Establishing the risk related to hormone replacement therapy and cardiovascular disease in women
Hormone replacement therapy (HRT) has a profound impact on the cardiovascular system in postmenopausal women, achieved through its effects on metabolic risk factors for coronary heart disease (CHD) and on arterial function. Observational studies have consistently shown an association between postmenopausal HRT use and a reduced incidence of CHD. However, the largest randomised trial initially reported no overall benefit on CHD risk. Subsequent analyses and follow-up of the study have demonstrated a significant benefit for CHD risk in healthy women initiating oestrogen therapy soon after the onset of menopause. This benefit of early initiation of HRT has been confirmed in more recent trials and in meta-analyses. The dose and type of hormones at initiation of therapy appears crucial to obtaining CHD benefit.
Keywords: Arterial function, coronary heart disease, glucose metabolism, hormone replacement therapy, insulin metabolism, lipids, lipoproteins, observational studies, randomised clinical trials.
Source: Cristina Pedrazzini / Science Photo Library
Publication of the Women’s Health Initiative had a major impact on the management of postmenopausal women, as prescribers and regulatory bodies disparaged the use of HRT. However, subsequent data analysis has cast doubt on the initial interpretation of the results.
- Metabolic effects show beneficial changes in lipids and lipoproteins, glucose and insulin metabolism and body fat distribution.
- Vascular effects show beneficial changes in vasodilator and vasoconstrictor mechanisms, pro-inflammatory cytokines, and in vascular remodelling mechanisms.
- Observational studies of hormone replacement therapy (HRT) have consistently shown benefits in coronary heart disease (CHD) reduction. However, several randomised clinical trials (RCTs) have shown no overall benefit in CHD event reduction, probably owing to inappropriate dosing of hormones.
- More recent studies and meta-analyses have shown benefit with early postmenopause initiation of HRT. With the appropriate dose and hormone type, HRT appears to be beneficial for the prevention of CHD in postmenopausal women.
Hormone replacement therapy (HRT) has been used for more than 60 years to treat menopausal oestrogen deficiency and increase longevity in postmenopausal women. However, following the publication of the Women’s Health Initiative (WHI) study, the effect of HRT on cardiovascular risk has been the subject of much debate. The initial findings of this study demonstrated an increased risk of coronary heart disease (CHD) in older women, without increasing mortality. Numerous observational studies have shown the benefits of HRT on CHD,, while some randomised controlled trials (RCTs) have demonstrated negligible effects. Since there are clear biological effects of oestrogen on the cardiovascular system, with studies showing beneficial effects on classical risk factors for CHD (e.g. dyslipidaemia and insulin resistance, as well as arterial endothelial function), it seems unlikely that HRT should not benefit CHD in postmenopausal women. However, it has been shown that inappropriately high doses of oestrogen may cause cardiovascular harm owing to transient disturbances in thrombogenesis and vascular remodelling. It is interesting to note that almost all randomised studies using clinical outcomes failed to show benefits in older women,. In these studies, the average starting age of the participants was mid-60s, and included inappropriately high doses of HRT. Conversely, there were trends to benefit women in the observational studies, comprising participants in their early 50s, the average onset age for menopause, who started on an appropriate dose of HRT. In addition, a pilot study using older women on lower dose HRT did not demonstrate any cardiovascular harm. While benefits seem to be shown in younger women on HRT early in the postmenopause, there may be similar benefits in older women with an appropriate low-dose therapy.
Sources and selection criteria
A literature review was conducted using PubMed (Medline) using the key words and MeSH headings “hormone replacement therapy” and/or “oestrogen and cardiovascular disease”. Further searches were conducted of reference lists of relevant publications. High quality evidence from systematic reviews and meta-analyses and large RCTs were favoured.
Coronary heart disease – a battle of the sexes
Cardiovascular disease (CVD) includes diseases of the heart, and arterial and venous blood vessels. CHD is now recognised as the leading single cause of mortality in women in northern Europe and the Americas. Lifestyle factors (e.g. smoking, diet and physical inactivity) all contribute to CHD. Genetic and environmental factors also play a part in the incidence of CHD. While mortality rates increase with age for both sexes, and the overall incidence is similar regardless of sex, women are more likely to develop CHD at a later age than men,. Premenopausal women are at relatively low risk from the disease, but following the menopause, the incidence of CHD in women steadily increases to a level similar to that seen in men. It appears that female sex hormones may play a preventative role, since other risk factors for the disease are the same in both sexes. This is further supported by the fact that early-onset menopause leads to premature CHD.
