Schizophrenia

Alzheimers disease psychosis (ADP) is a common and serious condition with substantial unmet need for safe and effective treatments. Pimavanserin is approved in the US to treat Parkinsons disease hallucinations and delusions. This post-hoc analysis of randomized, double-blind, placebo-controlled, phase 2 trial of nursing-home-residents with ADP evaluated the efficacy of pimavanserin by improvements (least squares mean change) in the Neuropsychiatric Inventory-Nursing Home Version Psychosis Score (NPI-NH PS). The clinical significance of the primary endpoint was assessed using responder analyses ([≥] 30% and [≥] 50%); numbers needed to treat (NNT); cumulative response (0-100% improvement); All Patients and Severe Psychosis (NPI-NH PS [≥] 12) subgroups were evaluated; improvements in hallucinations and delusions by NPI-NH-PS frequency (3 or 4 points) and severity (2 or 3 points); and [≥] 50% responder analysis at earlier timepoints (weeks 2 and 4). Among 345 patients screened, 181 patients were randomized to pimavanserin (n = 90) and placebo (n = 91). Patients were elderly (mean age: 86 years) and frail (baseline mean NPI-NH PS scores of 9.5 and 10 for pimavanserin and placebo, respectively). Pimavanserin significantly improved NPI-NH PS relative to placebo (-3.76 versus -1.93, respectively; p = 0.0451). In responder analyses, pimavanserin demonstrated a significantly greater reduction in NPI-NH PS versus placebo at both the [≥] 30% (p = 0.0159) and [≥] 50% (p = 0.0240) thresholds with NNTs of 6 and 7, respectively. Furthermore, pimavanserin demonstrated significantly earlier reductions in NPI-NH PS compared with placebo for the [≥] 30% (p = 0.0336) and [≥] 50% (p = 0.0044) thresholds. The cumulative response analysis demonstrated significantly greater efficacy of pimavanserin for All Patients (p = 0.052) and Severe Psychosis (p = 0.004), and Severe Patients versus All Patients demonstrated a greater reduction in NPI-NH PS (p = 0.011; effect size = 0.73). Pimavanserin also demonstrated numerically greater improvements for the frequency and severity of delusions and hallucinations. Responder analyses at earlier timepoints demonstrated significantly greater response rates with pimavanserin versus placebo at week 2 (p = 0.016) but not 4 (p = 0.051). These findings support pimavanserin as safe and effective in a population of nursing-home-resident patients with ADP.

VII.ObjectivesThinning of retinal layers has been documented in patients with chronic schizophrenia using standard metrics of optical coherence tomography (OCT) devices. We demonstrate the effectiveness of machine learning (ML) techniques to differentiate between schizophrenia patients and healthy controls using OCT images. MethodsFeatures extracted from a convolutional neural network (CNN) designed to segment retinal layers from OCT images represented abstracted data from the OCT images of 14 first episode (FEP) and 18 chronic schizophrenia patients, and their respective 20 and 18 age-matched controls. The abstracted data and OCT machine metrics were used separately to train support vector classification (SVC) models to differentiate between control and schizophrenia samples and test them. ResultsSVCs operating on OCT machine metrics did not classify unseen samples of FEP schizophrenia patients and controls with performance better than chance, while those looking at chronic schizophrenia did, paralleling results obtained using parametric statistics. In contrast, SVCs operating on OCT image data extracted from the CNN classified unseen samples from both populations with performance greater than chance. ConclusionThese results suggest that ML techniques can detect patterns in patients with FEP schizophrenia with greater performance using features extracted from OCT images than metrics provided by OCT machines.

