurocognition are related to functioning, and social cog- nition mediates in this relationship. Introduction and at 2 years of follow-up; (3) to analyze the influence of social cognition on functioning; and (4) to analyze the influence of social Neurocognition, cognitive reserve, and social cognition have all cognition as a mediator between cognitive reserve and cognitive been related to functioning in first-episode psychosis (FEP). domains on functioning. Specifically, numerous studies have found that cognitive deficits are present early in psychosis, manifesting in multiple cognitive domains including working memory, executive function, atten- Method tion, processing speed, and/or learning and memory (Zabala Subjects et al., 2010; González-Ortega et al., 2013; Bora and Murray, 2014). These deficits have been related to poorer psychosocial This work was part of the ‘Phenotype-genotype and environmen- functioning (Leeson et al., 2011; González-Ortega et al., 2013; tal interaction. Application of a predictive model in first psychotic Green and Harvey, 2014; Torgalsbøen et al., 2015), but some stud- episodes’ study (PEPs study, from its acronym in Spanish) ies have not found this association (Kravariti et al., 2003; Stirling (Bernardo et al., 2013). et al., 2003). Overall, there is a lack of consensus on how cognitive Patients were aged between 18 and 35 years old at the time of functioning changes over the course of the disease and its relation first evaluation and were required to have fluent Spanish and to to functional outcome. give written informed consent. If patients were not capable of giv- Cognitive reserve has been associated with the functional ing consent or minors, a legally authorized representative was course of FEP, patients with a higher cognitive reserve showing asked to provide consent on their behalf. Moreover, it was a better outcome (Leeson et al., 2011; Amoretti et al., 2016). required that patients had a less than 12-month history of psych- Cognitive reserve has been defined as the ability of a brain to otic symptoms. cope with brain pathology and thereby minimize symptoms Exclusion criteria for patients were: mental retardation accord- (Stern, 2002) and variables used as measures of cognitive reserve ing to the DSM-IV (American Psychiatric Association, 1994) include premorbid IQ, educational level, and occupational attain- (including not only an IQ below 70, as determined using the ment (Anaya et al., 2016). test described in Table 1, but also poor functioning), history of Social cognition has also been associated with functional out- head trauma with the loss of consciousness, and organic illness come. Indeed, it has been suggested that social cognition is a bet- affecting mental health. The study was approved by the clinical ter predictor of functional outcome of FEP patients than research ethics committees of all participating centers in the neurocognition (Fett et al., 2011; Ohmuro et al., 2016). Social cog- PEPs Group. nition includes theory of mind, social perception, social knowl- Initially, 335 FEP patients were included in the study. Of these, edge, attributional biases, and emotion processing, areas that are 53 patients were excluded for not having seven or more neuro- affected in FEP patients (Horan et al., 2012; Bora and Pantelis, psychological tests completed. Thus, the final study sample con- 2013; Bliksted et al., 2014). It has also been associated with neu- sisted of 282 patients. rocognition and premorbid IQ (Bliksted et al., 2014) and there is evidence of a relationship between social cognition and neurocog- Data collection nition and that this may have a negative influence on functional outcome of FEP (Schmidt et al., 2011). Sociodemographic and clinical data were collected at baseline and Although cognitive reserve, neurocognitive functioning, social at 2 years, while cognitive assessment was performed at 2 months cognition, and functional outcome are probably related, the direc- of follow-up, in order to ensure the psychopathological stability of tion of their associations is not clear (Fett et al., 2011; Schmidt patients, and at 2 years of follow-up. If patients were actively et al., 2011; Pinkham and Harvey, 2012; Hedman et al., 2013; psychotic at the time of the 2-year follow-up, we delayed the Lin et al., 2013). In a systematic review, Schmidt et al. (2011) assessment until their condition was stabilized. The assessment found that social cognition acts as a mediator between neurocog- protocol was fully described by Bernardo et al. (2013). nition and functioning in schizophrenia (Schmidt et al., 2011), Adult patients were diagnosed with FEP using the Structured though in individuals at clinical high risk of psychosis this was Clinical Interview for DSM-IV Axis I and II Disorders (First not observed (Barbato et al., 2013). et al., 1997), while the Schedule for Affective Disorders and Therefore, for predicting the outcome of FEP patients after the Schizophrenia for school-age children-Present and Lifetime ver- onset of the disease, it is important to explore the potential medi- sion (Kaufman et al., 1997) was used for those under 18 years ating role of social cognition in the association between neurocog- old, following DSM-IV criteria. nition and functioning. In this context, our study had the following Functional impairment of patients was evaluated using the objectives: (1) to assess changes from baseline in cognitive function- functional assessment short test (FAST) (Rosa et al., 2007), for ing, social cognition, and functioning at 2 years of follow-up in FEP which high scores suggest poorer functioning. The FAST is a patients; (2) to analyze the influence of cognitive reserve and cog- measure of functioning that is valid, reliable, and sensitive to nitive domains on social cognition and functioning at baseline change and has been validated and widely used in bipolar https://doi.org/10.1017/S0033291719002794 Published online