Comprehensive Psychiatry 109 (2021) 152258 Contents lists available at ScienceDirect Comprehensive Psychiatry journal homepage: www.elsevier.com/locate/comppsych Group and sex differences in social cognition in bipolar disorder, schizophrenia/schizoaffective disorder and healthy people Guillem Navarra-Ventura a, b, c, Muriel Vicent-Gil a, Maria Serra-Blasco a, d, Carmen Massons a, Josep Maria Crosas a, Jesús Cobo a, d, Abigail Jubert a, Mercè Jodar d, e, f, *, Sol Fernández-Gonzalo a, d, f, Ximena Goldberg a, d, Diego Palao a, b, d, Guillermo Lahera d, g, Eduard Vieta d, h, Narcís Cardoner a, b, d, * a Department of Mental Health, Hospital Universitari Parc Taulí, Institut d'Investigació i Innovació Parc Taulí I3PT, Universitat Autònoma de Barcelona, Sabadell, Catalonia, Spain b Department of Psychiatry and Forensic Medicine, Universitat Autònoma de Barcelona, International Excellence Campus, Bellaterra, Cerdanyola del Vallès, Catalonia, Spain c Centro de Investigación Biomédica en Red de Enfermedades Respiratorias (CIBERES), Instituto de Salud Carlos III, Madrid, Spain d Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Instituto de Salud Carlos III, Madrid, Spain e Department of Neurology, Hospital Universitari Parc Taulí, Institut d'Investigació i Innovació Parc Taulí I3PT, Universitat Autònoma de Barcelona, Sabadell, Catalonia, Spain f Department of Clinical and Health Psychology, Universitat Autònoma de Barcelona, International Excellence Campus, Bellaterra, Cerdanyola del Vallès, Catalonia, Spain g Faculty of Medicine and Health Sciences, Universidad de Alcalá, Instituto Ramón y Cajal de Investigación Sanitaria IRYCIS, Madrid, Spain h Institute of Neuroscience, Hospital Clínic, Institut d'Investigacions Biomèdiques August Pi i Sunyer IDIBAPS, Universitat de Barcelona, Barcelona, Catalonia, Spain A R T I C L E I N F O A B S T R A C T Keywords: Background: Impairment of social cognition is documented in bipolar disorder (BD) and schizophrenia/schizo­ Sex differences affective disorder (SCH). In healthy individuals, women perform better than men in some of its sub-domains. Emotion recognition However, in BD and SCH the results are mixed. Our aim was to compare emotion recognition, affective The­ Theory of mind ory of Mind (ToM) and first- and second-order cognitive ToM in BD, SCH and healthy subjects, and to investigate Bipolar disorder sex-related differences. Schizophrenia Healthy people Methods: 120 patients (BD = 60, SCH = 60) and 40 healthy subjects were recruited. Emotion recognition was Subthreshold depression assessed by the Pictures of Facial Affect (POFA) test, affective ToM by the Reading the Mind in the Eyes Test (RMET) and cognitive ToM by several false-belief stories. Group and sex differences were analyzed using parametric (POFA, RMET) and non-parametric (false-belief stories) tests. The impact of age, intelligence quotient (IQ) and clinical variables on patient performance was examined using a series of linear/logistic regressions. Results: Both groups of patients performed worse than healthy subjects on POFA, RMET and second-order false- belief (p < 0.001), but no differences were found between them. Instead, their deficits were related to older age and/or lower IQ (p < 0.01). Subthreshold depression was associated with a 6-fold increased risk of first-order false-belief failure (p < 0.001). Sex differences were only found in healthy subjects, with women out­ performing men on POFA and RMET (p ≤ 0.012), but not on first/second-order false-belief. Limitations: The cross-sectional design does not allow for causal inferences. Conclusion: BD and SCH patients had deficits in emotion recognition, affective ToM, and second-order cognitive ToM, but their performance was comparable to each other, highlighting that the differences between them may be subtler than previously thought. First-order cognitive ToM remained intact, but subthreshold depression altered their normal functioning. Our results suggest that the advantage of healthy women in the emotional and affective aspects of social cognition would not be maintained in BD and SCH. * Corresponding authors at: Departments of Neurology and Mental Health, Hospital Universitari Parc Taulí, Institut d'Investigació i Innovació Parc Taulí I3PT, Universitat Autònoma de Barcelona, Sabadell, Catalonia, Spain. E-mail addresses: mjodar@tauli.cat (M. Jodar), ncardoner@tauli.cat (N. Cardoner). https://doi.org/10.1016/j.comppsych.2021.152258 Available online 5 July 2021 0010-440X/© 2021 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). G. Navarra-Ventura et al. Comprehensive Psychiatry 109 (2021) 152258 1. Introduction results for first- and second order ToM are contradictory. While some studies found better performance in BD patients than in SCH patients Social cognition, defined as the ability to recognize, understand, and [37,40], at least on second-order skills [32], others found no difference interpret one's own and others' emotions, thoughts, beliefs, and feelings [46,47]. [1], is crucial for successful functioning at work and in the community In patients with BD, little is known about the effect of sex on emotion [2,3]. In patients with bipolar disorder (BD) and schizophrenia/schiz­ recognition and ToM. To date, only one meta-analysis [25] and another oaffective disorder (SCH), mild to severe deficits in this domain have more recent study [12] have specifically explored the effect of sex on been described throughout the course of the illness, including phases of emotion recognition, providing conflicting results, while no study has clinical stability [4,5]. There is evidence that these deficits tend to yet examined the relationship between sex, affective ToM and cognitive remain fairly stable in most patients [6,7]. However, variables such as ToM. In patients with SCH, one review [55] and one meta-analysis [28] age, general intelligence, residual symptoms, psychotropic drugs and, in found no effect of sex on social cognition. However, a study not included BD patients, history of psychosis, may modulate their severity [8–12]. in these publications and two others published shortly thereafter show Social cognition is a multidimensional construct that includes that women perform better than men in emotion recognition [56,57] different sub-domains, namely emotional processing, Theory of Mind and affective ToM [58]. The overrepresentation of male patients in some (ToM), social perception, social knowledge, and attributional bias. cohorts [9,12,50,57,59] is a recurrent limitation in the literature [55] Within these sub-domains, facial emotion recognition (emotional pro­ that could compromise the generalizability of current knowledge to fe­ cessing), mental state decoding (ToM) and mental state reasoning (ToM) male patients. The use of a composite variable including measures of have been the most explored functions so far [1,4–6], and are considered emotion recognition, affective ToM, and cognitive ToM in a single index three of the main predictors of social functioning [2,3]. In addition, they may also be a limitation [42,59], as it does not allow testing whether sex are key domains of the ISBD Battery for Assessment of Neurocognition differences in social cognition are domain- or construct-specific. [13] and the MATRICS Consensus Cognitive Battery [14] and have It has recently been speculated that the advantage of healthy women recently been included as treatment targets in both disorders [15]. in emotion recognition and affective ToM may be maintained in women In healthy individuals, sex plays a crucial role in emotion recognition with SCH [56,58,60], and that this may be related to their better clinical and ToM, which, in turn, can be dissociated into two independent but outcomes compared to men with SCH [58,61]. However, this contrasts interacting processes [1,16]: affective ToM (decoding of complex emo­ with two meta-analyses in patients with BD and SCH that found that the tions, i.e., feelings) and cognitive ToM (reasoning about thoughts and effect of disease outweighs the effect of sex on emotion recognition beliefs). In general, studies agree that women tend to perform better [25,28]. Several neuroimaging studies indicate that emotion recognition than men in emotion recognition and affective ToM. However, the and ToM share a common neural substrate [1,16]. Therefore, it is extent to which sex modulates cognitive ToM remains unknown possible that the advantage of