Theory of Mind (ToM), the ability to infer others' mental states, is central to social interaction and may be sensitive to age-related differences. This study explored variations in cognitive and affective ToM across adulthood and examined the roles of executive functions (EFs), general cognition, and Cognitive Reserve (CR). Ninety-six participants were assigned to three groups: Young Adults (20-40), Older Adults (65-75), and Senior Older Adults (76-86). Cognitive ToM was assessed with the Strange Stories and II Order Stories tasks, and affective ToM with the Reading the Mind in the Eyes Test. Measures of processing speed, working memory, long-term memory, EFs (updating, shifting, inhibition), and CR were included. Both cognitive and affective ToM varied by age group, with younger adults performing better than older groups. Updating and shifting predicted ToM performance, accounting for group differences in cognitive ToM and partially in affective ToM. CR, particularly occupational activity, moderated the association between age group and cognitive ToM: differences emerged at low CR but not at high CR levels. Findings suggest that age-related differences in ToM are associated with executive functioning and that higher CR may mitigate these differences, supporting the role of cognitive engagement in social-cognitive maintenance.
Various mental disorders are associated with impaired cognitive control, which is crucial for effective emotion regulation. Cognitive control training has demonstrated promise in enhancing emotion regulation and alleviating distress in disorders characterized by repetitive negative thinking, such as depression and anxiety. Given the importance of cognitive control and emotion regulation across mental disorders, this study investigates the efficacy of a mobile cognitive control training in a transdiagnostic outpatient sample awaiting psychotherapy. In this randomized clinical superiority trial with 2 parallel arms, 80 patients with various mental disorders from an outpatient waiting list received either 10 sessions of mobile cognitive control training using the Paced Auditory Serial Addition Test (PASAT) or an active control training using a speed of response task. The primary outcome was mental distress, measured by the Hopkins Symptom Checklist-11 (HSCL-11). Secondary outcomes included measures of cognitive control, rumination, repetitive negative thinking, difficulties in emotion regulation, cognitive emotion regulation, and disorder-specific symptoms. Outcomes were measured at baseline, post training, and at 3-month and 6-month follow-up. Contrary to our primary hypothesis, cognitive control training was not superior in improving global mental distress directly after training (B=-.03, 95% CI -0.21, 0.16; t179.60=-0.26; P=.80; d=-0.04, 95% CI -0.35, 0.28); however, it led to greater improvements in cognitive control (B=-0.56, 95% CI -0.59,-0.54; z=-18.02; P<.001; d=-1.23, 95% CI -1.30,-1.20). This effect was similar at the 3-month and 6-month follow-up. Furthermore, at 3-month follow-up, cognitive control training resulted in fewer difficulties in emotion regulation (B=4.73, 95% CI 0.52, 9.12; t177.99=2.09; P=.04; d=0.34, 95% CI 0.04, 0.65), and anxiety symptoms (B=2.94, 95% CI 0.38, 5.82; t66.51=2.09; P=.04; d=0.70, 95% CI 0.09, 1.38), although the latter refers to a small subsample of patients with anxiety disorders. At 6-month follow-up, cognitive control training led to more adaptive cognitive emotion regulation (B=-5.18, 95% CI -9.74,-0.41; t180.90=-2.16; P=.03; d=-0.40, 95% CI -0.75,-0.03), and less social anxiety (B=2.00, 95% CI 0.14, 3.81; t43.43=2.08; P=.04; d=0.66, 95% CI 0.05, 1.24). The groups did not differ in any other outcome at any point in time. This study is the first to assess the efficacy of a mobile cognitive control training using the PASAT in a transdiagnostic outpatient sample. There was no evidence for the training's efficacy on global mental distress and only weak evidence for the superiority in measures of emotion regulation and anxiety at follow-ups. Potential mediating pathways and moderating factors, such as the number of training sessions, should be investigated in larger studies.
The environment, especially social features, plays a key role in shaping the development of the brain, notably during adolescence. To better understand variation in brain-environment coupling and its associated outcomes, we identified 'social envirotypes', or different patterns of social environment experience, in the Adolescent Brain Cognitive Development Study by hierarchically clustering subjects. Two focal clusters, which accounted for 89.3% of all participants, differed significantly on eight out of nine youth-report social environment quality measures, representing almost perfect complements. We then applied tools from network neuroscience to show different social envirotypes are associated with different patterns of whole brain functional connectivity. Differences were distributed across the brain but were especially prominent in Default and Somatomotor Hand systems for these focal clusters. Finally, we examined how social envirotypes change over development and how these patterns of change are associated with a suite of outcomes. The resulting dynamic social envirotypes differed along dimensions of stability and quality, but outcomes diverged based on stability. Altogether, our findings represent significant contributions to both social developmental neuroscience and network neuroscience, emphasizing the variability and dynamicity of brain-environment coupling and its consequences.
