Biological Psychiatry: Cognitive Neuroscience and Neuroimaging

Background. Psychosis spectrum symptoms (PSS) are prevalent in youth and are associated with increased risk for psychotic disorder, suicidality, and functional impairment. Computationally, PSS may stem from altered predictive coding of basic sensory surprises and environmental volatility. Formalized as hierarchical precision-weighted prediction errors (pwPEs), this altered processing is a proposed mechanistic substrate of aberrant perceptual inference across disorders, including psychosis-risk populations. While the auditory mismatch negativity (MMN) provides an electrophysiological index of pwPEs, it remains unknown if distinct hierarchical pwPE components distinguish youth who endorse PSS. Methods. A sample of 131 participants (PSS-=66, PSS+=65; ages 11-24) from the ongoing Toronto Adolescent and Youth (TAY-CAMH) Cohort study were stratified by PSS status using the PRIME Screen-Revised and were assessed for their psychosocial functioning. 64-channel EEG was recorded during an auditory oddball paradigm with stable and volatile phases. A hierarchical Bayesian model applied to the stimulus stream generated trajectories of low-level sensory and high-level volatility-related pwPEs. Alongside standard phase-averaged event-related potentials (ERPs), Bayesian trajectories derived model-based ERPs. Results. Replicating prior findings in non-clinical controls, stable-phase MMN significantly exceeds volatile-phase MMN and lower psychosocial functioning was associated with reduced volatile-phase MMN amplitude. Age significantly modulated oddball MMN and unweighted prediction errors ({delta}1, {delta}2). Group differences between PSS+ and PSS- were statistically significant for volatility-level pwPE ({epsilon}3), peaking at ~180 ms Peri-Stimulus Time (pFWE-peak =.024). Conclusions. Independent of age-related developmental effects, volatility-level pwPE learning ({epsilon}3) constitutes a more sensitive EEG marker associated with PSS status in help-seeking youth than low-level sensory pwPE.

ObjectivesOpioid use disorder (OUD) impacts millions of people worldwide. The prevalence and debilitating effects of OUD present a pressing need to understand its neural mechanisms to provide more targeted interventions. Prior studies have linked altered functioning in large-scale brain networks with clinical symptoms and outcomes in OUD. However, these investigations often do not consider how brain responses change over time. Time-varying brain network engagement can convey clinically relevant information not captured by static brain measures. MethodsWe investigated brain dynamic alterations in individuals with OUD by applying a new multivariate computational framework to movie-watching (i.e., naturalistic; N=76) and task-based (N=70) fMRI. We further probed the associations between cognitive control and brain dynamics during a separate drug cue paradigm in individuals with OUD. ResultsCompared to healthy controls (N=97), individuals with OUD showed decreased variability in the engagement of recurring brain states during movie-watching. We also found that worse cognitive control was linked to decreased variability during the rest period when no opioid-related stimuli were present. ConclusionsThese findings suggest that individuals with OUD may experience greater difficulty in effectively engaging brain networks in response to evolving internal or external demands. Such inflexibility may contribute to aberrant response inhibition and biased attention toward opioid-related stimuli, two hallmark characteristics of OUD. By incorporating temporal information, the current study introduces novel information about how brain dynamics are altered in individuals with OUD and their behavioral implications.

Background: While counterfactual thinking ('what could have been') guides adaptive decision-making, it remains unclear how this process is altered by the negative biases and motivational deficits characteristic of Major Depressive Disorder (MDD). Methods: We used a sequential economic decision-making task designed to emulate a volatile stock market to assess choice behavior in adults with or without MDD (Total N=178); a subset of these participants completed the task during functional MRI (N=53). The task allowed participants to make either positive ('invest') or negative ('short') bets, under either positive or negative contextual valence, defined by whether the immediately preceding stock price change was positive or negative. Fictive errors were defined as the difference between realized and best-possible outcomes. Results: Across the full cohort, group differences in behavioral adjustments to fictive error signals emerged exclusively under negative contextual valence, when stock prices decreased. Compared with controls, participants with MDD showed heightened sensitivity to invest-and-loss fictive errors, reflected in a greater reduction in subsequent bets (interaction beta = -0.63, p < .001), but blunted adjustment to short-and-gain fictive errors (beta = -0.86, p < .001). In the imaging cohort, blunted short-and-gain adjustment was accompanied by heightened anterior cingulate (ACC) activity and attenuated ventromedial prefrontal (vmPFC)-to-ACC coupling in MDD. vmPFC activity following negative market returns also tracked depression symptom severity. Conclusions: Depression selectively disrupts the use of counterfactual outcomes to guide adaptive choice under negative contextual valence, implicating altered frontocingulate function in maladaptive decision-making.

