全球正在加速老化，如何在老化过程中，克服伴随年龄增长而发病率逐渐增高的阿尔兹海默病（Alzheimer’s Disease，AD）等重大脑疾病的发生发展，延缓认知障碍恶化速度，是减小老龄化社会带来的威胁，促进老年人大脑健康，实现成功老化的关键。 在老化过程中伴随的AD对老年人身心健康危害重大，但是目前尚未研发出有效的治疗方法，现在共识是将治疗阶段前移至临床前期。现有的生物标记物，如PET、脑脊液等检测成本高，对病人造成有创损伤，磁共振脑影像学技术的兴起为我们开发一种便捷无创的生物标记物提供新的可能，该技术使我们可以在活体中对大脑的解剖结构、功能活动进行测量，探索疾病背后的大脑改变模式，揭示神经退行性疾病背后的未解之谜。越来越多的研究者认为这些早期的影像标记物对于AD的早期诊断和干预具有重要的临床意义。通过多模态神经影像学技术，早期鉴别和诊断AD早期甚至前期的轻度认知障碍（mild cognitive impairment，MCI）群体，将他们从正常老化群体中筛查出来，在疾病发展早期阶段及时发现AD疾病对大脑的影响，区分AD病理老化与正常老化轨迹的差异，找到最敏感的脑区，一方面探讨不同模态神经影像指标中发现的敏感脑区之间的内在联系，解释疾病发生发展的可能路径；另一方面对这些脑区展开一系列有针对性的治疗和干预，对于预防或者延缓AD发生发展均起到关键作用。 过去的认知老化研究已经发现各项认知功能随着年龄变化的轨迹，并提出了多种理论假设加以解释。然而以往针对于老化过程中伴随的老年性痴呆等脑病研究，多是研究AD或MCI患者与正常老年人在认知表现上的组间差异，并未将疾病群体放在动态、变化的过程中去观察其各项指标随年龄的变化。因此，我们提出以下研究问题：病理老化是否加速了原有的认知老化进程？病理老化对认知老化的损伤模式是否具有一定的脑神经基础？病理因素影响认知老化的神经机制是什么，在脑结构、脑功能以及脑网络上的影响是怎样的？这些跨模态脑影像学数据在考察病理老化与正常老化之间存在什么共同规律以及差异之处？这些都是认知老化与脑老化研究领域中亟待解决的问题。 既往认知老化研究中多是仅使用了单一模态的神经影像学数据，或仅仅考察MCI群体、老化群体以及健康年轻群体的组间差异，在系统评价认知老化的机制研究中存在方法学上的局限，也不能很好地考察正常老化与AD病理老化在年龄维度上的变化差异，造成了疾病研究与老化过程割裂、认知老化与脑老化过程割裂以及脑老化的不同角度割裂。另外，目前针对汉族人群的大样本相关研究相对较少，本研究所选取的被试均来源于北京社区，最终纳入269名具有完整认知能力相关指标及T1结构像、DTI及静息态功能磁共振等脑影像学数据的老年人被试进行分析，其中，MCI患者97名，172名认知正常老化对照组。采用了全套、标准化的认知测查量表，包括记忆、执行功能、注意力、言语功能等多方面认知能力；从脑灰质结构、白质结构、白质结构网络、以及静息态功能网络等多个维度考察了轻度认知障碍与认知正常老人在认知老化过程中的变化异同。研究使用协方差、偏回归以及一般线性模型分析等统计方法，将认知状态和年龄作为自变量，计算两个变量在上述各个模态脑影像学数据中的交互作用，揭示两组人群在年龄增长过程中的神经影像学变化机制。本研究总共分为4项具体研究。 1、正常老化与病理老化对脑灰质结构的影响。研究发现，首先，MCI表现出于正常老化相似的灰质皮层萎缩模式，老化模式相似的脑区主要分布在大脑前部的额叶脑区之中。其次，MCI在表现出于正常老化相似的灰质皮层萎缩模式基础之上，比正常老化更严重地影响了大脑后内部脑区结构，比如颞叶、扣带回等脑区，这些脑区在MCI组呈现出了与正常老化组相比加速萎缩的情况，预示着这些脑区的灰质结构萎缩可能与MCI老化有关。最后，这些有交互效应的脑区中，左侧颞平面脑区以及左侧后扣带回脑区两个脑区的灰质体积萎缩与总体认知能力、记忆力、执行功能等领域的认知测验成绩下降显著相关，在MCI组中这种相关性更强。 2、正常老化和MCI老化对脑白质结构的影响。研究发现，正常老化群体年龄相关白质完整性下降主要表现在前额叶相关的纤维束及部分枕叶脑区，包括双侧丘脑前辐射，右侧额枕下束，以及小钳等纤维束。而在MCI群体中， 则主要表现在内侧颞叶以及后部脑区附近，主要包括双侧扣带回（海马段），双侧钩束，小钳以及左侧丘脑前辐射等脑区。仅仅在双侧扣带回（海马段）发现了分组×年龄交互作用，即正常老化组该纤维束随年龄增长并未发生显著的老化损伤，而MCI患者组则发生了显著的年龄相关损伤。 3、正常老化和MCI老化对脑结构网络的影响。对于白质网络节点属性，正常老化群体年龄相关白质网络节点全局效率（global efficiency，gE）下降的脑区为顶叶以及边缘系统节点，主要包括左侧旁中央小叶、左侧楔前叶、左侧中扣带回以及右侧中扣带回和右侧舌回等脑区，而在MCI群体中，则年龄相关显著下降的网络节点则主要分布在额颞叶脑区，主要包括左侧额下回眶部、左侧额中回眶部、右侧前扣带回、右侧旁海马回、以及右侧旁中央小叶等脑区。而分组×年龄交互作用显著的白质脑网络节点主要包括左侧距状裂、左侧中央前回、左侧额上回背外侧沟、左侧前扣带回、以及右侧楔回等脑区。其中，只有左侧前扣带该节点的网络全局效率下降与几个负责记忆、执行功能测验的成绩呈显著相关。左侧前扣带的网络全局效率成为区分正常老化与MCI老化最显著的白质结构网络指标。 4、正常老化和MCI老化对脑功能网络的影响。本研究通过中国大样本静息态功能磁共振数据对正常群体和MCI群体在老化过程中静息态脑功能网络随年龄变化趋势的差异进行了详细考察，结果发现默认网络中的后背侧默认网络在正常老化组和MCI群体中存在显著的分组×年龄交互效应。围绕后背侧默认网络，还发现了它和默认网络其他成分以及突显网络之间的功能连接的组间变化模式的差异。 综合以上结果，本研究发现MCI老化能够影响老年人的认知功能随年龄变化的轨迹，同时这种认知老化的改变有其内在的神经病理学机制。具体体现在：1）MCI老化会加重额叶和皮质下脑区的萎缩，并特异性损伤了颞叶和后部边缘系统的老化过程，脑萎缩模式与部分认知能力的表现相关；2）MCI老化损伤了正常老化中免于损伤的扣带（海马段）等脑白质纤维束的完整性，同时引起了相关脑区在结构网络中的效率降低；3）MCI老化过程中加重损伤的眶额皮层，旁海马，后扣带，脑岛，前扣带等区域及与其相连的白质纤维束等脑默认网络和突显网络中的关键节点，并且MCI老化对默认网络产生了额外的损伤，并加重了突显网络的老化过程；4）MCI老化过程加重了后背侧默认网络功能连接的损伤，同时引起了与该网络相关的默认网络其他子网络以及突显网络之间功能连接的异常。综合以上四个方面的内在神经机制，MCI老化的脑神经机制大致可以体现为局部结构异常——结构网络节点异常——功能网络异常的整个脑结构和功能机制的加重和异常老化系统现象。 总之，本研究首次在中国汉族的大样本人群中，通过多模态脑影像学指标结合认知行为学测验系统阐述了MCI老化与正常老化在认知老化过程中的变化异同。我们的研究全面系统地诠释了MCI老化区别于正常老化的脑结构、脑功能、脑网络以及行为表现上的特征，对认知老化基础研究以及阿尔兹海默病的临床前期防治都有重要的理论价值和现实意义。

