JAMA Psychiatry

ObjectiveWe examined the clinical utility of resting state electroencephalography (rsEEG) by evaluating its temporal stability, discriminant validity for B-SNIP psychosis Biotypes, and suitability as a treatment target for brain stimulation. MethodsWe collected 5 minutes of eyes-open rsEEG from 1401 participants with psychosis and 750 healthy persons. A subset of participants was re-tested after 6 months and 12 months (N=109). In a pilot target engagement study (n=5) we collected rsEEG before and after 2 high-definition transcranial direct current stimulation (HD-tDCS) interventions targeting the left dorsolateral prefrontal cortex (dlPFC) and temporoparietal junction (TPJ). Data were reduced with principal component analyses to delta/theta, alpha, beta, and gamma frequency bands, and compared between groups and timepoints. ResultsrsEEG frequency bands displayed good-to-excellent stability and significantly distinguished psychosis Biotypes with large effect sizes. Compared to healthy, Biotype-1 had low activity (average ES=-.58), Biotype-2 had high activity (ES=1.07), and Biotype-3 had slightly elevated activity (ES=.33). There were no rsEEG differences between DSM psychosis groups. After anodal dlPFC stimulation, alpha and gamma power slightly increased while positive symptoms and verbal fluency improved. After cathodal TPJ stimulation, delta/theta power slightly increased while psychoticism and digit sequencing improved. ConclusionsResting state brain activity is a trait-like marker that differentiates B-SNIP psychosis Biotypes, suggesting differing underlying neurophysiology. The pilot intervention supports the feasibility of targeting this underlying neurophysiology with HD-tDCS. Integrating rsEEG in diagnostic procedures and stratified intervention selection may be beneficial for psychosis patients.

Background and HypothesisAbnormal default mode network (DMN) connectivity was observed in both tobacco use and psychotic spectrum disorders, but it remains unknown how psychosis impacts the relationship between connectivity and tobacco use. Interventions targeting the left lateral parietal DMN node (LLPDMN) have modulated DMN connectivity and nicotine craving in psychosis. We aimed to investigate relationships between DMN connectivity, psychotic illness, and tobacco use. Study Design336 participants (psychosis: n=161, control: n=175) reported their tobacco use history and underwent resting-state functional magnetic resonance imaging. We calculated connectivity within DMN and salience network (SN), between DMN-SN, and from LLPDMN to other DMN and SN nodes. Logistic and LASSO regression with bootstrapping were performed to investigate diagnosis-by-connectivity interactions on lifetime tobacco use. Exploratory brainwide analysis was conducted by regressing brainwide connectivity to LLPDMN against daily cigarette use. Study ResultsWe observed a significant diagnosis-by-DMN connectivity interaction for lifetime tobacco use (p=0.0281, coefficient=0.457, OR=1.579, 95% CI=[1.063, 2.411]); in the psychosis group, higher DMN connectivity was associated with higher odds of lifetime tobacco use. LASSO regression yielded four predictors of lifetime tobacco use: age, diagnosis, LLPDMN connectivity to a prefrontal SN node, and the interaction between diagnosis and LLPDMN connectivity to a right parietal DMN node. Brainwide analysis identified bilateral somatomotor clusters where higher connectivity to LLPDMN correlated with higher daily cigarette use (voxel-wise p<0.001, cluster p<0.05). ConclusionsPsychosis diagnosis modified relationship between DMN connectivity and tobacco use. Modulating DMN connectivity may provide a psychosis-specific treatment target for tobacco dependence.

