Background: Polygenic scores (PGSs) are increasingly used to investigate the genetic architecture of complex traits. In genetics, family study designs are often used to adjust for confounders such as population structure and shared environment. However, family studies may also be particularly vulnerable to to non-random ascertainment, for example when individual case status affects the probability of inclusion, leading to differential representation of sibling pairs. In sibling samples, PGS associations can be decomposed into within-family and between-family components, where the within-family estimate captures associations between sibling differences in PGS and differences in outcome, thereby providing an estimate that is less affected by shared familial confounding. In this study, we examined the impact of non-random sampling on estimated genetic effects in family-based studies using both simulations and real-world data. Further, we leveraged the iPSYCH study design to estimate within-family PGS effects for six common mental health outcomes and whether accounting for these can improve prediction accuracy. Methods: We conducted simulations and applied the same framework to real-world data to evaluate the impact of selection bias on within- and between-family PGS estimates. Selection bias was modelled through differential sampling of sibling pairs based on case status, and inverse probability weighting (IPW) was applied to adjust for known heterogeneous inclusion probabilities. Analyses were replicated in the iPSYCH cohort using registry-based sampling weights and PGSs for six major psychiatric disorders. Predictive performance of models was assessed using five-fold cross-validation. Results: In simulation studies, biased sampling led to deviations in estimated PGS effects, with greater distortion observed for between-family components. IPW adjustment reduced the discrepancy between estimates obtained from biased and true underlying data. In the iPSYCH cohort, between-family estimates from unweighted models were larger than within-family estimates across traits. IPW weighted attenuated several of these estimates. Prediction analyses comparing models using total PGS versus decomposed within- and between-family components showed minimal differences in area under the curve and scaled R2 in the iPSYCH data, while modest gains were observed in selected simulation scenarios. Conclusions: Non-random ascertainment distorts effect estimates in family-based models, with particular sensitivity when estimating between-family effects. Incorporating IPWs derived from known or estimable inclusion probabilities can reduce this bias. Our findings highlight the importance of accounting for selection bias in family studies when estimating genetic effects

Clinical deployment of automated brain MRI analysis faces a fundamental challenge: clinical data is heterogeneous and noisy, and high-quality labels are prohibitively costly to obtain. Self-supervised learning (SSL) can address this by leveraging the vast amounts of unlabeled data produced in clinical workflows to train robust foundation models that adapt out-of-domain with minimal supervision. However, the development of foundation models for brain MRI has been limited by small pretraining datasets and in-domain benchmarking focused on high-quality, research-grade data. To address this gap, we organized the FOMO25 challenge as a satellite event at MICCAI 2025. FOMO25 provided participants with a large pretraining dataset, FOMO60K, and evaluated models on data sourced directly from clinical workflows in few-shot and out-of-domain settings. Tasks covered infarct classification, meningioma segmentation, and brain age regression, and considered both models trained on FOMO60K (method track) and any data (open track). Nineteen foundation models from sixteen teams were evaluated using a standardized containerized pipeline. Results show that (a) self-supervised pretraining improves generalization on clinical data under domain shift, with the strongest models trained out-of-domain surpassing supervised baselines trained in-domain. (b) No single pretraining objective benefits all tasks: MAE favors segmentation, hybrid reconstruction-contrastive objectives favor classification, and (c) strong performance was achieved by small pretrained models, and improvements from scaling model size and training duration did not yield reliable benefits.

Background Currently available cardiovascular disease (CVD) risk prediction tools may underestimate the risk in individuals with schizophrenia.Objective To develop and externally validate 5-year CVD risk prediction models for people with schizophrenia using large-scale register data in Sweden and Denmark with a machine learning (ML) approach.Methods Individuals with a diagnosis of schizophrenia, aged 30 and older and without prior CVD, were followed for up to 5 years. We investigated whether adding additional health-related and socio-demographic predictors to the established CVD risk factors improved predictions and compared ML models with logistic regression. External validation was performed across countries.Findings A lasso penalised logistic regression including additional predictors achieved the highest predictive performance, both on Swedish and Danish data, while complex ML models with interaction terms did not provide additional improvements. The area under the receiver operating characteristic curve (AUC) on the internal validation data was 0.745 (95% CI (0.742 to 0.749)) in the Swedish model, and 0.722, 95% CI (0.719 to 0.726) in the Danish model. External validation showed similar performance, yielding an AUC of 0.746, 95% CI (0.741 to 0.751) using the Danish model on the Swedish data, and an AUC of 0.720, 95% CI (0.712 to 0.726) using the Swedish model on the Danish validation data.Conclusions Incorporating additional health-related information, such as psychiatric comorbidities and medication use, improved 5-year CVD risk prediction for people with schizophrenia in both countries.Clinical implications The models can be deployed between Denmark and Sweden without loss of performance compared with training a model on each country.

