From the Department of Medicine, Division of Cardiology, Tufts-New England Medical Center, Tufts University School of Medicine, Boston, Massachusetts
Endothelial cells perform crucial functions within the vascular system. Abnormal vasoactivity largely results from decreased nitric oxide production by the endothelium.1 Endothelial dysfunction can be detected in coronary and peripheral arteries before the development of atherosclerotic plaque,2 and its presence has been shown to predict future cardiac events.3 High-resolution ultrasonography of the brachial artery during reactive hyperemia is a noninvasive method of assessing endothelium-dependent vasomotion. Abnormalities in peripheral vasomotor function correlate with the presence of coronary arterial endothelial dysfunction.4 In some studies, maximal ability of the artery to vasodilate via an endothelium-independent pathway also appears to correlate with the presence of coronary artery disease (CAD).5 Abnormalities in low-density lipoprotein (LDL) cholesterol, along with other risk factors for atherosclerosis, correlate with impaired vasorelaxation of coronary and peripheral arteries.
The beneficial effects of high-density lipoprotein (HDL) cholesterol on the vasculature are numerous. HDL cholesterol inhibits oxidation of LDL cholesterol, up-regulates prostacyclin formation, and exerts direct anti-inflammatory effects.6 HDL cholesterol also has direct beneficial effects on the vascular endothelium, including rapid activation of endothelial nitric oxide synthase (eNOS)7 and reversal of oxidized LDL-mediated inhibition of eNOS activity.8 Based on these observations, we sought to determine whether HDL levels correlate with peripheral endothelial function in a population of adults with CAD.
Patients and methods
The study population consisted of consecutive adult subjects under-going nuclear stress testing for evaluation of CAD. Subjects with a recent myocardial infarction or unstable angina, congestive heart failure, or valvular heart disease were excluded. The presence or absence of cardiovascular risk factors was assessed in each subject. All subjects underwent clinically indicated, symptom-limited nuclear stress testing with intravenous technetium-99m sestamibi using a standard Bruce exercise protocol. Single-
photon emission computed tomography (SPECT) scans were analyzed in
a blind fashion and considered positive for CAD if perfusion abnormalities were found and negative for CAD if no perfusion abnormalities were noted.
Brachial artery ultrasonography testing was performed to evaluate peripheral endothelial function as previously described by us and others.9 Longitudinal brachial artery images of subjects were obtained with a high-resolution (10 MHz)
linear array vascular transducer during the morning hours after
at least a 6-hour fast. After an equilibrium period of 10-minutes, base-
line 2-dimensional images of the brachial artery were obtained 2 cm above the antecubital fossa. A blood pressure cuff placed on the upper arm was inflated to suprasystolic pressure for 5 minutes. After cuff release, the vessel was imaged continuously, and peak hyperemia brachial artery diameter was obtained after 60 seconds. After a return to the baseline brachial artery diameter, peak nitroglycerin brachial artery diameter was determined
5 minutes after administration of sublingual nitroglycerin.
Percent flow-mediated dilation was defined as the difference between the maximal brachial artery diameter during reactive hyperemia and the baseline brachial artery diameter divided by the baseline brachial ar-
tery diameter. Percent nitroglycerin-mediated dilation was defined as the difference between the maximal brachial artery diameter after nitroglycerin administration and the baseline brachial artery diameter divided by the baseline brachial artery diameter. Fasting serum samples were drawn for analysis.
One hundred fifty-one subjects (87 men, 64 women) were enrolled in the study. The average age (± SD) was 58 ± 11 years. Forty-three percent had hypertension, and 25% had diabetes mellitus. Approximately one fourth of the subjects were taking aspirin, one third were taking angiotensin-converting enzyme inhibitors, and nearly one half were receiving statin therapy. The average values for lipid levels were: total cholesterol, 188 ± 48 mg/dL; HDL cholesterol, 47 ± 13 mg/dL; LDL cholesterol, 108 ± 37 mg/dL; triglycerides, 154 ± 88 mg/dL; and non-HDL cholesterol, 141 ± 43 mg/dL. CAD was diagnosed in 66 subjects by SPECT imaging.
