| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Center for Laboratory Medicine and Laboratory of Atherosclerosis Genetics (H.J., R.R., A.S., T.L.) and Departments of Diagnostic Radiology (P.D.) and Gynecology and Obstetrics (K.T., R.P.), Tampere University Hospital; and Department of Clinical Chemistry, University of Tampere Medical School (R.R., T.L.), FIN-33521 Tampere, Finland
Address all correspondence and requests for reprints to: Hannu Jokela, Ph.D., Department of Clinical Chemistry, Tampere University Hospital, Center for Laboratory Medicine, P.O. Box 2000, FIN-33521 Tampere, Finland. E-mail: hannu.jokela{at}tays.fi.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
. The study group comprised 15 women receiving estradiol valerate HRT (mean age, 56 yr; treatment duration, 10.5 yr) and 15 women receiving combined HRT with estradiol valerate and levonorgestrel (mean age, 58 yr; treatment duration, 11.3 yr). The peak flow velocity (PFV) and pulsatility index of the common carotid and internal carotid artery and the abdominal aorta were measured by ultrasonography after long-term HRT (baseline), after a 4-wk pause and again 3 wk after reintroducing HRT. A statistically significant interaction between the groups and time points was observed in the PFV of the internal carotid artery (P = 0.011). In women taking estradiol valerate, the PFV values decreased significantly after withdrawal of HRT (P = 0.007) and increased again to the baseline level after reintroduction of therapy (P < 0.001). In women receiving combined HRT, the PFV remained stable over all study periods. At baseline, the PFV of women taking estradiol valerate correlated with the plasma nitrate concentration in the common carotid artery (r = 0.646; P = 0.009) and in the abdominal aorta (r = 0.579; P = 0.024). For pulsatility index and urinary 8-iso-prostaglandin F2 excretion, there were no significant differences between the groups. Our results suggest that the favorable effects of long-term estrogen treatment on blood flow are at least partly mediated through NO. The addition of levonorgestrel to the treatment regimen appears to abolish this effect. |
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Although several observational studies have previously indicated that estrogen is beneficial to cardiovascular health (16, 17), randomized secondary prevention trials have shown that conjugated equine estrogen plus medroxyprogesterone acetate (MPA) is of no proven benefit in reducing coronary events or the incidence of cerebrovascular events (18, 19, 20). Such discrepant results from observational and randomized studies suggest that the beneficial effects of estrogen are counterbalanced by harmful effects. There would also appear to be a complex relationship between estrogens and progestins with regard to vascular reactivity. Measurements of endothelial function in response to combination therapy with different progestins and estrogen in healthy women have produced varying results (21, 22, 23, 24, 25), indicating that the benefits of estrogen may be preserved or negated depending on which progestin is used simultaneously. Concurrent progestin therapy may thus also oppose the beneficial effects of estrogen on the endothelium (26).
We here studied the effect of long-term HRT with estradiol valerate and with a combination of estradiol valerate and levonorgestrel on the hemodynamics of the carotid arteries and the abdominal aorta and on plasma NO, the predominant vasodilatory factor released from the endothelium. The aim was to establish whether these two therapies differ in their ability to affect vasomotor function. A further aim was to find out whether HRT prevents the formation of the in vivo lipid peroxidation marker 8-iso-prostaglandin F2 (8-iso-PGF2) in the urine (27), which could be related to the antioxidative effects of HRT.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The study population consisted of 30 healthy, normotensive, nonsmoking postmenopausal women. Of this cohort, 15 were hysterectomized, and their HRT consisted of 2 mg/d estradiol valerate. Their mean age was 56.4 ± 7.0 yr, and mean duration of treatment was 10.5 ± 3.2 yr. For the other 15 women, the hormone regimen was 2 mg/d estradiol valerate for 11 d, followed by 2 mg/d estradiol valerate and 0.25 mg/d levonorgestrel for 10 d. These women were 58.0 ± 4.1 yr old, and the duration of their treatment was 11.3 ± 3.6 yr. The mean body mass index was 24.6 ± 2.5 kg/m2 in women receiving estradiol valerate therapy and 26.2 ± 3.5 kg/m2 in women receiving combined HRT. There were no women with diagnosed diabetes or cardiovascular diseases. The study was approved by the ethical committee of Tampere University Hospital, and subjects gave written informed consent.