BackgroundIn 2020, Oregon became the first U.S. state to establish a regulated framework for adults to access psilocybin services using naturally-derived mushroom products. No studies have examined mental health outcomes among individuals receiving psilocybin in this context. AimsTo evaluate changes in self-reported symptoms of depression, anxiety, and well-being 30-days post-psilocybin session under the Oregon state-regulated model , and document session-related adverse events and doses consumed. MethodsThis was a naturalistic study (March 2024-April 2025) among adults [&ge;]21 years participating in a legal psilocybin services program. Online surveys were completed pre-session, 1-day, and 30-days post-session. Primary outcomes were change in depression, anxiety, and well-being symptoms pre-session to 30-days post-session evaluated using linear mixed-effects models (random effect: timepoint; fixed effects: sex, concurrent psychiatric medication use, age, session dose [total psilocybin equivalents, TPE, mg: psilocybin mg + 1.39 * psilocin mg]). Adverse events (e.g., hallucinogen persisting perception disorder [HPPD]) were assessed at 1-day and 30-days post-session. ResultsParticipants (n=88; median age 43 years; 52% male) were predominantly Oregon residents (53.4%), psychedelic-experienced (64.8%), and concurrently using psychiatric medication (46.6%). All outcomes improved significantly at 30-days post-session (p<0.001), including in sensitivity analyses stratified by concurrent psychiatric medication usage (p<0.001 all outcomes, both groups). Two participants (2.3%) reported symptoms consistent with HPPD at 1-day post-session, but none at 30-days. Mean dose was 27.8 mg (SD 8.2) TPE. ConclusionsPsilocybin sessions delivered under the Oregon regulatory model were associated with clinically meaningful improvements in depression, anxiety, and well-being 30-days post-session, supporting therapeutic effectiveness of legal psilocybin services.

Psychiatric disorders are fundamentally challenged by symptom heterogeneity, high comorbidity, and the absence of objective biomarkers, which together result in substantial variability in clinical assessment and treatment selection. Patient-generated language captures rich information about subjective experience and symptom severity, which can be systematically encoded and analyzed using computational models, making it a scalable signal for psychiatric assessment. We compare two approaches: (i) a domain-specialized transformer fine-tuned on clinical language, based on the Bio-ClinicalBERT encoder architecture, and (ii) a large-scale instruction-tuned generalist encoder (Instructor-XL) used as a frozen feature extractor with a shallow classification head. A corpus of N = 151,228 de-identified texts was compiled from five public sources, covering four psychiatric phenotypes: anxiety, depression, schizophrenia, and suicidal intention. Models were evaluated using stratified 10-fold cross-validation with cost-sensitive training, prioritizing imbalance-aware metrics, including Macro-F1 and Matthews Correlation Coefficient (MCC), over accuracy. Bio-ClinicalBERT achieved superior overall performance (Macro-F1 = 0.78, MCC = 0.6752), indicating more reliable separation of diagnostically overlapping affective categories. In contrast, Instructor-XL achieved its highest class-specific performance for schizophrenia (F1 = 0.798). Explainability analyses suggest that the domain-specialized model places greater weight on clinically relevant terms, whereas the generalist model relies on a broader set of lexical features.

Background At present, there are no approved pharmacological treatments for people at clinical high risk for psychosis (CHR-P). We sought to assess the acceptability of cannabidiol (CBD): a promising candidate treatment for this population. Methods CHR-P individuals completed a survey which assessed their views on the acceptability of CBD, its expected effectiveness and side effects, and on formulation preferences. Results The sample comprised 55 CHR-P individuals (24.3 years and 69% female). Most (91%) were familiar with CBD, and had previously used cannabis (64%), and around half (42%) had tried over-the-counter CBD. 75% were willing to take CBD as an intervention for mental health problems. Most participants anticipated fewer side effects with CBD than with existing medications, and preferred tablet or capsule formulations over liquids. Discussion CBD is perceived as a highly acceptable treatment among CHR-P individuals.

Psychotic disorders are increasingly recognized as the extreme end of a progressive psychopathology continuum, with less advanced stages including the asymptomatic familial high-risk state (FHR), the help-seeking clinical high-risk state (CHR), and first episode psychosis (FEP). However, we lack a comprehensive study of clinical, cognitive, functional, and neuroanatomical markers across all three early stages of psychosis, limiting our understanding of how the multimodal phenotypes which define psychotic disorders emerge in the broader course of psychopathology. We leveraged a sample of 70 FEP, 40 CHR, 43 FHR, and 41 healthy participants recruited from the same clinical and sociodemographic setting - the first such dataset to be described in the literature. Several markers were elevated in CHR but did not worsen in FEP, including depression/anxiety and difficulties functioning, while FEP was uniquely defined by cognitive impairments and cortical thickness reductions characteristic of those seen in schizophrenia. Across the sample, the dominant axis of joint brain-behaviour variability captured a relationship between reduced cortical thickness and lower cognitive performance, a pattern which was equally established in both CHR and FEP. Initial longitudinal data revealed that depressive and negative symptoms best predicted lower functioning at 6-month follow-up, regardless of group status. Together, our analysis suggests that affective and functional disturbances emerge in earlier stages of psychosis, while cognitive and anatomical abnormalities characterize more advanced ones - though the overlap we observed across groups demonstrates that clinically relevant phenotypes can cut across group boundaries, requiring personalized care to manage.