healthy women in affective ToM is not [17–19]. One possible explanation for this difference is that female maintained in these disorders. However, this hypothesis has not yet been brains are less lateralized than male brains [20], which allows for tested, at least in BD patients. greater communication between the two cerebral hemispheres and, In this study, we tried to overcome some of the limitations of pre­ therefore, better integration between emotional-intuitive and cognitive- vious research by including only clinically stable patients in a male:fe­ analytical processing modes [21]. A complementary explanation comes male ratio of 1:1, using a comprehensive battery with tests of emotion from the different gender roles that men and women have played recognition, affective ToM, and first- and second-order cognitive ToM, throughout evolution. From this perspective, sex differences in empathic and by analyzing the emotional, affective, and cognitive aspects of social behaviors could also stem from the greater prosocial and caregiving cognition separately. Our hypotheses were that BD patients will perform roles that women typically adopt in most cultures [17,18]. intermediate between healthy subjects and SCH patients, that the In patients with SCH, sex differences have been observed in several of advantage of healthy women in emotion recognition and affective ToM its clinical features. Compared to women, men tend to show a higher will be lost in BD and SCH patients, and that clinical variables will incidence of the disorder, an earlier age of onset and a more severe modulate their performance. Our aim was threefold: first, to compare course of the disease [22]. In patients with BD, these differences are emotion recognition, affective ToM, and first- and second-order cogni­ much more diffuse [23]. However, some data suggest that manic epi­ tive ToM in BD, SCH and healthy subjects; second, to examine sex- sodes are more frequent in men and depressive symptoms in women related differences in emotion recognition, affective ToM, and first- [24]. and second-order cognitive ToM in each of the three groups; and third, Deficits in emotion recognition and ToM are a well-established to explore the modulatory effect of clinical variables on these sub- finding in BD [25,26] and SCH [27,28]. One of the most salient issues domains of social cognition. among studies comparing the two disorders is the question of whether these deficits are of equal magnitude [29–31]. In general, studies agree 2. Material and methods that BD patients tend to perform intermediate between SCH patients and healthy individuals [32–46]. However, results are not always concor­ 2.1. Participants and procedure dant [8,47–49], possibly due to methodological differences and short­ comings such as small sample size (n < 20 [8,32,34,35,41,46,47] or n < Sixty patients with BD (30 men, 30 women) and sixty patients with 30 [39,42,44,50] in at least one subgroup of the study), the mix of SCH (30 men, 30 women) participated in this cross-sectional study. They clinically stable and acute patients [32–35,43,48], and the use of were recruited at the outpatient mental health clinic of the Parc Taulí different instruments to assess social cognition [51], in particular University Hospital in Sabadell, Catalonia (Spain), between 2016 and cognitive ToM. 2019. To be enrolled, patients had to be clinically stable, which was In this regard, it should be noted that cognitive ToM is not a unitary defined as: having been on follow-up treatment for the past 3 months, domain. In fact, it encompasses different sub-processes [52–54], not having suffered any exacerbation of symptoms during that period, including first-order skills (what I think another person thinks or be­ and not having changed psychotropic drug regimen during the past lieves), second-order skills (what I think another person thinks about month (including antipsychotics and mood stabilizers/anticonvulsants). what a third party thinks or believes) and other higher order skills (e.g., Inclusion criteria for patients with BD were: DSM-IV-TR diagnosis of understanding metaphors, irony, or sarcasm). To date, most studies BD type I/II [62], score ≤ 6 on the Young Mania Rating Scale (YMRS), comparing the two disorders have focused on analyzing higher order and score ≤ 14 on the Hamilton Depression Rating Scale (HAM–D) ToM [36,41–43,45,48]. All but two of these studies show that BD pa­ [63]. For patients with SCH, inclusion criteria were: DSM-IV-TR diag­ tients tend to perform better than SCH patients [42,48]. However, the nosis of schizophrenia/schizoaffective disorder [62], score ≤ 3 on items 2 G. Navarra-Ventura et al. Comprehensive Psychiatry 109 (2021) 152258 P1 (delusions), P2 (conceptual disorganization) and P3 (hallucinatory The examiner read each story aloud and participants had to answer two behavior) of the Positive and Negative Syndrome Scale (PANSS) [63], questions. The first (ToM) question concerned the subject's false belief and score ≤ 7 on the Calgary Depression Scale for Schizophrenia (CDSS) about the situation. The second (control) question was intended to assess [64]. the subject's comprehension of the story. Results are presented in per­ The following were considered exclusion criteria: age outside the centages of failure vs. non-failure. A higher non-failure score means 18–64 range, any concomitant Axis I/II disorder, substance abuse/ better reasoning about the thoughts and beliefs of others (range dependence in the past 6 months (excluding nicotine and caffeine), any 0–100%). medical or neurological disorder associated with cognitive impairment (including brain damage), electroconvulsive therapy in the past 12 2.4. Data preprocessing months, or intelligence quotient (IQ) ≤70. Complementarily, forty healthy subjects (20 men, 20 women) were For the purposes of statical analysis, daily doses of antipsychotic recruited, matched by age and years of education with the patients. They drugs were converted to chlorpromazine equivalents [76], antidepres­ were recruited from healthy companions of non-psychiatric patients sant drugs to fluoxetine equivalents [77], and benzodiazepine drugs to attending the Parc Taulí University Hospital and from other community diazepam equivalents [78]. sources. Exclusion criteria were the same as for the patients, with the At the time of evaluation, about one-fifth of patients (n = 26/120) addition that they had no history of any Axis I/II disorder. Individuals had subthreshold depressive symptoms. To analyze the impact of these with first-degree relatives diagnosed with bipolar disorder type I/II, symptoms on social cognition we created a dichotomous variable schizophrenia/ schizoaffective disorder, or autism spectrum disorder [63,64]: “Subthreshold depression” was defined by a HAM-D score of were also excluded. 8–14 or a CDSS score of 4–7, and “No depression” by a HAM-D score of The study was approved by the Institutional Review Board of the <8 or a CDSS score of <4. Parc Taulí University Hospital (#2017/579) and was conducted in To measure the percentages of failure vs. non-failure in cognitive accordance with the latest version of the Declaration of Helsinki. All ToM, we created two dichotomous variables (one for each order of participants were informed about the characteristics of the study and cognitive ToM) [56]. These variables were categorized as 0 (no ToM gave written informed consent prior to enrollment. Inclusion/exclusion failure) when the participant correctly answered all ToM and control criteria were confirmed by reviewing electronic medical records and questions of the two same order false-belief stories, or 1 (ToM failure), interviewing all participants using a semi-structured clinical interview when the participant incorrectly answered the ToM questions but based on DSM-IV-TR criteria [62]. correctly answered the control questions of the two same order false- belief stories. No participant failed the control questions of the first- 2.2. Clinical evaluation order false-belief stories. However, 5 BD patients (4 men, 1 women), 8 SCH patients (6 men, 2 women) and 1 healthy woman failed the control In addition to collecting demographic data such as sex, age and years question(s) of the second-order false-belief stories and were excluded of education, the clinical evaluation included administration of the from the corresponding analyses. This strict categorization allowed us to YMRS [65] and HAM-D [66] in BD patients and the PANSS [67] and control for the possible confounding effect of neurocognitive deficits (e. CDSS [68] in SCH patients. In patients with SCH, the CDSS was used g., comprehension difficulties), as only participants who correctly instead of the HAM-D because it allows better discrimination of answered the control questions were analyzed. depressive symptoms from negative symptoms [69]. Age of onset of the disorder, duration of illness, total number of hospitalizations, history of 2.5. Statistical analysis psychosis (only in BD patients) and psychotropic drugs were also collected. All analyses were performed using SPSS v.19.0. Statistical signifi­ cance was set at p < 0.05. The normal distribution of data was explored 2.3. IQ and social cognition assessment using the Shapiro-Wilk test. Skewness and kurtosis were also checked as indicators of deviation from normality. When necessary, log10 and IQ was estimated using the Vocabulary subtest of the Wechsler Adult square root transformations were performed to normalize data Intelligence Scale, 3rd edition [70], as it is highly correlated with gen­ distribution. eral intelligence (r = 0.80) [71]. Group and sex differences in continuous demographic, clinical and The Pictures of Facial Affect (POFA [72]; Cronbach's alpha = 0.810 cognitive variables were analyzed using parametric (ANOVA, Student's [73]) was used to assess emotion recognition. Participants were shown t-test) and non-parametric (Kruskal-Wallis H, Mann-Whitney U) tests, as 60 monochromatic photographs of adult male and female faces, each appropriate. For categorical clinical and cognitive variables, the Chi- depicting one of the six basic emotions (disgust, sadness, anger, fear, square test (X2) was used. To control for possible type I errors, un­ surprise, happiness). All stimuli were presented for 5 s along with a 6-op­ planned post hoc analyses (for group differences) and planned multiple tion multiple-choice question and subjects were asked to identify the comparison tests (for sex differences) were corrected with the Bonfer­ emotion displayed in each photograph. A higher score means better roni method. Effect sizes (Cohen's d or Cramer's V) are reported for all recognition of the basic emotions of others (range 0–60 points). significant outcomes. The revised version of the Reading the Mind in the Eyes Test (RMET The impact of clinical variables on patients' social cognitive perfor­ [74]; Cronbach's alpha = 0.735 [75]) was used to assess affective ToM. mance was analyzed using a series of bivariate/binomial regressions, as Participants were shown 36 monochromatic photographs of male and appropriate. Age and estimated IQ were also included in these analyses female gazes without a preset time limit. All stimuli were presented because of their clinical relevance to cognitive performance [9,56]. Each along with a 4-option multiple-choice question and subjects were asked independent variable that reached statistical significance in these to discriminate what the individual in each photograph is thinking or screening analyses was included as a possible factor in their corre­ feeling (e.g., playful, comforting, irritated, or bored). A higher score sponding multiple linear/logistic regression model. To obtain more means better decoding of the complex emotions of others (range 0–36 consistent models, non-significant variables were excluded step by step points). starting with the parameters with the highest p-value. To control the Four false-belief stories were used to assess cognitive ToM skills. The stability of the models, multiple collinearity diagnostics were per­ first-order cognitive ToM was measured using the Sally & Anne [52] and formed. A final model was constructed for each sub-domain of social The Box of Chocolate [54] stories. The second-order cognitive ToM was cognition assessed that included all variables that independently influ­ measured using The Burglar [54] and The Ice-Cream Van [53] stories. enced the test score. 3 G. Navarra-Ventura et al. Comprehensive Psychiatry 109 (2021) 152258 Table 1 Demographic data and estimated IQ of the total sample (n = 160). Mean (SD) are reported Patients with BD (n = 60) Patients with SCH (n = 60) Healthy subjects (n = 40) Statistics (Kruskal-Wallis H test) 1, Men 2, Women 3, Men 4, Women 5, Men 6, Women H p Post hoc tests n 30 30 30 30 20 20 Age, years 47.5 (8.3) 46.9 (9.2) 44.8 (8.7) 45.1 (8.8) 46.1 (11.2) 45.6 (9.9) 2.619 0.758 Education, years 11.7 (3.1) 11.4 (2.5) 11.1 (2.3) 10.9 (3.1) 12.4 (2.7) 11.3 (3.0) 4.367 0.498 Estimated IQ*** 99.2 (7.1) 98.5 (7.4) 87.7 (9.6) 85.4 (9.7) 100.8 (7.8) 99.5 (7.8) 52.340 <0.001 3, 4 < 1, 2, 5, 6 Note: Estimated IQ is presented in standard scores, which have a mean of 100 and a SD of ±15. Abbreviations: IQ, Intelligence quotient; BD, Bipolar disorder; SCH, Schizophrenia/Schizoaffective disorder. *** p < 0.001. 3. Results 3.2. Clinical variables 3.1. Demographic data and estimated IQ Table 2 summarizes the clinical characteristics of the patient groups. Patients with BD had an earlier age of onset (d = 0.596) and a longer Table 1 summarizes the demographic data and estimated IQ of the duration of illness (d = 1.377) than patients with SCH. In contrast, SCH total sample. The groups did not differ in age (p = 0.420) or years of patients received higher doses of chlorpromazine equivalents (d = education (p = 0.307). However, they differed in estimated IQ (H = 7.012). 51.495, p < 0.001, d = 1.796). Post hoc analysis showed that SCH pa­ Most patients (78.3%) were free of subthreshold depressive symp­ tients had a lower estimated IQ than BD patients (U = 606.5, p < 0.001, toms at the time of evaluation (HAM–D: 3.8 ± 2.2; CDSS: 1.3 ± 1.3). d = 1.819) and healthy subjects (U = 375.5, p < 0.001, d = 1.800), while However, 13 patients with BD (6 men, 7 women) and 13 patients with there were no significant differences between BD patients and healthy SCH (6 men, 7 women) suffered from subthreshold depression subjects (U = 1.090, p = 0.431). (HAM–D: 9.2 ± 1.8; CDSS: 4.2 ± 0.4). In the BD group, women had more pronounced depressive symptoms Table 2 Clinical characteristics of the patient groups (n = 120). Mean (SD) are reported unless otherwise specified Patients with BD (n = 60) Patients with SCH (n = 60) Statistics t/U/X2 p Diagnostic subtype, n Type I/II 46/14 Schizophrenia/Schizoaffective disorder 40/20 Symptom rating scales YMRS Total score 1.0 (1.5) HAM-D Total score 5.0 (3.1) PANSS Total score 53.8 (12.6) Positive scale 9.9 (3.1) Negative scale 17.4 (5.7) General scale 26.5 (6.0) CDSS Total score 2.0 (1.7) Course of the disease Age of onset, years* 28.3 (11.1) 32.1 (9.2) 2.286a 0.024 Duration of illness, years*** 19.6 (11.5) 10.7 (9.4) 980.0b <0.001 Number of hospitalizations 1.7 (2.0) 2.1 (2.6) 1608.0b 0.296 History of psychosis, n (%) 34 (56.7) Type of psychotropic drugs, n (%) Antipsychotics (AP)*** 0 (0.0) 18 (30.0) 21.176c <0.001 Mood stabilizers/Anticonvulsants 3 (5.0) 0 (0.0) 3.077c 0.079 AP + Mood stabilizers/Anticonvulsants 13 (21.7) 6 (10.0) 3.064c 0.080 Other combinations (including AD and BZD) 44 (73.3) 36 (60.0) 2.400c 0.121 Doses of psychotropic drugs (milligrams/day) Chlorpromazine equivalents (n = 109)*** 252.5 (293.9) 510.8 (382.8) 4.836a <0.001 Fluoxetine equivalents (n = 55) 37.0 (19.7) 42.1 (29.4) 0.461a 0.647 Diazepam equivalents (n = 58) 24.0 (30.8) 19.6 (11.4) − 0.315a 0.754 Abbreviations: BD, Bipolar disorder; SCH, Schizophrenia/Schizoaffective disorder; YMRS, Young Mania Rating Scale; HAM–D, 17-item Hamilton Depression Rating Scale; PANSS, Positive and Negative Syndrome Scale; CDSS, Calgary Depression Scale for Schizophrenia; AD, Antidepressants; BZD, Benzodiazepines. a Student's t-test. b Mann-Whitney U test. c Chi-square test (X2). * p < 0.05. *** p < 0.001. 