Biochemical measurements
Venous blood samples were drawn after a 12-h overnight fast using a normal blood-sampling technique and from women receiving combination therapy during the third week of the treatment cycle. Serum total cholesterol and triglycerides were measured by the dry slide technique (Ektachem 700 analyzer, Johnson & Johnson Clinical Diagnostics, Rochester, NY). The high density lipoprotein (HDL) concentration was determined after precipitation of LDL and very low density lipoprotein with dextran sulfate/magnesium chloride (28) using the same technique. The LDL cholesterol concentration was calculated by Friedewald’s formula (29). Apolipoproteins (apo) AI and B were analyzed using an immunonephelometric method (Behring, Behringwerke AG, Marburg, Germany) and lipoprotein(a) [Lp(a)] by two-site immunoradiometric assay (Amersham Pharmacia Biotech, Uppsala, Sweden).
Plasma and urinary nitrate (NO3), which are used as indexes of NO formation, were determined by capillary electrophoresis (30) using the HP 3DCapillary Electrophoresis System (Hewlett-Packard, Palo Alto, CA). For the measurement of 8-iso-PGF2, the total volume of 24-h urine was mixed, and an aliquot was stored frozen at -70 C until analyzed. Thawed urine samples were centrifuged at 10,000 x g for 10 min, and after appropriate dilution, the supernatant was used for the determination of 8-iso-PGF2 by a competitive ELISA according to the manufacturer’s instructions (R&D Systems, Inc., Minneapolis, MN). Urinary 8-iso-PGF2 was expressed as the total amount excreted in 24 h. The 8-iso-PGF2/creatinine ratio was used alternatively in the analysis. Urinary creatinine was measured with the creatinine test of the Ektachem 700 analyzer (Johnson & Johnson Clinical Diagnostics).
Ultrasound examinations
Ultrasound examinations of arteries were performed using the color Doppler transducer ultrasound imaging system Acuson 128 XP (Acuson Corp., Mountain View, CA), with a 7.5-MHz phased array pulsed Doppler transducer. The internal carotid and common carotid arteries were examined on right and left sides, and the mean value of these measurements was used in the analysis. The examination of the internal carotid arteries included the first centimeter of the carotid bifurcation. The abdominal aorta was examined at three levels, i.e. pancreas, infrapancreatic level, and immediately above the bifurcation. The mean value of these measurements was again used in the analysis. The pulsatility index (PI), which is thought to represent impedance to blood flow downstream from the point of measurement (31), and the peak flow velocity (PFV), expressed in centimeters per second and defined as the midpoint of the upper portion of the Doppler signal at the time of maximal velocity (32), were used as indexes of hemodynamics and vascular tone. These measurements were made at baseline (time point 1) and were repeated 4 wk after the treatment was discontinued (time point 2) and again 3 wk after reintroducing HRT (time point 3). The investigator was unaware of the HRT status of the women. All recordings were performed by the same certified sonographer and radiologist (P.D.). This guaranteed the highest validity and reproducibility.
Statistics
Means of continuous variables between treatment groups were compared using one-way ANOVA. Statistical analyses of longitudinal data were carried out using ANOVA for repeated measures (RANOVA) to establish any treatment period by treatment group interaction. In the case of a significant interaction, the least significant differences post hoc test was used after RANOVA to ascertain the differences in the treatment groups between two time points. Nonnormally distributed data were analyzed after logarithmical transformation. Pearson’s correlation coefficient was used to evaluate correlations. Values are expressed as the mean ± SD unless otherwise stated. P < 0.05 was considered statistically significant. Statistica for Windows (version 5.1 software package, Statsoft, Inc., Tulsa, OK) was used for statistical analyses.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
shows the baseline characteristics of the subjects after 10 yr of HRT. For plasma lipids and apo, subjects who had used only estradiol valerate had higher concentrations of HDL cholesterol (P < 0.001) and its major apolipoprotein, apo AI (P < 0.001). The LDL/HDL ratio was lower in the estradiol group compared with the combined therapy group (P = 0.004). Other lipid values did not differ significantly between the two treatment groups at baseline, although concentrations of triglycerides and Lp(a) tended to be lower in subjects receiving combination therapy than in those taking estrogen.