BackgroundSchizophrenia is characterised by widespread neural dysconnectivity and impaired temporal coordination. Despite these pervasive disruptions, patients can maintain coherent perception and functional behaviour, particularly during remission. This paradox is underexplored in the literature, which has primarily focused on symptom emergence. We propose a dual-role hypothesis of sensorimotor temporal recalibration, a process known for adapting to temporal discrepancies in sensorimotor integration. We hypothesise that temporal recalibration may serve both as a compensatory mechanism mitigating neural incoherence and as a trigger for positive symptoms. MethodTo investigate its compensatory role, we compared 20 clinically stable schizophrenia patients to 20 matched healthy controls using a visuomotor temporal order judgement task following adaptation to varying sensorimotor asynchronies (0, 150 and 300 ms). ResultsOur findings revealed that patients with schizophrenia exhibited significantly greater temporal recalibration compared to controls. Across participants, recalibration was stronger for longer delays. No group differences were observed in just noticeable differences (JNDs) suggesting comparable task difficulty and temporal precision. ConclusionsThese findings suggest that patients not only have an intact temporal recalibration mechanism but also may engage it more to counteract neural delays and preserve temporal coherence. We further propose that overreliance on this adaptive mechanism may paradoxically contribute to symptom emergence under conditions of temporal instability by distorting the perceived order of actions and their sensory consequences. This dual-role hypothesis offers a novel perspective for understanding how the same temporal mechanism can sustain perceptual coherence, yet under certain conditions, contribute to the breakdown of causality and agency, which underlie control delusions and hallucinations.

Abnormalities in visual exploration affect the daily lives of patients with schizophrenia; however, its origin is unknown. In this study, we examined whether such abnormalities reflect aberrant processing of visual salience. Eye movements of 82 patients and 252 healthy individuals viewing natural and/or complex images were examined using saliency maps for static images to determine the contributions of low-level visual features to salience-guided eye movements. The results showed that the gazes of the participants with schizophrenia were attracted to position in the images with high orientation salience but not luminance or color salience. Further analyses revealed that orientation salience defined by the L+M channel of the DKL color space is specifically affected in schizophrenia, suggesting abnormalities in the magnocellular visual pathway. These results suggest aberrant processing of visual salience in schizophrenia, thereby connecting the dots between abnormalities in early visual processing and the aberrant salience hypothesis of psychosis.

Psychosis is frequently associated with the occurrence of visual hallucinations (VH), but their etiology remains largely unknown. While patients with psychosis show deficits on various behavioral visual and attentional tasks, previous studies have not specifically related these deficits to the presence of VH. This suggests that tasks used in these studies do not target the visual-cognitive neural mechanisms that mediate VH, which in turn limits the development of effective therapies. We therefore designed a study to target these mechanisms directly. In this case control study we asked patients with psychosis who had previously experienced VH to indicate when they recognized objects that were gradually emerging from dynamic visual noise, while scanning their brains using functional Magnetic Resonance Imaging. In a previous study, this recognition task was used to identify the neural basis of VH in patients with Parkinsons Disease. Based on this earlier work, we decided to test the following hypothesis: when compared to psychosis patients not experiencing VH and age-matched healthy controls, psychosis patients with VH show reduced occipital activity and frontal activity around the moment of recognition (known as pop-out). For all groups, neuroimaging revealed increased activity in all examined visual areas around pop-out. However, psychosis patients with VH showed reduced occipital responsiveness, especially in the inferior part of the bilateral lateral occipital complex, a region known to play a key role in object recognition. We did not observe altered frontal or prefrontal activity before pop-out in this group. A possible explanation is that the relatively sustained activation of the visual memory-related angular gyri around pop-out may have compensated for the impaired early visual processing in psychosis patients with VH. We discuss our results in terms of current theories of visual hallucinations, such as predictive coding and contextual modulation. Our study is the first to show that visual processing deficits contribute to the occurrence of VH in psychosis. These findings could be used to develop tests to identify the visual-cognitive mechanisms that mediate VH in this group.

Patients with schizophrenia show strong impairments in visual backward masking, which are associated with reduced EEG N1 responses. However, it is currently unclear whether reduced N1 amplitudes in patients reflect attenuated neural responses or increased inter-trial latency variability, as both can lead to reduced trial-averaged responses. Previous studies using trial-averaged data cannot distinguish between these two possibilities. Here, we estimated inter-trial latency variability of the visual N1 component and found significantly increased variability in patients compared to controls. Inter-trial latency-variability was a strong predictor of the N1 amplitude in both groups. Importantly, after accounting for the effect of latency variability in the group comparison, patients continued to exhibit significantly reduced N1 amplitudes, although the effect size diminished from large to medium. These findings indicate that both higher latency variability and attenuated neural responses contribute to visual processing deficits in schizophrenia.