QUESTION:To what extent are psychiatry guidelines supported by high-level evidence? STUDY SELECTION AND ANALYSIS:Guidelines from the American Psychiatric Association, European Psychiatric Association, WHO and World Federation of Societies for Biological Psychiatry (2014-2024) were selected. Recommendations were graded by guideline authors' levels of evidence (LOE) appraisal (standardised to the Grading of Recommendations, Assessment, Development and Evaluations framework (high, moderate, low, very low)) and by the highest-level study referenced (meta-analysis, randomised controlled trial (RCT), observational study, expert opinion, etc.). FINDINGS:24 guidelines, containing 545 recommendations, were included. Of 82 guidelines screened, 29 (35%) had not been updated in a decade. 63 (11.6%) recommendations were rated by guideline authors as based on high LOE. The proportion was the highest for pharmacotherapies (41/281 (14.6%)) and the lowest for somatic assessment (0/13 (0%)). The proportion of high LOE recommendations varied between publishers (European Psychiatric Association: 20 %, WHO: 1.6 %). For high LOE recommendations, only those concerning pharmacotherapies cited meta-analyses based on double-blind studies using adequate controls. A large proportion (n=241 (44.2%)) of recommendations cited either a meta-analysis of RCTs (n=155 (28.4%)) or ≥two RCTs (n=86 (15.8 %)). There were few recommendations primarily addressing self-harm (n=2), autism (n=3), attention-deficit/hyperactivity disorder (n=3), prevention (n=3), patient involvement (n=3) or discontinuation (n=6). CONCLUSIONS:Clinical guidelines in psychiatry frequently cite RCTs, but the evidence is often downgraded by guideline authors, highlighting the need for better quality trials. LOE varies across areas, with pharmacotherapies supported by the highest quality evidence. Organisations should commit to a timely update of guidelines covering all areas of psychiatry.

Large-scale mega-analyses of worldwide combined Magnetic Resonance Imaging (MRI) studies have demonstrated brain differences between individuals with mental disorders and controls. However, the potential of large-scale observational studies using population-based clinical MRI data remains unexplored. We analyzed clinical MRI data from 23,545 patients in the Eastern half of Denmark (Capital Region of Denmark and Region Zealand). 2774 patients with mental disorders and 2062 non-psychiatric controls fulfilled our inclusion and exclusion criteria. Patients with mental disorders exhibited smaller thalamic (d = -0.298) and amygdala volumes (d = -0.250), with larger ventricles (d = 0.272), and thinner insula (d = -0.177), all p < 0.0001. Analysis across all ROIs revealed a widespread pattern of thinner cortex (d = -0.180), especially in the temporal pole (d = -0.234) and superior frontal (d = -0.212) regions, and increased extracerebral cerebrospinal fluid (d = 0.264). For volumetric measurements, findings were consistent across different inclusion and exclusion criteria but varied for cortical thickness measurements. Utilizing this currently largest population-based MRI cohort for mental disorders, we demonstrate that clinical MRI scans can detect brain structural differences among patients with mental disorders in real-world clinical settings, aiding in the stratification of patients without mental disorders. Cross-disorder analyses reveal shared neuroanatomical changes, including globally smaller brain volumes and thinner cortex. Integrating large-scale clinical MRI data with electronic health records holds promise for improved patient stratification and tracking of disease progression for future longitudinal cross-disorder studies, bridging real-world MRI data with clinical trajectories for further biological subgrouping.

