Estrogen Enhances Basal Nitric Oxide Release in the Forearm Vasculature in Perimenopausal Women

Estrogens are secreted in a cyclical fashion in women between the menarche and menopause. During the perimenopausal period, the production of these hormones declines to very low levels, reflecting a decrease in the size of the primary follicle pool in the ovary.¹ The reduced estrogen levels in postmenopausal years are believed to be in large part responsible for the increased predisposition in later years to coronary artery disease.² Although not yet unequivocally shown in large-scale randomized trials, many observational studies of chronic estrogen replacement therapy in postmenopausal women show a significant reduction in the risk of coronary events, lending support to the hypothesis that estrogen may exert cardioprotective effects.³ However, the precise mechanism underlying its benefits are unclear. Estrogens appear to have beneficial effects on lipoprotein profiles³ ; however, Bush et al⁴ have estimated that estrogen-induced changes in lipoprotein levels account for only 25% to 50% of the observed risk reduction, suggesting that additional factors are involved.

It is being increasingly recognized that some steroid hormones, including estrogens, exert direct effects on the vasculature. Steroid receptors have been identified in the aorta, left atrial appendage, and coronary and internal carotid arteries.⁵ 17β-Estradiol has been reported to induce relaxation of precontracted rabbit coronary rings in vitro⁶ and coronary vasodilation in vivo in dogs.⁷ Physiological levels of estrogen have been shown to enhance acetylcholine-induced vasorelaxation in the forearm⁸ and coronary⁹ vasculatures in postmenopausal women. Short intravenous¹⁰ and intracoronary¹¹ infusions of estrogen in women attenuate acetylcholine-induced vasoconstriction in vivo in atherosclerotic coronary arteries. In primate models of atherosclerosis, more long-term estrogen supplementation has been shown to preserve endothelial function,¹² a finding that may be relevant to the cardioprotective effects of estrogen in humans. Estrogen supplementation has also been shown to augment endothelium-dependent flow-mediated vasodilation in the brachial artery in hypercholesterolemic postmenopausal women.¹³ However, the mechanisms by which estrogen replacement modulates endothelial function are not well defined.

In the present study, using local infusion of an inhibitor of nitric oxide (NO) synthase in the forearm vasculature, we sought to determine whether 8 weeks of estrogen supplementation in perimenopausal women influences basal NO release. In addition, we examined the effect of estrogen supplementation on blood pressure (BP), serum lipoproteins, and endothelium-dependent and endothelium-independent vasorelaxation in the forearm vasculature.

Methods

Eleven perimenopausal women were recruited through advertisement in a local newspaper. All had definite menstrual irregularities, were within 2 years of their last period, and were actively experiencing vasomotor symptoms of menopause. Subject characteristics at baseline in both groups are shown in Table 1. Women with one or more cardiovascular risk factors, thyroid disease or diabetes mellitus, or clinical evidence of vascular disease were excluded. Individuals on vasoactive medications, such as angiotensin-converting enzyme inhibitors or calcium channel blockers, were also excluded. The study was approved by the Alfred Hospital Ethics Committee, and all subjects gave full written informed consent.

Subjects were randomized to receive 8 weeks of either estrogen supplementation as estradiol valerate (Progynova, Schering; 2 mg daily, n=6) or placebo (n=5). Hemodynamic studies and assessment of forearm vascular reactivity were performed on two separate occasions, 8 weeks apart. Subjects were unaware of the treatment they were receiving, and all measurements of hemodynamics and vascular reactivity were made by an investigator who was blinded to the treatment regimen. On each study day, subjects underwent the following procedures.

Hemodynamic Measurements

Subjects rested in the supine position throughout each study in a quiet, temperature-controlled room maintained at 22°C. After subjects had rested 20 minutes, baseline supine systolic and diastolic BPs were measured with an automated sphygmomanometer (Dinamap, Critikon, Johnson and Johnson Medical). The brachial artery of the left arm was then cannulated with a 21-gauge, 5-cm catheter (Cook) under strict aseptic conditions after local anesthesia (1% lignocaine, Astra) for intra-arterial BP measurement (Spacelabs Inc) and drug infusion. Heart rate was continuously monitored by electrocardiography. After brachial cannulation, subjects rested for 30 minutes before the study was begun.

Assessment of Forearm Vascular Reactivity

Forearm vascular responsiveness to vasoactive agents was assessed by venous-occlusion plethysmography with a sealed alloy-filled (gallium and indium), double-stranded strain gauge (Medasonic). Hand blood flow was excluded via a wrist cuff inflated to 200 mm Hg, and venous-occlusion pressure on the arm was 50 mm Hg. Before each drug dose, basal blood flow was obtained from an average of at least three measurements. Drugs were infused at the rate of 2 mL/min via an infusion pump.

