【作者】 马菲；

【导师】 曲成毅；

【作者基本信息】 山西医科大学 ， 流行病与卫生统计学， 2009， 博士

【摘要】 研究背景:随着世界人口老龄化发展,痴呆正在成为一个全球性公共卫生问题。轻度认知功能障碍(MCI)是介于正常老化和痴呆或阿尔兹海默病(AD)之间过渡阶段,处于此阶段个体存在超出其年龄范围的记忆或其它认知障碍,但仍能维持功能完好,且达不到痴呆的诊断标准。研究MCI的人群分布特征和转归影响因素对于实现痴呆的早期防治和临床前预警有重要价值。目的1.了解太原市城区65岁以上老年人轻度认知障碍的现患率及人群分布特征。2.探明不同人口学特征、疾病史、遗传特征人群(不同ApoEε4)及人格特征人群(依据16种基本人格与4个次元因素)MCI转归结果及影响因素的差异。3.探明全血APP基因、血脂及ApoEε4在由MCI发展为痴呆中独立作用、相互关系和联合作用。4.探明老年痴呆早期认知功能障碍的特点,在MCI阶段中哪些指标是老年痴呆的危险信号?5.提出判定发展为老年性痴呆敏感、特异、廉价、实用的检测手段及指标。对象和方法本研究采用传统、规范流行病学设计方案,分为现况研究、病例对照研究、巢氏病例对照研究和队列研究四个部分。1.研究对象在太原市范围内,采用整群随机抽样方法调查社区65岁以上老年人6192人,在基底人群中筛选600 MCI对象构成队列人群,按半年一次随访计划,完成三次随访。2.研究方法①基线调查对符合纳入标准的6192名老年人进行集中和入户调查,MCI诊断标准参考DSM-IV“轻微神经认知障碍”,采用MMSE、GDS、ADL,对研究人群进行认知功能评定,可疑MCI及老年痴呆患者由神经科医师进一步做出诊断,排除特定原因引起的MCI,从中筛查MCI者600人。②病例对照研究病例和对照采用n:m不等比匹配方法进行选取,按照前述标准入选患MCI者423人。对照年龄与病例年龄在±2岁,性别相同,文化程度一致,无严重躯体疾病能完成神经心理测验,上述记忆、认知等检查均正常,最终入选995人。应用cox回归模型进行不等比匹配Logistic回归分析,分析社会人口学、生活方式、疾病史、人格特征、生理、血清生化及基因型别指标对MCI发生的影响作用。③巢氏病例对照研究由600名患轻度认知功能损害的社区老年居民组成随访队列,与基线调查相比,智商降低1个SD被认为认知减退,按年龄、性别、文化程度1:1匹配后形成认知减退组和对照组后进行影响因素分析。④队列研究筛选600 MCI对象构成队列人群,按半年一次随访预定计划,分别于2007年11月、2008年6月、2008年12月完成三次随访。中文版MMSE、GDS、ADL、WAIS-RC比较MCI和认知正常受试者转化为痴呆、AD或VD的年转化率。采用cox回归模型分析上述基线指标(一般人口学特征、疾病史、基因标记、人格特征、血清生化指标)对MCI者发生认知减退转化的预测价值。⑤实验室检查对全部MCI和老年痴呆患者抽取静脉血,检测血中ApoEε4、全血APP基因表达、血脂系列。所有指标均设正常对照进行群组比较研究。对所有MCI和老年痴呆患者采用16PF测定人格,同时采用韦氏成人智力量表(WAIS-RC)进行智商评估。⑥统计分析EpiDate3.0软件建立数据库进行两次数据录入。采用SPSS13.0软件进行分析,首先进行单因素分析,在此基础上结合专业知识,用多因素分析方法对某些可能引起混杂作用变量进行调整。多因素分析采用1:1匹配Logistic回归模型及n:m不等比匹配Logistic回归模型分析。随访资料按照人年法计算发病密度及比较转化为认知减退RR和95 %CI ;采用Log-rank检验对每一指标不同水平随访对象认知减退转归比较。所有统计检验均为双侧检验。转归影响因素筛选若是不可控制变量采用Cox Regression过程,可控制变量采用Cox w/Time-Dep Cov过程完成。结果1.现况研究MCI人群发生率9.70% (95% CI: 9.62%- 9.77%),单因素结果显示年龄、性别、文化程度、月经济收入和婚姻状况( p < 0.01),但与职业无关( P > 0.05)。多因素Logistic回归分析结果显示:年龄、性别、文化程度、婚姻状况和职业、年龄和性别交互作用具有显著统计学意义( p < 0.05)。2.病例对照研究最终入选病例423人,对照925人。通过单因素和多因素cox回归分析,得出MCI发生危险因素,其OR值及95%CI分别是:从事体力劳动:1.396(1.092-1.785);吸烟:1.551(1.021-2.359);血清中较高血糖浓度:1.354(1.102-1.664);较高高密度脂蛋白水平:1.543(1.232-1.932);较高低密度脂蛋白水平:1.299(1.060-1.592);低雌激素水平:1.263(1.031-1.547);高血压:1.967(1.438-2.689);糖尿病:1.381(1.139-1.675);抑郁症:1.406(1.110-1.780);脑血栓:1.593(1.307-1.943);较高收缩压:1.331(1.129-1.569);ApoEε4型等位基因:1.462(1.140-1.873);保护因素有:常读书看报:0.610(0.503-0.740);常参加公益活动:0.617(0.502-0.757);常做家务:0.804(0.665-0.973);退休后有第二职业:0.759(0.636-0.906);嗅觉敏锐:0.900(0.845-0.958);外向人格:0.829(0.699-0.984);果断人格:0.811(0.662-0.993)。3.巢氏病例对照研究危险因素有从事体力劳动(OR:1.9,95%CI:1.0-3.6)、吸烟(OR:2.1,95%CI:1.0-4.4)、喜欢呆在家里(OR:2.3,95%CI:1.0-4.9)、血清中较高血糖(OR:3.6,95%CI:1.9-6.8)、胆固醇(OR:2.2,95%CI:1.1-4.3)、低雌激素水平(OR:1.9,95%CI:1.1-3.4),高血压(OR:4.0,95%CI:1.8-4.6),糖尿病(OR:3.0,95%CI:1.9-4.2),高血脂(OR:4.1,95%CI:1.7-9.6),脑血栓(OR:2.3,95%CI:1.3-4.5),脑出血(OR: 2.7,95%CI:1.6-4.8),较高收缩压(OR:2.2,95%CI:1.3-3.6),ApoEε4型等位基因(OR:2.7,95%CI:1.1-6.7)、ApoEε4型等位基因*胆固醇(OR:1.6,95%CI:1.0-2.6);保护因素有:常读书看报(OR:0.2,95%CI: 0.1-0.4)、常做家务(OR:0.2,95%CI:0.1-0.5)、外向人格(OR: 0.5,95%CI: 0.3-0.9)。4.队列研究112对认知减退和正常对照纳入研究,认知减退组人年发病密度14.70%(14.52%,15.29%)和正常对照组3.75%(3.56%,3.67%)两组随访对象转化为认知减退结局的生存曲线经Log-rank检验差异有统计学意义(χ2=11.643,P < 0.01) ,600随访对象,最终557人纳入研究,三年随访后人年发病密度15.31%(13.99%,16.87%)。采用cox回归模型进行单因素及多因素分析结果显示:MCI发生认知减退转归结局因素如下:年龄(RR:1.957;95%CI:1.916-1.999)、女性(RR:2.713;95%CI:1.616-4.554)、高文化程度(RR:0.662;95%CI:0.500-0.877)、糖尿病(RR:2.890;95%CI:1.635-5.107)、脑血栓(RR:1.898;95%CI: 1.157-3.114)、ApoEε4携带者(RR:1.876;95%CI: 1.139-3.090)、内向人格(RR:1.876;95%CI:1.139-3.090)、焦虑人格(RR:2.515;95%CI:1.342-4.711)、高血糖(RR:1.3236;95%CI: 1.193-1.470)、高胆固醇(RR:2.390;95%CI:1.288-4.436)。结论1.社区65岁以上老年人群存在较高MCI发生率,此部分人群作为老年痴呆的高危人群给予足够关注,加强防范。