【作者】 聂坤；

【导师】 王丽娟；

【作者基本信息】 南方医科大学 ， 神经病学， 2013， 博士

【摘要】 帕金森病认知功能障碍是帕金森病一种较为常见的非运动症状,包括帕金森病轻度认知功能障碍(Mild cognitive impairment in Parkinson’s disease, PD-MCI)和帕金森病痴呆(Parkinson’s disease dementia, PDD),其中PD-MCI是PDD发生的独立危险因素,2012年1月国际运动障碍协会(The Movement Disorder Society, MDS)也正式公布了PD-MCI的诊断标准。研究表明PD-MCI的患病率为18.9%-38.2%,在非痴呆PD中甚至高达52.8%。PD-MCI每年进展为痴呆的几率为6%-15%,62%的PD-MCI患者在4年内进展为PDD。帕金森病认知障碍患者加之本身严重的运动障碍,给患者本人及照料者造成了沉重的生活和经济负担。因此及早发现,及时干预,延缓病程进展是当前治疗PD认知功能障碍最有效的手段。PD轻度认知功能障碍患者存在明显的异质性,发病可以在PD病程晚期也可以在PD诊断之初；累及认知领域广泛,包括执行功能、记忆力、注意力、视空间能力、语言流畅性等；内在病理生理改变复杂,可能合并有皮层路易小体(Lewy body)形成、阿尔茨海默病(Alzheimer’s disease,AD)样病理改变和脑微血管病变等病理改变,同时伴有多种神经递质参与和广泛的神经元变性。PD-MCI作为PD认知功能正常(Cognitively normal Parkinson’s disease, PDCN)和PDD之间的一种中间状态,对其进行研究将有助于我们更加全面深入地了解PD认知功能障碍的特点和其发展演变过程。尽管制定了PD-MCI的诊断标准,但是MDS专家委员会同时认为,目前关于PD-MCI的系统性研究较少。现有研究由于采用的诊断标准和评估量表有所差异,其患病率在17.4%-38.2%不等,PD-MCI临床表型的研究同样存在着争议,一些研究认为PD-MCI患者以单项认知功能受损为主,且单项额叶/执行功能受损占多数,而另外一些研究却提出异议,一项多中心研究1346名PD患者中PD-MCI表型以记忆力损害最常见。目前仅有两项研究显示PD-MCI的发生可能与高龄、起病晚、男性、抑郁以及运动症状的严重程度相关,血管危险因素与PD-MCI的相关性目前还未见报道。而且以上这些方面的研究均为国外报道,目前还没有针对中国PD人群的研究。目前PD认知功能障碍包括PD-MCI和PDD的诊断乃至PD的诊断,仍以临床诊断为主,缺乏有效简便易行的分子生物学和神经影像学标记物。而且PD认知障碍病理发病机制的研究也多局限于少量尸检结果和动物细胞试验。随着影像设备的更新和图像处理技术的发展,脑结构和功能影像学为研究者提供了体外观察脑组织形状、体积、密度、结构完整性、代谢变化和分析认知功能活动的重要手段。然而以往采用的定量或半定量的分析手段,存在主观性强,研究部位局限,无法提取细微结构变化以及无法进行横向比较等局限性。在脑结构的研究中,新的分析方法基于体素对整个大脑灰质、白质和白质纤维完整性进行自动化定量分析,并在统计参数图软件(SPM)上完成分析比较。目前这种方法已成功应用于多种神经变性疾病、头痛、精神疾病、睡眠障碍等；其中关于PDD的研究已经较多,研究发现PDD患者存在广泛的脑萎缩和白质纤维的改变,而对其前期阶段PD-MCI的研究甚少,尤其是对整个大脑脑白质纤维的研究还未见报道。正如像AD等伴有痴呆的神经变性疾病一样,大脑结构的萎缩和白质纤维的病变可以成为其神经病理改变的标记物甚至在疾病的早期。因此本研究假设：大脑结构的萎缩和/或白质完整性的损害可能是导致PD-MCI发生的重要因素。本课题组在国内率先制定了一系列针对PD认知功能的评估体系,并应用该体系进行了多项影像学和分子生物学研究,科研成果多次在国际杂志和会议上发表交流,得到国际国内同行的认可。本课题在以往评估体系的基础上进一步完善,采用整体认知功能评估与四个认知域评估相结合,参考日常生活能力受损情况,排除痴呆及严重抑郁,依据2012年PD-MCI诊断标准,针对中国PD人群,大样本量研究帕金森病轻度认知障碍患者的临床特征、认知功能特点,进行亚组分型,分析其认知障碍表型特点,并对PD轻度认知障碍相关危险因素包括血管危险因素等进行分析,为全面认识PD-MCI并对其早期干预治疗提供支持和依据。在此基础上联合应用基于体素的形态学测量(voxel based morphometry, VBM)和整个大脑基于体素分析(voxel-based analysis, VBA)磁共振弥散张量成像(diffusion tensor imaging, DTI)技术,对PD-MCI患者脑灰质、白质密度和白质纤维的完整性进行比较分析,探讨各认知功能领域的损害与脑结构变化的相关性,寻找特定认知功能损害可能的病理改变和内在机制,以期发现PD-MCI早期识别和诊断的特征性影像学标记物。第一章帕金森病轻度认知障碍的临床特征、认知特点及相关因素分析目的：探讨帕金森病轻度认知障碍患者临床特征、认知功能障碍特点及亚组分型,对PD轻度认知障碍相关危险因素进行分析。方法：共收集234例非痴呆PD患者,收集相关临床资料,行运动功能、认知功能、焦虑、抑郁、日常生活能力的评估；采用整体认知评估和四个认知域评估相结合,排除痴呆及严重抑郁,依据PD-MCI诊断标准将非痴呆PD患者分为PDCN、PD-MCI两组,再依据认知功能受损情况将PD-MCI患者分为单项非记忆力受损型、多项非记忆力受损型、单项记忆力受损型和多项伴记忆力受损型等四个亚型,分析各亚型的临床和认知特点,运用二分类的logistic回归分析PD-MCI相关危险因素,包括性别、年龄、发病年龄、病程、教育年限、UPDRS-Ⅲ、H-Y分级和血管危险因素高血压、糖尿病、吸烟、饮酒、高同型半胱氨酸血症等因素。结果：(1)234例非痴呆PD患者中PD-MCI患者106例,占45.3%。与PDCN组比较,两组在年龄、起病年龄、教育年限及运动症状的严重程度上有显著性差异(P<0.05)。(2)PD-MCI亚组分型显示PD-MCI患者以单项认知功能受损为主,其中单项执行功能受损占22.64%,单项记忆力受损占20.75%,单项视空间受损占6.6%,单项注意力受损占2.83%；四个亚组所占比例分别为：单项非记忆力受损型占32.08%、多项非记忆力受损型占10.38%、单项记忆力受损型占20.75%和多项伴记忆力受损型36.79%。(3)PD-MCI患者四个亚型之间在运动症状的严重程度上未见显著差异,而受教育程度的高低可能与记忆力损害相关,四组中多项伴记忆力受损型PD-MCI患者整体认知功能下降更严重。(4)回归分析发现高龄、UPDRS-III评分高和高同型半胱氨酸血症是PD-MCI发生的危险因素,而受教育程度高为保护因素,教育年限越长,发生认知障碍的风险越小。结论：首次针对中国PD-MCI患者进行分析研究,发现非痴呆PD人群中轻度认知功能障碍广泛存在,其以单项认知功能受损为主,各分组亚型之间运动症状的严重程度上没有显著差异,受教育程度高是PD-MCI的保护因素,而高龄、严重的运动症状以及血管危险因素中同型半胱氨酸血症可能促进了PD-MCI发生。第二章应用VBM技术分析帕金森病轻度认知障碍患者脑结构的研究目的：应用VBM技术分析帕金森病非痴呆患者脑灰质体积和白质密度的改变,探索帕金森病轻度认知障碍患者大脑结构的变化特点,揭示其内在的病理变化。方法：根据帕金森病轻度认知障碍的诊断标准,入组帕金森病轻度认知障碍患者(PD-MCI)和帕金森病认知正常患者(PDCN)组患者各23例及年龄和性别匹配的健康老年人21例,所有被试者行3.0T的磁共振扫描获取脑结构磁共振图像,在Matlab R2008a平台上采用SPM8软件包中的VBM8工具进行图像预处理,包括标准化、分割、调制、平滑等,所得图像数据利用SPM自带统计软件包进行统计分析,比较每两组之间的灰质体积/白质密度的差异,分析与不同整体认知功能评估相关的脑结构变化情况。