ERs have been demonstrated in canine peripheral and coronary arteries, in cultured rat aortic smooth muscle cells, in baboon myocardium and aorta, and in human endothelial cells. Vasodilatation in response to estrogen administration was first noted in the rabbit ear artery and later in primate coronary arteries, human umbilical and human coronary arteries. It appears that the effect of estrogen on the cardiovascular system (CVS) is complex, acting via at least eight different mechanisms that are outlined in Table Effects of estrogen on the cardiovascular system.
Table Effects of estrogen on the cardiovascular system
- Beneficial effect on blood lipids and lipoproteins
- Positive effect on endothelial cell function
- Relaxation of the vascular smooth muscle
- Decreased platelet adhesion and aggregation
- Effects on coagulation and fibrinolysis
- Effects on inflammatory markers
- Promotion and maintenance of gynecoid body fat distribution
- Improved glucose and insulin metabolism
Progesterone receptors are present in myocardium but the effect of progesterone on the CVS is less well studied. The influence of estrogen on the CVS will be discussed in further detail below.
The response of vessels to estrogen is largely one of vasodilatation. Under the influence of estrogen, cardiac output increases, systemic vascular resistance decreases and blood pressure remains unchanged or falls slightly. In healthy pregnant women estrogen causes increased angiotensinogen levels and renin activity, and these lead to increased blood volume by a direct effect on volume through enhanced sodium retention. In the cerebrovascular circulation, there are gender differences. Before the menopause, cerebrovascular blood flow is greater in women than in men of the same age. Cerebrovascular blood flow is increased during pregnancy, when estrogen levels are also increased, but decreases after the menopause. Estrogen affects the reactivity of cerebral arteries to vasoactive stimuli such as serotonin.
As yet, the mechanism of estrogen’s effects on the vasculature has not been fully elucidated, but several explanations have been proposed, involving either immediate or delayed effects. Immediate effects are likely to result from nitric oxide (NO) production and prostaglandin formation from the endothelium and effects on calcium channels in the smooth muscle cells of the vessel wall. Longer-lasting effects may involve reducing angiotensin-converting enzyme activity, inhibition of smooth muscle cell proliferation and increasing the smooth muscle cell prostaglandin production by increasing prostacyclin synthetase and cyclo-oxygenase.
The normal effect of endogenous estrogen in premenopausal women, as described above, is found in postmenopausal women given physiologic doses of estrogen. Angiographic and ultrasound studies have shown enhancement in the coronary, brachial and internal iliac artery blood flows and in the cerebrovascular circulation in response to estrogen. There are reports that estrogen decreases the size of atherosclerotic plaques in the common carotid artery as measured by ultrasound scan (USS). The results of angiographic studies of coronary arteries affected by atherosclerosis are not consistent: after treatment with estrogen some studies show either no progression or reversal of the disease, while other fail to demonstrate any effect.
Lipids and lipoprotein metabolism
Premenopausal women have an overall more favorable lipid profile than men, with lower low-density lipoprotein (LDL) until the sixth decade, lower very low-density lipoprotein (VLDL) throughout life, and higher highdensity lipoprotein (HDL) during postpubertal life. The differences are due to increased apolipoprotein B100 receptor activity in the cell membrane leading to lower LDL, more efficient clearance of VLDL, an increased rate of synthesis and a reduced clearance of HDL. Hepatic lipase activity is higher in men than in women, hence HDL is lower in men. Estrogen deficiency following the menopause is associated with adverse changes in blood lipids and lipoproteins. Many studies examining these changes have shown that there is a significant increase in total cholesterol, LDL cholesterol and triglycerides and a decrease in HDL cholesterol after adjustment for age, body mass index and smoking, though not all studies are consistent. The important HDL2 subtraction was found to be reduced as a direct result of the menopause and there is an age-related increase of lipoprotein (a) (Lp (a)) after 50.
There are over 50 randomized studies on the effect of estrogen or estrogen/ progestogen on lipid and lipoprotein metabolism in healthy women and in women with risk factors for (or with clinically established) coronary artery disease. Some of the studies have compared hormone replacement therapy with a placebo and some with statins (simvastatin). The studies have consistently demonstrated the ability of estrogen replacement therapy (ERT) to improve the cholesterol profile significantly. Typical percentage changes in lipids are given in Table Typical changes in blood lipids (%) following oral hormone replacement therapy. There is little difference between oral conjugated equine estrogens (CEE) and oral estradiol. Progestogens have to be added on a cyclical or continuous basis in order to protect against endometrial cancer. Progestogens have an effect on lipids that, like estrogens, depends on their chemical structure, dose, regiman and route of administration. Well designed control trials in menopausal women show that progesterone and dydrogesterone do not oppose the favorable changes in lipid metabolism induced by estrogen. The antagonism of norethisterone is minimal and the effect of levonorgestrel seems most pronounced. On this basis, it seems preferable to use hormone replacement therapy combinations containing dydrogesterone or norethisterone. The case of medroxyprogesterone is peculiar. In clinical controlled trials where the endpoint was changes in lipid metabolism, it showed minimal antagonizing effect on the entrogen-induced lipid changes. In trials where the end point was cardiovascular morbidity and mortality, the group treated with CEE and medroxyprogesterone fared significantly worse than those on placebo, so preparations with medroxyprogesterone are best avoided. We believe that hormone replacement therapy prescribing should be individualized according to the clinical condition and the nature of the dyslipidemia, to maximize the benefits and improve compliance. For example, transdermal estradiol/norethisterone may be preferable in diabetics, and continuous combined hormone replacement therapy is best avoided in patients with established ischemic heart disease.
