Nature Mental Health

Conduct disorder (CD) is among the most prevalent and burdensome disorders in early adolescence. Over the past decade, there has been growing interest in identifying reliable and localized neurobiological markers of conduct disorder (CD). However, recent meta-analyses have highlighted the weak reliability of these so-called markers, thereby limiting the ability to draw firm conclusions. Using normative network mapping (598 healthy subjects), we rather sought to investigate whether the heterogeneous findings across studies may map unto a common brain network. A meta-analysis of 38 fMRI studies involving adolescents with a CD (932 cases, 975 controls) was first conducted and showed only a very weak spatial convergence in brain activity alterations in the anterior temporal lobe (5 out of 38 studies). In turn, network mapping revealed that findings across studies show a consistent connectivity pattern across the whole brain, with regional overlap reaching up to 94.7% (36 out of 38 studies). This network was primarily driven by functional connectivity of brainstem nuclei, subcortical structures (i.e., thalamus, ventral striatum), cingulate cortex (i.e., anterior to posterior midcingulate), superior temporal sulcus, and visual cortices. We further describe the neurochemicals and genetic markers of this CD-Network with emphasis on midbrain serotoninergic, dopaminergic and cholinergic projections. Our findings suggest that our understanding of the neurobiological markers of CD could be enhanced by viewing the brain as a complex interconnected system rather than reducing its complexity to a limited number of brain structures. More importantly, this CD-Network may serve as evidence that the various theories of CD can be reconciled rather than seen as conflicting.

The increased vulnerability to mental health problems in adolescence is frequently reported but poorly understood, hampered by a rigid diagnostic system which fails to capture intertwining symptoms and only loosely aligns with biological axes of variability. Here, we reconceptualised the mental health symptoms of young adolescents in the ABCD cohort (N=11862) as a latent topology of overlapping symptom dimensions, using an unsupervised machine learning algorithm to establish how transdiagnostic dimensions co-occur and overlap within individuals. Combining this with a novel classification approach, we delineated zones within this landscape, within which specific profiles of symptoms were robustly represented. These data-driven profiles were leveraged to establish associated resting-state functional connectivity and genetic characteristics. In doing so we recaptured the commonly reported p-factor axis as well as further symptom-subtype dimensions. Gene ontology analysis revealed that shared neurobiological and cellular mechanisms embedded in both the genome and transcriptome may confer risk for psychopathology.

World Trade Center (WTC) responders exposed to traumatic and environmental stressors during rescue and recovery efforts have higher prevalence (23%) of persistent, clinically significant WTC-related post-traumatic stress disorder (WTC-PTSD). Here, we applied eigenvector centrality (EC) metrics and data driven methods on resting state functional magnetic resonance (fMRI) outcomes to investigate neural mechanisms underlying WTC-PTSD and to identify how EC shifts in brain areas relate to WTC-exposure and behavioral symptoms. Nine brain areas differed significantly and contributed the most to differentiate functional neuro-profiles between WTC-PTSD and non-PTSD responders. The association between WTC-exposure and EC values differed significantly between WTC-PTSD and non-PTSD in the right anterior parahippocampal gyrus and left amygdala (p= 0.010; p= 0.005, respectively, adjusted for multiple comparisons). Within WTC-PTSD, the index of PTSD symptoms was positively associated with EC values in the right anterior parahippocampal gyrus and brainstem. Our understanding of functional changes in neural mechanisms underlying WTC-related PTSD is key to advance intervention and treatment.

An important aim in psychiatry is the establishment of valid and reliable associations linking profiles of brain functioning to clinically relevant symptoms and behaviors across patient populations. To advance progress in this area, we introduce an open dataset containing behavioral and neuroimaging data from 241 individuals aged 18 to 70, comprising 148 individuals meeting diagnostic criteria for a broad range of psychiatric illnesses and a healthy comparison group of 93 individuals. These data include high-resolution anatomical scans, multiple resting-state, and task-based functional MRI runs. Additionally, participants completed over 50 psychological and cognitive assessments. Here, we detail available behavioral data as well as raw and processed MRI derivatives. Associations between data processing and quality metrics, such as head motion, are reported. Processed data exhibit classic task activation effects and canonical functional network organization. Overall, we provide a comprehensive and analysis-ready transdiagnostic dataset, which we hope will accelerate the identification of illness-relevant features of brain functioning, enabling future discoveries in basic and clinical neuroscience.

