It is now well established that cardiovascular disease represents the major cause of death in women just as in men, but at a later age. In women the incidence of cardiovascular disease increases after the menopause and it has been shown that the risk of atherosclerosis increases 3 — 4-fold after natural menopause.

Among the main recognized cardiovascular risk factors for both men and women, cigarette smoking, high cholesterol levels, hypertension, diabetes mellitus and obesity are preventable causes of coronary heart disease. The fall in oestrogens following the menopause may affect several of these risk factors and it is now well established that oestrogen replacement therapy will improve cholesterol levels, diastolic blood pressure, insulin sensitivity and some of the clotting factors. The beneficial effects of oestrogens on the vasodilating endothelial factors, hence on vasomotor tone,suggest that oestrogens may play a major role in the primary prevention of coronary heart disease in women. Better understanding of the mechanisms underlying the protective effect of sex steroids has been provided by animal studies using the monkey modelas well as studies of postmenopausal women, using surrogate markers of cardiovascular risk.

There is now evidence that oestrogens will improve endothelial function and hence vascular tone,however, the effect of progestogens on cardiovascular disease risk is controversial. Progestins are prescribed together with oestrogens to protect the endometrium, and some of the most commonly prescribed progestins have been shown to oppose partially the beneficial effect of oestrogens on cardiovascular disease risk markers. However, several categories of progestins may be prescribed, and there are striking differences according to the type of molecules used.

However, evidence-based medicine does not rely on surrogate markers of risk and the relationship between hormone replacement therapy and heart disease can be demonstrated only by randomized controlled trials with cardiovascular events as main outcomes.

Epidemiological data

Results from a large number of observational studies over the last two decades have strongly suggested that postmenopausal oestrogen replacement therapy protects against cardiovascular disease. However, this evidence is not based on randomized trials.

A number of epidemiological studies performed since the early 1980s indicate around a 50% reduction in cardiovascular morbidity and mortality in women using oestrogens after the menopause. However, a selection bias in the studies may account for this observation, women who take oestrogens in the long term are generally healthier than those who do not.

The selection bias hypothesis has been reinforced by the recent results from the Heart and Estrogen/Progestin Replacement Study (HERS) research group, the first ever published randomized controlled trials of Hormone replacement therapy for secondary prevention of cardiovascular disease. This study suggested that in postmenopausal women with established coronary heart disease an Hormone replacement therapy regimen with conjugated oestrogens and medroxyprogesterone acetate does not protect from further cardiovascular disease.

In contrast, in most observational studies of oestrogen replacement therapy and coronary heart disease, including studies conducted in women with established coronary disease, lower cardiovascular morbidity and mortality were shown in users of oestrogen replacement therapy. Observational studies of Hormone replacement therapy incorporating oestrogen and progestogen for primary prevention of coronary heart disease are fewer than studies of oestrogen replacement therapy, but some show a reduction in the risk of coronary heart disease comparable to the reduction in the risk of coronary heart disease in users of Hormone replacement therapy. The results from the observational data showing a lower risk of coronary heart disease in hormone users caused considerable debate about the ethics of randomization of women to hormone therapy, in which some subjects would receive a placebo.

Table Key studies of combined Hormone replacement therapy and ischaemic heart disease in postmenopausal women.

Type of study Treatment Relative risk 95% Cl
Cohort EE+ P 0.690.53 0.54-0.86 0.30-0.87
Case-control EE+ P 0.690.68 0.47-1.02 0.38-1.22
Cohort EE+ P 0.60.39 0.43-0.83 0.19-0.78
Meta-analysis 25 studies* EE+ P 0.70.66 0.65-0.75 0.53-0.84

In her comments on the HERS study, Diana Petitti urged that caution be exercised towards observational results for coronary heart disease in users of oestrogen replacement therapy and Hormone replacement therapy. She argued that women with healthy behaviour, such as following a low-fat diet and exercising regularly, may selectively use postmenopausal hormones. These differences in behaviour may not be taken into account in the analysis of observational studies because they are not measured. Estimates of the relative risk of coronary heart disease in hormone users will then be biased towards finding a protective effect of hormone use for coronary heart disease. Barrett-Connor demonstrated such a prevention bias. In two different randomized trials of drugs to prevent coronary heart disease, the subjects who faithfully took their placebo had a lower risk of coronary heart disease than subjects who were noncompliant in taking placebo. The subjects compliant in taking placebo had 50% fewer cardiovascular events than noncompliant subjects in the Coronary Drug Project and in the Beta-Blocker Heart Attack Trial. A similar reduction in the relative risk of coronary heart disease was found in compliant subjects compared with noncompliant subjects taking active drug. Women included in epidemiological studies using oestrogen replacement therapy or Hormone replacement therapy

