第一部分 血管危险因素与不同类型痴呆

【研究背景与目的】既往研究证实多种血管危险因素包括高血压、糖尿病、高脂血症和吸烟等与全因认知功能下降密切相关；血管危险因素参与许多痴呆病因的发生、发展。了解血管危险因素在不同类型痴呆中的作用有助于理解疾病的机制及预防措施，甚至作为治疗靶点。基底节区密集点灶病灶，又称作“奶酪征”，被认为提示潜在的小血管机制，不仅仅见于脑小血管病，在阿尔茨海默病（Alzheimer’s Disease, AD）、额颞叶变性（Frontotemporal Lobar Degeneration, FTLD）等神经系统退行性疾病中也可见到。本研究目的是描述各血管危险因素在北京协和医院痴呆队列中不同类型痴呆患者的发生率，描述奶酪征的影像学特征和危险因素。【研究方法】本研究是一项回顾性研究。纳入2007年至2022年就诊于北京协和医院痴呆与脑白质病门诊，并加入北京协和医院痴呆队列的痴呆患者，同时收集临床资料。（1）痴呆队列：依照相应诊断标准对患者进行痴呆病因分类，包括AD、FTLD、血管性认知功能障碍等，对比不同群体高血压、糖尿病、高脂血症及脑卒中的发病率，描述奶酪征在各疾病人群中的检出率。（2）根据奶酪征的有无将全因痴呆人群分为两组，对比探讨奶酪征的危险因素；半定量统计描述基底节区不同性质病灶（基底节高信号、血管周围间隙、腔隙及陈旧性梗死）数量，对比不同程度奶酪征中各病灶的占比。【研究结果】本研究共纳入812例痴呆患者，其中生物学诊断的AD 133例，FTLD 68例，淀粉样脑血管病（Cerebral Amyloid Angiography，CAA）36例，伴皮质下梗死和白质脑病的常染色体显性遗传性脑动脉病（Cerebral Autosomal Dominant Arteriopathy with Subcortical Infarcts and Leukoencephalopathy，CADASIL）21例，其他类型脑小血管病（Cerebral Small Vessel Disease，CSVD）95例。在AD、FTLD、CAA、CADSIL中高血压的发生率为27.7%、20.59%、44.44%、42.86%；糖尿病发生率为9.02%、10.59%、8.33%、4.76%；高脂血症发生率为27.82%、22.06%、19.44%、9.53%；脑血管事件在AD和FTLD中的发生率为3.01%、5.88%。奶酪征检出率在AD、FTLD组分别为2.25%、8.82%，CADASIL组9.25%；在CAA组（27.8%）和其他CSVD组（50.52%）中最高。奶酪征的危险因素为年龄、高血压、脑血管事件；其主要构成为基底节高信号、血管周围间隙，偶见腔隙、梗死灶。此外，我们建立了奶酪征的分级方法，在不同程度的奶酪征中，基底节高信号仍为主要成分（轻度组69.96%，中度组62.58%，重度组57.11%）；但血管周围间隙占比与奶酪征程度成正相关（轻度组22.53%，中度组29.94%，重度组37.37%，P=0.006）；而腔隙/梗死在三组间无统计学差异。【研究结论】认知障碍患者常存在多种共病，包括高血压、糖尿病和高脂血症，以及脑血管事件。但上述共病在不同痴呆疾病的人群中发生率存在较大差异，如高血压在AD、FTLD及非动脉粥样硬化性脑小血管病中的出现率在波动在27.7-44.4%。奶酪征可见于脑小血管病及其他变性病痴呆，主要危险因素为年龄、高血压及脑血管事件；影像学主要构成为基底节高信号及血管周围间隙，偶见腔隙/梗死或微出血，基底节高信号占比最高，而血管周围间隙占比与奶酪征程度相关。我们的研究结果不再支持把奶酪征简单等同于密集血管周围间隙。

