Hypertension in midlife is an independent risk factor of late life cognitive dysfunction or dementia. Chronic hypertension cause vascular damage and cerebral ischemia, which ultimately gives rise to the cognitive dysfunction or dementia.

A recent study showed that high visit-to-visit variability in clinic systolic blood pressure (BP) was a strong independent predictor of stroke. This finding suggests that high clinic systolic BP variability itself as well as chronic hypertension may cause vascular damage and cerebral ischemia. Therefore, high clinic SBP variability may be also an independent risk factor of cognitive dysfunction or dementia.

Vascular damage leads to the diminished autoregulatory capacities of cerebral arteries. The brain with the reduced autoregulatory capacity may be more vulnerable to BP fluctuation. Therefore, high BP variability may be more harmful in patients with damaged vessels (for example, in patients with cerebral small vessel disease).

Previous data about BP variability and cognition revealed very controversial. Some studies showed poor cognition in patients with high BP variability, but others did not.

The previous studies were mostly based on cross-sectional designs, and performed in small-sized heterogeneous population for primary prevention. The harmful effect of high BP variability may be clearer in the population with damaged vascular bed, such as cerebral small vessel disease. The previous studies usually used ambulatory BP monitoring (ABPM). However, recent data suggested that variability in BP on ABPM may be a weaker predictor of vascular events than be visit-to-visit variability in clinic BP.

The investigators sought to find whether high visit-to-visit variability in clinic BP is related with poor cognitive function in patients with cerebral small vessel disease.

This is a retrospective cohort study.

We include patients with cerebral small vessel disease, documented on MRI from Jan 2006 to Dec 2010, who have been regularly followed up.

We evaluate the patients' cognitive function after written informed consent.

We independently review patients' medical record and analyze MRI data. BP variability parameters include standard deviation(SD, primary measuring parameter), coefficient of variation, successive variation, average real variability (ARV), SD independent of mean(SDIM), SV independent of mean(SVIM), and ARV independent of mean (ARVIM). We will adjust following confounding variables: age, sex, level of education, vascular risk factors, mean SBP and DBP, NIHSS score, and white matter lesion burden on T2-weighted MRI.

Inclusion Criteria:

• Patients with cerebral small vessel disease previously documented on MRI (definition of small vessel disease - symptomatic lacunar infarction or white matter ischemic lesion with one or more asymptomatic lacunes)

Exclusion Criteria:

• Incomplete clinic BP data (less than 6 BP readings during recent one year)
• History of cardiovascular or cerebrovascular events during recent one year
• Documented cerebral infarction from large artery atherosclerosis
• Atrial fibrillation or cardiac disease with high risk of embolism
• Significant medical, neurological, or psychiatric disease affecting cognition
• Known dementia treated with acetylcholine esterase inhibitor or memantine
• Patients without informed consent

Outpatients of stroke clinic in secondary or tertiary hospitals