HRT diehards continue to effuse over the long-term benefits of HRT, but the latest evidence shows that long-term hormones offer a greater risk of dying early from cancer or stroke.
As little as 10 years ago, it was considered heresy to criticise the miracle of hormone replacement therapy (HRT). Doctors were confidently claiming that it kept women's bodies and minds younger, and that it prevented heart failure, brittle bones and even some forms of cancer.
Such boasts, however, were made on the basis of some very poor research data. Indeed, the swiftness with which the medical profession embraced HRT was mostly a triumph of enthusiasm over evidence - and some would say, ethics. Diehards still insist that the flood of new research proving the dangers of HRT are exceptional and should be ignored, since giving them credence will only 'panic women unnecessarily'.
But women have a right to be worried. The most recent research, all carefully conducted, suggests that exogenous (externally produced) hormones do not prevent brittle bones or heart attacks, but instead cause cancer in healthy women. Women taking HRT are also at greater risk of dying.
In fact, almost everything that we thought we knew about HRT is wrong.
This view was given further credibility when the International Position Paper on Women's Health and Menopause - an 'evidence-based' report financed by the US National Centers for Disease Control (CDC) - was published this year. The authors, 28 doctors and scientists from the US, Italy, Sweden, Switzerland and Australia, warned that while HRT can ease hot flushes and night sweats, recent trials cast doubt on its efficacy in treating heart disease, Alzheimer's disease, depression, and 'leaky' and broken bones due to osteoporosis.
Of course, the results of these trials are still being written off by enthusiasts as insignificant compared with the great relief HRT brings to women suffering from 'menopausal symptoms'. But these symptoms - most commonly hot flushes - while uncomfortable, are ultimately trivial and self-limiting. For the majority of women, they persist for no more than two years (Kronenberg F, 'Hot flashes', in Lobo RE, ed, Treatment of the Postmenopausal Woman: Basic and Clinical Aspects, New York: Raven Press, 1981).
It is difficult to understand the rationale behind giving such strong and potentially dangerous drugs to treat what may be troublesome, but essentially minor, complaints.
The progestogen myth
It is now generally accepted by even the most stalwart of HRT proponents that giving women oestrogen alone (called 'unopposed oestrogen') causes breast and endometrial cancer. Indeed, this was one of the misgivings raised by the authors of the International Position Paper on Women's Health and Menopause.
This cancer risk led to the production of oestrogen-progestogen (synthetic progesterone, called progestin in the US) combination drugs in the belief that progestogen would counteract the rapid proliferation of endometrial cells seen with oestrogen on its own. These combinations were introduced onto the market without long-term testing. Although physicians acknowledged that our knowledge of the long-term health risks was unknown, they nevertheless maintained that HRT was safe. But was it?
Oestrogen is an internationally recognised carcinogen and progestogens are rapidly being recognised as having a similar action. Studies of oral contraceptives show that progestogen alone has the greatest effect on breast cell proliferation (Hum Pathol, 1989; 20: 1139-44).
Thus, it is not surprising that combined and sequential regimes of HRT show evidence of increased cancer rates. A recent study of postmenopausal women with and without breast cancer, half of whom had used some form of HRT, found an astonishing 51 per cent increased risk of breast cancer with 10 or more years of oestrogen-progestogen use, a risk level never reached with 15 years of using oestrogen alone (J Natl Cancer Inst, 2000; 92: 328-32).
Even before HRT became de rigueur for women in menopause, studies into the use of other hormonal therapies, such as oral contraceptives, showed an increased risk of breast cancer (Cancer, 1971; 28: 1395-9; Lancet, 1983; ii; 926-30; Clin Obstet Gynecol, 1984; ii: 769-86; Lancet, 1989; i: 973-82).
We also know that rises in breast cancer rates have closely matched the rise in hormone consumption. This in itself is not proof of a link. But, as long-time opponent of hormonal therapies Dr Ellen Grant notes, when hormone use fell dramatically in the late 1970s and early 1980s due to warnings about thrombosis and cancer, rates of breast cancer also fell dramatically (Lancet, 1996; 348: 682).
Despite initial enthusiasm during the early 1970s, the research soon began to show that adding progestogens offered no protection but, in fact, only seemed to increase the incidence of cancer. Data from the long-term US Nurses' Health Study showed that the excess risk of cancer among those taking oestrogen alone was 32 per cent, but 41 per cent for those taking an oestrogen-progestogen combination.