Abstract Schizophrenia is highly heritable, yet the molecular mechanisms linking genetic risk to abnormal human brain development remain poorly understood. To address this, we generated dorsal forebrain organoids from 17 individuals with idiopathic schizophrenia and 17 age- and sex-matched controls and profiled them across multiple molecular layers, including single-nucleus transcriptomics, quantitative proteomics, metabolomics and deep post-translational modification (PTM) analysis. The organoids reproducibly modelled early cortical development and showed largely similar cellular composition between schizophrenia and control groups. Surprisingly, transcriptomic differences were relatively limited, with the strongest cell-type-specific changes observed in Cajal–Retzius neurons. In contrast, proteomic and particularly PTM-level analyses revealed widespread molecular disruption affecting pathways involved in neuronal migration, neurite development, synaptic function, protein kinase signalling, extracellular matrix organisation and lipid metabolism. Many of the earliest disease-associated changes emerged at the level of protein phosphorylation, consistent with altered neuronal maturation and neurite dynamics. At later developmental stages, schizophrenia organoids showed reduced abundance of synaptic proteins, fewer synaptic puncta and evidence of dysregulated retinoic acid and YAP1 signalling. Notably, most disease-associated alterations occurred independently of changes in transcript or protein abundance, indicating that key aspects of schizophrenia biology are encoded in protein state rather than expression level. These findings identify sex-specific dysregulation of protein state as a major molecular feature of schizophrenia and demonstrate the value of multi-layer proteomic approaches for uncovering disease mechanisms missed by transcriptomics alone.

Neural organoids are invaluable model systems for studying neurodevelopment, generated by either guided or unguided approaches. Despite the importance for the field, the resulting differences between these models are unclear. To obtain an unbiased comparison, we performed a multi-omic analysis of forebrain organoids generated in parallel with two widely applied guided and unguided protocols. The guided forebrain organoids contained a larger proportion of neurons, including GABAergic interneurons, whereas the unguided organoids contained significantly more choroid plexus, radial glia, and astrocytes at later stages. Substantial differences in metabolic profiles were identified, pointing to increased levels of oxidative phosphorylation and fatty acid β-oxidation in the unguided forebrain organoids and a higher reliance on glycolysis in the guided forebrain organoids. Overall, our study comprises a thorough description of the multi-omic differences between these guided and unguided forebrain organoids and provides an important resource for the neural organoid field studying neurodevelopment and disease.

We present FOMO260K, a large-scale, heterogeneous dataset of 260,927 brain Magnetic Resonance Imaging (MRI) scans from 77,589 MRI sessions and 55,378 subjects, aggregated from 910 publicly available sources. The dataset includes both clinical- and research-grade images, multiple MRI sequences, and a wide range of anatomical and pathological variability, including scans with large brain anomalies. Minimal preprocessing was applied to preserve the original image characteristics while reducing entry barriers for new users. Companion code for self-supervised pretraining and finetuning is provided, along with pretrained models. FOMO260K is intended to support the development and benchmarking of self-supervised learning methods in medical imaging at scale.

Objectives:We aimed to summarise findings on blood cytokine levels in patients with paediatric-onset inflammatory bowel disease (PIBD) compared to healthy controls. Methods:This systematic review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines and was registered at Prospero (CRD42024579684). A literature search was performed on 1 August 2024 in PubMed, EMBASE, Web of Science and Scopus. We included studies that reported levels of cytokines in plasma or serum. Pooled effect sizes of mean values were calculated using Hedges' g for any cytokine reported by two or more studies. Results:The search revealed 5529 papers. Twenty-three articles reporting on 58 cytokines were included, totalling 1018 patients with PIBD and 634 healthy controls. Meta-analysis was feasible for three cytokines: interleukin (IL)-1β, IL-6 and tumour necrosis factor (TNF)-α. IL-6 levels were elevated in PIBD (n = 482, standardised mean difference [SMD]: 1.71; 95% confidence interval [CI]: 1.00-2.43), although with high heterogeneity (I 2: 96.69%). The effect size for patients with Crohn's disease (n = 336) was SMD 1.93 (95% CI: 0.85-3.00, I 2: 97.81%), and for patients with ulcerative colitis (n = 85) the SMD was 1.58 (95% CI: 0.34-2.83, I 2: 93.98%). In contrast, levels of IL-1β (SMD: 1.07, 95% CI: -1.01 to 3.14, I 2: 93.8%) and TNF-α (SMD: 0.22; 95% CI: -0.17 to 0.60; I 2: 77.0%) did not differ significantly between patients and controls. Additionally, the associations with the remaining 55 cytokines not eligible for meta-analysis were reviewed. Conclusions:IL-6 was increased in the peripheral blood of patients with PIBD. Further proteomic studies, including correlation analyses of clinical data, are encouraged.