Neuroimaging methods rely on models of neurovascular coupling that assume hemodynamic responses evolve seconds after changes in neural activity. However, emerging evidence reveals noncanonical BOLD (blood oxygen level dependent) responses that are delayed under stress and aberrant in neuropsychiatric conditions. To investigate BOLD coupling to resting-state fluctuations in neural activity, we simultaneously recorded EEG and fMRI in people with schizophrenia and psychiatrically unaffected participants. We focus on alpha band power to examine voxelwise, time-lagged BOLD correlations. Principally, we find diversity in the temporal profile of alpha-BOLD coupling within regions of the default mode network (DMN). This includes early coupling (0-2 seconds BOLD lag) for more posterior regions, thalamus and brainstem. Anterior regions of the DMN show coupling at canonical lags (4-6 seconds), with greater lag scores associated with self-reported measures of stress and greater lag scores in participants with schizophrenia. Overall, noncanonical alpha-BOLD coupling is widespread across the DMN and other non-cortical regions, and is delayed in people with schizophrenia. These findings are consistent with a "hemo-neural" hypothesis, that blood flow and/or metabolism can regulate ongoing neural activity, and further, that the hemo-neural lag may be associated with subjective arousal or stress. Our work highlights the need for more studies of neurovascular coupling in psychiatric conditions.

Deficits in motivation and pleasure are common across many psychiatric disorders, and manifest as symptoms of amotivation and anhedonia, which are prominent features of both mood and psychotic disorders. Here we provide evidence for a shared transdiagnostic mechanism underlying impairments in motivation and pleasure across major depression, bipolar disorder, and schizophrenia. We found that value signals in the ventromedial prefrontal cortex (vmPFC) during decision-making were dampened in individuals with greater motivational and hedonic deficits, regardless of the primary diagnosis. This relationship remained significant while controlling for diagnosis-specific symptoms of mood and psychosis, such as depression as well as positive and negative symptoms. Our results demonstrate that dysfunction in the vmPFC during value-based decision-making is specifically linked to motivational and hedonic impairments across various psychiatric conditions. These findings provide a quantitative neural target for the potential development of novel treatments for amotivation and anhedonia.

Major Depressive Disorder (MDD) is a leading cause of disability worldwide. Core cognitive deficits, such as impaired selective attention, may contribute to its pathogenesis, yet their neurophysiological basis remains unclear. We here employed an additional-singleton paradigm and examined lateralized ERPs and EEG-based effective connectivity in individuals with MDD (n = 33), as compared to healthy controls (HCs; n = 31) to examine impaired statistical learning (SL) of distractor locations, a mechanism supporting proactive attention suppression. While HCs demonstrated typical SL effects including faster reaction times and reduced suppression-related Pd amplitudes at high-probability distractor locations, MDD patients failed to benefit behaviorally from SL and instead exhibited enhanced N2pc amplitudes reflecting attention-related selection to high-probability distractor locations, and increased fronto-parietal-occipital connectivity in alpha and theta bands, suggesting maladaptive attentional engagement and compensatory network hyperactivation. Crucially, more negative N2pc amplitudes and earlier latencies were significantly associated with slower cognitive flexibility and greater suicidal symptomology in MDD, independent of medication status. These findings identify a behavioral and neurophysiological candidate mechanism of impaired SL-guided proactive selective attention in MDD. Dysfunctions in SL-guided attentional suppression may represent a neurophysiological phenotype of cognitive vulnerability in depression with potential utility for biomarker and treatment development.