are compliant. Compliance bias is large enough to explain entirely reductions in the relative risk of coronary heart disease between users and non-users of oestrogen replacement therapy and Hormone replacement therapy of the magnitude found in observational studies. In the study of Horwitz et al adjustment for multiple known predictors of coronary disease did not eliminate the decreased risk for coronary disease associated with good adherence to medication. Therefore, although the results published by several observational studies are consistent, they may reflect consistent biases.

Coronary heart disease

A recent meta-analysis by Barrett-Connorof 25 studies on the relationship between Hormone replacement therapy and coronary heart disease demonstrated a relative risk for oestrogen users of 0.7 (95% confidence interval, 0.65-0.75) and 0.66 (95% confidence interval, 0.53-0.84) for studies that assessed combined Hormone replacement therapy. A recent epidemiological study analysing the relative risk of myocardial infarction of oestrogen/progestin users versus non-users indicated a protective effect for all therapies as compared to no treatment, even in the group using levonorgestrel, one of the most androgenic progestins used in Hormone replacement therapy.

Grodstein et al reported the relationship between cardiovascular disease and Hormone replacement therapy in 59,337 women followed for up to 16 years. Comparing the risk of major coronary heart disease for women who did not use hormones, the relative risk was 0.6 (95% confidence interval, 0.43 — 0.83) for women using oestrogens alone and 0.39 (95% confidence interval, 0.19-0.78) for those using combined hormones. The authors found no association between stroke and use of combined hormones. For this study, conducted in the United States, the oestrogen used was mainly conjugated equine oestrogens, and the progestin most commonly used was medroxyprogesterone acetate.

Although this large observational study is in accord with the previous results of Falkeborn et al in Sweden and Psaty et al in the United States, selection bias is possible.

The long-term randomized controlled trials in progress have been designed to eliminate such bias.

Primary and secondary prevention trials

Several ongoing randomized trials focus on primary or secondary prevention of coronary heart disease.Of the large ongoing studies, the secondary prevention trial of the HERS research group is the first published. The other trials will be completed early in the 21st century.

Table Key ongoing trials focusing on primary or secondary prevention of coronary heart disease.

Steroidsstudied Population Expected results
Primary prevention
WHI conjugated oestrogens/medroxyprogesterone acetate 164,000 2005
WISDOM conjugated oestrogens/medroxyprogesterone acetate 34,000 2006
Secondary prevention
HERS conjugated oestrogens/medroxyprogesterone acetate 2,763 with coronary heart disease 1998
WEST E2 652 ?
Angiographic trials
WELL-HART E2/medroxyprogesterone acetate with coronary heart disease 2001?
ERA Placebo E/E+ P 309 2000

Long-term primary prevention trials

The Women’s Health Initiative (WHI), in the United States, and the Women’s International Study of Long Duration Oestrogen after Menopause (WISDOM), in Europe, are designed as long-term primary prevention trials and will enroll postmenopausal women without coronary disease. Long-term follow-up under treatment or placebo will help to answer questions about the presumed protective effects of Hormone replacement therapy to prevent the occurrence of cardiovascular disease and coronary heart disease.

The WHI is a large multicentre trial enrolling more than 160,000 postmenopausal women. In one subgroup of 16,500 participants the effects of Hormone replacement therapy will be assessed. Women with a uterus will receive oestrogen plus a progestin and hysterectomized women will receive only oestrogen. The WHI will conclude around 2005 but may give earlier interim data on cardiovascular events.

The WISDOM trial is another primary prevention study. Thirty-four thousand women will be included from 12 European countries. They will receive for 10 years either conjugated equine oestrogen with or without medroxyprogesterone acetate depending on uterine status, or a placebo.

Secondary prevention trials

The results of HERS do not support the beneficial effect of Hormone replacement therapy in secondary prevention of cardiovascular disease. This well-designed randomized, double-blind, placebo-controlled study evaluated the role of continuous oestrogen (conjugated equine oestrogen) plus progestin (medroxyprogesterone acetate) in preventing recurrent cardiovascular disease events in 2763 women with documented coronary disease.

