Importance:Inflammation is increasingly implicated in the pathophysiology of mood and psychotic disorders. Integrating blood biomarkers and brain imaging may help uncover mechanistic pathways and guide targeted interventions. Objective:To identify shared and distinct multivariate patterns of peripheral inflammation and gray matter volume (GMV) in early-stage depressive and psychotic disorders using a transdiagnostic machine learning approach. Design, Setting, and Participants:The naturalistic multicenter PRONIA study was conducted between February 2014 and May 2019 with a follow-up period of up to 36 months; baseline data were analyzed between August 2021 and April 2024. Eight sites, including inpatient and outpatient facilities, in 5 European countries (Germany, Italy, Switzerland, Finland, and the United Kingdom) were included. The study included individuals with recent-onset depression (ROD, n = 163) or psychosis (ROP, n = 177) or clinical high-risk states for psychosis (CHR-P, n = 172), all with minimal medication exposure, and healthy control (HC) individuals (n = 166). Exposures:Structural magnetic resonance imaging (MRI), peripheral assays of cytokines (eg, interleukin [IL] 6, IL-1β, tumor necrosis factor [TNF] α, C-reactive protein [CRP], brain-derived neurotrophic factor [BDNF], S100 calcium-binding protein B [S100B]); clinical assessments; neurocognitive testing. Main Outcomes and Measures:After data collection, sparse partial least squares was used to identify latent brain-blood signatures. Support vector machine classification evaluated psychosocial and neurocognitive predictors of signature expression using repeated nested cross-validation. Results:A total of 678 participants (346 [51.0%] female; median [IQR] age, 24.0 [20.9-28.9] years) were included. Four signatures were identified. A psychosis signature (ρ = 0.27; P = .002) differentiated ROP from CHR-P with elevated IL-6, TNF-α, and reduced CRP, alongside GMV shifts in corticothalamic circuits. A depression signature (ρ = 0.19; P = .02) differentiated ROD from HC individuals with elevated IL-1β, IL-2, IL-4, S100B, and BDNF and GMV reductions in limbic regions. Additional signatures reflected age (ρ = 0.67) and sex or MRI quality (ρ = 0.53). Psychosocial features, including a differential childhood trauma pattern, predicted both the psychosis (balanced accuracy [BAC] = 67.2%) and depression (BAC = 78.0%) signatures. Cognitive performance predicted only the psychosis signature (BAC = 65.1%). Conclusions and Relevance:In this study, early-stage depression and psychosis exhibited distinct neurobiological signatures involving immune and neuroanatomical markers, challenging fully dimensional disease models. These signatures are shaped by childhood trauma and cognition and may support biologically informed early interventions.

Individuals with psychosis and depression show widespread alterations in brain resting-state functional connectivity (rs-FC), affecting both sensory and higher-order brain regions. In this study, we investigate disruptions in the hierarchical organization of brain functional networks in individuals with psychotic and affective disorders. We derived functional brain gradients, low dimensional representations of rs-FC that capture cortical hierarchy, in a sample of 1071 (56.3% female) participants, including clinical high-risk for psychosis (CHR-P) individuals, recent-onset psychosis (ROP) patients, recent-onset depression (ROD) patients, and healthy controls (HC). We examined regional alterations, network-level alterations and functional differentiation and their relationship to clinical symptoms. In addition, we linked case-control differences to receptor expression maps to explore underlying neurobiological mechanisms. All clinical groups exhibited alterations in the visual-to-sensorimotor gradient, while only ROP patients showed alterations in the sensory-to-association gradient. CHR-P and ROP individuals exhibited lower values in the ventral attention network. Clinical groups combined showed higher values in the somatomotor network, a reduced gradient range and altered between-network dispersion. ROD patients showed reduced within-network dispersion in the attentional networks and a reduced range. Correlational analysis revealed weak associations of gradient measures with functioning, visual dysfunctions and cognition. Case-control differences showed associations to receptor expression maps, suggesting the involvement of neurotransmitter systems in these disruptions. Our findings reveal transdiagnostic and disease-specific alterations of hierarchical brain organization. These alterations indicate deficits in functional integration across psychiatric diseases, highlighting the role of attentional and sensory networks in disease processes.