Improving diagnostic accuracy of obsessive-compulsive disorder (OCD) using models of brain imaging data is a key goal of the field, but this objective is challenging due to the limited size and phenotypic depth of clinical datasets. Leveraging the phenotypic diversity in large non-clinical datasets such as the UK Biobank (UKBB), offers a potential solution to this problem. Nevertheless, it remains unclear whether classification models trained on non-clinical populations will generalise to individuals with clinical OCD. This question is also relevant for the conceptualisation of OCD; specifically, whether the symptomology of OCD exists on a continuum from normal to pathological. Here, we examined a recently published "meta-matching" model trained on functional connectivity data from five large normative datasets (N=45,507) to predict cognitive, health and demographic variables. Specifically, we tested whether this model could classify OCD status in three independent clinical datasets (N=345). We found that the model could identify out-of-sample OCD individuals. Notably, the most predictive functional connectivity features mapped onto known cortico-striatal abnormalities in OCD and correlated with genetic brain expression maps previously implicated in the disorder. Further, the meta-matching model relied upon estimates of cognitive functions, such as cognitive flexibility and inhibition, to successfully predict OCD. These findings suggest that variability in non-clinical brain and behavioural features can discriminate clinical OCD status. These results support a dimensional and transdiagnostic conceptualisation of the brain and behavioural basis of OCD, with implications for research approaches and treatment targets.

Multivariate machine learning techniques are a promising set of tools for identifying complex brain-behavior associations. However, failure to replicate results from these methods across samples has hampered their clinical relevance. This study aimed to delineate dimensions of brain functional connectivity that are associated with child psychiatric symptoms in two large and independent cohorts: the Adolescent Brain Cognitive Development (ABCD) Study and the Generation R Study (total n=8,605). Using sparse canonical correlations analysis, we identified three brain-behavior dimensions in ABCD: attention problems, aggression and rule-breaking behaviors, and withdrawn behaviors. Importantly, out-of-sample generalizability of these dimensions was consistently observed in ABCD, suggesting robust multivariate brain-behavior associations. Despite this, out-of-study generalizability in Generation R was limited. These results highlight that the degree of generalizability can vary depending on the external validation methods employed as well as the datasets used, emphasizing that biomarkers will remain elusive until models generalize better in true external settings.

Task-based functional magnetic resonance imaging (tb-fMRI) has advanced our understanding of brain-behavior relationships. Standard tb-fMRI analyses suffer from limited reliability and low effect sizes, and machine learning (ML) approaches often require thousands of subjects, restricting their ability to inform how brain function may arise from and contribute to individual differences. Using data from 9,024 early adolescents, we derived a classifier ( neural signature) distinguishing between high and low working memory loads in an emotional n-back fMRI task, which captures individual differences in the separability of activation to the two task conditions. Signature predictions were more reliable and had stronger associations with task performance, cognition, and psychopathology than standard estimates of regional brain activation. Further, the signature was more sensitive to psychopathology associations and required a smaller training sample (N=320) than standard ML approaches. Neural signatures hold tremendous promise for enhancing the informativeness of tb-fMRI individual differences research and revitalizing its use.

Current psychiatric diagnoses are not defined by neurobiological measures which hinders the development of therapies targeting mechanisms underlying mental illness 1,2. Research confined to diagnostic boundaries yields heterogeneous biological results, whereas transdiagnostic studies often investigate individual symptoms in isolation. There is currently no paradigm available to comprehensively investigate the relationship between different clinical symptoms, individual disorders, and the underlying neurobiological mechanisms. Here, we propose a framework that groups clinical symptoms derived from ICD-10/DSM-V according to shared brain mechanisms defined by brain structure, function, and connectivity. The reassembly of existing ICD-10/DSM-5 symptoms reveal six cross-diagnostic psychopathology scores related to mania symptoms, depressive symptoms, anxiety symptoms, stress symptoms, eating pathology, and fear symptoms. They were consistently associated with multimodal neuroimaging components in the training sample of young adults aged 23, the independent test sample aged 23, participants aged 14 and 19 years, and in psychiatric patients. The identification of symptom groups of mental illness robustly defined by precisely characterized brain mechanisms enables the development of a psychiatric nosology based upon quantifiable neurobiological measures. As the identified symptom groups align well with existing diagnostic categories, our framework is directly applicable to clinical research and patient care.