The world is encountering accelerated aging, how to overcome the occurrence of major brain diseases such as dementia, cognitive impairment is the key to reduce the threat posed by the ageing problem, and to promote the elderly brain health, and to realize successful aging. With the multimodal neuroimaging technology development, we can measure anatomic structure and function of brain activity in vivo, to explore the brain behind the superficial cognitive change of neurodegenerative diseases, revealing the mystery behind them.These early imaging markers are of great clinical significance for early diagnosis and intervention of Alzheimer's Disease (AD). Through multimodal neuroimaging techniques for early detecting and diagnosis of AD and even to the early stage of AD (mild cognitive impairment, MCI) group, will help us to screen out the patients from normal aging population. To distinguish between AD pathological aging and normal aging trajectory difference, and find out the most sensitive brain area, on the one hand could help explore different modal neuroimaging index found in sensitive potential relationship between brain areas and explain the possible paths of the development of the disease; on the other hand, could play a key role on a series of targeted therapies and interventions on these brain regions in preventing or delaying the development of AD. In the past, cognitive aging studies have found that cognitive function changes with age, and several theoretical hypotheses have been proposed to explain it. However aging with dementia, in the process of brain research, mostly focus on the study of AD and MCI patients and normal old people on the cognitive performance differences between groups, did not put concentration on the disease group dynamically to find out the process of how the indicators changes with age. Therefore, we put forward the following research questions: Does the pathological aging speed up the process of the original cognitive aging? Does pathological aging have a certain neural basis for cognitive aging damage pattern? What are the neurological mechanisms that affect cognitive aging, and what are the effects on brain structure, brain function, and brain networks? What are the common and differences patterns between these multimodal brain imaging data in the study of pathological aging and normal aging? These are urgent problems in the field of cognitive aging and brain aging. Previous studies on cognitive aging, usually only basedon either cognitive behavior or a single modality of neuroimaging data, therefore led to limitations in the mechanism of systematic evaluation of cognitive aging. It also results in the fragmentation of the disease research and aging process, the fragmentation of cognitive aging and brain aging, and the different angles of brain aging. In addition, current large sample studies of the Han population is relatively small, while the participants in our study were from communities in Beijing, adopted a full battery of neuropschological tests, including memory, attention, executive function and speech function aspects of cognitive ability. From the level of cognitive behavior, the structure of the gray matter, , white matter, structural network , and function of resting state network investigation form a multiple dimensions of mild cognitive impairment and cognitively normal elderly in cognitive similarities and differences of the changes in the aging process. In the result analysis and comprehensive, we applied two independent variables, cognitive status and age at various to reveal their interaction on cognitive performance in the process of two groups in age and neuroimaging changes mechanism.This study is divided into five specific studies. 