结果：(1)PDCN组较对照组灰质体积萎缩的区域主要集中于前额叶、边缘叶及左侧颞中回等,且白质密度减少的区域主要是基底节和额叶等。(2)PDMCI组较对照组灰质体积萎缩的区域有额叶、边缘叶、基底节区及小脑等部位,且白质密度减少的区域有额叶及尾状核等。(3)PDMCI组较PDCN组灰质体积萎缩的区域有左侧颞中回/颞下回及额叶等,且白质密度萎缩的区域有右侧额下回,中脑和枕叶舌回等。(4)与记忆力评估为主MMSE评分正相关的脑灰质区域有右侧扣带回及边缘叶,与之正相关的脑白质区域有额叶、海马旁回及边缘叶等。(5)与非记忆力评估为主MoCA评分正相关的脑灰质区域有额叶、基底节区、海马旁回、枕叶及小脑等。与之正相关的脑白质区域有额叶、顶叶、楔前叶、枕叶、楔叶、岛叶、边缘系统及小脑等。结论：PD-MCI的发生与局部脑区如额叶、边缘叶、内侧颞叶及基底节的萎缩相关；非痴呆PD脑结构的变化与整体认知功能相关,其脑萎缩的部位集中于额叶和边缘系统等,且以皮质下萎缩为主。非痴呆PD患者脑边缘叶的萎缩与其记忆力受损相关,而额叶的萎缩与执行功能障碍相关；PD认知功能障碍累及了广泛的大脑皮层及皮层下结构的萎缩。VBM技术可探查PD认知障碍早期-PD-MCI时期大脑结构的细微改变,局部脑结构的萎缩有可能成为PD-MCI早期诊断的神经影像学标记物,但仍需更多的大样本研究加以证实。第三章帕金森病轻度认知障碍患者基于体素分析的磁共振弥散张量研究目的：应用基于体素分析的DTI图像处理技术分析帕金森病轻度认知障碍患者认知功能与脑白质纤维变化的关系,以期发现与帕金森患者认知功能相关的特异的脑白质改变。方法：根据帕金森病轻度认知障碍的诊断标准,入组帕金森病轻度认知障碍患者(PD-MCI)15例,帕金森病认知正常患者(PDCN)组患者17例及年龄和性别匹配的健康老年人15例,所有被试者行3.0T的磁共振扫描获取磁共振弥散张量(DTI)图像,采用DTI studio软件计算出各向异性分数(Fractional Anisotropy, FA)图。然后将各组受试者图像进行基于体素的全脑分析比较(Voxel Based Analysis, VBA),分析比较每两组之间FA值有显著差异的脑区,分析与不同整体认知功能评估相关的脑白质纤维完整性的变化情况。结果：(1)PDCN组较对照组FA值下降的区域主要包括右侧额叶,中央前回等；(2)PDMCI组较对照组FA值下降的区域集中于左侧颞中回和顶下小叶等,其次还包括丘脑,豆状核,岛叶,扣带后回,楔前叶,枕叶,梭状回等部位；(3)PDMCI组较PDCN组FA值下降的区域集中于右侧顶上小叶,楔前叶和左侧枕中回,楔叶等,其次还包括丘脑,后扣带回、顶下回,边缘叶,海马旁回,颞上回等；(4)PD患者中MMSE、MoCA认知功能评分与脑白质FA值成正相关；其中与记忆力评估为主MMSE评分相关的脑区集中于右侧丘脑、小脑前叶,楔回,边缘叶,海马旁回和左侧中脑,丘脑和壳核等,其次还有颞上回、舌回,扣带回,距状裂,额叶中央前回等；与非记忆力评估为主MoCA评分相关的脑区包括边缘叶,海马旁回、颞中回、顶下回,顶叶,楔前叶,枕中回舌回、丘脑、小脑前叶以及基底节区等。结论：PD-MCI的发生与特征性局部脑白质纤维完整性异常相关,部位主要包括顶上小叶、楔前叶、枕中回,楔叶及颞中回等；与整体认知功能评估正相关的FA值下降区包括脑室周围、扣带回、楔前叶、楔叶、舌回等,可能与胆碱能纤维的受损相关；PD患者脑白质纤维完整性的改变可能是加速PD认知功能障碍的重要原因；基于体素的全脑分析DTI技术可以全面准确地探查的脑白质改变,有助于PD轻度认知障碍的早期诊断。更多还原

【Abstract】 Cognitive dysfunction is a pretty common non-motor symptom of Parkinson’s disease, including mild cognitive impairment in Parkinson’s disease (PD-MCI) and Parkinson’s disease dementia (PDD), PD-MCI is an independent risk factor for PDD. The Movement Disorder Society (MDS) formalized diagnostic criteria for PD-MCI in January2012. Studies have shown that the prevalence of PD-MCI was18.9%-38.2%, even as high as52.8%in the non-demented PD. Annual progression rate from PD-MCI to dementia was6%-15%,62%of PD-MCI patients progress into PDD in four years. Patients with cognitive impairment in PD combined with severe motor symptoms resulted in a heavy social and economic burden. Therefore, early detection and timely intervention to delay the progression may be the most effective treatment for cognitive deficits in PD.There has been significant clinical heterogeneity in PD-MCI. It can initiate either at advanced or at the beginning of the PD diagnosis, involving a wide range of cognitive domains, including executive function, memory, attention, visuospatial ability, language fluency. Complex intrinsic pathophysiological changes may be associated with the formation of Lewy body, Alzheimer’s disease (AD)-like pathological changes and cerebral microvascular disease, accompanied by a variety of neurotransmitters participation and extensive neuronal degeneration. PD-MCI is an intermediate state between cognitively normal Parkinson’s disease (PDCN) and PDD, the study of which will help us to look insight into the characteristics and evolution of PD cognitive dysfunction.Although MDS has developed the PD-MCI diagnostic criteria, there are few systematic researches about it. Because of the diversity of assessment scale and diagnostic criteria