The role of oestrogen – metabolic effects
Oestrogen clearly has beneficial effects on the metabolic risk factors for CHD, as well as on arterial function.
Lipids and lipoproteins
Oestrogen reduces overall cholesterol levels and the effect is maintained in the long term during treatment. This effect is considered to be beneficial in reducing the risk of CHD. This results primarily from a decrease in low density lipoprotein (LDL) cholesterol levels owing to an upregulation of apolipoprotein B100 (apoB100) receptor. Small dense LDL particles are more readily cleared through scavenger mechanisms rather than by the apoB100 receptors, and are more likely to become embedded in the subendothelial space. While HRT may increase the level of small dense LDL particles, it also increases their clearance from the circulation. This may reduce the likelihood of their retention in the arterial wall. Small dense LDL particles may be more prone to oxidative damage, leading to foam cell production and eventual atheroma, however, oestrogen may protect LDL against oxidative damage. Oestrogen also increases high density lipoprotein (HDL) cholesterol, particularly the HDL2 subfraction. It inhibits hepatic lipase activity and increases the hepatic synthesis of apolipoprotein AI, the main protein component of HDL and HDL2.
While the effects on LDL cholesterol levels are unaffected by the addition of progestogen, the increase in HDL cholesterol are reversed or greatly reduced with the addition of androgenic progestogens owing to an increase in hepatic lipase activity. The effects of HRT are clearly linked to the dose and route of administration, which determine its effects on triglycerides. Oral oestrogen increases triglycerides, while this is reduced or reversed by the addition of androgenic progestogens. Conversely, transdermal oestradiol reduces triglycerides, which should reduce the risk of CHD.
Tissue resistance to insulin action increases the future risk of developing both CHD and type 2 diabetes mellitus. Oestrogen has beneficial effects on the metabolism of glucose and insulin, which results in a reduction in insulin resistance,,. Oral oestrogen increases the effect more compared with transdermal oestrogen, and may help prevent the onset of diabetes. However, the effect is impaired with the addition of androgenic progestogens (e.g. norgestrel or medroxyprogesterone acetate [MPA]). Non-androgenic progestogens (e.g. dydrogesterone) do not have this unwanted effect.
Central body fat accumulation is an important risk factor for CVD. It is a common misconception that HRT results in weight gain; it has little overall effect on body weight. In fact, as many women lose weight with HRT as gain weight. Central fat distribution is linked to insulin resistance and the metabolic syndrome and, therefore, is an increased risk for CHD. The menopause contributes to an increase in central fat distribution, but HRT reverses the changes in body fat distribution associated with menopause. This results in a reduction in central fat accumulation.
Oral oestrogen is associated with an increase in coagulation activation and a transient increase in venous thromboembolism (VTE). This occurs even though oral oestrogen decreases the levels of certain clotting factors (e.g. fibrinogen and factor VII) and decreases plasminogen activator inhibitor-1 (PAI-1). This adverse effect may be avoided by the use of transdermal oestrogen, or low doses of oral oestrogen.
The role of oestrogen – vascular function
Oestrogen has wide effects throughout the vasculature, including endothelial function.
Oestrogen increases levels of endothelial nitric oxide synthase (eNOS) and subsequently increases the production of nitric oxide (NO). This potent vasodilator regulates blood pressure and platelet function, and inhibits vascular smooth muscle proliferation and the expression of adhesion molecules. Oestrogen also reduces the release of endothelin-1, a potent vasoconstrictor. Oral and transdermal oestrogen both lower levels of cell adhesion molecules, which suggests an anti-inflammatory effect on blood vessels. Oestrogen inhibits calcium channels and activates BKCa channels, increasing vasodilatation and improving arterial function. Oestrogen also reduces angiotensin-converting-enzyme (ACE) activity, which benefits cardiovascular health.