Table Typical changes in blood lipids (%) following oral hormone replacement therapy
|Total cholesterol||↓ 5-10||↓ 3-8|
|LDL||↓ 15-20||↓ 10-15|
|HDL||↑ 10-15||↑ 3-10|
|Triglycerides||↑ 20-30||↓ 10-15|
LDL, low-density lipoprotein; HDL, high-density lipoprotein
Body fat distribution
Estrogen promotes and maintains the deposition of adipose tissue in the classic gynecoid pattern of postpubertal females. The ideal waist-to-hip ratio is 0.8 or less, while a ratio over 1 is abnormal. A waist circumference for women of < 89 cm and for men of < 102 cm is considered healthy. Based on cohort studies some researchers suggest lower number (< 80 cm for women and < 94 cm for men). Waist-to-hip ratio measurements and dual-energy X-ray absorptiometry (DXA) measurements show that there is a postmenopausal shift in fat deposition, with a significant increase in abdominal (android) and intra-abdominal fat distribution. This may result from the decline in the estrogemandrogen ratio after the menopause. This increase in waist-to-hip ratio over time has been shown in women to be associated with a significant increase in cardiovascular morbidity and mortality. Reversal of these unfavorable changes has been observed in randomized controlled trials of hormone replacement therapy.
Glucose and insulin metabolism
Estrogen deficiency leads to decreased insulin output from the pancreas, decreased insulin elimination and an increase in relative insulin resistance. While hyperglycemia itself may cause injury to the vascular endothelium, it is insulin resistance with accompanying hyperinsulinemia that may be a pivotal metabolic disturbance in the pathogenesis of CHD. Fasting blood glucose and glycated hemoglobin do not change after menopause. However, hormone replacement therapy decreases insulin resistance and circulating insulin levels and therefore may be beneficial.
It has been shown that increased levels of Factor VII and fibrinogen are risk factors for cardiovascular disease. The association of those two factors with cardiovascular death appears to be at least as strong as the association between cholesterol and cardiovascular death. The menopause itself leads to a 6% increase in the level of Factor VII and a 10% increase in fibrinogen. Orally administered 17p-estradiol decreases the level of fibrinogen. Estrogen enhances fibrinolysis in postmenopausal women by reducing plasminogen activator inhibitor-1 (PAI-1) by 50%. Little is known about the effect of the menopause on platelet function, but it appears that exogenous estrogen, or hormone replacement therapy, decreases platelet aggregation adenosine triphosphate release. Transdermal hormone replacement therapy does not appear to lead to any detrimental changes in the coagulation profile. The effects of the menopause and hormone replacement therapy on coagulation and fibrinolysis are summarized in Table Menopause, hormone replacement therapy and coagulation and fibrinolysis.
Table Menopause, hormone replacement therapy and coagulation and fibrinolysis
|Factor||Menopause||Hormone Replacement Therapy|
|Factor VII||Increase 7%||No change|
|Factor VIII/von Willebrand||Increase||Increase|
|Platelets||No change||Decrease aggregation|
PAI-1, plasminogen activator inhibitor-1; t-PA, tissue plasminogen activator
The complex relationship between estrogen and coagulation is not fully understood, and the studies of coagulation are frustrated by:
- (1) The use of different laboratory methods;
- (2) The effect of tourniquet on sampling;
- (3) Disagreement between researchers on the relative importance of coagulation variables;
- (4) Difficulties in dissociating the effects of aging from the menopause;
- (5) The participation of some coagulation factors in the ‘acute phase response’;
- (6) The continuous stream of new discoveries, such as Factor V Leiden, prothrombin gene mutation, etc.;
- (7) The effects of progestogens; and
- (8) Differences betwen synthetic and natural estrogens, route of administration and doses.
