【作者】 王贺波；

【导师】 吕佩源；

【作者基本信息】 河北医科大学 ， 神经病学， 2009， 博士

【摘要】 血管性痴呆(vascular dementia, VD)是由脑血管疾病引起的缺血性、缺血-缺氧性或出血性脑损害所导致的认知功能障碍,是继阿尔茨海默病(Alzheimer’s disease, AD)之后老年期痴呆的第二常见原因。随着对认识损害研究的深入,以AD为模型的VD诊断标准存在的不足之处日趋明显。以往VD的定义以AD为模型,主要强调学习新知识障碍和记忆的缺陷,而不是典型VD所伴有的精神运动迟缓和执行功能障碍;其诊断要求患者认知损害达到影响日常生活能力程度,这样就会影响脑血管性疾病所致轻度认知功能损害的早期诊断,低估其发病率,耽误患者的早期预防和治疗。1993年Hachinski和Bowler提出用血管性认知障碍(vascular cognitive impairment, VCI)这一更为宽泛的概念代替或涵盖VD。VCI概念的提出对于痴呆的早期诊断、早期预防和治疗具有重要的临床意义。VCI是指由血管因素引起的或与之伴随的认知功能损害。与以往的VD概念相比,其含义更加宽泛,包括各种脑血管及其危险因素导致的各种水平的认知功能损害,尤其是包括了未到达痴呆标准的轻度认知功能损害。VCI为一大类综合征,可分为三个主要亚型,即皮质型VCI、关键部位梗死型VCI和皮质下型VCI。不同类型VCI的病因和发病机制不同,其临床表现和治疗也不尽相同。因为血管性危险因素是可以治疗的,所以应该有可能预防、延缓或减轻VCI,但是由于缺乏统一的、令人满意的诊断标准,已使VCI的研究进展受到阻碍。故而有关VCI研究的重点逐渐向临床特征相对一致的皮质下型VCI转移。皮质下缺血性脑血管病(subcortical ischemic vascular disease, SIVD)是造成皮质下型VCI最常见原因。目前认为SIVD的主要病因是小血管病变(cerebral small vessel disease, SVD),其发病可能是在多种因素的共同作用下引起脑小血管病变,最终导致小动脉阻塞和腔隙性脑梗死(lacunar infarcts, LI);或是由于深穿支动脉临界狭窄和低灌注引起广泛的白质不全梗死。SIVD的临床症状包括运动和认知执行速度减慢、健忘、构音障碍、情绪改变、小便失禁和步态异常等。高血压是其主要的危险因素,其他导致脑血管疾病的血管性危险因素如糖尿病、高脂血症和心脏疾病、贫血、吸烟、酗酒等,均可促进SIVD的发生。而并非所有SIVD均出现认知损害,徐群等报道44.5%SIVD患者可有血管性认知功能损害,那么以上SIVD危险因素与其认知损害的关系如何,尚少见报道。积极探讨和研究皮质下VCI的危险因素,进一步明确其与各种血管性危险因素的关系,有助于VCI患者的早期预防和治疗。SIVD特征性病理表现为多发性LI和皮质下缺血性脑白质病变(white matter lesions, WML),后者以往称为白质疏松。WML在磁共振(magnetic resonance imaging, MRI)T2加权像和液体衰减反转恢复像(fliud attenuated inversion recovery, FLAIR )表现为脑白质高信号( white matter hyperintensities,WMH)。多年来,人们一直没有停止影像学改变的程度和类型与认知功能损害关系的研究,但没有明确的结论。早期研究结果表明,白质疏松与认知功能的相关性较弱。这与以往研究多采用定性分析不无关系。而近年来发展兴起的自动图像分割定位、定量技术,为神经影像学改变与认知功能损害关系的研究翻开了崭新的一页。Bowler综述近年来有关WML与认知功能关系的研究结果,提出"如忽视病变部位的重要性,而试图找到梗死、白质病变体积与认知改变的密切联系必然会失败"。综合定位、定量分析才有可能对SIVD患者早期认知损害,乃至其他类型VD的早期诊断提供可靠的帮助。为此,本研究拟通过病例对照研究,从危险因素、结构性神经影像学病灶定位、定量等方面对SIVD进行对比观察,以便掌握皮质下VCI临床及辅助检查特征、LI和WML在认知损害中的作用及其相互影响。为建立统一的、可操作的诊断标准提供部分帮助,最终能为易感患者的早期预防、为可疑患者的早期诊断和早期治疗提供临床及辅助检查证据,避免或减缓痴呆的发生。第一部分:皮质下缺血性脑血管病的临床特征及其认知损害的危险因素目的:观察SIVD患者的临床特征,确定SIVD患者认知损害的危险因素,同时分析不同亚型SIVD的危险因素。方法:急性缺血性脑血管病发病3个月以上的患者,由经过培训的专职医师、应用标准化语言对患者和家属进行访谈,应用标准化方法对患者进行详细查体,记录患者症状和体征,进行神经心理学评估并行头MRI扫描。根据Erkinjuntti提出MRI诊断标准最终入选SIVD患者70例,其中白质病变型(white matter lesions type SIVD, WML-SIVD)43例,腔隙性脑梗死突出型(lacunar infarcts type SIVD, LI-SIVD)27例。根据神经心理学评估分为认知正常组(no cognitive impairment, NCI)和血管性认知障碍组(VCI),详细记录其血管性危险因素;应用单因素和多因素Logistic逐步回归进行分析。结果:70例SIVD患者平均年龄(66.89±7.03)岁,男性患者33例,女性37例,平均受教育时间为(2.81±2.37)年。NCI组和VCI组患者简明精神状态检查(mini-mental status examation, MMSE)得分分别为(24.45±2.95)和(19.81±2.56),VCI患者MMSE得分明显降低(P<0.05)。临床症状以肢体肌力减退为主,占81.43%(57例),行走不稳者40% (28例),构音障碍患者20%(14例),但症状较轻;其次为饮水呛咳和尿失禁各占14.29%(10例)。神经系统查体71.4%患者检出上运动神经元受损体征,14.9%假性延髓麻痹(10例),7例发现锥体外系体征(10%),而共济失调步态异常者仅占7.14%。不同亚型SIVD单因素分析显示WML-SIVD患者高血压发生率明显高于LI-SIVD患者(P<0.05),LI-SIVD患者糖尿病、高脂血症和冠心病发生率明显高于WML-SIVD患者(P<0.05),两种亚型在贫血、吸烟和饮酒史比率无统计学差异(P>0.05)。非条件Logistic逐步回归分析结果显示高血压病为WML-SIVD患者的危险因素[odds ratio (OR) 8.531 (1.676-43.41); P=0.008];而糖尿病[OR 0.082 (0.016 to 0.436); P=0.003]、高脂血症[OR 0.158 (0.035 to 0.720) ; P=0.019]和冠心病[OR 0.005 (0.007 to 0.336); P=0.002]在LI-SIVD的发生率更高。SIVD患者认知损害血管性危险因素单因素分析显示,VCI组患者高血压、糖尿病和高脂血症病史的发生率明显高于NCI组(P<0.05),提示存在高血压、糖尿病和高脂血症病史SIVD患者出现认知损害的危险增加,而心脏病史、贫血、饮酒和吸烟史在两组之间无显著差异(P>0.05)。