Subclinical Magnetic Resonance Imaging Markers of Cerebral Small Vessel Disease in Relation to Office and Ambulatory Blood Pressure Measurements

Abstract

Background: Twenty-four-hour and nighttime blood pressure (BP) levels are more strongly associated with cardiovascular risk than office or daytime BP measurements. However, it remains undocumented which of the office and ambulatory BP measurements have the strongest association and predictive information in relation to the presence of type I, or arteriolosclerosis type, cerebral small vessel diseases (CSVD).

Methods: A subset of 429 participants from the Maracaibo Aging Study [aged ≥40 years (women, 73.7%; mean age, 59.3 years)] underwent baseline brain magnetic resonance imaging (MRI) to visualize CSVD, which included log-transformed white matter hyperintensities (log-WMH) volume and the presence (yes/no) of lacunes, cerebral microbleeds (CMB), or enlarged perivascular spaces (EPVS). Linear and logistic regression models were applied to examine the association between CSVD and each +10-mmHg increment in the office and ambulatory systolic BP measurements. Improvement in the fit of nested logistic models was assessed by the log-likelihood ratio and the generalized R ² statistic.

Results: Office and ambulatory systolic BP measurements were related to log-WMH (β-correlation coefficients ≥0.08; P < 0.001). Lacunes and CMB were only associated with ambulatory systolic BP measurements (odds ratios [OR] ranged from 1.31 [95% confidence interval, 1.10-1.55] to 1.46 [1.17-1.84], P ≤ 0.003). Accounted for daytime systolic BP, both the 24-h (β-correlation, 0.170) and nighttime (β-correlation, 0.038) systolic BP measurements remained related to log-WMH. When accounted for 24-h or daytime systolic BP levels, the nighttime systolic BP retained the significant association with lacunes (ORs, 1.05-1.06; 95% CIs, ≥1.01 to ≤ 1.13), whereas the 24-h and daytime systolic BP levels were not associated with lacunes after adjustments for nighttime systolic BP (ORs, ≤ 0.88; 95% CI, ≥0.77 to ≤ 1.14). On top of covariables and office systolic BP, ambulatory systolic BP measurements significantly improved model performance (1.05% ≥ R ² ≤ 3.82%). Compared to 24-h and daytime systolic BP, nighttime systolic BP had the strongest improvement in the model performance; for WMH (1.46 vs. 1.05%) and lacunes (3.06 vs. ≤ 2.05%).

Conclusions: Twenty-four-hour and nighttime systolic BP were the more robust BP measurements associated with CSVD, but the nighttime systolic BP level had the strongest association. Controlling ambulatory BP levels might provide additional improvement in the prevention of CSVD.

Keywords: ambulatory blood pressure monitoring; cerebral microbleeds; cerebral small vessel disease; enlarged perivascular spaces; lacunes; nighttime blood pressure; white matter hyperintensities.

Figure 1

Heat maps displaying the probability (%) of having subclinical markers of cerebral small vessel disease according to office and ambulatory blood pressure Categories. BP indicates blood pressure; WMH, white matter hyperintensities; HT, hypertension. Normal, elevated BP, stage-1, stage-2, and severe hypertensions thresholds were 120/130/140/160 mm Hg for office systolic BP, 115/125/130/145 for 24-h, 120/130/135/145 for daytime, and 100/110/120/140 for nighttime BP. Numbers inside the box indicates the probability of having top-90th total log-WMH (A), lacunes (B), cerebral microbleeds (C), or enlarged perivascular spaces (D). The probability was derived by multivariable logistic regression and was standardized to the average of the distributions in the whole study population of sex, age, education, cephalic circumference, body mass index, diabetes mellitus, use of antihypertensive treatment, glomerular treatment, and history of cardiovascular diseases.

Figure 2

Association between total log-transformed white matter hyperintensities and standardized ambulatory systolic blood pressure measurements. The ambulatory blood pressure (BP) measurements were all standardized by sex, age, education, cephalic circumference, body mass index (BMI), diabetes mellitus, high-density serum cholesterol, use of antihypertensive treatment, glomerular filtration rate, and history of cardiovascular diseases. The 24-h systolic BP was additionally adjusted for daytime (A) and nighttime (B) systolic BP levels. The daytime systolic BP was additionally adjusted for 24-h (C) and nighttime (D) systolic BP levels. Finally, the nighttime systolic BP was additionally adjusted for 24-h (E) and daytime (F) systolic BP levels.