by Cambridge University Press 2704 I González-Ortega et al. Table 1. Description of test and measures used for cognitive domain summary scores Cognitive domain Neuropsychological subtests Premorbid IQ Vocabulary subtest of WAIS-IIIa/WISC-IVb Processing speed Trail Making Testc-form A Stroop Test word and color conditiond Attention CPT-IIe Verbal memory TAVECf Working memory Digit Span subtest of the WAIS-IIIa/WISC-IVb Letter and number sequencing of the WAIS-IIIa/ WISC-IVb Fig. 1. Hypothesized mediation model. g Executive FAS test function Animal wordsh differences in cognitive performance and functioning of patients c Trail Making Test -form B between baseline and 2 years of follow-up. To assess the cognitive domains, we calculated the average of WCSTi the z-scores of the neuropsychological variables which corres- Stroop Testd-interference condition pond to each domain (Table 1). As the scores of the Trail a Wechsler Adult Intelligence Scale III (WAIS-III) (Wechsler, 1997). Making Test (Reitan and Wolfson, 1985) are inverse, they b Wechsler Intelligence Scale for Children-IV (WISC-IV) (Wechsler, 2003). were multiplied by −1 to facilitate the interpretation (higher c Trail Making Test (Reitan and Wolfson, 1985). scores indicating better performance). The same approach was d Stroop Test (Golden, 1978). e Conners’ Continuous Performance Test (CPT-II) (Conners, 2004). taken for cognitive reserve, that is, averages of the z-scores f g Test de Aprendizaje Verbal España-Complutense (TAVEC) (Benedet and Alejandre, 1998). were calculated for estimated premorbid IQ, educational level, Controlled Oral Word Association Test (FAS test) (Loonstra et al., 2001). h Animal words (Peña-Casanova, 1990). and occupational attainment. i Wisconsin Card Sorting Test (WCST) (Heaton et al., 1993). To explore whether social cognition mediates the influence of cognitive reserve and cognitive domains on functioning, a three- step procedure was used (Frazier et al., 2004). Figure 1 represents disorder (Rosa et al., 2011; Bonnín et al., 2018; Vieta et al., 2018), the hypothesized mediation model. Firstly, we analyzed the rela- attention-deficit hyperactivity disorder (Rotger et al., 2014), and tionship between cognitive reserve and cognitive domains with FEP (González-Ortega et al., 2010). social cognition with a linear regression (path a). Secondly, linear The substance use was assessed using the European Adaptation regression models were built to explore the association of cognitive of a Multidimensional Assessment Instrument for Drug and reserve and cognitive domains (independent variables) with func- Alcohol Dependence (EuropASI) and patients were categorized tioning (dependent variable) (path c). All models were adjusted according to EuropASI scores into four groups (no use, use, for potential confounders (namely, age, sex, civil status, socio- abuse, and dependence) (Kokkevi and Hartgers, 1995). economic level, educational level, diagnosis, substance use, and The following cognitive domains are assessed: processing family history) with a stepwise procedure which only included vari- speed, attention, verbal memory, working memory executive ables in the final model if they produced a significant change in the function, and premorbid IQ. The tests and measures used for coefficient of the independent variable. Finally, social cognition was domain summary scores are described by Bernardo et al. (2013) included as a mediator in the multiple regressions found to be sig- and Cuesta et al. (2015) (Table 1). nificant in the second step (paths b and c´). If the relation of social Estimated premorbid IQ, educational level, and occupational cognition with functioning was significant and the path c associ- attainment were used as measures of cognitive reserve (Anaya ation became non-significant, conditions for mediation would be et al., 2016). IQ was evaluated with the Vocabulary subtest of satisfied, this suggesting that social cognition is a mediating vari- Wechsler Adult Intelligence Scale (Wechsler, 1997) or Wechsler able. The statistical significance of the mediation was examined Intelligence Scale for Children-IV (Wechsler, 2003), while educa- by bootstrap analysis (95% bootstrapped confidence interval (CI)). tional level and occupational attainment variables were assessed Data analyses were conducted with Stata 12.1. The level of sig- with the Hollingshead–Redlich Scale (Hollingshead and Redlich, nificance was set at p < 0.05, except for the analyses of the cogni- 1958). tion domains for which we applied the Bonferroni correction for Social cognition was assessed with understanding emotions, multiple comparisons. For these, the level of significance was set managing emotions, and total emotional intelligence scores of at p < 0.01, calculated by dividing 0.05 by five, the number of cog- the Mayer–Salovey–Caruso Emotional Intelligence Test, which nitive domains. rate the helpfulness of certain moods and assess the effectiveness of strategies to manage emotions (Mayer et al., 2009). Results Sociodemographic characteristics, cognitive, and functional Statistical analyses course of the sample Demographic variables are described with means (±S.D.) and per- Baseline characteristics of the sample are summarized in Table 2. centages. t Tests for paired samples were used to analyze The total sample comprised 192 men (68.1%) with a mean age of https://doi.org/10.1017/S0033291719002794 Published online by Cambridge University Press Psychological Medicine 2705 Table 2. Baseline characteristics of the sample Table 3. Means of cognition and FAST during follow-up (n = 186) Variable Total sample (n = 282) Variable Baseline 2 years Sex Attention −0.38 (0.99) −0.21 (0.87) Male 192 (68.1%) Processing speed −0.96 (1.06) −0.69 (1.00) Age, years 23.64 (5.93) Verbal memory −1.07 (1.24) −0.65 (1.11) Civil status Working memory −0.71 (0.76) −0.67 (0.86) Single 245 (86.9%) Executive function −0.43 (0.50) −0.16 (0.62) Married 20 (7.1%) Social cognition −0.72 (0.89) −0.58 (0.95) Other 17 (6%) Cognitive reserve −0.45 (0.56) 0.04 (0.56) Socioeconomic level FAST 26.01 (16.16) 19.40 (15.40) Low 124 (44.6%) p value <0.05 except for the cognitive domains for which the Bonferroni correction was applied ( p < 0.01). Bold values represent the significant values. Medium 69 (24.8%) High 85 (30.6%) At 2 years of follow-up, after including social cognition at baseline in the final model, all significant relations of baseline 23.64 (5.93) years. Of these, 123 (43.6%) had a diagnosis of cognitive domains and cognitive reserve with FAST at 2 years schizophrenia and 59 (20.9%) had a diagnosis of bipolar disorder. remained significant. Further, their coefficients increased after The drug most frequently used was tobacco (66.4%) and most the inclusion of social cognition, and therefore, baseline social patients were single (86.9%). cognition did not mediate in these relations. Nevertheless, rele- During 2 years of follow-up, all the cognitive domain scores vant results were obtained after including social cognition at 2 improved significantly, except working memory and executive years in the model. Social cognition at 2 years was a significant function. Social cognition, cognitive reserve, and functioning mediator between cognitive reserve and functioning at 2 years were also significantly better at 2 years (Table 3). (Fig. 2). In fact, its association with functioning was significant Correlations between cognitive and functional variables are (path b), while the relation between cognitive reserve and func- presented in a Supplementary Table. tioning became non-significant (path c´). These findings suggest a significant full mediation of social cognition, it mediating 20.8% of the effect (95% bootstrapped CI −10.215 to −0.337). With respect to cognitive domains, after including social cog- Relationship between cognitive reserve, social cognition, nition at 2 years in the regression model between baseline atten- general cognition, and functioning tion and functioning at 2 years, social cognition had a significant Table 4 reports the results of the analysis of the basic path models association with functioning (path b) and attention remained sig- at both baseline and 2 years. At baseline, cognitive reserve was sig- nificant, its coefficient increasing (b = 3.445, p = 0.006). Therefore, nificantly associated with social cognition. Regarding baseline the conditions for the mediation of social cognition were not sat- cognitive domain scores, after the Bonferroni correction, only isfied. Regarding baseline verbal memory, this domain lost its sig- processing speed and verbal memory showed a significant associ- nificant association with functioning at 2 years when including ation with social cognition at baseline (path a). Social cognition social cognition at 2 years in the model (path c´), while the latter was not significantly associated with functioning at baseline had a significant association with functioning (path b) (Fig. 3). (path b). Neither cognitive reserve nor the cognitive domains Thus, social cognition at 2 years appears to be a significant full were related to functioning at baseline (path c). mediator, it mediating 25.2% of the effect (95% bootstrapped At 2 years of follow-up, in addition to baseline cognitive CI −4.731 to −0.605). reserve and verbal memory, baseline executive function was also associated with social cognition at 2 years, while processing speed became non-significant at this time point (path a). Discussion Furthermore, social cognition at both baseline and 2 years was significantly associated with functioning at the end of the The main finding of this study is that social cognition acted as a follow-up (path b). Cognitive reserve was significantly related to mediator between cognitive reserve and functioning at 2 years of functioning at 2 years. Regarding baseline cognitive domain follow-up. Likewise, social cognition was a mediator between ver- scores, after applying the Bonferroni correction, only attention bal memory and functional outcome. These findings suggest that and verbal memory were significantly related to functioning at social cognition is involved and indeed plays a crucial role in the 2 years (path c). functional outcome of FEP patients. There is little previous research on this mediation, although in a systematic review Schmidt et al. (2011) also found that social cognition mediated the relationship between neurocognition and functioning in The influence of social cognition as a mediator between schizophrenia. Our findings are from a sample of patients with cognitive reserve and cognitive domains on functioning FEP and take into account the role of cognitive reserve. At baseline, as neither cognitive reserve nor cognitive domain Moreover, our study has a longitudinal design. scores were related to functioning (path c), the conditions for There is extensive evidence of social cognitive deficits in FEP mediation were not satisfied. (Horan et al., 2012; Bora and Pantelis, 2013; Bliksted et al., https://doi.org/10.1017/S0033291719002794 Published online by Cambridge University Press 2706 I González-Ortega et al. Table 4. Statistical analyses of the paths a and b Path a (cognitive domains/CR at baseline → social cognition) Social cognition at baseline Social cognition at 2 years Cognitive reserve b = 0.462, p < 0.001 b = 0.491, p < 0.001 Attention b = −0.016, p = 0.461 b = 0.057, p = 0.463 Processing speed b = 0.159, p = 0.004 b = 0.138, p = 0.051 Verbal memory b = 0.148, p = 0.002 b = 0.185, p = 0.003 Working memory b = 0.146, p = 0.051 b = 0.206, p = 0.039 Executive function b = 0.420, p = 0.059 b = 0.910, p = 0.006 Path b (social cognition → FAST) FAST at baseline FAST at 2 years Social cognition at baseline b = −1.862, p = 0.077 b = −3.410, p = 0.006 Social cognition at 2 years – b = −3.347, p = 0.007 Path c (cognitive domains/CR at baseline → FAST) FAST at baseline FAST at 2 years Cognitive reserve b = −2.105, p = 0.195 b = −3.849, p = 0.048 Attention b = 0.125, p = 0.725 b = 3.257, p = 0.005 Processing speed b = −1.347, p = 0.125 b = −0.961, p = 0.364 Verbal memory b = −1.459, p = 0.059 b = −2.451, p = 0.007 Working memory b = −2.881, p = 0.014 b = −2.062, p = 0.162 Executive function b = −6.669, p = 0.057 b = −9.594, p = 0.016 p value <0.05 except for the cognitive domains for which the Bonferroni correction was applied ( p < 0.01). Bold values represent the significant values. Fig. 2. Mediation of social cognition between cognitive reserve and functioning at 2 Fig. 3. Mediation of social cognition between verbal memory and functioning at 2 years. years. 