Impulsivity is a core dimension of ADHD and a transdiagnostic vulnerability factor for a wide range of adverse psychiatric and somatic outcomes, that could be mitigated through more effective screening of at-risk individuals. However, laboratory-based measures of impulsivity show weak convergence across paradigms and limited prediction of real-world behavior, constraining their utility. We tested whether combining repeated ecological assessment with computational modeling of response-time (RT) dynamics improves measurement of impulsivity and its cross-paradigm validity. Sixty participants, including adolescents with ADHD, individuals with 22q11.2 deletion syndrome, and healthy controls, completed a total of 1347 smartphone-based Balloon-Analogue-Risk-Task (D-BART) assessments repeatedly in daily life, alongside a single-session Conners CPT-3. RT was modeled using linear mixed-effects models as a function of objective risk and subjective uncertainty, with random effects capturing between- and within-person variability. Dynamic RT parameters were integrated with conventional performance metrics and related to CPT-3 variables using partial least squares analysis. External validity was evaluated against parent-rated behavioral symptoms. RT increased with both risk and uncertainty, consistent with adaptive modulation of speed-accuracy trade-offs. These effects varied substantially across individuals and repeated assessments. Dynamic RT parameters differentiated clinical from control participants, whereas traditional aggregate metrics did not. A PLS latent component linked D-BART and CPT-3 patterns and was associated with real-world hyperactivity/impulsivity, whereas CPT-3-derived scores alone were not. Experimental manipulation of ecological sampling density directly impacted D-BART predictive accuracy. These findings show that ecological repetition combined with parsimonious RT-dynamics modeling enhances construct validity, cross-paradigm convergence, and behavioral relevance of impulsivity measures, providing a scalable framework for capturing dynamic cognitive-control processes.

BackgroundThis study aims to understand learning and decision-making in suicidal behavior by investigating temporal dynamics of reward value encoding in individuals with late-life depression and a history of suicide attempts. MethodsIn a retrospective case-control study, 134 older adults (33 with depression and history of suicide attempts, 29 with depression and suicidal ideation but no past attempts, 32 with depression and no suicidal ideation/past attempts, and 40 psychiatrically healthy controls) completed an explore-exploit decision-making task during functional MRI (fMRI). Multilevel models of deconvolved fMRI time series, time-locked to trial events, interrogated whether temporal patterns of reward value encoding was associated with suicidal behavior. ResultsSuicidal behavior in general was associated with blunted ventral PFC (vPFC) value signals, but profiles varied as a function of attempt lethality. Specifically, low-lethality suicide attempts and excessive behavioral shifts were associated with abolished phasic responses to value updates in the default network vPFC and its connected regions, including the striatum, amygdala, and hippocampus. Additionally, an unexpected pattern of sustained negative value responses was observed in vPFC and striatal control subregions, and hippocampus of high-lethality suicide attempters. ConclusionsDiverging patterns of decision-related responses may reflect different paths toward suicidal behavior. Impaired value updating in individuals with low-lethality suicide attempts suggests a failure to integrate recent outcomes alongside prior experience, potentially relating to over-reactivity to stressors and a lower threshold for suicide attempts. In contrast, increased control network responses to difficult choices in high-lethality attempters may underlie cognitive constriction and consideration of a narrow set of potential solutions.

BackgroundBorderline personality disorder (BPD) is associated with altered activity in the prefrontal cortex (PFC) and amygdala, yet no studies have examined fronto-limbic circuitry in borderline adolescents. Here, we examined the contribution of fronto-limbic connectivity to the longitudinal stability of emotion-related impulsivity (ERI), a key feature of BPD, in symptomatic adolescents and young adults. MethodsWe compared resting-state effective connectivity (EC) in 82 adolescents and emerging adults with and without clinically significant borderline symptoms (n BPD = 40, ages 13-30). Group-specific directed networks were estimated amongst fronto-limbic nodes including PFC, ventral striatum (VS), central amygdala (CeN), and basolateral amygdala (BLA). We calculated directed centrality metrics and examined if these values were associated with initial levels and rates of change in ERI symptoms over a one-year follow-up using latent growth curve models (LGCMs). ResultsIn the healthy group, ventromedial prefrontal cortex (vmPFC) and dorsal ACC had a directed influence on CeN and VS respectively. In the borderline group bilateral BLA had a directed influence on CeN, whereas in the healthy group CeN influenced BLA. LGCMs revealed that in borderline adolescents, ERI remained stable across follow-ups. Further, higher output of R CeN in controls was associated with stronger within-person decreases in ERI. ConclusionsFunctional inputs from BLA and vmPFC appear to play competing roles in influencing CeN activity. In borderline adolescents BLA may predominate over CeN activity, while in controls the ability of CeN to conversely influence BLA activity is associated with more rapid reductions in ERI.

