Comparison Of Sleep Structure And Prediction Modeling In Patients With Severe Mental Disorders

Objective:In this paper,we will investigate the variability of polysomnography(PSG)among different patients with severe mental disorder(SMD),namely schizophrenia(Sch),bipolar disorder(BD)and major depressive episode(MDE),and construct a prediction model to provide a reference.Method:From March 2021 to February 2023,223 patients with Sch,100 patients with BD,and 100 patients with MDE who met the diagnostic criteria of the International Statistical Classification of Diseases and Related Health Problems 10 th Revision(ICD-10)and were seen at the First Hospital of Shanxi Medical University were collected.General demographic data and PSG parameters were collected from each group of patients.Data were analyzed using R4.1.1 and subjected to one-way ANOVA,and correlation heat maps were drawn for variables that were significant for one-way analysis,and factors influencing the three SMDs were explored using LASSO and logistic regression,and nomogram models were constructed to visualize the results.The overall effect of the constructed regression models was evaluated using ROC curves,calibration curves,and decision curves.Results:The differences in sleep efficiency(SE),REM sleep latency(REML),N2%,and N3% among the three groups of SMD patients were statistically significant(P < 0.05),where the results of the two comparisons showed that there were differences in sleepefficiency and REM latency between Sch and MDE(P < 0.05);Body Mass Index(BMI),duration of illness,REML,N2% and N3% were statistically significant(P <0.05)in BD compared with MDE;and SE and N3% were statistically significant(P <0.05)in Sch compared with BD.there was no significant correlation between BMI and the duration of illness and sleep architecture.PSG parameters were found to be useful for AUC for identifying MDE and Sch was 0.658 and for identifying MDE and BD was 0.6991.calibration curve with decision curve response model had good clinical utility.Conclusion:There maybe differences in PSG parameters among patients with Sch,BD and MDE,and we can establish a prediction model to identify the three diseases by sleep structure,which has high clinical application value and is worth promoting.