Background and HypothesisClozapine is the only antipsychotic with protective effects against suicide in schizophrenia (SCZ). Newer third-generation antipsychotics (TGA) have better tolerability and modulate serotonin, dopamine, and N-methyl-d-aspartate neurotransmission pathways implicated in suicide. We aimed to investigate the effects of TGAs on suicide in SCZ. MethodsWe searched seven databases up to December 2023 for SCZ studies that reported suicide data. The primary outcome was suicide deaths and attempts; suicidal ideation was added as a secondary outcome. Random effects meta-analyses quantified suicide risk in randomized controlled trials (RCT) while single proportion meta-analyses assessed longitudinal suicide risk in open label extension trials (OLE). For RCTs, sensitivity analyses were conducted and subgroup analyses explored the impact of dose, drug type, and comparator arm. Study ResultsTwenty articles were included; thirteen excluded higher suicide risk participants. Compared to placebo control, TGAs did not significantly change the risk of primary [RR = 0.65, p = 0.38] or secondary [RR = 0.63, p = 0.15] suicide outcomes. Subgroup and sensitivity analyses were not statistically significant. For OLEs, there was a significant increase in the incidence of primary [Ip = 0.004, p = 0.048] and secondary [Ip = 0.024, p = 0.0013] suicide outcomes, but there was marked study heterogeneity. ConclusionThere is no current trial evidence to show that TGAs significantly impact suicide outcomes in SCZ. The signal from OLEs should be interpreted cautiously due to heterogeneity and requires replication. An effective clozapine alternative is needed for suicide prevention in SCZ.

The COVID-19 pandemic had substantial impact on healthcare systems across the globe, including psychiatric services. Use of electroconvulsive therapy (ECT), a lifesaving intervention for severe mental illness, was reported to have declined during the pandemic in several countries, but nationwide data remain scarce. Using nationwide data from the Danish National Patient Register, we examined all ECT treatments administered in Denmark from September 2019 to May 2025. Weekly treatment numbers were visualized across the three national COVID-19 lockdowns to descriptively assess changes in ECT use. A notable reduction in ECT treatments was observed in the weeks preceding and during the first lockdown (March 11 to May 18, 2020). A post-hoc estimation indicated approximately 1,366 "missed" treatments during the initial pandemic phase in 2020. When these were added to the 27,033 treatments delivered in 2020, the adjusted total approximated annual treatment volumes in 2019 and 2022, suggesting a temporary disruption rather than sustained decline. In contrast, ECT activity during the second and third lockdowns appeared largely unaffected. These findings suggest that ECT provision in Denmark was temporarily reduced during the initial phase of the pandemic but remained resilient thereafter. In the case of a future pandemic, safeguarding timely access to ECT--particularly in early phases-- should be prioritized given its critical role in the treatment of severe mental illness.

