Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Aug 11.
Published in final edited form as: Psychiatry Res. 2025 Aug 15;352:116694. doi: 10.1016/j.psychres.2025.116694

Social cognitive deficits in schizophrenia, bipolar disorder, and first-degree relatives: A large-sample, multi-task investigation

Scott D Blain a,b,*, Jerillyn S Kent c, Chloe A Peyromaure de Bord c, Scott R Sponheim d,e, Ivy F Tso a,b
PMCID: PMC13456643  NIHMSID: NIHMS2181734  PMID: 40849997

Abstract

Background:

Social cognitive deficits are well documented in schizophrenia (SZ) and bipolar disorder (BD). These deficits extend to first-degree relatives, indicating potential genetic liability. However, most studies have used single tasks with small samples, limiting generalizability.

Methods:

We examined social cognition across patients with SZ (n = 105), BD with psychotic features (n = 37) first-degree relatives of these patients (n = 101; 60 siblings, 33 parents, 8 offspring), and healthy controls (HC; n = 53). Participants completed four tasks: Penn Emotion Recognition Task (ER-40; facial-emotion recognition), Performance-based Prosody Identification Test (PROID; vocal-emotion recognition), Triangles Task (implicit mentalizing), and Social Attribution Task-Multiple Choice (SAT-MC; explicit mentalizing). We used Bayesian confirmatory factor analysis to derive a social cognition factor and assessed relationships with symptom severity using the Scales for the Assessment of Positive/Negative Symptoms (SAPS/SANS).

Results:

Groups differed significantly on the social cognition factor (F = 18.10, p < .001), with HC showing highest performance and SZ lowest. BD and relatives showed intermediate deficits, with all pairwise comparisons significant except relatives vs. BD. A group-by-task interaction (F = 3.70, p = .015) revealed larger group differences for mentalizing tasks (SAT-MC, Triangles) than emotion recognition tasks. In patients, social cognition correlated negatively with SAPS (r = −0.217) and SANS (r = −0.255).

Conclusions:

Social cognitive deficits scale with genetic liability for psychosis, being most pronounced in SZ, intermediate in BD and relatives, and minimal in controls. Deficits are particularly marked for higher-level mentalizing versus emotion recognition, suggesting domain-specific vulnerability. These findings support social cognition as an endophenotype for psychotic disorders with implications for early identification and intervention.

Keywords: Social cognition, Schizophrenia, Bipolar

Introduction

Individuals with schizophrenia/schizoaffective disorder (SZ) show robust deficits in social cognition that are associated with symptom severity and functional outcomes (Fett et al., 2011). These deficits span both lower-level processes like facial and vocal emotion recognition, and higher-level processes like mentalizing (Green et al., 2015; Kent and Pinkham, 2024). Given limited treatment options for social cognitive deficits (Harvey et al., 2006), understanding their mechanisms and risk factors has become a research priority. As SZ has a strong genetic component (Legge et al., 2021; Sullivan et al., 2003), examining how social cognitive deficits manifest across the spectrum of genetic liability could inform strategies for early detection and intervention.

Social cognitive deficits also characterize bipolar disorder (BD), particularly in patients with psychotic features, though they are typically less severe than in SZ (Bora and Pantelis, 2016; Samamé et al., 2012). A meta-analysis of 26 studies confirmed that BD patients outperform SZ patients in social cognition, with psychotic features associated with greater impairment (Thaler et al., 2013). These patterns mirror broader neurocognitive findings, suggesting shared vulnerability across psychotic disorders (Bortolato et al., 2015).

Critically, social cognitive deficits extend to biological relatives of both patient groups, supporting their role as markers of genetic liability. Meta-analyses document emotion recognition and mentalizing deficits in first-degree relatives of SZ patients (29 studies; Lavoie et al., 2013) and BD patients (17 studies; Bora and Özerdem, 2017). However, these studies have been constrained by small samples (median n = 27–34 per group) and reliance on single tasks, which are vulnerable to measurement error and cannot distinguish domain-general from domain-specific deficits (Apperly, 2012; Friedman et al., 2008).

Despite extensive research establishing social cognitive deficits across the psychosis spectrum, fundamental questions remain unanswered. To our knowledge, no previous study has simultaneously compared all four groups (SZ, BD, relatives, and controls) using multiple identical social cognition measures, preventing direct quantification of effect sizes across groups. Moreover, the field’s reliance on single tasks introduces substantial measurement error and limits our understanding of how deficits manifest across different social cognitive domains. Latent variable modeling allows us to extract shared variance across tasks—reducing measurement error and providing more reliable estimates of overall social cognitive functioning—while simultaneously examining whether specific domains (e.g., emotion recognition versus mentalizing) are differentially affected across groups.

