Alzheimer's disease (AD) and vascular dementia (VaD) are the most common forms of senile dementia. Although the animal research of dementia has made remarkable progress, clinical trials of drugs for AD pathology have failed in recent years. The study of dementia based on cell and animal model generally aims at a single mechanism and target, and its results are quite different from the real clinical environment. More and more studies suggest that investigators should shift the focus of research to the early stage of cognitive impairment before dementia. Prevention is more important than cure, and intervention against multi-factors and multi-targets has become an important consensus. A large number of studies have shown that the mechanism of vascular brain injury plays an important role in the pathogenesis of AD and VaD, and many vascular risk factors are interventionable to some extent. Therefore, based on the clinical cohort, in-depth study of vascular cognitive impairment (Vascular cognitive impairment, VCI) has important clinical significance for the effective prevention and treatment of AD and VaD.

The leading team of the project has focused on VCI research for a long time. After nearly 20 years of experimental research and preliminary clinical observation, it is proposed that chronic cerebral ischemia can not only be a clinical disease entity, but also an important pathological basis for the early onset of VCI. This view has recently been supported by a number of authoritative international research evidence. Big data's study of 1171 patients with AD reported by Nature Commun in 2016 shows that the early pathological changes of AD may not be a cascade of amyloid protein (Aβ), but a decrease in cerebral blood flow. Therefore, this project intends to establish an early clinical research cohort of VCI to focus on three key issues in VCI research and clinical practice: (1) the theory that cerebral hypoperfusion may be an important pathological basis for the occurrence and development of VCI needs direct evidence support from clinical studies, and its mechanism needs further elucidation. (2) Based on the fusion of multimodal MRI of VCI vascular brain injury pathology and PET imaging markers of Aβ molecular pathology, a multivariate VCI cognitive evaluation model is constructed, and its sensitivity and specificity may be better than the existing VCI diagnostic standards. (3) the protective effect of early comprehensive intervention of vascular risk factors on cognitive decline in VCI may be more effective than that of single risk factor.

The first part of this project is to establish a study cohort of non-demented vascular cognitive impairment(VCIND). Neurocognitive function assessment combined with multimodal MRI including ASL, DCE, DTI and BOLD techniques were used to observe the role of cerebral hypoperfusion in the early stage and progression of VCI. At the same time, the relationship between the changes of blood-brain barrier and neural network and cognitive decline was dynamically observed to verify and explore the effect and mechanism of cognitive impairment caused by cerebral hypoperfusion. The second part studies the pathology of vascular brain injury based on MRI and the molecular pathology of A β based on PET and the relationship between Aβ molecular pathology and cognitive impairment, including the main factors affecting cognitive function, and uses artificial intelligence (AI) algorithm to develop a multiple quantitative evaluation system of VCI cognitive function, which is mainly based on the fusion of MRI and PET image markers. In the third part, a multicenter randomized controlled clinical cohort study was conducted to observe the cognitive protective effect of comprehensive intensive intervention of vascular risk factors on early VCI, so as to provide direct clinical evidence and intervention model for the prevention and treatment of VCI. The topics of the above three aspects covered by this project are closely related, which is not only a key scientific problem, but also an important clinical problem to be solved in the diagnosis and treatment of VCI.

The study of this project is expected to further clarify the role and mechanism of cerebral hypoperfusion in VCI, provide a new theoretical basis for the prevention and treatment of dementia, and develop a quantitative evaluation system of VCI cognitive function mainly based on imaging technology and AI algorithm, so as to provide a more accurate and convenient diagnostic tool for early clinical identification and scientific research of VCI. Draw up the early comprehensive intervention paradigm of VCI based on vascular risk factors and popularize it in clinic, gradually form an expert consensus, enrich and update the guidelines for diagnosis and treatment of dementia, and effectively improve the level of prevention and treatment of dementia related to VCI.

The main research contents of this project focus on the above three key issues: to study the role and mechanism of cerebral hypoperfusion in the early pathogenesis of VCI; to establish a predictive model for early diagnosis of MRI vascular brain damage and PET-A β pathology; to explore the clinical effect of early comprehensive intensive intervention of VRF on the prevention and treatment of VCI.