|
Changes in PFV values in carotid arteries and aorta according to treatment periods and groups are shown in Table 2. In the internal carotid artery, there was a statistically significant interaction between the treatment periods and groups (P = 0.011). Post hoc analysis showed the result to be due to the group of women receiving estradiol valerate therapy. These PFV decreased significantly when the therapy was discontinued for 4 wk (P = 0.007) and increased again after reintroduction (P < 0.001; Fig 1). In the combination therapy group, the PFV values remained stable over all study periods. In this group, PFV was also lower at baseline after 10 yr of therapy (P = 0.002) as well as after resumption of therapy (P < 0.001) than in the estradiol valerate group (Fig 1). The difference between treatment groups in PFV values in the internal carotid artery remained statistically significant when baseline HDL cholesterol concentration (P = 0.038), LDL/HDL ratio (P = 0.004), and Lp(a) level (P = 0.042) were used as covariates. The difference was of borderline significance when the baseline triglyceride level was used as a covariate (P = 0.064). When the apo AI concentration at baseline was used as a covariate, the difference between groups was nonsignificant (P = 0.155). In common carotid artery and abdominal aorta, there were no statistically significant differences in PFV between treatment groups or periods.
|
|
As shown in Table 1
, NO3 in plasma was higher in women taking estradiol valerate after 10 yr of HRT compared with women receiving combined therapy (P = 0.043). Urinary NO3, in turn, was similar in the two groups. Table 3 shows the correlations between PFV and plasma NO3 according to treatment groups in different arteries. At baseline, PFV correlated statistically significantly with plasma NO3 in all subjects and in all three arteries. In the common carotid artery and in the abdominal aorta, this correlation was emphasized in subjects receiving estradiol valerate therapy, and in the common carotid artery it persisted during the treatment break as well as after reintroduction of therapy. There were no significant correlations between PI and NO3 in plasma.
|
, 24-h 8-iso-PGF2 excretion was not statistically significantly different at baseline in women receiving estrogen therapy compared with women receiving combination therapy, although the excretion was slightly lower after long-term estrogen use. Similarly, after the 24-h excretion was adjusted with creatinine, the groups did not differ in this respect. Figure 2 shows the 24-h 8-iso-PGF2 excretion in the two HRT groups at baseline, after the 4-wk pause in HRT, and 3 wk after reintroducing HRT.
|
excretion and plasma NO3 in any time point in either therapy. However, the 24-h 8-iso-PGF2 correlated statistically significantly with urinary NO3 at all time points when all women regardless of group were included in the analysis. In women receiving estrogen therapy, the correlation was statistically significant when the therapy was discontinued (r = 0.905; P < 0.001) and after reintroduction of therapy after the break (r = 0.756; P = 0.002). In women receiving combination therapy, the 8-iso-PGF2 excretion correlated with NO3 in the urine at time point 1 (r = 0.845; P < 0.001) and at time point 3 (r = 0.812; P < 0.001). In all subjects, urinary and plasma NO3 values correlated positively after long-term HRT (r = 0.700; P < 0.001), when therapy was discontinued (r = 0.320; P = 0.085), and after reintroducing HRT (r = 0.383; P = 0.037). This correlation was most clearly seen at baseline after long-term HRT both in women receiving estrogen therapy (r = 0.849; P < 0.001) and in women using combination therapy (r = 0.582; P = 0.023).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
tended to be lower in women receiving estrogen HRT compared with women receiving combined HRT, however, not to a statistically significant extent. 8-Iso-PGF2 excretion correlated with the amount of NO3 in urine, but not with the amount in plasma. Over the past few decades HRT has been widely used by postmenopausal women in western countries. The findings in observational studies have been important in promoting the belief that HRT prevents cardiovascular disease. Previous epidemiological studies of estrogen therapy (most using conjugated equine estrogens) have suggested that estrogen use reduces the overall relative risk of coronary artery disease by approximately 50% (17). However, the majority of those women who have used hormones have been taking combined HRT (33). Contrary to observational studies focusing on estrogen therapy, secondary prevention trials with conjugated equine estrogen plus MPA have provided no evidence that combined HTR can reduce coronary events, the progression of angiographically defined coronary atherosclerosis, or the incidence of cerebrovascular events (18, 19, 20). This suggests that concurrent progesterone therapy may oppose the beneficial effects of estrogen on cardiovascular health. Therefore, evidence that HRT is of benefit in either primary or secondary prevention of major cardiovascular events is currently lacking.