校正其他因素影响进行非条件Logistic逐步回归分析,结果证实高血压、糖尿病、高脂血症SIVD患者更易出现认知损害(P<0.05),其OR值分别为高血压5.265(1.563-17.731),糖尿病3.445(1.008-11.772),高脂血症3.649(0.974-11.466);结论:SIVD患者主要临床表现为肢体肌力减退、行走不稳、构音障碍、尿失禁和饮水呛咳,以及上运动神经元受损和假性延髓麻痹体征,部分患者可见锥体外系体征。两种SIVD亚型的危险因素不尽相同。高血压为WML-SIVD患者的危险因素,而LI-SIVD患者糖尿病、高脂血症和冠心病发生率明显高于WML-SIVD患者。SIVD患者出现认知损害的危险因素为高血压、糖尿病,其危险程度分别增加5.265倍和3.445倍。高脂血症对SIVD患者出现认知损害所起作用尚不能明确。年龄可能并不是SIVD认知损害的危险因素。第二部分:皮质下血管性认知障碍与脑白质病变体积的关系目的:确定SIVD患者缺血性WML体积与认知损害之间的关系。方法:急性缺血性脑血管病发病3个月以上的患者,由经过培训的专职医师、应用标准化语言对患者和家属进行访谈,应用标准化方法对患者进行详细查体,记录患者症状和体征,进行神经心理学评估并行头MRI扫描。根据Erkinjuntti提出MRI诊断标准最终入选SIVD患者70例,根据神经心理学评估分为NCI组和VCI组。采用半自动定量测定方法,测定患者WMH体积。应用相关分析和分层多元回归分析认知功能与WMH体积的关系。结果:70例患者平均WMH体积(29.56±15.58)cm3,范围4.20-86.01cm3。NCI组和VCI组平均WMH体积分别为(23.63±10.52)cm3和(34.84±17.50)cm3,VCI组WMH体积明显高于NCI组,统计学分析显示显著性差异(Z=-2.83,P=0.005)。单因素相关分析显示MMSE与教育呈正相关(r=0.49,P<0.001),而与年龄(r=-0.235,P=0.025)、WMH体积(r=-0.398,P<0.001)呈负相关,同时WMH体积与年龄呈正相关(r=0.279,P=0.01),而教育与年龄呈负相关,具有统计学差异。分层多元线性回归分析显示,首先进入模型的年龄、性别和受教育程度,仅受教育程度与简明精神状态检查得分有关(P<0.001)。当WMH体积进入模型后,简明精神状态检查得分仍与受教育程度有关(P<0.001)。校正年龄、性别与受教育程度的影响后,WMH体积与MMSE得分呈负相关(B=-0.077, t=-3.243,P=0.002),具有统计学显著性差异;WMH体积变化能解释简明精神状态评分改变的10.2%。结论:ImageJ结合其Voxel Counter Plugins测定和计算脑白质高信号体积的方法应用方便,省时、准确,测定结果可靠,可操作性强。SIVD患者认知损害与白质病变程度密切相关,随着WML程度逐渐增加,SIVD患者认知功能将不断恶化。在校正年龄、性别和受教育程度的影响后,WMH体积仅能解释认知损害变化的10.2%。第三部分:皮质下血管性认知障碍与脑白质病变部位的关系目的:分析SIVD患者缺血性脑WML部位与认知功能损害的关系,同时对WMH体积测定与分级评分的相关性进行评价。方法:急性缺血性脑血管病发病3个月以上的患者,由经过培训的专职医师、应用标准化语言对患者和家属进行访谈,应用标准化方法对患者进行详细查体,记录患者症状和体征,进行神经心理学评估并行头MRI扫描。根据Erkinjuntti提出MRI诊断标准最终入选SIVD患者70例,根据神经心理学评估分为NCI组和VCI组。本研究在前一部分研究结果的基础上,应用年龄相关白质改变分级方法,对WMH评分,应用相关分析和分层多元回归分析认知功能与WMH部位的关系,以及WMH体积和部位的相互影响。结果:70例皮质下缺血性脑血管病患者白质高信号总平均分为(12.37±3.93)分,范围3-20。WMH体积和总评分相关分析结果显示,随着WMH体积增加,WMH总评分增多,两者呈正相关(r=0.879, P<0.001)。以白质高信号体积为自变量绘制散点图,显示WMH体积与WMH总评分呈曲线关系。VCI组WMH总平均分(13.65±3.51),较NCI组WMH总平均分(10.94±3.94)明显升高,具有统计学显著性差异(t=-3.043,P=0.003)。VCI组额区、顶枕区和基底节区WMH评分均较NCI组明显升高,具有显著性差异(P<0.05);而颞区和幕下区WMH评分两组无显著性差异(P>0.05)。分层多元线性回归分析显示结果。首先进入模型的年龄、性别和受教育程度,结果同前。当WMH总评分进入模型后,简明精神状态检查得分仍与受教育程度有关(P<0.001)。在校正年龄、性别与受教育程度的影响后,WMH总平均分与MMSE得分呈负相关(B=-0.270,t=-2.842, P=0.006),WMH总平均分仅能解释简明精神状态评分改变的8.1%,低于WMH体积的10.2%;而各脑区评分能解释简明精神状态评分改变的19.3%,明显高于WMH体积和总评分。其中额区和基底节区WMH评分与MMSE得分有关,具有统计学显著性差异(P<0.05),而顶枕区、颞区和幕下区WMH评分与MMSE得分无统计学差异(P>0.05)。结论:白质病变体积与总评分呈正相关,两种方法均可用来评价白质病变的程度;而与评分方法相比,WMH体积测定能更加敏感地反映认知功能改变。SIVD患者认知功能与白质病变的部位密切相关,白质病变部位评价认知损害改变较体积测定更敏感。SIVD患者认知功能损害与额区、基底核区白质病变有关,而与顶枕区、颞区和幕下区白质病变无关。第四部分:皮质下血管性认知障碍与腔隙性脑梗死的关系目的:探讨腔隙性脑梗死与皮质下缺血性脑血管病认知损害的关系,及腔隙性脑梗死和脑白质病变之间的相互影响。方法:急性缺血性脑血管病发病3个月以上的患者,由经过培训的专职医师、应用标准化语言对患者和家属进行访谈,应用标准化方法对患者进行详细查体,记录患者症状和体征,进行神经心理学评估并行头MRI扫描。根据Erkinjuntti提出MRI诊断标准最终入选SIVD患者70例,根据神经心理学评估分为NCI组和VCI组。应用相关分析和分层多元回归分析认知功能与LI数量和部位的关系,以及WMH和LI的相互影响。结果:VCI组LI总平均数为(4.38±1.82),较NCI组的(3.39±1.71)明显升高,具有统计学显著性差异(Z=-2.202,P=0.028)。VCI组额区和基底节区LI数量均较NCI组明显升高,具有显著性差异(P<0.05);而顶枕区、颞区和幕下区LI数量两组无显著性差异,分层多元线性回归分析以简明精神状态量表得分为因变量,第一个模型:分析LI数量对认知损害的影响,首先进入模型的年龄、性别和受教育程度结果同前。当第二步LI总平均数进入模型后,简明精神状态检查得分仍与受教育程度有关(P<0.001)。在校正年龄、性别与受教育程度的影响后, LI数仅能解释简明精神状态评分改变的9.1%。第二个模型:分析LI部位对认知损害的影响;第一步同前,第二步不同部位LI同时进入模型,结果显示LI部位能解释简明精神状态评分改变的25.7%,明显高于第一模型中LI数量的作用。