BACKGROUND:Adolescence and early adulthood are periods of increased vulnerability for psychiatric disorders, when trauma and personality development converge on shared and distinct, often unknown, brain signatures. METHODS:We used sparse partial least squares (SPLS) to identify multivariate signatures between voxelwise gray matter volume (GMV) and the following 3 domains: childhood trauma, personality, and depressivity. We performed structural equation modeling (SEM) among these domains, predicted functional outcome at 9-month follow-up via support vector machine classification, and correlated the SPLS signatures with resilience, coping, and visual dysfunctions. All models were cross-validated in the discovery sample [n = 633; 52.9% female, mean [SD] age = 25.41 [5.98] years] and validated in the replication sample (n = 343; 53.0% female, 24.69 [5.72] years) of the multisite prospective PRONIA (Personalised Prognostic Tools for Early Psychosis Management) cohort, comprising individuals with recent-onset depression or psychosis, psychosis risk syndromes, and healthy control individuals. RESULTS:We identified the following 3 signatures of interest: 1) depressivity, linked to reduced GMV in limbic regions; 2) childhood trauma, associated with GMV in thalamic, frontotemporal, and parietal regions; and 3) a trauma-personality-depressivity signature, relating childhood trauma, personality, and depressivity to GMV in thalamic, occipital, temporal, and limbic regions. SEM revealed that childhood trauma was associated with depressivity directly and also indirectly via a maladaptive personality structure. The trauma-personality-depressivity signature was the strongest predictor of poor functional outcome (balanced accuracy [BAC]Discovery = 75.8%, BACReplication = 83.2%). The depressivity and trauma-personality-depressivity signatures were linked to deficient resilience and coping styles as well as visual dysfunctions. CONCLUSIONS:Childhood trauma, personality, and depressivity are associated with shared and distinct brain signatures spanning the affective-psychotic spectrum. If these factors converge, current and future mental health may be compromised.

The latent structure and longitudinal stability of cognitive heterogeneity during the early course of mood and psychosis spectrum illness has not been well-studied. We determined the presence, stability and characteristics of latent cognitive profiles underlying a transdiagnostic sample of individuals at high risk for psychosis (CHR) or with a recent onset of psychosis (ROP) or depression (ROD). The sample comprised 666 CHR, ROP or ROD individuals. Latent Profile Analysis identified transdiagnostic cognitive profiles in baseline and 11-month follow up data. Latent Transition Analysis established the stability of these profiles and their transition probabilities. Profiles were characterised across several clinical factors and those indexing functioning, neurodevelopment, stress exposure, and physical/brain health. A 3-profile model was most optimal at both timepoints, with profile equivalence metrics indicating a low likelihood of transition between them, and thus, temporal stability. The profiles were labelled 'Average Cognition', Moderately Impaired', 'Severely Impaired', with all diagnoses represented in each. No changes in cognition were observed for any profile over the follow-up despite clinical symptom improvement. The Severely Impaired profile had the lowest premorbid adjustment, brain and cognitive reserve, and the highest levels of functional impairment. A higher number and burden of recent stressful life events were reported in the Average Cognition profile. The findings suggest that stable transdiagnostic latent cognitive profiles are observable even at very early stages of manifest mood and psychotic illness. The Severely Impaired profile appears to map to indices of abnormal neurodevelopment, while the Average Cognition profile appears to represent a more stress-resilient phenotype.

BACKGROUND:Specific cognitive difficulties are common in major depressive disorder, impacting functioning and quality of life. Yet, the timing of their emergence and longitudinal course remains poorly understood. This study aimed to characterise longitudinal cognitive functioning following recent onset depression and its association with changes in depressive symptoms. METHODS:Longitudinal data from the PRONIA (Personalised Prognostic Tools for Early Psychosis Management) cohort recruited from ten European sites were used to evaluate trajectory differences between Healthy Controls (HC) and individuals experiencing recent onset depression (ROD). Linear mixed effect models were used with group-by-time interaction term for trajectory differences between baseline and nine-month follow-up, and the associations between changes in depression symptoms and cognitive functioning among ROD. RESULTS:The sample comprised 420 participants (ROD, N = 151; HC, N = 269) aged 15-40 years (M = 25.4, SD = 6.1; 55% female). Two distinct group-level cognitive trajectories were observed. First, a similar trajectory (i.e., no difference) to HC in visual memory, attention span, verbal learning and memory, visuospatial working memory, emotion recognition, and processing speed. A stable deficit trajectory was observed in mental flexibility, auditory verbal working memory, phonetic and semantic verbal fluency among the ROD group. Analysis within ROD group suggested that these outcomes were unrelated to reductions in depressive symptoms. Changes in visual memory, visuospatial working memory, sustained attention, and processing speed were associated with changes in depressive symptoms, despite being unrelated to baseline variations in depressive symptoms, possibly suggesting a sensitivity to state effects of illness, regardless of baseline severity. CONCLUSIONS:Specific cognitive difficulties are already evident at the first depressive episode and may endure in the short-medium term, irrespective of depressive course. Tailored treatment addressing cognition should be provided early to promote cognitive health and functional recovery.