The present study addresses these gaps through several innovations. First, we provide the first direct comparison across all four groups within a single study (N = 296), enabling valid effect size quantification. Second, our latent variable approach extracts shared variance across multiple tasks, reducing measurement error and providing more reliable estimates. Third, we test for domain-specific patterns to determine whether deficits are general or specific to certain social cognitive processes. These advances provide the precision needed for power calculations, intervention targeting, and early identification efforts.

We hypothesized that social cognitive ability—modeled as shared variance across emotion recognition and mentalizing tasks—would show a graded pattern: HC > relatives > BD > SZ. We anticipated associations with symptom severity within patient groups. Finally, we explored whether groups show differential deficits in lower- versus higher-level social cognition, as these processes have distinct neural correlates and may be differentially affected in psychotic disorders (Chung et al., 2013; Green et al., 2015).

Methods

Participants and procedure

Participants (n = 296) were recruited as part of the Psychosis Human Connectome Project (P-HCP), a study on the neurobiology of psychosis funded by the NIH (Demro et al., 2021). The sample included 105 individuals with schizophrenia/schizoaffective disorder (SZ; 79 schizophrenia, 16 schizoaffective, 10 other/not specified), 37 with bipolar I disorder with psychotic features (BD), 101 first-degree biological relatives, and 53 healthy controls (HC).

Recruitment utilized multiple strategies to ensure a representative sample: mental health clinics (n = 74), Veterans Affairs medical centers (n = 40), University of Minnesota databases and prior studies (n = 37), other research studies or databases (n = 51), community mental health organizations (n = 24), ResearchMatch registry (n = 23), National Alliance on Mental Illness (NAMI) support groups (n = 21), provider or participant referrals (n = 15), and public flyers (n = 11). First-degree relatives included siblings (59.4 %, n = 60), parents (32.7 %, n = 33), and adult offspring (7.9 %, n = 8). Of the 101 relatives, 77 could be definitively classified based on their proband’s diagnosis: 50 were relatives of SZ patients and 27 were relatives of BD patients.

All participants were 18–65 years old (18–69 for relatives to accommodate parents), English-speaking, and capable of providing informed consent as assessed by the University of California Brief Assessment of Capacity to Consent (Jeste et al., 2007). Exclusion criteria included current substance abuse/dependence, IQ 〈 70, neurological conditions, head injury with loss of consciousness 〉 30 min, ECT within the past year, or sensory/motor impairments preventing task completion.

Clinical diagnosis was established through the Structured Clinical Interview for DSM-IV-TR (SCID; First et al., 2002) supplemented by the psychosis section of the Diagnostic Interview for Genetic Studies (DIGS; Nurnberger et al., 1994). Current symptoms were assessed using the Scales for the Assessment of Positive/Negative Symptoms (SAPS/SANS; Andreasen and Grove, 1986) and Brief Psychiatric Rating Scale – 24-item version (BPRS; Ventura et al., 2000). Clinical interviews and supporting materials were reviewed by teams of at least two doctoral-level clinicians to reach diagnostic consensus. The study was approved by the University of Minnesota Institutional Review Board (#1607M90781).

Social cognition measures

A variety of social cognition measures were administered, spanning domains of emotion recognition and mentalizing. Specific measures included the ER-40 (Gur et al., 2001b), POSIT Science Performance-based Prosody Identification Test (Russ et al., 2008), triangle animation task (Abell et al., 2000), and Social Attribution Task-Multiple Choice (Johannesen et al., 2018).

Penn emotion recognition task (ER-40)

In the ER-40 (Gur et al., 2001a, 2001b; Kohler et al., 2003), participants were shown a series of 40 face images, each of which displayed a particular emotion. They were asked to select the appropriate emotion for each face, from the options of “Happy,” “Sad,” “Angry,” “Scared,” and “No Feeling.” Eight stimuli were presented for each of the emotion conditions, split between mild and extreme expressions of each emotion. The faces included Black, White, and Asian individuals, as well as individuals of Hispanic origin. Participants indicated which emotion each face was expressing in a forced choice format. Participants’ accuracy was recorded and scored out of 40.

Performance-based prosody identification test (POSIT-PROID)

We also administered the PROID test (Russ et al., 2008) from the POSIT Science Company’s online test battery (see https://www.brainhq.com). This test of emotion recognition has participants listen to 21 auditory stimuli. For each trial, a recorded speaker reads a neutral sentence aloud in a way that is intended to convey one of six emotions (happiness, sadness, anger, fear, surprise, disgust). Participants were asked to identify each emotion and rate its intensity on a Likert scale. A composite accuracy score based on how many stimuli each participant selected as the correct emotion category was used for the current analyses.

