Estrogen improves endothelial function

Estrogen improves endothelial function S u b o d h Arora, M D , Aristidis Veves, MD, A. E n r i q u e Caballaro, M D , Paula Smakowski, MS, PT, and F r a n k W. L o G e r f o , M D , Boston, Mass.

Purpose: To determine the effect of estrogen on endothelium-dependent rdaxation in the cutaneous microcirculation of women. Methods: Three groups of women participated in the study. Group 1 (n = 20) was premenopausal and had a mean age of 39 years (range 24-50 years). Group 2 (n = 9) was postmenopausal and had a mean age of 58 years (range 53-65 years). Group 3 (n = 11) was postmenopausal and taking estrogen replacement therapy; the mean age was 53 years (range 43-58 years). Eleven women in group 1 underwent testing twice, once during menstruation (mean serum estradiol level 73 _+30 pg/ml) and once during midcycle (mean serum estradiol level 268 -+ 193 pg/ml; p = 0.003). Single-point laser Doppler ultrasound and laser Doppler imaging with a Scanner were used to measure vasodilatation in the forearm skin in response to iontophoresis of 1% acetylcholine (endothelium dependent) and 1% sodium nitroprusside (endothelium-independent smooth muscle relaxant). Results: All three groups were matched for body mass index and fasting glucose, total, high-density lipoprotein, and low-density lipoprotein cholesterol and triglyceride levels. All women had normal blood pressure, and none smoked. Mean serum estradiol levels were 196 ± 170 p g / m l (group 1), 35 ± 12 p g / m l (group 2), and 107 _* 78 p g / m l (group 3) (p = 0.004). Maximum microvascular vasodilatation (percentage increase over baseline) in response to acetylcholine was reduced in group 2 (93% ± 43%) compared with group 1 (187% _+ 63%) and group 3 (142% ± 56%) (p = 0.001). The response to sodium nitroprusside also was diminished in group 2 (73% ± 27%) compared with group 1 (126% _+45%) and group 3 (100% _+32%) (p = 0.02). Within group 1 the acetylcholine response was higher during the midcycle phase (186% ± 31%) compared with the menstrual phase (147% ± 57%) (p < 0.05). The sodium nitroprusside response also was higher during the midcycle phase (144% ± 31%) compared with the menstrual phase (94% ±

41%) (p < 0.05) Conclusion: The results indicate that estrogens might enhance endothelium-dependent and endothelium-independent vasodilatation in the microcirculation of women. (J Vasc Surg 1998;27:1141-7.)

The lower incidence o f cardiovascular events among premenopausal w o m e n and postmenopausal w o m e n taking estrogen replacement therapy has been ascribed to a cardioprotective effect o f estrogens. >4 Cardiovascular disease accounts for 23% o f deaths among w o m c n in thc United Statcs, and the incidence is increasing, especially a m o n g postFrom the Microcirculation Laboratory, Beth Israel Deaconess Medical Center, Harvard Medical School. Presented at the Twenty-fourth Annual Meeting of the New England Society for Vascular Surgery, Bolton Landing, N.Y., Sep. 18-I9, 1997. Reprint requests: Subodh Arora, MD, or Aristidis Veves, MD, Beth Israel Deaconess Medical Center (West Campus), 110 Francis St., Suite 5B, Boston, MA 02215. Copyright © 1998 by The Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/98/$5.00 + 0 24/6/89388

menopausal women.5 The mechanisms that underlie the cardioprotective effect o f estrogen are the subjcct o f m u c h interest. One o f the mechanisms appears to be a reduction in low-density lipoprotein cholesterol levels and an increase in high-density lipoprotein cholesterol levels, which may account for only 25% to 50% of the reduction in cardiovascular events. A direct effect o f estrogen on the vessel wall may account for the remainder o f the beneficial effcct on cardiovascular function. Endothelial dysfunction has been suggested as one o f the initial stages in the dcvelopment o f atherosclerosis and might be the result o f reduced release o f nitric oxide. 6 Studies have suggested that estrogen might enhance the activity o f endotheliumderived relaxing factor (nitric oxide) in the macrocirculation, thcrcby prcvcnting coronary vasoconstriction and thrombosis. In these studies duplcx 1141