1. To explore the role and mechanism of cerebral hypoperfusion in the early brain damage of VCI.

1. The relationship between vascular risk factor burden and VCI Vascular risk factor (VRF) is not only an important factor in the occurrence and development of VCI, but also the main intervention target. Elucidating the relationship between VRF and VCI can provide an important theoretical basis for its early prevention and treatment. Some clinical studies have shown that long-term VRF burden is closely related to the decrease of global cerebral blood flow in young and middle-aged people without stroke, suggesting that there is a close relationship between VRF, cerebral hypoperfusion and cognitive impairment, but it needs to be further clarified in clinical cohort studies. In this project, a clinical cohort of VCIND was established, including 500 middle-aged and elderly people with VRF burden, and their complete demographic information was collected. The overall VRF burden was assessed by the Framingham Stroke risk Assessment scale, and the subjects were divided into low risk group (10-year stroke risk < 10%), medium risk group (10-year stroke risk 10%) and high-risk group (10-year stroke risk > 20%) for follow-up. The changes of cognitive function in baseline and 3-year follow-up period were measured. The relationship between VRF burden and cognitive decline in VCI was analyzed by mixed effect model, and whether decreased cerebral perfusion mediated this correlation was analyzed by mediating effect model.
2. The role of cerebral hypoperfusion in the early brain damage of VCI. Many experimental studies have shown that chronic cerebral hypoperfusion may play an important role in the occurrence and development of VaD and AD. Big data of 1171 patients with late-onset AD in the last study showed that the early pathological changes in AD may not be an A β cascade, but a decrease in cerebral blood flow, suggesting that cerebral hypoperfusion may play an important role in cognitive decline. However, there is still a lack of systematic prospective study on the role of cerebral hypoperfusion in the early pathogenesis of VCI. Based on the establishment of a cohort of early VCI population, arterial spin labeled magnetic resonance imaging (ASL-MRI) was used to observe the changes of whole brain and regional cerebral blood flow in 300 VCIND subjects during the 3-year follow-up period. The subjects were divided into 3 groups for follow-up according to the tertile value of the relative total cerebral blood flow (rCBF) at baseline. At the same time, 100 patients with normal cognitive function matching gender, age, and education level were included as controls. The changes of cognitive function of the subjects were evaluated with a complete neuropsychological scale. The follow-up outcome indicators were as follows: VCIND progressed to VaD,VCIND progressed to AD or other types of dementia, declining VCIND, stable VCIND, and cognition remained normal. The mixed effect model was used to analyze the role of cerebral hypoperfusion in the early brain damage of VCI.
3. Damage of blood-brain barrier and neural network in the early stage of VCI. The integrity of neural network is the neural basis of cognitive function, and the blood-brain barrier is an important part of neurovascular units. Previous experimental studies of investigators team have shown that chronic cerebral hypoperfusion can lead to damage of neural networks and blood-brain barrier in many brain regions, including hippocampus. The subjects were followed up longitudinally to observe the changes of blood-brain barrier and neural network. The outcome index of 3-year follow-up was as follows: VCIND progressed from VaD, to AD and other types of dementia, VCIND, stable VCIND and cognition remained normal. By comparing the multi-dimensional relationship between blood-brain barrier, neural network and cognitive outcome in time and space, the mechanism of brain function and structure damage in patients with early VCI was further revealed.

2. Early VCI diagnosis prediction model based on imaging technology and its key techniques.

Previous studies have shown that the occurrence and development of VCI is not only closely related to structural MRI indexes such as cerebral infarction, cerebral atrophy and cerebral microvascular disease, but also to functional MRI indexes such as cerebral hypoperfusion and brain ultrastructural damage. However, a VCI quantitative evaluation system based on the above imaging index fusion has not been established at present. On the other hand, up to 40-80% of VCI patients have AD-like pathological changes characterized by Aβ. The impact of Aβ pathology on cognitive decline in patients with VCI and whether A β pathological indicators need to be included in the diagnosis of VCI still need to be clarified in cohort studies. This project intends to use artificial intelligence (AI) algorithm, combined with MRI vascular brain pathological damage and related factors affecting cognitive function, to build a VCI diagnosis prediction model suitable for clinical application, and verify it in the cohort study. At the same time, the Aβ molecular pathological images based on PET were integrated to establish a VCI diagnosis prediction model suitable for clinical scientific research, and further clarify the possible role of Aβ pathology in the occurrence and development of VCI.