BackgroundStudies applying machine learning to obsessive-compulsive disorder (OCD) typically report accuracy in homogeneous samples but rarely assess model reliability, generalizability, and interpretability needed for clinical use. MethodsWe applied a transformer-based deep learning model, the Multi-Band Brain Net, to the ENIGMA-OCD cohort - the largest available resting-state functional magnetic resonance imaging (rs-fMRI) dataset in OCD with 1,706 participants (869 cases with OCD, 837 controls) across 23 sites worldwide. We evaluated model reliability by calculating calibration - the models ability to "know what it doesnt know". We assessed generalizability using leave-one-site-out validation to test performance on unseen sites with different scanners, acquisition protocols, and patient populations. Finally, we examined interpretability by analyzing model attention weights to identify the neural connectivity patterns that influence model predictions. ResultsThe model achieved modest but competitive classification performance (AUROC = .653 {+/-} .039). Crucially, while large-scale pretraining on the UK Biobank (N = 40,783) did not boost accuracy, it significantly enhanced model calibration by reducing overconfident predictions. Leave-one-site-out validation showed a generalization gap across sites (AUROC = .427-.819). Pretraining did not close this gap but removed scanner manufacturer bias. Finally, attention-based mapping identified biologically plausible patterns of widespread hypoconnectivity in OCD relative to healthy controls, particularly in low-frequency bands involving the default mode, salience, and somatomotor networks. These findings aligned with known OCD neurobiology. ConclusionsThis study provides a framework for developing more reliable and trustworthy clinical artificial intelligence for OCD.

Major depressive disorder has a point prevalence of 5% of the world population and is the main cause of disability, with up to a third of patients not responding to first-line treatments. Surgical neuromodulation is offered to only an anecdotal proportion of these patients, because while these methods are curative in some individuals, the proportion of responders rarely exceeds 50%. Recent efforts to establish reliable brain circuit-symptom relationships and thus predict response have involved mapping with multiple intracranial electrodes, but the impracticality of this approach currently prevents its application at scale. In the present study (ClinicalTrials.gov identifier NCT05697172; FDA Q220192) we begin to address this gap by leveraging low-intensity focused ultrasound (LIFU), a novel noninvasive technique, to modulate the anterior limb of the internal capsule, which is an established surgical deep white matter target for depression. We based our study on burgeoning in vitro evidence that LIFU attenuates axonal conduction by operating mechanosensitive channels in nodes of Ranvier. Compared with sham stimulus, active LIFU produced a functional disconnection of gray matter hubs reached by the sonicated axonal tracts, an increase in positive emotion, and top-down effects on the cardiovascular autonomic balance. Our results using LIFU of deep-brain white matter tracts in humans open three potential avenues to understand the mechanisms and improve the outcome of depression, namely attaining a personalized definition of brain circuit-symptom relationships, serving as a noninvasive probe for neuromodulation before irreversible procedures in a "try before you buy" approach, and ultimately emerging as a therapeutic intervention itself.

As the global burden of addiction intensifies, the neurobiological commonalities and distinctions between substance use disorders (SUDs) and behavioral addictions (BAs) remain poorly characterized. This coordinate-based meta-analysis of 59 fMRI articles (n = 2,951) mapped the neural signatures of visual cue-reactivity across the addictive disorders. Our results revealed a universal core network shared by SUDs and BAs centered in the bilateral opercular inferior frontal gyrus, suggesting a shared disruption in inhibitory control. Distinctively, SUDs exhibited a stronger recruitment of a subcortical salience pathway, with greater involvement of the left thalamus ventral anterior nucleus than BAs, potentially reflecting pharmacologically amplified bottom-up salience attribution. Notably, recovery-related patterns diverged in the left medial superior frontal gyrus. Alcohol use disorder was associated with neural restoration, whereas heroin use disorder showed neural decompensation. These neural signatures establish a rigorous neurobiological basis for differentiating substance and behavioral phenotypes, supporting tailored circuit-based precision treatments.

Interoception--the sensing and perception of the internal viscera--is widely cast as a transdiagnostic mechanism linking brain-body interaction to mental illness. Prevailing models propose that altered interoceptive ability reflects a core liability underlying diverse psychiatric symptoms. We tested this hypothesis in a large community sample (N = 547) using psychophysically optimised tasks spanning cardiac and respiratory domains, combined with hierarchical Bayesian modelling and comprehensive symptom profiling. Contrary to this central prediction, objective interoceptive sensitivity, precision, and metacognitive insight were largely unrelated to general symptom burden or specific mental health dimensions. In contrast, self-reported interoceptive sensibility showed moderate associations with symptoms, but semantic similarity analyses suggest these reflect conceptual rather than mechanistic overlap. These findings challenge the prevailing view that objective interoceptive sensitivity is a broad marker of psychopathology, prompting a reconsideration of how we measure interoception in mental health research.