adopted, the prevalence ranges from17.4%to38.2%and the clinical phenotype of the disease is also controversial. Some researchers believe that the single frontal/executive function impairment is the most common in PD-MCI. To the contrary, a multicenter study enrolling1346patients indicated that the most common clinical phenotype of PD-MCI is memory deficit in. Currently, there are only two studies showed that PD-MCI might be associated with older age, male, late onset, depression and severe motor symptoms, the correlation between vascular risk factors and PD-MCI has not been reported yet. And studies on these aspects are foreign publications, there is no research on Chinese PD population at present.At present, the diagnosis of PD-MCI, PDD and even PD is still in clinical diagnosis, lacking effective molecular biological and neuroimaging markers. And the study on pathogenesis of cognitive impairrment of PD is limited to autopsy of small samples, animal and cell experiments. With the development of imaging equipment and image processing techniques, the structural and functional brain imaging provides an important way in vitro to observe the brain shape, volume, density, structural integrity, metabolic changes and to analyze the cognitive activities. Previous semi-quantitative or quantitative analysis method is not only subjective and limited to small study areas, but also can not extract subtle structural changes and can not be compared with other similar studies. In the study of brain structure, a new analysis method based on voxelwas adopted to conduct automated quantitative analysis for gray matter, white matter and white matter fiber integrity of the whole brain in the statistical parametric mapping software (SPM). This method has been successfully applied in the the studies of neurodegenerative disease, headache, mental disease, and sleep disorder. Among the studies, researches on PDD are common, which found that PDD patients have widespread brain atrophy and white matter fiber changes.However, the study on the early-PD-MCI, especially the study on the entire brain white matter fiber has not been reported. Brain atrophy and white matter fiber lesions can be used as neuropathological markers of PD cognitive dysfunction, even in the early stage, as in other neurodegenerative diseases with dementia, such as AD. So here is my hypothesis:brain atrophy and/or white matter integrity lesions may bean important factor leading to PD-MCI.Our research group has developed a series of cognitive function assessment system for PD patients in China, which were applied to a number of neuroimaging and molecular biological studies,.many research results were published in international journals and conference communications and recoguzed by the domestic and international peers. In this study all non-demented PD patients were classified into cognitively normal Parkinson’s disease (PDCN) or PD-MCI considering global cognitive and four cognitive domains (executive function/planning skills, memory, attention and visual-spatial skill), according to the diagnostic criteria for PD-MCI (MDS), excluding dementia and severe depression. The objective of this study was to determine the prevalence, subtype, clinical characteristics and risk factors of mild cognitive impairment in Parkinson’s disease (PD-MCI).,which will contribute to a