Depressed individuals show an attentional bias toward negatively valenced stimuli and thoughts. Here we present a novel closed-loop neurofeedback procedure that seeks to remediate this bias. Internal attentional states were detected by applying machine learning techniques to fMRI data in real-time, and externalized using a visually presented stimulus that the participant could learn to control. We trained 15 depressed and 12 healthy control participants over three fMRI sessions, preceded and followed by behavioral and clinical assessments. Initially, depressed participants were more likely than non-depressed participants to get "stuck" in negative attentional states, but this diminished with neurofeedback training relative to controls. Depression severity also decreased from pre- to post-training. These results demonstrate that our method is sensitive to the negative attentional bias in depressed individuals, and its reduction after training showcases the potential of this method as a treatment in the future.

The association of unipolar depression (UD), relative to healthy controls (HC), with cortical myelin is underexplored, despite growing evidence of associations with white matter tract integrity. We characterized cortical myelin in the 360 Glasser atlas regions using the T1w/T2w ratio in 39 UD and 47 HC participants (ages=19-44, 75% female). A logistic elastic net regularized regression with nested cross-validation and a subsequent linear discriminant analysis conducted on held-out samples were used to select brain regions and classify UD vs. HC. True-label model performance was compared against permuted-label model performance. Cortical myelin distinguished UD from HC with 68% accuracy (p<0.001; sensitivity=63.8%, specificity=71.5%). Brain regions contributing to this classification performance were located in the orbitofrontal cortex, anterior cingulate, extended visual, and auditory cortices, and showed statistically significant decreases and increases in myelin levels in UD vs. HC. The patterns of cortical myelin in these regions may be a biomarker of UD.

ObjectiveTo assess whether cognitive prediction models trained on multimodal neuroimaging from a population-based cohort generalize to children with and without ADHD across internal and external datasets. MethodsThis cross-sectional study used task-based and resting-state fMRI, structural MRI, and diffusion tensor imaging from the Adolescent Brain Cognitive Development (ABCD) Study (n = 11,747; mean age = 9.5 years) to train models predicting cognitive functioning. ADHD diagnoses were stratified into four tiers (n = 1,034 to 61), with the remaining participants classified as non-ADHD (n = 10,713). Models were trained using either single neuroimaging feature sets (e.g., task-based fMRI contrasts or cortical thickness) or combined feature sets via stacking. External generalizability was tested using an independent ADHD dataset (Lytle et al.; ADHD, n = 35; non-ADHD, n = 44; mean age = 9.0 years). ResultsIn ABCD, the stacked model integrating 81 neuroimaging types achieved comparable predictive performance for non-ADHD (r =.57) and ADHD (r =.51-.56 across tiers) groups. Key contributors included (a) fMRI contrasts from the NBack task, particularly in the ventral occipital cortex and anterior cingulate cortex (ACC), and (b) task-based functional connectivity from the dorsal attention and posterior multimodal networks. In external validation, combining different fMRI contrasts from the NBack task yielded similar performance for non-ADHD (r =.36) and ADHD (r =.42) participants. ConclusionsPopulation-trained neuroimaging models generalized well to both ADHD and non-ADHD children, underscoring the translational potential of multimodal brain-based models for predicting cognitive functioning in clinical populations.

BackgroundImpairments in language comprehension and production are core symptoms of schizophrenia spectrum disorders (SSD). Predictive processing, a Bayesian framework of brain function, provides an explanatory account of both language processing and psychotic symptom formation. Within SSD, neurobiological correlates of alterations in predictive language processing are poorly understood. Here, we investigate if glutamatergic neurotransmission and functional connectivity (FC) within the language network are associated with (1) altered predictive language processing, (2) task-based hallucinations (TBH), and (3) schizotypy. MethodsIn 53 healthy individuals (26 women), we combined a predictive language task analyzed using a Bayesian belief update model to estimate the individuals reliance on prior knowledge versus sensory input depending on the their schizotypal trait expression, resting-state functional MRI, and MR spectroscopy to measure glutamate levels in the anterior cingulate cortex (ACC) and left dorsolateral prefrontal cortex (DLPFC). Seed-based FC analysis focused on the language network including the left inferior frontal gyrus (IFG), posterior superior temporal gyrus (pSTG), and posterior middle temporal gyrus (pMTG). ResultsIncreased FC between left IFG and frontal areas predicted schizotypy-informed prior overweighting. Enhanced left pSTG-medial frontal FC was associated with the frequency of TBH. Finally, schizotypy was predicted by interactions between ACC/DLPFC Glx and left pSTG-frontal FC. ConclusionsThese results provide novel insights into neurobiological and computational mechanisms underlying altered language-related predictive processing and subclinical symptom formation. They highlight potential early indicators of schizotypal symptom development and underscore the value of multimodal approaches in elucidating pathophysiology of SSD.