Markers of inflamation
The current understanding is that the process of atherosclerosis is to a large extend an inflammatory process. Some blood markers of inflammation are C-reactive protein (CRP), interleukins 1 and 6, E-selection, serum amyloid protein A and adhesion molecules. These markers correlate well with the severity of CHD, the transition of stable into unstable angina and the rate of myocardial infarction (MI) and stroke. The best studied marker is CRP. The reasons for that are as follows: (1) the development of highly sensitive assays for this protein allowing the stratification of subjects within the normal range; (2) the fact that CRP levels are stable over long periods of time with the exception of the time during an acute infection (approximately 2 weeks); (3) there is no diurnal variation. CRP was found to be a strong independent predictive factor for cardiovascular events in healthy women, in women at risk of CVD and in women with established CVD. In studies, the relative risk of events for women in the highest as compared to the lowest quartile of CRP was 2.3. It is now accepted that the predictive values of CRP blood levels for cardivascular events is at least as good as that of low-density liproprotein. CRP is synthesized in the liver in response to inflammation. The signal for synthesis is usually mediated via inflammatory cytokines. The level of CRP in apparently healthy subjects correlates with the body mass index (BMI). The effect of age and menopause on CRP is not known. However, the effect of estrogen or hormone replacement therapy is well studied. In all studies so far 0.625 mg of CEE and 2.5 mg of MPA per day invariably lead to a substantial (80-85%) increase in CRP levels. Oral CEE increase the levels by 48-65% and transdermal estrogen by 3-10%. It is not clear if hormone replacement therapy stimulates hepatic synthesis of CRP or systemic inflammation or both. In any case the effect of hormone replacement therapy on inflammation may help explain the increase in cardiovascular events observed in hormone replacement therapy users during the first few years of treatment.
Cardiovascular system: Summary
CVD (cardiovascular disease) is the main cause of death in the UK: more than one in three people (40%) die from it. The main forms of CVD are CHD (coronary heart disease) and stroke. About half of all deaths are caused by CHD and about a quarter by stroke.
A 50-year-old woman has a 31% lifetime probability of developing CHD and a 17% lifetime probability of dying from coronary heart disease. This is the leading cause of death in postmenopausal women. The prevalence of CHD increases progressively with age and this may be partly due to estrogen deficiency. This assumption is biologically plausible and supported by experimental and epidemiologic evidence. A large meta-analysis of observational epidemiologic data showed a 35% reduction in the incidence of CHD and a 37% reduction in mortality from coronary heart disease in users of ERT. Moreover, this protective effect seemed greater in women who already had CHD. This wealth of encouraging data prompted a number of randomized controlled trials. The designs, endpoints and regimens were varied: some were primary prevention studies, others were secondary. Some used surrogate endpoints such as changes in coronary artery diameter, others incidence, morbidity and mortality of coronary heart disease. Oral or transdermal estradiol or CEE were used as the estrogen component either alone or as sequential or continuous combined hormone replacement therapy with MPA or NETA. In all but one study ERT or hormone replacement therapy were no better than placebo. In the one trial with positive results, new cases of CHD in the hormone replacement therapy group were observed. What are we to make of this divergent results between observational and controlled studies? In randomized controlled trials subjects are randomly assigned to treatment and thus possible biases are minimized. In observational trials subjects who choose to take the treatment may be very different from those that do not. This fundamental methodologic difference helps to explain the results at least partially. One possible bias is the so-called ‘healthy-user effect’. This means that observational trials fail to control fully for lifestyle and other health-related factors that may differ in hormone users and not-users. In observational studies subjects who choose to take ERT or hormone replacement therapy may be generally healthier and/or have healthier lifestyles than non-users and this imbalance may lead to an overestimation of the effect of the treatment and an underestimation of its risks. Observational studies are also susceptible to ‘compliance bias’. It is known that subjects who are compliant with their treatment tend to have improved outcome even if the tretment is placebo. It is possible that hormone taking is simply a marker for better compliance with other lifestyle advice and/or treatments of cardiovascular disease. It is also possible that cohort studies do not capture fully early clinical events. Let us imagine that women are enrolled in a cohort study, information is collected at baseline and update questionnaires are sent to them every 2 years. The subject is not on hormone replacement therapy and returns the questionnaire saying so. If she starts taking hormone replacement therapy shortly after that, say in 2-3 months’ time, and proceeds to have an MI within the next 12-18 months for the purpose of the study she would be misclassified as a non-user having had an MI thus underestimating the hazard of the treatment. As the Women’s Health Initiative (WHI) and Heart and Estrogen/ Progestin Replacement Therapy Follow-up (HERS) trials show, the risk for MI early in the treatment is higher than overall. There are also biological explanations that may account for some of the discrepancy in the data between randomized controlled and observational studies. One explanation is that different regimens may have different effects. In the nurses’ Health Study mose hormone users were on estrogen and those that were on estrogen/progestin were taking progestins for 10-14 days. The estrogen arm of the WHI trial continues. The implication is that estrogen may still be beneficial and that further research into different regimens and preparations is necessary. Another biological explanation is that estrogen may need to be taken early after the menopause in order to prevent the development of atherosclerosis, rather than after the establishment of atherosclerotic plaques when estrogen may raise the levels of inflammatory markers, especially CRP, and destabilize the plaques leading to clinical events. This possibility is supported by randomized controlled trials on monkeys. Whatever the explanations are for the current data on hormone replacement therapy and CHD we have to accept the supremacy of experimentation over observation. On the basis of this, hormone replacement therapy regimens cannot be prescribed for primary or secondary prevention of CHD in the populations specified in site. However, randomized controlled trials did not show any benefit in women with established heart disease. Postmenopausal estrogen use does not affect the blood pressure and is safe to be administered to hypertensive women, when the hypertension is under control.