Background Lipidomic alterations have been reported across schizophrenia (SCZ) and bipolar disorder (BD), but findings are heterogeneous and often overlap across diagnoses, limiting diagnostic specificity. Associations between lipid profiles and illness severity have also been inconsistent when assessed using single symptom scales, raising the possibility that unidimensional measures fail to capture biologically relevant variation. Whether plasma lipidomic alterations relate to multidimensional psychosis severity, and how they relate to polygenic liability, remains unclear. Methods We examined associations among psychiatric and cognitive polygenic risk scores (PRS), plasma lipidomics (361 species across 16 classes), and a machine-learning-derived severe psychosis probability score in a transdiagnostic cohort of individuals with SCZ or BD (PRS n=1,320; lipid subset n=428). Regression and lipid class enrichment analyses tested severity associations. Mediation and canonical correlation analyses assessed integrated genetic-lipid-severity relationships. Results SCZ-PRS (positive), BD-PRS (negative), and educational attainment PRS (negative) showed modest associations (beta = |0.02|) with severe psychosis probability. Lipid class enrichment analysis identified nine classes associated with severity, including increased sphingolipids (dSM, dCer), phosphatidylcholines (PC), triacylglycerides (TAG), and phosphatidylethanolamine plasmalogens (PE-P), alongside decreased phosphatidylcholine plasmalogens (PC-P). Most lipid class associations were robust to adjustment for diagnosis and medication. No significant mediation or shared multivariate genetic-lipid structure was observed. Conclusions Plasma lipidomic variation tracks multidimensional psychosis severity across diagnostic boundaries. These findings suggest that lipidomic alterations may reflect transdiagnostic biological processes linked to illness burden that are not fully captured by categorical diagnoses, single symptom scales, or common variant polygenic risk. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement This study did not receive any funding ### 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: University Hospital Munich's ethics committee (Project number 17-13) 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 A unique feature of the PsyCourse Study is that it has been conceptualized as a continuously growing data resource available to the scientific community. Data sharing will be based on mutually agreed research proposals and within the Open Science framework of the PsyCourse Study.

Abstract Background The severity of positive psychotic symptoms largely defines emerging psychosis syndromes. However, depressive and negative symptoms are strongly psychologically and biologically interlinked. A transdiagnostic exploration of symptom severity across early illness syndromes could enhance the understanding of shared common factors and future trajectories of mental illness. We aimed to identify subgroups based on the severity of positive, negative, and depressive symptoms and assess relationships with: 1) premorbid functioning, 2) longitudinal illness course, 3) genetic risk, and 4) brain volume differences. Methods We analysed 749 participants from a multisite, naturalistic, longitudinal (18 months) cohort study of: clinical high risk for psychosis (n=147), recent onset psychosis (n=161), and healthy controls (n=286), and recent onset depression (n=155). Participants were stratified into subgroups based on severity of baseline positive, negative, and depression symptoms. Baseline and longitudinal differences between groups for clinical, functioning, and polygenic risk scores (schizophrenia, depression, cross-disorder) were assessed with ANOVAs and linear mixed models. Voxel-based morphometry was used to examine whole-brain grey matter volume differences. Discovery findings were replicated in a held-out sample (n=610). Results Participants were stratified into no (n=241), mild (n=50), moderate (n=182), and severe symptom (n=254) subgroups. The mean (SD) age was 25.3 (6.0) and 344 (47.3%) were male. Symptom severity was associated with poorer premorbid functioning and illness trajectory, greater genetic risk, and lower brain volume. Findings were not confounded by the original study groups or symptoms and were largely replicated. Conclusions and relevance Transdiagnostic symptom severity is linked to shared aetiologies, prognoses, and biological markers across diagnoses and illness stages. Such commonalities could guide therapeutic selection and future research aiming to detect unique contributions to specific psychopathologies.

