痴呆是一种以记忆与认知功能渐进性衰退为主要特征的疾病，其发病风险随年龄增长而增加，严重危害老年人群的日常生活能力，阿尔茨海默病（Alzheimer’s Disease, AD）与血管性痴呆（Vascular Dementia, VaD）是两种最常见的痴呆类型。痴呆不仅给患者、家庭与医护人员带来巨大的压力和痛苦，还给社会增加了巨大的经济负担。然而，目前有关AD的确切病因尚不清楚，学界普遍认同的淀粉样级联假说使Aβ成为主要的治疗靶点，尽管在新药研发领域针对该靶点投入了大量的资源，但几乎所有的以Aβ为靶点的AD临床试验都相继失败，目前还没有有效的治疗手段。同时，由于VaD的疾病分类和诊断标准的不确定性，VaD缺乏可识别的、易于处理的治疗靶点，至今没有得到许可的治疗方法。因此，迫切需要重新思考针对这种危害性巨大的疾病的治疗目标和干预策略。

目前，大多数由多种因素调控的复杂性疾病都采用三级预防战略进行管理，从而有效调控疾病的进展以改善患者的预后与生活质量。这种阶段性的分级治疗模式已经在脑卒中的干预治疗方面取得了相当大的成功，例如，仅针对高血压这一项血管危险因素的有效控制就能降低脑卒中的40%的发病率。痴呆作为一种复杂的慢性疾病，考虑到其日益增长的患病人数以及难治性，针对不同疾病阶段的多维度干预管理可能是一种有前途的干预策略，国际阿尔茨海默病协会也提出了针对老年痴呆风险因素的多级干预，强调了通过调整生活方式与控制风险因素来预防认知衰退的可能性。

因此，为了阻止痴呆的流行，如何有针对性地防控痴呆的危险因素，实施有效的干预措施是至关重要的。血管危险因素（Vascular Risk Factor, VRF），作为诱发痴呆发病的重要风险，主要包括糖尿病，高血压，高脂血症，吸烟和肥胖，是可提前干预并有潜力延缓甚至逆转痴呆发病的重要可调节因素。尽管血管危险因素和痴呆之间存在既定的联系，但两者之间的机制联系尚未确定。因此，全面的探索，有效的评估血管危险因素对大脑认知功能的影响和对大脑结构的改变逐渐成为近年研究的热点。厘清血管危险因素与脑健康之间的关系既可以作为认知障碍相关疾病的预警，也可以作为干预的目标，从而为维护老年人的脑健康提供了一个实用性强、可普及率高且相对经济的战略基础。

尽管针对该领域已有了较为广泛的研究，但至少两个关键问题尚未得到很好的解决。首先，不同血管危险因素对认知功能障碍、脑结构损伤的风险权重差异尚未得到充分研究。由于每一个血管危险因素对各领域认知功能以及大脑结构指标的损伤权重并不相同，因此，应当明确各个血管风险因素的风险权重，才能将干预重点放在风险权重更大的因素，从而使针对认知功能障碍相关疾病的预防和治疗更加有效和经济。第二，在实际临床情况下，个体可能协同合并有两个或多个风险因素，这种交互作用的模式对脑健康可能存在特定的影响，还没有研究对血管风险因素的不同负担组合进行充分的研究和比较。

因此，基于以上两个问题，本研究假设每一个血管危险因素或不同的血管危险因素组合的风险权重存在差异，并会表现出特定的认知功能和脑结构损伤模式，应用北京社区老年脑健康（BABRI）计划基线行为数据与多模态磁共振数据，从多个维度描述血管危险因素对各领域认知功能以及大脑灰质、白质结构和结构网络影响的特征模式。本研究共分为3个部分。

研究1. VRF对高级认知功能的影响 本研究通过对BABRI计划共4667名社区非痴呆受试者实施测评，利用社区卫生站医疗记录或口头报告记录的形式确定受试者个体糖尿病、高血压、高脂血症患病状态以及肥胖、吸烟的情况；利用成套神经心理学量表评估受试者认知能力损伤与否，其中包括一般精神状态、记忆功能、视空间能力、语言能力、注意功能和执行功能；其次，将每一领域认知得分换算为标准化Z分数，计算不同VRF合并模式下认知功能的表现。结果发现，糖尿病和高血压是认知功能的风险因素，尤其是在整体认知功能与记忆领域，糖尿病相对于高血压权重更大。相比之下，高脂血症（其中87.1%规律服用他汀类药物）对认知功能的影响表现为保护作用。吸烟和肥胖对认知功能影响并不显著。此外，兼并负担VRF分组也显示出认知障碍的差异，合并糖尿病的亚组认知表现更差，合并高脂血症的亚组认知表现较好。VRF的累加会增加认知功能的损伤程度。