It has been reported that estrogen reduces LDL cholesterol and increases HDL cholesterol concentrations (34). However, preparations with androgenic properties or combined therapies containing androgenic progestogens may have no effect on,or may even reduce HDL cholesterol. This was clearly seen in the present study, where estrogen valerate therapy, but not combination therapy, reduced the LDL to HDL ratio. In the combination therapy group, the HDL cholesterol concentration was lower than in the estradiol valerate therapy group and increased statistically significantly when the therapy was discontinued. Despite the negative effects on the HDL cholesterol level, combination therapy may have beneficial effects on some aspects of the lipoprotein profile. Levels of Lp(a), which is similar in composition to LDL and is considered particularly atherogenic, are markedly reduced by progestogens, whereas estrogen has a minimal effect (35, 36). This was also seen in our subjects receiving combination therapy, as Lp(a) and also triglycerides tended to be lower than in the subjects receiving estrogen valerate therapy.
Short-term estrogen replacement therapy has been shown to reduce peripheral resistance and improve blood flow in the carotid arteries (37), whereas combination therapy with different progestins and estrogen have produced varying results (21, 22, 23, 24, 25). Our study shows that a constant dose of estrogen affects arterial reactivity after 10 yr of treatment. This is likely to contribute to the protective cardiovascular effects of estrogens. The positive correlation seen here between PFV and plasma nitrate suggests that the effects of long-term estrogen treatment on blood flow are at least partly mediated through NO, a powerful vasodilator released by vascular endothelial cells. Concomitant administration of progestogen might, however, attenuate this effect on endothelial function (38). This was indeed seen in our study, where addition of levonorgestrel to the treatment regimen appeared to abolish the beneficial effect of estradiol valerate on PFV. In the combination therapy group, the correlation between PFV and plasma NO3 was also lost. Our findings thus support the evidence that combined HRT might not necessarily be cardioprotective. We observed no significant changes in PI, possibly due to the short discontinuance of the therapy. Effects of estradiol take some weeks to manifest themselves and they carry over weeks after treatment is withdrawn (37).
The susceptibility of LDL to oxidative modifications is held to be an independent risk factor for atherosclerosis (39). Reports of an antioxidant effect of estrogen have, however, remained contradictory. Short-term studies have reported that treatment with transdermal estradiol significantly increased (40), moderately increased (41), or had no effect on (42) the lag time to copper-mediated oxidation of LDL in vitro. In addition, one short-term study reported that oral estrogen reduced the susceptibility of LDL to oxidation (43), which conflicts with the result from two studies that found no antioxidant effect of estrogen given orally in both unopposed and combined preparations (42, 44). The only study to focus on the effect of long-term HRT on oxidation of lipids showed that there were equal amounts of antibodies against oxidized LDL in subjects receiving estradiol valerate alone, combined estradiol valerate and levonorgestrel, and combined estradiol valerate and MPA (45).
Here we measured the excretion of 8-iso-PGF2 in urine. 8-Iso-PGF2 is the most abundant member of the F2-isoprostane family of prostaglandin isomers derived from arachidonic acid and undergoes peroxidation by a mechanism catalyzed by free radicals to yield arachidonyl radical intermediates and, finally, prostaglandin F2-like compounds (27). The levels of F2-isoprostanes increase dramatically in smokers (46), experimental animals subject to oxidant injury (47), and patients with acute myocardial infarction (48). This suggests that urinary 8-iso-PGF2 could be a useful noninvasive in vivo index of free radical generation. In the present study the levels of 8-iso-PGF2 in urine were moderately decreased in subjects receiving estradiol valerate therapy compared with those receiving combined therapy. Although the trend was not statistically significant, it suggests that oxidative stress might be reduced with long-term estrogen use. Concurrent levonorgestrel may thus abolish at least part of this effect.