1. Effects of normal aging and mild cognitive impairment on cognition. This part explains the interaction of two aging processes from the perspective of cognitive behavior. The study of this paper is all from Beijing Aging Brain Rejuvenation Initiative database. Using inclusion and exclusion criteria, finally 269 paticipants with complete cognitive and brain imaging data were included in our analysis. Among them, 97 patients belonged to MCI group, 172 cognitively normal aging belonged to healthy control group. In this study, by using covariance, partial regression and statistical methods such as regression analysis, we found that the cognitively normal elderly and MCI patients showed the trend of a significant decline in cognitive performance along with the age, exists the interaction between grouped by age in the executive function. 2. Effects of normal aging and mild cognitive impairment on gray matter structure. Firstly, MCI was found to be in the same gray cortical atrophy pattern as normal aging, with a similar pattern of brain regions in the frontal lobes of the brain. Secondly, MCI similar gray matter atrophy due to normal aging model basis, is more serious than normal aging impacts on the internal structure of the brain, such as temporal lobe, cingulate gyrus and other brain regions. These brain regions in the MCI group presented accelerated atrophy, compared with normal aging group, which indicates the atrophy of gray matter in these regions structure may be associated with MCI pathological aging. Finally, interaction effect were found at such brain areas, the left posterior temporal lobe and left cingulate gyrus area. The volume of gray matter in two regions are correlated with general cognitive ability, memory, executive function test performance decrease significantly , and the correlation is stronger in MCI group than NC group. 3. Effects of normal aging and mild cognitive impairment on white matter structure. The study found that normal aging population age related decline in the fiber integraty at frontal and the occipital part , including bilateral thalamus anterior radiation, right fronto-occipital fasciculus, and forceps minor. While in MCI group, white matter at the medial temporal lobe and posterior part of the brain, including bilateral cingulate (hippocampus part ), bilateral fasciculus uncinatus, forceps minor and the left thalamus anterior radiation. Only on bilateral cingulate (hippocampus part) the groupe by age interaction were discovered, showing the normal aging of the white matter did not occur significant aging damage, and MCI patients suffered significant age-related damage. 4. Effects of normal aging and mild cognitve impairment on brain structure network. For white matter network node properties, normal aging population age-related white matter network nodal global efficiency (global efficiency, gE) decline in areas of the brain as the parietal lobe, and the limbic system node, mainly including the left paracentral louble, left the precuneus, the left middle cingulate cortex and right in the right lingual gyrus. In MCI group, a significant reduction of age-related network nodes decline were found mainly in the frontal temporal lobe brain regions, including the orbital part of the left inferior frontal lobe, posterior orbital part of left inferior frontal love, right anterior cingulated cortex, right parahippocampus gyrus, and right paracentral louble. Significant group by age interaction on network nodal efficiency were found mainly on the left calcarine fissure, left paracentral louble, left anterior cingulated gyrus, and right cuneus. Among them, the decrease of the network global efficiency of the left anterior cingulated gyrus was significantly correlated with the scores of the memory and executive function score. The network global efficiency of the left anterior cingulate is the most significant index of white matter structure which distinguish normal aging from MCI pathological aging. 