2014; Langdon et al., 2014; Ho et al., 2015) and related conditions could be a better predictor of functional outcome than neurocog- (Lahera et al., 2015; Varo et al., 2017). Impairments in social nition with a significant impact on general functioning of patients functioning have been related to deficits in emotion recognition, (Fett et al., 2011; Ohmuro et al., 2016). In our sample of FEP emotional intelligence/social perception, and social knowledge patients, although baseline social cognition was related to func- of FEP patients (Addington et al., 2010), because of their difficul- tioning at 2 years, it was not a mediator between cognition and ties managing specific situations and predicting social conse- functional outcome; however, social cognition 2 years later was quences, as well as in recognizing changes in behavioral a significant mediator. This finding is probably related to the meaning (Montreuil et al., 2010). Moreover, it is considered state and trait nature of social cognition. Social dysfunction has that deterioration in social cognition has been related to poor been described as a state marker related to symptoms of schizo- functioning, especially poor work and social functioning as well phrenia, and some studies suggest that it is not impaired after as interpersonal and problem solving skills (Addington et al., symptom recovery (Corcoran et al., 1995; Drury et al., 1998). 2010; Fett et al., 2011; Horan et al., 2012). Nonetheless, several studies have demonstrated deficits in social A second major finding of our study is that social cognition cognition in remitted patients (Herold et al., 2002; Janssen may be a potential marker for predicting the functional outcome et al., 2003; Inoue et al., 2006). It is possible that patients have of patients. Previous studies have also found that social cognition social cognition deficits as a trait marker, and in addition, these may be a trait marker of psychosis (Bora et al., 2008; Bora and deficits may worsen with symptoms, indicating that it is both a Pantelis, 2013). In fact, it has been seen that social cognition trait and a state variable. Moreover, social cognitive deficits are https://doi.org/10.1017/S0033291719002794 Published online by Cambridge University Press Psychological Medicine 2707 already present even in patients at risk of psychosis (Thompson capacity to use compensatory forms and inhibit abnormal neural et al., 2012; Ohmuro et al., 2016; Zhang et al., 2016) and deficits processing, which translates into greater insight and treatment have even been found in unaffected first-degree relatives (Bora adherence. Other studies also found that cognitive reserve can and Pantelis, 2013), this highlighting its trait properties. predict both the risk of illness as symptoms remit and functional On the other hand, another relevant finding of our study is improvement (Barnett et al., 2006; Forcada et al., 2015; Grande that cognitive reserve and neurocognition predicted social cogni- et al., 2017). Amoretti et al. (2016) have shown that cognitive tion, both at baseline and 2 years. Bliksted et al. (2014) also found reserve predicts neuropsychological outcome, negative symp- a correlation between social cognition and IQ, though they distin- toms, and functioning at 2 years of follow-up, and hence, cogni- guish between a complex version of social cognitive deficits linked tive reserve can be used as a reliable indicator for the course of with IQ and another version related to simpler forms of social FEP patients (Amoretti et al., 2018). cognition that are independent of IQ. Regarding cognitive The findings of our study have important implications for domains, processing speed and verbal memory showed a signifi- clinical practice. An understanding of the association between cant association with social cognition at baseline. At 2 years of neurocognition, cognitive reserve, and functioning and of the follow-up, verbal memory and executive function were associated mediating role of social cognition is important for predicting with social cognition. Previous studies have also found that cog- the outcome of FEP patients after illness onset. It is essential nitive functioning is related to social cognition. Theory of mind to consider the relationship between illness outcome in patients disabilities have been related to general cognitive impairment and their IQ, both premorbid values and changes therein, to (Bora et al., 2008) and more specifically, Fernandez-Gonzalo identify the need for a specific treatment to improve the func- et al. (2013, 2014) showed that executive function was related to tional outcome of FEP patients. Specially, patients with lower theory of mind in FEP patients. Further, Sachs et al. (2004) cognitive reserve may be more likely to need functional rehabili- found a correlation between emotion discrimination and verbal tation to help to improve their prognosis. In addition, people memory and language processing cognitive domains. with higher cognitive reserve can be oriented to more ‘normal’ As we hypothesized, at 2 years of follow-up there was an work or education. Likewise, it is essential that the treatment improvement in social cognition, cognitive functioning, and gen- of these patients focuses on social cognitive