Overall, women taking hormones for five or more years had a 46 per cent greater risk of breast cancer than those who were not. Worse, those taking oestrogen for five or more years had a 45 per cent increased risk of death due to breast cancer (N Engl J Med, 1995; 332: 1589-93).
More bad news came from other US data - specifically, the 11-year Iowa Women's Health Study - which followed 37,105 randomly selected postmenopausal women to see whether HRT increases the risk of three different types of breast cancer: benign lesions of the ducts, highly malignant cancers of the ducts or lobes, and invasive cancers with what physicians claim is a 'favourable prognosis' (JAMA, 1999; 281: 2091-7).
The use of HRT increased the overall risk of invasive cancers with a 'favourable prognosis' by nearly four and a half times. Women who used HRT for five years or less had an 81 per cent greater risk of developing these slow-growing, supposedly curable cancers than women who had never used hormones, while those who had used HRT for more than five years had more than two and a half times the risk.
Although other studies have shown a small increase in breast cancer risk with long-term hormone use, this was the first to show that HRT also causes certain types of cancer more readily than others.
Another study carried out by the Collaborative Group on Hormonal Factors in Breast Cancer was a major reanalysis of the worldwide evidence linking breast cancer with HRT, based on more than 160,000 women. This analysis found that the average length of time a woman takes HRT is 11 years, and that the risk of breast cancer increases by 2.3 per cent for every year of use. Using HRT for five years or more raised the risk of breast cancer by 35 per cent (Lancet, 1997; 350: 1047-59).
Like the Iowa study, this analysis showed that those who developed cancer while on HRT tended to develop so-called treatable, more localised tumours. But the researchers acknowledged that this was not necessarily a protective effect of HRT, but was more likely due to the emphasis on regular breast examinations throughout the study and, thus, the likelihood of earlier detection. Although doctors argue that the risk is acceptable, it is difficult to understand how giving one in every three healthy woman cancer could ever be considered 'acceptable'.
A recent study by the US National Cancer Institute used follow-up data collected in 1980-1995 as part of a nationwide breast screening programme from more than 46,000 healthy women using HRT. The findings confirmed what we already know. Women using combined oestrogen-progestogen within the last four years had a 40 per cent greater risk of breast cancer - or twice the risk of those taking oestrogen alone. The likelihood of developing breast cancer increased with every year the women took HRT. Also, thinner women were more at risk than comparatively heavier ones (JAMA, 2000; 283: 485-91, 534-5).
While the incidence of breast cancer has received attention from the media, HRT users are at risk from other cancers too, such as endometrial and ovarian cancer. One early study found that women taking oestrogen for reasons other than contraception (which has its own level of risk) had a 20 per cent greater chance of developing epithelial ovarian cancer compared with those who never took outside oestrogens (Am J Epidemiol, 1989; 130: 1142-51). The risk doubled with use for five to nine years, with use of higher doses of oestrogen and for those who had never had children. Similar findings have been reported elsewhere (Gynecol Oncol, 1996; 63: 254-7).
In a US study of oestrogen replacement therapy and fatal ovarian cancer involving more than 240,000 healthy women, oestrogen use at any point in life increased the overall risk of death from ovarian cancer by 15 per cent. The longer a woman had been using oestrogen prior to the study, the higher the risk: use of oestrogen for six to 10 years had a 40 per cent greater risk of death, while using the drug for more than 11 years increased the risk by 71 per cent (Am J Epidemiol, 1995; 141: 828-35).
Recently, researchers from the American Cancer Society studied former and current oestrogen use among more than 200,000 disease-free women over a 14-year period. Their conclusion? Taking oestrogen replacement for 10 years or longer could more than double the chances of dying from ovarian cancer (JAMA, 2001; 285: 1460-5). Although the risk did begin to tail off when HRT was stopped, excess death from ovarian cancer persisted for up to 29 years after stopping the drug.
The heart-protection myth
A decade ago, doctors were trumpeting the protective benefits of HRT on the heart. Since heart disease is the number one killer of women, this sounded like good news. Women were being prescribed HRT as a just-in-case means of preventing both primary and secondary heart disease.
While it is biologically plausible that oestrogen provides some protection against heart disease, it is only part of the picture of heart health. As one Swedish study found, the overall health of the women being studied can significantly influence trial results and falsely reassure those who interpret them (BMJ, 1999; 319: 890-3).
