Eye-Movement Features And Related Brain Function Mechanism In Adolescents With Schizophrenia

Objective: This study aims to explore the eye-movement features and related brain function mechanism in adolescent schizophrenia(SCZ)patients at the early disease stage,with adolescent first-episode SCZ patients at acute phase and adolescents at Ultra High Risk for Psychosis(UHR)state as the research subjects.Methods:(1)12 to 17 years old adolescents were recruited,including patients in the first episode acute phase of SCZ(SCZ group),UHR population(UHR group),and healthy controls(HC group).Demographic information and clinical data were collected,and an eye tracker was used to collect eye-movement data on all participants completing eye-movement tasks,namely prosaccade/antisaccade and free-viewing.We compared the eye-movement performance of the three groups to find indicators with inter-group differences,and conducted correlation analysis between the eye-movement indicators and the clinical characteristics of the SCZ group.Using machine learning algorithms,we explored the effectiveness of eyemovement indicators in distinguishing adolescent SCZ patients from UHR populations and healthy controls.(2)A multimodal method of synchronously recording of functional near-infrared spectroscopy(f NIRS)and eye tracking was used to measure the activity of the frontotemporal cortex during eye-movement tasks in adolescent SCZ patients and healthy controls,and functional connectivity(FC)intensity between different frontotemporal regions were calculated to explore the functional mechanisms of the frontotemporal cortex related to the eye-movement features of adolescent SCZ patients.Results:(1)The data included in the analysis of eye-movement features were from 113 adolescent SCZ patients(SCZ group),51 UHR adolescents(UHR group)and 97 healthy controls(HC group),the demographic information of adolescents in all groups were matched.(1)Compared with the HC group,the SCZ group exhibited multiple dimensions of defects in eye-movement tasks: in the prosaccade task,the SCZ group exhibited shorter saccade amplitude,and slower average acceleration of saccade(all P<0.05);In the antisaccade task,the SCZ group had a higher error rate,longer saccade latency,shorter saccade duration and saccade amplitude,and slower saccade velocity,acceleration,and peak velocity(all P<0.01);In the free-viewing task,the SCZ group had fewer fixation number,shorter saccade amplitude and scanpath length(all P<0.01)(2)In the SCZ group,Spearman correlation analysis showed a positive correlation between the average saccade amplitude of the antisaccade task and the severity of positive symptoms(rho=0.404,P=0.005),and a negative correlation between the scanpath length during the free-viewing task and the severity of negative symptoms(rho=-0.404,P=0.009)after FDR correction.(3)The eye-movement features of the UHR group were partially similar to those of the SCZ group: compared with the HC group,the UHR group had slower saccade acceleration in the prosaccade task(P=0.002);In the antisaccade task,the UHR group had a higher error rate and longer saccade latency(whereas significantly shorter than saccade latency of the SCZ group)(all P<0.001);The fixation number,average saccade amplitude,and scanpath length in free-viewing task were all decreased(all P<0.05)(4)The machine learning model constructed using eye-movement indicators can effectively distinguish between the SCZ group and the HC group(the optimal model has a sensitivity of 0.923,a specificity of 0.826,and an accuracy of 0.857).(2)The data included in the f NIRS analysis were from 61 adolescent SCZ patients and 50 HC matched with demographic information.In adolescent SCZ patients,the FC within brain regions of frontal cortex was decreased during the antisaccade task(P<0.05),and increased during the free-viewing task(P<0.05)when compared with the HC group.There was no significant difference between the two groups during the prosaccade task.During all eye-movement tasks,there were no significant intergroup differences in FC within the temporal cortex,and between brain regions of temporal and frontal cortex.In adolescent SCZ patients,correlation was found between the FC intensity within the frontal cortex and the average saccade velocity of prosaccade,the error rate of antisaccade,and total saccade amplitude during free-viewing task.Conclusion: Both adolescent SCZ patients and adolescent UHR populations exhibited multi-dimensional eye-movement abnormalities,and eye-movement measurement may serve as an effective detection tool for adolescent-onset SCZ at the early stage of the disease.The potential brain functional mechanism involved in eye-movement features of adolescent SCZ patients is frontal cortex dysfunction,especially in the dorsolateral prefrontal cortex and frontopolar area.