deficits to improve eral functioning of patients. There is extensive evidence of the their functional outcome. involvement of cognitive impairment in the functional outcome This study has some limitations and strengths that should be of FEP patients (Lepage et al., 2014; Torgalsbøen et al., 2015). considered. Our study had a follow-up of 2 years, and a longer Neurocognitive deficits have been identified as a core symptom follow-up period should be considered in future research. of schizophrenia and may be more related to functional recovery Another limitation is the variability across studies in how IQ is of patients than clinical symptoms (Lepage et al., 2014). It has measured. Although we used criteria based on previous studies, been suggested that cognitive impairment appears soon after or there is no specific instrument validated to measure cognitive even before the first episode of psychosis (Bora and Murray, reserve. Finally, future studies should include a control group. 2014), and while some studies argue that there is deterioration The study also has important strengths. First, the results can be in cognitive functions in the follow-up (Morrison et al., 2006; considered generalizable because it is a multicenter study that Arango et al., 2012), others postulate an improvement in cognitive included a large sample of patients recruited in numerous function (Rodríguez-Sánchez et al., 2008; Becker et al., 2010). Our Spanish psychiatric admission centers for acute psychosis. findings suggest that cognitive impairment in FEP is not progres- Second, the wide age range of the patients means that the sample sive after the onset of illness, but rather improves in the follow-up, is representative of the target population, with an average age which lends support to the neurodevelopmental model of cogni- (23.63 ± 5.9 years), lower than that for other studies with large tion. Although the effects of treatments were not analyzed in the FEP cohorts which did not include child and adolescent patients study, it is possible that treatment effects and/or clinical stabiliza- (OPUS trial: 26.6 ± 6.4; Bertelsen et al., 2008; and EUFEST trial: tion contribute to cognitive improvements in FEP. In our sample 26 ± 5.6; Kahn et al., 2008). Third, the neuropsychological battery of FEP patients, cognitive functioning improved significantly at used was extensive and covered the areas proposed by the the 2-year follow-up in all cognitive domains, except working NIMH-MATRICS consensus (except visual memory) (Kern memory and executive function. Similarly, Lepage et al. (2014) et al., 2008; Nuechterlein et al., 2008; Nuechterlein and Green, found that verbal and working memory are the cognitive domains 2009). most affected in FEP. This suggests that specific cognitive To conclude, although further studies are required, our find- domains may be more sensitive than others for predicting long- ings suggest that cognitive reserve and neurocognition are related term functional outcome, that is, some may be more impaired to functioning, and social cognition plays a mediating role. and may be more closely related to patient outcome. Specifically, in our FEP sample, verbal memory predicted the Supplementary material. The supplementary material for this article can be found at https://doi.org/10.1017/S0033291719002794 functional outcome at the 2-year follow-up. In relation to this, Torgalsbøen et al. (2015) found that attention/vigilance and ver- Acknowledgements. We would like to thank the authors of the PEPs group bal learning predicted remission at 6 months of follow-up, who participated in the development of this manuscript. whereas attention/vigilance and working memory predicted func- tional outcome. Financial support. This work was supported by the Carlos III Institute of Health and European Fund for Regional Development (PI08/1213, PI11/ Cognitive reserve was also associated with functional out- 01977, PI14/01900, PI08/01026, PI11/02831, PI14/01621, PI08/1161, PI16/ come. This result has also been found in other studies. 00359, PI16/01164, PI18/00805), the Basque Foundation for Health Innovation Specifically, it has been concluded that patients with a better and Research (BIOEF), the Secretaria d´Universitats I Recerca del cognitive reserve will have better outcomes in follow-up. Departament d´Economia I Coneixement (2017 SGR 1365), and R&D activities Barnett et al. (2006) concluded that patients with better cogni- in Biomedicine, Madrid Regional Government and Structural Funds of the tive reserve could have better reasoning skills and functional European Union (S2017/BMD-3740 (AGES-CM 2-CM)). https://doi.org/10.1017/S0033291719002794 Published online by Cambridge University Press 2708 I González-Ortega et al. Conflict of interest. A González-Pinto has received grants from or served as Bernardo M, Bioque M, Parellada M, Saiz-Ruiz J, Cuesta MJ, Llerena A, consultant, advisor or CME speaker for the following entities: Almirall, Sanjuán J, Castro-Fornieles J, Arango C and Cabrera B, PEPs Group AstraZeneca, Bristol-Myers Squibb, Cephalon, Eli Lilly, Glaxo-Smith-Kline, (2013) Assessing clinical and functional outcomes in a gene-environment Janssen-Cilag, Ferrer, Johnson & Johnson, Lundbeck, Merck, Otsuka, Pfizer, interaction study in first episode of psychosis (PEPs). Revista de Sanofi-Aventis, Servier, Shering-Plough, Solvay, the Spanish Ministries of Psiquiatría y Salud Mental 6, 4–16. Science and Innovation (through CIBERSAM) and of Science (through the Bertelsen M, Jeppesen P, Petersen L, Thorup A, Ø hlenschlaeger J, le Carlos III Institute of Health), the Basque Government, the Stanley Medical Quach P, Christensen TØ, Krarup G, Jørgensen P and Nordentoft M Research Institute, and Wyeth. (2008) Five-year follow-up of a randomized multicenter trial of intensive C Arango has been a consultant to or has received fees or grants from Abbot, early intervention vs. standard treatment for patients with a first