There was a significant correlation between flow-mediated dilation and the HDL level (r = 0.3; P < .001). The average flow-mediated dilation for the group was 9.9 ± 5.2%, and the mean baseline brachial artery diameter was 3.8 ± 0.7 mm. Subjects with an HDL level below 40 mg/dL had lower flow-mediated dilation (n = 39; 7.4 ± 3.6%) compared with those with an HDL level 40 mg/dL or above (n = 112; 11.0 ± 5.5%; P < .001). The baseline brachial artery dimension was also different between the two groups (4.2 ± 0.7 mm for those with HDL cholesterol < 40 mg/dL compared with 3.7 ± 0.7 mm for those with HDL cholesterol ≥ 40 mg/dL; P < .001). The absolute difference between the brachial artery diameter during hyperemia and baseline in subjects with HDL cholesterol below 40 mg/dL was significantly less than for those with an HDL cholesterol level 40 mg/dL or above (P = .014). As expected, subjects with CAD (n = 66) had lower HDL cholesterol (40 ± 9 mg/dL) and flow-mediated dilation (7.4 ± 4.2%), whereas subjects without CAD
(n = 85) had higher HDL cholesterol (53 ± 13 mg/dL) and flow-mediated dilation (11.7 ± 5.2%; P < .001). Multivariate analysis showed that HDL cholesterol was an independent
predictor of flow-mediated dilation (P = .015). Endothelium-independent vasodilation was decreased in subjects with HDL cholesterol below 40 mg/dL (16.0 ± 6.3%) compared with those with HDL cholesterol 40 mg/dL or above (19.9 ± 9.6%; P = .02).
This study shows that HDL cholesterol levels directly correlate with peripheral endothelial function in a population of patients undergoing evaluation for atherosclerosis. Interestingly, other lipid parameters, including total cholesterol and LDL cholesterol along with hyperten-sion and diabetes mellitus, did not correlate independently with flow-mediated dilation in this popula-tion. On average, brachial artery vasomotion during reactive hyperemia was blunted in subjects with low HDL cholesterol levels and well preserved in those with HDL cholesterol levels 40 mg/dL or above.
In addition, the relationship between flow-mediated dilation and HDL cholesterol was stronger in men compared with women in this study, although this may be because of the small sample size of women.
Atherosclerosis results in remodeling of the arterial wall and may lead to arterial expansion.10 Thus, arterial size alone may be an indicator of atherosclerosis. In our study, baseline brachial artery size was larger in subjects with HDL cholesterol below 40 mg/dL compared with those who had higher levels of this lipoprotein. Absolute changes in vessel diameter (reactive hyperemia minus baseline diameter) correlated with HDL levels. Endothelial-independent vasomotion was also decreased in subjects with low HDL levels compared with those with high levels.
The data presented in this study are consistent with previous studies relating HDL cholesterol to endothelial function. Kuhn and colleagues studied subjects undergoing elective angiography and performed intracoronary acetylcholine injections to evaluate vasomotor responsiveness.11 In their study, subjects with a normal vasodilatory response to acetylcholine had higher levels of HDL cholesterol compared with those with paradoxical vasoconstriction. In a study by Toikka and colleagues, a similar correlation between HDL cholesterol levels and peripheral vasomotor function was noted. Low HDL levels in young healthy subjects were related to brachial artery endothelial dysfunction and increased oxidative stress.12 HDL cholesterol has also been shown to correlate with peripheral endothelial function in hyperlipidemic subjects.13 In another analysis, reconstituted HDL cholesterol infused into patients with hyperlipidemia resulted in a transient increase in HDL levels and improvement in acetylcholine-induced brachial artery blood flow.14 Infusion of an eNOS inhibitor in this population prevented the HDL-induced improvement in blood flow, indicating that a likely interaction between HDL cholesterol and eNOS was responsible for the vessel dilation. Finally, we have recently shown that brachial artery flow-mediated dilation improves after administration of HDL-raising therapy,9 thus supporting the importance not only of basal levels of HDL cholesterol, but also of approaches to raise these levels.
HDL cholesterol levels independently predicted peripheral vasomotor function in a population
undergoing evaluation for atherosclerosis. These findings may indicate a significant effect of HDL levels on endothelial function and help to unmask previously unrecognized mechanisms for the beneficial effects of HDL cholesterol. HDL levels may provide mechanistic insight into vasomotor dysfunction and perhaps offer a therapeutic target to improve cardiovascular risk. Further studies focusing on the effects of HDL-raising therapies on endothelial function are warranted.