comprehensive understanding of the PD-MCI and provide theoretical basis for earlyintervention. Besides,we analyzed the changes of structural brain and white matter fiber in PD-MCI patients by combining VBM and DTI magnetic resonance imaging technology We aimedto explore the relevance between cognitive dysfunction and brain structure changes, to find potential pathological changes and the internal mechanism and tofind the characteristic imaging markers for the early identification and diagnosis of PD-MCI.Part I:Mild Impairment in Parkinson’s disease:Motor and Cognitive Characteristics, Risk FactorsObjective:The objective of this study was to determine the prevalence, subtype, clinical characteristics and risk factors of mild cognitive impairment in Parkinson’s disease (PD-MCI)Methods:We performed a retrospective study of234non-demented PD patients, who completed a standardized neurological assessment, including demographic and clinical information and a full neuropsychological battery. All non-demented PD patients were classified into cognitively normal Parkinson’s disease (PDCN) or PD-MCI considering global cognitive and four cognitive domains (executive function/planning skills, memory, attention and visual-spatial skill) according to the diagnostic criteria for PD-MCI (MDS), excluding dementia and severe depression. We further divided participants with MCI into four subtypes:amnestic single domain, amnestic multiple domain, nonamnestic single domain or nonamnestic multiple domain and analyzed the characteristics of the four subtypes. The risk factors for PD-MCI were examined using binary logistic regression.Results:(1) A total of45.3%of subjects (106) were classified as MCI. The MCI group had older age at assessment and at disease onset, shorter years of education and more severe motor symptoms than the cognitively normal group(P<0.05).(2) Single-domain impairment is the most common subtype of PD-MCI, Of which executive ability impairment was the most common (22.64%), followed by memory (20.75%), visual-spatial (6.6%) and attention impairment (2.83%). Regarding cognitive profiles,32.08%were classified as nonamnestic single domain MCI,20.75%as amnestic single-domain,36.79%as amnestic multiple domain, and10.38%as nonamnestic multiple-domain MCI.(3) Differences were not significant in the severity of motor symptoms among the four subtypes of PD-MCI. The education level may be associated with memory deficits, the decline of global cognition in amnestic multiple domain-MCI patients was more severe.(4)Regression analysis showed that advanced age, high score of UPDRS-Ⅲ and hyperhomocysteinemia were risk factors for PD-MCI and the high level of education was a protective factor, with lower risk of cognitive impairment in patients with a high level of education.Conclusions:MCI is common in patients with PD without dementia, affecting a range of cognitive domains, including memory, visual-spatial, and attention/executive abilities, of which single-domain impairment is the most common subtype. Differences were not significant in the severity of motor symptoms among the four subtypes of PD-MCI. MCI was associated with older age at assessment,lower levels of education, more severe motor symptoms, and vascular risk factors of hyperhomocysteinemia. Part II:A magnetic resonance imaging study of patients with Parkinson’s disease