The coronavirus disease 2019 (COVID-19) pandemic and lockdowns prompted a major concern for mental health effects. Comprehensive nationwide studies are lacking on the indirect effect of the COVID-19 pandemic on the mental health of the population. We aimed to determine whether the COVID-19 pandemic and lockdowns affected mental health service usage, suicide attempts and suicides. This comprehensive nationwide register-linked study followed all individuals in Denmark from 1990. The main outcomes were rates of psychiatric admissions, use of psychotropic medication, suicide attempts, suicides, patients in community-based private psychiatry or psychology practices and referrals to psychiatric hospitals. The impact of the pandemic (11 March 2020-30 June 2023) and lockdowns was assessed with log-normal models adjusted for pre-pandemic trends (1 January 2017-10 March 2020). We reported rate ratios (RR) of the observed and counterfactual rates. We identified the 5 807 714 (50.3% female) individuals living in Denmark on 1 March 2020. The rates of psychiatric admissions [RR: 0.95, 95% confidence interval (CI): 0.91 to 0.99, P-value: 0.017] and suicide attempts (RR: 0.85, 95% CI: 0.76 to 0.95, P-value: 0.007) were lower during the pandemic compared with the pre-pandemic trend. The rates of suicides (RR: 0.89, 95% CI: 0.75-1.05, P-value: 0.173), patients in private practices (RR: 1.00, 95% CI: 0.96-1.04, P-value: 0.986) and referrals (RR: 1.06, 95% CI: 0.95-1.18, P-value: 0.307) were not significantly different during the pandemic compared with the pre-pandemic trend. During the first lockdown, rates were lower for psychiatric admissions (RR: 0.85, 95% CI: 0.80 to 0.90, P-value <0.001), suicide attempts (RR: 0.80, 95% CI: 0.69 to 0.94, P-value: 0.007), suicides (RR: 0.67, 95% CI: 0.52 to 0.86, P-value: 0.002), patients in private practices (RR: 0.88, 95% CI: 0.82 to 0.93, P-value <0.001) and referrals (RR: 0.69, 95% CI: 0.60 to 0.81, P-value <0.001) compared with the pre-pandemic trend. However, during the pandemic, the rate of psychotropic medication users increased by 6% compared with the pre-pandemic trend (RR: 1.06, 95% CI: 1.05 to 1.06, P-value < 0.001). The COVID-19 pandemic and lockdowns did not severely influence pre-pandemic trends of the mental health burden in the population of Denmark on a nationwide level.

Background:Psychiatric disorders represent a significant global health burden with complex etiologies involving both genetic and environmental factors. However, while the main effects of genetic and environmental factors are frequently studied, their interplay in shaping psychiatric disorder risk remains poorly understood. This study investigates the interaction between polygenic scores (PGS) as proxies for genetic risk and environmental factors in the risk of psychiatric disorders. Method:This study utilized the iPSYCH case-cohort sample (n = 141,265). Environmental factors, including region, urbanicity, parental socioeconomic status, parental age, parental psychiatric history and early-life exposures: autoimmune disease, brain injury, and central nervous system (CNS) infections, were obtained from Danish nationwide registers. Logistic regression was used to examine the effects of targeted disorder-specific PGSs, environmental factors and their interactions on psychiatric disorders. Analyses were adjusted for age, sex and ancestral principal components to account for population stratification. Results:Both PGS and environmental factors were associated with psychiatric disorders. We found limited evidence of gene-environment interactions across the investigated psychiatric disorders. Most interaction terms were small and not statistically significant, but a few remained significant. These included a smaller PGS association with attention deficit hyperactivity disorder in Southern Denmark compared with the Capital Region, a reduced PGS association with bipolar disorder in the medium and lowest parental income groups compared with the highest income group, and a weaker PGS association with major depressive disorder in individuals with parental psychiatric history compared with those without such history. In contrast, a larger PGS association with schizophrenia was observed in the paternal age group 31-35 years compared with the 26-30 years reference group, and a stronger PGS association with anorexia nervosa in North Denmark compared with the Capital Region. Conclusion:Genetic liability for psychiatric disorders, captured through PGS, was associated with mental disorder risk across environmental contexts. Only a few gene-environment interactions were significant, and these were modest and mental disorder-specific. Overall, the findings highlight the difficulty of detecting robust gene-environment interactions and the need for studies of sufficient scale and statistical power to identify subtle gene-environment effects and improve understanding of psychiatric disorder etiology.