Triangle animation task

In the triangles task (Abell et al., 2000; Barch et al., 2013; Castelli et al., 2002; White et al., 2011), participants were presented with a series of animations of shapes interacting in a way that was random, physical, or social. In the random condition, the shapes did not interact with each other, but rather moved around purposelessly (e.g., bouncing or drifting). In the physical condition, the shapes moved in a goal-directed manner without invoking mentalizing (e.g., mirroring one another). In the social condition, shapes enacted a social sequence, such as surprising or playing. Participants were instructed that there was no “right” or “wrong” answer, and to select the type of interaction they thought had occurred. Participants were tasked with indicating whether each animation was random, physical, or social in nature, then scored for their accuracy. Accuracy was recorded as percent correct across conditions. This version of the task is described in detail by Demro and colleagues (2021) and was adapted from the original Human Connectome Project task-fMRI battery (Barch et al., 2013). Though participants completed this task during fMRI, only the behavioral data were considered in the present study.

Social attribution task-multiple choice (SAT-MC-II)

For the SAT-MC-II (Johannesen et al., 2018, 2013), which is similar to the triangles task, participants first view a short (approximately 1 min) animated video depicting three shapes (triangle, oval, and rectangle) acting out a social drama, which is then played for participants again. Next, the video is shown again in shorter segments, after which a multiple-choice question is asked regarding the clip that was just viewed. Each segment presents unique response options that assess participants’ ability to accurately attribute specific mental states. A total accuracy score was used for the current analyses.

Statistical analyses

A latent social cognition variable was derived via Bayesian confirmatory factor analysis (Muthén and Muthén, 2017) using accuracy scores from all four tasks. We chose Bayesian estimation for its robust handling of missing data through full information methods, flexibility with non-normal distributions common in mixed clinical/control samples, and superior reliability compared to traditional maximum likelihood approaches (Muthén and Muthén, 2017). To test whether the social cognition construct operated equivalently across groups, we conducted measurement invariance testing using robust maximum likelihood estimation in Mplus 8.0. We grouped participants into clinical (SZ and BD, n = 142) and non-clinical (Relatives and HC, n = 154) groups to ensure adequate sample sizes for multi-group modeling, testing configural, metric, and scalar invariance sequentially (Cheung and Rensvold, 2002).

Group differences for the social cognition factor were examined using linear regression (with age and sex as covariates) followed by post-hoc pairwise t-tests with false discovery rate (FDR) correction (Benjamini & Hochberg, 1995). A supplementary analysis separated relatives by proband diagnosis. Associations between social cognition and symptom severity (SAPS/SANS/BPRS) were evaluated using Pearson correlations. Finally, a general linear model tested group-by-task interactions on z-transformed accuracy scores. See supplement for technical details.

Results

Descriptive statistics and measurement model

A summary of demographic differences between the groups is shown in Table 1.

Table 1.

Demographic and symptom characteristics.

Group SZ (N = 105) BD (N = 37) Rel (N = 101) HC (N = 53)
Demographic data
Mean age (SD) 40.25 (12.47) 33.46 (11.48) 44.41 (15.00) 38.70 (13.06)
Sex
Female ( %) 34 (32.38) 27 (72.97) 66 (65.35) 27 (50.94)
Male ( %) 71 (67.62) 10 (27.03) 34 (33.66) 26 (49.06)
Race/Ethnicity
White, not of Hispanic Origin ( %) 67 (63.81) 29 (78.38) 85 (84.16) 47 (88.68)
Black, not of Hispanic Origin ( %) 23 (21.90) 5 (13.51) 8 (7.92) 3 (5.66)
Asian or Pacific Islander ( %) 4 (3.81) 1 (2.70) 1 (0.99) 1 (1.89)
Hispanic ( %) 6 (5.71) 0 (0) 4 (3.96) 1 (1.89)
Other ( %) 5 (4.76) 1 (2.70) 2 (1.98) 1 (1.89)
American Indian or Alaskan Native ( %) 0 (0) 1 (2.70) 0 (0) 0 (0)
Symptoms
SAPS 14.85 (12.36) 4.78 (7.08) — —
SANS 22.34 (11.02) 11.11 (11.77) — —
BPRS 48.75 (12.31) 38.76 (9.08) 33.26 (7.41) 27.75 (4.13)
Open in a new tab

Note. Demographic information was not available for one participant (which is why some percentage values are not standardized). SAPS/SANS = Scales for the Assessment of Positive/Negative Symptoms; BPRS = Brief Psychiatric Rating Scale.

Our social cognition latent variable showed excellent fit (see supplement). Loadings for each of the social cognition tasks onto the latent variable are displayed in Fig. 1. Before examining group differences, we tested measurement invariance between clinical (SZ and BD patients) and non-clinical (relatives and controls) groups. Configural invariance was supported with excellent fit (χ2 = 36.821, df = 38, p = .524, CFI = 1.000), indicating the same factor structure across groups. Metric invariance was also supported (Δχ2 = 4.151, Δdf = 6, p = .656, ΔCFI = 0.000), confirming equivalent factor loadings. However, scalar invariance was not tenable (Δχ2 = 55.615, Δdf = 6, p < .001, ΔCFI = 0.098), indicating differential item intercepts between groups.