The abnormal deposition and remodelling of vascular extracellular matrix are important processes involved in the pathogenesis and progression of atheroma. The normalisation of these processes may inhibit atherogenesis. The matrix metalloproteinases (MMPs) and their tissue inhibitors are central to vascular remodelling and may contribute to the development of cardiovascular disease. A study has shown that oestradiol increases the release of MMPs in a dose-dependent manner. Therefore, increases in MMPs induced by low-dose oestrogen may normalise vascular remodelling, while high doses of oestrogen may produce large increases in MMPs and produce excessive remodelling. Therefore, the oestrogen dose at the onset of therapy is vital in terms of beneficial or harmful effects on vascular remodelling.
Renin-angiotensin-aldosterone system (RAAS)
Oestrogen plays a role in the renin-angiotensin-aldosterone (RAAS). Oral and transdermal HRT both contribute to a reduction in the activity of ACE, which reduces the risk of CVD,,. The progestogen drospirenone has antimineralocorticoid effects and also influences the RAAS by blocking the effects of aldosterone. HRT containing drospirenone therefore results in significant reductions in blood pressure in women with mild hypertension.
Clinical studies of HRT on CHD outcomes
Observational studies consistently show a benefit of HRT on CHD, but some randomised studies do not agree with these findings.
Since 1983, many observational studies have shown a relational benefit between the use of postmenopausal HRT and a reduction in CHD. The biggest and most influential research from the Nurses’ Health Study demonstrated a 40% reduction in CHD, and persisted for up to ten years. When the Nurses’ Health Study is analysed like an RCT, the findings are more similar to the Women’s Health Intiative (WHI) but supports a benefit from early initiation of therapy. Women initiating HRT within ten years of menopause have a lower, albeit non-significant, hazard ratio (0.54) for coronary disease. Furthermore, observational research has also demonstrated benefit in women with established CHD. However, there are concerns about observational studies owing to the non-randomisation to treatment. For example, women who choose HRT are healthier and less susceptible to CHD risk factors than those who do not take HRT. Concerns have also been raised that observational studies might miss early events, which could negate the benefits. However, a study from the Finnish national register used coronary death as the end point and, for this end point, early events would not be missed. In this study of more than 91,000 women, those aged 60 years and under initiating HRT showed a significant 70% reduction in coronary death, while those aged over 60 years initiating HRT showed a non-significant reduction of 30%.
The WHI, conducted a prospective RCT of HRT, using either conjugated equine oestrogens alone or conjugated equine oestrogens with MPA in more than 27,000 (16,608 in the oestrogen plus MPA arm) postmenopausal women aged 50–79 years. The study demonstrated no overall difference between treatments and placebo in outcomes of CHD. The study has subsequently been used as “evidence” that HRT does not lower the risk of CHD, however, there are problems with the results of this study. It demonstrated some possible initial adverse cardiovascular effects, perhaps owing to the high dose of oestrogen in the older women, but suggested an eventual benefit. This appeared to be greater in those women on oestrogen alone than in those on combined HRT, suggesting a possible adverse effect of MPA. Benefits were seen in women aged 50–59 years taking oestrogen alone. This was statistically significant for a composite outcome of myocardial infarction, death and coronary interventions. With 13 years post-trial observational follow-up, the reduction in myocardial infarction in these oestrogen users became statistically significant compared with the placebo users. No benefits were seen in women aged 70–79 years. The results suggested a “window of opportunity” to reduce the risk of CHD during the early postmenopausal years using HRT. Further analyses have shown no decreased risk of CHD in the first two years of treatment but a possible cardioprotective effect in women initiating oestrogen plus progestogen after six years, with a halving of the hazard ratio.
This age-dependency is supported by the findings of the Danish Osteoporosis Prevention Study (DOPS), in which women around the age of menopause were randomised to oral oestradiol, with or without cyclical norethisterone acetate (NETA), or no treatment for ten years, with a further six years observational follow-up. This showed a significant reduction in a composite outcome of myocardial infarction, death or hospital admission for heart failure in the HRT users compared with placebo users. One limitation of the study was that the number of events was very small owing to the young age of the women involved – there were five myocardial infarctions in the randomised trial (one on HRT), which increased to 16 in the observational follow-up (five on HRT).
Studies of cynomolgus macaques,, have also shown an age-dependency relationship between HRT and a reduced risk of CHD. Intervention with HRT at the time of menopause in the macaques resulted in a 70% reduction of dietary-induced atheroma compared with placebo, but a delay in initiating HRT resulted in a loss of this effect.
