David E. J. Linden

Differences between clinical questionnaires limit the pooling of symptom data across studies, particularly when participants have completed different instruments. We developed a framework that uses questionnaire item content to harmonize neuropsychiatric measures without requiring overlapping response data. Experts mapped items to shared symptom dimensions, which informed score transformation and the training of models to map new items automatically. We evaluated the framework in a multisite Parkinson’s disease sample and an independent depression cohort. Transformation improved score alignment across questionnaires, while additional analyses supported clinical relevance and cross-questionnaire convergence. Our fine-tuned embedding models performing the semantic mapping task generalize to unseen questionnaires with a mean accuracy of 85.3%. This semantic, dimensional harmonization approach supports retrospective pooling of heterogeneous questionnaire data and offers a practical and systematic path toward mega-analyses on retrospective datasets.

People with severe mental illness (SMI), including schizophrenia (SCZ) and bipolar disorder (BD), show elevated risk for somatic conditions and premature mortality. Both genetic liability and modifiable lifestyle factors have been implicated in this comorbidity, yet large-scale studies examining their independent and combined effects across multiple somatic conditions are lacking. We therefore examined whether modifiable lifestyle factors (i.e. smoking, alcohol intake, diet, and exercise) and polygenic risk scores (PRS) for SCZ and BD are associated with the risk of developing somatic illnesses. Data were drawn from 315,313 unrelated individuals in the UK Biobank cohort, with a median follow up of 14 years. Cox proportional hazards models were used to estimate the independent, joint, and interaction effects of PRS-SCZ and PRS-BD and lifestyle factors on the incidence of nine chronic somatic illnesses. Healthy lifestyle was associated with lower risks of type 2 diabetes (T2D), cardiovascular disease (CVD), dementia, rheumatoid arthritis, asthma, chronic liver disease (CLD), and cancer (HRs ranging from 0.43 to 0.80). In addition, higher PRS for SCZ was associated with increased risks of asthma and dementia and lower risk of T2D, while higher PRS for BD was associated with increased risk of T2D. Moreover, likelihood ratio tests confirmed that PRS-SCZ and PRS-BD each contribute independent information to the joint Cox model for T2D, indicating that the opposing directions of their associations with T2D risk are not driven by shared genetic variance between the two scores. Finally, individuals with a healthy lifestyle consistently showed lower risks of CLD, T2D, CVD, and dementia across genetic risk strata; however, no significant additive interaction was observed between genetic risk and lifestyle factors. In conclusion, these findings indicate that genetic liabilities for SCZ and BD show outcome-specific associations with somatic illnesses, while a healthy lifestyle is associated with lower somatic illness risk regardless of genetic risk level.

Identifying informative voxels is a critical, yet challenging step in functional magnetic resonance imaging (fMRI), particularly for multivariate analyses involving multiple related conditions. Existing approaches often rely on predefined regions of interest (ROIs) or activation-based criteria, which may be insufficient for capturing fine-grained representational differences. This challenge becomes particularly relevant in experimental settings and interventions such as neurofeedback training, where voxels are not only measured as neural responses but also used as targets for intervention based on their previously observed activity patterns. In this study, we propose a subject-level searchlight optimization framework that integrates voxel-wise general linear model (GLM)–based univariate analysis with representational similarity analysis (RSA)–based multivariate refinement to identify voxels that are both task-relevant and condition-sensitive. To enhance practical applicability, the framework further incorporates a data-driven hyperparameter tuning step based on Bayesian optimization, enabling efficient identification of high-performing configurations from small pilot datasets, with consistent performance when applied to larger samples. The proposed framework was evaluated using an emotion imagery fMRI dataset with four affective conditions. Results demonstrate that the multivariate refinement improves alignment between empirical and target representational structures compared with univariate selection alone. Compared with a classifier-based voxel selection approach, the RSA-based approach better preserves the representational geometry of emotional states while maintaining discriminative capacity. These findings highlight the effectiveness, efficiency, and robustness of the proposed RSA framework, providing a practical solution for identifying condition-sensitive voxels and supporting more precise multivariate investigation of affective brain states in multi-condition fMRI studies.