Endothelium-dependent vasodilation was assessed by intra-arterial infusion of acetylcholine (BDH Chemicals) at sequential doses of 9.25, 18.5, and 37 μg/min. Each dose was infused for 2 minutes. Basal NO release was assessed by intra-arterial infusion of N^G-monomethyl-l-arginine (L-NMMA, Calbiochem-Novabiochem) at sequential doses of 1, 2, and 4 μmol/min. Each dose was infused for at least 5 minutes. Finally, endothelium-independent vasodilation was assessed by intra-arterial infusion of sodium nitroprusside (David Bull Laboratories) at 1.6 μg/min for 2 minutes. The peak response was determined as the average of three consecutive steady-state measurements. A 15-minute rest period between interventions was sufficient for flow to return to resting levels. Intra-arterial brachial mean BP was recorded both during measurement of basal flows and immediately after each intervention.

Measurement of Hormone and Lipid Levels

In all subjects, venous blood was sampled on both study days for measurement of estradiol, total and high-density lipoprotein cholesterols, triglycerides, and glucose.

Calculations and Statistical Analysis

Results are expressed as mean±SE. Vascular reactivity data are expressed as both absolute values and the percent change from basal forearm blood flow measured before each administration. Vascular reactivity dose-response curves before and after estrogen or placebo were compared by two-way repeated measures ANOVA. Other data were compared by Student's t test for paired observations. Significance was defined at a value of P<.05.

Results

Effect of Estradiol on BP

In subjects receiving estradiol supplementation, both systolic and diastolic BPs dropped significantly (P<.01). Systolic and diastolic BPs did not change significantly in subjects who received placebo (Table 2).

Effect of Estradiol on Serum and Plasma Measurements

Estradiol levels increased in subjects receiving estradiol (P<.001) and did not change in subjects on placebo. In subjects who received estrogen, cholesterol had a tendency to decrease, although this change did not reach statistical significance (P=.09). High-density lipoprotein cholesterol, serum triglycerides, and plasma glucose in subjects receiving estrogen also did not change significantly. In subjects who received placebo, total cholesterol, high-density lipoprotein cholesterol, triglycerides, and glucose did not change significantly (Table 2).

Effect of Estradiol on Acetylcholine-Induced Vasorelaxation

Acetylcholine induced a dose-dependent increase in forearm blood flow. The acetylcholine dose-response relationship after supplementation with either estrogen or placebo did not change significantly (Fig 1). BP and heart rate were unchanged during acetylcholine infusions, at baseline, and after both estrogen and placebo.

Effect of Estradiol on L-NMMA–Induced Vasoconstriction

L-NMMA induced a dose-dependent decrease in forearm blood flow. After estrogen supplementation, the degree of vasoconstriction induced by L-NMMA was considerably enhanced (F=9.21, P=.04 from two-way ANOVA, before versus after estrogen), suggesting an increase in basal NO release. The L-NMMA dose-response relationship after administration of placebo did not change significantly (Fig 2). BP and heart rate were unchanged during L-NMMA infusions, at baseline, and after estrogen or placebo.

Effect of Estradiol on Nitroprusside-Induced Vasorelaxation

The increase in forearm blood flow induced by nitroprusside remained unchanged after administration of either estradiol (before estradiol: 372±39%; after estradiol: 362±41%) or placebo (before placebo: 380±51%; after placebo: 373±48%). BP and heart rate were unchanged during nitroprusside infusions, at baseline, and after estrogen or placebo.

Discussion

This randomized, double-blind, placebo-controlled study demonstrates that 8 weeks of estrogen supplementation enhances basal NO release in the vasculature of perimenopausal women. This was associated with a significant reduction in systolic and diastolic BPs in estrogen-treated subjects. Endothelium-dependent relaxation in response to acetylcholine and endothelium-independent relaxation in response to nitroprusside were unchanged by estrogen supplementation.

Previous studies in postmenopausal women have shown that acetylcholine-induced vasomotion is modulated by short-term administration of estrogen.⁸⁹¹⁰¹¹ Our finding of an unchanged response to acetylcholine differs from those studies. However, there were several differences between our subjects and those previously studied. In the previous studies, estrogen was administered as infusions into the brachial artery,⁸ the coronary artery,⁹¹¹ or a peripheral vein,¹⁰ unlike the present study in which estradiol was administered orally. Furthermore, compared with subjects in previous studies, the women in the present study were perimenopausal, younger by about a decade, had experienced a shorter period of estrogen deficiency than the postmenopausal subjects in previous studies, and had higher baseline levels of circulating estrogen than those in some previous studies.⁸⁹¹⁰¹¹ In addition, in several previous studies, the women examined had coronary vascular disease and/or multiple risk factors,⁹¹⁰¹¹ which are likely to have resulted in impaired endothelium-dependent vasorelaxation at baseline.¹⁴ By contrast, our subjects were largely free of major cardiovascular risk factors and had no evidence of vascular disease. These factors may have resulted in greater baseline acetylcholine-induced vasorelaxation than observed in previous studies and thus the lack of an effect after estrogen supplementation. As with previous studies, the current study showed no change in nitroprusside-induced endothelium-independent vasodilation after estrogen supplementation.