At the onset of menopause, women still have relatively healthy arteries, but arterial disease increases with age and there is no robust clinical evidence for the repair of already diseased arteries in reponse to oestrogen. As a result, any preventative treatment for CHD, including HRT, is more effective for younger women with less established atheroma. However, this does not mean that older women receive no benefit. Some studies have demonstrated that older women can still benefit from an appropriate dose of HRT, even those with established CHD. Evidence for this comes from the Women’s Hormone Intervention Secondary Prevention (WHISP) pilot study, in which low dose oral oestradiol plus NETA was given to older women with established CHD. The results demonstrated a trend to decreased CHD events over 12 months compared with placebo. Consequently, the benefits of HRT may still be possible outside the window of opportunity. Women with CHD should not be denied HRT if there is a good clinical indication, but the dose at initiation is crucial.
In 2009, Salpeter et al. conducted a Bayesian meta-analysis of HRT observational and RCTs, including the WHI data, and found that HRT decreases the mortality rate for younger postmenopausal women. The study built on their earlier analyses,, which demonstrated that HRT reduced the risk of CHD in younger postmenopausal women (by 32% in the 2006 study and 40% in the 2004 study). Salpeter et al. separated the data into younger women (those less than ten years from menopause or aged under 60 years) and older women (those ten years or more from menopause or aged over 60 years), and highlighted that the risk of CHD events in older women increased initially but then decreased over time. These results concur with the observational Nurses’ Health Study, which related the age at which women began HRT with a reduction in the risk of CHD events. More recently, a Cochrane review of RCTs compared oral HRT (oestrogen alone or combined with progestogen) with either a placebo or a no treatment control. It identified 19 trials, including 40,410 postmenopausal women. In agreement with the Salpeter studies, it was found that women initiating HRT within ten years of the onset of menopause had a significantly reduced risk of coronary disease and all-cause mortality. Together, these data suggest that HRT may have a primary preventative effect, particularly if it is taken before the development of atherosclerosis.
The Heart and Estrogen/Progestin Replacement Study (HERS) included 2,763 postmenopausal women (aged 67 years on average) with established CHD to assess the impact of HRT on CHD events, specifically heart attack. Participants were randomly assigned to an oestrogen/progestin combination or placebo, and treated for around four years. The study found that the use of oestrogen plus progestin in postmenopausal women with heart disease did not reduce the risk of heart attack. These results were surprising given previous observational studies that found lower rates of CHD events in women with established CHD who undertook HRT. When the researchers examined the results by year, however, they found that there was a trend towards a higher risk for CHD “events” (e.g. heart attack) during the first year of therapy. This trend gradually declined over the course of the study.
The majority of randomised trials with clinical cardiac end points have studied the effects of combined hormone therapies. The oEStrogen in the Prevention of ReInfarction Trial (ESPRIT) assessed the effect of unopposed oestradiol valerate on the risk of CHD events or death in postmenopausal women who had survived their first myocardial infarction. The results of the ESPRIT study indicated no overall difference in the frequency of reinfarction or cardiac death between treatment groups (unopposed oestradiol valerate or placebo) at 24 months. The study also suggested that the risk of all-cause mortality and cardiac mortality was insignificantly lower in the active treatment group at three months post-recruitment to the trial. While the results of ESPRIT suggest that the decision to use unopposed oestradiol valerate in a RCT was not inappropriate, the study provided insufficient evidence to endorse its use and change the considered viewpoint at that time against the use of HRT for the secondary prevention of cardiovascular disease.
The Papworth HRT Atherosclerosis Survival Enquiry (PHASE) randomised trial considered the possible benefits of transdermal oestradiol or combined oestradiol/norethisterone in postmenopausal women with ischaemic heart disease. The study analysed the results from 255 women, of whom 134 were treated with transdermal HRT and 121 acted as controls. The study found that treatment with transdermal oestrogen or combined oestrogen/progestogen did not reduce the incidence of acute CHD events (i.e. cardiac mortality, non-fatal myocardial infarction, or hospitalisation with unstable angina). In fact, the study suggested that the treatment could initially increase the incidence of these events by between 30% and 50%, although this was also followed by a decline in events. In conclusion, the study recommended against the use of transdermal HRT for this purpose. Noting the trend towards early harm indicated by the PHASE study, and a possible later benefit suggested by the HERS trial, it was suggested that it may be appropriate for women already receiving HRT to continue. However, it should be noted that the dose of transdermal oestradiol (80mg) was inappropriately high for the age of the women.