The women were followed up for 4.1 years on average. The main objective of the study was to determine whether Hormone replacement therapy would reduce the morbidity and mortality of cardiovascular disease. The primary outcome was the occurrence of non-fatal myocardial infarction or coronary heart disease death.

During follow-up the same number of events were recorded in both the treatment and the placebo groups. The relative hazard for a further event was 0.99 (95% confidence interval, 0.8-1.2). There was no significant difference between groups in any of the secondary outcomes. This lack of difference in outcomes was found despite a net decrease in low density lipoprotein and an increase in high density lipoprotein cholesterol levels in the hormone group.

The authors concluded that they do not recommend starting Hormone replacement therapy for secondary prevention of coronary heart disease. However, they considered it appropriate for women receiving Hormone replacement therapy to continue, as there appeared to be a reduction of coronary heart disease in the treated group in the latter part of the trial.

Among the numerous comments published on the HERS study, Ong et al highlighted the fact that most women included in the HERS trial were enrolled toward the end of the study period (mean 4.1 years; range 3.5 — 5.3 years). Since Hormone replacement therapy was found to induce early thrombotic events and late beneficial effects (from year 3 onward), insufficient long-term follow-up of the patients recruited late in the study may have biased the results.

Although at first glance the results of HERS are disappointing, the following points are important:

  1. 1. The findings of observational research cannot be extrapolated to direct secondary prevention.
  2. 2. Although the lipid hypothesis has dominated our thinking, the mechanisms of cardiovascular disease are likely to be much more complex.
  3. 3. A negative result in secondary prevention trials does not imply such an effect in primary prevention. Postmenopausal women without previous coronary heart disease may respond to Hormone replacement therapy differently to those with coronary heart disease.

A more recent study has challenged the assumption that Hormone replacement therapy is cardioprotective.

Duke University researchers performed an observational analysis of 1857 postmenopausal women with coronary artery disease. Nearly one-third of the women who started Hormone replacement therapy after myocardial infarction were hospitalized with unstable angina within a year, compared with 21% of those who were already taking Hormone replacement therapy and only 17% of those who were never on Hormone replacement therapy. The authors concluded, as Hulley et alpreviously did, that women who have heart disease should not start using Hormone replacement therapy, but that there is no reason to suggest that women should stop using hormones if they develop heart disease while on therapy.

Angiographic trials

Angiographic trials are ongoing to compare oestrogens with and without progestins to placebo on the vessels of women with previously documented coronary heart disease. These studies may answer the question of whether Hormone replacement therapy can slow or even reverse the accumulation of atherosclerotic plaque in the coronary arteries by providing direct evidence of anatomic changes in the arteries. The Estrogen Replacement and Atherosclerosis (ERA) study is a three-arm trial comparing a placebo group, an oestrogen-only group (conjugated oestrogens), and an oestrogen plus continuous progestin group (medroxyprogesterone acetate) in a total of 309 women. Angiographic information will be obtained at baseline and after three years of follow-up. The Women’s Estrogen/Progestin Lipid Lowering Hormone Atherosclerosis Regression Trial (WELL-HART) is examining the effects of oral oestradiol with or without cyclic medroxyprogesterone acetate on progression of coronary stenosis.

Mechanisms of the protective effect of sex steroids on the arterial wall

From studies performed in the primate model, it is now established that sex hormones may prevent plaque formation and modulate the vasomotor response of arteries, particularly the coronary arteries. On the other hand, once the vessels are atherosclerotic, oestrogen treatment may have no effect on intimal hyperplasia or arterial remodelling in the injured artery.In the cynomolgus monkey, it has been shown that 17β-oestradiol preserves the normal endothelium-mediated dilatation of coronary arteries. However, whereas addition of cyclic or continuous medroxyprogesterone acetate to oestrogens caused inhibition of vasomotor responses by 50%, natural progesterone and other non-androgenic progestins did not reverse this effect. Miyagawa et al compared the effects of progesterone and medroxyprogesterone acetate on coronary artery vasospasm, and showed that progesterone plus oestradiol protected against vasospasm but medroxyprogesterone acetate plus estradiol did not. The results of the study of Williams etal indicated that a non-androgenic progestin, nomegestrol acetate, does not diminish the beneficial effects of oestrogen on the coronary dilator response in monkeys.