The major new finding in the present study is the enhancement in forearm vasoconstriction induced by L-NMMA after estrogen supplementation. This observation is unlikely to be due to a nonspecific increase in the response to vasoconstrictor agents: In a preliminary study, we recently demonstrated that after estrogen supplementation, vasoconstrictor responsiveness to norepinephrine in the forearm vasculature was significantly attenuated.¹⁵ Thus, the most likely explanation for our observation of an accentuated vasoconstrictor response to L-NMMA is that basal NO synthesis and release were enhanced after estrogen supplementation. Our results are consistent with studies performed in vitro in rat aortic rings by Hayashi and colleagues,¹⁶ who observed that acetylcholine-induced vasorelaxation was not different in male and female rats but that basal NO release was higher in female rats. They also showed that this difference was related to ovarian sex hormones, since oophorectomy abolished the difference. Studies examining the effect of NO synthase inhibition on estrogen-induced vasodilation show conflicting results in different vascular beds: Nitro-l-arginine methyl ester attenuated uterine arterial vasodilation¹⁷ but did not influence coronary vasodilation.⁷ Preliminary studies of the effects of estrogen on NO synthesis in cultured bovine endothelial cells have also shown conflicting results, with one study showing an increase in the expression of constitutive NO synthase¹⁸ and another showing no change.¹⁹ Roselli et al²⁰ recently showed that serum nitrite and nitrate levels are higher in postmenopausal women on hormone replacement therapy, providing indirect evidence of enhanced NO production. Our study supports the hypothesis that estrogen induces an increase in basal NO release from the vasculature. Recent studies have suggested that NO may retard a variety of atherogenic processes, including smooth muscle proliferation,²¹ monocyte adhesion,²² and platelet aggregation.²³ Estrogen-induced enhancement of NO release may thus be an important underlying mechanism in the cardioprotective effect of estrogen.

The present study demonstrated a striking effect of estrogen on arterial BP that might be related to increased basal NO release. Such an effect is in accord with epidemiological data showing a lower diastolic BP in premenopausal women than in their postmenopausal counterparts.²⁴ On the other hand, studies reporting effects of postmenopausal estrogen therapy on BP have been conflicting, with some showing a rise in BP in subjects on conjugated estrogens,²⁵²⁶ others reporting lower BP in women on estrogen replacement therapy,²⁷²⁸ and still others, including the recently concluded Postmenopausal Estrogen/Progestin Interventions (PEPI) trial, showing no effect.²⁹³⁰³¹

Endothelium-dependent relaxation is impaired in individuals with hypercholesterolemia,³² and lipid-lowering therapy has been shown to restore endothelial function.³³ In the present study, there was a trend toward a decrease in total cholesterol with estrogen supplementation, which did not reach statistical significance. However, given the evidence that estrogen favorably influences lipid profile,³ a contribution to the observed increase in vascular release of NO by the lipid-lowering effect of estrogen cannot be excluded.

In the current study, forearm vascular responses to both acetylcholine and L-NMMA were very similar before and after placebo, suggesting that in our laboratory, venous-occlusion plethysmography yields reproducible results within subjects over time. The reproducibility of casual forearm blood flow measurements by plethysmography has been previously examined. Roberts et al³⁴ found a between-day mean coefficient of variation of 12%, and Bassett et al³⁵ have reported that the average difference between two within-day measurements was 8.8%. In the present study, baseline blood flows before and after intervention within either the estrogen-treated or placebo-treated groups did not differ significantly. However, overall, there was a tendency toward a lower baseline forearm blood flow in the placebo-treated group than in the estrogen-treated group. It is unlikely that such differences in baseline flow were the primary reason for the differences in L-NMMA response obtained in the two groups after treatment, because the increased response to L-NMMA in the estrogen group did not occur before estrogen supplementation, despite a tendency toward higher baseline flows. Furthermore, the L-NMMA responses after treatment, expressed as percent changes, were also significantly different: It has been previously suggested that in studies using forearm plethysmography, if differences in response to an infused drug are due to baseline differences, percent changes should be similar.³⁶³⁷

In conclusion, the present observations in postmenopausal women have demonstrated that basal but not acetylcholine-induced NO release is enhanced after 8 weeks of estrogen supplementation. The relationship of this observation to the apparently substantial risk reduction induced by estrogen supplementation in epidemiological studies requires further examination.

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Footnotes