Background and HypothesisIndividuals with schizophrenia report that to them time sometimes feels discontinuous. Previous work has shown a link between the sense of self, and the ability to prepare to react to a target and to benefit from the passage of time, at the level of a few hundreds of milliseconds. However, the sense of time continuity requires a higher time resolution than examined in previous studies. Here we investigate to which extent individuals with schizophrenia and controls benefit from an increased delay when detecting asynchronies in the order of tens of milliseconds (stimulus onset asynchronies, i.e. SOA) between two successive visual stimuli. Study DesignWe re-analyzed three datasets and contrasted performance when the SOA increases, remains identical or decreases from trial t-1 to trial t. Study ResultsPerformance of all participants improved with an increase in the SOA, as the task became easier, but for controls more so than individuals with schizophrenia. These results are replicated across datasets, and were specific to the condition when SOA increased from one trial to the next. There was no significant group difference when the SOA decreased or remained identical. This was true even when the latter condition was more frequent, and when SOA magnitude was equalized across the conditions. Importantly, the difference between the two groups was specific to temporal judgements, and was not observed in a control masking task. ConclusionsWe suggest the results reveal a difficulty for individuals with schizophrenia to experience time at the level of milliseconds.

ObjectiveTo describe the clinical outcomes related to the introduction of paliperidone palmitate in a specialty hospital in the Philippines DesignCross-sectional study among patients seen at the psychiatry service of a specialty hospital catering to veterans who were initiated on paliperidone palmitate. We reviewed and abstracted baseline patient data from the medical record of eligible patients. Outcome of treatment was collected through a one-time objective assessment of the patient by a third-party psychiatrist using the Structured Clinical Interview for Symptoms of Remission (SCI-SR) tool. Main ResultA total of 30 patients were recruited for the study from August 2020 and June 2021, the majority of whom were males (80%), residents of the National Capital Region (50%), and single (20%). The median duration from schizophrenia diagnosis to initiation of paliperidone treatment was 19.50 years (IQR: 16.60 - 33.50). In eight patients (22.67%), other antipsychotic drugs were discontinued following initiation of paliperidone treatment; in the remaining 22 participants (73.33%), paliperidone was taken concurrently with other antipsychotic drugs. The median duration from the initiation of paliperidone treatment to follow-up assessment was 27.20 months (IQR: 24.73 - 30.50), with all participants having at least 6 months of treatment. At follow-up assessment, all participants were classified to be in remission. ConclusionIn this study among patients with schizophrenia seen in a specialty hospital in the Philippines, we found evidence that clinical outcomes with paliperidone palmitate were comparable to those given a combination of oral and long-acting antipsychotics.

Cognitive impairment is a central characteristic of schizophrenia. Executive functioning (EF) impairments are often seen in mental disorders, particularly schizophrenia, where they relate to adverse outcomes. As a heterogeneous construct, how specifically each dimension of EF to characterize the diagnostic and prognostic aspects of schizophrenia remains opaque. We used classification models with a stacking approach on systematically measured EFs to discriminate 195 patients with schizophrenia from healthy individuals. Baseline EF measurements were moreover employed to predict symptomatically remitted or non-remitted prognostic subgroups. EF feature importance was determined at the group-level and the ensuing individual importance scores were associated with four symptom dimensions. EF assessments of inhibitory control (interference and response inhibitions), followed by working memory, evidently predicted schizophrenia diagnosis (area under the curve [AUC]=0.87) and remission status (AUC=0.81). The models highlighted the importance of interference inhibition or working memory updating in accurately identifying individuals with schizophrenia or those in remission. These identified patients had high-level negative symptoms at baseline and those who remitted showed milder cognitive symptoms at follow-up, without differences in baseline EF or symptom severity compared to non-remitted patients. Our work indicates that impairments in specific EF dimensions in schizophrenia are differentially linked to individual symptom-load and prognostic outcomes. Thus, assessments and models based on EF may be a promising tool that can aid in the clinical evaluation of this disorder.

