Pulse pressure and endothelial dysfunction in hypertension

Cardiology Review® Online, February 2004, Volume 21, Issue 2

From the Internal Medicine and Cardiovascular Diseases Unit, Department of Medicina Sperimentale e Clinica G. Salvatore, University Magna Græcia of Catanzaro, Italy

Pulse pressure is an independent predictor of cardiovascular morbidity and mortality.1,2 The concept of bidirectionality may explain the association between pulse pressure and adverse cardiovascular events, because an increase in pulse pressure causes and affects atherosclerosis.2

Endothelial dysfunction, which is associated with traditional and emerging risk factors, occurs early in vascular disease. It is characterized by a loss of endothelium-dependent vasodilation because of a decrease in the bioavailability of nitric oxide.3-5 This condition is clinically relevant because it may be considered a marker of subsequent cardiovascular outcomes.6,7

The association between pulse pressure and endothelial dysfunction in an experimental model was recently described.8 There are no data, however, on the association between pulse pressure and acetylcholine-stimulated vasodilation in hypertensive humans. We investigated the possible association between pulse pressure and endothelial dysfunction in a group of untreated hypertensive patients.

Patients and methods

Our study included 130 male and 132 female hypertensive outpatients, ages 30 to 55 years (mean ± SD, 46.1 ± 5.7 years), at Catanzaro University Hospital. Of the 262 participants, 225 were included in our previous report.7 None of the participants had been treated with an antihypertensive drug.

Clinic blood pressure readings were obtained in the left arm of the supine patients with a mercury sphygmomanometer after 5 minutes of quiet rest. Patients with a clinic systolic blood pressure of 140 mm Hg or above or a diastolic blood pressure of 90 mm Hg or above were considered hypertensive.

All vascular function studies were performed at 9:00 am. Patients had fasted overnight and were in the supine position in a quiet, air-conditioned room (22°C—24°C). We used the study protocol previously described by Panza and colleagues and subsequently used by our group.5,7 Forearm blood flow was measured as the slope of the change in forearm volume. We calculated the mean of three measurements or more at each time point. Forearm vascular resistance was expressed in arbitrary units (U). This value was calculated by dividing mean blood pressure by forearm blood flow.

All participants rested for at least 30 minutes after artery cannulation to obtain a stable baseline before data collection. Forearm blood flow and vascular resistance measurements were repeated every 5 minutes until stable. We determined endothelium-dependent and -independent vasodilation by assessing the dose-response curve to intra-arterial infusions with increasing doses of acetylcholine (7.5, 15, and 30 µg/mL—1/min–1 for 5 minutes each) and sodium nitroprusside (0.8, 1.6, and 3.2 µg/mL–1/min–1 for 5 minutes each). The drug infusion rate, adjusted for the forearm volume of each subject, was 1 mL/min.

We performed standard descriptive and comparative analyses. The dose-response curve to acetylcholine and sodium nitroprusside were compared by analysis of variance

for repeated measurements and by

a post-hoc Tukey test. The effects

of different covariates on forearm blood flow were evaluated by a

simple linear regression analysis and, successively, by a stepwise

multiple linear regression. One at a time, we entered the different clinic and monitored blood pressure components—systolic blood pressure, diastolic blood pressure, mean blood pressure, and pulse pressure—into a baseline model including age, sex, body mass index (BMI), glucose, cholesterol, triglycerides, and smoking as independent variables. Parametric data were reported as mean ± SD, and results were considered significant if the P value was below .05.


Table 1 summarizes the clinical, humoral, and hemodynamic characteristics of the study patients. Intra-arterial acetylcholine infusions induced a significant dose-dependent increase in forearm blood flow (P < .001) and a decrease in forearm vascular resistance. The forearm blood flow increments at the three incremental doses of acetylcholine were 2.1 ± 1.3 ( 65%), 5.1 ± 2.7 ( 153%), and 10.2 ± 4.9 mL/100 mL—1 of tissue/min–1 ( 307%).

An intra-arterial infusion of sodium nitroprusside also induced a

significant increase in forearm blood flow (P < .001) and a decrease in forearm vascular resistance. The increments from baseline were 80% ± 38%, 169% ± 63%, and 354% ± 84%. An intra-arterial infusion of vasoactive agents did not change blood pressure or heart rate values.

As reported in table 2, a significant correlation was detected between the peak increase in acetylcholine-stimulated forearm blood flow and different clinical variables. A stepwise multivariate analysis showed that adding blood pressure components to the baseline model, which included age, sex, and BMI, explained the 10.5% variation and significantly increased the forearm blood flow variation (P < .001). Clinic systolic blood pressure increased by up to 22.6%, clinic pulse pressure increased by up to 29.9%, monitored systolic blood pressure increased by up to 31.3%, and monitored pulse pressure increased by up to 42.8%. Adding monitored

systolic blood pressure explained another 1.1% of the variation. Therefore, the total model explained 43.9% of the forearm blood pressure variation (P < .001). It shows that acetylcholine-stimulated forearm blood pressure decreases by 8.7% for each 1-mm Hg increase in monitored pulse pressure.