Fig. 1. Social cognition latent variable.

Fig. 1.

Open in a new tab

*Credible at ≥ 99 %. SCog = Social Cognition; Triangles = Triangles Task; SAT-MC = Social Attribution Task-Multiple Choice; PROID = POSIT-Science Performance-based Prosody Identification Test Task; ER40 = Penn Emotion Recognition Task. All social cognition tasks loaded strongly onto the social cognition latent variable, and all loadings were credible using 99 % intervals.

Examination of specific task indicators revealed that the failure of scalar invariance was primarily driven by the mentalizing tasks. By examining divergences between the configural and scalar models, we identified the following mean differences between groups: SAT-MC showing the largest standardized mean difference (ΔM = 0.673), followed by Triangles (ΔM = 0.425) then ER40 (ΔM = 0.279) and PROID (ΔM = 0.277). This pattern converges with our substantive findings below, providing psychometric validation that clinical groups experience mentalizing as systematically more difficult.

Group differences

Results of a linear regression predicting social cognition by group—with HC as the reference group—showed a significant main effect of the group factor (F = 18.10, R2 = 0.157, p < .001). Performance was highest in HC and lowest in SZ, with the relatives and BD groups showing intermediate levels of performance (Fig. 2). All pair-wise differences were significant, except for relatives versus BD; that test showed a trend toward worse performance in BD (Table 2).

Fig. 2. Social cognition by group.

Fig. 2.

Open in a new tab

* = significant pair-wise differences at p < .05, †= statistical trend for pair-wise differences at p < .1. Error bars represent standard errors for each group mean. SCog = estimated values of the social cognition latent variable; HC = healthy controls; Rel = first-degree relatives of patients; BD = bipolar disorder with psychotic features; SZ = schizophrenia/schizoaffective disorder. There was a significant effect of group on social cognition (F = 18.10, R2 = 0.157, p < .001), and all pair-wise differences were significant, expect for relatives versus patients with bipolar disorder.

Table 2.

Pairwise comparisons of social cognition across groups.

Group Mean Difference t p p’ Cohen’s d
HC vs. Rel 0.03 2.04 .022 .026 0.34
HC vs. BD 0.06 3.10 .001 .003 0.67
HC vs. SZ 0.10 6.77 < 0.001 < 0.001 1.07
Rel vs. BD 0.03 1.65 .052 .052 0.32
Rel vs. SZ 0.07 5.53 < 0.001 < 0.001 0.77
BD vs. SZ 0.04 2.56 .006 .009 0.46
Open in a new tab

Note. HC =healthy controls; Rel = first-degree relatives of patients; BD = bipolar disorder with psychotic features; SZ = schizophrenia/schizoaffective disorder; p’ = p-values adjusted using false discovery rate correction.

To address whether relatives of SZ and BD patients differed, we conducted a supplementary regression separating relatives by proband diagnosis (F = 14.32, R2 = 0.164, p < .001). This analysis included the 77 relatives who could be definitively classified (50 relatives of SZ patients, 27 relatives of BD patients). Follow-up t-tests (Table S1) showed that there were no significant differences in performance between relatives of SZ patients and relatives of BD patients (t = − 0.93, p = .179). Given this finding and loss of statistical power from excluding 24 unclassifiable relatives, we focus on interpreting our primary analyses of combining all relatives.

Follow-up analyses

When examined across all patients with SZ or BD (Fig. 4), social cognition was correlated negatively with both SAPS (r = − 0.217, p < .01) and SANS (r = − 0.255, p < .01). Although the focus of this study was on positive/negative symptoms—as we were interested primarily in symptoms of psychosis—we also found a negative correlation with mania (r = − 0.184, p = .029) but not depression (r = 0.112, p = .186).

Fig. 4. Relations of social cognition with symptom severity.

Fig. 4.

Open in a new tab

SANS = Scale for the Assessment of Negative Symptoms, SAPS = Scale for the Assessment of Positive Symptoms, SZ = Schizophrenia, BD = Bipolar Disorder.

There was a significant group-by-task interaction on z-transformed values for task accuracy (F = 3.70, p = .015). Percent variance explained by group, for each task, is visualized in Fig. 3. Group was significantly related to accuracy on the SAT-MC, triangles task, and ER-40 but not significantly related to accuracy on the PROID (Table 3).

Fig. 3. Percent variance explained by group for each social cognition task.

Fig. 3.