Backgroundspeech carries cues to variation in mental state in schizophrenia spectrum disorders/psychotic disorders, typically indexed with clinician-rated scales such as the PANSS. Progress in the automation of speech-based symptom modelling has been constrained by data scale and the underrepresentation of low-resource languages. In this study, we aggregate multi-center recordings to assemble a large corpus and assess symptom-prediction models at scale, to enable more objective and efficient assessments and the early detection of relapse-related signals from speech. MethodsWe compiled data from 453 patients with schizophrenia spectrum disorders, recruited from ten global sites, and clipped their speech recordings into 6,664 segments. Across three feature sets, acoustic-prosodic profile, pretrained multilingual embeddings, and their concatenation, we compared 16 algorithms to predict eight relapse-related PANSS items, including three positive (P1, P2, P3), three negative (N1, N4, N6), and two general (G5, G9) items, on speaker-disjoint splits (80% train, 10% test, and 10% validation). Performance was assessed by root-mean-squared-error (RMSE) at both segment and participant (median aggregation) levels. Best model per item underwent bias checks for age, sex, education, and symptom severity. OutcomesBest-performing models predicted symptoms with prediction errors of 1{middle dot}5 PANSS points or lower: P1 1{middle dot}494/1{middle dot}527, P2 1{middle dot}318/1{middle dot}107, P3 1{middle dot}407/1{middle dot}542, N1 1{middle dot}029/1{middle dot}030, N4 1{middle dot}452/1{middle dot}430, N6 0{middle dot}860/0{middle dot}855, G5 0{middle dot}850/0{middle dot}882, G9 1{middle dot}213/1{middle dot}282 (segment/participant). Performance of the pretrained multilingual embeddings surpassed acoustic-prosodic features and their concatenation. Results were comparable in low-resource languages (e.g., Czech). We found no bias by age, sex, or education, aside from reduced N4 accuracy in males; but performance degraded with higher symptom severity. InterpretationSpeech can support automatic assessment of schizophrenia symptoms using pretrained multilingual embeddings, even without the use of transcripts. Such models show promise as clinically meaningful, efficient, and low-burden tools for real-time monitoring of symptom trajectories. FundingEU Horizon research and innovation programme. Research in contextO_ST_ABSEvidence before this studyC_ST_ABSAutomatic assessment of disease severity is a key issue in schizophrenia research, for which spontaneous speech offers a cost-effective, automatable solution. To evaluate existing evidence for speech-based symptom assessment, two reviewers (RHe, MK) searched PubMed, IEEE Xplore, arXiv, bioRxiv, and medRxiv for publications from inception to Aug 25, 2025, using the terms: ("symptom" OR "PANSS" OR "Positive and Negative Syndrome Scale") AND ("psychosis" OR "schizophrenia") AND ("language" OR "speech" OR "spontaneous speech") AND ("prediction" OR "machine learning" OR "deep learning" OR "algorithm" OR "neural network" OR "AI" OR "artificial intelligence"). Fourteen studies on symptom-level modelling were identified. Ten studies dichotomized clinical scores (e.g., PANSS) into low vs high for classification: five used conventional ML (e.g., random forests) and five used neural networks, with F1 scores ranging from 0{middle dot}60-0{middle dot}85. The remaining four studies, and two of the ten studies as mentioned above, modelled raw scores directly as regression tasks. Two relied solely on conventional regressors and the rest used neural networks, with errors from 0{middle dot}487 for single items (scale 1-7) to 8{middle dot}04 for summed scores (scale 18-126). All studies used free speech for elicitation, except one study, which used a reading task. Three studies incorporated additional tasks, such as picture description and immediate recall. None were multilingual: nine were in English, three in Chinese, one in Swiss German, and one in Brazilian Portuguese. Features spanned a wide range, including acoustic-prosodic profiles, morpho-syntactic structure, semantic organization, pragmatics (including sentiments), and even visual features capturing movement during talking. Representations from pretrained language models were also widely employed. Sample sizes (counting patients with schizophrenia) were generally small: eleven studies enrolled <50 patients, one had 65, and only two exceeded 100 patients. Some increased their effective sample size via multiple recordings per patient or by adding healthy controls and/or patients with other psychiatric disorders (e.g., depression). Added value of this studyTo our knowledge, this is the first multilingual, speech-based study modelling schizophrenia symptom severity with machine learning approach, and it includes the largest cohort of patients with schizophrenia to date. We further increased effective sample size by using diverse elicitation tasks and segmenting recordings into clips. This multilingual corpus empowers the usage of complex models and supports transfer learning from high-resource languages (e.g., English) to low-resource ones (e.g., Czech). For each of eight selected relapse-related PANSS items, the best audio-only models achieved RMSE < 1{middle dot}5, underscoring clinical relevance. We assessed potential biases: no effects were found for age, sex, or education (except poorer N4 performance in males), though performance declined at higher symptom severity. Trained models are released for use. Implications of all the available evidenceWe show that speech is a powerful signal for automatic assessment of schizophrenia symptom severity and holds promise for relapse prediction, even without transcripts. The approach readily extends to incorporate textual features (from manual or automatic transcripts) and more advanced models. Prospective studies with repeated recordings across relapse episodes are needed to validate the utility of our models on relapse prediction, for the sake of supporting precision psychiatry while reducing clinician burden.