OBJECTIVE:Identifying those at highest risk for making a first suicide attempt during adolescence is crucial to inform early suicide prevention. Our study aimed to predict the first ideation-to-attempt transition during adolescence among children with suicidal ideation at baseline using 187 sociodemographic, clinical, neurocognitive, functional, and structural brain predictors. METHOD:Data were obtained from the multisite, longitudinal Adolescent Brain Cognitive Development℠ (ABCD) study, conducted in 21 US sites among 11,864 children 9 to 10 years of age at baseline, with 4 follow-up waves measured between 2018 and 2022. The primary outcome was suicide attempt reported at any of the follow-up waves among children with suicidal ideation at baseline. Machine learning models were trained using 70% of the sample from 14 sites, and were validated in participants from 7 holdout sites. RESULTS:The final sample included 660 children with suicidal ideation at baseline (no previous suicide attempt; mean age = 9.91 years, SD = 0.63 years; 42% female at baseline), of whom 83 children had a first suicide attempt within 4-year follow-up. The final model, which excluded the brain imaging feature as its inclusion did not improve performance, generalized well to the external holdout sites (area under the receiver operating characteristic curve [95% CI] = 0.75 [0.68, 0.83], sensitivity = 0.65 [0.61, 0.75], specificity = 0.69 [0.50, 0.80], positive predictive value = 0.23 [0.15, 0.34], negative predictive value = 0.94 [0.88, 0.97]), p <. 01) with good expected calibration error of 0.03. The model was unbiased across race and sex subgroups. The top contributing features included female sex, presence of self-harm, access to means, generalized anxiety disorder, social anxiety, impulsivity, severity of suicidal ideation, parental income, and clinical treatment history. CONCLUSION:Our model using clinically accessible features predicts the first-onset suicide attempt in children. Most predictors (eg, suicidal ideation severity, impulsivity, anxiety symptoms) are modifiable, highlighting the potential intervention targets. Findings provide longitudinal evidence for key risk factors for the ideation-to-attempt transition in current suicide theories.

Modern research management, particularly for publicly funded studies, assumes a data governance model in which grantees are considered stewards rather than owners of important data sets. Thus, there is an expectation that collected data are shared as widely as possible with the general research community. This presents problems in complex studies that involve sensitive health information. The latter requires balancing participant privacy with the needs of the research community. Here, we report on the data operation ecosystem crafted for the Accelerating Medicines Partnership® Schizophrenia project, an international observational study of young individuals at clinical high risk for developing a psychotic disorder. We review data capture systems, data dictionaries, organization principles, data flow, security, quality control protocols, data visualization, monitoring, and dissemination through the NIMH Data Archive platform. We focus on the interconnectedness of these steps, where our goal is to design a seamless data flow and an alignment with the FAIR (Findability, Accessibility, Interoperability, and Reusability) principles while balancing local regulatory and ethical considerations. This process-oriented approach leverages automated pipelines for data flow to enhance data quality, speed, and collaboration, underscoring the project’s contribution to advancing research practices involving multisite studies of sensitive mental health conditions. An important feature is the data’s close-to-real-time quality assessment (QA) and quality control (QC). The focus on close-to-real-time QA/QC makes it possible for a subject to redo a testing session, as well as facilitate course corrections to prevent repeating errors in future data acquisition. Watch Dr. Sylvain Bouix discuss his work and this article: https://vimeo.com/1025555648 .

OBJECTIVE:Suicide is one of the leading causes of death among youth worldwide, yet existing studies that aimed to predict the first onset of suicidal thoughts and behaviors (STB) included a limited number of data modalities and/or focused on adult populations. This study aimed to prospectively predict first-onset STB across 4-year follow-ups in adolescents using an existing STB history classification model that was previously applied to baseline data and a new machine learning model with 195 biopsychosocial features. METHOD:Participants were 7,503 unrelated adolescents (54.5% female, ages 9-11 years at baseline) from the multisite, longitudinal Adolescent Brain Cognitive Development (ABCD) Study. An existing baseline STB history classification model was applied to predict longitudinal first-onset STB in adolescents compared with healthy controls and clinical controls (individuals with a mental health disorder but no STB). A new elastic net logistic regression model with 195 features was trained on data from 14 sites (n = 5,220), and the resulting top 15 features were validated at 7 independent sites (n = 2,283). RESULTS:The previously developed model to classify STB lifetime history also prospectively predicted first-onset STB in adolescents with an area under the curve (AUC) [95% CI] of 0.73 [0.70, 0.75], p < .001, compared with healthy controls and AUC [95% CI] of 0.63 [0.60, 0.66], p < .001, compared with clinical controls. The newly trained model with top 15 features performed similarly with AUC [95% CI] of 0.73 [0.71, 0.76], p < .001, and AUC [95% CI] of 0.64 [0.60, 0.66], p < .001, for the same comparison groups. The most consistent predictors across models included female sex, sleep disturbances, and maladaptive home and school environments. CONCLUSION:The models predicted first-onset STB in adolescents with moderate accuracy. This study also confirmed the roles of well-established psychological risk factors for STB and identified several novel neurocognitive and brain imaging risk factors. Future studies should validate these models in large-scale diverse samples before clinical translation. PLAIN LANGUAGE SUMMARY:This study followed over 7,500 adolescents for 4 years and tested 2 machine learning models using psychological, social, and brain data to identify those at risk of experiencing suicidal thoughts or behaviors. Both models predicted first-time suicidal thoughts or behaviors with moderate accuracy. Key risk factors that were identified included being female, experiencing sleep problems, and negative home and school environments. DIVERSITY & INCLUSION STATEMENT:We worked to ensure sex and gender balance in the recruitment of human participants. We worked to ensure race, ethnic, and/or other types of diversity in the recruitment of human participants. We worked to ensure that the study questionnaires were prepared in an inclusive way. Diverse cell lines and/or genomic datasets were not available. One or more of the authors of this paper self-identifies as a member of one or more historically underrepresented racial and/or ethnic groups in science. One or more of the authors of this paper self-identifies as a member of one or more historically underrepresented sexual and/or gender groups in science. We actively worked to promote sex and gender balance in our author group. One or more of the authors of this paper received support from a program designed to increase minority representation in science. We actively worked to promote inclusion of historically underrepresented racial and/or ethnic groups in science in our author group. While citing references scientifically relevant for this work, we also actively worked to promote sex and gender balance in our reference list. While citing references scientifically relevant for this work, we also actively worked to promote inclusion of historically underrepresented racial and/or ethnic groups in science in our reference list. The author list of this paper includes contributors from the location and/or community where the research was conducted who participated in the data collection, design, analysis, and/or interpretation of the work.

