Hormone replacement therapy should have only a very limited role in the management of menopausal hot flushes, should not be used for the long-term prevention of heart disease, and should not be used for the long-term prevention of hip fractures, except in exceptional circumstances.

This message has been clearly articulated in the US, following the publication of early results from the Women’s Health Initiative (WHI) trial. The WHI trial is one of the largest prevention trials ever conducted in the US.

The WHI trial was stopped early by the National Heart, Lung, and Blood Institute of the National Institutes of Health, because after 5.2 years the risks of heart disease, stroke, venous thromboembolism (VTE) and invasive breast cancer were increased in healthy women using continuous combined HRT, and these risks outweighed the benefits of reduced fractures and colorectal cancer. These results were valid, said the investigators, for women of all sub-groups defined by prior health status, age or ethnicity. The high drop-out rate in the treatment arm of the study (42%) could mean that the magnitude of effects, both adverse and beneficial, have been underestimated among women who adhere to treatment.1

One of the most startling results of the WHI trial was that even short-term use carries serious risks. The risks of VTE and heart attacks appeared immediately use was initiated; the stroke risk in the second year; and the breast cancer risk in the fourth year. This calls into question the wisdom of prescribing HRT for hot flushes, unless these are severe and women are fully informed of the risks. It is particularly ethically unsustainable to continue prescribing HRT for such unproven benefits as looking younger or improved wellbeing.

The oestrogen-only arm of the WHI trial is continuing, because the balance of risks and benefits is not clear at this stage. However, earlier communications from the investigators stated that there were increased numbers of cardiovascular events and VTE in all users, but these did not reach statistical significance.2 In 2001, they reported there were ‘still more heart attacks, strokes and blood clots in women taking active hormones’.3 There are fewer participants in the oestrogen-only arm of the WHI trial compared to the combined therapy arm, so that the balance of risks and benefits will take longer to emerge.

No risk of breast cancer has emerged in users of oestrogen alone as yet. However, a very large meta-analysis (52 705 women with breast cancer and 108 411 controls) found an increased risk of breast cancer after five years of HRT use and no difference according to whether combined or oestrogen-only therapy was used.4

In an editorial accompanying the WHI paper, Fletcher and Colditz concluded that ‘the WHI provides an important health answer for generations of healthy postmenopausal women – do not use estrogen/progestin to prevent chronic disease.’5

US healthcare officials have moved quickly to relabel HRT with the study results. The Food and Drug Administration (FDA) signalled that the label for these drugs could be further altered to recommend limited treatment for a limited amount of time, as well as a black box warning that is used to warn of potentially fatal risks of drugs. The FDA has required Wyeth (the company that produces the HRT products used in the WHI trial) to remove all references to ‘replacement’. As part of a wider assessment, the FDA is examining the issue of whether or not pharmaceutical companies have encouraged women to believe menopause is a condition to be treated, rather than an inevitable and natural set of events to be managed.6

In New Zealand, women and their doctors have not received such clear guidance. The Ministry of Health took little action to warn women and/or their doctors of the risk. The Ministry informed around 1000 doctors of the results electronically, but otherwise relied on letters sent to doctors by Wyeth.7 The downplaying by the Ministry of the risks was echoed in media statements made by the Family Planning Association and the New Zealand Medical Association, and in articles in local doctors’ newspapers.8

There was a lack of consistent advice from medical opinion leaders, clinicians or public health specialists. Media statements were contradictory and these were reflected in stories from women who called the Women’s Health Action HRT Hotline. In particular, there was silence from a number of local and Australian medical opinion leaders who, over many years, have promoted HRT use to GPs and specialists.

The lack of action from the Ministry of Health contrasted with the proactive advice it gave on the risks of third generation oral contraceptives (OCs), and Diane-35/Estelle-35, where the risks are far lower. Compared to women not using oral contraceptives, there are 5 more cases of VTE per 10 000 women using third generation OCs, and 7 more cases in women using Diane-35.7 On the basis of these risks, the Ministry wrote to all doctors, and developed consumer fact sheets. For third generation OCs, the Ministry also advertised in newspapers and established a telephone helpline for women.

By contrast, the WHI found there would be 18 more cases of VTE per 10 000 women each year, as well as 7 more heart attacks, 8 more strokes and 8 more cases of breast cancer. Taking into account the benefits of 5 fewer hip fractures and 6 fewer cases of colorectal cancer, there would be 19 excess serious adverse events per 10 000 women using HRT each year. The investigators reported that within five years 1 in 100 women using HRT would have a serious adverse event.

Considering that women using HRT are well women, this is an unacceptable level of harm. It is doubtful that a drug with this risk profile would be accepted for marketing to healthy women were such approval to be sought today.

Yet the response from the Ministry has been, firstly, that the results are nothing new, and secondly, that the results are not applicable to New Zealand women, as New Zealand women using HRT are younger and less obese and use HRT for menopausal symptoms.9 The Ministry claimed that the New Zealand Guideline Group Guideline on the appropriate prescribing of hormone replacement therapy,10 released in May 2001, promulgated advice that mirrors that of the WHI trial.7 This has been used to justify the Ministry’s present inaction.