Adolescence is a period characterized by social and biological changes that are associated with increases in sensation seeking and risk-taking behaviors. Unpredictable environments may influence neurocognitive development and are associated with increases in risky behaviors and poor social functioning. The majority of research linking adversity and risky behaviors has focused on negative risk taking; however, research has started to distinguish between negative and positive risk-taking across different domains. Thus, the current study used seven years of longitudinal data to examine how unpredictable life events are associated with positive social risk taking through changes in neural cognitive control in frontoparietal circuitry. Results of conditional growth curve modeling indicated that unpredictability predicted higher levels of frontoparietal activation in late adolescence, which, in turn, were associated with lower positive social risk taking. Significant indirect effects revealed that unpredictability was associated with lower positive social risk taking in young adulthood through higher frontoparietal activation during cognitive control in late adolescence. The findings have important implications for understanding the antecedents of risk-taking behaviors by highlighting the role of neurocognitive functioning in linking environmental unpredictability to positive social risk outcomes.
Empathy allows us to infer the affective state of another individual and resonate with it. Accumulating evidence suggests that on a neural level, empathizing relies on 'shared neural representations', i.e., patterns of neural activation recruited both during the first-hand and the empathic experience of a specific affective state. Studies employing placebo analgesia have shown consistent reductions in behavioral ratings and neural activity for both firsthand and empathic pain. The mechanistic interpretation of such effects, however, remains elusive, as placebo analgesia could exert its effects on empathy either via pharmacological actions or via top-down cognitive processes on pain and empathy beliefs. To address this limitation, this double-blind placebo-controlled study (N = 35, 21 female) administered the opioid antagonist naltrexone and tested its effects on firsthand pain and affective and cognitive ratings of empathy for pain. While we predicted that naltrexone would increase both electrocutaneous and cold pain, as well as cognitive and affective aspects of empathy for pain, the results instead pointed in the other direction. While these hypo- rather than hyperalgesic effects were unexpected, the coherence in their directionality fits with previous findings and suggests the involvement of shared opioidergic mechanisms in the firsthand experience of pain and empathy for pain.
Autism spectrum disorder (ASD) and schizophrenia spectrum disorder (SSD) both involve social-cognitive difficulties but may rely on opposite predictive mechanisms. Predictive-coding and diametrical accounts propose weak priors in ASD versus strong priors in SSD. We tested whether autistic and positive schizotypal traits differentially shape the use of social priors when interpreting ambiguous social cues. Two studies with non-clinical adults employed an adapted Animated Triangles Task (ATT) with three cueing conditions: Random-Uncued, Random-Cued, and ToM-Cued. Study 1 (n = 71) assessed behavior; Study 2 (n = 45) combined behavioral data with task-based functional magnetic resonance imaging (fMRI), including region-of-interest (ROI) and psychophysiological interaction (PPI) analyses of the temporoparietal junction (TPJ), medial prefrontal cortex (mPFC), and cerebellar Crus II. Social-attribution ratings increased across conditions, confirming the cueing manipulation. At the uncorrected level, higher positive schizotypal traits were associated with greater social attribution without cues, whereas higher autistic traits were associated with lower intention ratings despite strong cues. ToM-Cued trials activated the mentalizing network. Both trait dimensions showed exploratory associations with reduced mentalizing-network connectivity. These preliminary findings suggest potentially shared neural patterns coupled with divergent behavioral responses across the autism-psychosis trait continuum, pending further validation with larger sample sizes and adjusted statistical analyses.