第二部分 血管危险因素对早老性阿尔茨海默病的影响

【研究背景与目的】血管危险因素包括高血压、糖尿病、高脂血症等是阿尔茨海默病（Alzheimer’s Disease, AD）主要可调节的危险因素，通过影响大脑灌注、血脑屏障破坏、血管微结构损伤等影响β淀粉样蛋白（Amyloid-β，Aβ）的清除及其他炎症机制促进AD的发生及发展。老年性AD和早老性AD可能存在不同发病机制，目前血管危险因素在AD中的作用还存在争议，针对血管危险因素在早老性AD人群的研究也较少。本研究旨在对比早老性AD有/无合并血管危险因素人群间的临床及生物学特征，探索血管危险因素在早老性AD中的作用。【研究方法】本研究为回顾性研究。对2007年至2022年就诊于北京协和医院痴呆门诊，符合2011年NIA-AA临床诊断标准的AD患者进行生物标志物诊断：1）脑脊液生物标志物（Aβ42、Aβ40、T-tau、P-tau）检测；或2）Aβ淀粉样蛋白PET；或3）全外基因组测序筛选APP/PSEN1/PSEN2突变。纳入脑脊液生物标志物符合AD病理改变或携带APP/PSN1/PSN2致病突变者进行进一步的血管相关因素分析。采集所有入组患者临床资料，对所有患者进行影像学评估，包括Fazekas评分、血管周围间隙、微出血分级、颞叶萎缩评分（Medial Temporal Lobe Atrophy，MTA）及后部皮层萎缩Koedam评分；部分患者完善脑脊液生物标志物检测，利用联影系统测量入组患者各脑区体积。按照有无合并血管危险因素（高血压、糖尿病、高脂血症和脑卒中）将早老性AD患者分为有血管性危险因素组和无血管性危险因素组两组，对比两组间发病年龄、基线认知水平、脑脊液标志物水平及影像学特征。【研究结果】共纳入133例AD患者，其中符合早老性痴呆108例（男性48例，女性60例）。合并血管危险因素（阳性组）46例，平均起病年龄55.03±5.86岁，未合并血管危险因素（阴性组）62例，平均起病年龄54.41±6.20岁。两组人群在起病年龄、病程、教育年限、认知水平、APOEε4携带情况均无明显差异。两组人群在各脑脊液标志物水平、影像学特征（白质高信号、血管周围间隙、微出血及颞叶/后部萎缩程度）均无差异。基于3D-T1序列测定不同脑区、脑脊液体积；两组人群总脑体积无明显差异，但阴性组较阳性组灰质总体积更小，而脑脊液体积及占比包括外周脑脊液、三脑室体积较阳性组更高（外周脑脊液占比：P=0.006，三脑室体积占比：P=0.042）。【研究结论】高血压、糖尿病、高脂血症及脑血管事件对早老性AD的起病年龄、疾病严重程度、脑脊液生物标志物及影像学特征并无明显影响。不伴血管危险因素的AD患者脑脊液体积及三脑室体积较合并危险因素的AD组更大，占全脑容积百分比更高。第一部分 血管危险因素与不同类型痴呆

第三部分 血管危险因素对额颞叶痴呆的影响

【研究背景与目的】额颞叶变性（Frontotemporal Lobar Degeneration, FTLD）是最为常见的早老性痴呆病因之一；约40%患者由基因变异导致，遗传因素在FTLD发病机制中起着重要作用。而散发性患者致病机制尚不清楚；对FTLD危险因素的探索有助于提高对疾病的认识和预防。年龄相关的血管危险因素参与诸多神经系统退行性变包括阿尔茨海默病和路易体痴呆，但对于FTLD相关研究较少。本研究旨在基于北京协和医院痴呆队列探索血管危险因素在FTLD人群中的影响。【研究方法】本研究为单中心回顾性研究，纳入2007年至2022年北京协和医院痴呆与脑白质专病门诊就诊的FTLD患者，符合很可能的/确诊的行为变异性FTD（Behavioral Variant Frontotemporal Dementia，bvFTD）及很可能的/确诊的原发性进行性失语（Primary Progressive Aphasia，PPA）。对所有患者详细采集一般资料及既往史情况，完善神经心理学及影像评估，部分患者完善腰椎穿刺送检脑脊液标志物检测。按是否合并高血压、高脂血症、糖尿病及脑血管事件将FTLD患者分为两组，对两组的临床及生物学特征进行比较。【研究结果】本研究共纳入66例FTLD患者，平均起病年龄55.94±10.11岁；其中34例诊断很可能的bvFTD，32例诊断为PPA，并有6例患者合并肌萎缩侧索硬化。合并血管危险因素患者28例（阳性组），未合并血管危险因素38例（阴性组）。阴性组起病年龄较小(53.76±11.39岁，P=0.041)，日常生活能力受损程度更重（40.26±12.23分，P=0.049）。两组患者影像学所示白质高信号、腔隙、血管周围间隙、颞叶萎缩评分（Medial Temporal Lobe Atrophy，MTA）和Koedam评分均无明显差异。共48例FTLD患者完善脑脊液生物标志物检测，Aβ平均水平为710.5±307.7pg/ml，T-tau为183.6（115.8，362.2）pg/ml；各脑脊液生物标志物水平组间并无差异。28例FTLD患者完善脑体积测定，血管危险因素阴性及阳性组间总脑体积及各脑叶体积并无差异。【研究结论】FTLD按有无血管危险因素分组对比，阳性组患者发病年龄小，日常生活能力受累重，而两组人群在认知水平、脑脊液生物标志物、脑小血管病相关影像特征和不同脑区体积无明显差异；血管危险因素在FTLD中的作用还需进一步探索。

