Blood Pressure and Brain Lesions in Patients With Atrial Fibrillation

Abstract

The association of blood pressure (BP) and hypertension with the presence of different types of brain lesions in patients with atrial fibrillation is unclear. BP values were obtained in a multicenter cohort of patients with atrial fibrillation. Systolic and diastolic BP was categorized in predefined groups. All patients underwent brain magnetic resonance imaging and neurocognitive testing. Brain lesions were classified as large noncortical or cortical infarcts, small noncortical infarcts, microbleeds, or white matter lesions. White matter lesions were graded according to the Fazekas scale. Overall, 1738 patients with atrial fibrillation were enrolled in this cross-sectional analysis (mean age, 73 years, 73% males). Mean BP was 135/79 mm Hg, and 67% of participants were taking BP-lowering treatment. White matter lesions Fazekas ≥2 were found in 54%, large noncortical or cortical infarcts in 22%, small noncortical infarcts in 21%, and microbleeds in 22% of patients, respectively. Compared with patients with systolic BP <120 mm Hg, the adjusted odds ratios (95% CI) for Fazekas≥2 was 1.25 (0.94-1.66), 1.41 (1.03-1.93), and 2.54 (1.65-3.95) among patients with systolic BP of 120 to 140, 140 to 160, and ≥160 mm Hg (P for linear trend<0.001). Per 5 mm Hg increase in systolic and diastolic BP, the adjusted β-coefficient (95% CI) for log-transformed white matter lesions was 0.04 (0.02-0.05), P<0.001 and 0.04 (0.01-0.06), P=0.004. Systolic BP was associated with small noncortical infarcts (odds ratios [95% CI] per 5 mm Hg 1.05 [1.01-1.08], P=0.006), microbleeds were associated with hypertension, but large noncortical or cortical infarcts were not associated with BP or hypertension. After multivariable adjustment, BP and hypertension were not associated with neurocognitive function. Among patients with atrial fibrillation, BP is strongly associated with the presence and extent of white matter lesions, but there is no association with large noncortical or cortical infarcts. Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT02105844.

Keywords: atrial fibrillation; blood pressure; brain; magnetic resonance imaging; white matter.

Figure 1.

Brain magnetic resonance imaging (fluid-attenuated inversion recovery) of 2 different patients with atrial fibrillation. A, Normotensive man, 72 y old with a small amount of white matter lesions (Fazekas score <2, arrows). B, Uncontrolled hypertensive man, 74 y old with white matter lesions inferior (Fazekas score ≥2, arrows), lacunar lesions (circle), and cortical infarcts (right bottom).

Figure 2.

Prevalence of brain lesions stratified by systolic blood pressure categories. Faz indicates Fazekas; LNCCI, large noncortical or cortical infarcts; MB, microbleeds; and SNCI, small noncortical infarcts.

Figure 3.

Scatterplot of the age and sex-adjusted association of blood pressure with volume of white matter lesions. The solid line represents the model-based predicted values, and the dotted lines represent the 95% pointwise CI.