2.MCI是介于认知正常和认知减退之间的一种过渡阶段的认知障碍,MCI转化为认知减退结局的危险性远远大于认知正常受试者。3.横断面研究显示:人口学特征中高龄、老年女性、较低文化程度、体力劳动、独身是MCI发生的危险因素。年龄与性别存在明显交互作用,提示高龄女性是MCI发生的高危人群,应引起重视。4.回顾性研究显示:体力劳动、不健康生活方式、罹患糖尿病、高血压、高血糖、高血脂、抑郁症、脑血管病变发生风险是老年人MCI发生的可疑危险因素,嗅觉减退、内向人格、携带ApoEε4型等位基因可作为预示MCI发生的早期指征,ApoEε4基因型和LDL存在交互作用。5.前瞻性研究显示:人口学特征(高龄、女性、较低文化程度)、疾病史(罹患糖尿病和脑出血)、ApoEε4基因型携带者、血清较高血糖、胆固醇及低密度脂蛋白及内向、焦虑人格特征是影响转归危险因素,坚持脑力劳动、健康生活方式、降低糖尿病、高血压、高血糖、高血脂、脑血管病变发生风险是预防老年人认知减退主要手段,内向人格、携带ApoEε4型等位基因可作为预示认知减退的早期指征。6.通过对社区老年人群轻度认知功能障碍发生的影响因素与认知减退转归的影响因素的对比,发现具有较高的一致性,提示对社区MCI人群早期干预可能对认知减退甚至老年痴呆预防起到重要作用。更多还原

【Abstract】 BackgroundWith the rapid increase in life expectancy in the population,Dementia is becoming one of the important healthy problems.Mild cognitive impairment(MCI)refers to the transition clinical state between normal aging and dementia or Alzheimer?disease (AD) in which individuals experience memory loss or other cognitive deficit to a greater extent than one would expect for age,yet they do not meet currently accepted criteria for dementia.It?s valuable to study the cognitive characterization and natural history of MCI for the early detection and prevention of dementia.Objects1.To determine the prevalence and distribution of MCI in the aged population and analyze socio-demographic factors.2..To understand the influencing factor and the difference in the outcome of the conversion to cognitive decline among the MCI subjects with different socio-demographic and personality characteristics、disease history、and hereditary feature.3.To study independent role,united role of amyloid precursor protein blood fat apolipoprotein e during the period of conversion to cognitive decline among the MCI subjects and the relationship between them.4. To find out the characteristic of MCI during the earlier period of dementia and manifestation of memory decline associated with the conversion to dementia.5.To provide sensitive、Specific、cheap and pragmatic detection means and index for assessment of the conversion to dementia. Subjects and methodsThe initial phase 1 assessment took place in May 2007.A cross-sectional study was conducted among the community-dwelling elders aged over 65. A total of 6192 subjects (91.06% of the eligible subjects) participated in the baseline survey (2007.4-2007.6). According to the Petersen’s diagnostic standard on mild cognitive impairment(MCI), 600 subjects with MCI were screened from the baseline population.all of which were treated as the cohort population. The follow-up interviews were performed at the following 2 years during phase 2 after the baseline assessment. Follow-up visits occurred annually, with the last examination occurring in November 2008.1.Baseline assessmentBy cluster random sampling,6192 aged people over 65 were ultimately involved. This survey composed of face-to-face interviews and self-administered questionnaires including questions on socio-demographic features, subjective ratings of memory and physical and emotional health and medical conditions such as medical history, current medication, and a subjective assessment of memory disturbances or depression. The scale such as MMSE、GDS and ADL have been performed to measure cognitive function.Information about the subjects was evaluated by a panel of neurologists, neuropsychologists, a geriatrician, and clinical nurses. Of all the 6192 community-dwelling participants aged 65 and older who underwent psychometric testing,, 600 subjects met our criteria for MCI.2.Case-control studyA n:m matched case-control study was