ObjectiveAs cannabis use has increased in the United States, so has cannabinoid hyperemesis syndrome (CHS), a disorder characterized by severe nausea, vomiting, and abdominal pain among heavy cannabis users. We previously showed that CHS symptoms are associated with several behavioral and psychological characteristics linked to psychosocial impairment. We examined links between CHS symptoms and suicidal thoughts, behaviors, and proximal suicide risk factors. MethodsWe used data from the National Firearms, Alcohol, Cannabis, and Suicide survey, a nationally representative survey of 7,034 US adults. Items assessed symptoms of CHS and suicidal thoughts and behaviors. Comparisons focused on: those with daily cannabis use and CHS symptoms (n = 191), those with daily cannabis use without CHS symptoms (n = 882), those with past year cannabis use but not daily use (n = 1288), and those without past year cannabis use (n = 4673). ResultsThose with CHS symptoms reported the highest prevalence of suicidal thoughts and behaviors with most lifetime rates being significantly higher than those with daily cannabis use without CHS symptoms. Those with CHS symptoms also reported higher mean-levels of thoughts and feelings associated with suicide (i.e., perceived burdensomeness, thwarted belongingness, defeat, entrapment) than all the other groups. ConclusionsThose with CHS symptoms reported especially high rates of suicidal thoughts, behaviors, and attempts even when compared to others with daily cannabis use. People with CHS symptoms appear to be at high risk of suicide, possibly related to distress from their gastrointestinal symptoms and psychiatric, substance use, and medical comorbidities.

Opioid use disorder (OUD) is characterized by compulsive drug seeking and impaired executive control arising from maladaptive plasticity within cortico-striatal circuits. While transcriptomic studies have identified coding gene alterations in the nucleus accumbens (NAc) and dorsolateral prefrontal cortex (DLPFC), the contribution of the noncoding genome remains poorly defined. Here, we performed integrative transcriptomic analysis of postmortem human NAc and DLPFC to systematically identify and characterize long noncoding RNAs (lncRNAs) in OUD. We identified 36,225 lncRNA loci expressed across reward and executive regions, approximately half of which were previously unannotated. OUD was associated with widespread lncRNA dysregulation in NAc and DLPFC, with lncRNA-centered co-expression modules enriched for neuroimmune signaling, phosphorylation-dependent synaptic pathways, and intracellular receptor cascades. Notably, OUD disrupted circadian rhythmicity of lncRNAs to a degree comparable to or exceeding mRNAs, implicating temporal reorganization of noncoding networks in addiction pathology. Integration with single-nucleus transcriptomic data revealed pronounced neuronal and glial cell type specificity among OUD-associated lncRNAs. Together, these findings demonstrate that lncRNAs represent a critical regulatory layer in reward and executive circuits and suggest that spatial, temporal, and cellular remodeling of the noncoding transcriptome contributes to circuit dysfunction in OUD.

BackgroundCannabinoid hyperemesis syndrome (CHS) is characterized by episodes of severe nausea, vomiting, and abdominal pain among those with heavy cannabis use. We estimated differences between those reporting CHS symptoms and other daily and less frequent cannabis users on drug use, psychiatric problems, other health problems, antisocial behavior, and personality. MethodsThe National Firearms, Alcohol, Cannabis, and Suicide survey was administered to 7034 US adults in 2025. Survey items assessed substance use, common psychiatric symptoms, personality traits, and symptoms of CHS. ResultsThose with CHS symptoms reported the highest rates and greatest variety of drug use compared to others who used cannabis. Those with CHS symptoms reported higher rates of other drug use than those who used cannabis daily without CHS symptoms across a variety of drug classes, including opioids, hallucinogens, and sedatives, higher rates of drug overdoses, and greater use of all drug classes than those with less-than-daily cannabis use. Those with CHS symptoms also reported more depression, anxiety, sleep problems, chronic pain, antisocial behavior, intimate partner violence, and disinhibited personality traits than those who used daily (mean d = 0.58) and less frequently (mean d = 0.69) and those with no cannabis use in the past 12 months (mean d = 0.99). ConclusionsThose with CHS symptoms exhibit a variety of psychological and behavioral problems including higher rates of other drug use, psychiatric symptoms, antisocial behavior, and dysfunctional personality traits. Results highlight the importance of understanding and addressing the broader psychosocial challenges faced by people experiencing CHS symptoms. Highlights O_LICHS symptoms are linked to greater polysubstance use and overdose risk C_LIO_LICHS symptoms are associated with depression, anxiety, sleep, and pain problems C_LIO_LICHS tied to antisocial behavior and intimate partner violence C_LIO_LICHS shows disinhibited personality traits and low well-being C_LIO_LINational survey identifies high-risk psychosocial CHS profile C_LI

BackgroundLipidomic 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. MethodsWe 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. ResultsSCZ-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. ConclusionsPlasma 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.