Schizophrenia is associated with cognitive impairments, including deficits in implicit learning. Probabilistic serial reaction time tasks (SRTT) offer an objective approach to characterizing these deficits through both motor execution (ME) and motor imagery (MI), the mental simulation of movement without physical action. Whether implicit probabilistic sequence learning is impaired across both modalities in schizophrenia remains poorly understood. Thirty individuals with schizophrenia (ME: n=12; MI: n=13) and 40 healthy controls (ME: n=20; MI: n=20) completed an auditory probabilistic SRTT. Symptom severity was assessed with the PANSS and cognitive functioning with the MCCB. Healthy controls demonstrated a robust signature of implicit probabilistic sequence learning, whereas participants with schizophrenia exhibited weaker and less consistent learning signatures, particularly during motor imagery. Sensitivity to probabilistic structure differed significantly between groups during motor execution but not motor imagery. Participants with schizophrenia also showed significantly longer reaction times than controls across both modalities, consistent with generalized psychomotor slowing. Greater PANSS-General severity was associated with greater deviation from the probabilistic learning patterns observed in healthy controls during ME, whereas higher MCCB verbal learning scores were associated with greater similarity to these learning patterns during MI. These findings indicate that implicit probabilistic sequence learning is impaired in schizophrenia across both motor execution and motor imagery, and that these deficits are meaningfully associated with clinical symptom severity and cognitive functioning.

Perceptual distortions are core features of psychosis. Weakened surround suppression has been proposed as a neural mechanism of such atypical perceptual experiences. While previous work has measured suppression by asking participants to report the perceived contrast of a low-contrast target surrounded by a high-contrast surround, it is possible to modulate perceived contrast solely by manipulating the orientation of a matched-contrast center and surround. Removing the bottom-up segmentation cue of contrast difference and isolating the orientation-dependent suppression may clarify the neural processes responsible for atypical surround suppression in psychosis. We examined surround suppression across a spectrum of psychotic psychopathology including people with schizophrenia (PSZ; N=31) and bipolar disorder (PBD; N=29), first-degree biological relatives of these patient groups (PBDrel, PSZrel; N=28, N=21, respectively), and healthy controls (N=29). Surround suppression deficits in PSZ, while observable under many stimulus conditions, were absent under the condition that produced the strongest suppression. PBD and PSZrel exhibited intermediate suppression, while PBDrel performed most similarly to controls. Intriguingly, group differences in surround suppression magnitude were moderated by visual acuity. We propose a potential model by which visual acuity and/or focal attention interact with untuned gain control that reproduces the observed pattern of results including the lack of group differences when orientation of center and surround are the same. Our findings further elucidate perceptual mechanisms of impaired center-surround processing in psychosis and provide insights into the effects of visual acuity on orientation-dependent suppression in PSZ.

Background and HypothesisAntipsychotic medications are the first-line treatment for schizophrenia. However, around 40% of people with schizophrenia who are treated with antipsychotics could develop extrapyramidal side-effects (EPSE) including: 1) Dyskinesias, 2) Parkinsonism, 3) Akathisia, and 4) Dystonia. Study DesignWe conducted Genome-wide association (GWAS) and Epigenome-wide association (EWAS) meta-analysis of EPSE utilising data from previous schizophrenia case control studies. We integrated significant EWAS findings to an EPSE GWAS meta-analysis to enhance our understanding of the functional impact of common variants on EPSE. We also investigated whether polygenic risk scores (PRS) for schizophrenia, Parkinsons disease, and Lewy-body dementia could be predictive of EPSE development. Study ResultsThe top index SNP rs2709733 (A/G) from EPSE GWAS (p=2.214x10-7) mapped to a long intergenic non-protein coding RNA, LINC01162 with consistent effects across all cohorts. We identified 9 differentially methylated positions (DMPs) associated with EPSE when controlling for methylation age, sex, derived estimates of cell composition, smoking score, and schizophrenia PRS. Four of the DMPs cg14531564, cg20647656, cg12004641, cg22845912, and their affiliated genes (SDF4, ANKMY1, TNS1, SLA) were associated with the risk of developing EPSE and not with schizophrenia risk. Another DMP (cg12044923) which mapped to the STK32B gene, showed significant enrichment for association with risk of EPSE. ConclusionsOur study sheds new light on the potential biological mechanisms underlying EPSE development in schizophrenia, highlighting the importance of exploring both methylation shifts and common SNP associations. Further research with larger samples sizes and a focus on the role of STK32B are encouraged.

