BACKGROUND:Anxiety disorders are highly prevalent yet lack objective biomarkers. Whereas threat-related attentional biases are well documented, less is known about broader eye movement alterations that may characterise anxiety. AIMS:To characterise multi-paradigm eye movement profiles in anxiety disorders and evaluate their potential as behavioural markers for disorder differentiation. METHOD:Eye movements were recorded in 91 patients with anxiety disorders, 118 with depressive disorders and 98 healthy controls during free viewing of neutral-stimuli, smooth-pursuit and fixation-stability tasks. Principal component analysis was applied to derive latent eye movement dimensions, which were then tested for group differences, associations with symptom severity and classification performance. RESULTS:Compared with both patients with depression and healthy controls, patients with anxiety disorders exhibited hyper-scanning during free viewing, characterised by increased saccade frequency and path length, and hyper-pursuit during smooth pursuit, reflected in increased velocity gain, fewer intrusive saccades and more catch-up saccades. Principal component analysis identified six latent components, among which active visual exploration, pupillary arousal and smooth-pursuit control demonstrated robust group differences. Machine learning models trained on 6 components yielded areas under the receiver operating characteristic curve of 0.82 for anxiety versus healthy controls, 0.83 for depression versus healthy controls and 0.61 for anxiety versus depression. CONCLUSIONS:Hyper-scanning and hyper-pursuit emerge as defining eye movement signatures of anxiety, linking core mechanisms of vigilance and prediction with measurable behavioural markers. These insights position eye-tracking as a promising behavioural modality for mechanism-informed differentiation across affective disorders.

Although eye movement abnormalities are documented in schizophrenia (SZ), their translation into objective diagnostic biomarkers remains limited. In this study, we propose a novel identification framework that integrates a Sparsity-Scoring Kernel Entropy Component Analysis (SSKECA) algorithm with a multidimensional eye movement feature set. A total of 40 patients with SZ and 50 healthy controls (HC) completed a free-viewing task involving 100 distinct semantic images. The proposed SSKECA algorithm optimizes multidimensional feature representations to capture latent eye movement patterns characteristic of SZ. The SSKECA-AdaBoost model achieved competitive performance, with an accuracy of 0.933 and an area under the receiver operating characteristic curve (AUC) of 0.960. Notably, when restricted to only 25 highly discriminative images, the SSKECA-XGBoost model achieved an accuracy of 0.922. Feature ablation analyses not only reproduced previously reported eye movement findings but also highlighted additional atypical patterns. Misclassification analyses revealed more pronounced eye movement deficits in incorrectly classified SZ patients. Overall, the proposed framework translates complex eye movement patterns into robust indicators for subject-level identification, offering a practical and efficient tool to support objective assessment in SZ.

BACKGROUND Impaired insight is a core feature of schizophrenia and an established predictor of psychosis in individuals at clinical high risk (CHR). However, the neurocognitive mechanisms underlying impaired insight in CHR populations remain unclear. AIM To investigate the role of neurocognitive deficits in impaired insight and examine their combined influence on psychosis risk over a six-year follow-up. METHODS A total of 312 CHR individuals were assessed for insight using the G12 item of the Positive and Negative Syndrome Scale. Participants were categorized into the low-impairment insight group (n = 151, G12 score < 3) and the high-impairment insight group (n = 161, G12 score >= 3). Neurocognition was evaluated using the MATRICS Consensus Cognitive Battery. RESULTS Performance on the Brief Visuospatial Memory Test-Revised (BVMT-R) was the only cognitive domain differentiating the insight groups after controlling for positive symptoms. The effect of BVMT-R on insight was most pronounced at moderate symptom levels. Moreover, risk-curve analyses indicated that higher BVMT-R scores were associated with a reduced conversion risk linked to impaired insight. CONCLUSION Incorporating visuospatial memory assessment may improve the identification of those at greatest risk and inform targeted interventions aimed at enhancing insight and reducing conversion to psychosis.

