Determinants of medium-term blood pressure variability and the related risks of stroke and dementia

Thesis

Visit-to-visit variability in blood pressure (BP) increases stroke risk, independent of mean BP. However, its physiological validity, the ideal method of measurement and the mechanisms increasing cardiovascular risk are unclear.

In meta-analyses of individual patient data, I pooled associations between BP variability and risk of stroke, all cardiovascular events and death. I then determined antihypertensive drug-class differences in cardiovascular risk, intra-individual (I-VR) and inter-individual BP variability (M-VR). In 500 Oxford Vascular Study (OXVASC) patients undergoing thrice-daily home (HBPM) and awake ambulatory monitoring (ABPM), associations between mean, maximum or variability in BP (CV-BP) were determined with premorbid BP, hypertensive arteriopathy (creatinine, aortic stiffness, cognitive impairment, stroke versus TIA and leukoaraiosis) and cardiovascular events. In 200 patients, I determined associations with pulsatility or stiffness (pulse wave velocity) in cerebral and aortic vessels.

There was a 21% and 27% increased risk of stroke and myocardial infarction per standard deviation of CV-SBP in 318700 patients, independent of mean SBP. In 244,479 patients, SBP variability was reduced by CCBs and diuretics within (I-VR=0.89, 95% CI=0.82-0.96, p=0.0001) and between individuals (M-VR 0.83, 0.77-0.89, p<0.0001), especially in the first year of treatment, explaining drug class differences in stroke risk (OR=0.76, 0.68-0.87, p<0.0001). In OXVASC, drug class differences on day-to-day SBP variability were greatest immediately after waking.

Residual hypertension after treatment on HBPM but not ABPM (BP>135/85) predicted recurrent cardiovascular events (HR 2.82, 1.44-5.51, p=0.002 vs. 1.48, 0.68-3.23, p=0.33), reflecting stronger associations with premorbid BP and hypertensive arteriopathy, due largely to inaccuracy of ABPM in patients aged >65 years. Furthermore, day-to-day maximum and CV-SBP were associated with premorbid BP, hypertensive arteriopathy and cardiovascular events, with no additional predictive value of mean SBP when analysed with maximum SBP. Maximum SBP was greater in men and CV-SBP in women, whilst age and creatinine determined both. Increased stroke risk may partly be due to the association between BP variability and cerebral pulsatility, which was correlated with leukoaraiosis (p=0.01) and determined by aortic stiffness (p=0.016) and pulsatility (p<0.001).

BP variability is clinically significant and physiologically valid, and is treatable with CCBs and diuretics. After TIA or minor stroke, HBPM best identifies residual hypertension and demonstrates the predictive value of BP variability and maximum BP, but associated arterial changes might explain some of the increased stroke risk.

Authors: Webb, AJS

• Publication date: 2014
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: Blood Pressure Variability; Meta-analysis; Dementia; Physiology; Hypertension; Stroke
• Subjects: Cognitive Neuroscience; Cardiovascular disease; Epidemiology; Physiology; Stroke