The Kronos Early Estrogen Prevention Study (KEEPS) was another randomised study, involving more than 700 women. The main outcome findings have been presented but are yet to be published. The main aim of the study was to compare low-dose oral oestrogen therapy with transdermal oestradiol, both with cyclical micronised progesterone in relation to atherosclerosis progression. Both therapy arms were compared with a placebo group. The average age of the participants was 52 years and they were all within three years of their final menstrual period. Overall, the study found that no progression in atheroma was observed in either the treatment or placebo groups. The oral route provided increased benefits for mood, depressive symptoms, anxiety and tension for menopausal women compared with transdermal oestradiol. Low-dose oral oestrogen therapy also offered increased benefits for lowering LDL and increasing HDL cholesterol levels. Conversely, the KEEPS study suggested that the transdermal route offered greater reductions in insulin resistance and increased libido. In conclusion, the study suggested that overall outcome for HRT varies according to the individual, with advantages for transdermal therapy in some women, and advantages for oral in others.
The Early versus Late Intervention Trial with Estradiol (ELITE) looked specifically to test the timing hypothesis for atherosclerosis progression. In total, 643 healthy postmenopausal women were stratified according to time since menopause, either <6 years or >10 years. They were randomised to either oral estradiol 1mg daily plus sequential vaginal progesterone or placebo for a median of five years. Measurements of carotid artery intima-media thickness were made every six months. The main finding was that women initiating HRT within six years of menopause onset had significantly less atheroma progression than those initiating HRT beyond ten years past menopause. These findings lend further support to a benefit of HRT in the primary prevention of CHD, particularly when initiating close to the onset of menopause.
The WHI had a major impact on the management of postmenopausal women. Following initial publication, prescribers and regulatory bodies disparaged the use of HRT. Subsequent data analysis has cast doubt on the initial interpretation of the results.
Numerous epidemiological observational studies indicate a beneficial effect when the treatment is started in the early postmenopause, including that of the WHI. In most countries, this is exactly the target age group for HRT use. It may be possible to develop HRT regimens around the time of menopause to reduce the risk of CHD in such women. Appropriate starting doses should ensure that any risks of stroke or VTE are minimised. The observational findings are supported by some, but not all, RCTs that have also highlighted the importance of dose at initiation in older women.
As yet, there is no firm support for the use of HRT solely for the prevention of CHD, but the evidence for HRT use in the primary prevention of CHD in postmenopausal women continues to accumulate. When given appropriately in terms of dose at initiation and types of hormones, HRT is not harmful to the cardiovascular system and may well prove to be beneficial.
Further clinical trials of different HRT types, doses and routes of administration are urgently needed. For example, the Royal Brompton and Harefield NHS Foundation Trust in the UK is currently sponsoring a study entitled OPTIMISE (Oral vs. Patch Trial in Menopause – Individualization of oeStrogEn therapy) to compare the effect of ultra-low-dose oral oestradiol/dydrogesterone combination versus a low dose transdermal therapy on the risks of VTE and CHD. Different metabolic effects of oral versus transdermal HRT administration routes may have different effects on CHD and VTE. Few studies have performed direct comparisons, and none have compared a transdermal regimen with a new ultra-low-dose oral oestradiol/dydrogesterone combination. Studies of the effects of newer HRT regimens (e.g. combinations of oestrogens classed as tissue selective receptor modulators [TSECs]) will also be needed. In the meantime, we refer clinicians to the evidence-based NICE menopause guidance documents and evidence tables.
Financial and conflicts of interest disclsoure:
Marie-Odile Gerval is partly supported by a grant from Abbott Laboratories.
John C Stevenson has received consulting fees and research support from Abbott and Pfizer, and honoraria from Abbott, Amgen, Bayer, Novo Nordisk, Pfizer and Theramex.
Citation: Clinical Pharmacist DOI: 10.1211/CP.2017.20202066
Recommended from Pharmaceutical Press
Pharmaceutical Press is the publishing division of the Royal Pharmaceutical Society, and is a leading provider of authoritative pharmaceutical information used throughout the world.Visit rpharms.com