Polygenic prediction has yet to make a major clinical breakthrough in precision medicine and psychiatry, where the application of polygenic risk scores is expected to improve clinical decision-making. Most widely used approaches for estimating polygenic risk scores are based on summary statistics from external large-scale genome-wide association studies, which rely on assumptions of matching data distributions. This may hinder the impact of polygenic risk scores in modern diverse populations due to small differences in genetic architectures. Reference-free estimators of polygenic scores are instead based on genomic best linear unbiased predictions and model the population of interest directly. We introduce a framework, named hapla, with a novel algorithm for clustering haplotypes in phased genotype data to estimate heritability and perform reference-free polygenic prediction in complex traits. We utilize inferred haplotype clusters to compute accurate heritability estimates and polygenic scores in a simulation study and the iPSYCH2012 case-cohort for depression disorders and schizophrenia. We demonstrate that our haplotype-based approach robustly outperforms standard genotype-based approaches, which can help pave the way for polygenic risk scores in the future of precision medicine and psychiatry.

Introduction: Increasing evidence has highlighted bidirectional associations between mental disorders and general medical conditions, with underlying causes ranging from lifestyle habits and side effects from medications to genetic contributions. Novel methods now provide a way to estimate the shared genetic underpinnings and the possibility of a causal relationship between conditions. Methods: Using summary statistics from large genome-wide association studies of 16 categories of general medical conditions and 12 categories of mental disorders, we estimated pairwise genetic correlations between general medical conditions and mental disorders using LD score regression. For conditions with significant, positive genetic correlations, we used the latent causal variable (LCV) model to assess the evidence for a causal relationship between them. Results: Ninety-five out of 192 pairs of conditions were significantly genetically correlated (q <= 0.05). Strong and significant correlations were found between conditions such as infections and a psychiatric cross-disorder phenotype (r(g) = 0.50, p = 1.33 x 10(-6)) and irritable bowel syndrome and depression (rg = 0.58, p = 1.50 x 10(-16)). In the causality analyses, statistically significant evidence for causality was obtained for seven pairs of conditions, including infections being causal to the psychiatric cross-disorder phenotype, metabolic disorders being causal to attention deficit/hyperactivity disorder (ADHD), post-traumatic stress disorder (PTSD) being causal to bone and cartilage disorders, arthropathies and epilepsy, obsessive-compulsive disorder (OCD) being causal to irritable bowel syndrome (IBS), and ADHD being causal to arthropathies. Conclusions: Multiple pairs of general medical conditions and mental disorders were significantly genetically correlated, and for some pairs, there was genetic evidence for a causal relationship. Our findings can inform further molecular studies and clinical practice, raising awareness of the possible co-occurrence of these conditions.

BACKGROUND: Immune dysregulation has been implicated in Alzheimer's disease; however, precise mechanisms and timing have not been established. OBJECTIVE: To investigate the concurrent and longitudinal associations of serum C-reactive protein (CRP) and dietary inflammatory index (DII) with cognitive decline as observed in Alzheimer's disease. METHODS: The study was based on 7613 individuals who participated in Tromsø6 (2007-2008) and Tromsø7 (2015-2016). We analyzed the relationship between CRP levels, DII, and cognitive function cross-sectionally using linear regression. We used mediation analysis to examine if CRP mediates the effects of DII on cognitive function. Further, we related baseline serum CRP to cognitive function and to change in cognitive function after 7 years of follow up. We used linear mixed models to relate changes in CRP levels to changes in cognitive function measured at two time points with 7 years apart. RESULTS: Both CRP level and DII were cross-sectionally inversely associated with cognitive function (psychomotor speed, executive function). There was no prospective relationship between CRP level at baseline and cognitive function after 7 years of follow up. Increase in CRP levels was associated with decrease in cognitive function (psychomotor speed, executive function, and verbal memory) observed between two measurements 7 years apart. The mediation model did not show convincing evidence of a mediating effect of CRP in the association between diet and cognitive function. CONCLUSIONS: After comprehensive analysis of associations between CRP, DII and cognitive function, we conclude that CRP is likely to reflect the changes in inflammatory environment occurring in parallel with cognitive decline.