conducted to analyze influencing factors of mild cognitive impairment among the elderly community-based population. Four hundred and twenty-three cases together with nine hundred and twenty-five controls were interviewed with a uniformed questionnaire .cases were matched with controls by age decade,education group(i.e.,less than 10 years of education or greater than or equal 10 years of education),and gender in a n:m ratio.All the subjects can accomplish psychological tests independently with adequate cognition and memory. Cox regression model of survival analysis was selected to deal with non-geometric proportional matched data which is difficult to analyze by Logistic regression model.the influencing factors analyzed included socio-demographic、life style disease history、personality characteristics、physiological functions、biochemical index in the serum and hereditary feature.3.Nested case-control study600 subjects with MCI were screened from the baseline population.all of which were treated as the cohort population.In each follow-up,cognitive decline was defined as at least a 1 SD drop in IQ Compared with the baseline survey. Subjects who conform to the cretia above can be included in cognitive decline group,while subjects who were normally cognitive can be included in stable group Each case was matched with one control by age decade,education group(i.e.,less than 10 years of education or greater than or equal 10 years of education),and gender.Exclusion criteria for controls were the same as those for cases. Of all these subjects,114 cognitive decline individuals and 114 cognitive stable individuals were selected to participate in the experiment reported here. they formed the final samles of the present study.Analysis are based on two groups.Cox regression model of survival analysis was applied to analyze influencing factors with SPSS 13.0 Software 4.Cohort study600 subjects with MCI were screened from the baseline population.All of which were treated as the cohort population. we employed the Chinese version of the Wechsler Adult Intelligence Scale-Revised (WAIS-R) to assess general intellectual ability and used the IQ value to assess cognitive variation among cohort population.The conversion rates to cognitive decline of MCI and cognitive normal subjects are calculated respectively. Cox regression model was performed to analyze prediction value of socio-demographic and personality characteristics、disease history、and hereditary feature on whether can be converted into cognitive decline or not.5.Experimental examinationBlood samples were drawn in the morning after overnight fasting, amyloid precursor protein、blood fat、apolipoprotein E were measured . We employed the Chinese version of the Wechsler Adult Intelligence Scale-Revised (WAIS-R) to assess general intellectual ability and 16PF to determine personality.6.Statistical AnalysisDatabase was constructed by EpiDate3.0 software,All the data information were input for two times. Statistical Analysis were performed by SPSS13.0 software..Results were analyzed in two ways:First, we examined the distribution of continuous and categorical variables using t-test andχ2-test.Second,we used logistic regression in univariate and multivariate modeling to estimate the association between influencing factors and cognitive decline as measured by domain-specific cognitive tests,adjust for possible confounding variables ,Cox regression model of survival analysis was selected to deal with non-geometric proportional