Visit-to-Visit Blood Pressure Variability and Arterial Stiffness Independently Predict Cardiovascular Risk Category in a General Population: Results from the SEPHAR II Study.

INTRODUCTION The aim of our study was to evaluate visit-to-visit blood pressure variability (BPV) and the association of this parameter with cardiovascular risk determinants, according to the SEPHAR II survey. METHODS Following a selection based on the multi-stratified proportional sampling procedure, a total of 1975 subjects who gave informed consent were evaluated by means of a questionnaire, anthropometric, blood pressure (BP) and arterial stiffness measurements (pulse wave velocity and augmentation index), 12-lead ECG recordings, and blood and urine analysis. BPV was quantified in terms of the standard deviation (SD) of the mean systolic blood pressure (SBP) and high BPV was defined as SBP-SD above the 4th quartile. Total cardiovascular risk was assessed by the 2013 ESH/ESC risk stratification chart. RESULTS Mean BP was 132.37/82.01 mmHg. Mean systolic BPV was 6.16 mmHg, with 24.62% of values above the 75th percentile (8.48 mmHg). Factors found to be associated with high systolic BPV were age, SBP, pulse pressure, total and LDL-cholesterol, triglycerides, visceral obesity, diabetes mellitus, metabolic syndrome and increased aortic stiffness. In addition, in the hypertensive group high BPV was associated with the severity of hypertension and a lack of treatment control. Both visit-to-visit systolic BPV and aortic stiffness proved to be positively and independently correlated with the risk category. Based on these parameters it was possible to predict with 72.6% accuracy the probability of finding subjects in a high and very high cardiovascular risk category. CONCLUSIONS The results of our study indicate a notable prevalence of high BPV, affecting almost a quarter of the Romanian adult population. Visit-to-visit systolic BPV and arterial stiffness are strongly correlated and together might contribute to the improvement of cardiovascular risk prediction models.