BACKGROUND:Almost 40% of individuals at ultra-high risk (UHR) for psychosis experience persistent attenuated psychotic symptoms (APS) yet it is unclear (1) whether they share overlapping clinical and functional outcomes compared to individuals who transition to psychosis, (2) when symptom and functioning trajectories begin to diverge between UHR individuals with different clinical outcomes, and (3) whether non-remission (persistent APS or transition) can be predicted using baseline and/or longitudinal data. STUDY DESIGN:Participants were drawn from 2 randomized clinical trials: Neurapro (n = 220; discovery sample) and STEP (n = 180; external validation sample). First, 12-24 month symptoms and functioning were compared between UHR individuals with persistent APS, sustained remission, or transition to psychosis. Next, short-term changes in symptoms and functioning were compared between groups to determine timepoints at which trajectories began to diverge. Finally, we used support vector machines to predict non-remission (persistent APS or transition) vs sustained remission using data from baseline, 6-month follow-up, and combined baseline and 6-month follow-up. RESULTS:Individuals with persistent APS had substantially poorer outcomes compared to those who remitted, and more closely resembled individuals who later transitioned to psychosis. Despite few baseline differences between groups, clinical and functional trajectories of the persistent APS and transition groups rapidly diverged from those who remitted. Prediction of non-remission was poor using baseline data but improved substantially when using 6-month follow-up or combined baseline-6-month data. CONCLUSIONS:Ultra-high-risk individuals with persistent APS display similar clinical and functional trajectories to transitioned cases, suggesting that more intensive and sustained intervention is required for this subgroup. However, prospective identification of individuals with poor clinical outcomes (ie, persistence or deterioration of attenuated psychotic symptoms) may require longitudinal monitoring of symptom and functioning trajectories for several months.

This Viewpoint explores how suicide risk prediction models could be improved by prioritizing dynamic data, allowing clinicians insight into actionable modifiable risk factors.

BACKGROUND:Early recovery of functioning is critical for favorable outcomes in psychotic and affective disorders. Transdiagnostic brain activity patterns may capture pathways for poor outcomes before clinical manifestation, thereby supporting timely prevention and intervention. METHODS:Using machine learning, we evaluated the transdiagnostic prognostic value of resting-state functional magnetic resonance imaging fractional amplitude of low-frequency fluctuations (fALFF) (slow-5 and slow-4 sub-bands) for functional outcomes in patients at clinical high risk for psychosis (n = 217) or with recent-onset depression (n = 198) from the multisite PRONIA (Prognostic Tools for Early Psychosis Management) study. Leave-site-out cross-validation assessed the geographic generalizability of models across disability and symptom domains, with outcomes defined as snapshots at 9- or 18-month follow-up or across both time points. We examined diagnosis-specific performance, generalization to recent-onset psychosis (n = 140), and negative symptoms and the added value of fALFF over clinical prognostication. RESULTS:Transdiagnostic models predicting stable good functioning across follow-ups showed up to 10% higher balanced accuracy (BAC) than snapshot models. Decreased slow-5 fALFFs in the default mode network, executive control network (ECN), and dorsal attentional network (DAN) and increased fALFF in the salience network, ECN, and DAN predicted impairment with BAC = 67% (sensitivity = 65%, specificity = 70%, p < .001). This model generalized to recent-onset psychosis (BAC = 62%, sensitivity = 64%, specificity = 59%, p < .001) and predicted (BAC = 65%, sensitivity = 66%, specificity = 65%, p < .001) and was mediated by negative symptoms. Slow-5-based models improved prognostic accuracy over expert ratings in disability (BACraters = 66%, BACraters+slow-5 = 75%, W = 1680, p < .001) and symptom (BACraters = 61%, BACraters+slow-5 = 71%, W = 1444, p < .001) domains. CONCLUSIONS:We highlighted the prognostic value of fALFF for functional impairment in psychosis risk and early depression. Leveraging trajectorial information, we identified candidate imaging biomarkers to improve prognostication, thereby supporting personalized prevention and recovery strategies.