matched data which is difficult to analyze by Logistic regression model.According to the follow-up information,incidence density of cognitive decline in both MCI group and normal cognitive group were calculated by person-years approach. RR ,together with 95%CI of the conversion to the outcome of cognitive decline in MCI group were compared with that in the normal cognitive group.Log-rank test was performed to make comparison with the different level in follow-up subjects? each index. All tests were two-sided and our level of significance wasα= 0.05.Cox Regression was adopted if the influencing factor was treated as uncontrollable variable,while Cox w/Time-Dep Cov was adopted if the influencing factor was treated as controllable variable.Results1.Baseline assessmentResults showed that an overall prevalence of MCI was 9.70% (95% CI: 9.62% to 9.77%).Univariate analyses showed that the prevalence of MCI were significantly different among different groups assigned according to age, gender, education level ,monthly household income and marital status( p < 0.01) ,but not among different occupation groups(P > 0.05). Considering the effect of the interactions between age and gender,educational level and occupation on multiple logistic regression,Significant predictor variables were as follows:age, gender, marital status, educational level and occupation( p < 0.01).However,OR of MCI between groups with different monthly household income was similar( P > 0.05).2.Case-control studyFour hundred and twenty-three cases together with nine hundred and twenty-five controls were interviewed with a uniformed questionnaire. By univariate and multivariate cox regression analysis, OR and 95%CI of these risk factors were physical labour(OR: 1.396 ,95%CI: 1.092-1.785);smoking(OR:1.551,95%CI:1.021-2.359);higher level of blood glucose(OR:1.354,95%CI:1.102-1.664);HDL in the serum(OR:1.543,95%CI:1.232-1.932);LDL in the serum(OR:1.299,95%CI:1.060-1.592);lower level of estrogen in the serum(OR:1.263 ,95%CI:1.031-1.547);hypertension(OR:1.967,95%CI:1.438-2.689);diabete(OR:1.381,95%CI:1.139-1.675);depressive disorder(OR:1.406,95%CI:1.110-1.780);cerebral thrombosis(OR:1.593,95%CI:1.307-1.943);higher SBP(OR:1.331,95%CI:1.129-1.569);ApoEε4 carrier(OR:1.462,95%CI:1.140-1.873). OR and 95%CI of protection factors were reading newspaper frequently:(OR:0.610,95%CI:0.503-0.740);doing household duties frequently(OR:0.804,95%CI:0.665-0.973);frequent social activities(OR:0.617,95%CI:0.502-0.757);reemployment after retirement(OR:0.759,95%CI: 0.636-0.906);acumen olfaction(OR:0.900,95%CI:0.845-0.958); extroversion character(OR:0.829,95%CI:0.699-0.984); decisive character(OR: 0.811,95%CI:0.662-0.993).3.Nested case-control studyUnivariate and multivariate cox regression analysis revealed that risk factors were physical labour(OR: 1.949,95%CI:1.041-3.637),smoking(OR: 2.062,95%CI:1.029-4.445),to stay alone(OR: 2.254,95%CI:1.029-4.937),higher level of blood glucose(OR:3.584 ,95%CI:1.891-6.791)、cholesterol in the serum(OR:2.204,95%CI:1.137-4.275),lower level of estrogen in the serum(OR:1.946,95%CI:1.087-3.411),hypertension(OR:3.951,95%CI:1.822-4.637),diabete(OR:3.016,95%CI:1.886-4.157),hyperlipemia(OR:4.061,95%C:1.724-9.568),cerebral thrombosis(OR:2.347,95%CI:1.329-4.533),cerebral hemorrhage(OR:2.668,95%CI:1.579-4.802),higher SBP (OR:2.208,95%CI:1.343-3.629),ApoEε4 carrier(OR:2.717,95%CI:1.084-6.743),ApoEε4 carrier* cholesterol in the serum(OR:1.626,95%CI:1.011-2.618); Protection factors were reading