5. Effects of normal aging and mild cognitive impairment on brain function network. This study using large sample resting state fMRI data of normal group and MCI group resting state in the process of aging brain functional network change tendency along with the age difference, found that the posterior default network default network in normal aging and MCI groups exist significant group by age interaction effect. We also found the functional connectivity between posterior default network and other components of the default network and between it and other functional network such as salience network was changed. In sum of the above results, this study found that the mild cognitive impairment can affect cognitive function the trajectory of change with age in the elderly, at the same time, the change of cognitive aging has its specific mechanism of neuropathology. We found: 1) pathological aging can aggravate the atrophy of frontal cortex and subcortical cerebral area, particularly on the temporal lobe and the posterior part of the limbic system in the aging process, brain atrophy patterns is correlated with some cognitive performance. 2) pathological aging damaged the integrity of white matter fasciculus, such as the cingulum (hippocampus part), which was not damaged in normal aging, and caused the related brain regions to be less efficient in the structural network. 3) in the process of pathological aging increase orbitofrontal cortex injury, hippocampus, cingulate, posterior insula, anterior cingulate region and its adjacent white matter fiber tracts, such as the default network. It highlight the key role of the network, and the pathological aging has extra damage on the default network, and aggravated aging process of the network. 4) pathological aging process increased posterior default network connection damage, caused the default network associated with the other subnet of DMN and also cause abnormal functional connectivity between DMN and othernetworks. To sum up the above four aspects inherent neural mechanisms, the pathological mechanism of aging of the brain can be roughly embodied in the local structural abnormalities, abnormal structure network nodes, the entire brain function network abnormal structure and function mechanism . In conclusion, this study for the first time, in large sample population of Han population in China through multimodal brain imaging and combined with cognitive behavior test system elaborated the pathological changes in the process of normal and pathological aging. Our research from three aspects: structure, function, behavior, and systematically explained the pathological aging differ from normal aging brain structure and brain function, the network and the characteristics of the behavior, not only for basic research of cognitive aging, but also for clinical prevention and treatment of senile dementia could serve important theory value and practical significance.