北京城乡55岁或以上居民简易智能状态检查测试结果的分布特征

The minimental state examination in the Chinese residents population aged 55 years and over in the urban and rural areas of Beijing

目的研究简易智能状态检查（ＭＭＳＥ）测试结果的分布特征，为制定适合我国国情的筛查分界值的标准提供科学的依据。方法通过分层多级整群抽样，以北京市５５岁或以上５３６７名居民为样本。总结其简易智能状态检查（ＭＭＳＥ）测试结果的百分位数值和均值的年龄、性别和受教育程度的分布；比较我国不同的ＭＭＳＥ筛查分界值检测痴呆的效度。按城市和农村建立多元线性回归方程，分析影响测试结果的因素。结果５５～５９岁至８５岁或以上年龄组的居民，在文盲组、小学组、初中及以上组，中位ＭＭＳＥ成绩分别为２３～１８，２７～２４，２９～２７分。性别、年龄和受教育程度诸因素均对ＭＭＳＥ得分有显著影响（Ｐ＜０．０１）。ＭＭＳＥ的分界值文盲组≤１９分、小学组≤２２分、初中及以上组≤２６分检测痴呆的敏感度（城乡分别为９０．７％和９７．１％）高于既往常用的分界值（城乡分别为６６．７％和７８．６％）。结论ＭＭＳＥ得分是判断认知功能是否正常的非特异性指标，不能取代临床诊断。我们调整的分界值标准，有助于早期发现痴呆病人，减少漏诊。我国既往报道的痴呆患病率远低于西方国家的原因之一可能和筛查分界值较低有关。

Objective To make a study of the distribution of miniental state examination (MMSE) scores in terms of ages and educational levels in the Chinese residents aged 55 years and over and to offer a benchmark for a cutoff score. Methods A total of 5 367 residents aged 55 years or older was drawn by means of stratified multistage cluster sampling in the urban and rural areas of Beijing.The summary scores for the MMSE were given in the form of mean percentage distributions specific for age, sex and educational level; to compare the validity of the MMSE using different cutoff scores by different studies in China. By a fitting multiple regression model to assess the influence of the factors on MMSE score in our urban and rural population. Results We found the bivariate correlation between the MMSE scores and all three factors of age, sex, and educational level (P<0.01). The optimal cutoff values were taken from the 10th percent lowest score in the age group of 60 to 65 years for each educational level: illiterate19, elementary22 and high26. When compared with cutoff score used by other studies, our values achieved high sensitivity (90.7% Vs 66.7% in the urban area,97.1% Vs 78.6% in the rural area). Conclusion The MMSE scores should be used to identify current cognitive difficulties and not to make formal diagnoses. Our adjustment in the cutoff scores would improve the disease detection by reducing the number of false negatives.It is suggested that the reported low prevalence of dementia in China may be related to the use of low cutoff MMSE scores.