IntroductionSchizophrenia is a psychiatric disorder hypothesized to result from disturbed brain connectivity. Structural covariance networks (SCN) describe the shared variation in morphological properties emerging from coordinated neurodevelopmental processes, This study evaluates the potential of SCNs as diagnostic biomarker for schizophrenia.MethodsWe compared the diagnostic value of two SCN computation methods derived from regional gray matter volume (GMV) in 154 patients with a diagnosis of first episode psychosis or recurrent schizophrenia (PAT) and 366 healthy control individuals (HC). The first method (REF-SCN) quantifies the contribution of an individual to a normative reference group’s SCN, and the second approach (KLS-SCN) uses a symmetric version of Kulback-Leibler divergence. Their diagnostic value compared to regional GMV was assessed in a stepwise analysis using a series of linear support vector machines within a nested cross-validation framework and stacked generalization, all models were externally validated in an independent sample (NPAT=71, NHC=74), SCN feature importance was assessed, and the derived risk scores were analyzed for differential relationships with clinical variables.ResultsWe found that models trained on SCNs were able to classify patients with schizophrenia and combining SCNs and regional GMV in a stacked model improved training (balanced accuracy (BAC)=69.96%) and external validation performance (BAC=67.10%). Among all unimodal models, the highest discovery sample performance was achieved by a model trained on REF-SCN (balanced accuracy (BAC=67.03%). All model decisions were driven by widespread structural covariance alterations involving the somato-motor, default mode, control, visual, and the ventral attention networks. Risk estimates derived from KLS-SCNs and regional GMV, but not REF-SCNs, could be predicted from clinical variables, especially driven by body mass index (BMI) and affect-related negative symptoms.DiscussionThese patterns of results show that different SCN computation approaches capture different aspects of the disease. While REF-SCNs contain valuable information for discriminating schizophrenia from healthy control individuals, KLS-SCNs may capture more nuanced symptom-level characteristics similar to those captured by PCA of regional GMV.

Several multivariate prognostic models have been published to predict outcomes in patients with first episode psychosis (FEP), but it remains unclear whether those predictions generalize to independent populations. Using a subset of demographic and clinical baseline predictors, we aimed to develop and externally validate different models predicting functional outcome after a FEP in the context of a schizophrenia-spectrum disorder (FES), based on a previously published cross-validation and machine learning pipeline. A crossover validation approach was adopted in two large, international cohorts (EUFEST, n = 338, and the PSYSCAN FES cohort, n = 226). Scores on the Global Assessment of Functioning scale (GAF) at 12 month follow-up were dichotomized to differentiate between poor (GAF current < 65) and good outcome (GAF current ≥ 65). Pooled non-linear support vector machine (SVM) classifiers trained on the separate cohorts identified patients with a poor outcome with cross-validated balanced accuracies (BAC) of 65-66%, but BAC dropped substantially when the models were applied to patients from a different FES cohort (BAC = 50-56%). A leave-site-out analysis on the merged sample yielded better performance (BAC = 72%), highlighting the effect of combining data from different study designs to overcome calibration issues and improve model transportability. In conclusion, our results indicate that validation of prediction models in an independent sample is essential in assessing the true value of the model. Future external validation studies, as well as attempts to harmonize data collection across studies, are recommended.