For a long time, from the nineteenth century to most of the twentieth century, the cerebellum was thought to be an organ that regulates movement. Towards the end of the twentieth century, the brain functions associated with the cerebellum began to extend beyond motor control. Now, there is a consensus that the cerebellum is involved not only in motor functions but also in the most basic autonomic functions and the most complex cognitive and emotional functions, with a focus on predictions and internal models. A new functional model of the cerebellum is needed to explain all layers of brain functions by extending predictive computations in the cerebellum. On the other hand, the cerebellum and the basal ganglia were believed to be independent and complementary motor centers that lacked direct neural connections. For example, in neurophysiology classes in the 1980s, the characteristics of cerebellar ataxia were summarized as hyperkinetic and hypotonia, while the characteristics of Parkinson's disease (traditionally classified as "basal ganglia disorder") were summarized as hypokinetic and hypertonia, and therefore their functions were assumed at opposite poles, without interactions between the two main subcortical systems. The cerebellum and the basal ganglia were also assigned contrasting models regarding their learning mechanisms. Namely, the cerebellum was assumed to employ supervised learning with error signals, while the basal ganglia were assumed to employ reinforcement learning with reward prediction errors. However, recent neuroanatomical studies have demonstrated a number of novel connections between them, questioning their independence. Moreover, recent single-neuron recording and inactivation studies provided evidence that the cerebellum may also be involved in reinforcement learning. The cerebellum is neither independent of the basal ganglia nor exclusively specialized for supervised learning. We now need a new, general model to explain the contradiction between the known uniformity of the cerebellar cortex's structure and the newly added diversity of brain functions to which the cerebellum contributes. This consensus paper summarizes many of the seeds of such a new theory. The panel of experts (1) highlights the importance of the anatomical connectivity between cerebellar circuitry and basal ganglia, (2) points out that the anatomy of the cerebellum is unique and allows predictive computations in motor and extra-motor domains such as cognition, affect, social interactions and reward processes, (3) underlines the need to further elucidate the nature of interactions between cerebellar cortex and cerebellar nuclei to better understand cerebellar and psychiatric disorders and (4) suggests that common operations may underlie the motor and non-motor functions of the cerebellar circuitry. Cerebellar models remain a major topic of research to improve our understanding of the numerous cerebellar activities and to better understand the complexity of cerebellar disorders.
Social exclusion threatens belongingness, triggering adaptive reward processing changes. It remains unknown whether exclusion selectively alters social (vs. monetary) reward processing via hypervigilance or compensatory affiliation mechanisms. To address this, we recruited 61 participants to complete a Cyberball-based social exclusion or inclusion induction, with an electroencephalogram recorded during subsequent social and monetary reward tasks. Results revealed that positive versus negative feedback elicited stronger reward positivity in the social inclusion condition, while such an effect was eliminated by social exclusion. However, this interaction effect was absent for P300 during the consummatory stage or the stimulus preceding negativity during the anticipation stage. Meanwhile, social exclusion reduced the delta power, enhanced positive social evaluations, and led to a dissociation between this positive evaluation and the delta power in response to others' social feedback. In contrast, for monetary reward processing, no effects of social exclusion were found. This domain-specific effect was further validated by cross-task comparison. These results suggest that social exclusion selectively attenuates the initial cognitive processing of social rewards, disrupts the association between neural feedback and behavior, and may trigger a dual coping mechanism involving withdrawal self-protective neuroadaptations, and approach-driven prosocial behavioral tendencies.
Perceivers' beliefs are critical to form social judgments. However, it remains unclear how the strength of these beliefs or predictions shapes social perception. Here, we investigated how the representation of social perceptions (trustworthiness and likeability ratings) and bodily responses (skin conductance response; SCR) are modulated by prediction strength, using representational similarity analysis. In two samples of university students, stronger predictions led participants to judge identities as more trustworthy and likeable when identities' facial reactions conformed to their predictions compared to when they were violated. Furthermore, the representational structure of social judgments and SCR were associated with models of prediction fulfillment/violation. More importantly, the similarity between social judgments, SCR representations, and models of prediction violation were positively linked to predictions strength. These results demonstrate that predictions strength modulates how fulfilled/violated predictions influence social judgments and bodily responses, providing new insights into the mechanisms underlying social perception.