4 G. Navarra-Ventura et al. Comprehensive Psychiatry 109 (2021) 152258 Fig. 1. Sex differences in the POFA (emotion recognition) and RMET (affective ToM) tasks in BD, SCH and healthy subjects. Note: Box plots showing the median (bold line), first and third quartiles (middle lines between the bold line and the whiskers), and the minimum and maximum values (whiskers) of the POFA and RMET tasks separated by group and sex. Abbreviations: POFA, Pictures of Facial Affect; RMET, Reading the Mind in the Eyes Test; ToM, Theory of Mind; BD, Bipolar disorder; SCH, Schizophrenia/Schizoaffective disorder; ns, not significant. (A) POFA (emotion recognition) performance. BD men vs. BD women: 43.5 ± 5.9 vs. 44.9 ± 5.7 (t(58) = − 0.936, p = 0.353); SCH men vs. SCH women: 44.8 ± 5.8 vs. 43.9 ± 6.4 (t(58) = 0.529, p = 0.599); Healthy men vs. Healthy women: 49.4 ± 4.2 vs. 52.9 ± 2.7 (t(32.1) = − 3.078, p = 0.004). (B) RMET (affective ToM) performance. BD men vs. BD women: 20.8 ± 4.0 vs. 21.1 ± 4.4 (t(58) = − 0.276, p = 0.783); SCH men vs. SCH women: 19.4 ± 4.8 vs. 20.1 ± 4.4 (t(58) = − 0.562, p = 0.576); Healthy men vs. Healthy women: 24.9 ± 3.4 vs. 28.9 ± 3.1 (t(38) = − 3.820, p < 0.001). Statistics: ** p < 0.01; *** p < 0.001. than men (5.7 ± 2.8 vs. 4.2 ± 3.3; U = 307.5, p = 0.034, d = 0.429), but 1.165) and SCH (p < 0.001, d = 1.131), while there were no significant men had suffered more manic episodes than women (2.5 ± 2.4 vs. 1.0 ± differences between patient groups (p = 1.000). Sex differences were 1.2; U = 273.0, p = 0.007, d = 0.559). No other sex-related differences only found in healthy subjects, with women performing better than men were found. (Fig. 1, A). The significance of this difference remained even after strict Bonferroni correction (p = 0.012, d = 0.942). The mean RMET (affective ToM) score was 21.0 ± 4.2 in BD patients, 3.3. Group and sex differences in social cognition 19.8 ± 4.6 in SCH patients and 26.9 ± 3.8 in healthy subjects (F(2,157) = 36.608, p < 0.001, d = 1.397). Post hoc analysis showed that healthy The mean POFA (emotion recognition) score was 44.2 ± 5.8 in BD subjects performed better than patients with BD (p < 0.001, d = 1.115) patients, 44.4 ± 6.1 in SCH patients and 51.1 ± 3.9 in healthy subjects and SCH (p < 0.001, d = 1.342), while there were no significant dif­ (F(2,157) = 23.431, p < 0.001, d = 1.284). Post hoc analysis showed that ferences between patient groups (p = 0.385). Again, sex differences were healthy subjects performed better than patients with BD (p < 0.001, d = Fig. 2. Sex differences in first- and second-order false-belief tasks (cognitive ToM) in BD, SCH and healthy subjects. Note: Bar chart showing the percentage of non- failure in first- and second-order false-belief tasks separated by group and sex. Abbreviations: ToM, Theory of Mind; BD, Bipolar disorder; SCH, Schizophrenia/ Schizoaffective disorder; ns, not significant. (C) First-order false-belief non-failure/failure. BD men vs. BD women: 26/4 vs. 27/3 (X2(1) = 0.162, p = 0.688); SCH men vs. SCH women: 23/7 vs. 26/4 (X2(1) = 1.002, p = 0.317); Healthy men vs. Healthy women: 19/1 vs. 19/1 (X2(1) = 0.000, p = 1.000). (D) Second-order false-belief non- failure/failure. BD men vs. women: 13/13 vs. 16/13 (X2(1) = 0.147, p = 0.701); SCH men vs. SCH women: 13/11 vs. 13/15 (X2(1) = 0.310, p = 0.578); Healthy men vs. Healthy women: 18/2 vs. 18/1 (X2(1) = 0.308, p = 0.579). Statistics: No significant differences at p < 0.05. 5 G. Navarra-Ventura et al. Comprehensive Psychiatry 109 (2021) 152258 only found in healthy subjects, with women performing better than men 3.4. Impact of age, estimated IQ, and clinical variables on patients' social (Fig. 1, B). The significance of this difference also remained after strict cognitive performance Bonferroni correction (p = 0.001, d = 1.109). Additional analyses exploring the effect of group, sex, and group-sex Table 3 (A) shows that older age, lower estimated IQ, and younger interaction on a variable combining the POFA and RMET tests into a age of onset were associated with worse POFA performance in bivariate single index can be found in the Supplementary Material. These analyses analyses. All other variables had p ≥ 0.05 and were discarded. In the were repeated adjusting for age and estimated IQ. The results confirm multiple linear regression model age and estimated IQ, but not age of that the two groups of patients perform worse than healthy subjects on onset, remained significant factors. The final model included age (B = emotional and affective aspects of social cognition (Online Resource 1 − 0.199, 95% CI = − 0.31 to − 0.09, p = 0.001) and estimated IQ (B = and 2), that there are no differences between BD and SCH patients 0.217, 95% CI = 0.12 to 0.31, p < 0.001) and explained 18.5% of the (Online Resource 1 and 2), and that the differences between men and variance in emotion recognition (adjusted R2 = 0.171, F(2,117) = 13.303, women are limited to healthy subjects (Online Resource 2 and 3). p < 0.001). As for first- and second-order cognitive ToM, 88.3% (n = 53/60) of Table 3 (B) shows that older age, lower estimated IQ, younger age of BD patients, 81.7% (n = 49/60) of SCH patients and 95.0% (n = 38/40) onset, and higher doses of chlorpromazine and fluoxetine equivalents of healthy subjects responded correctly to the first-order false-belief task were associated with worse RMET performance in bivariate analyses. All and no significant differences were found between them (X2(2) = 3.962, other variables had p ≥ 0.05 and were discarded. Again, only age and p = 0.138). However, only 52.7% (n = 29/55) of BD patients and 50.0% estimated IQ remained significant factors in the multiple linear regres­ (n = 26/52) of SCH patients responded correctly to the second-order sion model. Non-significant variables were extracted from the model in false-belief task compared to 92.3% (n = 36/39) of healthy subjects the following order: age of onset (p = 0.472), chlorpromazine equiva­ (X2(2) = 20.454, p < 0.001, V = 0.374). Post hoc analysis showed that lents (p = 0.304) and fluoxetine equivalents (p = 0.148). The final model healthy subjects performed better than patients with BD (X2(1) = 16.757, included age (B = − 0.153, 95% CI = − 0.24 to − 0.07, p < 0.001) and p < 0.001, V = 0.422) and SCH (X2(1) = 18.372, p < 0.001, V = 0.449), estimated IQ (B = 0.174, 95% CI = 0.10 to 0.24, p < 0.001) and while there were no significant differences between patient groups explained 20.9% of the variance in affective ToM (adjusted R2 = 0.196, (X2(1) = 0.080, p = 0.778). No sex-related differences were found in first- F(2,117) = 15.460, p < 0.001). (Fig. 2, C) and second-order false-belief tasks (Fig. 2, D). An additional analysis exploring the impact of age, estimated IQ, and To control for the possible confounding effect of BD women's more clinical variables on the “POFA and RMET variable” can be found in pronounced depressive symptoms, sex-related analyses were repeated in Online Resource 4 (see Supplementary Material). The results of this the BD group, including HAM-D as a covariate. Sex remained non- analysis are consistent with the data reported so far. significant in all sub-domains of social cognition, including the “POFA Table 3 (C) shows that lower estimated IQ and subthreshold and RMET variable” (p ≥ 0.165). depression were associated with higher first-order false-belief failure (cognitive ToM) in binomial analyses. All other variables had p ≥ 0.05 and were discarded. In the multiple logistic regression model, only subthreshold depression remained a significant factor (Nagelkerke R2 = Table 3 Impact of age, estimated IQ, and clinical variables on patients' social cognitive performance. (A) POFA (emotion recognition) (B) RMET (affective ToM) Bivariate analyses Multiple linear regression model Bivariate analyses Multiple linear regression model B (95% CI) p B (95% CI) p B (95% CI) p B (95% CI) p Age* ¡0.141 (¡0.26 to 0.022 ¡0.184 (¡0.30 to 0.002 ¡0.106 (¡0.20 to 0.021 ¡0.141 (¡0.28 to 0.049 ¡0.02) ¡0.07) ¡0.02) ¡0.00) Estimated IQ* 0.180 (0.08 to 0.28) <0.001 0.203 (0.11 to 0.30) <0.001 0.146 (0.07 to 0.22) <0.001 0.165 (0.04 to 0.29) 0.011 Subthreshold depression − 0.428 (− 3.03 to 2.17) 0.745 − 0.065 (− 2.00 to 1.87) 0.947 Age of onset ¡1.382 (¡2.50 to 0.016 − 0.618 (− 1.71 to 0.48) 0.266 ¡0.848 (¡1.70 to 0.048 − 0.376 (− 1.42 to 0.67) 0.472 ¡0.26) ¡0.01) Duration of illness 0.008 (− 0.09 to 0.10) 0.875 − 0.002 (− 0.07 to 0.07) 0.958 Chlorpromazine − 1.025 (− 3.45 to 1.40) 0.404 ¡2.537 (¡4.29 to 0.005 − 1.344 (− 3.98 to 1.30) 0.310 equivalents ¡0.78) Fluoxetine equivalents − 4.940 (− 10.95 to 0.105 ¡4.805 (¡8.99 to 0.025 − 2.336 (− 6.39 to 1.72) 0.252 1.07) ¡0.62) Diazepam equivalents − 2.375 (− 7.49 to 0.357 − 0.071 (− 3.86 to 3.72) 0.970 − 2.74) (C) First-order false-belief failure (cognitive ToM) (D) Second-order false-belief failure (cognitive ToM) Binomial analyses Multiple logistic regression model Binomial analyses Multiple logistic regression model OR (95% CI) p OR (95% CI) p OR (95% CI) p OR (95% CI) p Age 1.064 (1.00 to 1.13) 0.051 1.023 (0.98 to 1.07) 0.304 Estimated IQ 0.948 (0.90 to 1.00) 0.034 0.949 (0.90 to 1.00) 0.062 0.953 (0.92 to 0.99) 0.016 Subthreshold depression* 6.719 (2.30 to 19.63) <0.001 6.398 (2.14 to 19.15) 0.001 2.057 (0.74 to 5.72) 0.167 Age of onset 1.586 (0.92 to 2.74) 0.099 1.171 (0.79 to 1.75) 0.441 Duration of illness 1.000 (0.96 to 1.05) 0.987 1.005 (0.97 to 1.04) 0.776 Chlorpromazine equivalents 1.526 (0.50 to 4.67) 0.459 1.593 (0.67 to 3.77) 0.290 Fluoxetine equivalents 1.460 (0.07 to 30.68) 0.808 2.243 (0.24 to 21.42) 0.483 Diazepam equivalents 5.792 (0.58 to 58.22) 0.136 2.000 (0.33 to 12.20) 0.452 (A), (B) and (C) sample size: n = 120, except chlorpromazine equivalents (n = 109), fluoxetine equivalents (n = 55) and diazepam equivalents (n = 58). (D) sample size: n = 107, except chlorpromazine equivalents (n = 96), fluoxetine equivalents (n = 49) and diazepam equivalents (n = 51). Abbreviations: IQ, Intelligence quotient; POFA, Pictures of Facial Affect; RMET, Reading the Mind in the Eyes Test; ToM, Theory of Mind. Statistics: Bold = Statistically significant at p < 0.05. * p < 0.05 in the multiple linear/logistic regression model. 