JOURNAL OF VASCULAR SURGERY

1142 Arora et al.

ultrasound was used to examine the effect of estrogen on flow-mediated vasodilatation of the brachial artery.7, 8 Flow-mediated vasodilatation is known to be an endothelium-dependent phenomenon and can be measured during reactive hyperemia by means of high-resolution ultrasound scanning of superficial arteries. The effect of estrogen on the cutaneous microcirculation has not been evaluated selectively in the past. We hypothesized that endothelium-dependent vasodilatation in the cutaneous microcirculation is affected by serum estrogen (endogenous or exogenous) levels. To test this hypothesis we studied the effect of normal ovarian estrogen (endogenous) fluctuations during the menstrual cycle and the effect of estrogen replacement (exogenous) therapy on endothelial function in the cutaneous microcirculation after menopause.

METHODS Subjects. Three groups of female subjects were studied. Group 1 was premenopausal (n = 20, mean age 39 years, range 18 to 45 years). This group underwent testing at two points in the menstrual cycle, once during the menstrual period (low serum estrogen levels) and once midcycle (10 to 14 days after menses, high serum estrogen levels). Group 2 was postmenopausal (n = 9, mean age 58 years, range 53 to 65 years). Group 3 was postmenopausal and taking estrogen replacement therapy (n = 11, mean age 53 years, range 43 to 58 years). Subjects were recruited randomly by means of intrahospital advertisement. For comparisons between the three groups midcycle estrogen levels and the midcycle vasodilatory response for group 1 were used for analysis. All subjects were healthy and had no symptoms. The premenopausal group had regular menstrual cycles (26 to 30 days) and the postmenopausal women had been postmenopausal for at least 2 years. P0stmenopausal women taking estrogen replacement therapy had been taking the medication for at least 1 year. The exclusion criteria included current smoking, diabetes, hypertension, anemia (hematocrit 600 mg/dl, cholesterol level >250 mg/dl, and use of the following medications: lipid-lowering agents, glucocorticoids, bronchodilators, antineoplastic agents, antibiotics, and oral contraceptives if premenopausal. The study was approved by the institutional review board. All subjects signed a written informed-consent form after listening to an explanation of the study design and methods.

June 1998

All tests were performed on the same day in a warm environment (room temperature 25 to 26 ° C) under fasting conditions. Cutaneous perfusion was measured on the flexor aspect of the left forearm for all subjects at the same distance from the elbow, midway between the elbow and the wrist joint. There was no history of peripheral vascular disease, and all subjects had a palpable radial pulse. Women in group 3 were taking the following hormone replacement therapy: estrogen only Estraderm patch (n = 4), estrogen only Premarin (n = 2), estrogen plus progesterone Premarin/Provera (n = 3), and estrogen only Estrase (n = 1).

LASER DOPPLER IMAGING AND IONTOPHORESIS. The term "iontophoresis denotes the introduction of soluble ions into the human skin by means of application of electric current. It is a noninvasive technique that avoids any systemic effects of the drugs used. Endothelium-dependent vasodilatation was measured with application of acetylcholine chloride (acetylcholine induces release of nitric oxide from endothelium). Endothelium-independent vasodilatation was measured with sodium nitroprusside (sodium nitroprusside causes direct relaxation of smooth muscle cells). The MIC1 iontophoresis system (Moor Instruments, Millwey, Devon, England) was used in this study. One percent acetylcholine and 1% sodium nitroprusside in deionized water were delivered subdermally through an iontophoresis chamber with an anodal current of 200 microamperes for 30 seconds. Laser Doppler perfusion imaging measurements before and after iontophoresis of acetylcholine and sodium nitroprusside were made for all participants. A laser Doppler perfusion imager (Lisca PIM 1.0; Lisca Development, Linkoping, Sweden) was used to map skin perfusion. A 1 mW helium-neon laser beam of 633 nm wavelength sequentially scans an area of the forearm. The maximum number of measured points is 4096, and the apparatus produces a color-coded image of skin erythrocyte flux on a computer monitor. Each measurement lasts 20 to 30 seconds. The usual response to iontophoresis consists of transient erythema directly under the iontophoresis chamber. This technique has been shown by our group to have satisfactory reproducibility with a coefficient of variation less than 15%. 9 Blood specimens. At the end of the noninvasive test a 10 to 15 ml blood specimen was drawn from the same arm as the