Understanding the neural mechanisms underlying the impact of psychedelics on social perception and cognition may be instrumental to unravel their therapeutic potential. We conducted a pharmacoimaging study to examine ayahuasca's effects on a key theory of mind region, at the core of the third visual pathway (TVP)-the posterior superior temporal sulcus (pSTS), which is involved in facial emotion recognition and social perception. Twelve healthy participants (mean age: 40 ± 6.6 years; four females) completed a crossover design with three conditions: 0.5 mg/kg N, N-dimethyltryptamine (DMT), 0.8 mg/kg DMT, and placebo, with 1-2 months washout intervals. Resting-state functional magnetic resonance imaging (fMRI) was used to assess pSTS functional and effective connectivity. The highest dose significantly increased right pSTS connectivity and directed modulation from visual (primary and extrastriate cortices) and mirror neuron regions (supplementary motor cortex; SMC). Subjectively, this enhanced social cognitive states, with a strong positive correlation between pSTS-SMC connectivity and perspective-taking experiences. Additionally, ayahuasca produced positive psychological effects, including improved perceived social relationships, at 1-week follow-up despite minimal acute effects. Our findings reveal a novel mechanism of action of psychedelics at early stages of social information processing, with enhanced integration of the TVP and mirror neuron systems. The pSTS emerged as a critical hub supported by top-down and bottom-up evidence, providing a basis for understanding ayahuasca's prosocial therapeutic effects.
Traditional assessments of depressive symptoms often rely on retrospective self-reports, which may be affected by cognitive and memory biases. Few studies have compared retrospective and dynamic (real-time) assessments to examine the consistency and structure of depressive symptom reporting. This study aimed to compare retrospective and dynamic assessments of depressive symptoms in youth using network analysis to explore symptom-level associations and clustering. Ninety Brazilian adolescents and young adults (mean age = 18 years), with and without depression, completed the Short Mood and Feelings Questionnaire (SMFQ) every other day for 14 days via a smartphone-based chatbot (dynamic assessment). At the end of the 2-week period, they completed the same questionnaire retrospectively. Network analyses were conducted using Exploratory Graph Analysis (EGA) and Dynamic Exploratory Analysis (DEA) to identify symptom communities and compare network structures across both assessment methods. Both retrospective and dynamic assessments revealed three symptom communities; however, the composition and structure of these communities differed. Retrospective assessments showed stronger connections among cognitive symptoms, while dynamic assessments displayed a more balanced distribution, with stronger associations between somatic and affective symptoms. Findings highlight significant differences in depressive symptom networks between retrospective and dynamic assessments. Dynamic methods may offer less biased and more ecologically valid insights into youth depression, underscoring the importance of real-time data collection in clinical assessment and research.
Fairness is essential for balancing interests and mitigating social conflict. People's rejection of unfairness is influenced by contextual factors (e.g. empathic concern) and individual traits including social value orientation (SVO). This study examines how individuals with different SVOs make trade-offs between empathic concern and fairness without involving their own interests. Participants played a modified ultimatum game, in which they made decisions on behalf of either a beneficiary of a public welfare project (empathy) or a stranger (non-empathy), choosing whether to accept or reject allocation offers. Results showed that empathy increased participants' tolerance for unfair offers, particularly among prosocials, who accepted more disadvantageous offers than proselfs did. EEG results showed that proselfs exhibited reduced N1 amplitudes in empathic conditions, reflecting attentional avoidance. Moreover, in the empathy condition, an unfairness-related MFN was observed only in proselfs. The absence of this typical MFN response in prosocials provides neural evidence for their active downregulation of unfairness aversion to prioritize the interests of empathic targets. In addition, prosocials showed stronger parietal-occipital alpha suppression and reduced P3 amplitude in empathy contexts, indicating heightened attentional arousal and greater allocation of cognitive resources. These findings highlight the crucial role of empathic concern and SVO in fairness decision-making.
A central unresolved issue in affective neuroscience is whether human emotions have unique biological signatures in the brain and body. Despite decades of debate, a consensus is lacking on whether emotions are patterned (similar across contexts and characterized by distinct features) or flexible (variable across contexts and lacking distinct features). Studies in other species have revealed the ubiquity of autonomic and motor patterns, and these investigations have elucidated the biology of central pattern generation at the levels of networks, neurons, and synapses. Here, we integrate the knowledge gained from this research and introduce the Dynamic Emotion Fabric Theory (DEFT), a new model of emotions biology. DEFT proposes that, just as fabrics are comprised of both patterns and textures, emotions are accompanied by stereotyped responses that are distinct and recognizable, yet also nuanced and malleable. A neurobiological system that is both automatic-generating patterned reactions essential for survival-and flexible-texturizing each response in context-is equipped to produce the spectrum of human emotions. DEFT provides new insights into the biological basis of emotions and identifies novel areas that warrant further study in humans and other species. This framework has implications for basic affective neuroscience and clinical studies of affective symptoms.
