摘要： 目的 调查浙江省杭州市农村社区老年人轻度认知损害患病率，并探讨轻度认知损害筛查量表（sMCI）在低受教育程度老年人中的筛查准确性。方法 2010年4-9月招募浙江省杭州市江干区（现为上城区）四季青街道360例老年人，采用简易智能状态检查量表（MMSE）、蒙特利尔认知评价量表（MoCA）、sMCI量表和临床痴呆评价量表（CDR）判断痴呆和轻度认知损害；参照美国精神障碍诊断与统计手册第4版及相关临床评价量表判断认知功能障碍类型及其他神经精神疾病（包括抑郁、焦虑、双相情感障碍、精神分裂症、精神发育迟滞）。绘制受试者工作特征（ROC）曲线并计算曲线下面积，比较sMCI量表与MoCA和CDR量表筛查轻度认知损害的准确性。结果 最终171例完成全部调查。（1）55例（32.16%）诊断为轻度认知损害，其中遗忘型轻度认知损害25例（14.62%）、非遗忘型轻度认知损害30例（17.54%），11例（6.43%）诊断为痴呆，10例（5.85%）为抑郁障碍，4例（2.34%）为焦虑障碍，1例（0.58%）为双相情感障碍，1例（0.58%）为精神分裂症，1例（0.58%）为精神发育迟滞。154例认知功能障碍患者，经MMSE量表筛查痴呆25例（16.23%），CDR量表筛查痴呆8例（5.19%），临床确诊11例（7.14%）；经MoCA量表筛查轻度认知损害138例（89.61%），sMCI量表筛查轻度认知损害117例（75.97%），CDR量表筛查轻度认知损害70例（45.45%），临床确诊55例（32.16%）。（2）以临床诊断作为参考标准，ROC曲线显示，CDR量表、sMCI量表和MoCA量表筛查轻度认知损害的曲线下面积分别为0.90 ±0.03（95% CI：0.844~0.957，P=0.000）、1.00 ±0.00（截断值为评分23；95% CI：1.000~1.000，P=0.000）和0.67 ±0.05（截断值为22；95% CI：0.578~0.756，P=0.001）、0.53 ±0.05（95% CI：0.430~0.621，P=0.603）。（3）以CDR评分作为参考标准，ROC曲线显示，在受教育程度为4~6年的受试者中sMCI和MoCA量表筛查轻度认知损害的曲线下面积分别为0.50 ±0.17（截断值为评分23；95% CI：0.161~0.839，P=1.000）和0.49 ±0.17（95% CI：0.157~0.824，P=0.955），在受教育程度为0~3年的受试者中sMCI和MoCA量表的曲线下面积分别为0.64 ±0.05（截断值为22；95% CI：0.535~0.744，P=0.011）和0.51 ±0.06（95% CI：0.402~0.617，P=0.858）。结论 受教育程度较低的农村社区老年人存在轻度认知损害较常见，sMCI量表诊断轻度认知损害的准确性高于MoCA量表，其截断值分别为评分23（受教育程度4~6年）和22（受教育程度0~3年），值得临床推广应用。

关键词: 认知障碍, 神经心理学测验, ROC曲线, 农村卫生, 老年人, 流行病学研究, 横断面研究

Abstract: Objective To investigate the prevalence of mild cognitive impairment (MCI) among the elderly in rural communities in Hangzhou, Zhejiang, and to explore the screening accuracy of Screening Scale for Mild Cognitive Impairment (sMCI) in the elderly with low education. Methods From April 2010 to September 2010, 360 elderly people in Sijiqing street, Jianggan district (now Shangcheng district), Hangzhou, Zhejiang were recruited. Dementia and MCI were judged by Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), sMCI and Clinical Dementia Rating Scale (CDR). Draw receiver operating characteristic (ROC) curve and calculate the area under the curve (AUC), and compare the accuracy of sMCI, MoCA and CDR scores in screening MCI. Results Finally, 171 cases completed all investigations. 1) 55 cases (32.16%) were diagnosed as MCI, including 25 cases (14.62%) of amnestic MCI (aMCI) and 30 cases of non-aMCI, 11 cases (6.43%) of dementia, 10 cases (5.85%) of depression, 4 cases (2.34%) of anxiety disorder, one case (0.58%) of bipolar disorder, one case (0.58%) of schizophrenia and one case (0.58%) of mental retardation. Among 154 patients with cognitive impairment, 25 cases (16.23%) were screened for dementia by MMSE, 8 cases (5.19%) were screened for dementia by CDR, and 11 cases (7.14%) were clinically confirmed; 138 cases (89.61%) of MCI were screened by MoCA, 117 cases (75.97%) by sMCI, 70 cases (45.45%) by CDR, and 55 cases (32.16%) were clinically confirmed. 2) Taking clinical diagnosis as reference standard, the ROC curve showed CDR score had the highest accuracy in screening MCI, and the AUC was 0.90 ±0.03 (95%CI:0.844-0.957, P=0.000); the AUC of MoCA score was 0.53 ±0.05 (95%CI:0.430-0.621, P=0.603); when the cut-off value of sMCI score was 23, the AUC was 1.00 ±0.00 (95%CI:1.000-1.000, P=0.000). The cut-off value of subjects with education level of 0-3 years was adjusted to 22, and the AUC was 0.67 ±0.05 (95%CI:0.578-0.756, P=0.001). 3) According to education level, they were divided into 0-3 years group (113 cases) and 4-6 years group (47 cases). Taking CDR score as the reference standard, ROC curve showed the AUC of MoCA score in screening MCI in 4-6 years group was 0.49 ±0.17 (95%CI:0.157-0.824, P=0.955), the cut-off value of sMCI score was 23, the AUC of sMCI score was 0.50 ±0.17 (95%CI:0.161-0.839, P=1.000); the AUC of MoCA score in the 0-3 years group was 0.51 ±0.06 (95%CI:0.402-0.617, P=0.858), and the cut-off value of sMCI score was adjusted to 22, and the AUC was 0.64 ±0.05 (95%CI:0.535-0.744, P=0.011). Conclusions It is more common for the elderly in the rural communities with low education to have MCI, the accuracy of sMCI in screening MCI is higher than MoCA, and the cut-off value is 23 (education level 4-6 years) and 22 (education level 0-3 years), which is worthy of clinical application.

Key words: Cognition disorders, Neuropsychological tests, ROC curve, Rural health, Aged, Epidemiologic studies, Cross-sectional studies