While there is no solid evidence that HRT prevents heart disease in healthy women but, in some, it appears to lower bad LDL cholesterol while increasing good HDL. This is because the studies that can be found only included women with heart disease, for whom HRT appears to offer some benefit.
But there is now a wealth of information suggesting that, far from keeping the heart healthy, HRT may not only provide no benefit but, under some circumstances, may even raise a woman's risk. Some women suffer from cardiovascular problems soon after taking HRT. Hormones can sometimes cause an increase in heart-damaging triglycerides (Am J Clin Nutr, 2000; 72: 389-94; JAMA, 1995; 273: 199-208) and double levels of C-reactive protein (CRP), a biochemical marker of systemic inflammation. High levels of CRP can increase the risk of vascular inflammation and heart disease (Circulation, 1999; 100: 713-6, 717-22).
The first real seeds of doubt as to the heart-protective effects of HRT were sown by the now well-known Heart and Estrogen Replacement Study (HERS) - which turned on its head everything we wanted to believe about HRT (JAMA, 1998; 280: 605-13).
Unlike many other trials of HRT, the HERS had a placebo group for comparison. It also had clearly defined clinical endpoints - disease or death. The HERS showed that Prempro (an oestrogen-progestogen combination) ultimately had no advantage over placebo in preventing coronary events in women with heart disease. Indeed, early in the study, the incidence of heart disease increased slightly among those taking the drug. Overall, taking HRT led to 50 per cent more heart attacks than did placebo during the first year of a four-year follow-up, although the placebo figures nearly caught up by the end of the study.
Other amazing findings of the HERS trial include a threefold greater risk of thromboembolic events, a 38 per cent rise in gallbladder disease, and 24 per cent more deaths due to heart disease.
The HERS and other trials that followed were bombshells, met with bewilderment and feelings of betrayal by physicians and women alike. How was it that a drug that appeared to lower LDL and raise HDL could induce heart failure and even death? Could it be that there was more to heart failure than a mechanistic cause and effect, and that using HRT to prevent heart disease was akin to using a hammer to pound the wrong piece of the jigsaw into the wrong space in the puzzle?
Then came the Estrogen Replacement and Atherosclerosis (ERA) study, which reported no difference in the progression of coronary atherosclerosis among postmenopausal women with established heart disease regardless of whether they took Premarin (oestrogen), Prempro (a combination) or a placebo (N Engl J Med, 2000; 343: 522-9).
In April 2000, when early results from the ongoing Women's Health Initiative study (due to end in 2005) came in, participants without heart disease randomly given Premarin, Prempro or placebo were notified of a small, early increase in heart attacks, strokes and blood clots in the lungs among those taking HRT. However, scientists considered the risk small and assured women that the risk would decrease with extended use, so the trial carried on.
In June 2001, the women were again notified that the risk remained. This time, the scientists admitted that they didn't know if the risk 'might change in future years' (for details, see WHI website www.nhlbi.nih.gov/whi/hrt.htm). The study continues.
Most recently, another study (originally to test the effectiveness of warfarin against aspirin) reanalysed data according to whether a woman used HRT or not (J Am Coll Cardiol, 2001; 38: 1-7). It found that women who started HRT after a heart attack were more likely to have a further coronary event within two years than those who were already taking HRT before the heart attack. The overall increased risk of death was 44 per cent, with new users at greatest risk.
As a result, some experts now advise extreme caution in prescribing HRT as a just-in-case measure for both healthy women and those with a history of heart disease. An editorial in the New England Journal of Medicine (2001; 345: 34-40) advised that cardiovascular protection should no longer be included among the benefits that women and their doctors weigh against HRT's increasing list of risks.
And while proponents continue to dispute these weighty findings, the reaction of the American Heart Association is unequivocal. The AHA now advises doctors against prescribing HRT (in particular, regimes based on horse oestrogen combined with medroxyprogesterone acetate) as a protective measure against cardiovascular events (Circulation, 2001; 104: 499-503).
Stroke risk and more
Although the risk of stroke and thromboembolism is well known with the oral contraceptive pill, scientists have refused to accept that HRT may carry the same risk. A link between these diseases and HRT has long been suspected in view of the incidental findings of other studies (see above), but early and smaller trials had failed to prove it.