Background and HypothesisAltered cortical folding is a putative marker of neurodevelopmental disruption in schizophrenia spectrum disorders (SSD). Patients with SSD have been hypothesized to exhibit an accelerated decline in age-related cortical folding, quantified with the Local Gyrification Index (LGI). Here, we assessed longitudinal and cross-sectional LGI differences in patients with chronic SSD relative to healthy controls across 13 years. Study DesignThe sample comprised patients with SSD (mean baseline age=41.28 years) and healthy controls (mean baseline age=41.56 years) with MRI acquisitions at baseline (103 SSD; 99 controls) and follow-up after 5 (50 SSD; 57 controls) and 13 years (42 SSD; 60 controls). T1-weighted images were processed with the longitudinal pipeline in FreeSurfer. Spatio-temporal linear mixed-effects models were used to test for longitudinal and cross-sectional case-control differences in LGI, as well as the impact of symptom severity and antipsychotic medication dose among patients. Study ResultsAlthough cross-sectional LGI was lower in patients in extensive frontal, parietal, and occipital regions, we observed no significant differences in longitudinal trajectories between patients and controls after FDR correction. Medication dose was linked cross-sectionally to lower LGI of the anterior cingulate and orbitofrontal cortex, and the postcentral gyrus. ConclusionIn the longest longitudinal study on cortical folding in SSD to date, we found no accelerated progressive decline in cortical folding in patients. Thus, chronic SSD appears to be characterized by a state of stable hypogyria relative to healthy controls, supporting the interpretation of LGI as a marker of neurodevelopmental or early-life disturbances in SSD.

BackgroundUsing machine learning methods based on neurocognitive deficits and neuroimmune biomarkers, two distinct classes were discovered within schizophrenia patient samples. The first, major neurocognitive psychosis (MNP) was characterized by cognitive deficits in executive functions and memory, higher prevalence of psychomotor retardation, formal thought disorders, mannerisms, psychosis, hostility, excitation, and negative symptoms, and diverse neuroimmune aberrations. Simple neurocognitive psychosis (SNP) was the less severe phenotype. AimsThe study comprised a sample of 40 healthy controls and 90 individuals diagnosed with schizophrenia, divided into MNP and SNP based on previously determined criteria. Soft Independent Modelling of Class Analogy (SIMCA) was performed using neurocognitive test results and measurements of serum M1 macrophage cytokines, IL-17, IL-21, IL-22, and IL-23 as discriminatory/modelling variables. The model-to-model distances between controls and MNP+SNP and between MNP and SNP were computed, and the top discriminatory variables were established. ResultsA notable SIMCA distance of 146.1682 was observed between MNP+SNP and the control group; the top-3 discriminatory variables were lowered motor speed, an activated T helper-17 axis, and lowered working memory. This study successfully differentiated MNP from SNP yielding a SIMCA distance of 19.3. M1 macrophage activation, lowered verbal fluency, and executive functions were the prominent features of MNP versus SNP. DiscussionBased on neurocognitive assessments and the immune-linked neurotoxic IL-6/IL-23/Th-17 axis, we found that MNP and SNP are qualitatively distinct classes. Future biomarker research should always examine biomarkers in the MNP versus SNP phenotypes, rather than in the combined MNP + SNP or schizophrenia group.