研究2. VRF对全脑灰质结构的影响 本研究应用BABRI计划核磁样本690名非痴呆受试者的T1结构像数据（研究1的子集），利用基于体素的形态测量学研究方法，计算全脑灰质体积，分析不同VRF合并模式的作用下全脑灰质体积差异，寻找各VRF易感特定灰质损伤区域。结果发现糖尿病和高血压与全脑多个区域灰质体积萎缩有关。相比之下，高脂血症与较大的海马体积相关。肥胖与全脑灰质体积降低显著相关，吸烟也伴随着一定的灰质体积下降。累加的VRF会加剧全脑灰质体积萎缩的区域重叠，显示出全脑灰质体积模式的差异。除高脂血症外，合并存在的糖尿病与高血压对大脑灰质结构损伤最为严重。

研究3. VRF对脑白质微结构与拓扑网络的影响 本研究应用BABRI计划660名非痴呆老年人的弥散张量成像数据，经过计算得到全脑20根纤维束的平均各向异性分数 FA、平均弥散率MD等白质微结构指标，并应用纤维追踪技术构建脑白质拓扑网络，分析不同VRF合并模式的作用下白质纤维束完整性与拓扑网络属性差异。结果表明，两种VRF合并模式均不能很好的区分脑白质微结构指标的差异；相比之下，在兼并分组的脑白质拓扑网络分析中，我们发现了糖尿病合并高血压组在全局效率、局部效率显著降低，糖尿病合并高血压组全局效率较三高组有所下降。
    因此，本研究的结果表明，各血管危险因素对认知功能与脑结构指标的损伤有不同的风险权重，其中糖尿病是所有VRF中风险权重最大的，高血压次之，而高脂血症（或降脂药应用）对认知功能与大脑灰质结构表现为保护性作用；累加VRF负担越大，认知功能与脑结构表现越差；与单纯的VRF累加组合相比，兼并分组能够更好的区分不同负担模式的VRF对大脑健康指标的影响，表现为糖尿病合并高血压组的风险权重最大，甚至高于三高组。通过结合多领域认知功能测评与多模态神经影像技术，本研究从多个维度描述了血管危险因素对脑健康的影响，进一步理清了单一血管危险因素、多血管危险因素负担联合效应影响脑健康的认知功能及神经影像基础，为AD和其他认知障碍相关疾病的早期预警和分级临床治疗提供重要的依据。

Dementia is a disease characterized by progressive decline in memory and cognitive function. Its risk increases with age and seriously endangers the daily living ability of the elderly. Alzheimer's disease(AD) and vascular dementia (VaD) are the two most common types of dementia. Dementia not only brings enormous pressure and pain to patients, family and medical staff, but also adds huge economic burden to society. However, the exact etiology of AD is not yet clear. The commonly recognized amyloid cascade hypothesis has made Aβ a major therapeutic target, although it has invested many resources in the field of new drug research and development, however almost all AD clinical trials targeting Aβ have failed, and there is no effective treatment. At the same time, due to the uncertainty of VaD's classification and diagnostic criteria, VaD lacks recognizable and easily targeted therapeutic targets. So far, there is an urgent need to rethink the therapeutic objectives and intervention strategies for this highly dangerous disease.

At present, most complex diseases regulated by multiple factors are managed with the Tertiary Prevention Strategy, thereby effectively regulating the progress of the disease to improve the prognosis and quality of life of patients. This staged graded treatment model has achieved considerable success in the intervention treatment of stroke, for example, effective control of hypertension alone can reduce the incidence of stroke by 40%. Dementia as a complex chronic disease, taking into account its increasing number of patients and intractable, multidimensional intervention management for different stages of disease may be a promising intervention strategy. The Alzheimer's Association has also put forward multi-level interventions against dementia risk factors, emphasizing the possibility of preventing cognitive decline by adjusting lifestyle and controlling risk factors.

Therefore, in order to prevent the epidemic of dementia, it is vital that to explore how to specifically prevent and control the risk factors of dementia and effective interventions. Vascular Risk Factor (VRF), as an important risk for the onset of dementia, mainly includes diabetes, hypertension, hyperlipidemia, smoking and obesity. It is an important regulatory factor for early intervention and has potential to postpone or even reverse the onset of dementia. Although there is an established link between VRF and dementia, the mechanistic link between them has not yet been determined. Therefore, a comprehensive exploration and effective assessment of the impact of VRF on brain cognitive function and changes in brain structure has become a hot topic in recent years. Clarifying the relationship between VRF and brain health can be used both as an early warning of and as targets of intervention for these diseases and thus provide the basis for a huge and relatively economical strategy for the preservation of health in the elderly.