BACKGROUND:Abnormalities in event-related potential P300 have been reported in clinical high-risk (CHR) populations, reflecting deficits in cognitive processes. Studies suggest that task-induced changes in the exponent are associated with working memory capacity. The present study aimed to compare visual P300 amplitude and task-related exponent modulation during a visual oddball paradigm among CHR individuals, and to examine whether the exponent could serve as potential biomarker for clinical outcomes. METHODS:CHR and healthy controls (HC) were enrolled and completed a visual oddball task while EEG was recorded. CHR participants were divided into psychosis converters (CHRC) and non-converters (CHR-NC) group according to the clinical outcomes of two-year follow-up. P300 amplitude and the pre- to post-stimulus change in aperiodic exponent (Δ-exponent) were analyzed. Logistic regression modeling was performed to evaluate the predictive power of Δ-exponent. RESULTS:P300 amplitude at parietal region was significantly lower in the CHR-C group compared to the HC (p = 0.01) group. The Δ-exponent was significantly attenuated in CHR-C relative to HC (p < 0.01) and CHR-NC (p = 0.01). Logistic regression showed that Δ-exponent may have better predictive power than P300 amplitude (mean AUC difference 0.13, 95%CI: 0.002-0.38), and a reduction in Δ-exponent was associated with psychosis conversion (OR = 0.43(0.23-0.75), p < 0.01). CONCLUSION:Attenuation of the visual oddball task-induced change in aperiodic exponent is observed in CHR individuals who later convert to psychosis and may hold better predictive power compared to P300 amplitude. Δ-exponent may serve as a potential neurobiological biomarker in the CHR population.

BACKGROUND:Deficits in visual attention and emotional processing are core to schizophrenia (SZ). Although oculomotor abnormalities during static viewing are well documented, it remains unclear whether they persist or differ by valence under dynamic emotional stimuli that better mimic real-world social situations. Identifying emotion-related differences in oculomotor responses may help characterize candidate behavioral markers in SZ. METHODS:We enrolled 104 SZ patients and 104 demographically matched healthy controls (HCs). All participants viewed stimuli within the Dynamic Emotional Video Paradigm (DEVP), from which 14 oculomotor metrics were extracted. Using two feature selection strategies and nested 5-fold cross-validation (CV), we compared all-emotional and emotion-specific machine learning models for SZ classification. RESULTS:SZ patients showed oculomotor abnormalities across all-emotional conditions, including altered fixation duration, changed saccade patterns, and decreased pupil responsiveness. These findings extend existing static-task evidence to more ecologically valid dynamic conditions. Oculomotor impairments varied by valence, with relatively more evident changes under sad stimuli. Random forest (RF) selection identified pupil dynamics and fixation duration as key features, whereas correlation-variance filtering retained a wider range of oculomotor indices. The sad-specific model achieved accuracy of 0.79 and AUC of 0.86, with a marginal advantage over global models. CONCLUSION:Our findings highlight that oculomotor impairments in SZ exhibit clear valence specificity under dynamic emotional stimuli. Oculomotor metrics measured under the sad emotion condition show favorable biomarker potential for supporting auxiliary early screening of SZ.

Clinical subtypes of individuals at clinical high risk for psychosis (CHR), classified by baseline symptoms and cognitive performance, may follow distinct trajectories of symptom progression and functional outcomes. Baseline symptom, cognitive and EEG data were collected from 204 CHR individuals aged 13-38 years, who subsequently completed clinical assessments at 2 months, 1 year, and 2 years. Although baseline positive symptom scores and highest past-year global functioning did not differ significantly, the subtypes demonstrated divergent progression patterns across follow-ups. Baseline cognitive performance exerted sustained influence over subsequent assessments, with visuospatial learning and working memory showing significant long-term effects on positive symptoms or global functioning, highlighting them as core cognitive domains with longitudinal, cross-subtype impact on CHR progression. Brain network features also contributed to long-term outcomes. The involvement of microstate D was sustained over time, with its occurrence linked to functional outcomes in the most impaired subtype and its coverage associated with positive symptom trajectories in the moderately impaired subtype. We further constructed a predictive model incorporating the interaction between microstate D and subtype that effectively identified individuals who converted to psychosis, underscoring its potential as a dynamic biomarker for more precise CHR clinical profiles.