The ability to search for and detect social cues, such as facial expressions of emotion, is critical to the understanding of complex dynamic social situations. This ability involves the coordinated actions of multiple cognitive domains, including face-emotion processing, mentalization, and visual attention. Individuals with schizophrenia are generally impaired in social cognition, and have been shown to have deficits in all of these domains. However, the whether the neural substrates of these impairments are shared or separate remains unclear. One candidate region for a shared substrate is the right temporoparietal junction/posterior superior temporal sulcus (TPJ-pSTS), which contains areas belonging to all of the cortical networks underlying these domains. Here we use functional MRI to examine differences in cortical activity evoked by a naturalistic movie, and link these results to impaired visual scanning and social cognition. 27 schizophrenia participants and 21 healthy controls watched a 15-minute clip of the movie \"The Good, the Bad, and the Ugly\" while high resolution multiband BOLD-fMRI activity was recorded. Inter-subject correlation was used to measure the evoked activity. BOLD-fMRI activity was also correlated with motion content in the movie, with the average activity in other cortical areas, and with frequency of saccades made during the movie. Visual scanning performance was measured in a separate behavioral experiment, and social cognition measured by The Awareness of Social Inference Test (TASIT). Contrasting the groups revealed that the TPJ-pSTS has the largest engagement deficit in both cortical hemispheres in schizophrenia patients versus healthy controls. Follow-up analyses find that brain activity in this region is less correlated with the motion content of the movie, that this region is abnormally synchronized to the other cortical areas involved in the cognitive domains underlying visual scanning of social scenes, and that activity this region is less correlated with the saccades made during the movie. Lastly, schizophrenia participant visual scanning performance of this clip was impaired compared to healthy controls, and correlated across the two groups with social cognition. These results indicate that the TPJ-pSTS plays less of an integral role in the coordination of face-emotion processing, mentalization, and visual attention in schizophrenia participants versus healthy controls. This functional deficit then impacts the visual scanning of a complex dynamic visual scenes, which in turn affects the comprehension of that scene. These findings indicating that the TPJ-pSTS is potentially the shared substrate for all of these deficits will lead to new treatments targeting this region to improve social cognition in individuals with schizophrenia.

ObjectiveSubthreshold depression (StD) in adolescence is a potent risk factor for major depressive disorder, yet its neurobiological mechanisms remain unclear. The glymphatic system, a brain-wide waste clearance pathway, has been implicated in psychiatric disorders but has not been investigated integratively in StD. This study aimed to characterize glymphatic function in adolescents with StD using a multimodal neuroimaging approach. MethodsA total of 107 adolescents (71 with StD and 36 healthy controls) underwent multimodal MRI. Glymphatic function was assessed using three complementary metrics: the diffusion tensor imaging analysis along the perivascular space (DTI-ALPS) index for clearance capacity, BOLD-CSF coupling for neurovascular-driven circulation, and Hurst exponent for CSF flow dynamics. An integrated Mahalanobis Distance (MD) was calculated to quantify individual deviation from the healthy glymphatic profile. ResultsCompared with healthy controls, adolescents with StD exhibited significant reductions in the left DTI-ALPS index and global BOLD-CSF coupling (p < 0.05). No significant group differences were found in structural volumes of the choroid plexus or perivascular spaces. The integrated MD metric significantly differentiated StD from controls and demonstrated stronger associations with clinical severity than single biomarkers. Specifically, glymphatic dysfunction was significantly correlated with greater depressive symptom severity, sleep disturbances, and general psychopathology. ConclusionsAdolescents with StD exhibit distinct, lateralized impairments in glymphatic clearance and neurovascular driving mechanisms. These findings provide early physiological evidence for a dimensional model of depression, suggesting that glymphatic dysfunction is present before diagnostic criteria for major depression are met. The multivariate MD metric offers a sensitive biomarker for tracking early pathophysiological changes and potential intervention targets.