1. Establishment of VCI diagnosis and prediction model. Three hundred non-dementia subjects (including pathology of VRF and / or vascular brain damage) were enrolled. According to the clinical data, imaging data and systematic neuropsychological evaluation results, all subjects were divided into VCIND and non-VCIND (including cognitive normal or other types of cognitive impairment). The changes of cognitive function were systematically evaluated during the 3-year follow-up. Two types of outcome indicators are defined as cognitive decline and cognitive retention. The indexes of different degrees and forms of brain injury (including vascular risk factors, structural MRI, ASL-MRI and DTI) were recorded at baseline, and the corresponding quantitative analysis was carried out with the outcome index. AI algorithm was used to establish an early VCI diagnosis model suitable for clinical application. According to the cognitive outcome of the 3-year follow-up period, a model was established to predict the cognitive outcome of VCIND by imaging indicators at the baseline. In addition, 50 subjects in the VCIND group and 50 subjects in the non-VCIND group at the baseline were tested by 11C-PiB PET, and the Aβ pathological images were fused with the above-mentioned MRI images to establish a VCI diagnosis and evaluation system suitable for clinical scientific research.
2. Verification of VCI diagnostic and predictive model. At the end of the 3-year follow-up, the brain damage and cognitive function of the subjects were re-evaluated. According to the newly diagnosed VCIND, the reliability and validity of the diagnostic and predictive model were verified, and the clinical application model and scientific research model were verified.The sensitivity and specificity of type A were compared and analyzed.

3. A cohort study on the effect of comprehensive intervention of vascular risk factors on vascular cognitive impairment.

1. Study on the effect of early comprehensive intensive intervention on vascular cognitive impairment.

   The purpose of this study was to establish a cohort of VCIND (n = 250) and a cohort of cognitive normal people with VRF (n=250). Each cohort was divided into two groups: comprehensive intensive intervention group and general intervention group. Comprehensive intensive intervention provides guidance on blood pressure, blood sugar, lipid regulation, exercise and other aspects according to the corresponding VRF, and formulates specific intervention programs and quality control standards. Medical staff implement tracking management and regular reminders through smart wearable devices and cloud platform data to achieve the goal of gradual intervention. General data, vascular risk factors and neuropsychological assessment were completed at baseline and 1-year, 2-year, 3-year follow-up. Mixed random effect regression model and Cox proportional hazard model were used to analyze the effects of VRF and comprehensive intervention on the outcome of end events (main outcome: dementia, secondary outcome: stroke, myocardial infarction, cognitive impairment, dementia subtype and mortality). According to the Fuminghan stroke risk subgroup, the differences of clinical effects of comprehensive intervention in different subgroups were statistically analyzed.

2. Effect of comprehensive intensive intervention on early cerebral perfusion and Aβ in patients with vascular cognitive impairment.

At the end of baseline and 3-year follow-up, 50 patients in the VCIND comprehensive intensive intervention group and 50 patients in the general treatment group completed the detection of ASL cerebral blood flow and 11C-PiB PET cerebral Aβ. The effects of comprehensive intensive intervention on cerebral blood flow, Aβ and cognitive function were analyzed. To further analyze the relationship between the changes of cerebral blood flow and A β pathology, as well as its influence on cognitive function, to clarify the target and mechanism of early comprehensive intensive intervention, and to establish a perfect evaluation standard of imaging effect.

Inclusion Criteria:

• ①Healthy volunteers or VCIND patients（Diagnostic criteria comply with the "Guidelines for the Diagnosis and Treatment of Vascular Cognitive Impairment" formulated by the Dementia and Cognitive Impairment Group of the Neurology Branch of the Chinese Medical Association）;
• ②Age between 50-70 years old;
• ③Able to accept and cooperate in completing various instrumental examinations and neuropsychological examinations Scale testing;
• ④Agree to sign an informed consent form.

Exclusion Criteria:

• ①Combined with other serious diseases, life expectancy is less than 3 years;
• ②Patients with vision and hearing impairment that significantly affect cognitive testing;
• ③Severe heart, brain, lung, kidney and other diseases;
• ④Have a history of psychoactive substance abuse;
• ⑤Have other serious physical diseases that affect cognitive testing, such as coma, epilepsy, hypothyroidism , Hypoxemia, etc.;
• ⑥have a history of mental illness;
• ⑦have MRI contraindications, unwilling to participate in research and unconditional follow-up.