Background and Hypothesis Antipsychotics (APs) are often used among individuals with clinical high risk (CHR) for psychosis and affect cortical thickness (CT). Whether clinical and CT changes after initial AP use correlate with long-term clinical outcomes remains largely unknown. Study Design One hundred and thirty-eight CHRs and 65 healthy controls accepted 2 MRI scans at an interval of 2 months. CHRs were categorized as responders (n = 53) and non-responders (n = 69) based on their response to APs after 2-month treatment. According to 2-year outcomes, they were also subdivided into converters (n = 26) and non-converters (n = 96). The relationships among short-term CT changes, AP effects, and long-term outcomes were explored. Study Results At baseline, CHRs had CT reduction in the right inferior temporal cortex with a correlation with clinical symptoms. At 2 month, CHRs showed steeper gray matter loss in bilateral frontotemporal regions than healthy controls. Cortical thickness change rates of the clusters located in bilateral middle temporal and right lateral orbitofrontal cortex were negatively correlated with the cumulative AP dose. Furthermore, 2-year psychosis conversion rate was significantly higher in non-responders than responders (33.3% vs 5.1%). A random forest model based on demographic, clinical, baseline, and longitudinal CT variables predicted 2-year conversion with an AUC of 0.90 (accuracy: 0.83, sensitivity: 0.78, and specificity: 0.89), with model predictive power driven primarily by symptom and CT variables. Conclusions These findings contribute valuable insights into the potential impact of early AP treatment on brain morphology and clinical trajectories and highlight the importance of monitoring the initial treatment responses.

BACKGROUND:Cognitive impairment is common in depression, yet most studies examine cognitive domains separately and rarely characterize patterns of cognitive coupling at the behavioral network level. This study investigated cognitive network architecture in first-episode depression and examined patterns of altered cognitive coupling among cognitive measures. METHODS:201 First-episode depression patients and matched healthy controls completed MCCB subtests (excluding social cognition). Group differences in cognitive performance were examined, and symptom-cognition associations were tested using Spearman correlations. Cognitive networks were constructed from pairwise correlation matrices. Global cognitive coupling was quantified as the mean absolute value of pairwise correlations among cognitive measures, and edge-wise differences were further examined. RESULTS:Patients showed poorer performance in processing speed and verbal/visual memory, while attention/vigilance was relatively preserved. Greater depressive severity was associated with poorer performance in several cognitive measures, particularly processing speed, working memory, and attention/vigilance; anxiety severity was primarily associated with attention/vigilance measures, while psychotic symptoms showed domain-specific associations. Network analysis revealed higher global cognitive coupling in patients than controls (mean absolute correlation = 0.358 vs. 0.251), with increased pairwise coupling involving processing-speed and memory-related measures. CONCLUSIONS:These findings suggest that altered cognitive coupling may reflect reduced differentiation among cognitive processes at the behavioral network level and highlight cognitive network analysis as a complementary framework for understanding early-stage cognitive dysfunction in depression.

Background and Hypothesis Given that fixation stability is closely linked to cognition, we investigated fixation stability in patients at different stages of schizophrenia, its relationship with cognitive impairments, and its predictive role for conversion to psychosis. Study Design Fixation stability was measured by bivariate contour ellipse area (BCEA), and cognition was assessed by the Measurement and Treatment Research to Improve Cognition in Schizophrenia Consensus Cognitive Battery for 75 patients with first-episode schizophrenia (FES), 75 patients with clinical high-risk (CHR) syndrome, and 75 healthy controls (HCs). After a 1-year longitudinal study, CHR follow-up outcomes were classified as CHR-converters and CHR-nonconverters. Diagnostic model for clinical stages and prediction model for conversion were constructed using logistic regression and Cox regression, respectively. Study Results Patients exhibited fixation instability and cognition impairments compared to HC, with impairments increasing from CHR to FES. In CHR, BCEA negatively correlated with Measurement and Treatment Research to Improve Cognition in Schizophrenia Consensus Cognitive Battery scores, but this correlation was absent in FES. Diagnostic model effectively discriminated HC and FES, with an area under the receiver-operating characteristic curve of 0.914. Among 66 CHR followed up for 1 year, 13 have converted to schizophrenia, with a conversion rate of 19.70%. When divided into large and small BCEA groups (33 each), the conversion rate was 27.27% and 12.12%. Conversion prediction model achieved an area under the receiver-operating characteristic curve of 0.708. Conclusions Our results indicate that fixation instability worsens with schizophrenia progression, which is associated with cognitive impairments. Additionally, BCEA may serve as a biomarker for predicting conversion to psychosis.

