OBJECTIVE:The aim of the present study was to assess the predictive role of changes inarterial stiffness for the incidence of stroke in a cohort of essential hypertensive patients. DESIGN AND METHOD:We followed up 1082essentialhypertensives (mean age 55.9 years, 562 males, office blood pressure (BP) = 145/91 mmHg)for a mean period of 8 years. All subjects had at least one annual visit and arterial stiffness was evaluated on the basis of carotid to femoral pulse wave velocity (PWV), by means of a computerized method at the initial and last visit. The distribution of baseline PWV was split by the median (8.2 m/sec) and accordingly subjects were classified into those with high (n = 546) and low values (n = 536). Stroke was defined as rapid onset of a new neurological deficit persisting at least 24 hours unless death supervened confirmed by imaging findings. RESULTS:The incidence of stroke over the follow-up period was 2.2%. Hypertensives who had stroke (n = 24) compared to those without stroke at follow-up (n = 1058) were older at baseline (65 ± 9vs56 ± 12 years, p = 0.032), had higher office BP levels (155 ± 13vs145 ± 15 mmHg, p = 0.014) and prevalence of high PWV levels (67% vs 40%, p = 0.021). No difference was observed between hypertensives with stroke and those without stroke with respect to baseline renal function and lipid levels (p = NS for all). By univariate Cox regression analysis it was revealed that changes in PWV levels between baseline and last visit predicted stroke (hazard ratio = 1.352, p = 0.004). Moreover, in multivariate Cox regression model, baseline age (hazard ratio = 1.087, p = 0.03), changes in PWV (hazard ratio = 1.115, p = 0.024) but not changes in office BP levels turned out to be independent predictors of stroke. CONCLUSIONS:In essential hypertensive patients, changes in PWV predict future development of stroke, independently of established confounders, including BP. These findings support that PWV constitutes a potent prognosticator of cerebrovascular events and its estimation is essential in order to improve risk stratification in hypertension.