Vasopressin, a key molecular regulator of social behavior, is implicated in promoting self-protective responses to threats against physical safety and resources. However, its role in defending against self-view threats-common in social interactions and critical to well-being-remains unclear. This study investigates the neural mechanisms through which vasopressin modulates self-protective responses to spontaneous social comparison. In a double-blind, placebo-controlled neuroimaging study, 48 healthy male participants were randomized to receive intranasal vasopressin or placebo. Participants then underwent fMRI while rating their satisfaction with social evaluations and monetary outcomes assigned to a stranger, friend, or themselves. Compared to placebo, vasopressin selectively intensified contrastive emotional responses to social evaluations of the stranger, decreasing satisfaction with positive evaluations and increasing satisfaction with negative evaluations, relative to those of the self or friend. At the neural level, vasopressin reduced the neural distinction between the stranger and the self/friend in medial prefrontal cortex (mPFC), an effect that was stronger in more socially dominant individuals. Complementing this regional effect, vasopressin attenuated whole-brain multivariate differentiation and reduced functional connectivity from ventral mPFC to the temporoparietal junction and precuneus. Together, these converging neuroimaging findings suggest that vasopressin may increase the psychological salience of strangers and thereby modulate their neural representations, potentially engaging self-protective affective processes. Our findings elucidate novel neuropsychological mechanisms through which vasopressin amplifies emotional defense against social threats and offer insights into potential clinical applications for psychiatric conditions characterized by impaired self-protection.
Traumatic events can profoundly impact cognition, mood, and behavior, manifesting in traumatic stress symptoms (TSS). Perceived injustice (PI) reflects cognitive appraisals involving severity of loss and irreparability, blame, and unfairness. We examined the relationship between PI and TSS in an Israeli convenience sample following the October 7th attack and subsequent war. We pre-registered our hypotheses that PI would positively associate with current and predict future TSS severity, beyond emotional distress-related symptoms, namely depression, anxiety, and anger. Data were collected via electronic surveys distributed through social media three months after the attack in January 2024 (TP1: n = 1693; age = 38.45 ± 13.62; 56 % female), and follow-up collected nine months after the attack in July 2024 (TP2: n = 613; age = 38.39 ± 12.69; 61 % female). TSS were measured using the PTSD Checklist-5, PI with the Injustice Experience Questionnaire, and emotional distress-related measures using NIH's Patient Reported Outcomes Measurement Information System scales. At both timepoints, PI correlated with TSS (TP1: r = 0.74, p < 0.001; TP2: r = 0.71, p < 0.001). Emotional distress-related measures similarly correlated with TSS and PI (r = 0.50-0.78; p-values<0.001). Regression analysis demonstrated that PI explained variance in TSS above and beyond emotional distress-related measures: 9.6 % at TP1 (p < 0.001) and 6.4 % at TP2 (p < 0.001). PI at TP1 correlated with TSS at TP2 (r = 0.61, p < 0.001) with explained variance of 4.1 % (p < 0.001) above and beyond TP1 emotional distress-related measures. Findings supported our pre-registered hypotheses, highlighting PI as a crucial yet understudied risk-factor for the development and maintenance of TSS. Clinically, addressing appraisals of injustice seems crucial to better understand and support coping with trauma.
Neuroticism is significantly associated with various psychiatric disorders. Individuals exhibiting high levels of neuroticism are more susceptible to experiencing anxiety, depression, and other negative emotional responses. Research on the differences in macroscopic functional connectivity gradients among neuroticism levels and their associations with microscopic transcriptomics remains scarce. This study explores the associations between functional gradient and transcriptional expression in neuroticism across 109 individuals with low neuroticism (LNG) and 210 with medium-high neuroticism (MHNG). We analyzed functional gradient alterations in MHNG and their correlations with neurotransmitters, meta-analytic cognitive terms, and transcriptional patterns using partial least squares regression (PLS), involving similarity with major psychiatric disorders, functional enrichments, developmental stages, cortical layers, and specific cell types. MHNG exhibited functional gradient changes within the default, limbic, and visual networks, which correlated with higher-order cognitive terms and alterations in several neurotransmitters. We identified significant overlaps between PLS1 weighted genes and those dysregulated in schizophrenia and autism. Genes linked with gradient alterations were enriched in synaptic signaling, infection and metabolism, astrocytes, specific cortical layers, and developmental phases from early fetal to young adulthood. These findings offer a critical theoretical foundation for understanding the complex relationship between macroscopic functional gradients and microscopic transcriptional patterns across various neuroticism levels.