Evidence-based psychotherapies are first-line treatments for psychiatric disorders, yet response rates remain suboptimal. Noninvasive brain stimulation (NIBS) may augment psychotherapy by modulating treatment-engaged circuits. We conducted a systematic review and meta-analysis of randomized controlled trials comparing active NIBS plus evidence-based psychotherapy versus sham NIBS plus psychotherapy. Following Cochrane methods, we searched six databases through February 2025, screening 1,017 records. Twenty-eight trials (31 treatment arms; 1,506 participants) met inclusion criteria. Active NIBS combined with psychotherapy produced significantly greater symptom improvement than sham NIBS with psychotherapy (standardized mean difference = -0.38, 95% confidence interval [-0.68, -0.08]), with substantial heterogeneity. Moderator analyses revealed critical implementation parameters: repetitive transcranial magnetic stimulation (rTMS) showed significant benefit while transcranial direct current stimulation did not. Non-concurrent delivery--stimulation before or after psychotherapy sessions--was significantly effective, whereas concurrent administration was not. Among psychotherapy modalities, cognitive behavioral therapy combined with NIBS produced significant benefit. Human-delivered psychotherapy, but not computerized formats, significantly enhanced outcomes. By diagnosis, significant effects were observed only for anxiety disorders. Secondary analyses revealed significant anxiety symptom reduction specific to rTMS. Treatment integrity was under-reported: only 39.3% of studies used fully manualized protocols and 10.7% documented therapist adherence. Non-concurrent rTMS paired with human-delivered, manualized cognitive behavioral therapy emerges as the most effective strategy, particularly for anxiety disorders. These findings provide an evidence-based framework for optimizing combined treatment protocols and highlight the need for standardized psychotherapy fidelity monitoring in future trials.

Patients with bipolar depression are at the highest risk for suicidal behavior, comprising [~]10% of all deaths. In the critical period preceding attempts, most are not in contact with mental health professionals to effect antisuicidal strategies. There is an urgent need for decision support tools to help nonspecialist providers identify those at elevated risk to facilitate prevention. However, we lack robust, performant predictive models to form the core of such tools. Here, we build a high-precision predictive model of 30-day risk for suicidal behavior using unique electronic health record data from >220,000 patients with bipolar depression. We show that optimized machine learning approaches offer very strong clinical utility, delivering high Standardized Net Benefit in the context of near-perfect calibration and smooth, threshold-robust decision curves. Our results break the longstanding performance ceiling in suicide risk prediction and highlight the importance of training models for clinical utility as well as discriminative skill.

In-scanner head motion is a recognized source of bias in structural magnetic resonance imaging (sMRI), yet it remains under-addressed in psychiatric neuroimaging where structural difference in patient populations are considered foundational. We examined motion-related bias in grey matter volume estimates across eight independent cohorts comprising 9,664 individuals, including 8,979 neurotypical controls (NC), 497 patients with schizophrenia (SCZ), and 188 patients with bipolar disorder (BD). Motion estimates were derived from multiple fMRI scans acquired within the same scanning session and summarized using principal component analysis. In NC, motion accounted for 1-6% of regional grey matter variance, a magnitude comparable to reported psychiatric case-control effect sizes. Adjusting for motion attenuated SCZ-NC group differences, reducing effect sizes in 85% of brain regions and yielding 5% fewer significant ROIs (pFDR<0.05). In BD, motion correction reduced effect sizes in 97% of regions, with a 24% reduction in significant ROIs. Cross-diagnostic spatial patterns were significantly correlated (r=0.63, p=3x10-{superscript 1}3), explaining a sizable portion of SCZ-BD commonalities. Critically, a falsification analysis in UK Biobank (N=5,123) showed that stratifying NC by motion alone produced grey matter differences accounting for 45-62% of SCZ case-control effect magnitude, underscoring how difficult it is to interpret SCZ-like morphometric differences as tissue properties rather than as motion-driven patterns. These findings urge caution in interpretations of sMRIdifferences in patient-control comparisons and use of systematic fMRI based motion control as standard practice in sMRI analyses.

