摘要
目的:探究不同重性精神障碍病人之间睡眠结构的差异性,并构建预测模型。方法:选取2021年3月—2023年2月就诊于山西医科大学第一医院的精神分裂症病人100例、双相情感障碍病人100例和重度抑郁发作病人223例,收集病人一般人口学资料以及3种疾病病人多导睡眠监测参数并进行比较。使用LASSO和Logistic回归分析构建列线图预测模型。结果:3种疾病病人的体质指数(BMI)、病程、睡眠效率、快速动眼期(REM)潜伏期、N2期睡眠比例、N3期睡眠比例比较差异有统计学意义(P<0.05)。重度抑郁发作与双相情感障碍病人睡眠效率和REM潜伏期比较差异有统计学意义(P<0.05),精神分裂症与重度抑郁发作病人REM期潜伏期、N2期睡眠比例、N3期睡眠比例比较差异有统计学意义(P<0.05),精神分裂症与双相情感障碍病人睡眠效率和N3期睡眠比例比较差异有统计学意义(P<0.05)。以重度抑郁发作为参考,构建的预测模型识别精神分裂症和重度抑郁发作的受试者工作特征曲线下面积(AUC)为0.658,识别双相情感障碍和重度抑郁发作的AUC为0.699,校正曲线与决策曲线反映模型具有较好的临床效用。结论:精神分裂症、双相情感障碍和重度抑郁发作病人的睡眠结构存在差异,医护人员可以通过建立预测模型鉴别3种疾病。
Objective:To explore the difference of sleep structure among patients with different types of major mental disorders,and to build prediction model.Methods:A total of 100 patients with schizophrenia,100 patients with bipolar disorder and 223 patients with major depressive episode were selected from the First Hospital of Shanxi Medical University from March 2021 to February 2023.The general demographic data of patients and the parameters of polysomnography in patients with 3 diseases were collected and compared.Nomogram prediction model was constructed using LASSO and Logistic regression.Results:There were significant differences in BMI,disease duration,sleep efficiency,REM latency,N2 sleep proportion and N3 sleep proportion among patients with 3 kinds of diseases(P<0.05).There were significant differences in sleep efficiency and REM latency between patients with bipolar disorder and major depressive episode(P<0.05).There were significant differences in REM latency,N2 sleep proportion and N3 sleep proportion between patients with schizophrenia and major depressive episode(P<0.05).There were significant differences in sleep efficiency and N3 sleep proportion between patients with schizophrenia and bipolar disorder(P<0.05).Major depressive episode as a reference,the area under the subject working characteristic curve(AUC)for identifying schizophrenia and major depressive episode was 0.658,and the AUC for identifying major depressive episode and bipolar disorder was 0.699.The response model of correction curve and decision curve had good clinical effectiveness.Conclusion:There were differences in sleep structure among patients with schizophrenia,bipolar disorder and major depressive episode.Medical staff can identify three diseases by building predictive model.
作者
栗恺雯
郎小娥
刘彩
郝萌
LI Kaiwen;LANG Xiao'e;LIU Cai;HAO Meng(The First Clinical College of Medicine,Shanxi Medical University,Shanxi 030001 China)
出处
《护理研究》
北大核心
2023年第19期3453-3460,共8页
Chinese Nursing Research
基金
中央引导地方科技发展资金项目,编号:YDZJSX2022A064
山西医科大学第一医院“136”专项经费科研项目,编号:Y2022136006。