with mild cognitive impairment using voxel-based morphometryObjective:Voxel-based morphometry was used to examine the changes of gray and white matter in the brains of patients with mild cognitive impairment in Parkinson’s disease (PD-MCI) and to identify the specific regions responsible for cognitive dysfunction in PD.Methods:Patients were classified into PDCN (23) and PDMCI (21) groups according to the diagnostic criteria for PD-MCI (MDS), and21age-matched healthy control subjects recruited from an outpatient setting, who underwent3.0T MRI and neuropsychological assessment. The statistical MRI analyses were carried-out using SPM8(Wellcome Department of Imaging Neuroscience, London, UK) running under MatlabR2008a (Math Works, Natick, MA). A standard VBM analysis was used to assess the pattern of gray and white matter changes according to previously described methods. The preprocessing steps included normalization of the images to a template, segmentation into tissue classes, modulation and smoothing. The resulting smoothed and modulated images were used in the statistical analysis to assess gray matter volume and white matter density changes. Regional volume differences were determined using statistics at every voxel in the GM and WM from patients with PD-MCI and PDCN and healthy controls. In the PD patients, a correlation was also performed between local gray matter volume, white matter density and the score of MMSE and MoCA, two tests of global cognitive function.Results:(1) In a comparison between PDCN and control subjects, GM volume was significantly reduced in prefrontal, limbic system and middle temporal gyrus, WM density was significantly decreased in the basal ganglia and frontal areas.(2) PD-MCI showed grey matter atrophy in the frontal, limbic system, basal ganglia and cerebellum areas compared to control, WM density was significantly decreased in the frontal lobe and caudate nucleus.(3) PD-MCI patients had greater gray matter atrophy in the inferior/middle temporal gyrus and frontal lobe compared to PDCN, patients with PD-MCI exhibited decreasedWM density throughout the right inferior frontal gyrus, midbrain, and occipital lobe/lingual gyrus compared with PDCN.(4) Right cingulate and limbic of GM, frontal lobe, limbic and parahippocampal gyrus of WM were correlated with performance on the memory-based test of MMSE.(5) Regions related to non-memory-based MoCA scores were gray matter regions of the frontal lobe, basal ganglia, parahippocampal gyrus, occipital lobe and cerebellum, white matter regions of the frontal lobe, parietal lobe, precuneus, occipital lobe, cuneus, the insular cortex, limbic system and cerebellum.Conclusions:These findings show that mild cognitive impairment in Parkinson’s disease is associated with structural neocortical changes in the brain, such as the frontal lobe, limbic system, the medial temporal lobe and basal ganglia, and that cognitive impairment in patients with PD without dementia may be associated with structural changes in the brain. Memory deficits related to the limbic lobe atrophy, frontal lobe atrophy associated with executive dysfunction. The atrophy of the cerebral cortex and subcortical structures may coexist in brain of PD cognitive function disorders, the underlined mechanisms may be the Lewy body degeneration and AD-like pathological changes. VBM can