Cognitively impaired and spared patient subgroups were identified in psychosis and depression, and in clinical high-risk for psychosis (CHR). Studies suggest differences in underlying brain structural and functional characteristics. It is unclear whether cognitive subgroups are transdiagnostic phenomena in early stages of psychotic and affective disorder which can be validated on the neural level. Patients with recent-onset psychosis (ROP; N = 140; female = 54), recent-onset depression (ROD; N = 130; female = 73), CHR ( N = 128; female = 61) and healthy controls (HC; N = 270; female = 165) were recruited through the multi-site study PRONIA. The transdiagnostic sample and individual study groups were clustered into subgroups based on their performance in eight cognitive domains and characterized by gray matter volume (sMRI) and resting-state functional connectivity (rsFC) using support vector machine (SVM) classification. We identified an impaired subgroup ( N ROP = 79, N ROD = 30, N CHR = 37) showing cognitive impairment in executive functioning, working memory, processing speed and verbal learning (all p < 0.001). A spared subgroup ( N ROP = 61, N ROD = 100, N CHR = 91) performed comparable to HC. Single-disease subgroups indicated that cognitive impairment is stronger pronounced in impaired ROP compared to impaired ROD and CHR. Subgroups in ROP and ROD showed specific symptom- and functioning-patterns. rsFC showed superior accuracy compared to sMRI in differentiating transdiagnostic subgroups from HC (BAC impaired = 58.5%; BAC spared = 61.7%, both: p < 0.01). Cognitive findings were validated in the PRONIA replication sample ( N = 409). Individual cognitive subgroups in ROP, ROD and CHR are more informative than transdiagnostic subgroups as they map onto individual cognitive impairment and specific functioning- and symptom-patterns which show limited overlap in sMRI and rsFC. Clinical trial registry name German Clinical Trials Register (DRKS). Clinical trial registry URL: https://www.drks.de/drks_web/ . Clinical trial registry number: DRKS00005042.

This article describes the rationale, aims, and methodology of the Accelerating Medicines Partnership® Schizophrenia (AMP® SCZ). This is the largest international collaboration to date that will develop algorithms to predict trajectories and outcomes of individuals at clinical high risk (CHR) for psychosis and to advance the development and use of novel pharmacological interventions for CHR individuals. We present a description of the participating research networks and the data processing analysis and coordination center, their processes for data harmonization across 43 sites from 13 participating countries (recruitment across North America, Australia, Europe, Asia, and South America), data flow and quality assessment processes, data analyses, and the transfer of data to the National Institute of Mental Health (NIMH) Data Archive (NDA) for use by the research community. In an expected sample of approximately 2000 CHR individuals and 640 matched healthy controls, AMP SCZ will collect clinical, environmental, and cognitive data along with multimodal biomarkers, including neuroimaging, electrophysiology, fluid biospecimens, speech and facial expression samples, novel measures derived from digital health technologies including smartphone-based daily surveys, and passive sensing as well as actigraphy. The study will investigate a range of clinical outcomes over a 2-year period, including transition to psychosis, remission or persistence of CHR status, attenuated positive symptoms, persistent negative symptoms, mood and anxiety symptoms, and psychosocial functioning. The global reach of AMP SCZ and its harmonized innovative methods promise to catalyze the development of new treatments to address critical unmet clinical and public health needs in CHR individuals.

The concept of ultra-high risk for psychosis (UHR) has been at the forefront of psychiatric research for several decades, with the ultimate goal of preventing the onset of psychotic disorder in high-risk individuals. Orygen (Melbourne, Australia) has led a range of observational and intervention studies in this clinical population. These datasets have now been integrated into the UHR 1000+ cohort, consisting of a sample of 1,245 UHR individuals with a follow-up period ranging from 1 to 16.7 years. This paper describes the cohort, presents a clinical prediction model of transition to psychosis in this cohort, and examines how predictive performance is affected by changes in UHR samples over time. We analyzed transition to psychosis using a Cox proportional hazards model. Clinical predictors for transition to psychosis were investigated in the entire cohort using multiple imputation and Rubin's rule. To assess performance drift over time, data from 1995-2016 were used for initial model fitting, and models were subsequently validated on data from 2017-2020. Over the follow-up period, 220 cases (17.7%) developed a psychotic disorder. Pooled hazard ratio (HR) estimates showed that the Comprehensive Assessment of At-Risk Mental States (CAARMS) Disorganized Speech subscale severity score (HR=1.12, 95% CI: 1.02-1.24, p=0.024), the CAARMS Unusual Thought Content subscale severity score (HR=1.13, 95% CI: 1.03-1.24, p=0.009), the Scale for the Assessment of Negative Symptoms (SANS) total score (HR=1.02, 95% CI: 1.00-1.03, p=0.022), the Social and Occupational Functioning Assessment Scale (SOFAS) score (HR=0.98, 95% CI: 0.97-1.00, p=0.036), and time between onset of symptoms and entry to UHR service (log transformed) (HR=1.10, 95% CI: 1.02-1.19, p=0.013) were predictive of transition to psychosis. UHR individuals who met the brief limited intermittent psychotic symptoms (BLIPS) criteria had a higher probability of transitioning to psychosis than those who met the attenuated psychotic symptoms (APS) criteria (HR=0.48, 95% CI: 0.32-0.73, p=0.001) and those who met the Trait risk criteria (a first-degree relative with a psychotic disorder or a schizotypal personality disorder plus a significant decrease in functioning during the previous year) (HR=0.43, 95% CI: 0.22-0.83, p=0.013). Models based on data from 1995-2016 displayed good calibration at initial model fitting, but showed a drift of 20.2-35.4% in calibration when validated on data from 2017-2020. Large-scale longitudinal data such as those from the UHR 1000+ cohort are required to develop accurate psychosis prediction models. It is critical to assess existing and future risk calculators for temporal drift, that may reduce their utility in clinical practice over time.