Aberrant functional connectivity (FC) between the left dorsolateral prefrontal cortex (DLPFC) and subgenual anterior cingulate cortex (sgACC) is a replicated neural correlate of major depressive disorder (MDD). Emerging evidence suggests that individualized DLPFC-sgACC peak connectivity profiles may optimize transcranial magnetic stimulation (TMS) targeting and therapeutic outcomes. We hypothesized that the heterogeneity of DLPFC-sgACC peak FC locations could serve as a neurological basis for classifying distinct MDD subgroups. We recruited 120 patients with MDD and used resting-state functional magnetic resonance imaging (MRI) to identify the peak of DLPFC-sgACC FC. Using the personalized peaks’ spatial distribution, we clustered patients with MDD into subgroups and compared between-subgroup depressive and anxiety profiles. The classification performance of different clinical profiles between the subgroups was evaluated. The TMS therapeutic outcomes were retrospectively compared between these two subgroups in a small subsample of 37 patients who completed TMS treatment. The personalized DLPFC-sgACC peaks of patients with MDD were spread widely within the left DLPFC, clustering into the anterior (73.3

Individuals at clinical high risk (CHR) for psychosis exhibit both baseline and progressive brain structural abnormalities. However, the extent to which these changes reflect neurobiological trajectories of illness progression versus iatrogenic effects of antipsychotic (AP) treatment remains unresolved. A total of 148 AP-naïve CHRs and 65 healthy controls (HCs) underwent baseline structural magnetic resonance imaging (MRI) scans. One hundred thirty CHRs received second-generation AP treatment and completed 2-month follow-up scans. HCs also completed the follow-up scans. We compared baseline and longitudinal brain volume changes between CHRs and HCs and explored the relationship between AP treatment and brain structural changes in CHR. At baseline, CHRs showed enlarged third and inferior lateral ventricles compared to HCs. Within CHRs, larger ventricular, as well as smaller hippocampus and amygdala volumes, were associated with more severe symptoms and poorer functioning. No cortical volume differences were observed between groups at baseline, nor were cortical volumes related to clinical symptoms. After 2-month AP treatment, CHRs exhibited continued ventricular enlargement, reduced accumbens volume, and widespread cortical volume loss relative to HCs. Notably, cortical volume reductions were dose-dependent, with higher AP dose correlating with more pronounced cortical reductions. Additionally, cortical volume changes were linked to treatment response, with high-dose responders showing more significant HC-referenced changes compared to high-dose non-responders, low-dose responders, and low-dose non-responders. Our findings underscore the complex, region-specific, and clinically relevant neuroanatomical changes in CHR individuals, emphasizing the critical need to account for AP exposure in CHR neuroimaging studies.

Schizophrenia is a severe mental disorder characterized by abnormal eye movements. However, existing methods for detecting these abnormalities rely primarily on static stimuli that lack ecological validity. This study explores the potential of video paradigms for capturing eye movement patterns specific to schizophrenia. Forty patients with schizophrenia (SZ) and forty healthy controls (HCs) completed eye movement tests and cognitive assessments. The participants watched two types of videos: an animated video with low cognitive load and a documentary with high social cognitive load, from which ten eye movement features were extracted. Group differences were compared using analysis of covariance (ANCOVA). Support vector machine (SVM) and random forest (RF) algorithms constructed the classification models, employing leave-one-out cross-validation to optimize performance. Compared to HCs, the SZ group showed significant differences in eye movement metrics, including longer fixation durations, smaller saccade amplitudes, and a reduced pupil size ratio. Furthermore, the SZ group exhibited greater variability in fixation distribution, particularly during the documentary video. When combining eye movement features from both videos, the SVM algorithm achieved a classification accuracy of 84%, while the RF algorithm reached 80%. The video-based eye movement paradigm effectively detected abnormalities in the SZ group, highlighting its potential for objective schizophrenia detection.