Open in a new tab

ER40 = Penn Emotion Recognition Task; PROID = Performance-based Prosody Identification Test; SAT_MC = Social Attribution Task-Multiple Choice; Triangles = Triangles Task. There was a significant group-by-task interaction on z-transformed values for task accuracy (F = 3.70, p = .015). While all tasks load strongly on the social cognition factor (Fig. 1), groups differed more on mentalizing tasks (SAT-MC, Triangles) than emotion recognition tasks (PROID, ER40). This pattern was validated by measurement invariance testing, which showed metric invariance (equal loadings) but scalar non-invariance (differential task difficulty between clinical and non-clinical groups—driven by larger performance differences on SAT-MC and Triangles).

Table 3.

Effects of group on social cognition by task.

Task F R2 p
ER40 2.71 .034 .046
PROID 1.09 .028 .358
SAT-MC 10.00 .129 < 0.001
Triangles 3.90 .046 .009
Open in a new tab

Note. ER40 = Penn Emotion Recognition Task; PROID = POSIT-Science Performance-based Prosody Identification Test Task; SAT-MC = Social Attribution Task-Multiple Choice; Triangles = Triangles Task. There was a significant group- by-task interaction on z-transformed values for task accuracy (F = 3.70, p = .015). The effect of group on accuracy was significant for SAT-MC, Triangles, and ER-40 but not PROID.

Discussion

This study demonstrates that social cognitive deficits follow a gradient across the psychosis spectrum: HC > relatives > BD > SZ. By directly comparing all groups within a single study using latent variable modeling, we provide precise effect size benchmarks that advance beyond previous meta-analytic findings. Our approach, which reduces measurement error by extracting shared variance across tasks, establishes social cognition as a quantitative endophenotype for psychotic disorders with immediate implications for intervention development and early identification.

Domain-Specific vulnerabilities

A key finding is that mentalizing tasks showed systematically larger group differences than emotion recognition tasks. Group explained 12.9 % of variance in explicit mentalizing (SAT-MC) but only 2.8 % in prosody recognition (PROID). This pattern received psychometric validation from measurement invariance testing, where scalar non-invariance was driven by mentalizing tasks, confirming that clinical groups experience these as systematically more difficult.

This differential pattern aligns with evidence that SZ and BD show greater impairment in tasks involving cognitive-linguistic versus primarily perceptual components (Chung et al., 2013; Samame et al., 2012). The explicit mentalizing required by SAT-MC—identifying specific mental states—may be particularly vulnerable compared to the implicit mentalizing of the triangles task. These findings suggest social cognitive remediation programs, many emphasizing emotion recognition, may benefit from increased focus on higher-level mentalizing skills.

Developmental context

Our findings gain additional significance when viewed through a developmental lens. Emotion recognition deficits like those measured by our tasks emerge years before psychosis onset, appearing in children who later develop schizophrenia (Dickson et al., 2013), clinical high-risk individuals (Addington et al., 2008), and unaffected siblings (Bediou et al., 2007). Our cross-sectional data showing intermediate deficits in relatives likely captures the endpoint of a developmental cascade beginning with early neurodevelopmental disruptions.

Meta-analyses reveal that children who develop schizophrenia show IQ deficits of as much as 0.43 standard deviations below controls, with accelerated decline during adolescence particularly affecting processing speed and executive function—domains supporting social cognitive development (Khandaker et al., 2011; Rajji et al., 2018). This trajectory aligns with our finding that mentalizing tasks, requiring greater cognitive resources, showed larger group differences than emotion recognition. Our observed correlations between social cognition and negative symptoms (r = − 0.255) may reflect this cascade’s cumulative impact, where early cognitive limitations impair social skill acquisition, potentially contributing to withdrawal and negative symptom emergence.

Genetic liability

While our results support social cognitive deficits as markers of genetic liability, the present design cannot disentangle genetic from environmental influences. Future research incorporating polygenic risk scores (PRS) could provide more direct evidence of genetic contributions (Wray et al., 2014). Previous work has linked schizophrenia PRS to reduced emotion identification speed (Germine et al., 2016), suggesting molecular genetic approaches could clarify whether deficits reflect shared or disorder-specific genetic factors.

The associations between social cognition and symptom severity in both SZ and BD underscore the clinical relevance of these deficits. As most previous work focused on SZ, finding similar patterns in BD is noteworthy, though replication in larger BD samples is warranted. The heterogeneity in psychiatric disorders suggests precision medicine approaches may benefit from tailoring interventions to specific symptom profiles.

Limitations and future directions

Several limitations warrant consideration. Our cross-sectional design cannot establish temporal relationships between social cognitive deficits and symptom development. The chronic nature of our patient sample and medication exposure preclude conclusions about illness chronicity effects. BD patients were selected for psychotic features, limiting generalizability to BD without psychosis. Future studies should examine medication-naive first-episode and clinical high-risk samples to clarify developmental trajectories.