Depression is a prevalent psychiatric condition and a common comorbidity across neurological disorders. Common symptoms include anhedonia, negative emotional biases, and cognitive dysfunction. Beta (15-30 Hz) neural oscillations have been shown to increase during reward-based learning within fronto-striatal reward networks. Corticostriatal beta oscillations have also been implicated in cognitive functions including working memory. However, the relationship between beta oscillations and depression remains unknown. Using intracranial recordings, we aimed to investigate how depression modulates the spectral power of neural oscillations in corticostriatal structures during reward feedback in a working memory task. Thirty movement disorder patients undergoing awake deep brain stimulation surgery with electrode trajectories traversing the caudate or dorsolateral prefrontal cortex (DLPFC) participated in this study. We recorded local field potential data intraoperatively as subjects completed a 2-back verbal working memory task where they identified whether a word matched the word presented two trials prior. Subjects received reward in the form of visual feedback for correct answers. Word stimuli had either a positive, negative, or neutral emotional valence. Subjects completed the Beck Depression Inventory-II preoperatively, and we used a cut-off score of 14 to identify patients with depression. We found that caudate and DLPFC power increased in the alpha (8-15 Hz) and beta range during reward feedback and that this increase was significantly greater for subjects without depression compared to depressed subjects. In non-depressed patients, positive feedback stimuli evoked significantly higher beta power in the caudate during reward compared to neutral and negative stimuli. In depressed patients, emotional valence did not affect reward-related caudate spectral power, while DLPFC alpha power was significantly higher following positive emotional stimuli in comparison to neutral but not negative stimuli. We additionally found that anti-depressant medications (ADMs) generally blunted alpha and beta reward signaling processes in the DLPFC. This blunting effect on reward-related alpha power in the DLPFC, however, was reversed in depressed patients, indicating that the effects of ADMs on reward signaling processes may depend on whether a patient is exhibiting depression symptoms. Our findings suggest that depression suppresses the alpha and beta power response to both reward and emotional stimuli during working memory, indicating power attenuation in these frequency bands may contribute to emotional and cognitive depression symptoms.

Large-scale brain network function is critical for healthy cognition, yet links between such network function, neurochemistry, and smaller-scale neurocircuitry are unclear. Here, we evaluated 59 healthy individuals using resting-state fMRI to determine how network-level temporal dynamics were impacted by two well-characterized pharmacotherapies targeting catecholamines: methylphenidate (20mg) and haloperidol (2mg). Network dynamic changes were tested for links with drug-induced alterations in complex corticostriatal connections as this circuit is a primary site of action for both drugs. A randomized, double-blind, placebo-controlled design was used. Methylphenidate enhanced time spent in the default mode network (DMN p<0 .001) and dorsal attention network (DAN p<0.001) and reduced time in the frontoparietal network (p<0.01). Haloperidol increased time in a sensory motor-DMN state (p<0.01). The magnitude of change in network dynamics induced by methylphenidate vs. placebo was correlated with the magnitude of methylphenidate-induced rearrangement of complex corticostriatal connectivity (R=0.32, p=0.014). Haloperidol did not alter complex corticostriatal connectivity. Methylphenidate increased time in networks involved in internal (DMN) and external attention (DAN), aligning with methylphenidates established role in attention. Methylphenidate also significantly changed complex corticostriatal connectivity by altering the relative strength between multiple corticostriatal connections, indicating that methylphenidate may shift which corticostriatal connections are prioritized relative to others. Findings further show that these local circuit changes are linked with large scale network function. Collectively, these findings provide a deeper understanding of large-scale network function, set a stage for mechanistic understanding of network engagement, and provide needed information to potentially guide medication use based on network-level effects.