Abstract Background Cytokines play a central role in the aetiology, disease severity and treatment of adult-onset inflammatory bowel disease (IBD)1. However, despite the aggressive phenotype reported in paediatric-onset IBD, little is known about the cytokine levels in paediatric-onset IBD. This systematic review and meta-analysis aimed to summarize findings of cytokine levels in peripheral blood in patients with paediatric-onset IBD compared to healthy controls. Methods This systematic review followed the PRISMA guidelines2 and was registered at Prospero [CRD42024579684]. A literature search was performed on the 1st of August 2024 in PubMed, EMBASE, Web of Science and Scopus. We included studies that reported levels of cytokines in plasma or serum, in patients with paediatric-onset IBD and healthy controls. To ensure a complete overview, we did not exclude studies based on patients’ treatment status. Pooled effect sizes of mean values were calculated using Hedges’ g for any cytokine reported by two or more studies. The Newcastle-Ottawa scale was used for quality assessment. Results The literature search revealed 10,110 papers (4,582 duplicates), and 5,528 articles were independently screened by two authors. Twenty-one articles met the inclusion criteria including a total of 950 patients with paediatric-onset IBD, and 481 healthy controls. Studies reported on 57 different cytokines. Meta-analysis was performed for interleukin-6 (IL-6) (12 studies, of which 7 provided mean values) and tumour necrosis factor-α (TNF-α) (7 studies, of which 2 provided mean values). Mean values of other cytokines were not provided by two or more studies, and meta-analysis could therefore not be performed. Pooled effect sizes showed increased levels of IL-6 (standardized mean difference: 1.99, 95% confidence interval: 1.15-3.26). However, the heterogeneity between studies was high (Figure 1). Levels of TNF-α did not differ between patients and controls (standardized mean difference: 0.36 95% confidence interval: -0.08-0.79). Overall, the included studies had moderate-good quality. Conclusion IL-6 was increased in the peripheral blood of patients with paediatric-onset IBD. Data on other cytokines were scarce. References 1)Neurath MF. Cytokines in inflammatory bowel disease. Nat Rev Immunol. 2014;14(5):329-342. doi:10.1038/nri3661 2)Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. Published 2021 Mar 29. doi:10.1136/bmj.n71

The chemokine receptor variant CCR5delta32 is linked to HIV-1 resistance and other conditions. Its evolutionary history and allele frequency (10%-16%) in European populations have been extensively debated. We provide a detailed perspective of the evolutionary history of the deletion through time and space. We discovered that the CCR5delta32 allele arose on a pre-existing haplotype consisting of 84 variants. Using this information, we developed a haplotype-aware probabilistic model to screen 934 low-coverage ancient genomes and traced the origin of the CCR5delta32 deletion to at least 6,700 years before the present (BP) in the Western Eurasian Steppe region. Furthermore, we present strong evidence for positive selection acting upon the CCR5delta32 haplotype between 8,000 and 2,000 years BP in Western Eurasia and show that the presence of the haplotype in Latin America can be explained by post-Columbian genetic exchanges. Finally, we point to complex CCR5delta32 genotype-haplotype-phenotype relationships, which demand consideration when targeting the CCR5 receptor for therapeutic strategies.

Post-stroke depression is a common consequence of stroke with an estimated prevalence of approximately 30% in stroke patients. It negatively impacts both rehabilitation and quality of life after stroke. Stroke induces an acute activation of the immune system in the central nervous system with concomitant immunologic alterations in the periphery. Immunologic alterations have been associated with non-stroke-related depression, and the evidence points to both central and peripheral immune activation with bidirectional interactions. By identifying and evaluating the current evidence of immunologic alterations associated with depression, stroke, and post-stroke depression, we outline the current knowledge and hypotheses on stroke-related immunologic alterations and associations with the subsequent risk of post-stroke depression. This includes immune system alterations in the cerebrospinal fluid and blood; pre- and post-stroke infections; the blood-brain barrier; autoimmunity of the central nervous system; brain imaging; the spleen-brain, gut-brain, and neuroendocrine-immune axes; and immunogenetic studies. All these topics are discussed within the context of post-stroke depression, pointing to a potential involvement of a multifactorial immunologic pathophysiology. In this narrative review, we identify key directions for future research and conclude by offering perspectives related to the therapeutic potential and associated challenges of this underinvestigated but important topic in neuropsychiatry.