ObjectiveCognitive deficits are a leading cause of disability in schizophrenia and are linked to poor functional outcomes. There are no first line treatments for these deficits, and their neural basis is poorly understood. While schizophrenia is associated with widespread cognitive deficits, information processing speed is most profoundly impaired. Processing speed deficits have been associated with hyperconnectivity in the Default Mode Network (DMN). We therefore tested if modulating DMN connectivity with single or multiple sessions of transcranial magnetic stimulation (TMS) applied to an individualized DMN target would affect processing speed. MethodsIn the first study, 10 individuals with schizophrenia received single TMS sessions and underwent resting-state neuroimaging and processing speed assessment (Brief Assessment of Cognition in Schizophrenia digit symbol coding) acutely before and after each session. These sessions included excitatory (intermittent theta burst stimulation, iTBS); inhibitory (continuous theta burst stimulation, cTBS); and sham stimulation sessions. In the second study, 29 individuals (17 schizophrenia, 12 non-psychosis controls) received 5 accelerated sessions of cTBS with resting-state neuroimaging and processing speed assessment before and after the course of TMS sessions. ResultsIn the accelerated, multi-session DMN-targeted TMS trial, cTBS improved processing speed in the schizophrenia group (p=0.0124). In individuals with schizophrenia, reduction in DMN connectivity was linked to improvement in processing speed (p=0.021). These changes were dependent on age, where younger participants experienced greater processing speed improvements than older participants (p=0.006). ConclusionsIn sum, personalized network targeted TMS is a novel method for reducing cognitive impairment associated with schizophrenia.

ObjectiveTo delineate the phenotype of juvenile myoclonic epilepsy (JME) with a focus on obsessive-compulsive personality disorder (OCPD) using multimodal psychiatric, neuropsychological, quantitative EEG (qEEG), and structural MRI markers within a predictive-processing/free-energy framework. MethodsWe prospectively studied 65 patients with JME and 68 matched healthy controls (HC). Participants completed DSM-IV SCID I/II interviews and a neuropsychological battery assessing working memory, psychomotor speed, mental flexibility, divided attention, inhibition, and phasic/tonic alertness; standard EEG and high-resolution structural MRI were acquired. Groups comprised HC and JME subgroups without psychiatric comorbidity, with non-OCPD Axis I/II diagnoses, and with OCPD. Welchs t-tests (FDR-corrected) and Hedges g quantified neuropsychological and alpha-band coherence differences. Surface-based analyses assessed cortical thickness/surface area. Exploratory regressions tested associations of OCPD, seizure freedom, and antiseizure medication (ASM) load with cognition; Kendalls tau tested coherence-cognition associations. ResultsCompared with HC, JME showed broad executive-attentional impairment, most pronounced in patients with psychiatric comorbidity. The OCPD subgroup exhibited particularly large slowing in psychomotor speed, inhibition (reaction time), and tonic alertness versus HC, while OCPD versus non-OCPD JME differences did not survive multiple-comparison correction. qEEG showed increased interhemispheric frontal and decreased temporal alpha coherence in JME, with temporal hypo-coherence strongest in those with psychiatric comorbidity; within JME, OCPD was linked to increased left fronto-temporal alpha coherence. In the MRI subsample, JME-OCPD demonstrated increased cortical thickness in left medial orbitofrontal and anterior cingulate regions (vs HC and vs JME without OCPD) and additional posterior occipito-temporal clusters versus HC. Regression and coherence-cognition associations were weak and non-significant after FDR correction. SignificanceJME features syndrome-level executive-attentional dysfunction and altered fronto-temporal network organization. Comorbid OCPD marks a subgroup with accentuated cognitive slowing and distinct medial prefrontal/cingulate structural and left fronto-temporal connectivity signatures, aligning with predictive-processing accounts of rigid, over-precise high-level priors. Key pointsJME is linked to broad executive-attentional impairment versus healthy controls. Psychiatric comorbidity amplifies cognitive deficits in JME. JME with OCPD shows particularly large slowing/inhibitory-control deficits versus controls, while OCPD vs non-OCPD differences within JME are modest. Alpha-band EEG coherence indicates altered network organization in JME and an OCPD-related increase in left fronto-temporal coherence within JME Surface-based MRI suggests an OCPD-related structural phenotype in JME, involving medial orbitofrontal/anterior cingulate cortical thickening