The in vivo activity of NO must be monitored indirectly, because unstable NO rapidly oxidizes to nitrite and predominantly NO3. Measurement of plasma NO3 reflects the level of systemic NO production. The NO3 in urine, in turn, is a more complex parameter, also reflecting the level of renal function and plasma volume (49). It has been observed in experimental animals that urinary NO3 production is enhanced in hypercholesterolemia and, in particular, when there is already evidence of atherosclerotic plaques (49). As NO and superoxide radicals can combine to be excreted as nitrates (50), increased urinary NO3 excretion could reflect an increased production of these entities. At the same time, cytotoxic hydroxyl radicals may be formed (51). Another reason for enhanced excretion of NO3 could be increased NO synthesis by inducible NO synthase (iNOS). In humans, iNOS expression is correlated with atherosclerotic lesion size (52), and activation of iNOS is probably the major source of vascular wall bioactive NO in atherosclerosis (53). It may thus be hypothesized that in increased oxidative stress, perhaps simultaneously with existing atherosclerotic plaques, both the formation of NO3 and the excretion of the oxidative stress marker 8-iso-PGF2 are increased.
The limitation of the present study was a rather small number of subjects. Women were not randomized to hysterectomy and thus were not randomly divided into estrogen and combined HRT groups. This can serve as a source of selection bias. This bias is, however, very difficult to avoid, because hysterectomy is almost always a clinical decision that cannot be randomized. Also, it is well known fact that cardiovascular risk factors differ in hysterectomized and nonhysterectomized women. The results of the present study cannot thus be generalized to all women receiving HRT.
Our findings support the evidence that estrogen valerate therapy, but not combined therapy, reduces the LDL to HDL ratio. In addition to the beneficial effects on the plasma lipid profile, long-term estrogen treatment may improve blood flow; this is at least in part due to the increased availability of NO, a powerful vasodilator released from the endothelium. Cardiovascular advantages of estrogen may also include decreased oxidative stress, as in the present study excretion of the in vivo oxidative marker 8-iso-PGF2 tended to be decreased in estrogen users. Addition of levonorgestrel to the treatment regimen appears to abolish these beneficial effects.
|
Footnotes |
|---|
Abbreviations: apo, Apolipoprotein; HDL, high density lipoprotein; HRT, hormone replacement therapy; iNOS, inducible nitric oxide synthase; 8-iso-PGF2, 8-iso-prostaglandin F2; LDL, low density lipoprotein; Lp(a), lipoprotein(a); MPA, medroxyprogesterone acetate; NO, nitric oxide; NO3, nitrate; PFV, peak flow velocity; PI, pulsatility index; RANOVA, ANOVA for repeated measures.
Received January 6, 2003.
Accepted June 10, 2003.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
in vivo is not affected by cyclooxygenase inhibition. Adv Prostaglandin Tromboxane Leukotriene Res 23:233–236[Medline]
This article has been cited by other articles:
 |
|
 |
|
  B. C. Jacobson, B. Moy, G. A. Colditz, and C. S. Fuchs Postmenopausal Hormone Use and Symptoms of Gastroesophageal Reflux Arch Intern Med, September 8, 2008; 168(16): 1798 - 1804. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. N. Torgrimson, J. R. Meendering, P. F. Kaplan, and C. T. Minson Endothelial function across an oral contraceptive cycle in women using levonorgestrel and ethinyl estradiol Am J Physiol Heart Circ Physiol, June 1, 2007; 292(6): H2874 - H2880. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Rogers and D. D. Sheriff Role of estrogen in nitric oxide- and prostaglandin-dependent modulation of vascular conductance during treadmill locomotion in rats J Appl Physiol, August 1, 2004; 97(2): 756 - 763. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
), 4 wk after cessation of treatment (
), and 3 wk after its reintroduction (
). Interaction between treatment periods and groups was determined by RANOVA. By the least significant differences post hoc test in estradiol valerate group: *, P = 0.007 (between time points 1 and 2); **, P < 0.001 (between time points 2 and 3). Values are the mean ± SD.