episode Amgen, AstraZeneca, Bristol-Myers-Squibb, Caja Navarra, CIBERSAM, the of psychotic illness: the OPUS trial. Archives of General Psychiatry 65, Alicia Koplowitz Foundation, the Carlos III Institute of Health, Janssen Cilag, 762–771. Lundbeck, Merck, the Spanish Ministries of Science and Innovation, of Bliksted V, Fagerlund B, Weed E, Frith C and Videbech P (2014) Social cog- Health, and of Economy and Competitiveness, Mutua Madrileña, Otsuka, nition and neurocognitive deficits in first-episode schizophrenia. Pfizer, Roche, Servier, Shire, Takeda, and Schering-Plough. Schizophrenia Research 153, 9–17. E Vieta has received grants from or served as consultant, advisor or CME Bonnín CM, Martínez-Arán A, Reinares M, Valentí M, Solé B, Jiménez E, speaker for the following entities: AB-Biotics, Allergan, Angelini, Ferrer, Montejo L, Vieta E and Rosa AR (2018) Thresholds for severity, remission Gedeon Richter, Janssen, Lundbeck, Otsuka, Sanofi-Aventis, Sunovion, the and recovery using the functioning assessment short test (FAST) in bipolar Carlos III Institute of Health, the European Union’s 7th Framework Program disorder. Journal of Affective Disorders 240, 57–62. and Horizon 2020, the Brain and Behaviour Foundation (NARSAD), and the Bora E and Murray RM (2014) Meta-analysis of cognitive deficits in ultra- Stanley Medical Research Institute. high risk to psychosis and first-episode psychosis: do the cognitive deficits R Rodriguez-Jimenez has been a consultant for, spoken in activities of, or progress over, or after, the onset of psychosis? Schizophrenia Bulletin 40, received grants from: the Carlos III Institute of Health, the Spanish Health 744–755. Research Fund (FIS), CIBERSAM, Madrid Regional Government (S2010/ Bora E and Pantelis C (2013) Theory of mind impairments in first-episode BMD-2422 AGES), JanssenCilag, Lundbeck, Otsuka, Pfizer, Ferrer, Juste, and psychosis, individuals at ultra-high risk for psychosis and in first-degree rela- Takeda tives of schizophrenia: systematic review and meta-analysis. Schizophrenia The other authors have no conflicts of interest to declare. Research 144, 31–36. Bora E, Gokcen S, Kayahan B and Veznedaroglu B (2008) Deficits of social- cognitive and social-perceptual aspects of theory of mind in remitted patients with schizophrenia. Effect of residual symptoms. The Journal of References Nervous and Mental Disease 196, 95–99. Addington J, Girard TA, Christensen BK and Addington D (2010) Social Conners CK (2004) Continuous Performance Test II. Toronto, Canada: cognition mediates illness-related and cognitive influences on social func- Multi-Health Systems Inc. tion in patients with schizophrenia-spectrum disorders. Journal of Corcoran R, Mercer G and Frith CD (1995) Schizophrenia, symptomatology Psychiatry and Neurosciences 35, 49–54. and social inference: investigating ‘theory of mind’ in people with schizo- American Psychiatric Association (1994) Diagnostic and Statistical Manual of phrenia. Schizophrenia Research 17, 5–13. Mental Disorders (DSM-IV), 4th Edn. Washington, DC: APA. Cuesta MJ, Sánchez-Torres AM, Cabrera B, Bioque M, Merchán-Naranjo J, Amoretti S, Bernardo M, Bonnin CM, Bioque M, Cabrera B, Mezquida G, Corripio I, González-Pinto A, Lobo A, Bombín I, de la Serna E, Solé B, Vieta E and Torrent C (2016) The impact of cognitive reserve in Sanjuan J, Parellada M, Saiz-Ruiz J and Bernardo M, PEPs Group the outcome of first-episode psychoses: 2-year follow-up study. European (2015) Premorbid adjustment and clinical correlates of cognitive impair- Neuropsychopharmacology 26, 1638–1648. ment in first-episode psychosis. The PEPsCog Study. Schizophrenia Amoretti S, Cabrera B, Torrent C, Mezquida G, Lobo A, González-Pinto A, Research 164, 5–73. Parellada M, Corripio I, Vieta E, de la Serna E, Butjosa A, Contreras F, Drury VM, Robinson EJ and Birchwood M (1998) ‘Theory of mind’ skills dur- Sarró S, Penadés R, Sánchez-Torres AM, Cuesta M, Bernardo M and ing an acute episode of psychosis and following recovery. Psychological PEPsGroup (2018) Cognitive reserve as an outcome predictor: first-episode Medicine 28, 1101–1112. affective versus non-affective psychosis. Acta Psychiatrica Scandinavica 138, Fernandez-Gonzalo S, Pousa E, Jodar M, Turon M, Duño R and Palao D 441–455. (2013) Influence of the neuropsychological functions in theory of mind Anaya C, Torrent C, Caballero FF, Vieta E, Bonnin CM and Ayuso- in schizophrenia: the false-belief/deception paradigm. The Journal of Mateos JL, CIBERSAM Functional Remediation Group (2016) Cognitive Nervous and Mental Disease 201, 609–613. reserve in bipolar disorder: relation to cognition, psychosocial functioning Fernandez-Gonzalo S, Jodar M, Pousa E, Turon M, Garcia R, Rambla CH and quality of life. Acta Psychiatrica Scandinavica 133, 386–398. and Palao D (2014) Selective effect of neurocognition on different theory of Arango C, Rapado-Castro M, Reig S, Castro-Fornieles J, Gonzalez-Pinto A, mind domains in first-episode psychosis. Journal of Nervous & Mental Otero S, Baeza I, Moreno C, Graell M, Janssen J, Parellada M, Moreno D, Disease 202, 576–582. Bargallo N and Desco M (2012) Progressive brain changes in children and Fett AK, Viechtbauer W, Dominguez MD, Penn DL, van Os J and adolescents with first-episode psychosis. Archives of General Psychiatry 69, Krabbendam L (2011) The relationship between neurocognition and social 16–26. cognition with functional outcomes in schizophrenia: a meta-analysis. Barbato M, Liu L, Penn DL, Keefe RSE, Perkins DO, Woods SW and Neuroscience & Biobehavioral Reviews 35, 573–588. Addington J (2013) Social cognition as a mediator between neurocognition First MB, Spitzer R and Gibbon M (1997) Structured Clinical Interview for and functional outcome in individuals at clinical high risk for psychosis. DSM-IV Axis I Disorders. Washington, DC: American Psychiatric Press Inc. Schizophrenia Research 150, 542–546. Forcada I, Mur M, Mora E, Vieta E, Bartrés-Faz D and Portella MJ (2015) Barnett JH, Salmond CH, Jones PB and Sahakian BJ (2006) Cognitive The influence of cognitive reserve on psychosocial and neuropsychological reserve in neuropsychiatry. Psychological Medicine 36, 1053–1064. functioning in bipolar disorder. European Neuropsychopharmacology 25, Becker HE, Nieman DH, Wiltink S, Dingemans PM, van de Fliert JR, 214–222. Velthorst E and Linszen DH (2010) Neurocognitive functioning before Frazier PA, Tix AP and Barron KE (2004) Testing moderator and mediator and after the first psychotic episode: does psychosis result in cognitive effects in counseling psychology research. Journal of Counseling Psychology deterioration? Psychological Medicine 40, 1599–1606. 