Background and HypothesisIdentifying schizophrenia spectrum disorders (SSD) from spontaneous speech features is a key focus in computational psychiatry today. Study DesignWe present a task-voting procedure using different speech-elicitation tasks to predict SSD in Spanish, followed by ablation studies highlighting the roles of specific tasks and feature domains. Speech from five tasks was recorded from 92 subjects (49 with SSD and 41 controls). A total of 319 features were automatically extracted, from which 24 were pre-selected based on between-feature correlations and ANOVA F-values, covering acoustic-prosody, morphosyntax, and semantic similarity metrics. Study ResultsExtraTrees-based classification using these features yielded an accuracy of 0.840 on hold-out data. Ablating picture descriptions impaired performance most, followed by story reading, retelling, and free speech. Removing morphosyntactic measures impaired performance most, followed by acoustic and semantic measures. Mixed-effect models suggested significant group differences on all 24 features. In SSD, speech patterns were slower and more variable temporally, while variations in pitch, amplitude, and sound intensity decreased. Semantic similarity between speech and prompts decreased, while minimal distances from embedding centroids to each word increased, and word-to-word similarity arrays became more predictable, all replicating patterns documented in other languages. Morphosyntactically, SSD patients used more first-person pronouns together with less third-person pronouns, and more punctuations and negations. Semantic metrics correlated with a range of positive symptoms, and multiple acoustic-prosodic features with negative symptoms. ConclusionsThis study highlights the importance of combining different speech tasks and features for SSD detection, and validates previously found patterns in psychosis for Spanish.

Background and HypothesisMachine Learning (ML) models have been argued to reliably predict diagnosis and symptoms of schizophrenia based on voice data only. However, it is unclear to what extent such ML markers would generalize to different clinical samples and different languages, a crucial assessment to move towards clinical applicability. In this study, we systematically assessed the generalizability of ML models of vocal markers of schizophrenia across contexts and languages. Study DesignWe trained models relying on a large cross-linguistic dataset (Danish, German, Chinese) of 217 patients with schizophrenia and 221 controls, and used a conservative pipeline to minimize overfitting. We tested the models generalizability on: (i) new participants, speaking the same language; (ii) new participants, speaking a different language; (iii) further, we assessed whether training on data with multiple languages would improve generalizability using Mixture of Expert (MoE) and multilingual models. ResultsModel performance was comparable to state-of-the-art findings (F1-score [~] 0.75) within the same language; however, models did not generalize well - showing a substantial decrease - when tested on new languages. The performance of MoE and multilingual models was also generally low (F1-score [~] 0.50). ConclusionsOverall, the cross-linguistic generalizability of vocal markers of schizophrenia is limited. We argue that more emphasis should be placed on collecting large open cross- linguistic datasets to systematically test the generalizability of voice-based ML models, and on identifying more precise mechanisms of how the clinical features of schizophrenia are expressed in language and voice, and how different languages vary in that expression.

Medication resistance in psychotic disorders represents a critical challenge in forensic psychiatry, where up to 50% of patients show poor treatment response, leading to increased risk of relapse, violence, and rehospitalization. Feature Transfer, a novel machine learning framework based on rank aggregated feature selection, transfers predictive features identified for one outcome to related outcomes while maintaining clinical interpretability, a critical advantage over conventional transfer learning approaches that obscure feature level insights by transferring complex model parameters. Applied to psychotic disorders, this methodology identified key predictors for medication resistance and assessed their transferability to related clinical outcomes. Analyzing data from 893 patients across 11 forensic psychiatric institutions, we compared Feature Transfer models (using the top 25 features discriminating medication resistance from responders) with full feature models (95 features) for predicting clinical relapse, treatment non adherence, and escape behaviors. In the broader psychotic disorders sample, Feature Transfer achieved statistically equivalent performance to full feature models for clinical relapse and treatment non adherence (F1 score differences with confidence intervals overlapping zero), though performed less effectively for escape behaviors (AUC: 0.736 vs 0.838). In schizophrenia patients (n=634), Feature Transfer showed statistically significant improvement in F1 score for clinical relapse prediction compared to full feature models (difference: 0.119, 95% CI: 0.025 to 0.213), with notably higher sensitivity (0.912 vs 0.802) while maintaining comparable discriminative ability (AUC: 0.912 vs 0.925, difference not statistically significant). Treatment history features, particularly previous medication unresponsiveness and duration of clinical care, maintained high predictive importance across multiple clinical outcomes (relapse, non adherence, and escape behaviors), suggesting they represent fundamental risk indicators regardless of the specific outcome being predicted. While our retrospective design limits causal inference and relies on historical indicators as proxies for secondary outcomes (relapse and escape behaviors), the demonstrated utility of medication resistance features across different clinical outcomes reveals potential shared risk dimensions in psychotic disorders, particularly for relapse prediction in schizophrenia. Feature Transfer offers a transparent approach for identifying common predictive factors that could advance personalized intervention strategies in complex psychiatric populations.