Cognitive impairment plays a crucial role in the development from clinical high-risk(CHR) to schizophrenia-spectrum psychosis. Understanding its dynamic change trajectories has significant preventive value. This study aims to comprehensively explore the dynamic cognitive trajectories of clinical high-risk(CHR) individuals over a three-year period, compare the trajectories among CHR individuals who convert to psychosis(CHR-C), those who do not(CHR-NC), and healthy controls(HC). This prospective study was part of the Shanghai At Risk for Psychosis(SHARP)-extended program. CHR participants, from psychological counseling outpatient clinics and medication - naive, were followed up at baseline (T1), 2 months ± 1 week (T2), 1 year ± 1 month (T3), 2 years ± 1 month (T4), and 3 years ± 1 month (T5). All participants completed the Chinese version of the MATRICS Consensus Cognitive Battery(MCCB) at these five strictly fixed time points to assess neurocognitive function. Conversion to psychosis was determined as main outcome. A total of 43 CHR(median age, 16[IQR, 15–18] years; 14 men[32.6

BACKGROUND:Clinical high risk for psychosis (CHR) is often managed with antipsychotic medications, but their effects on neurocognitive performance and clinical outcomes remain insufficiently explored. This study investigates the association between aripiprazole and olanzapine use and cognitive and clinical outcomes in CHR individuals, compared to those receiving no antipsychotic treatment. METHODS:A retrospective analysis was conducted on 127 participants from the Shanghai At Risk for Psychosis (SHARP) cohort, categorized into three groups: aripiprazole, olanzapine, and no antipsychotic treatment. Neurocognitive performance was evaluated using the MATRICS Consensus Cognitive Battery (MCCB), while clinical symptoms were assessed through the Structured Interview for Prodromal Syndromes (SIPS) at baseline, 8 weeks, and one year. RESULTS:The non-medicated group demonstrated greater improvements in cognitive performance, clinical symptoms, and functional outcomes compared to the medicated groups. Among the antipsychotic groups, aripiprazole was associated with better visual learning outcomes than olanzapine. Improvements in neurocognition correlated significantly with clinical symptom relief and overall functional gains at follow-up assessments. CONCLUSIONS:These findings suggest potential associations between antipsychotic use and cognitive outcomes in CHR populations while recognizing that observed differences may reflect baseline illness severity rather than medication effects alone. Aripiprazole may offer specific advantages over olanzapine, underscoring the importance of individualized risk-benefit evaluations in treatment planning. Randomized controlled trials are needed to establish causality.

BACKGROUND:Individuals at clinical high risk (CHR) for psychosis exhibit reduced P300 responses, particularly in auditory tasks. While N200 abnormalities have been reported in CHR individuals, findings are inconsistent, and research on visual oddball tasks is limited. This study compares event-related potentials (ERPs) in auditory and visual oddball paradigms between CHR individuals and healthy controls (HC), aiming to explore these components as potential biomarkers for CHR and its clinical outcomes. METHODS:Baseline auditory and visual oddball N200 and P300 were obtained from CHR participants (N = 96) and HC participants (N = 60). All CHR participants were followed up for five years and stratified into CHR converters and non-converters. The differences in N200 and P300 amplitude in auditory and visual oddball tasks were compared between the clinical outcome subgroups and HC participants. Prediction models were developed using Cox proportional hazards regression to identify baseline predictors. RESULTS:CHR converters showed significantly reduced N200 and P300 amplitudes in both oddball paradigms relative to CHR non-converters and HC participants. Furthermore, Prediction models including visual N200 amplitudes (β = -0.233, p = 0.012) and auditory P300 amplitudes (β = 0.219, p = 0.039) demonstrated effective discrimination between CHR converters and non-converters. CONCLUSION:These results suggest that N200 and P300 amplitude deficits across auditory and visual modalities precede the onset of full psychosis. Moreover, the combined assessment of visual N200 and auditory P300 amplitudes may serve as a more sensitive prognostic biomarker for predicting clinical outcomes in individuals at clinical high risk.