6 G. Navarra-Ventura et al. Comprehensive Psychiatry 109 (2021) 152258 0.168, X2(1) = 15.733, p = 0.001). Consistent with this result, patients sex on the neural substrate of social cognition has been little studied with subthreshold depressive symptoms had a higher failure rate on the [81], especially in patients with BD. Therefore, these ideas point only to first-order false-belief task than patients without subthreshold depres­ hypothesis generation. sive symptoms (38.5% vs. 8.5%; X2(1) = 14.329, p < 0.001, V = 0.346). The two groups of patients had deficits in emotion recognition, af­ This difference was also significant in the BD (38.5% vs. 4.3%; X2(1) = fective ToM and second-order cognitive ToM, but not in first-order 11.562, p = 0.001, V = 0.439) and SCH (38.5% vs. 12.8%; X2(1) = 4.491, cognitive ToM. This result is consistent with previous data indicating p = 0.034, V = 0.274) groups. that social cognition is altered in BD [5,12,26], while confirming that Table 3 (D) shows that lower estimated IQ was associated with this deficit is a core feature of SCH [4,6,27]. However, the mental higher second-order false-belief failure (cognitive ToM) in binomial workload required to successfully elaborate a first-order false-belief is analyses (Nagelkerke R2 = 0.076, X2(1) = 6.277, p = 0.016). However, lower than that required for a second-order false-belief [52–54]. the Hosmer-Lemeshow test indicated poor fit of the model (p = 0.026). Therefore, it is possible that first-order cognitive ToM is more resistant Thus, caution is advised when interpreting this result. All other variables to social brain changes than second-order cognitive ToM. Consistent had p ≥ 0.05 and were discarded, so multiple logistic regression analysis with this idea and with other research in BD [32,37,40,46,47] and SCH was not performed. [32,37,46,56], first-order cognitive skills were not impaired in the pa­ Follow-up analyses in the BD group showed that diagnostic subtype tient groups of our study. Furthermore, we observed that no participant (type I/II) and history of psychosis were not significant factors for social responded correctly to the second-order false-belief task if they had cognition performance, including the “POFA and RMET” variable (p ≥ failed the first-order false-belief task, which, in turn, is compatible with 0.132). In the SCH group, diagnostic subtype (schizophrenia/schizo­ a hierarchical relationship between the different sub-domains of social affective disorder) and residual negative symptoms (measured by cognition already discussed in previous studies [1,56]. PANSS Negative) were also not significant factors (p ≥ 0.275). Unlike other research in which BD patients had similar but less se­ vere social cognitive deficits than SCH patients [32–46], we found that 4. Discussion the two disorders were equally impaired in both the ability to recognize others' basic emotions and the ability to identify others' cognitive and This is the first study to compare emotion recognition, affective ToM, affective mental states [8,47–49]. This suggests that, at least in these and first- and second-order cognitive ToM in a sample of BD patients, sub-domains of social cognition, the severity of impairment might be SCH patients, and healthy subjects from a sex-related perspective. Our comparable between them. In view of this finding, one might speculate results show that patients with BD and SCH performed worse than that the lack of gradation in the severity of these deficits could be related healthy subjects in all sub-domains of social cognition assessed, except to between-group differences in disease course or psychotropic drugs. for first-order cognitive ToM, which remained preserved in up to 85.0% However, in the final regression models, none proved significant for of cases. However, no differences were found between the two disorders. performance in social cognition. Indeed, there is evidence that deficits in Instead, patients' deficits were related to older age, lower estimated IQ emotion recognition and ToM show little or no relationship to these and/or subthreshold depression. Our results also show that healthy fe­ variables in clinically stable patients [12,36,44,50,58,60,61]. males had a marked advantage in emotion recognition and affective In contrast, older age and lower estimated IQ were associated with ToM tasks compared to healthy males, but that this difference was not more severe deficits in specifics aspects of social cognition. Whereas age evident in the patient groups. influenced emotion recognition and affective ToM, estimated IQ was Looking at sex differences in more detail, it is striking that healthy related to emotion recognition, affective ToM, and (possibly) second- females showed better emotion recognition and affective ToM skills than order cognitive ToM, but not to first-order cognitive ToM. On one healthy males, but performed just as poorly as their male counterparts hand, this is consistent with the lower cognitive workload required by when diagnosed with BD or SCH. Similar results have been reported in this task compared to the other three. On the other hand, it should be two recent cross-sectional studies in BD patients [59], different psy­ kept in mind that IQ is one of the proxy measures of cognitive reserve chotic disorders and healthy subjects [57], suggesting that female pa­ (the brain's resistance to pathological changes). From this perspective, tients may not retain the advantage in emotion recognition observed in our finding could also indicate a possible protective effect of this vari­ healthy females. Along these lines, a review [55] and two meta-analyses able, as discussed in a previous study [82]. in BD [25] and SCH [28] found that the effect of disease outweighs the Finally, we observed that first-order cognitive ToM was only deter­ effect of sex on this ability and that sex differences in emotion recog­ mined by the presence of mild depressive symptoms, but not by the nition remain isolated in healthy individuals. Our results add to the severity of residual negative symptoms or by any other clinical or de­ existing literature that the loss of healthy women's advantage in patients mographic variable. More specifically, patients suffering from sub­ with BD and SCH might also involve affective ToM skills (the ability to threshold depression were up to 6 times more likely to have impaired decode others' complex emotions). However, other cross-sectional first-order skills than non-depressed individuals. Therefore, given that studies using assessment tools similar, although not identical, to those this sub-process of cognitive ToM remained preserved in up to 85.0% of in our study have shown opposite results [12,50,58,60,61]. Therefore, cases, this finding is of notable clinical interest because it points to an the present findings should be confirmed in future studies with larger increased risk of severe ToM deficits in this sub-group of individuals. samples. Similar results have been found in a previous study [11]. To provide an explanation why the advantage of healthy females in This latter finding highlights that, along with impaired emotion emotion recognition and affective ToM might be lost in BD and SCH recognition, affective ToM, and second-order cognitive ToM in BD and patients, we turned