Nevertheless, it was only a matter of time before HRT research caught up, and revealed that women who take either oestrogen or oestrogen-progestogen combinations have at least a threefold greater risk of fatal stroke (N Engl J Med, 2001; 345: 1243-9; Ann Intern Med, 2000; 132: 689-96).
The data are clearly consistent. A small study from the Radcliffe Infirmary in Oxford suggested that women on HRT increase their chances of thrombosis significantly and that the risk seems greatest among new and short-term users. After studying 135 women, the Oxford researchers concluded that those on HRT were 3.5 times more likely to suffer a venous thrombosis (Lancet, 1996; 348: 977-80).
These findings were confirmed by researchers at the Boston University Medical Center, who put the risk at 3.6 times after analysing 210 women. These risks can increase if doses of HRT are raised, and reached nearly sevenfold if a woman took 1.25 mg/day or more of oestrogen (Lancet, 1996; 348: 981-3). Neither study showed any difference in venous thrombosis risk between oestrogen and oestrogen-progestogen regimes.
The HERS data also show stroke risk. Women taking oestrogen-progestogen during a four-year follow-up had a 61 per cent increased risk of fatal strokes and an 18 per cent increased risk of nonfatal strokes - outcomes that the authors dismissed as not significant (Circulation, 2001; 103: 638-42).
Figures from the Nurses' Health Study suggest that higher doses increase risk. While a low dose (0.3 mg/day) halved the risk of stroke, a higher dose (0.625-1.25 mg/day) led to a 35 per cent greater risk (Ann Intern Med, 2000; 133: 933-41, 999-1001).
Down and out
A little-publicised adverse effect of HRT use is depression - often severe enough to lead to death.
While various data show that HRT may lead to an improvement in depression, these are always qualified. Most suggest that HRT has no direct action on mental state, but its ability to relieve complaints such as flushing in the short-term may have a knock-on effect in improving mood (JAMA, 2002; 287: 591-7; CNS Drugs, 2001; 15: 175-84).
A recent study from the Royal College of General Practitioners following 46,000 women using oral contraceptives for 25 years found, among other things, a significant increase in neurotic depression and attempted suicide, and deaths due to accidents and violence (BMJ, 1999; 318: 96-100).
The first inklings that exogenous hormones could lead to depression came from early studies of the Pill. In one Oxford study, four times more Pill users were admitted to hospital for attempted suicide than those using a diaphragm (J Biosocial Sci, 1976; 8: 373-424).
In the British HRT Study, women given oestrogen were more than two and a half times as likely to die from suicide than those not taking oestrogen (Br J Obstet Gynaecol, 1987; 94: 620-35).
When these data were reanalysed a few years later, a further 112 women had died. The authors found that the risk of suicide or attempted suicide for HRT users (not all of whom had a history of emotional problems) was still nearly two and a half times greater than that of the general population (Br J Obstet Gynaecol, 1990; 97: 1080-6).
The addition of progestogens to HRT has only made the problem worse. The worrying link between progestogens and depression was already known. The Norplant implant, which contains levonorgestrel, has been implicated in several cases of major depression among healthy women with no history of emotional illness (J Clin Psychiatry, 1996; 57: 152-7; J Clin Psychiatry, 1994; 55: 478-80). Studies of combined HRT are now also beginning to show that progestogens can increase both anxiety and depression (J Women's Health Gend Based Med, 1999; 8: 637-46).
Quick and dirty
According to the UK Menopausal Helpline, women can and do die almost immediately after taking HRT. Reporting of this phenomenon is still low, but several cases have surfaced recently. In one case, a man was given an out of court settlement by a pharmaceutical company when his wife died within six weeks of beginning HRT.
The Menopausal Helpline files also include a woman in Cumbria who died nine days after starting a second course of HRT, and another woman in Bournemouth in a coma, allegedly due to HRT.
The Helpline continues to monitor the situation with a view to future legal action. Women who have suffered cancer, thrombosis, stroke, epilepsy, high blood pressure, lost limbs, itching, burning or any other adverse side-effects since taking HRT are encouraged to contact them (Menopausal Helpline, 228 Muswell Hill Broadway, London N10 3SH).
Women who take HRT are not independent, free spirits marching their way confidently to a youthful third age. Women who take HRT are patients who were generally healthy before they began taking these drugs. In vulnerable women, it may prove to be the final straw that breaks the camel's back. In healthy women, the use of HRT may strongly point to a future where they will end up being treated for serious, sometimes life-threatening, diseases that they would not have developed otherwise.