JOURNAL OF VASCULAR SURGERY Volume 27, Number 6

Arora et aL

1143

Table I. Demographic and serum profiles of study subjects Characteristic No. o f patients Age (yr;range) Body mass index Systolic blood pressure ( m m H g ) Diastolic blood pressure ( m m H g ) Fasting glucose ( m g / d l ) Total cholesterol ( m g / d l ) High-density lipoprotein cholesterol ( m g / d l ) Low-density lipoprotein cholesterol ( m g / d l ) Triglycerides ( m g / d l ) Estradiol ( p g / m l )

Group 1

Group 2

Group 3

20

9

11

39 (24-50) 25 + 4.5 119 • 10 71 +_ 7 84 • 14 168 • 21 45 • 9 108 • 17 75 • 37

58 (53-65) 29 ± 6 121 • 6 75 • 5 90 • 25 201 • 36 44 • 11 136 • 26 103 ± 22

53 (43-58) 26 • 6 121 • 11 74 • 5 88 • 9 184 • 43 54 • I2 118 • 25 112 • 63

186 • 31

35 • 12

107 • 78

p Value

Group 1 v s groups 2 and 3, p < 0.05 NS NS NS NS 0.37 0.38 0.09 0.83 Groups 1 and 3 vs group 2, p < 0.05

Values are mean value + SD. NS, N o statistical difference.

test from each subject. High-density lipoprotein, low-density lipoprotein, and total cholesterol, triglyceride, fasting glucose, and estradiol levels were measured.

History and physical examination. All subjects had a medical history taken. Blood pressure, weight, height, and body mass index were obtained.

Statistical analysis. The statistical power of the study was 0.80 at an level of 0.05 capable of showing a 20% difference in microvascular changes among the various groups. For statistical analysis the Minitab statistical package was used. Analysis of variance was used for comparisons among the three groups, and Fisher test was used for pairwise differences between level means. All variables were treated as independent variables. The Kruskal Wallis test was used for nonparametrically distributed data. The default error was 0.05. Results were expressed as mean _+SD. RESULTS Demographic data, body mass index, lipid profiles, and fasting glucose and estrogen levels are shown in Table I. All three groups were matched for body mass index, blood pressure, normal lipid profile, and fasting glucose level. There was no significant difference in lipid profiles of postmenopausal women using transdermal estrogen and those taking oral estrogen. Because it was not possible to match the age of the premenstrual group with that of the postmenstrual group, only group 2 and group 3 were matched for age. Blood pressure measurements among group 1 were similar. Mean serum estradiol

levels were significantly different for the three groups (p = 0.004). Eleven premenopausal women underwent testing during both the menstrual phase and the midcycle phase. Menstrual phase estrogen levels (73 _+30 pg/ml) were significantly lower than midcycle levels (268 + 193 pg/ml) (p < 0.05). The cutaneous temperature at which the iontophoresis and laser Doppler imaging were performed was not statistically different among the three groups (Table II). Mean baseline skin perfusion before iontophoresis was similar for group 1 and group 3 (0.6 + 0.06 volts and 0.67 + 0.10 volts) but significantly lower for group 2 (0.55 + 0.04 volts). Maximum microvascular vasodilatation (percentage increase over baseline) in response to iontophoresis of acetylcholine was reduced in group 2 (93% _+43%) compared with group 1 (187% _+63%) and group 3 (142% _+ 56%) (p = 0.001). The response to sodium nitroprusside also was diminished in group 2 (73% _+27%) compared with group 1 (126% _+ 45%) and group 3 (100% +_ 32%) (p = 0.02) (Fig. 1). There was no significant difference in vasodilatory response to acetylcholine and sodium nitroprusside between women using transdermal estrogen and women taking oral estrogen. Within group 1, acetylcholine response was higher during the midcycle phase (I86% ± 31%) compared with the menstrual phase (147% _+57%) (p < 0.05). The sodium nitroprusside response was also higher during the midcycle phase (144% _+31%) compared with the menstrual phase (94% + 41%) (p < 0.05) (Fig. 2).