Substance use disorder (SUD) stands as a critical public health concern, contributing to substantial morbidity, mortality and societal costs. The effects of SUD on structural brain changes have been well documented. However, the neural mechanisms underlying SUD and the spatial-temporal volumetric changes associated with SUD remained underexplored. In this investigation, neuroimaging, behavioral and genomic data across four large population cohorts jointly covering the full lifespan were harmonized, and whole-brain volumetric trajectories between substance use disorders (SUDs) and healthy controls (HCs) were compared, revealing the potential neurobiological mechanisms and the genomic basis underlying SUD. Results highlighted three distinct life stages critical for the development of SUD: 1) adolescence to early adulthood (before 25y), where SUD is suspected to be the consequence of prefrontal-subcortical imbalance during neurodevelopment; 2) early-to-mid adulthood (25y - 45y), where SUD was strongly associated with compulsivity-related brain volumetric changes; 3) mid-to-late adulthood (after 45y), where SUD-related brain structural changes could be explained by neurotoxicity. Results were externally validated both via longitudinal analysis of these population cohorts and in independent cross-sectional samples. In summary, our study demonstrated the lifespan whole-brain volumetric changes associated with SUD, revealed potential neurobehavioral mechanisms for the development of SUD, and suggested critical time window for effective prevention and treatment of SUD.

Resilience to emotional disorders is critical for adolescent mental health, especially following childhood abuse. Yet, brain signatures of resilience remain undetermined due to the differential susceptibility of the brains emotion processing system to environmental stresses. Analyzing brains responses to angry faces in a longitudinally large-scale adolescent cohort (IMAGEN), we identified two functional networks related to the orbitofrontal and occipital regions as candidate brain signatures of resilience. In girls, but not boys, higher activation in the orbitofrontal-related network was associated with fewer emotional symptoms following childhood abuse, but only when the polygenic burden for depression was high. This finding defined a genetic-dependent brain (GDB) signature of resilience. Notably, this GDB signature predicted subsequent emotional disorders in late adolescence, extending into early adulthood and generalizable to another independent prospective cohort (ABCD). Our findings underscore the genetic modulation of resilience-brain connections, laying the foundation for enhancing adolescent mental health through resilience promotion.

Neuroimaging has revealed that major depression is underpinned by dysfunctional brain networks, with symptom variability stemming from altered interactions within and between brain regions. While the effect of depression severity is well-studied, the effect of depression duration (chronicity) is relatively neglected, despite its clinical significance. This study examined how severity, chronicity, and their interaction affect brain network connectivity and grey matter volume. Forty-six patients (31 females, mean age 40.5) were assessed using whole-brain network modeling and voxel-based morphometry (VBM). Severity was measured via the Hamilton Depression Rating Scale, and chronicity was defined as an episode lasting over 24 months. The key finding was that chronicity moderated the impact of severity on functional connectivity between the Central Executive Network (CEN) and the precuneus (part of the Default Mode Network, DMN). Chronic versus non-chronic patients showed opposite patterns. Non-chronic patients showed stronger CEN-Default Mode Precuneus connectivity at low severity and weaker at high severity; chronic patients showed the reverse. This study reveals a novel impact of chronicity on CEN-DMN interactions, a neglected moderator of brain-symptom severity correlations in depression.

Harmonized phenotyping and diverse population-specific studies are crucial for advancing gene discovery in psychiatric genetics. We conducted a genome-wide association (GWAS) mega-analysis of DSM-defined lifetime major depressive disorder (MDD) in 64 941 participants (25.7% cases) from the Dutch BIObanks Netherlands Internet Collaboration (BIONIC) consortium. SNP-based heritability was estimated at 13.4%, exceeding recent global meta-analyses, with a high genetic correlation (rG = 0.89) to the latest major depression GWAS by the Psychiatric Genetics Consortium (PGC-MD). We identified a novel genome-wide significant locus in PALMD (p = 3.26 x 10-), that was confirmed by GWAS-by-subtraction. Polygenic scores (PGSs) based on BIONIC predicted MDD in UK Biobank, and PGSs from PGC-MD predicted into BIONIC, with within-family analyses indicating minimal confounding. Genetic causal inference revealed associations with over 30 phenotypes. Twin concordance for MDD increased with polygenic burden, reinforcing its genetic architecture. This study emphasizes the power of harmonized phenotyping and regional biobanks in uncovering the genetic architecture of MDD, highlighting the value of population-specific studies for improving risk prediction and advancing psychiatric genetics.