to several neuroimaging studies. So far, there is SCH patients, the presence of subthreshold depressive symptoms is agreement that ToM depends on an intact prefrontal cortex and that, likely to disrupt normal first-order cognitive ToM functioning [10,83]. while cognitive ToM is impaired by extensive prefrontal lesions, affec­ Difficulties in recognizing emotions and understanding the thoughts, tive ToM is impaired by localized damage to the ventromedial prefrontal beliefs and feelings of others have real-life consequences, such as cortex [16], which, in turn, shows broad connections with other areas problems in social relationships due to misinterpretation of the true involved in emotion recognition such as the amygdala [1]. There is intentions of others. In addition, a person who is not perceived as so­ evidence that the orbitofrontal cortex to amygdala ratio may be greater cially competent will not be a partner with whom one wants to interact, in healthy females than in healthy males and that this may be related to which could promote social distancing or even isolation [2,3]. Thus, it is the sex differences found in social cognition [79]. However, it has conceivable that BD and SCH patients with subthreshold depressive recently been found that this sexually dimorphic difference may be symptoms will be less likely to participate adequately in social situations altered in patients with SCH [80]. Beyond these studies, the influence of than their non-depressed counterparts, at least in part because of their 7 G. Navarra-Ventura et al. Comprehensive Psychiatry 109 (2021) 152258 higher failure rate in first-order cognitive ToM skills (38.5% vs. 8.5%). authors contributed to data interpretation, approved the final version Consistent with this hypothesis, it has been found that the severity of for publication, and participated sufficiently in the work to take public depressive symptoms modulates the relationship between social cogni­ responsibility for appropriate portions of the content. tion and social functioning [83]. Finally, our results point to the need for a systematic and regular Research data assessment of both social cognition and subthreshold depression in clinically stable outpatients with BD and SCH, as this will contribute to a The data that support the findings of this study are available from the more accurate determination of their cognitive deficit profile and to the corresponding authors upon reasonable request. The data are not pub­ identification of therapeutic targets aimed at improving their social licly available due to privacy or ethical restrictions. functioning. According to our results, existing cognitive rehabilitations programs would be useful for both BD and SCH patients, as well as for Funding men and women, since they all showed similar performance in emotion recognition and ToM. However, in patients with subthreshold depres­ The work was supported by the Instituto de Salud Carlos III (Madrid, sive symptoms, additional effort should be made to train first-order Spain) and the European Regional Development Fund (European Com­ cognitive ToM skills. mission) (grant number PI15/01478). The funders played no role in When interpreting the results of this study, the following limitations study design, data collection and analysis, manuscript preparation, or should be considered. First, the cross-sectional design, which does not decision to publish. allow any causal inference between patients' social cognitive deficits and their associated factors. Second, the inclusion of patients with BD type II Declaration of Competing Interest and schizoaffective disorder, which could help explain the lack of dif­ ferences in emotion recognition and ToM between patient groups, as The authors declare no conflict of interest in relation to the publi­ better cognitive outcomes have occasionally been found in these clinical cation of this article. populations than in those with BD type I and schizophrenia [15,47]. However, diagnostic subtype was not a significant factor for perfor­ Acknowledgements mance on emotion recognition and ToM in the regression analyses. Therefore, we do not expect this variable to have confounded the results. We would like to thank the staff of the Department of Mental Health Third, the relatively small sample size, especially when groups are of the Parc Taulí University Hospital in Sabadell, Catalonia (Spain), for divided by sex, which limits our ability to draw definitive conclusions their assistance in this study, specifically Dr. Javier Labad, Sara Crivillés, about whether sex differences in emotion recognition and affective ToM Laura Ros and Wanda Zabala for their help in patient recruitment, and in healthy individuals are lost in BD and SCH. In contrast, our results are Joan Carles Oliva for his technical assistance in statistical analysis. We strengthened by careful matching between the clinical samples and with thank the patients and healthy subjects who participated in the study for the healthy subjects. Fourth, our study included patients who did not their kind cooperation. necessarily meet criteria for full depression remission or who had re­ sidual negative symptoms. Although this could be considered a meth­ Appendix A. Supplementary data odological limitation, it allows the data to be generalized to most patients with BD and SCH that healthcare professionals treat in their Supplementary data to this article can be found online at https://doi. daily clinical practice. Finally, we did not include any tests of social org/10.1016/j.comppsych.2021.152258. functioning. The relationship between social cognition and social func­ tioning has already been described extensively in previous studies [2,3]. References 5. Conclusions [1] Mitchell RLC, Phillips LH. The overlapping relationship between emotion perception and theory of mind. Neuropsychologia 2015;70:1–10. https://doi.org/ 10.1016/j.neuropsychologia.2015.02.018. The following conclusions can be drawn. First, in healthy subjects sex [2] Fett AJ, Viechtbauer W, Dominguez M, Penn DL, Van Os J, Krabbendam L. The only influenced affective ToM, but not cognitive ToM, thus confirming relationship between neurocognition and social cognition with functional that the two ToM processes are somehow independent. Second, the outcomes in schizophrenia: a meta-analysis. Neurosci Biobehav Rev 2011;35: 573–88. https://doi.org/10.1016/j.neubiorev.2010.07.001. advantage of healthy women in emotion recognition and affective ToM [3] Vlad M, Raucher-Chéné D, Henry A, Kaladjian A. Functional outcome and social was not maintained in BD and SCH patients, so that disease, not sex, cognition in bipolar disorder: is there a connection? Eur Psychiatry 2018;52: would be the main factor related to the deficit in social cognition. 116–25. https://doi.org/10.1016/j.eurpsy.2018.05.002. [4] Green MF, Bearden CE, Cannon TD, Fiske AP, Hellemann GS, Horan P, et al. Social Finally, while replicating previous findings that BD and SCH patients are cognition in schizophrenia, part 1: performance across phase of illness. Schizophr characterized by mild to severe impairments in social cognition, we Bull 2012;38:854–64. https://doi.org/10.1093/schbul/sbq171. found that emotion recognition, affective ToM and first- and second- [5] Samamé C. Social cognition throughout the three phases of bipolar disorder: a state-of-the-art overview. Psychiatry Res 2013;210:1275–86. https://doi.org/ order cognitive ToM represent at least four sub-domains in which the 10.1016/j.psychres.2013.08.012. level of impairment may be comparable between the two disorders. It [6] Horan WP, Green MF, Degroot M, Fiske A, Hellemann G, Kee K, et al. Social may be that SCH patients only show more severe deficits than BD pa­ cognition in schizophrenia, part 2: 12-month stability and prediction of functional outcome in first-episode patients. Schizophr Bull 2012;38:856–72. https://doi.org/ tients in the more complex and sophisticated aspects of social cognition, 10.1093/schbul/sbr001. so future studies are encouraged to use more demanding tests when [7] Martino DJ, Samamé C, Strejilevich SA. Stability of facial emotion recognition comparing the two disorders. performance in bipolar disorder. Psychiatry Res 2016;243:182–4. https://doi.org/ 10.1016/j.psychres.2016.06.026. [8] Daros AR, Ruocco AC, Reilly JL, Harris MSH, Sweeney JA. Facial emotion Contributors recognition in first-episode schizophrenia and bipolar disorder with psychosis. Schizophr Res 2014;153:32–7. https://doi.org/10.1016/j.schres.2014.01.009. [9] Pinkham AE, Kelsven S, Kouros C, Harvey PD, Penn DL. The effect of age, race, and MJ and NC designed the study and wrote the protocol. CM, JMC, JC sex on social cognitive performance in individuals with schizophrenia. J Nerv Ment and AJ recruited the patients and conducted the clinical evaluation. Dis 2017;205:346–52. https://doi.org/10.1097/NMD.0000000000000654. GNV, MVG, MSB and SFG recruited the healthy subjects and conducted [10] Vieta E, Berk M, Schulze TG, Carvalho AF, Suppes T, Calabrese JR, et al. Bipolar the IQ and social cognition assessment. GNV and NC run the statistical disorders. Nat Rev Dis Prim 2018;4:18008. https://doi.org/10.1038/nrdp.2018.8. [11] Varo C, Jimenez E, Solé B, Bonnín CM, Torrent C, Valls E, et al. Social cognition in analysis. GNV drafted the first version of the manuscript. XG, DP, GL and bipolar disorder: focus on emotional intelligence. J Affect Disord 2017;217:210–7. EV critically reviewed the article for important intellectual content. All https://doi.org/10.1016/j.jad.2017.04.012. 