其中额区和基底节区LI与MMSE得分呈负相关,具有统计学显著性差异(P<0.05),而顶枕区、颞区和幕下区LI与MMSE得分无统计学差异。第三个模型:分析WMH体积和LI数量对认知损害的影响;在校正年龄、性别与受教育程度的影响后,WMH体积能解释简明精神状态评分改变的10.2%;而LI数量进入模型后R2变化0.112,高于LI数量独自进入模型的R2值(R2=0.091),与MMSE的相关性均有统计学显著性差异(P<0.001),两者均为SIVD患者认知损害的预测指标。第四个模型:分析白质病变部位和腔隙性脑梗死部位对认知损害的影响:结果显示仅基底节LI和额区缺血性WML为认知损害的独立预测因素(P<0.001)。结论:LI数量和部位均与SIVD患者的认知损害有关,LI部位在SIVD认知功能损害中的作用比LI数量更大。LI数量和WMH体积均为SIVD认知损害的独立危险因素。校正年龄、性别、受教育程度、WML和LI部位后,仅基底节区LI和额区WML为SIVD患者认知功能损害的独立预测指标。LI数量增加和缺血性脑白质病变程度加重,尤其是基底节区腔梗增加和额叶WML程度加重将预示认知功能损害的恶化。更多还原

【Abstract】 Vascular dementia (VD) is the second most common cause of dementia in elderly after Alzheimer’s disease (AD). It results from ischaemic, ischemic-hypoxic, or haemorrhagic brain injury that cuased by cerebrovascular disease, manifests a series of cognitive impairment syndromes. With the deeper development of cognitive impairment research, the criteria for VD based on AD presents its shortcomings. Based on the model of AD, the old criteria for VD emphasizes on new knowledge learning disability and memory loss, but not on psychomotor slowing and impairment of executive functioning which is the clinical features of the typical VD form. It also defines dementia as the level of cognitive impairment at which normal daily functions are impaired, therefore only late cases will be identified, so underestimating the prevalence of cognitive impairment due to vascular disease and denying early cases the benefit of timely preventative treatment. Vascular cognitive impairment (VCI), a new concept covering that of VD suggested by Hachinski and Bowler in 1993, is now widely accepted as a more appropriate alternative terminology than VD.Cognitive impairment that is caused by or associated with vascular factors has been termed as VCI. It covers a wide spectrum of cognitive dysfunction associated with and presumed to be caused by cerebrovascular disease, what most important is that it included subtle and clinically often undetected cognitive impairment. VCI presents a clinical syndrome. The main subtypes of VCI included in current classification are cortical VCI, strategic infarct VCI and subcortical VCI. The clinical manifest and treatment of the subtype are different from each other for its difference at etiological aspects and pathogenesis. Because vascular risk factors are treatable, it should be possible to prevent, postpone, or mitigate VCI. However, progress in VCI research has been hindered by lack of unified and satisfactory diagnostic criteria for the condition. Research emphasis of VCI gradually turned to the relatively homogeneous subcortical VCI.Subcortical ischemic vascular disease (SIVD) is the most common cause of subcortical VCI. It is accepted now that the pathogenesis of SIVD related to cerebral small vessel disease caused by multifactor. Two paths may be involved. First, occlusion of the arteriolar lumen due to arteriolosclerosis leads to the formation of lacunar infarcts (LI); second or critical stenosis and hypoperfusion of multiple medullary arterioles causes widespread incomplete infarction of deep white matter. Symptoms include motor and cognitive dysexecutive