Symptom heterogeneity characterizes psychotic disorders and hinders the delineation of underlying biomarkers. Here, we identify symptom-based subtypes of recent-onset psychosis (ROP) patients from the multi-center PRONIA (Personalized Prognostic Tools for Early Psychosis Management) database and explore their multimodal biological and functional signatures. We clustered N = 328 ROP patients based on their maximum factor scores in an exploratory factor analysis on the Positive and Negative Syndrome Scale items. We assessed inter-subgroup differences and compared to N = 464 healthy control (HC) individuals regarding gray matter volume (GMV), neurocognition, polygenic risk scores, and longitudinal functioning trajectories. Finally, we evaluated factor stability at 9- and 18-month follow-ups. A 4-factor solution optimally explained symptom heterogeneity, showing moderate longitudinal stability. The ROP-MOTCOG (Motor/Cognition) subgroup was characterized by GMV reductions within salience, control and default mode networks, predominantly throughout cingulate regions, relative to HC individuals, had the most impaired neurocognition and the highest genetic liability for schizophrenia. ROP-SOCWD (Social Withdrawal) patients showed GMV reductions within medial fronto-temporal regions of the control, default mode, and salience networks, and had the lowest social functioning across time points. ROP-POS (Positive) evidenced GMV decreases in salience, limbic and frontal regions of the control and default mode networks. The ROP-AFF (Affective) subgroup showed GMV reductions in the salience, limbic, and posterior default-mode and control networks, thalamus and cerebellum. GMV reductions in fronto-temporal regions of the salience and control networks were shared across subgroups. Our results highlight the existence of behavioral subgroups with distinct neurobiological and functional profiles in early psychosis, emphasizing the need for refined symptom-based diagnosis and prognosis frameworks.

Aims The specific and multifaceted service needs of young people have driven the development of youth-specific integrated primary mental healthcare models, such as the internationally pioneering headspace services in Australia. Although these services were designed for early intervention, they often need to cater for young people with severe conditions and complex needs, creating challenges in service planning and resource allocation. There is, however, a lack of understanding and consensus on the definition of complexity in such clinical settings.Methods This retrospective study involved analysis of headspace's clinical minimum data set from young people accessing services in Australia between 1 July 2018 and 30 June 2019. Based on consultations with experts, complexity factors were mapped from a range of demographic information, symptom severity, diagnoses, illness stage, primary presenting issues and service engagement patterns. Consensus clustering was used to identify complexity subgroups based on identified factors. Multinomial logistic regression was then used to evaluate whether these complexity subgroups were associated with other risk factors.Results A total of 81,622 episodes of care from 76,021 young people across 113 services were analysed. Around 20% of young people clustered into a 'high complexity' group, presenting with a variety of complexity factors, including severe disorders, a trauma history and psychosocial impairments. Two moderate complexity groups were identified representing 'distress complexity' and 'psychosocial complexity' (about 20% each). Compared with the 'distress complexity' group, young people in the 'psychosocial complexity' group presented with a higher proportion of education, employment and housing issues in addition to psychological distress, and had lower levels of service engagement. The distribution of complexity profiles also varied across different headspace services.Conclusions The proposed data-driven complexity model offers valuable insights for clinical planning and resource allocation. The identified groups highlight the importance of adopting a holistic and multidisciplinary approach to address the diverse factors contributing to clinical complexity. The large number of young people presenting with moderate-to-high complexity to headspace early intervention services emphasises the need for systemic change in youth mental healthcare to ensure the availability of appropriate and timely support for all young people.