Globally, depression represents one of the leading causes of years lived with disability. The effects of current pharmacological treatments are small-to-moderate and often delayed by weeks. Immunological disturbances have been associated with depression and meta-analyses have suggested that anti-inflammatory agents have moderate-to-large anti-depressant effects. The largest effects were reported for glucocorticoids. However, previous trials were too small to legitimize standard use of glucocorticoids as add-on treatment in depression. The purpose of the DEXA-PSYCH study is to investigate the effect and safety of short-term, oral dexamethasone compared to placebo as add-on therapy in moderate-to-severe depression. The DEXA-PSYCH trial is an investigator-initiated, double-blind, randomized, placebo-controlled, parallel-group superiority trial. Three-hundred participants meeting criteria for moderate-to-severe depression will be randomized in a 1:1 ratio to 1 week of either dexamethasone (4 mg/day for 4 days, subsequently 2 mg/day for 3 days) or placebo as add-on treatment to treatment as usual. Both in- and out-patients are eligible. Exclusion criteria include, but are not limited to, psychotic disorders, bipolar disorder, and diabetes. The primary outcome is change from baseline in Montgomery-Asberg Depression Rating Scale (MADRS) score on day 7 and will be analyzed through a Mixed Model for Repeated Measurements (MMRM) in an intention-to-treat analysis. Key secondary outcomes include response and remission rates, efficacy 3 weeks after the intervention, effects on quality of life, and safety outcomes. Other secondary outcomes include overall functioning, fatigue, anxiety, labor-market affiliation, and associations between inflammatory biomarkers and treatment response. The DEXA-PSYCH trial represents the largest trial of its kind globally. If the trial confirms findings from previous, smaller studies, short-term oral dexamethasone could become an attractive augmentation strategy if acute anti-depressant effects are warranted. Dexamethasone is an off-patent, well-known drug readily repurposed for new indications. EU Clinical Trials Number 2022–501428-45–00, Registered 25th of July 2022 in the EU Clinical Trials Information System ( https://euclinicaltrials.eu/search-for-clinical-trials/?lang=en EUCT=2022-501428-45-00 ). WHO Universal Trial Identifier U1111-1280–7614.

Large-scale mega-analyses of worldwide combined Magnetic Resonance Imaging (MRI) studies have demonstrated brain differences between individuals with mental disorders and controls. However, the potential of large-scale observational studies using population-based clinical MRI data remains unexplored. We analyzed clinical MRI data from 23,545 patients in the Eastern half of Denmark (Capital Region of Denmark and Region Zealand). 2,774 patients with mental disorders and 2,062 non-psychiatric controls fulfilled our inclusion and exclusion criteria. Patients with mental disorders exhibited smaller thalamic (d=-0.298) and amygdala volumes (d=-0.250), with larger ventricles (d=0.272), and thinner insula (d=-0.177), all p<0.0001. Analysis across all ROIs revealed a widespread pattern of thinner cortex (d=-0.180), especially in the temporal pole (d=-0.234) and superior frontal (d=-0.212) regions, and increased extracerebral cerebrospinal fluid (d=0.264). For volumetric measurements, findings were consistent across different inclusion and exclusion criteria but varied for cortical thickness measurements. Utilizing this currently largest population-based MRI cohort for mental disorders, we demonstrate that clinical MRI scans can detect brain structural differences among patients with mental disorders in real-world clinical settings, aiding in the stratification of patients without mental disorders. Cross-disorder analyses reveal shared neuroanatomical changes, including globally smaller brain volumes and thinner cortex. Integrating large-scale clinical MRI data with electronic health records holds promise for improved patient stratification and tracking of disease progression for future longitudinal cross-disorder studies, bridging real-world MRI data with clinical trajectories for further biological subgrouping. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement Stefano Cerri has received funding from the Lundbeck Foundation (R449-2023-1512). Mostafa Mehdipour Ghazi has received funding from the Lundbeck Foundation (R400-2022-617). Mads Nielsen is supported by the Pioneer Centre for AI, Danish National Research Foundation, grant number P1. Michael Eriksen Benros has received funding from the Lundbeck Foundation (R278-2018-1411 and R380-2021-1225). ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: Data Protection Agency and the Ethics Committee of the Capital Region of Denmark gave ethical and legal approval for this work (R-22002033, and P-2020-101). I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes All data is available to all researchers based in Denmark and employed in the Capital Region of Denmark after the relevant approvals from the Danish health authorities