newspaper(OR:0.203,95%CI:0.112-0.411) and doing household duties frequently(OR: 0.249,95%CI:0.135-0.528)、extroversion character(OR:0.544,95%CI: 0.327-0.938).4.Cohort studyAccording to changes and trends of cognitive function, 112 pairs of cognitive decline subjects and stable subjects were identified,both of which were matched on sex, age and educational level in a 1:1 ratio.Incidence density of cognitive decline subjects is 14.70%(14.52%,15.29%),while that of stable subjects is 3.75%(3.56%,3.67%).By Log-rank test ,there is significant difference in Survival curve on the outcome of cognitive decline between cognitive decline group and stable group(χ2=11.643,P < 0.01).Among 600 follow-up subjects ,557 were ultimately included into study,For two-year follow-up, incidence density is 15.31%(13.99%,16.87%),By univariate and multivariate cox regression analysis, RR and 95%CI of these risk factors were age(RR:1.957;95%CI:1.916-1.999)、sex(RR:2.713;95%CI:1.616-4.554)、educational level(RR:0.662;95%CI:0.500-0.877)、diabete(RR:2.890;95%CI:1.635-5.107)、cerebral thrombosis(RR:1.898;95%CI:1.157-3.114)、ApoEε4(RR:1.876;95%CI:1.139-3.090)、introversion character(RR:1.876;95%CI:1.139-3.090)、anxious characte(RR:2.515;95%CI:1.342-4.711)、higher level of blood glucose(RR:1.3236;95%CI:1.193-1.470)、higher level of cholesterol in the serum(RR:2.390;95%CI:1.288-4.436)。Conclusion1.This study confirms the high prevalence of MCI among the elderly community-based population aged above 65 in China, similar to previous epidemiological studies in other countries,2.MCI has been proved to be a step for the transition between normal aging and cognitive decline. The conversion rate of cognitive decline in MCI group is much higher than that in stable group,risk of the conversion to cognitive decline for MCI is much higher than that for normal control.3.Cross-sectional study shows that nearly all socio-demographic characteristics are associated with MCI. Age and sex can interact with each other, positive interaction existed between Age variable and sex variable in MCI,The putative risk factors identified merit further study4.Retrospective study shows that the major measures to prevent MCI are to go in for mental labour、healthy life style and decrease the risk to develop hypertension、diabete、depressive disorder、cerebrovascular disease. Olfactory hypoesthesia、cowardice and introvert character、ApoEε4 carrier can be treated as early indication to signify MCI.the interaction between ApoEε4 and LDL can be found.5.Perspective study shows that the risk factors of influencing conversion included socio-demographic (age、sex、educational level)、personality characteristics(introversion character、anxious character)、disease history(diabete、cerebral thrombosis)、and hereditary feature(ApoEε4)higher level of cholesterol in the serum and higher level of blood glucose. The major measures to prevent cognitive impairment among aged population with MCI are to go in for mental labour、healthy life style and decrease the risk to develop hypertension、hyperlipemia、diabete、cerebrovascular disease. introvert character、ApoEε4 carrier can be treated as early indication to signify cognitive impairment among subjects with MCI. 6. By comparison of influencing factors between the onset of MCI and the conversion of cognitive decline,higher concordance has been found.Which indicated that early intervention on the community-based elderly population with MCI may play an important role on cognitive decline even dementia.更多还原