8 G. Navarra-Ventura et al. Comprehensive Psychiatry 109 (2021) 152258 [12] Varo C, Jiménez E, Solé B, Bonnín CM, Torrent C, Lahera G, et al. Social cognition [38] Hwang HC, Kim SM, Han DH. Different facial recognition patterns in schizophrenia in bipolar disorder: the role of sociodemographic, clinical, and neurocognitive and bipolar disorder assessed using a computerized emotional perception test and variables in emotional intelligence. Acta Psychiatr Scand 2019;139:369–80. fMRI. J Affect Disord 2021;279:83–8. https://doi.org/10.1016/j.jad.2020.09.125. https://doi.org/10.1111/acps.13014. [39] Joshua NR. Face processing in schizophrenia: an investigation of configural [13] Yatham LN, Torres IJ, Malhi GS, Frangou S, Glahn DC, Bearden CE, et al. The processing and the relationship with facial emotion processing and neurocognition. international society for bipolar disorders-battery for assessment of neurocognition University of Melbourne; 2010 (Thesis). (ISBD-BANC). Bipolar Disord 2010;12:351–63. https://doi.org/10.1111/j.1399- [40] Sakarya A. Association of ToM deficits with insight and other cognitive functions 5618.2010.00830.x. among remitted schizophrenia and bipolar disorder patients. Ankara University; [14] Nuechterlein KH, Green MF, Kern RS, Baade LE, Barch DM, Cohen JD, et al. The 2012 (Thesis). MATRICS consensus cognitive battery, part 1: test selection, reliability, and [41] Caletti E, Paoli RA, Fiorentini A, Cigliobianco M, Zugno E, Serati M, et al. validity. Am J Psychiatry 2008;165:203–13. https://doi.org/10.1176/appi. Neuropsychology, social cognition and global functioning among bipolar, ajp.2007.07010042. schizophrenic patients and healthy controls: preliminary data. Front Hum Neurosci [15] Miskowiak KW, Seeberg I, Kjaerstad HL, Burdick KE, Martinez-Aran A, Bonnin C, 2013;7:661. https://doi.org/10.3389/fnhum.2013.00661. et al. Affective cognition in bipolar disorder: a systematic review by the ISBD [42] Thaler NS, Allen DN, Sutton GP, Vertinski M, Ringdahl EN. Differential impairment targeting cognition task force. Bipolar Disord 2019;21:686–719. https://doi.org/ of social cognition factors in bipolar disorder with and without psychotic features 10.1111/bdi.12834. and schizophrenia. J Psychiatr Res 2013;47:2004–10. https://doi.org/10.1016/j. [16] Kalbe E, Schlegel M, Sack AT, Nowak DA, Dafotakis M, Bangard C, et al. jpsychires.2013.09.010. Dissociating cognitive from affective theory of mind: a TMS study. Cortex 2010;46: [43] Rowland JE, Hamilton MK, Vella N, Lino BJ, Mitchell PB, Green MJ. Adaptive 769–80. https://doi.org/10.1016/j.cortex.2009.07.010. associations between social cognition and emotion regulation are absent in [17] Kret ME, de Gelder B. A review on sex differences in processing emotional signals. schizophrenia and bipolar disorder. Front Psychol 2013;3. https://doi.org/ Neuropsychologia 2012;50:1211–21. https://doi.org/10.1016/j. 10.3389/fpsyg.2012.00607. neuropsychologia.2011.12.022. [44] Guastella AJ, Hermens DF, Van Zwieten A, Naismith SL, Lee RSC, Cacciotti-Saija C, [18] Christov-Moore L, Simpson EA, Coudé G, Grigaityte K, Iacoboni M, Ferrari PF. et al. Social cognitive performance as a marker of positive psychotic symptoms in Empathy: gender effects in brain and behavior. Neurosci Biobehav Rev 2014;46: young people seeking help for mental health problems. Schizophr Res 2013;149: 604–27. https://doi.org/10.1016/j.neubiorev.2014.09.001. 77–82. https://doi.org/10.1016/j.schres.2013.06.006. [19] Thompson AEE, Voyer D. Sex differences in the ability to recognise non-verbal [45] Lee J, Altshuler L, Glahn DC, Miklowitz DJ, Ochsner K, Green MF. Social and displays of emotion: a meta-analysis. Cognit Emot 2014;28:1164–95. https://doi. nonsocial cognition in bipolar disorder and schizophrenia: relative levels of org/10.1080/02699931.2013.875889. impairment. Am J Psychiatry 2013;170:334–41. https://doi.org/10.1176/appi. [20] Fine JG, Semrud-Clikeman M, Zhu DC. Gender differences in BOLD activation to ajp.2012.12040490. face photographs and video vignettes. Behav Brain Res 2009;201:137–46. https:// [46] Andrews S. Exploring social cognition and mirror systems in schizophrenia and doi.org/10.1016/j.bbr.2009.02.009. bipolar disorder. Monash University; 2013 (Thesis). [21] Ingalhalikar M, Smith A, Parker D, Satterthwaite TD, Elliott MA, Ruparel K, et al. [47] Caponigro JM. Social cognitive deficits in schizophrenia, schizoaffective disorder Sex differences in the structural connectome of the human brain. Proc Natl Acad and bipolar disorder: similarities and differences. University of Pittsburgh; 2007 Sci 2014;111:823–8. https://doi.org/10.1073/pnas.1316909110. (Thesis). [22] Ochoa S, Usall J, Cobo J, Labad X, Kulkarni J. Gender differences in schizophrenia [48] Donohoe G, Duignan A, Hargreaves A, Morris DW, Rose E, Robertson D, et al. and first-episode psychosis: a comprehensive literature review. Schizophr Res Treat Social cognition in bipolar disorder versus schizophrenia: comparability in mental 2012;2012:916198. https://doi.org/10.1155/2012/916198. state decoding deficits. Bipolar Disord 2012;14:743–8. https://doi.org/10.1111/ [23] DiFlorio A, Jones I. Is sex important? Gender differences in bipolar disorder. Int bdi.12011. Rev Psychiatry 2010;22:437–52. https://doi.org/10.3109/ [49] Thonse U, Behere RV, Praharaj SK, Sharma PSVN. Facial emotion recognition, 09540261.2010.514601. socio-occupational functioning and expressed emotions in schizophrenia versus [24] López-Zurbano S, González-Pinto A, López P. Gender differences in bipolar bipolar disorder. Psychiatry Res 2018;264:354–60. https://doi.org/10.1016/j. disorder. In: Sáenz-Herrero M, editor. Psychopathol. Women Inc. Gend. Perspect. psychres.2018.03.027. into Descr. Psychopathol. Cham: Springer International Publishing; 2015. [50] Vaskinn A, Sundet K, Friis S, Simonsen C, Birfenaes AB, Engh JA, et al. The effect of p. 641–59. https://doi.org/10.1007/978-3-319-05870-2_28. gender on emotion perception in schizophrenia and bipolar disorder. Acta [25] Kohler CG, Hoffman LJ, Eastman LB, Healey K, Moberg PJ. Facial emotion Psychiatr Scand 2007;116:263–70. https://doi.org/10.1111/j.1600- perception in depression and bipolar disorder: a quantitative review. Psychiatry 0447.2007.00991.x. Res 2011;188:303–9. https://doi.org/10.1016/j.psychres.2011.04.019. [51] Li W, Zhou F-C, Zhang L, Ng CH, Ungvari GS, Li J, et al. Comparison of cognitive [26] Bora E, Bartholomeusz C, Pantelis C. Meta-analysis of theory of mind (ToM) dysfunction between schizophrenia and bipolar disorder patients: a meta-analysis impairment in bipolar disorder. Psychol Med 2016;46:253–64. https://doi.org/ of comparative studies. J Affect Disord 2020;274:652–61. https://doi.org/ 10.1017/S0033291715001993. 10.1016/j.jad.2020.04.051. [27] Bora E, Yucel M, Pantelis C. Theory of mind impairment in schizophrenia: meta- [52] Baron-Cohen S, Leslie AMM, Frith U. Does the autistic child have a “theory of analysis. Schizophr Res 2009;109:1–9. https://doi.org/10.1016/j. mind”? Cognition 1985;21:37–46. https://doi.org/10.1016/0010-0277(85)90022- schres.2008.12.020. 8. [28] Kohler CG, Walker JB, Martin A, Healey KM, Moberg PJ. Facial emotion perception [53] Baron-Cohen S. The autistic child’s theory of mind: a case of specific in schizophrenia: a meta-analytic review. Schizophr Bull 2010;36:1009–19. developmental delay. J Child Psychol Psyhiatry 1989;30:285–7. https://doi.org/ https://doi.org/10.1093/schbul/sbn192. 