Structural variants, including copy number variants (CNVs), confer substantial risk for neurodevelopmental and psychiatric disorders (NPDs), yet whether their cortical effects relate to those observed in the psychiatric conditions they predispose to remains unclear. Here, we present the first systematic comparison of cortical phenotypes across 18 NPD-associated CNVs and aneuploidies, disorder-associated common variants, and 8 psychiatric disorders. Rare CNVs preferentially affected total surface area (SA), with 11-fold larger effects than psychiatric diagnoses, while NPDs preferentially affected mean cortical thickness (CT), with most CT effects observed in medicated subgroups, suggesting non-genetic contributions. NPD-associated common variants showed enrichment in SA but not CT associations. Regionally, both rare and common genetic variants showed larger effects in sensorimotor regions, aligning with the sensorimotor-to-association cortical gradient as well as regional heritability estimates. In contrast, psychiatric diagnoses showed larger effects in association regions. Individual NPD-associated variants were evenly split between those increasing and decreasing surface area. This heterogeneity likely explains why aggregating variants using polygenic scores shows only weak associations with SA. Overall, cortical signatures of psychiatric diagnoses diverge from those associated with genetic risk. Genetic variants preferentially impact SA and sensorimotor regions through early developmental mechanisms, while psychiatric diagnoses are associated with CT and association regions likely reflecting medication, illness chronicity, and environmental factors.

Schizophrenia is increasingly recognized as a neurodevelopmental disorder arising from excitation-inhibition (E/I) imbalance within cortical and subcortical networks. The E/I imbalance model posits that converging genetic, epigenetic and environmental influences act through glutamatergic (Glu) and gamma-aminobutyric acid (GABA)-ergic dysfunction, redox and immune dysregulation to destabilize neural E/I homeostasis across sequential stages of the neurodevelopmental trajectory. This framework provides a mechanistically integrated explanation for the diverse symptom dimensions and progressive course of schizophrenia, while identifying modifiable targets for intervention. Grounded in the E/I imbalance model, we outline a novel mechanistically informed framework for treatment of schizophrenia with an emphasis on prevention and recovery. Across the lifespan, this approach advocates stage-specific preventive strategies that begin well before illness onset, extend through early and chronic phases of the illness and incorporate therapeutic strategies to modify underlying neurobiology rather than simply alleviating symptoms. By connecting molecular neuroscience with clinical medicine and precision health approaches, the model outlines a realistic path towards mechanism-based prevention and recovery that is applicable not only to schizophrenia and neuropsychiatric disorders, but also to other non-communicable systemic disorders.

The Clinical High Risk (CHR) state for psychosis is consistently associated with widespread cortical thinning. However, the underlying mechanisms driving this neuroanatomical phenotype remain poorly understood. Here, we integrated the ENIGMA CHR Working Group's large pooled dataset (N = 1782 CHR, N = 1333 healthy controls) with an open-source PET molecular atlas to identify, for the first time, potential neurochemical drivers of cortical thinning associated with psychosis risk, transition, and its core symptoms. Using multilinear model analysis, we show that local chemoarchitecture significantly explains CT differences associated with CHR case-control status, the severity of negative symptoms, and future psychosis transition after excluding medication confounds. PET-based maps of dopamine, GABA, glutamate, serotonin, and norepinephrine consistently emerged as the strongest predictors of lower CT in CHR and psychosis transition (total dominance range: 62-69% and 58-87%, respectively), with contributions of monoamine systems being especially sensitive to medication exposure (8-23% change in dominance range). Negative symptom-associated cortical thinning was best explained by PET-based maps of dopamine, histamine, serotonin and opioid systems (total dominance range: 60-81%), with contributions of histamine being sensitive to medication exposure (9-19% change in dominance range). Combined, these results uniquely identify specific neurochemical systems - particularly monoaminergic, glutamatergic, and GABAergic pathways - as key molecular mechanisms associated with cortical thinning in people at high risk of developing psychosis.

Real-time functional magnetic resonance imaging neurofeedback (rt-fMRI-NF) is a promising noninvasive brain computer interface (BCI) technique for enhancing self-regulation of affective brain states. However, conventional univariate rt-fMRI-NF approaches struggle to discriminate distributed neural patterns underlying distinct emotions. This study implemented an rt-fMRI semantic neurofeedback (rt-fMRI-sNF) paradigm incorporating real-time representational similarity analysis (rt-RSA) to enable navigation among emotional states. Four emotion-specific base patterns were first derived from functional localizer runs and then used as target patterns during neurofeedback. Using an RSA-informed circular semantic map (CSM), participants received real-time visual feedback indicating both the similarity and intensity of their current brain activity relative to target patterns. Participants were instructed to use mental imagery to shift their brain activity toward the specific target pattern and enhance its intensity. Analyses of localizer data revealed overlapping regional activations across emotions and demonstrated that RSA reliably distinguished between emotional states. Group-level mixed-effects modeling of neurofeedback performance indicated significant within-run improvements and higher initial performance in the second run. Together, these results demonstrate the methodological feasibility of an RSA-informed rt-fMRI-NF framework for multivariate brain-state modulation and establish a foundation for future studies examining its transferability and clinical relevance.