This functional near-infrared spectroscopy-based personalized multidomain intervention study aims to prevent cognitive impairment and reduce dementia and cerebrovascular events in 45-74 years old persons with high risk of stroke in China. The primary outcome is 6-months change in global cognitive score measured by a modified National Institute of Neurological Disorders and Stroke and Canadian Stroke Network-Canadian Stroke Network protocol. The investigators hypothesize that the intervention based on functional near-infrared spectroscopy will prevent cognitive decline by the initial 6-months intervention. The long-term primary outcome is the development of dementia and cerebrovascular events during a total of 2 years' follow-up. The investigators hypothesize that the functional near-infrared spectroscopy-based personalized intervention may reduce the 2-year risk of dementia and cerebrovascular events, mainly through the improvement in vascular risk factors control, social activity, and cognitive training activities.

Patients with ≥ 3 stroke risk factors (including hypertension, dyslipidemia, diabetes, atrial fibrillation or valvular heart disease, smoking history, obvious overweight or obesity, lack of exercise, family history of stroke), or with transient ischemic attack, are regarded as patients with high risk of stroke. Studies have indicated that these stroke risk factors might be associated with an increased risk of cerebral small vessel disease (CSVD) progress, glymphatic dysfunction, cognitive decline, dementia, and cerebrovascular events. However, prevention in these patients is largely unknown and the management of these patients is a very troublesome issue. Management on those patients mainly focused on improving lifestyle. Recently, cognitive training has emerged, with change from strategy-oriented paper-and-pencil and instructional training methods to computer-aided cognitive training that is difficulty adaptive and focuses on ability improvement. Studies have shown that the functional state of the brain is more sensitive and specific than the behavioral performance, so it provides a theoretical basis for personalized intervention. Functional near-infrared spectroscopy imaging converts signal values into oxygenated hemoglobin, deoxyhemoglobin, and changes in overall hemoglobin concentration according to Beer-Lambert's law to reflect brain activity levels, which is more convenient than EEG and functional magnetic resonance in collecting task-state signal. By pre-analyzing the correlation between the activation level of each brain region and cognition, the brain area related to cognitive function under each task can be preliminarily obtained, and the patient can be further judged whether the patient needs to be trained for the task according to the activity level of the brain region to achieve the purpose of personalization. The investigators hypothesize that the intervention based on functional near-infrared spectroscopy imaging will reduce cognitive impairment, CSVD progress, and reduce dementia and cerebrovascular events incidence in the study group compared to the control group.

Inclusion Criteria:

• Aged 45-74 years
• high risk of stroke (with≥3 of 8 stroke risk factors, including hypertension, dyslipidemia, diabetes, atrial fibrillation or valvular heart disease, smoking history, obvious overweight or obesity, lack of exercise, family history of stroke, or with transient ischemic attack)

Exclusion Criteria:

• previously diagnosed dementia
• previously diagnosed stroke (both cerebral infarction and hemorrhage)
• suspected dementia after clinical assessment by study physician at screening visit
• Mini-mental State Examination [MMSE] score<20
• disorders affecting safe engagement in the intervention (e.g., malignant disease, major depression, symptomatic cardiovascular disease, revascularization within 1 year previously)
• severe loss of vision, hearing, or communicative ability
• disorders preventing cooperation as judged by the study physician
• coincident participation in another intervention trial
• any MRI contraindications