BACKGROUND:While antipsychotic-induced eye movement alterations are well-documented in schizophrenia, their effects during the clinical high-risk (CHR) phase remain uncharacterized. This study examined the effects of two-month antipsychotic treatment on eye movement parameters in CHR individuals and their association with clinical outcome. METHODS:In this longitudinal cohort, 139 CHR individuals and 105 healthy controls completed baseline eye-tracking (fixation stability, free viewing, and smooth pursuit). CHR participants were reassessed at two months and followed for three years to track remission status. Linear mixed-effects models examined the effects of antipsychotic use, dose, and type on eye movement indicators, and a random forest model evaluated how changes in these indicators predicted clinical remission. RESULTS:In the antipsychotic - treated subgroup, fixation stability featured more microsaccades, free viewing showed reduced saccade amplitude and velocity, and smooth pursuit showed increased velocity gain with reduced saccade amplitude, and these changes scaled with dose and varied by agent with the most pronounced effects for aripiprazole. A random forest classifier using two treatment-induced eye movement change values predicted 3-year clinical non-remission with an area under the receiver operating characteristic curve of 0.80. CONCLUSIONS:Short-term antipsychotic exposure induced mixed eye movement alterations that were associated with non-remission at three-year follow-up. This finding provides a reference for the development of personalized risk stratification frameworks and targeted intervention strategies in CHR populations.

Objective Clustering individuals at the Clinical High-Risk(CHR) stage of psychosis often relies on single dimensions, and the independence or overlap of clustering results across different dimensions lacks sufficient evidence. Additionally, it remains unclear whether combining different dimensions—such as biological markers(e.g., cytokines) and symptomatic dimensions—can enhance predictive efficacy. Methods This study included 370 individuals with CHR and conducted a three-year follow-up, 50 CHR individuals transitioned to psychosis. The participants underwent thorough symptom assessments, encompassing both clinical symptoms and cognitive impairments. Baseline measurements of eight cytokines were obtained. Latent Class Analysis(LCA) was employed to construct clusters based on both symptom profiles and cytokine levels separately. Survival analysis was utilized to explore differences in conversion rates among different clusters. Results The LCA determined the selection of the four-cluster solution for symptoms, cytokines, and the integrated clusters. Symptom-Cluster-2 exhibited the most severe clinical symptoms and cognitive impairments, while Symptom-Cluster-4 displayed the mildest clinical symptoms and cognitive impairments. Cytokine-Cluster-1 was characterized by the highest levels of inflammatory cytokines, excluding vascular endothelial growth factor, whereas Symptom-Cluster-4 exhibited the lowest levels of cytokines. The clusters identified based on symptoms and cytokines showed substantial inconsistency. Survival analysis comparing conversion rates across four clusters revealed no significant difference in symptom(χ2 = 6.731, p = 0.081) and cytokine(χ2 = 7.139, p = 0.068) clusters but was significant in integrated clusters(χ2 = 9.234, p = 0.026). Conclusion The study emphasizes the distinct perspectives on psychosis risk offered by symptom and cytokine dimensions, advocating for the integration of these dimensions in a cross-modal approach to enhance predictive accuracy.