find subtle changes in brain even in early phase of PD cognitive impairment, the atrophy of the brain structures may become neuroimaging markers for the early diagnosis of PD-MCI. Further studies with larger groups of patients are needed to confirm these findings. Part III:Voxel-based Analysis of Diffusion Tensor Imaging in the Brain of Patients with Mild Cognitive Impairment in Parkinson’s diseaseObjective:To assess the microstructural damage to cerebral white matter and its relationship with the cognitive dysfunction in patients with Parkinson’s disease (PD) through diffusion tensor imaging (DTI).Methods:Fifteen patients with PD-MCI,17patients with PDCN, and15age-matched healthy controls underwent diffusion tensor imaging with a3T MR imager. Voxel-based DTI procedure:we calculated fractional anisotropy with DTI studio software version2.40(www.mristudio.org). The statistical MRI analyses were carried-out using SPM8(Wellcome Department of Imaging Neuroscience, London, UK) running under MatlabR2008a (Math Works, Natick, MA). We performed voxel-based analysis of variance with SPM8to detect white matter integrity differences among3groups using fractional anisotropy values as the dependent variables. In the PD patients, a correlation was also performed between the fractional anisotropy values and the score of MMSE and MoCA, two tests of global cognitive function.Results:(1) FA was significantly lower in patients with PDCN than in healthy controls in both the right frontal and precentral gyrus.(2) In a comparison between PD-MCI and control subjects, FA was significantly reduced in the left superior temporal gyrus and inferior parietal lobule, followed by the thalamus, lentiform nucleus, insular, posterior cingulate gyrus, precuneus, occipital lobe and fusiform gyrus.(3) The PD-MCI group showed significant F A reduction in the right superior parietal lobule, precuneus and left middle occipital gyrus and cuneus, followed by the thalamus, posterior cingulated, inferior parietal gyrus, limbic, parahippocampal gyrus and superior temporal gyrus compared with PDND.(4) There was a positive correlation between FA values and the performance of MMSE and MoCA in PD patients. FA values in the right thalamus, cerebellum anterior lobe, culmen occipital lobe, limbic, parahippocampal gyrus and the left midbrain, thalamus and putamen showed significant correlations with the score of MMSE. Regions related to non-memory-based MoCA scores were the limbic lobe, parahippocampal gyrus, middle temporal gyrus, inferior parietal gyrus, precuneus, occipital lobe lingual gyrus, thalamus, cerebellum anterior lobe and basal ganglia.Conclusions:These findings show that mild cognitive impairment in Parkinson’s disease is associated with local white matter fiber integrity changes in the brain, such as superior parietal lobule, precuneus, middle occipital gyrus, cuneus and middle temporal gyrus; FA values in the periventricular cingulate, precuneus, cuneus and lingual gyrus showed significant correlations with the global cognitive function, these changes were associated with cholinergic fiber damage. Integrity of white matter fiber changes may accelerate PD cognitive dysfunction. Whole brain voxel-based analysis of DTI can reflect white matter changes fully and accurately, which contributes to the early diagnosis of cognitive impairment in PD更多还原