When we placed patients into quartiles of monitored pulse pressure (< 53 mm Hg, 53—58 mm Hg, 58–65 mm Hg, and > 65 mm Hg) and used analysis of variance, we observed a significant increase in acetylcholine-stimulated forearm blood flow (P < .001; figure). We found that systolic rather than dia-stolic blood pressure contributed to the increase in pulse pressure. Systolic blood pressure increased from 141 ± 6 mm Hg, which was the lower quartile, to 158 ± 10 mm Hg, the upper quartile (analysis of variance P < .001). Diastolic blood pressure decreased from 93 ± 6 mm Hg, the lower quartile, to 87 ± 9 mm Hg, which defined the upper quartile (analysis of variance P < .001).


The major finding of our study was the inverse relationship between monitored pulse pressure

and acetylcholine-stimulated vasodilation in a group of untreated and uncomplicated hypertensive patients. This relationship persisted after adjusting for other covariates.

Under physiological conditions, the endothelial cells produce a variety of vasoactive substances that regulate vascular homeostasis.3,4,9 One of these is nitric oxide, which may be released after stimulation

by endogenous and pharmacologic agonists and physical stimuli, such as flow-mediated shear stress.9

Endothelial dysfunction has been implicated in the pathophysiology of essential hypertension, accounting for the increase in vascular resistance and for the vascular structural changes that occur.5,7 Endothelial dysfunction may be considered an early modification that promotes and develops coronary and extracoronary atherosclerosis. Chronic endothelial dysfunction may be associated with the erosion or rupture of the atherosclerotic plaque, which promotes plaque instability and acute vascular syndromes.3,4

A constant-flow—mediated shear stress modulates vascular diameter by vasoactive agents released by the endothelium. This stress interacts with experimental atherogenesis, modulating nitric oxide production.10 In human coronary and peripheral conductance arteries, low-flow–mediated shear stress interferes with endothelium-dependent vasodilation.11,12 Flow pulsatility has been documented at the sites of atherosclerotic lesions but turbulent flow has not.13 This finding confirms that oscillatory and steady laminar shear stress exert differential effects on endothelial cells. De Keulenaer and colleagues showed that continuous oscillatory shear stress or pulsatile flow cause the sustained activation of endothelial pro-oxidant processes.14 This activation results in redox-sensitive gene expression. A possible explanation is that the up-regulation of nitric oxide synthase III induces activation of nuclear factor-kappa B, resulting in an increase in oxidative stress. Thus, it is possible to affirm that laminar-flow–mediated shear stress, not oscillatory shear stress, protects against antiatherosclerotic vascular effects.

Local factors as well as vascular geometry modulate blood flow patterns, velocity, and flow-mediated shear stress.15 Vascular changes caused by hypertension may affect how much flow-mediated shear stress occurs and, thereby, the production of nitric oxide. In arterial blood pressure, we recognize mean blood pressure as a steady component. Pulse pressure is the pulsatile component. We speculate that elevated pulse pressure reduces endothelium-dependent vasodilation by increasing oxidative stress and reducing nitric oxide production as a consequence of low shear stress. Previous experimental findings support these possible mechanisms, even if they are still speculative.12,14 On the other hand,

it was reported recently that in vivo arterial distensibility may be in-

fluenced by endogenous nitric oxide, whereby increased endothelium-derived nitric oxide reduces arterial stiffness.16 In view of these findings, it is possible that low ni-

tric oxide bioavailability increases pulse pressure, thus reducing nitric oxide production. This idea supports the concept that pulse pressure is both a cause and a consequence

of atherosclerosis.2


Increased pulse pressure represents a good or better predictor of cardiac and vascular hypertrophy, coronary heart disease, and subsequent cardiovascular events than other blood pressure components.1,2 Pulse pressure may be a useful clinical predictor that can be used to help stratify total cardiovascular risk in hypertensive patients. Although it is too early to consider pulse pressure reductions as a therapeutic goal, cardiovascular risk may be further reduced by narrowing the pulsatile component of blood pressure at any reduction in mean blood pressure in hypertensive patients. Calcium channel blocking agents, angiotensin-converting enzyme inhibitors, as well as low-dose diuretics have been shown to improve arterial distensibility.2

We found pulse pressure strongly and independently predicted endothelium-dependent acetylcholine-stimulated vasodilation in hypertensive patients. The magnitude and type of shear stress may play a role in mediating this relationship.