slowing, forgetfulness, dysarthria, mood changes, urinary symptoms, and short-stepped gait. Hypertension serves as the main risk factor of SIVD. Other vascular risk factor of cerebrovascular disease, such as diabetes millitus, hyperlipidemia, heart disease, anemia, smoking and excessive drinking, rise the incidence of SIVD. VCI can not be diagnosed to all patients with SIVD. XU et al reported that total of 44.5% patients with SIVD presented cognitive impairment at a successive clinical research. There were few reports considering the relationship between cognitive impairment and risk factor of SIVD. It is benefit to VCI patients’timely preventively treatment to further study on the risk factor of subcortical VCI and identify the relationships between cognitive impairment and vascular risk factors of SIVD.SIVD is characterised by extensive cerebral white matter lesions (WML) and lacunar infarcts in deep grey and white matter structures. WML, which be called“leukoaraiosis”in the past, shows "white matter hyperintensities (WMH) on T2 and fliud attenuated inversion recovery (FLAIR) magnetic resonance imaging (MRI) sequence. Researches of the relationship between cognitive impairment and the degree or type of lesions on MRI had been going on for the past many years, but no definite consensus had been to reach. There were not significant difference between leukoaraiosis and cognitive impairment at initial research reports. Those differet results some extent due to the qualitative analysis used at their research. With the progress of technique on automated image segmentation to quantitation and location, researches on relationships between cognitive impairment and neuroimaging changes are stepping into a new era. Bowler reviewed reports considering WML and cognitive impairment in the past few years, and mentioned that“attempts to correlate closely infarct volumes with cognitive change routinely fail because of the overriding importance of lesion location”. Combinated quantitation with qualitation analysis, it is possible to gain reliable data which is benefit to early cases diagnosis of cognitive impairtment in SIVD, and even to other types of VD.The following 4 parts were designed to study on the risk factor, relationship between LI, WML and cognitive impairtment in SIVD.Part 1: Clinical Feature and Risk Factor of Cognitive Impairment in Patients with Subcortical Ischemic Vascular DiseaseObjectives:To investigate the clinical feature, the risk factor for cognitive impairment of SIVD, and the risk factor for its subtypes: white matter lesions type (WML-SIVD) and lacunar infarct type (LI-SIVD).Methods: Patients who had ever suffered at least an acute ischemic stroke for more than 3 months were invited to the pre-screening. After comprehensive clinical, neuropsychological profiles evaluation and magnetic resonance imaging (MRI) scanning, 70 patients with SIVD diagnosed according to the MRI criteria of Erkinjuntti and neuropsychological profiles were divided into no cognitive impairment(NCI) and vascular cognitive impairment(VCI) group. Symptomes and signs, vascular