Beyond accuracy, other aspects of social cognition merit investigation. Computational modeling could distinguish hypermentalizing from hypomentalizing errors and clarify whether deficits reflect perceptual limitations or response biases (Lasagna et al., 2024). Additional domains like empathic accuracy and social cue detection could provide a more comprehensive picture of social cognitive dysfunction across the psychosis spectrum.

Finally, our quantified effect sizes may enable evidence-based planning for intervention and early-detection studies. For example, the intermediate deficits in relatives (d = 0.61) occupy a clinically meaningful range—substantial enough to potentially inform early-screening/-intervention efforts without claiming to be deterministic or fully predictive at the individual level. Furthermore, the systematic gradient in deficits from HC to SZ—paralleling patterns for general cognition and negative symptoms—suggests shared neurodevelopmental mechanisms with the potential to be targeted preventively.

Conclusions

Social cognitive deficits represent a core feature of psychotic disorders that scales with genetic liability and relates to functional outcomes. Our multi-task, latent variable approach provides robust evidence for social cognition as an endophenotype spanning diagnostic boundaries. The domain-specific pattern of deficits, with mentalizing particularly affected, offers targets for intervention development. Future integration of molecular genetics and neuroimaging could illuminate biological mechanisms linking genetic risk to social cognitive dysfunction, ultimately informing preventive interventions for those at elevated genetic risk.

Supplementary Material

1

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.psychres.2025.116694.

Acknowledgements

The authors thank the research volunteers for their participation, as well as Jasmine Chao and Julia Hanson for their assistance in data collection and processing.

Funding

This research was supported by the National Institute of Health (U01MH108150 to S.R.S., R01MH122491 and R01MH135117 to I.F.T.). The study was approved by the University of Minnesota Twin Cities Institutional Review Board (1607M90781 Neural Disconnection and Errant Visual Perception in Psychotic Psychopathology).

Footnotes

Prior versions

Preliminary analyses were presented at the 2024 annual meeting of the Society for Research in Psychopathology. This work was posted online as a preprint: https://osf.io/preprints/psyarxiv/mue7h

Declarations of interest statement

The authors declare that they had no competing interests with respect to the authorship or publication of this article.

CRediT authorship contribution statement

Scott D. Blain: Writing – review & editing, Writing – original draft, Visualization, Software, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Jerillyn S. Kent: Writing – review & editing, Methodology, Data curation, Conceptualization. Chloe A. Peyromaure de Bord: Writing – review & editing, Methodology, Data curation. Scott R. Sponheim: Writing – review & editing, Resources, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization. Ivy F. Tso: Writing – review & editing, Supervision, Project administration, Methodology, Funding acquisition, Conceptualization.

Declaration of competing interest

The author(s) declare that there were no competing interest with respect to the authorship or the publication of this article.