Although there has been extensive research in this field, at least two key issues have not yet been well resolved. First, the contributing factor of single VRF for cognitive dysfunction and brain structural damage has not been extensively studied. Because different VRFs have various weights for cognitive dysfunctions and brain structural indicators in various fields, the risk weights of individual VRFs should be clarified to focus interventions on VRFs with greater risks, so as to target cognitive function. The prevention and treatment of dysfunction and AD are more effective and economical. Second, in actual clinical situations, individuals may have two or more VRFs in combination. The effect of this synergistic VRF on brain health may have a specific effect. There has not been sufficient research on the combination of different burden patterns of VRF. 

Therefore, based on the two abovementioned questions, this study assumes that a single VRF or different VRF combinations have different risk degrees and will show specific cognitive impairment and brain structure patterns. Using the baseline data of the Beijing Ageing Brain Rejuvenation Initiative (BABRI) community-based cohort study on cognitive function and multimodal MRI data, the status of VRFs in various cognitive fields and the characteristic patterns of the gray matter, white matter structure, and structural network were multi-dimensionally described. This study is divided into three parts.

Study 1. The impact of VRF on advanced cognitive function.

This study analyzed the data of 4667 community individuals without dementia in the BABRI program and used the medical records or oral report records of community health stations to determine the test results. In individuals with diabetes, hypertension, dyslipidemia, and obesity, and those who smoke, a set of neuropsychological scales was used to assess whether the elderly individual’s cognitive ability is impaired, including the general mental state, memory function, visual spatial ability, language ability, attention, and executive function. Then, the cognitive score of each field was converted into a Z score to calculate the cognitive function under the combined mode of different risk factors. The results found that diabetes and hypertension are risk factors for cognitive function. Especially in overall cognitive function and memory, diabetes plays a greater role than does hypertension. In contrast, dyslipidemia (87.1% of patients regularly used statins) has a protective effect on the cognitive function. The effects of smoking and obesity on cognitive function are insignificant. The combined burden VRF subgroup also showed differences in cognitive impairment; the subgroup with diabetes had worse cognitive performance, and the subgroup with dyslipidemia had better cognitive performance. In addition to dyslipidemia, the accumulation of multiple VRFs can increase the severity of cognitive dysfunction.
Study 2. The effect of VRF on the structure of the whole brain gray matter.

In this study, the T1 structure image data (subset of Study 1) of 690 subjects without dementia from the MRI sample of the BABRI program were used. Using voxel-based morphometry, the whole brain gray matter volume was calculated. The difference in the whole brain gray matter volume under the effect of different VRFs combined modes was analyzed, and the risk factors susceptible to specific gray matter damage areas were determined. It was found that diabetes and hypertension were related to the atrophy of gray matter in multiple regions of the brain. In contrast, dyslipidemia was found to be associated with larger hippocampal volume. Obesity was significantly associated with a decrease in the gray matter volume of the whole brain, and smoking also accompanied a certain gray matter volume injury. The cumulative VRF subgroup will aggravate the overlapping of areas of atrophy of the whole brain gray matter, showing the difference in the whole brain gray volume pattern. In addition to dyslipidemia, the combination of diabetes and hypertension causes the most severe damage to the gray matter structure of the brain.
Study 3. The effect of VRF on the white matter microstructure and topological network of the brain.

In this study, the diffusion tensor imaging data of 660 individuals without dementia in the BABRI program were used to analyze the average score of the white matter microstructure index, such as Fractional Anisotropy (FA) and Mean Diffusivity (MD) of 20 fiber bundles in the whole brain, and the fiber tracking technology was used to construct the white matter topological network, and the difference between the white matter fiber bundle integrity and topological network attribute under different VRF combined modes was analyzed. The results showed that the two VRF burden models cannot distinguish the differences in white matter microstructure indicators well. In contrast, in the analysis of the combined grouped white matter topology network, we found that the global and local efficiencies were significantly reduced in the diabetes combined with hypertension group, and the global efficiency of the three high groups with dyslipidemia was better than that of the diabetes combined hypertension group.
    Therefore, the results of this study show that each individual VRF has different risk weights for the impairment of cognitive function and brain structure indicators. Diabetes had the highest risk weight among all VRFs, followed by hypertension, and dyslipidemia (or use of statins) may have a protective effect on cognitive function and brain gray matter structure; the greater the combined VRF burden, the worse the cognitive function and brain structure performance. Compared with the simple VRF cumulative group, the combined group can better distinguish the impact of combined VRF on brain health indicators, and the risk of diabetes combined with hypertension is the highest, even higher than the three combination groups. This study describes the impact of VRF on brain health from multiple dimensions by combining cognitive function assessment and multimodal neuroimaging technology. It is important for the comprehensive assessment of the impact of VRF on cognitive function and brain structure, providing a crucial basis for the prevention and clinical treatment of AD and other cognitive impairment-related diseases.