risk factor profiles were recorded by interviewers and examiners, then risk factor profiles for cognitive impairment of SIVD or its subtypes were compared using univariate and multivariate logistic regression analysis.Results: The subjects consisted of 33 males and 37 females; the mean age was 66.89±7.03 years, the mean educational attainment was 2.81±2.37 years. Mini-mental status examation (MMSE) scores of VCI group (19.81±2.56) was significant lower than that of NCI group (24.45±2.95). Of the 70 SIVD patients, 81.43%(57) patients complainted of limbs weakness, 40% (28) of instability walking, 20%(14) of mild dysarthria, and 14.29%(10) of difficulty with drinding or urinary incontinence. On nervous system examination, 71.4% patients with upper motor neuron damaged signs, 14.9% with pseudobulbar palsy signs and 10% with extrapyramidal signs were detected, while only 7.14% patients with ataxia gait。On comparison the subtype of SIVD, univariate analysis demonstrated significant associations between WM and hypertension (P<0.05), and between LI and diabetes mellitus, hyperlipidemia, coronary artery disease (P<0.05), but no significant associations with anemia, smoking, alcohol drinking (P>0.05). These associations persisted after multivariate stepwise Logistic analysis, Hypertension was more frequent in WML-SIVD than LI-SIVD [odds ratio (OR) 8.531 (1.676 to 43.41); P=0.008], whereas diabetes mellitus [OR 0.082 (0.016 to 0.436); P=0.003], hyperlipidemia [OR 0.158 (0.035 to 0.720) ; P=0.019], coronary artery disease [OR 0.005 (0.007 to 0.336); P=0.002] more frequent in LI-SIVD. As to the risk factor for cognitive impairment, hypertension [OR 5.265 (1.563 to 17.731); P=0.007], diabetes mellitus [OR 3.445 (1.008 to 11.772); P=0.049] and hyperlipidemia [OR3.649 (0.974 to 11.466); P=0.027] were associated with VCI patients, even in the multivariate stepwise Logistic analysis after controling for the effects of other variables, but no correlations was found between coronary artery disease, anemia, smoking, alcohol drinking and cognitive impairment.Conclusions: Limbs weakness, instability walking, mild dysarthria, urinary incontinence, difficult with drinking, upper motor neuron damaged signs and pseudobulbar palsy are common clinical manifestation of subcortical ischemic vascular disease, although extrapyramidal signs may be detected. Hypertension associates with WML of SIVD, while diabetes mellitus, hyperlipidemia, coronary artery disease associates with LI-SIVD. Hypertension and diabetes mellitus are the main risk factors for cognitive impairment of SIVD, and the risk increase 5.265 and 3.445 times, respectively. The relationship between hyperlipidemia and cognitive impairment couldn’t be established from our data. It is likely that age is not correlated with cognitive impairment of SIVD patients.Part 2: Contributions of White Matter Lesions Volume to Subcortical Vascular CognitiveObjectives:To evaluate contributions of WML to cognitive impairment of patients with subcortical ischemic vascular disease (SIVD).Methods: Patients who had ever suffered