Background and HypothesisNiacin skin bluntness is a promising biomarker for schizophrenia, especially in precision medicine, as it helps identify a distinct subgroup (around 1/3) of patients with schizophrenia who experience severe functional impairment. However, research has not clarified if this phenotype is just a byproduct of the onset of the disease or is involved in the etiology of schizophrenia. We hypothesize that niacin bluntness reflects causal alterations in schizophrenia. Study DesignWe firstly conducted a quasi-genome-wide association analysis within schizophrenia patients to identify instrumental SNPs for niacin response. Then a two-sample bi-directional Mendelian randomization (MR) analysis was implemented to estimate the potential causal effects between niacin response and schizophrenia. Study Results25 independent SNPs showed a trend of genome-wide significant association (p<E-05) with niacin response in schizophrenia. In the MR analysis, the F statistics of the two instrumental SNPs for niacin response is 30.77, indicating a proper strength. As heterogeneity of the SNPs was detected (p<0.05), the result of inverse variance weighted method for multiplicative random effects was adopted for evaluation of the detected causal effect (OR=1.272, p=9.46E-03). In reverse MR analysis, none of the methods supported a causal effect of schizophrenia on niacin response (all p>0.05). ConclusionsOur results revealed that the attenuated niacin response caused schizophrenia but not vice versa. Etiological studies on niacin bluntness in schizophrenia are needed and will pave the way for biomarker-guided personalized treatment in future.

STRUCTURED ABSTRACTO_ST_ABSBackground and HypothesisC_ST_ABSCannabis use has been linked to psychotic-like experiences (PLEs). Amid increasing legalization, we examined the extent to which cannabis use is associated with PLEs after adjusting for other risk factors in a contemporary United States sample. Study DesignWe performed a cross-sectional analysis of self-reported cannabis use and four types of self-reported PLEs (auditory and visual perceptual distortions, referential ideation, and persecutory ideation) in the population-based biobank, the All of Us Research Program release 8 (maximum analytic N = 62,153). Study ResultsCannabis ever-use (ORs = 1.21 - 1.44, p-values < 2.7e-6) and more frequent past 3-month cannabis use (within lifetime ever-users) were associated with all four PLEs ({chi}2(4) = 21.06 - 70.09, p-values = 3.08e-4 to 2.17e-14), and these associations remained when adjusting for personal and family history of schizophrenia and polygenic liability for schizophrenia. The schizophrenia polygenic score, but not cannabis use frequency, was correlated with greater likelihood of being prescribed medication for the PLEs. When adjusting for lifetime ever-use of other substances, cannabis ever-use was no longer associated with PLEs, while methamphetamine use, cigarette use, and opioid use were associated with PLEs (ORs = 1.22 to 1.65, p-values < 1.68e-05). ConclusionsPrior associations between cannabis use and PLEs may have been confounded by comorbid use of other substances. Future studies that distinguish cannabis use from other substance use in the etiology of PLEs could provide insight into this transdiagnostic construct.

Machine learning approaches may support individual-level classification in psychiatry, but many EEG-based schizophrenia studies have relied on small samples or conventional summary features. We evaluated whether trial-level auditory oddball event-related potential (ERP) waveforms could support schizophrenia versus healthy-control classification using deep learning. The study included 258 patients with schizophrenia and 142 healthy controls. EEG recordings from an auditory duration oddball paradigm were segmented into -100 to 500 ms epochs, and trial-level mismatch waveforms were generated by subtracting each participant's mean standard response from accepted deviant trials. Models were trained using a fixed participant-level training, validation, and test split, with demographic residualization fit only in the training set. Five deep learning architectures were trained on full residualized ERP waveforms and compared with classical machine learning models trained on 18 conventional ERP summary features. Deep learning models achieved higher test set discrimination than classical feature-based models, with AUROC values ranging from 0.797 to 0.857 versus 0.705 to 0.720. Benchmark analyses suggested that performance depended on the combination of waveform-level input and deep learning architecture. These findings support trial-level auditory oddball ERP waveforms as promising classification inputs and candidate electrophysiological biomarkers of schizophrenia-related neural information processing.