Part 1. Vascular Risk Factors in Dementia with Various Etiology

[Background and objective] Previous studies indicated vascular risk factors factors including hypertension, diabetes, hyperlipidemia and smoking are related to all-cause cognitive impairment. vascular risk factors increase the risk of dementia and improve its development. Dense focal lesions in basal ganglia, called ‘cheese sign’, was seen as a marker of small vascular mechanism, they can also be detected in patients with degenerative diseases such as AD and FTLD. The aim of this study was to describe the incidence rate of hypertension, diabetes, hyperlipidemia and stroke in different groups, and to describe the characteristics and risk factors of cheese sign based on PUMCH dementia cohort. [Methods] This study was a retrospective study. Patients were enrolled consecutively from the dementia cohort, who visited the dementia clinic of our hospital from 2007 to 2022. (1) Dementia cohort was classified into different groups based on various causes including AD, FTLD, vascular cognitive impairment and others. The incidence rates of vascular risk factors and cheese sign in groups were described. (2) According to the presence or absence of cheese sign, people with all-cause dementia were divided into two groups. We use semiquantitative method to calculate the number of different lesions (basal ganglia hyperintensities, perivascular spaces, lacunae, and chronic infarcts) in basal ganglia. The risk factors and composition of cheese sign were identified. [Results] A total of 812 patients were included, 133 diagnosed with biological Alzheimer’s disease, 68 with FTLD, 36 with probable/possible CAA, 21 with definite CADASIL, and 95 with other CSVD. The prevalence of hypertension in AD, FTLD, CAA and CADASIL group were 27.7%, 20.59%, 44.44% and 42.86%; the prevalence of diabetes were 9.02%, 10.59%, 8.33% and 4.76%; the prevalence of hyperlipidemia were 27.82%, 22.06%, 19.44% and 9.53%. Stroke happened in 3.01% AD and 5.88% FTLD patients. The detection rates of cheese sign were 2.25% in AD, 8.82% in FTLD, 9.25% in CADASIL group and highest in CAA group (27.8%) and other CSVD group (50.52%). The risk factors of cheese sign were age, hypertension and cerebrovascular events. It mainly consists of basal ganglia hyperintensities and perivascular spaces, occasionally lacunae and chronic infarcts. We established a grading method for cheese sign. Basal ganglia hyperintensity remains the main component in different degrees (mild group 69.96%, moderate group 62.58%, severe group 57.11%). However, the proportion of perivascular spaces was positively correlated with the degree of cheese sign (mild group 22.53%, moderate group 29.94%, severe group 37.37%, P=0.006). There was no significant difference in proportion of lacuna/infarction among the three groups. [Conclusion] Vascular risk factors including hypertension, diabetes and hyperlipidemia and stroke can be seen in degenerative diseases. The prevalence differs in various dementia groups. The prevalence of hypertension was 27.7%-44.4% in AD, FTLD and non-atherosclerotic CSVD. Cheese sign can be seen in cerebrovascular diseases and other degenerative disorders. The main risk factors of cheese sign were age, hypertension, and cerebrovascular events. The main composition of cheese sign was basal ganglion hyperintensity, and the proportion of perivascular spaces was related to the degree of cheese sign. This study no longer supports the simple equivalence of cheese sign to dense perivascular spaces. 