at least an acute ischemic stroke for more than 3 months were invited to the pre-screening. After comprehensive clinical, neuropsychological profiles evaluation and magnetic resonance imaging (MRI) scanning, 70 patients with SIVD diagnosed according to the MRI criteria of Erkinjuntti and neuropsychological profiles were included and divided into NCI and VCI groups. The volume of WML, which shows white matter hyperintensities (WMH) on MRI fluid-attenuated inversion recovery images, was measured using a semi-automated quantitative method. Finally, correlation analysis and hierarchical multiple regression analysis was used to examine the relationship between general cognitive function and the volume of WMH.Results: The mean WMH volume of all SIVD patients was (29.56±15.58)cm3, ranged from 4.20cm3 to 86.01cm3. The mean WMH volume of VCI group was (34.84±17.50) cm3, increased remarkedly than that of NCI group (23.63±10.52) cm3, there was significant difference between groups(Z=-2.83, P=0.005). MMSE score was positively correlated with educational attainment(r=0.49, P<0.001), but negatively correlated with age(r=-0.235, P=0.025) and WMH volume(r=-0.398, P<0.001) in a univariate correlation analysis. And age was positively correlated with WMH volume(r=0.279, P=0.01), but negatively with educational attainment. In the hierarchical multiple regression analysis, age, gender and educational attainment was entered at the first step, but only the educational attainment was related to MMSE score(P<0.001). After the WMH volume entered at the second step, the relationship remain significant between MMSE score and eductional attainment(P<0.001). It was revealed that WMH volume was negatively correlated with MMSE score(B=-0.077, t=-3.243, P=0.002) after controlling for the effects of age, gender and educational attainment, WMH volume explained an additional 10.2% of the variance in MMSE.Conclusions: It is convenient to measure and count the volume of white matter hyperintensities using the semi-automated quantitative method which is based on ImageJ and its Voxel Counter Plugins. The result is accurary and reliable, the method is time-saving and easy to manipulate. There is a highly significant negative correlation between MMSE score and WMH volume, but WMH volume explains only an additional 10.2% of the variance in MMSE independently.Part 3: Contributions of White Matter Lesions Location to Subcortical Vascular Cognitive ImpairmentObjectives:To evaluate contributions of WML location to cognitive impairment of patients with subcortical ischemic vascular disease (SIVD), and to evaluate the correlation of volumetric quantitation and qualitation of WMH .Methods: Patients who had ever suffered at least an acute ischemic stroke for more than 3 months were invited to the pre-screening. After comprehensive clinical, neuropsychological profiles evaluation and magnetic resonance imaging (MRI) scanning, 70 patients with SIVD diagnosed according to the MRI criteria of Erkinjuntti and neuropsychological profiles were included and divided into no cognitive impairment(NCI) and