References

  1. Abell F, Happé F, Frith U, 2000. Do triangles play tricks? Attribution of mental states to animated shapes in normal and abnormal development. Cogn. Dev. 15, 1–16. 10.1016/S0885-2014(00)00014-9. [DOI] [Google Scholar]
  2. Addington J, Penn D, Woods SW, Addington D, Perkins DO, 2008. Facial affect recognition in individuals at clinical high risk for psychosis. Br. J. Psychiatry 192 (1), 67–68. 10.1192/bjp.bp.107.039784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Andreasen NC, Grove WM, 1986. Evaluation of positive and negative symptoms in schizophrenia. Psychiatry Psychobiol. 1, 108–122. 10.1017/S0767399×00003199. [DOI] [Google Scholar]
  4. Apperly IA, 2012. What is “theory of mind”? Concepts, cognitive processes and individual differences. Q. J. Exp. Psychol. 65, 825–839. 10.1080/17470218.2012.676055. [DOI] [PubMed] [Google Scholar]
  5. Barch DM, Burgess GC, Harms MP, Petersen SE, Schlaggar BL, Corbetta M, Glasser MF, Curtiss S, Dixit S, Feldt C, Nolan D, Bryant E, Hartley T, Footer O, Bjork JM, Poldrack R, Smith S, Johansen-Berg H, Snyder AZ, Van Essen DC, 2013. Function in the human connectome: task-fMRI and individual differences in behavior. NeuroImage Mapp. Connect. 80, 169–189. 10.1016/j.neuroimage.2013.05.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bediou B, Asri F, Brunelin J, Krolak-Salmon P, D’Amato T, Saoud M, Tazi I, 2007. Emotion recognition and genetic vulnerability to schizophrenia. Br. J. Psychiatry 191 (2), 126–130. 10.1192/bjp.bp.106.028829. [DOI] [PubMed] [Google Scholar]
  7. Bora E, Özerdem A, 2017. Social cognition in first-degree relatives of patients with bipolar disorder: a meta-analysis. Eur. Neuropsychopharmacol. 27, 293–300. 10.1016/j.euroneuro.2017.02.009. [DOI] [PubMed] [Google Scholar]
  8. Bora E, Pantelis C, 2016. Social cognition in schizophrenia in comparison to bipolar disorder: a meta-analysis. Schizophr. Res. 175, 72–78. 10.1016/j.schres.2016.04.018. [DOI] [PubMed] [Google Scholar]
  9. Bortolato B, Miskowiak KW, Köhler CA, Vieta E, Carvalho AF, 2015. Cognitive dysfunction in bipolar disorder and schizophrenia: a systematic review of meta-analyses. Neuropsychiatr. Dis. Treat 11, 3111–3125. 10.2147/NDT.S76700. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Castelli F, Frith C, Happé F, Frith U, 2002. Autism, Asperger syndrome and brain mechanisms for the attribution of mental states to animated shapes. Brain 125, 1839–1849. 10.1093/brain/awf189. [DOI] [PubMed] [Google Scholar]
  11. Cheung GW, Rensvold RB, 2002. Evaluating goodness-of-fit indexes for testing measurement invariance. Struct. Equ. Model. 9 (2), 233–255. [Google Scholar]
  12. Chung YS, Barch D, Strube M, 2013. A meta-analysis of mentalizing impairments in adults with schizophrenia and Autism spectrum disorder. Schizophr. Bull. 40, 602. 10.1093/schbul/sbt048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Demro C, Mueller BA, Kent JS, Burton PC, Olman CA, Schallmo M-P, Lim KO, Sponheim SR, 2021. The psychosis human connectome project: an overview. Neuroimage 241, 118439. 10.1016/j.neuroimage.2021.118439. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dickson H, Calkins ME, Kohler CG, Hodgins S, Laurens KR, 2013. Misperceptions of facial emotions among youth aged 9–14 years who present multiple antecedents of schizophrenia. Schizophr. Bull. 40 (2), 460–468. 10.1093/schbul/sbs193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fett A-KJ, Viechtbauer W, Dominguez M-G, Penn DL, van Os J, Krabbendam L, 2011. The relationship between neurocognition and social cognition with functional outcomes in schizophrenia: a meta-analysis. Neurosci. Biobehav. Rev. 35, 573–588. 10.1016/j.neubiorev.2010.07.001. [DOI] [PubMed] [Google Scholar]
  16. Friedman NP, Miyake A, Young SE, DeFries JC, Corley RP, Hewitt JK, 2008. Individual differences in executive functions are almost entirely genetic in origin. J. Exp. Psychol. Gen. 137, 201–225. 10.1037/0096-3445.137.2.201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Germine L, Robinson EB, Smoller JW, Calkins ME, Moore TM, Hakonarson H, Daly MJ, Lee PH, Holmes AJ, Buckner RL, Gur RC, Gur RE, 2016. Association between polygenic risk for schizophrenia, neurocognition and social cognition across development. Transl. Psychiatry 6, e924. 10.1038/tp.2016.147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Green MF, Horan WP, Lee J, 2015. Social cognition in schizophrenia. Nat. Rev. Neurosci. 16, 620–631. 10.1038/nrn4005. [DOI] [PubMed] [Google Scholar]
  19. Gur RC, Ragland JD, Moberg PJ, Bilker WB, Kohler C, Siegel SJ, Gur RE, 2001a. Computerized neurocognitive scanning:: II. The profile of schizophrenia. Neuropsychopharmacology 25, 777–788. 10.1016/S0893-133X(01)00279-2. [DOI] [PubMed] [Google Scholar]