Therefore, it appears that not all progestins act similarly on vasomotor tone. Different Hormone replacement therapy combinations may affect the vascular reactivity in different ways. Current data suggest that non-androgenic molecules appear to be safer in this respect.

Unfortunately, most of the large ongoing RCTs have selected the same Hormone replacement therapy regimen for their study design. In these circumstances we shall not get clear answers about possible beneficial effects of other treatment regimens.

Stroke and venous thromboembolism

Hormone replacement therapy has not been shown to be consistently associated with either a reduced or increased risk of stroke of either haemorrhagic or thromboembolic origin.

In recent observational studies current users of Hormone replacement therapy have been found to be at increased risk of venous Thromboembolism. All the studies showed an increased risk of deep venous thrombosis and/or pulmonary embolism in women currently taking Hormone replacement therapy. The relative risks were between 2.1 and 3.5.

The fact that risk appears to be concentrated in the first year suggests that some women, more sensitive or with predisposing factors, will develop thrombotic events with any Hormone replacement therapy and then stop therapy, whereas the rest will remain longer-term users.

The recent results of HERS indicate that combined treatment with conjugated equine oestrogens (conjugated oestrogens) and medroxyprogesterone acetate increased the rate of thromboembolic events in women with previous coronary heart disease, as compared with placebo (relative hazard 2.89; 95% confidence interval, 1.5-5.6). This double-blind, randomized, placebo-controlled study confirms the results of the observational studies discussed above.

A further secondary prevention trial on that specific end-point is ongoing; the Women’s Estrogen for Stroke Trial (WEST) includes 652 women and examines the effect of 17β-oestradiol in women with documented transient ischaemic attacks or stroke. The primary outcome of the trial is stroke and/or death during the 3 year study period.

Short-term trial using surrogate markers of risk

A large randomized controlled trial study on Hormone replacement therapy and cardiovascular risk factors was published recently. In the Postmenopausal Estrogen/Progestin Intervention trial (the PEPI trial), 875 postmenopausal women were followed for 3 years in a randomized, double-blind, placebo-controlled study. The three combined regimens of oestrogen and progestin induced an increase in high density lipoprotein levels and a decrease in serum low density lipoprotein levels. However, the increase in high density lipoprotein was partially reversed in the groups where oral medroxyprogesterone acetate was added to oestrogens, whereas oral micronized progesterone did not modify the oestrogen-induced rise. The results for low density lipoprotein with oestrogen were not modified by the addition of a progestin, either medroxyprogesterone acetate or progesterone.

However, lipid levels are only a surrogate marker of risk for cardiovascular disease, accounting for only part of the presumed preventive effect of Hormone replacement therapy. Large studies on indirect markers of risk will not give the most relevant information. This should be expected from the randomized controlled trials looking at cardiovascular disease events as primary outcomes.

Hormone replacement therapy and heart disease: Conclusion

The established indications for prescribing Hormone replacement therapy are well defined: symptom relief and consequent improvement of quality of life.

The prevention of bone fractures and coronary heart disease events remains controversial. Although several studies have demonstrated the benefits of oestrogen in decreasing postmenopausal bone loss, only a few prospective studies have shown actual prevention of fractures. For coronary heart disease the findings from observational studies are strong and consistent but several biases have been identified, the strongest of which is compliance bias. Although the secondary prevention trial of the HERS group showed no benefit for Hormone replacement therapy in women with coronary heart disease using a given hormonal regimen, we should not extrapolate and refute the benefits of treatment as a primary prevention in normal postmenopausal women. The ongoing trials in a low-risk population should provide answers for the same hormonal regimen as in HERS. Other large trials running in Europe may indicate the effects of different hormone regimens.

The mechanisms that mediate the rapid non-genomic effects of oestrogen and the longer-term effects of oestrogen on blood vessels are not fully understood. However, rapid progress in this field may lead to more specific therapies such as oestrogens with relative selectivity for the vasculature.

Until these studies are concluded, we should carefully evaluate risk factors for each individual woman and select the most appropriate therapy, favouring natural oestrogens and non-androgenic progesterone derivatives.


Selections from the book: “Hormone Replacement Therapy and the Menopause”, 2002

Share →