Part 2. Effect of Vascular Risk Factors in Presenile Alzheimer’s Disease

[Background and objective] Vascular risk factors (hypertension, diabetes, hyperlipidemia, etc.) are main adjustable risk factors of AD. These factors influence cerebral perfusion, blood-brain barrier destruction, vascular microstructure damage, β-Amyloid clearance and other inflammatory mechanisms, then promote onset and course of AD. Previous studies indicated heterogeneity in early and late-onset AD. The role of vascular risk factors in early-onset AD remains unclear. The purpose of this study was to compare the clinical and biological characteristics of presenile AD patients with and without vascular risk factors. [Methods] This study is retrospective. Biological diagnosis was performed on AD patients who met the 2011 NIA-AA clinical diagnostic criteria from cohort who visited the Dementia Clinic of our hospital from 2007 to 2022: 1) Cerebrospinal fluid (CSF) biomarkers (Aβ 42, Aβ40, T-tau, P-tau) detection; or 2) amyloid protein PET; or 3) Whole genome sequencing screening for APP/PSEN1/PSEN2 mutations. Patients whose CSF biomarkers met the criterion of AD pathological changes or carrying APP/PSN1/PSN2 pathogenic mutations were enrolled for further study of vascular risk factors. Detailed clinical data were collected and head MRI performed. Imaging data were analysed including cerebral small vessel disease markers, MTA and Koedam score. Test of Cerebrospinal fluid biomarkers (Aβ-42, T-tau, P-tau) and 3D-T1 imaging for brain volume measurement were performed in partial patients. These patients were divided into two groups according to the presence or absence of vascular risk factors (hypertension, diabetes, hyperlipidemia and stroke). The age of onset, baseline cognition, levels of cerebrospinal fluid markers, and imaging characteristics were compared. [Results] A total of 133 AD patients were enrolled, of which 108 were early-onset (48 males and 60 females). 46 patients with vascular risk factors (positive group), with an average onset age of 55.03±5.86 years; 62 cases without vascular risk factors (negative group), with an average onset age of 54.41±6.20 years. Negative group and positive group showed no significant difference in gender, age, course of disease, education, cognitive examination, APOE genotype and imaging features. 96 patients performed lumbar puncture, the levels of Aβ-42, T-tau, P-tau and ratio of T-tau/Aβ were similar in negative and positive groups. Volumes of various regions of brain and CSF were measured based on 3D-T1 imaging. The volume of gray matter in the negative group was smaller than that in positive group. The proportion of CSF volume including peripheral CSF (P=0.006) and the third ventricle (P=0.042), increased significantly in negative group than positive group. [Conclusion] Hypertension, diabetes, hyperlipidemia and stroke showed no significant impact on clinical feature, cerebrospinal fluid biomarkers and imaging characteristics of early-onset AD. AD patients without vascular risk factors had larger volume of cerebrospinal fluid, peripheral cerebrospinal fluid and third ventricle. 

Part 3. Effect of Vascular risk factors in Frontotemporal lobar degeneration

[Background and Objective] Frontotemporal lobar degeneration (FTLD) is one of the most common causes of presenile dementia. About 40% of FTLD are caused by genetic mutations, and genetic factors play an important role in the pathogenesis of FTLD. The mechanism and modified risk factors of sporadic FTLD are still unclear. Age related Vascular risk factors factors are involved in many neurodegenerative disorders like AD and DLB, but there are few studies in FTLD. The purpose of this study was to explore the impact of vascular risk factors in the FTLD based on PUMCH dementia cohort. [Method] This study is a single center retrospective study. Patients met probable bvFTD and imaging-confirmed PPA diagnosis were enrolled, some were definite FTLD with pathogenic mutation. Detailed general data and medical history were collected for all patients, and neuropsychological and imaging assessments were performed. Some patients underwent lumbar puncture for CSF marker test. FTLD patients were divided into two groups according to whether they were complicated with hypertension, hyperlipidemia, diabetes and cerebrovascular events. The clinical and biological characteristics were compared between two groups. [Results] A total of 66 patients with FTLD were enrolled in this study, with an average onset age of 55.94 ± 10.11 years. 34 cases were diagnosed with bvFTD and 32 cases with PPA, and 6 cases were complicated with amyotrophic lateral sclerosis. 28 patients with Vascular risk factors were divided into positive group and 38 cases without Vascular risk factors negative group. The average age at onset in negative group was younger (53.76 ± 11.39 years old, P=0.041), the impairment of daily living ability was more severe (40.26 ± 12.23, P=0.049). There were no differences in Fazekas score, perivascular space degree, MTA, and Koedam score between the two groups. 48 patients performed lumbar puncture, average Aβ level was 710.5 ± 307.7 pg/ml, and T-tau 183.6 (115.8, 362.2) pg/ml. There was no difference in the levels of cerebrospinal fluid biomarkers between groups. Brain volume measurements were completed in 28 FTLD patients. Total brain volume and each lobar volume also showed no difference between the two groups. [Conclusion] There was no significant difference in cognitive impairment, cerebrospinal fluid biomarkers, imaging features related to cerebral small vessel disease, and brain volumes. Patients with vascular risk factors had smaller age at onset and higher ADL. The role of Vascular risk factors in FTLD needs further researches.