51, 115–134. Benedet MJ and Alejandre MA (1998) Test de Aprendizaje Verbal España- Golden C (1978) The Stroop Color and Word Test: A Manual for Clinical and Complutense (TAVEC). Madrid, Spain: Tea Ediciones. Experimental Uses. Chicago: Stoelting Co. https://doi.org/10.1017/S0033291719002794 Published online by Cambridge University Press Psychological Medicine 2709 González-Ortega I, Rosa A, Alberich S, Barbeito S, Vega P, Echeburúa E, contribute to poor social functioning. Acta Psychiatrica Scandinavica. Vieta E and González-Pinto A (2010) Validation and use of the function- 131, 472–482. ing assessment short test in first psychotic episodes. The Journal of Nervous Langdon R, Still M, Connors MH, Ward PB and Catts SV (2014) and Mental Disease 198, 836–840. Theory of mind in early psychosis. Early Intervention in Psychiatry 8, González-Ortega I, de Los Mozos V, Echeburúa E, Mezo M, Besga A, Ruiz 286–290. de Azúa S, González-Pinto A, Gutierrez M, Zorrilla I and González- Leeson VC, Sharma P, Harrison M, Ron MA, Barnes TR and Joyce EM Pinto A (2013) Working memory as a predictor of negative symptoms (2011) IQ trajectory, cognitive reserve, and clinical outcome following a and functional outcome in first episode psychosis. Psychiatry Research first episode of psychosis: a 3-year longitudinal study. Schizophrenia 206, 8–16. Bulletin 37, 768–777. Grande I, Sanchez-Moreno J, Sole B, Jimenez E, Torrent C, Bonnin CM, Lepage M, Bodnar M and Bowie CR (2014) Neurocognition: clinical and func- Varo C, Tabares-Seisdedos R, Balanzá-Martínez V, Valls E, Morilla I, tional outcomes in schizophrenia. The Canadian Journal of Psychiatry 59, Carvalho AF, Ayuso-Mateos JL, Vieta E and Martinez-Aran A (2017) 5–12. High cognitive reserve in bipolar disorders as a moderator of neurocognitive Lin CH, Huang CL, Chang YC, Chen PW, Lin CY, Tsai GE and Lane HY impairment. Journal of Affective Disorders 208, 621–627. (2013) Clinical symptoms, mainly negative symptoms, mediate the influ- Green MF and Harvey PD (2014) Cognition in schizophrenia: past, present, ence of neurocognition and social cognition on functional outcome of and future. Schizophrenia Research: Cognition 1, e1–e9. schizophrenia. Schizophrenia Research 146, 231–237. Heaton RK, Chelune GJ, Talley JL, Kay GG and Curtiss G (1993) Wisconsin Loonstra AS, Tarlow AR and Sellers AH (2001) COWAT metanorms across Card Sorting Test Manual (Revised and Expanded). Odessa, FL: PAR age, education, and gender. Applied Neuropsychology 8, 161–166. Psychological Assessment Resources, Inc. Mayer JD, Salovey P and Caruso DR (2009) Mayer–Salovey– Hedman AM, van Haren NE, van Baal CG, Kahn RS and Hulshoff Pol HE Caruso Emotional Intelligence Test (Spanish Version). Madrid: TEA (2013) IQ change over time in schizophrenia and healthy individuals: a Ediciones. meta-analysis. Schizophrenia Research 146, 201–208. Montreuil T, Bodnar M, Bertrand MC, Malla AK, Joober R and Lepage M Herold R, Tenyi T, Lenard K and Trixler M (2002) Theory of mind deficit in (2010) Social cognitive markers of short-term clinical outcome in people with schizophrenia during remission. Psychological Medicine 32, first-episode psychosis. Clinical Schizophrenia & Related Psychoses 4, 1125–1129. 105–114. Ho KK, Lui SS, Hung KS, Wang Y, Li Z, Cheung EF and Chan RC (2015) Morrison G, O’Carroll R and McCreadie R (2006) Long-term course of cog- Theory of mind impairments in patients with first-episode schizophrenia nitive impairment in schizophrenia. The British Journal of Psychiatry 189, and their unaffected siblings. Schizophrenia Research 166, 1–8. 556–557. Hollingshead A and Redlich F (1958) Class and Mental Illness. New York: Nuechterlein KH and Green MF (2009) MATRICS. Consensus cognitive bat- Wiley. tery, norwegian version. In Rund BR and Sundet KS Trans (eds), MATRICS Horan WP, Green MF, DeGroot M, Fiske A, Hellemann G, Kee K, Kern RS, Assessment. Los Angeles, LA: Inc. Lee J, Sergi MJ, Subotnik KL, Sugar CA, Ventura J and Nuechterlein KH Nuechterlein KH, Green MF, Kern RS, Baade LE, Barch DM, Cohen JD, (2012) Social cognition in schizophrenia, Part 2: 12-month stability and Essock S, Fenton WS, Frese III FJ, Gold JM, Goldberg T, Heaton RK, prediction of functional outcome in first-episode patients. Schizophrenia Keefe RS, Kraemer H, Mesholam-Gately R, Seidman LJ, Stover E, Bulletin 38, 865–872. Weinberger DR, Young AS, Zalcman S and Marder SR (2008) The Inoue Y, Yamada K, Hirano M, Shinohara M, Tamaoki T, Iguchi H and MATRICS consensus cognitive battery, part 1: test selection, reliability, Tonooka Y (2006) Impairment of theory of mind in patients in remission and validity. American Journal of Psychiatry 165, 203–213. following first episode of schizophrenia. Journal of European Archives of Ohmuro N, Katsura M, Obara C, Kikuchi T, Sakuma A, Iizuka K, Psychiatry and Clinical Neuroscience 256, 326–328. Hamaie Y, Ito F, Matsuoka H and Matsumoto K (2016) Deficits of cog- Janssen I, Krabbendam L, Jolles J and Os JV (2003) Alterations in theory of nitive theory of mind and its relationship with functioning in individuals mind in patients with schizophrenia and non-psychotic relatives. Acta with an at-risk mental state and first-episode psychosis. Psychiatry Psychiatrica Scandinavica 108, 110–117. Research 