BACKGROUND: Restricted scan path mode is hypothesized to explain abnormal scanning patterns in patients with schizophrenia. Here, we calculated entropy scores (drawing on gaze data to measure the statistical randomness of eye movements) to quantify how strategical and random participants were when processing image stimuli. METHODS: Eighty-six patients with first-episode schizophrenia (FES), 124 individuals at clinical high risk (CHR) for psychosis, and 115 healthy control participants (HCs) completed an eye-tracking examination while freely viewing 35 static images (each presented for 10 seconds) and cognitive assessments. We compared group differences in the overall entropy score, as well as entropy scores under various conditions. We also investigated the correlations between entropy scores and symptoms and cognitive function. RESULTS: Increased overall entropy scores were noted in the FES and CHR groups compared with the HC group, and these differences were already apparent within 0 to 2.5 seconds. In addition, the CHR group exhibited higher entropy than the HC group when viewing low-meaning images. Moreover, the entropy within 0 to 2.5 seconds showed significant correlations with negative symptoms in the FES group, attention/vigilance scores in the CHR group, and speed of processing and attention/vigilance scores across all 3 groups. CONCLUSIONS: The results indicate that individuals with FES and those at CHR scanned pictures more randomly and less strategically than HCs. These patterns also correlated with clinical symptoms and neurocognition. The current study highlights the potential of the eye movement entropy measure as a neurophysiological marker for early psychosis.

BACKGROUND AND HYPOTHESIS:Visual fixation is a dynamic process, with the spontaneous occurrence of microsaccades and macrosaccades. These fixational saccades are sensitive to the structural and functional alterations of the cortical-subcortical-cerebellar circuit. Given that dysfunctional cortical-subcortical-cerebellar circuit contributes to cognitive and behavioral impairments in schizophrenia, we hypothesized that patients with schizophrenia would exhibit abnormal fixational saccades and these abnormalities would be associated with the clinical manifestations.STUDY DESIGN:Saccades were recorded from 140 drug-naïve patients with first-episode schizophrenia and 160 age-matched healthy controls during ten separate trials of 6-second steady fixations. Positive and negative symptoms were assessed using the Positive and Negative Syndrome Scale (PANSS). Cognition was assessed using the Measurement and Treatment Research to Improve Cognition in Schizophrenia Consensus Cognitive Battery (MCCB).STUDY RESULTS:Patients with schizophrenia exhibited fixational saccades more vertically than controls, which was reflected in more vertical saccades with angles around 90° and a greater vertical shift of horizontal saccades with angles around 0° in patients. The fixational saccades, especially horizontal saccades, showed longer durations, faster peak velocities, and larger amplitudes in patients. Furthermore, the greater vertical shift of horizontal saccades was associated with higher PANSS total and positive symptom scores in patients, and the longer duration of horizontal saccades was associated with lower MCCB neurocognitive composite, attention/vigilance, and speed of processing scores. Finally, based solely on these fixational eye movements, a K-nearest neighbors model classified patients with an accuracy of 85%. Conclusions: Our results reveal spatial and temporal abnormalities of fixational saccades and suggest fixational saccades as a promising biomarker for cognitive and positive symptoms and for diagnosis of schizophrenia.

Background and hypothesis Substantive inquiry into the predictive power of eye movement (EM) features for clinical high-risk (CHR) conversion and their longitudinal trajectories is currently sparse. This study aimed to investigate the efficiency of machine learning predictive models relying on EM indices and examine the longitudinal alterations of these indices across the temporal continuum.Study design EM assessments (fixation stability, free-viewing, and smooth pursuit tasks) were performed on 140 CHR and 98 healthy control participants at baseline, followed by a 1-year longitudinal observational study. We adopted Cox regression analysis and constructed random forest prediction models. We also employed linear mixed-effects models (LMMs) to analyze longitudinal changes of indices while stratifying by group and time.Study results Of the 123 CHR participants who underwent a 1-year clinical follow-up, 25 progressed to full-blown psychosis, while 98 remained non-converters. Compared with the non-converters, the converters exhibited prolonged fixation durations, decreased saccade amplitudes during the free-viewing task; larger saccades, and reduced velocity gain during the smooth pursuit task. Furthermore, based on 4 baseline EM measures, a random forest model classified converters and non-converters with an accuracy of 0.776 (95% CI: 0.633, 0.882). Finally, LMMs demonstrated no significant longitudinal alterations in the aforementioned indices among converters after 1 year.Conclusions Aberrant EMs may precede psychosis onset and remain stable after 1 year, and applying eye-tracking technology combined with a modeling approach could potentially aid in predicting CHRs evolution into overt psychosis.