10.1111/j.1469-7610.1989.tb00241.x. [29] Bora E, Pantelis C. Social cognition in schizophrenia in comparison to bipolar [54] Happé FG. An advanced test of theory of mind: understanding of story characters’ disorder: a meta-analysis. Schizophr Res 2016;175:72–8. https://doi.org/10.1016/ thoughts and feelings by able autistic, mentally handicapped, and normal children j.schres.2016.04.018. and adults. J Autism Dev Disord 1994;24:129–54. https://doi.org/10.1007/ [30] Mitchell RLC, Young AH. Theory of mind in bipolar disorder, with comparison to BF02172093. the impairments observed in schizophrenia. Front Psychol 2016;6:188. https://doi. [55] Mote J, Kring AM. Facial emotion perception in schizophrenia: does sex matter? org/10.3389/fpsyt.2015.00188. World J Psychiatry 2016;6:257. https://doi.org/10.5498/wjp.v6.i2.257. [31] van Neerven T, Bos DJ, van Haren NE. Deficiencies in Theory of Mind in patients [56] Navarra-Ventura G, Fernandez-Gonzalo S, Turon M, Pousa E, Palao D, Cardoner N, with schizophrenia, bipolar disorder, and major depressive disorder: a systematic et al. Gender differences in social cognition: a cross-sectional pilot study of recently review of secondary literature. Neurosci Biobehav Rev 2021;120:249–61. https:// diagnosed patients with schizophrenia and healthy subjects. Can J Psychiatr 2018; doi.org/10.1016/j.neubiorev.2020.11.011. 63:538–46. https://doi.org/10.1177/0706743717746661. [32] Doody GA, Götz M, Johnstone EC, Frith CD, Cunningham Owens DG. Theory of [57] Petrovic S, Jerotic S, Mihaljevic M, Pavlovic Z, Ristic I, Soldatovic I, et al. Sex mind and psychoses. Psychol Med 1998;28:S003329179700648X. https://doi.org/ differences in facial emotion recognition in health and psychotic disorders. Cogn 10.1017/S003329179700648X. Neuropsychiatry 2019;24:108–22. https://doi.org/10.1080/ [33] Addington J, Addington D. Facial affect recognition and information processing in 13546805.2019.1582411. schizophrenia and bipolar disorder. Schizophr Res 1998;32:171–81. https://doi. [58] Abu-Akel A, Bo S. Superior mentalizing abilities of female patients with org/10.1016/S0920-9964(98)00042-5. schizophrenia. Psychiatry Res 2013;210:794–9. https://doi.org/10.1016/j. [34] Goghari VM, Sponheim SR. More pronounced deficits in facial emotion recognition psychres.2013.09.013. for schizophrenia than bipolar disorder. Compr Psychiatry 2013;54:388–97. [59] Danaher H, Allott K, Killackey E, Hester R, Cotton S. An examination of sex https://doi.org/10.1016/j.comppsych.2012.10.012. differences in neurocognition and social cognition in first-episode psychosis. [35] Baez S, Herrera E, Villarin L, Theil D, Gonzalez-Gadea ML, Gomez P, et al. Psychiatry Res 2018;259:36–43. https://doi.org/10.1016/j.psychres.2017.09.053. Contextual social cognition impairments in schizophrenia and bipolar disorder. [60] Erol A, Putgul G, Kosger F, Ersoy B. Facial emotion recognition in schizophrenia: PLoS One 2013;8:e57664. https://doi.org/10.1371/journal.pone.0057664. the impact of gender. Korean Psychiatry Investig 2013;10:69–74. https://doi.org/ [36] Bora E, Veznedaroğlu B, Vahip S. Theory of mind and executive functions in 10.4306/pi.2013.10.1.69. schizophrenia and bipolar disorder: a cross-diagnostic latent class analysis for [61] Scholten MRM, Aleman A, Montagne B, Kahn RS. Schizophrenia and processing of identification of neuropsychological subtypes. Schizophr Res 2016;176:500–5. facial emotions: sex matters. Schizophr Res 2005;78:61–7. https://doi.org/ https://doi.org/10.1016/j.schres.2016.06.007. 10.1016/j.schres.2005.06.019. [37] Wang Y, Wang Y, Zou Y, Ni K, Tian X, Sun H, et al. Theory of mind impairment and [62] American Psychiatric Association. Diagnostic and statistical manual of mental its clinical correlates in patients with schizophrenia, major depressive disorder and disorders, fourth edition, text revision: DSM-IV-TR. Washingtion, DC: American bipolar disorder. Schizophr Res 2018;197:349–56. https://doi.org/10.1016/j. Psychiatric Association; 2000. schres.2017.11.003. 9 G. Navarra-Ventura et al. Comprehensive Psychiatry 109 (2021) 152258 [63] McIntyre RS, Fallu A, Konarski JZ. Measurable outcomes in psychiatric disorders: syndrome or high-functioning autism. J Child Psychol Psyhiatry 2001;42:241–51. remission as a marker of wellness. Clin Ther 2006;28:1882–91. https://doi.org/ https://doi.org/10.1111/1469-7610.00715. 10.1016/j.clinthera.2006.11.007. [75] Pinkham AE, Penn DL, Green MF, Harvey PD. Social cognition psychometric [64] Müller MJ, Brening H, Gensch C, Klinga J, Kienzle B, Müller K-M. The Calgary evaluation: results of the initial psychometric study. Schizophr Bull 2016;42: Depression Rating Scale for schizophrenia in a healthy control group: psychometric 494–504. https://doi.org/10.1093/schbul/sbv056. properties and reference values. J Affect Disord 2005;88:69–74. https://doi.org/ [76] Danivas V, Venkatasubramanian G. Current perspectives on chlorpromazine 10.1016/j.jad.2005.04.005. equivalents: comparing apples and oranges! Indian J Psychiatry 2013;55:207. [65] Young RC, Biggs JT, Ziegler VE, Meyer DA. A rating scale for mania: reliability, https://doi.org/10.4103/0019-5545.111475. validity and sensitivity. Br J Psychiatry 1978;133:429–35. https://doi.org/ [77] Hayasaka Y, Purgato M, Magni LR, Ogawa Y, Takeshima N, Cipriani A, et al. Dose 10.1192/bjp.133.5.429. equivalents of antidepressants: evidence-based recommendations from randomized [66] Hamilton M. Development of a rating scale for primary depressive illness. Br J Soc controlled trials. J Affect Disord 2015;180:179–84. https://doi.org/10.1016/j. Clin Psychol 1967;6:278–96. https://doi.org/10.1111/j.2044-8260.1967.tb00530. jad.2015.03.021. x. [78] Ashton CH. Benzodiazepines: how they work and how to withdraw. Newcastle, [67] Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale (PANSS) England, UK: Institute of Neuroscience, Newcastle University; 2002. for schizophrenia. Schizophr Bull 1987;13:261–76. https://doi.org/10.1093/ [79] Gur RC. Sex differences in temporo-limbic and frontal brain volumes of healthy schbul/13.2.261. adults. Cereb Cortex 2002;12:998–1003. https://doi.org/10.1093/cercor/ [68] Addington D, Addington J, Schissel B. A depression rating scale for schizophrenics. 12.9.998. Schizophr Res 1990;3:247–51. https://doi.org/10.1016/0920-9964(90)90005-R. [80] Gur RE, Kohler C, Turetsky BI, Siegel SJ, Kanes SJ, Bilker WB, et al. A sexually [69] Lako IM, Bruggeman R, Knegtering H, Wiersma D, Schoevers RA, Slooff CJ, et al. dimorphic ratio of orbitofrontal to amygdala volume is altered in schizophrenia. A systematic review of instruments to measure depressive symptoms in patients Biol Psychiatry 2004;55:512–7. https://doi.org/10.1016/j.biopsych.2003.10.009. with schizophrenia. J Affect Disord 2012;140:38–47. https://doi.org/10.1016/j. [81] Mendrek A, Mancini-Marïe A. Sex/gender differences in the brain and cognition in jad.2011.10.014. schizophrenia. Neurosci Biobehav Rev 2016;67:57–78. https://doi.org/10.1016/j. [70] Wechsler D. Wechsler adult intelligence scale, third edition: WAIS-III. Barcelona, neubiorev.2015.10.013. Catalonia (Spain): TEA Ediciones; 1999. [82] González-Ortega I, González-Pinto A, Alberich S, Echeburúa E, Bernardo M, [71] Johnstone E, Cunningham Owens D, Stephen L, McIntosh A, Sharpe M. Companion Cabrera B, et al. Influence of social cognition as a mediator between cognitive to psychiatric studies, eigth edition. Churchill Livingstone: Elsevier; 2010. reserve and psychosocial functioning in patients with first episode psychosis. [72] Ekman P, Friesen WC. Pictures of facial affect. Palo Alto, CA: Consulting Psychol Med 2019;50:2702–10. https://doi.org/10.1017/S0033291719002794. Psychologist Press; 1976. [83] Lahera G, Herrera S, Reinares M, Benito A, Rullas M, González-Cases J, et al. [73] Young A, Perret D, Calder A, Sprengelmeyer R, Ekman P. Facial expressions of Hostile attributions in bipolar disorder and schizophrenia contribute to poor social emotion - stimuli and tests (FEEST). Psychology manual v1.0. Bury St Edmunds, functioning. Acta Psychiatr Scand 2015;131:472–82. https://doi.org/10.1111/ England, UK: Thames Valley Test Company; 2002. acps.12399. [74] Baron-Cohen S, Wheelwright S, Hill J, Raste Y, Plumb I. The “Reading the Mind in the Eyes” test revised version: a study with normal adults, and adults with asperger 10