Accurately decoding emotions from facial and vocal cues is essential to successful social interaction. The human mirror neuron system (hMNS) is thought to support this through sensorimotor simulation of observed emotional expressions. While prior studies linked hMNS activity-indexed by mu rhythm desynchronization (mu-ERD)-to emotional action perception, causal evidence with dynamic, multimodal social stimuli remains limited. We investigated whether transcranial random noise stimulation (tRNS) to the inferior frontal cortex (IFC), a key node of the hMNS, enhances perception of static and dynamic emotional displays. Fifty-two participants received active or sham tRNS over the IFC. Consistent with pre-registered predictions, active tRNS led to better performance on static emotion perception tasks compared to sham, with no group difference on a static identity-matching control task-validating stimulation specificity. Extending prior work, active tRNS led to faster response times and greater mu-ERD measured by EEG during dynamic audio-visual emotion perception tasks, consistent with predictions relating to enhanced sensorimotor simulation. These findings suggest that tRNS to the IFC can augment rapid, embodied emotion perception-particularly when stimuli more closely approximate real-world social communication-and contribute to the causal mapping of the hMNS, opening new avenues for studying social-emotional function in neurotypical and clinical populations.
This study investigates the relationship between gestational age (GA) and resting-state functional connectivity (rsFC) in a cohort of very preterm children at school age, and how these neural patterns relate to cognitive and theory of mind (ToM) performance. Resting-state functional magnetic resonance imaging (fMRI) data were collected from 52 children (GA < 32 weeks, birth weight <1500 g) and independent component analysis was applied to extract the resting-state networks. Results showed that GA was positively associated with rsFC of the precuneus and the paracentral region within the left posterior cerebellar network (lpCER), while negatively associated with rsFC of the insula and putamen within the anterior default mode network (DMN), and with rsFC of the postcentral gyrus within the right frontoparietal network (rFPN). Cognitive and neuropsychological assessments revealed that increased connectivity involving the lpCER correlated with better verbal comprehension, visuospatial ability, fluid reasoning, working memory, and ToM performance. Conversely, increased aDMN connectivity was associated with lower working memory and decreased rFPN connectivity was found associated with lower intelligence quotient. These results underscore the influence of GA on intrinsic brain networks supporting cognitive and socio-cognitive functions, and highlight potential neural markers that could inform targeted intervention strategies for preterm children.
Adolescents' self-identity is shaped by online feedback's valence​ (positive/negative) and social Distance(near/far). This study tested their interactive effects on implicit self-identity and neural correlates.Experiment 1​used an implicit association measure (D-score) to assess self-identity across ages. Participants received near (peer) or far (stranger) positive/negative feedback. Results: Distant positive feedback reduced D-scores (weaker self-identity links) vs. near positive feedback; negative feedback showed the opposite. Late adolescents had higher D-scores for near vs. far feedback-indicating heightened sensitivity to proximal social cues during identity formation.Experiment 2​employed ERPs to explore neural mechanisms. Distant positive feedback elicited faster P3s (prioritized processing of socially distant positives). For near feedback, positive vs. negative stimuli showed smaller early LPP amplitudes-reflecting differential motivational attention.These findings reveal that feedback valence and social distance jointly modulate implicit self-identity. Neural markers (P3/LPP) uncover distinct processing priorities for proximal/distal positive/negative feedback-highlighting the interplay of social context and neural mechanisms in adolescent self-development.
Understanding others' intentions and emotions is supported by the mentalizing network, helping people navigate the inherently complex dynamics of social interactions. However, work investigating the mentalizing network often relies on non-naturalistic methods, limiting our understanding of individual differences most apparent within social interaction contexts. We took a naturalistic, multilevel approach to investigate relations between real-world affective expressions, social anxiety symptoms and mentalizing network connectivity. Same gender-identity friend pairs (40 participants; 62.5% Women; 87.5% White) engaged in a social interaction while videos from their individual visual perspectives were recorded. Participants were later scanned while they watched clips of the interaction from their own (self run) and their friend's (friend run) perspective. We found that expressed negative affect, not expressed positive affect, related to within-person variation in mentalizing network connectivity across perspectives. Further, participants higher in social anxiety symptoms showed more consistent mentalizing network connectivity across perspectives when expressed negative affect was higher than average, and particularly when friend negative affect expressions, rather than their own negative affect expressions, were higher. Mentalizing network connectivity may be responsive to shifts in negative affect during social interactions and individuals with higher social anxiety may find negative affect from a friend particularly salient.