DISCUSSION The cardioprotective effects of estrogen have been fairly well documented.l-4,10-13 The effect of

IOURNAL OF VASCULAR SURGERY

1144 A r o r a et al.

lune 1998

Table II. Cutaneous perfusion before and after iontophoresis of acetylcholine and sodium nitroprusside Characteristic

Group 1

Group 2

Group 3

p Value

Sldn temperature (o C) Perfusion before Ach (V)

31 + 0.8 0.63 +_0.06

30.7 + 0.8 0.56 ± 0.06

30.7 +_ 0.7

0.67 ± 0.10

Perfusion before NaNP (V)

0.65 -+ 0.07

0.55 -+ 0.04

0.70 -+ 0.11

Perfusion after Ach (V) Perfusion after NaNP (V)

1.89 _+ 0.49

1.17 _+0.25

1.66 e 0.44

1.49 -+ 0.30

1.05 ± 0.26

1.44 _+0.32

187 _+ 63

93 -+ 43

142 ± 56

NS Groups 1 and 3 vs group 2, 20 < 0.05 Groups 1 and 3 vs group 2, p < 0.05 Groups 1 and 3 vs group 2, p < 0.05 Groups 1 and 3 vs group 2, p < 0.05 Groups 1 and 3 vs group 2, p < 0.05

126 _+45

73 _+27

100 + 32

Percentage increase over baseline after Ach iontophoresis Percentage increase over baseline after NaNP iontophoresis

Groups i and 3 vs group 2, p < 0.05

Values are mean value + SD. NS, No statistical difference; Ach, acetylcholine; NaNP, sodium nitroprusside

estrogen on the microcirculation of women has not been studied. We performed a quantitative study of the vasodilatory response of the cutaneous microcirculation with a noninvasive method that has been validated.14,15 We found that both the endotheliumdependent and endothelium-independent vasodilatory response in the microcirculation was significantly greater during the midcycle phase than the menstrual phase ofpremenopansal women. The two phases were confirmed with appropriate serum estradiol levels, which were significantly higher during the midcycle phase than the menstrual phase (p < 0.05). These findings suggested that endothelial vasodilatory function may be positively modulated with endogenous ovarian hormones, especially estrogen. Our results among postmenopausal women clearly demonstrated a significant difference in both endothelitun-dependent and -independent vasodilatory response between women using estrogen replacement and those not using replacement. Women using hormone replacement had significantly higher serum levels of estradiol (p < 0.05) and had a much greater response to both acetylcholine (p < 0.01) and sodium nitroprusside (p < 0.05). These findings suggest that estrogen positively affects both endothelium-dependent and endothelium-independent vasodilatation in the microcirculation. Because all our study subjects had no obvious associated cardiovascular risk factors, this improved vasodilatory response can be ascribed to the presence of estrogen. Localization of estrogen receptors on vascular endothelium and vascular smooth muscle cells of several mammalian species also suggests that these hor-

moncs may influence vascular function directly.16,17 Previous studies on the macrocirculafion have demonstrated that estrogen augments endotheliumdependent vasodilatation, as-20 Lieberman et al. 7 showed that estrogen improves endothelium-dependent, flow-mediated vasodilatation in postmenopausal women. Those authors used duplex ultrasonography to measure changes in brachial artery diameter in response to reactive hyperemia. All their study subjects had mild hypercholesterotemia, and thus the results cannot necessarily be extrapolated to postmenopausal women with normal lipid profiles. It is also possible that the effect of estrogen on endothelium-dependent vasodilatation occurred as a result of the lipid-lowering effect of estrogen and not a direct effect of estrogen on the vessel wall. Hashimoto et al.8 used the same metlmd and demonstrated that endothelium-dependent vasodilatation in the brachial artery varies during the menstrual cycle. It is greater during the follicular and luteal phases than during the menstrual phase, suggesting that endothelial vasodilatory function is positively modulated by endogenous ovarian hormones, especially estradiol. Short-term administration of estrogen has been shown to improve endothelium-dependent vasodilatation of atherosclerotic coronary arteries of monkeys after ovariectomy and of postmenopausal women.21,22 The effect of estrogen on endotheliumindependent vasodilatation, however, is controversial. Reports in the literature both support and refute the aforementioned mechanism.2S, 24 In this study we found that estrogen enhanced endothelium-independent vasodilatation (direct smooth mus-

JOURNALOF VASCULARSURGERY Volume27, Number 6

Arora et al.

%

250

• Ach [] NaNP

a, C

200

,',

150

I,.