vascular cognitive impairment(VCI) group. Age-related white matter change rating scale, which is a visual rating scale developed by Wahlund, was used to qualitative measure and locate the WML. Integrated with the data of part 2, using correlation analysis and hierarchical multiple regression analysis to examine the relationship between general cognitive function and the location of WMH, and the correlation of volumetric quantitation and qualitation.Results: The mean WMH rating scores of all the SIVD patients was (12.37±3.93), ranged from 3 to 20. There was a highly significant positively correlation between total WMH volume and total WMH score(r=0.879, P<0.001), scatter diagram using WMH volume as independent variables reveal a curvilinear relationship between total WMH volume and total WMH score. The mean scores of WMH in all regions of VCI group was (13.65±3.51), higher than that of NCI group(t=-3.043,P=0.003), which was (10.94±3.94). As to each regions, the VCI group scores of frontal area, parieto-occipital area and basal ganglia area were remarkedly increased than that of NCI group(P<0.05), but no difference were noted at temporal area and infratentorial area. In the hierarchical multiple regression analysis, The result of first step was same to part 2. After the score of total WMH entered at the second step, the relationship remained significantly between MMSE score and eductional attainment(P<0.001). It was revealed that the score of total WMH was negatively correlated with MMSE score(B=-0.270,t=-2.842, P=0.006) after controlling for the effects of age, gender and educational attainment. The score of total WMH explained an additional 8.1% of the variance in MMSE,lower than the 10.2% of WMH volume. Moreover, when all the five regions enter instead of the score of total WMH at the second step, the R2 change was 19.3%. That is, all the five regions explained an additional 19.3% of the variance in MMSE,much higher than the effects of WMH volume and the score of total WMH. And the scores of frontal area and of basal ganglia area were the mainly variables(P<0.05), but no the scores of parieto-occipital area, infratentorial area and temporal area(P>0.05).Conclusions: The volumetric quantitation of WMH was positively correlated with its volumetric qualitation. Each of the methods can be used to tract any changes of WML, except that the volumetric quantitation is more sensitivity. WML located at frontal area and basal ganglia area predicts cognitive decline, but not suits for WML located at parieto-occipital area, infratentorial area and temporal area.Part 4: Contributions of Lacunar Infarcts to Subcortical Vascular Cognitive Impairment Objectives:To explore contributions of LI to cognitive impairment of patients with subcortical ischemic vascular disease (SIVD), and to evaluate the relationship of LI and WML and effects of each other.Methods: Patients who had ever suffered at least an acute ischemic stroke for more than 3 months were invited to the pre-screening. After comprehensive clinical, neuropsychological profiles evaluation and MRI scanning, 