BackgroundCannabis use is highly prevalent among people who use unregulated drugs. While daily cannabis use has been hypothesized to provide protective effects through substitution or tolerance mechanisms, the relationship between cannabis use frequency and overdose risk remains poorly understood, particularly for infrequent users. MethodsWe conducted a secondary analysis of cross-sectional interview data from people who use unregulated drugs in Vancouver, British Columbia, collected during the fentanyl crisis (November 2019-July 2021; n=657). Binary logistic regression examined associations between self-reported cannabis use frequency (five categories: less than monthly, 1-3 times per month, weekly, more than weekly and daily) and non-fatal overdose in the preceding six months. Daily use served as the reference category. Models adjusted for age, gender, ethnicity, homelessness, mental health, HIV status, incarceration and daily use of alcohol, opioids, fentanyl, cocaine and stimulants. ResultsAmong 657 participants, 95 (14.5%) reported non-fatal overdose in the past six months. In adjusted models with daily cannabis use as the reference, infrequent cannabis use was associated with significantly increased odds of overdose: use 1-3 times per month (aOR=3.17, 95% CI: 1.50-6.69, p=.002) and more than weekly use (aOR=3.13, 95% CI: 1.70-5.76, p<.001) showed approximately three-fold increased odds compared to daily use. Less frequent use showed non-significant trends in the same direction (less than monthly: aOR=1.73, 95% CI: 0.89-3.37, p=.109; weekly: aOR=1.44, 95% CI: 0.59-3.51, p=.421). Sensitivity analysis restricted to participants with daily stimulant or fentanyl use (n=148) revealed even stronger associations. ConclusionsInfrequent cannabis use was associated with substantially increased overdose risk compared to daily use. This frequency-dependent relationship, with infrequent users at highest risk, likely reflects tolerance differences: infrequent users lack tolerance to synergistic cannabis-opioid effects. These findings were completely obscured in preliminary analyses that dichotomized cannabis use as daily versus less-than-daily, demonstrating how analytical choices can mask critical public health insights. Current harm reduction approaches, including cannabis distribution programs, should incorporate frequency-dependent risk communication and develop strategies to protect infrequent users who may be at heightened overdose risk.

BackgroundEEG microstates provide a window into rapid, large-scale brain network dynamics. Despite showing alterations in schizophrenia, evidence in first-episode schizophrenia spectrum psychosis (FESSP) is limited. We assessed whether microstate temporal and transition features could identify a multivariate signature of FESSP, and whether these dynamics can track symptom severity. MethodsResting-state EEG was analysed in 69 participants (FESSP n=41, mean age: 22.49 years; healthy controls n=28, mean age: 21.33 years). Twenty-eight microstate temporal and transition features were extracted across microstate classes (A-D). Group classification accuracy was assessed using a linear support vector machine with stratified cross-validation and permutation testing. Within the FESSP group, we further assessed associations between microstate features and clinical scores using the Brief Psychiatric Rating Scale (BPRS), Scale for the Assessment of Positive Symptoms (SAPS), and Scale for the Assessment of Negative Symptoms (SANS). ResultsMultivariate microstate features provided above-chance discrimination of FESSP from controls (balanced accuracy=0.644; AUC=0.688; p=0.030). However, when comparing individual features between groups, no feature survived multiple-comparison correction consistent with characterisation of FESSP via a distributed multivariate pattern across correlated features. Within the FESSP group, microstate dynamics were most strongly linked to negative symptoms, with higher SANS scores associated with shorter microstate D durations ({rho}=-0.507, pFDR=0.020) and higher occurrence of microstates A and B ({rho}=0.434-0.443, pFDR=0.042). BPRS-18 and SAPS showed no associations with any features. ConclusionsUsing EEG microstate temporal and transition features with multivariate classification, we identified a pattern that differentiated FESSP from controls and showed selective associations with negative symptom severity.