Prognostic significance of blood pressure variability in essential hypertension.

BACKGROUND: Blood pressure variability is a determinant of target organ damage in essential hypertension, but its independent prognostic significance has not yet been assessed in prospective studies of cardiovascular morbidity and mortality. OBJECTIVE: To assess the relationship between blood pressure variability, assessed non-invasively using 24 h ambulatory blood pressure monitoring and subsequent incidence of cardiovascular morbid events in persons with essential hypertension. DESIGN: Prospective observational study. PATIENTS AND METHODS: We followed for up to 8.6 years (mean 2.92) 1372 individuals with essential hypertension whose initial off-therapy diagnostic work-up included 24 h non-invasive ambulatory blood pressure monitoring. Those with a standard deviation of daytime or night-time blood pressure below or above the group mean were classified as having low or high blood pressure variability, respectively. One hundred and eighty-two participants underwent repeated ambulatory blood pressure monitoring and echocardiography during follow-up, 2.7 years later. RESULTS: Target organ damage score was greater in the participants with high variability of daytime (P = 0.004) and night-time (P = 0.011) systolic blood pressure than in those with low blood pressure variability. In those who underwent repeated echocardiography, for every quartile of baseline ambulatory blood pressure, left ventricular mass at follow-up was greater (all P < 0.05) in those with high baseline blood pressure variability than in those with low baseline variability. During follow-up there were 106 major cardiovascular morbid events. Event rate was 1.99 and 3.26 events per 100 patient-years, respectively, in participants with low and high variability of daytime systolic pressure and 1.98 and 3.38 events per 100 patient-years, respectively, in those with low and high variability of night-time systolic pressure (log-rank test: both P < 0.05). However, in a Cox multivariate analysis, the variability score for daytime and night-time systolic pressure failed to enter the model (age, diabetes mellitus, previous cardiovascular events and average night-time systolic pressure were independently associated with cardiovascular events). CONCLUSION: Increased blood pressure variability, assessed with non-invasive monitoring, is associated with a higher incidence of cardiovascular morbid complications of hypertension, but also with a higher blood pressure, older age and a higher prevalence of diabetes mellitus. Because of the relevant predictive effect of these associated factors, the adverse prognostic significance of increased blood pressure variability is no longer detectable in multivariate analysis.