243, 318–325. Kahn RS, Fleischhacker WW, Boter H, Davidson M, Davidson M, Peña-Casanova J (1990) Test Barcelona. Barcelona: Masson. Vergouwe Y, Keet IP, Gheorghe MD, Rybakowski JK, Galderisi S, Pinkham AE and Harvey PD (2012) Future directions for social cognitive Libiger J, Hummer M, Dollfus S, López-Ibor JJ, Hranov LG, interventions in schizophrenia. Schizophrenia Bulletin 39, 499–500. Gaebel W, Peuskens J, Lindefors N, Riecher-Rössler A, Grobbee DE Reitan RM and Wolfson D (1985) The Halstead–Reitan Neuropsychological Test and EUFEST study group (2008) Effectiveness of antipsychotic drugs in Battery: Therapy and Clinical Interpretation. Tucson, AZ: Neuropsychological first episode schizophrenia and schizophreniform disorder: an open rando- Press. mized clinical trial. Lancet 371, 1085–1097. Rodríguez-Sánchez JM, Pérez-Iglesias R, González-Blanch C, Pelayo-Terán JM, Kaufman J, Birmaher B, Brent D, Rao U, Flynn C, Moreci P, Williamson D Mata I, Martínez O, Sánchez-Cubillo I, Vázquez-Barquero JL and and Ryan N (1997) Schedule for affective disorders and schizophrenia for Crespo-Facorro B (2008) 1-year follow-up study of cognitive function in first- school-age children-present and lifetime version (K-SADS-PL): initial reli- episode non-affective psychosis. Schizophrenia Research 104, 165–174. ability and validity data. Journal of the American Academy of Child and Rosa AR, Sánchez-Moreno J, Martínez-Aran A, Salamero M, Torrent C, Adolescent Psychiatry 36, 980–988. Reinares M, Comes M, Colom F, Van Riel W, Ayuso-Mateos JL, Kern RS, Nuechterlein KH, Green MF, Baade LE, Fenton WS, Gold JM, Kapczinski F and Vieta E (2007) Validity and reliability of the Keefe RS, Mesholam-Gately R, Mintz J, Seidman LJ, Stover E and Functioning Assessment Short Test (FAST) in bipolar disorder. Clinical Marder SR (2008) The MATRICS consensus cognitive battery, part 2: Practice & Epidemiology in Mental Health 7, 3–5. co-norming and standardization. American Journal of Psychiatry 165, Rosa AR, Reinares M, Amann B, Popovic D, Franco C, Comes M, 214–220. Torrent C, Bonnín CM, Solé B, Valentí M, Salamero M, Kapczinski F Kokkevi A and Hartgers C (1995) EuropASI: European adaptation of a multi- and Vieta E (2011) Six-month functional outcome of a bipolar disorder dimensional assessment instrument for drug and alcohol dependence. cohort in the context of a specialized-care program. Bipolar Disorders 13, European Addiction Research 1, 208–210. 679–686. Kravariti E, Morris RG, Rabe-Hesketh S, Murray RM and Frangou S Rotger S, Richarte V, Nogueira M, Corrales M, Bosch R, Vidal R, Marfil L, (2003) The Maudsley early onset schizophrenia study: cognitive function Valero S, Vieta E, Goikolea JM, Torres I, Rosa A, Mur M, Casas M and in adolescents with recent onset schizophrenia. Schizophrenia Research Ramos-Quiroga JA (2014) Functioning Assessment Short Test (FAST): 61, 137–148. validity and reliability in adults with attention-deficit/hyperactivity dis- Lahera G, Herrera S, Reinares M, Benito A, Rullas M, González-Cases J and order. European Archives of Psychiatry and Clinical Neuroscience 264, Vieta E (2015) Hostile attributions in bipolar disorder and schizophrenia 719–727. https://doi.org/10.1017/S0033291719002794 Published online by Cambridge University Press 2710 I González-Ortega et al. Sachs G, Steger-Wuchse D, Kryspin-Exner I, Gur RC and Katschnig H Varo C, Jiménez E, Solé B, Bonnín CM, Torrent C, Valls E, Morilla I, (2004) Facial recognition deficits and cognition in schizophrenia. Lahera G, Martínez-Arán A, Vieta E and Reinares M (2017) Social cog- Schizophrenia Research 68, 27–35. nition in bipolar disorder: focus on emotional intelligence. Journal of Schmidt SJ, Mueller DR and Roder V (2011) Social cognition as a mediator Affective Disorders 1, 210–217. variable between neurocognition and functional outcome in schizophrenia: Vieta E, Berk M, Schulze TG, Carvalho AF, Suppes T, Calabrese JR, Gao K, empirical review and new results by structural equation modeling. Miskowiak KW and Grande I (2018) Bipolar disorders. Nature Reviews Schizophrenia Bulletin 37(suppl. 2), S41–S54. Disease Primers 4, 18008. Stern Y (2002) What is cognitive reserve? Theory and research application of Wechsler D (1997) Wechsler Adult Intelligence Scale, 3rd Edn. (Administration the reserve concept. Journal of the International Neuropsychological Society and Scoring Manual). San Antonio, TX, USA: The Psychological 8, 448–460. Corporation. Stirling J, White C, Lewis S, Hopkins R, Tantam D, Huddy A and Wechsler D (2003) Wechsler Intelligence Scale for Children-IV (WISC-IV), 4th Montague L (2003) Neurocognitive function and outcome in first-episode Edn. San Antonio, TX: Psychological Corporation. schizophrenia: a 10-year follow-up of an epidemiological cohort. Zabala A, Rapado M, Arango C, Robles O, de la Serna González C, Schizophrenia Research 65, 75–86. Rodríguez-Sanchez JM, Andrés P, Mayoral M and Bombín I (2010) Thompson A, Papas A, Bartholomeusz C, Allott K, Amminger GP, Neuropsychological functioning in early-onset first-episode psychosis: com- Nelson B, Wood S and Yung A (2012) Social cognition in clinical “at parison of diagnostic subgroups. European Archives of Psychiatry and risk” for psychosis and first episode psychosis populations. Schizophrenia Clinical Neuroscience 260, 225–233. Research 141, 204–209. Zhang T, Yi Z, Li H, Cui H, Tang Y, Lu X, Xu L, Qian Z, Zhu Y, Jiang L, Torgalsbøen AK, Mohn C, Czajkowski N and Rund BR (2015) Relationship Chow A, Li C, Jiang K, Xiao Z and Wang J (2016) Faux pas recognition between neurocognition and functional recovery in first-episode schizo- performance in a help-seeking population at clinical high risk of psych- phrenia: results from the second year of the Oslo multi-follow-up study. osis. European Archives of Psychiatry and Clinical Neurosciences 266, Psychiatry Research 227, 185–191. 71–78. https://doi.org/10.1017/S0033291719002794 Published online by Cambridge University Press