BackgroundThe N400 semantic-priming event-related potential (ERP) is attenuated in schizophrenia and in youth at clinical high risk for psychosis (CHRP); however, the circuit mechanisms linking this abnormality to functional outcome remain unclear. MethodsWe recorded 32-channel EEG while 46 CHRP outpatients and 38 demographically matched healthy controls (HC) performed a word-pair priming task (80 related, 80 unrelated pairs; prime-to-target stimulus-onset asynchronies [SOA]=300ms or 750ms). Twenty-six CHRP participants were reassessed after one year. N400 difference waves (unrelated-related) were fit with a connectome-constrained Jansen-Rit neural-mass model in 200 cortical parcels. Local gains, synaptic time constants and effective connectivity parameters were optimised with the WhoBPyt framework, and principal-component trajectories of the inferred excitatory-inhibitory |E - I| balance were analysed with partial least squares. ResultsUnder the long-SOA (750ms)/unrelated condition, CHRP showed a sharply elevated early |E -I| peak at 70-100ms relative to HC (p=0.0004), driven by greater pyramidal excitatory gain (parameter A), stronger excitatory-to-pyramidal coupling and faster inhibitory decay, indicating cortical disinhibition. The amplitude of this early peak predicted poorer social functioning at one-year follow-up (r=-0.56,p=0.003). Conversely, in the short-SOA (300ms)/related condition CHRP exhibited a larger N400-window |E-I| peak (350ms) associated with enhanced inhibitory-to-pyramidal feedback (parameter C3) and lengthened inhibitory decay (parameter b); this putative compensatory inhibition correlated with better functional outcome (r=0.60,p=0.001). Network-level analyses revealed an amplified early sensory-network burst and attenuated default-mode and salience-network responses, consistent with a systems-wide shift toward disinhibition in CHRP. ConclusionsComputational modeling demonstrates that N400 abnormalities in CHRP arise from temporally specific E-I imbalances: early cortical disinhibition that forecasts functional decline and a later inhibitory reinforcement that may confer resilience. These time-resolved E-I metrics constitute low-burden, mechanistically interpretable biomarkers for stratifying psychosis risk and guiding early intervention.

BackgroundSymptoms of borderline personality disorder (BPD) often manifest in adolescence, yet the underlying relationship between these debilitating symptoms and the development of functional brain networks is not well understood. Here we aimed to investigate how multivariate patterns of functional connectivity are associated with symptoms of BPD in a large sample of young adults and adolescents. MethodsWe used high-quality functional Magnetic Resonance Imaging (fMRI) data from young adults from the Human Connectome Project: Young Adults (HCP-YA; N = 870, ages 22-37 years, 457 female) and youth from the Human Connectome Project: Development (HCP-D; N = 223, age range 16-21 years, 121 female). A previously validated BPD proxy score was derived from the NEO Five Factor Inventory (NEO-FFI). A ridge regression model with 10-fold cross-validation and nested hyperparameter tuning was trained and tested in HCP-YA to predict BPD scores in unseen data from regional functional connectivity, while controlling for in-scanner motion, age, and sex. The trained model was further tested on data from HCP-D without further tuning. Finally, we tested how the connectivity patterns associated with BPD aligned with agerelated changes in connectivity. ResultsMultivariate functional connectivity patterns significantly predicted out-of-sample BPD proxy scores in unseen data in both young adults (HCP-YA; pperm = 0.001) and older adolescents (HCP-D; pperm = 0.001). Predictive capacity of regions was heterogeneous; the most predictive regions were found in functional systems relevant for emotion regulation and executive function, including the ventral attention network. Finally, regional functional connectivity patterns that predicted BPD proxy scores aligned with those associated with development in youth. ConclusionIndividual differences in functional connectivity in developmentally-sensitive regions are associated with the symptoms of BPD.

BackgroundEvidence suggests a non-specific mapping between psychiatric disorders and underlying neurobiological substrates. A dimensional psychopathology framework may prove useful for organizing observed neurobiological alterations along broad psychopathological dimensions. MethodsWe applied latent class analysis, with an additional constraint on classification uncertainty, to identify clinical cohorts of symptomatic homogeneity to represent the high-risk end of specific psychopathological dimensions (i.e., internalizing/externalizing, p-factor), using baseline data (N = 11860) from the Adolescent Brain and Cognitive Development (ABCD) Study. These cohorts were compared against neurotypical individuals in deviations from the normality of cortical development, quantified using autoencoder-based normative models, to reveal cortical abnormalities. ResultsWe identified cortical thickness related to psychopathologies in the ABCD data, particularly to externalizing syndromes, and revealed distinct structural abnormalities to broad psychopathological dimensions. ConclusionThis study highlights the value of person-centered analytic techniques, combined with normative modeling, to complement traditional associational methodologies in revealing neurobiological correlates of dimensional psychopathologies.