  20. Gur RC, Ragland JD, Moberg PJ, Turner TH, Bilker WB, Kohler C, Siegel SJ, Gur RE, 2001b. Computerized neurocognitive scanning:: I. Methodology and validation in healthy people. Neuropsychopharmacology 25, 766–776. 10.1016/S0893-133X(01)00278-0. [DOI] [PubMed] [Google Scholar]
  21. Harvey PD, Patterson TL, Potter LS, Zhong K, Brecher M, 2006. Improvement in social competence with short-term atypical antipsychotic treatment: a randomized, double-blind comparison of quetiapine versus risperidone for social competence, social cognition, and neuropsychological functioning. Am. J. Psychiatry 163, 1918–1925. 10.1176/ajp.2006.163.11.1918. [DOI] [PubMed] [Google Scholar]
  22. Jeste DV, Palmer BW, Appelbaum PS, Golshan S, Glorioso D, Dunn LB, Kim K, Meeks T, Kraemer HC, 2007. A new brief instrument for assessing decisional capacity for clinical research. Arch. Gen. Psychiatry 64, 966–974. 10.1001/archpsyc.64.8.966. [DOI] [PubMed] [Google Scholar]
  23. Johannesen JK, Fiszdon JM, Weinstein A, Ciosek D, Bell MD, 2018. The social attribution task - multiple choice (SAT-MC): psychometric comparison with social cognitive measures for schizophrenia research. Psychiatry Res. 262, 154–161. 10.1016/j.psychres.2018.02.011. [DOI] [PubMed] [Google Scholar]
  24. Johannesen JK, Lurie JB, Fiszdon JM, Bell MD, 2013. The social attribution task- multiple choice (SAT-MC): a psychometric and equivalence study of an alternate form. Int. Sch. Res. Not. 2013, e830825. 10.1155/2013/830825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kent J, Pinkham A, 2024. Cerebral and cerebellar correlates of social cognitive impairment in schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry 128, 110850. 10.1016/j.pnpbp.2023.110850. [DOI] [PubMed] [Google Scholar]
  26. Khandaker GM, Barnett JH, White IR, Jones PB, 2011. A quantitative meta-analysis of population-based studies of premorbid intelligence and schizophrenia. Schizophr. Res. 132 (2–3), 220–227. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kohler CG, Turner TH, Bilker WB, Brensinger CM, Siegel SJ, Kanes SJ, Gur RE, Gur RC, 2003. Facial emotion recognition in schizophrenia: intensity effects and error pattern. Am. J. Psychiatry 160, 1768–1774. 10.1176/appi.ajp.160.10.1768. [DOI] [PubMed] [Google Scholar]
  28. Lasagna CA, Tso IF, Blain SD, Pleskac TJ, 2024. Cognitive mechanisms of aberrant self-referential social perception in psychosis and bipolar disorder: insights from computational modeling. Schizophr. Bull. sbae 147. 10.1093/schbul/sbae147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lavoie M-A, Plana I, Bédard Lacroix J, Godmaire-Duhaime F, Jackson PL, Achim AM, 2013. Social cognition in first-degree relatives of people with schizophrenia: a meta-analysis. Psychiatry Res. 209, 129–135. 10.1016/j.psychres.2012.11.037. [DOI] [PubMed] [Google Scholar]
  30. Legge SE, Santoro ML, Periyasamy S, Okewole A, Arsalan A, Kowalec K, 2021. Genetic architecture of schizophrenia: a review of major advancements. Psychol. Med. 51, 2168–2177. 10.1017/S0033291720005334. [DOI] [PubMed] [Google Scholar]
  31. Muthén LK, Muthen BO, 2017. Mplus User’s Guide, 8th ed. Muthén & Muthén, Los Angeles, CA. [Google Scholar]
  32. Russ JB, Gur RC, Bilker WB, 2008. Validation of affective and neutral sentence content for prosodic testing. Behav. Res. Methods 40, 935–939. 10.3758/BRM.40.4.935. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Samamé C, Martino DJ, Strejilevich SA, 2012. Social cognition in euthymic bipolar disorder: systematic review and meta-analytic approach. Acta Psychiatr. Scand. 125, 266–280. 10.1111/j.1600-0447.2011.01808.x. [DOI] [PubMed] [Google Scholar]
  34. Sullivan PF, Kendler KS, Neale MC, 2003. Schizophrenia as a complex trait: evidence from a meta-analysis of twin studies. Arch. Gen. Psychiatry 60, 1187–1192. 10.1001/archpsyc.60.12.1187. [DOI] [PubMed] [Google Scholar]
  35. Thaler NS, Allen DN, Sutton GP, Vertinski M, Ringdahl EN, 2013. Differential impairment of social cognition factors in bipolar disorder with and without psychotic features and schizophrenia. J. Psychiatr. Res. 47, 2004–2010. 10.1016/j.jpsychires.2013.09.010. [DOI] [PubMed] [Google Scholar]
  36. Ventura J, Nuechterlein KH, Subotnik KL, Gutkind D, Gilbert EA, 2000. Symptom dimensions in recent-onset schizophrenia and mania: a principal components analysis of the 24-item Brief Psychiatric Rating Scale. Psychiatry Res. 97, 129–135. [DOI] [PubMed] [Google Scholar]
  37. White SJ, Coniston D, Rogers R, Frith U, 2011. Developing the Frith-Happé animations: a quick and objective test of Theory of Mind for adults with autism. Autism. Res. 4, 149–154. 10.1002/aur.174. [DOI] [PubMed] [Google Scholar]
  38. Wray NR, Lee SH, Mehta D, Vinkhuyzen AAE, Dudbridge F, Middeldorp CM, 2014. Research review: polygenic methods and their application to psychiatric traits. J. Child Psychol. Psychiatry 55, 1068–1087. 10.1111/jcpp.12295. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2181734-supplement-1.docx (19.5KB, docx)

RESOURCES