70 patients with SIVD diagnosed according to the MRI criteria of Erkinjuntti and neuropsychological profiles were included and divided into NCI and VCI groups. The number of LI was counted and the location of it was recorded according to the method using at part 3. Integrated with the data of ahead, using correlation analysis and hierarchical multiple regression analysis to examine the relationship between general cognitive function and the number and the location of LI, and the relationship of LI and WML.Results: The mean number of LI in VCI group was (4.38±1.82), higher than that of NCI group(Z=-2.202,P=0.028), which was (3.39±1.71). As to each regions, the VCI group number of LI in frontal area and basal ganglia area were remarkedly increased than that of NCI group(P<0.05), but no difference was noted at temporal area, parieto-occipital area and infratentorial area. In the hierarchical multiple regression analysis, considered MMSE scores as dependent variables, the first step is same as part 2 and part 3. After the number of total LI entered at the second step, the relationship remain significant between MMSE score and eductional attainment(P<0.001). It was revealed that the number of total LI was negatively correlated with MMSE score(B=-0.270,t=-2.842, P=0.006) after controlling for the effects of age, gender and educational attainment, the number of total LI explained an additional 9.1% of the variance in MMSE. Another analysis then performed, all of the five regions enter instead of the number of total LI at the second step, the R2 change was 25.7%. That is, all of the five regions explained an additional 25.7% of the variance in MMSE,much higher than the effects of the number of total LI. And the number of LI at frontal area and basal ganglia area were the mainly negative variables(P<0.05) to predict cognitive decline of patients with SIVD, but no relationship was noted between MMSE and the number of LI at parieto-occipital area, infratentorial area or temporal area (P>0.05). The third hierarchical multiple regession analysis followed, at the first step, age, gender and education were entered; then WMH volume were entered at the second step, the result was same as part 2, and WMH volume make the model R2 change 10.2%; the number of total LI enter after WMH volume at step 3, an additional R2 change was 0.112, there was a negatively correlation between MMSE and WMH volume, MMSE and the number of LI, independently. At last, the fourth analysis performed, the location of WML and LI were entered the model at the second step after age, gender and education, it revealed that only WML located at frontal area and LI located at basal ganglia area predicted cognitive decline of patients with SIVD, independently.Conclusions: The number and location of LI were related to cognitive impairment, the effect of LI location is greater than than of its number. The number of LI and WMH volume both are predictors of cognitive impairment in patients with SIVD, independently; and LI located at basal ganglia area and WML at frontal area are the best predictors.更多还原