> O

®

100

P u •-

50 0

I

Group 1

T

Group 2

Group 3

Fig. 1. Increase in blood flow (percentage of increase over baseline, mean and standard deviation bars) after iontophoresis of acetylcholine (black bars) and sodium nitroprusside (white bars) for premenopausal women (group 1), postmenopausal women not taking hormone replacement therapy (group 2), and postmenopausal women taldng hormone replacement therapy (group 3). The response of group 2 during both measurements was significantly lower than that of groups 1 and 3 (p < 0.05), and no difference was observed between groups 1 and 3.

%

•

200

._

l:a.j

¶

¶

150

100

50

0 P phase

M ~

Fig. 2. Increase in blood flow (percentage of increase over baseline, mean and standard deviation bars) after iontophoresis of acetylcholine (black bars) and sodimn nitroprusside (white bars) among premenopausal women during menstruation (M phase) and during midcycle (P phase), The response during menstruation was significantly reduced compared with midcycle measurements of both acetylcholine and sodium nitroprusside (p < 0.05).

1145

1146

JOURNAL OF VASCULARSURGERY June 1998

Arora et al.

clc effect) significantly. T h e exact m e c h a n i s m o f a c t i o n o f e s t r o g e n r e m a i n s unclear. E s t r o g e n m a y s t i m u l a t e e n d o t h e l i a l n i t r i c o x i d e synthesis, a n d because it has a n t i o x i d a n t propcrtics, e s t r o g e n may release or i n h i b i t d e g r a d a t i o n o f nitric oxide.25, 26 O t h e r possiblc m c c h a n i s m s m a y be s t i m u l a t i o n o f release o f prostacyclin, i n h i b i t i o n o f vasoconstrictor prostanoids, a n d alteration o f calcium flux in vascular s m o o t h m u s c l c . T h e l i p i d - l o w e r i n g cffcct o f e s t r o g e n certainly plays a role i n i m p r o v i n g vascular f u n c t i o n a m o n g p e r s o n s w i t h dyslipidemia.2-4,27 E s t r o g e n also is l m o w n to i n h i b i t vasoconstriction in response to e n d o t h e l i n 1.28 E v a l u a t i o n o f the complete m e c h a n i s m s o f action o f e s t r o g e n as a vasoprotcctive a n d cardioprotective a g e n t rcquires f u r t h c r i n v e s t i g a t i o n . Because e n d o t h e l i u m - d e p e n d e n t vasodilatation is k n o w n to be impaired i n the early stages o f diabetes, h y p e r t e n s i o n , peripheral vascular disease, a n d athcrosclerosis, e s t r o g e n m i g h t have a role i n p r e v e n t i o n o f vascular complications a n d progression o f these diseases. This study demonstrated that e n d o t h e l i u m d e p e n d e n t a n d e n d o t h e l i u m - i n d e p e n d e n t vasodilatation in the m i c r o c i r c u l a t i o n is a u g m c n t c d d u r i n g thc midcycle phasc o f p r c m e n o p a u s a l w o m e n a n d i n p o s t m e n o p a u s a l w o m e n u s i n g estrogen r e p l a c e m e n t therapy as c o m p a r e d with p o s t m e n o p a u s a l w o m e n n o t u s i n g r e p l a c e m e n t therapy. This may be consist e n t with an estrogenic effect o n vascular dynamics. REFERENCES

1. Lerner DJ, Kannel WB. Patterns of coronary heart disease morbidity and mortality in the sexes: a 26 year follow up of the Framingham population. Am Heart J 1986;111:383-90. 2. Ross RK, Paganini-Hill A, Mack TM, Henderson AE. Menopausal estrogen therapy and protection from death from ischemic heart disease Lancet 1981 ;2:858-60. 3. Stampfer MJ, Willet WC, Colditz GA, Rosncr B, Speizer FE, Hennekens CH. A prospective study of postmenopausal estrogen therapy and coronary artery disease. N Engl J Med 1985;313:1044-9. 4. Manolio TA, Furberg CD, Shemanski L, Psaty BM, O'Leary DH, Tracy RP, et al. for the CHS Collaborative Research Group. Associations of post menopausal estrogen use with cardiovascular disease and its risk factors in older women. Circulation 1.993;88:2163-71. 5. Centers for Disease Control. Years of life lost from cardiovascular disease. MMWR Morb Mortal WklyPep 1986;35:653-4. 6. Ross R. The pathogenesis ofatherosclerosis: a perspective for the 1990s. Nature 1993;362:801-9. 7. Lieberman EH, Gerhard MD, Uehata A, Walsh BW, Selwyn AP, Ganz P, et al. Estrogen improves endothelium dependent, flow mediated vasodilatation in post menopausal women. Ann Intern Mcd 1994; 121:936-41. 8. Hashimoto M, Akishita M, Eto M, Ishikawa M, Kozaki K, Toba K, et al. Modulation of endothelinm dependent flow mediated dilation of the brachial artery by sex and menstrual cycle. Circulation 1995;92:3431-5. 9. VevesA, Saouaf R, Donaghue VM, Mullooly CA, Kistler JA,

10.