Puberty demarks a period of profound brain dynamics that orchestrates changes to a multitude of neuroimaging-derived phenotypes. This poses a dimensionality problem when attempting to chart an individuals brain development on a single scale. Here, we illustrate shifts in subject similarity of imaging data that relate to pubertal maturation and altered mental health, suggesting that dimensional reference spaces of subject similarity render useful to chart brain dynamics in youths.

Rumination, or perseverative negative self-referential thinking, is a hallmark of depression. In adults, a dynamic resting-state fMRI model of trait rumination was recently identified through predictive modelling. In adolescents, a development period during which rumination and depression increase, the neurobiological correlates of ruminative thinking are less clear. In the current preregistered study, we examine dynamic connectivity correlates of self-reported rumination in the largest sample of adolescents to date (n = 443, containing clinical and non-clinical individuals). Notably, the adult model failed to generalize to our sample. In addition, linear models trained on default-mode network (DMN) connectivity, as well as whole-brain connectome models, failed to generalize to held-out data. In an exploratory random forest analysis, we found significant prediction performance of a model where increased variability between DMN-cerebellum, DMN-dorsal attention network, and DMN-DMN connections was nominally associated with higher rumination. However, the model did not generalize to an external sample with lower rumination scores and a distinct scanner protocol. Our findings illustrate the difficulty of characterizing the neurodevelopment of risk factors for depression.

ObjectiveAssociations between psychiatric problems and white matter (WM) microstructure have been reported in childhood and adolescence. Yet, a deeper understanding of this relation has been hampered by a dearth of well-powered longitudinal studies and a lack of explicit examination of the bidirectional associations between brain and behavior. We investigated the temporal directionality of WM microstructure and psychiatric symptom associations from late childhood to early adolescence. MethodsIn this observational study, we leveraged the worlds largest single- and multi-site cohorts of neurodevelopment: the Generation R and Adolescent Brain Cognitive Development Studies (total n scans = 11,400). We assessed psychiatric symptoms with the Child Behavioral Checklist as broad-band Internalizing and Externalizing scales, and as syndrome scales (e.g., Anxious/Depressed). We quantified WM with Diffusion Tensor Imaging, globally and at a tract level. We used cross-lagged panel models to test bidirectional associations of global and specific measures of psychopathology and WM microstructure, meta-analyzed results across cohorts, and used linear mixed-effects models for validation. ResultsWe did not identify any robust longitudinal associations of global WM microstructure with internalizing or externalizing problems across cohorts (confirmatory analyses). We observed similar findings for longitudinal associations between tract-based microstructure with internalizing and externalizing symptoms, and for global WM microstructure with specific syndromes (exploratory analyses). ConclusionUni- or bi-directionality of longitudinal associations between WM and psychiatric symptoms was not robustly identified. We propose several explanations for these findings, including interindividual differences, the use of longitudinal approaches, and smaller effects than expected.

Reproducible patterns of atypical functional connectivity of sensorimotor and higher-order networks have been previously identified in the autistic brain. However, the neurosignalling pathways underpinning these differences remain unclear. The {micro}-opioid system is involved in sensory processing as well as social and reward behaviours and has been implicated in autism, suggesting a potential role in shaping the autistic brain. Hence, we tested the hypothesis that there is atypical involvement of the {micro}-opioid system in these networks in autism. We used a placebo-controlled, double-blind, randomised, crossover study design to compare the effects of an acute dose of the {micro}-opioid receptor agonist tianeptine in autistic and non-autistic participants on functional connectivity (FC) of sensorimotor and frontoparietal networks. We found that tianeptine increased FC of a sensorimotor network previously characterised by atypically low FC in autism. The connectivity of the frontoparietal network was not significantly shifted. Our findings suggest that {micro}-opioid neurosignalling might contribute to functional brain differences in the sensorimotor network in autism. Given that sensorimotor system alterations are thought to be core to autism and contribute to other core autistic features, as well as adaptability and mental health, further research is warranted to explore the translational potential of {micro}-opioid modulation in autism.