11.

12.

13.

14.

15. 16. 17.

18.

19.

20.

21.

22.

23.

24.

25. 26. 27.

28.

Giurini JM, et al. Aerobic exercise capacity remains normal despite impaired endothelial function in the micro and macrocirculation of physically active IDDM patients. Diabetes 1997;46:1846-52. Bush TL, Cowan LD, Barrett-Connor E, Criqui MH, Karon JM, Wallace RB, et al. Estrogen use and all-cause mortality. lAMA 1983;249:903-6. Criqui MH, Suarez L, Barrett-Connor, McPhillips J, Wingard DL, Garland C. Post menopausal estrogen use and mortality. Am J Epidemiol 1988;128:606-14. Stampfer JM, Colditz CA, Willett WC, Manson JE, Rosner B, Speizer FE, et al. Postmenopausal estrogen therapy and cardiovascular disease. N Engl J Med 1991;325:786-62. Barretr-Connor E, Wingard DL, Criqni MH. Postanenopausal estrogen use and heart disease risk factors in the 1980s. JAMA 1989;261:2095 ~100. Morris SJ, Shore AC, Tooke JE. Responses of the skin microcirculation to acetylcholine and sodium nitroprusside in patients with NIDDM. Diabetologia 1995;38:1337-44. Tooke JE. Methodologies used in the study of the microcirculation in diabetes mellitus. Diabetes Metab Rev 1993;9:57-70. Colbnrn P, BuonassisiV. Estrogen binding sites in endothelial cell cultures, Science 1978;210:817 9. Horwitz KB, Horwitz LD. Canine vasculartissues are targets for androgens, estrogens, progestins, and glucocorticoids. J Clin Invest 1982;69:750-8. Williams JK, Adams MR, Klopfenstein HS. Estrogen modulates responses to of atherosclerotic coronary arteries. Circulation 1990;81:1680-7. Gisclard V, Miller VM, Vanhoutte PM. Effect of 17 beta estradiol on endothelium dependent responses in the rabbit. J Pharmacol Exp Ther 1988;244:19-22. Miller VM, Gisclard V, Vanhoutte PM. Modulation of endothelium dependent and vascular smooth muscle responses by estrogens. Phlebology 1988;3:63-9. Williams JK, Adams MR, Herrington DM, Clarkson TB. Short term administration of estrogen and vascular responses of atherosclerotic coronary arteries. J Am Coil Cardiol 1992;20:452-7. Reis SE, Cloth ST, Blumenthal RS, Resar JR, Zacur HA, Gerstenblith G, et al. Ethynil estradiol acutely attenuates abnormal coronary vasomotor responses to acetylcholine in postmenopansal women. Circulation 1994;89:52-60. Jiang C, Sarrel PM, Lindsay DC, Poole-Wilson PA, Collins P. Endothelium independent relaxation of rabbit coronary artery by 17 beta estradiol in vitro. Br J Pharmacol 1991;104:1033-7. Gilligan DM, Badar DM, Panza JA, Quyyumi AA, Cannon RO III. Acute vascular effects of estrogen in postmenopausal women. Circulation 1994;90:786-91. Estradiol 17 beta increases the number of muscarinic receptors in hypothalmic nuclei. Brain Res 1980;198:239-43. Yagi I(, Komura S. Inhibitory effect of female hormones on lipid peroxidation. Biochem Int 1986;13:1051-5. Keany IF Jr, Shwaery GT, Xu A, Nicolosi RJ, Loscalzo J, Foxall TL, et al. 17 beta estradiol preserves endothelial vasodilator function and limits low density lipoprotein oxidation in hypercholesterolemic swine. Circulation 1994; 89:2251-9. Jiang C, Sarrel PM, Poole Wilson PA, Collins P. Acute effects of 17 beta estradiol on rabbit coronary artery contractile responses to endothelin 1. Am J Physiol 1992;263:H271-5.

Submitted Sep. 22, 1997; accepted Jan. 29, 1998.

JOURNAL OF VASCULARSURGERY Volume27, Number6

DISCUSSION Dr. Magruder C. Donaldson (Boston, Mass.). I congratulate the authors on an excellent paper. Others have observed that people who have a relatively high estrogen level are protected from atherosclerosis. The question is whether we should consider the idea of a protocol that puts men on estrogen at a certain point in their lifespan. Dr. S u b o d h Arora. You make an interesting point. It would be interesting to study men with prostate cancer who are on estrogen and see how they compare with agematched controls. The problem with studying men with prostate cancer on estrogen therapy is that most of them are started on fairly high doses. The complication rates with high doses can be high, but men on low-dose estrogen would certainly be a good group to compare with age-matched controls. Dr. Jeffrey L. K a u f m a n (Springfield, Mass.). I am impressed at the subtraction image that you showed. I think there is a positive effect, but I always have concern when I see laser Doppler data as to the issue of electrical noise and the noise generated in the sampling technology in terms of finding a real difference. I hope that the manuscript cxplains this phenomenon and the absolute differences rather than percentage comparisons in evaluating the data. Some papers have not emphasized mat there can be much electrical noise with these instruments. Are these differences between the groups reflective of something beyond the noise factor? Dr. Arora. The answer to your question is yes. We actually have looked at the biological zero in every subject studied. This is done by first scanning the forearm skin with a blood pressure cuff on the arm and occluding the blood flow to the forearm completely to see how much background noise is picked up by the scanner. Then, the test is done with the cuff deflated. Dr. Jack L. C r o n e n w e t t (Lebanon, N.H.). Dr. Arora, I thought your study was an elegant study; it is great to see this ldnd of research being performed in the human situation. I am having a little trouble understanding your conclusion in terms of whether this is a problem at the endothelial cell level or the smooth muscle cell level. Your data showed the same effect with nitroprusside and with acetylcholine, and so it appears that this may be a primary defect in the smooth muscle cell response. I wonder if that is your conclusion. That is, have you been able to separate whether this is an endothelial cell difference or really just a primary smooth muscle cell issue?

A r o r a et al.

1147

Dr. Arora. From our data, the response to acctylcholine or the endothelial response was much greater than the smooth muscle response, but both elements seem to be definitely involved. As you know, there are estrogen receptors both in the endothelial cells and smooth muscle cells. In fact, some studies have shown that there might be more receptors in the smooth muscle cells than in the endothelial ceils, but that is still debatable. We think both the smooth muscle cells and the endothelial cells are actually responsible for the vasodilatation. How much each contributes is still unknown. Dr. Martha D. McDaniel (White River Junction, Vt.). In trying to put this in a clinical context, I wonder how you would explain the phenomenon you observed in as much as we believe that migraine headaches are vasodilatory in nature and seem to be more prevalent during the menstrual period than during midcycle. Dr. Arora. We have thought about performing this study on women who have severe premenopausal and premenstrual symptoms, and I think that would be a very interesting study. It is hard to explain, with what we have found in normal healthy women, why some women get severe migraine during the menstrual period. You are right that it is vasodilatation that is thought to be the cause of these symptoms. This might be our next project. Dr. A n t h o n y W h i t t e m o r e (Boston, Mass.). In your reading, did you find any evidence that will allow you to correlate your findings with, perhaps, the incidence of coronary vasospasm? Dr. Arora. Data in the literature show a similar effect in the coronaries. Similar studies have been done, but these studies are all at the macrocirculation level where people have infused acetylcholine and estrogen into the coronary arteries and found profound vasodilatation. So, it certainly acts on the coronaries with profound vasodilatation. Dr. William W. Babson, Jr. (Plymouth, Mass.). I wondered if you tried this on testosterone. Does testosterone have a similar effect to estrogen? Dr. Arora. We have not tried it with testosterone, but testosterone apparently has the opposite effect as estrogen. There are data to suggest that testosterone might actually cause vasoconstriction and not vasodilatation, but these are anecdotal reports. There are no large studies in the literature that look at the effect of testosterone on endothelial function.