In reality, the New Zealand Guideline has been effectually rendered redundant by the WHI study. It was neutral on the issue of benefit or harm from using HRT for primary prevention of heart disease, did not mention any risk of strokes, and was equivocal about the risk of breast cancer. It recommended oral HRT for other long-term uses – such as vaginal dryness – which should not be contemplated now. It did not attempt any overall assessment of risks against benefits.

Worst of all, Pharmac figures show that the Guideline had no effect on levels of HRT prescribing.11 This throws into doubt the considerable effort that goes into developing guidelines for doctors, or at least the way they are promulgated, and emphasises the need to simultaneously get the same information out to the public.

It is difficult to come up with evidence to support the Ministry’s claim that New Zealand women are different (particularly with regard to age and weight), and use HRT differently, to women in the WHI study.

The women in the WHI study ranged in age from 50–79 years, with a mean age of 63.2 years at baseline. Nearly 80% of the study participants were aged under 70 years. Mean body mass index (BMI) was 28.5. Thirty five per cent of the participants had a BMI of 25–29, and 34% a BMI of 30 or over.

The only information we have on the characteristics of New Zealand users of HRT is the study by North and Sharples,12 which surveyed women aged 45–64 years, but unfortunately not older age groups. This found that between 1991 and 1997, the prevalence of HRT use in older age groups increased significantly. By 1997, an equal proportion of women in the 60–64 age group were using HRT to women 50–54 years. This suggests that the pattern of HRT use in New Zealand might not differ significantly from that of the WHI study.

North and Sharples’ study12 does not contain information about the BMI of the women using HRT, however, information about New Zealand women in this age group is available from other sources. Ministry of Health figures show that 26% of women aged 45–64 years and 22% of women aged 65–74 years have a BMI of 30 or over.13 While this is slightly lower than the WHI figures, the lack of exact comparability in the age groups makes conclusions difficult. On the other hand, there is more comparable data on some other indicators of health. For example, the prevalence of smoking among New Zealand women ranges from 22% for 50- to 54-year-olds, to 10% among 75- to 79-year-olds.13 In the WHI study, only 10% of the study participants overall were current smokers.

The Ministry maintained that NZ women used HRT for symptoms as opposed to prevention and implied that this would somehow make a difference to the risks. With regard to reasons for using HRT, the study by North and Sharples12 showed that 82% used HRT for symptoms, but nearly half were also using HRT for prevention purposes. Of these, nearly half were using HRT for prevention of osteoporosis and one quarter for prevention of heart disease – a proportion that had doubled between 1991 and 1997.

The WHI investigators were studying the safety and effectiveness of HRT when used for long-term prevention. Whether or not the women had menopausal symptoms was irrelevant to this task, however, as risks do not present themselves according to the reason HRT was initiated. A woman using HRT for hot flushes runs identical risks to a woman using HRT for prevention in a comparable year of use.

An interesting aspect to note from the North and Sharples study12 was the level of poor prescribing. Among women with an intact uterus, 11% were using unopposed therapy, whereas it is well known that progestogen is needed to protect the endometrium, and 15% of women who had had a hysterectomy were using combined therapy. Use of continuous combined HRT (the regimen used in the WHI trial) had increased from 0.4% in 1991, to 29% in 1997, despite the lack of evidence that this regimen produced the same level of benefits and risks as unopposed or sequential HRT. New Zealand is prone to huge shifts in prescribing habits, which can be put down to the activities of medical opinion leaders and the pharmaceutical industry.

The WHI results translate into a major public health concern. In the US, Jacques Rossouw, acting director of WHI, stated that ‘Considering that millions of American women might consider taking the estrogen plus progestin therapy, that could translate into tens of thousands of cases of breast cancer or cardiovascular disease over several years’.14 Nearly half a million New Zealand women are in the age group covered by the WHI study and are therefore potential users of HRT.15 In 2001, Pharmac estimated from the number of prescriptions filled that around 100 000 New Zealand women were currently using HRT. This is consistent with the finding from North and Sharples12 that in 1997, 20% of postmenopausal women were current users of HRT, up from 12 % in 1991, and 32% has used HRT at some time. Fifty four per cent were using combined therapy.

These figures would give rise each year to 40 extra cases of breast cancer (after three years of use), 35 more heart attacks, 40 more strokes and 90 more blood clots. It seems extraordinary to countenance such harm to health while at the same time putting in place expensive breast screening programmes and measures to reduce cardiovascular disease.

There are some larger lessons to be learned from the HRT saga. First, results from observational studies should be treated with great caution. Observational studies of HRT indicated that HRT might reduce the risk of heart disease by as much as 50%. Most studies showing a benefit of HRT on fractures were observational. More circumspect commentators argued that the heart benefit could well be the result of a ‘healthy user’ effect, as HRT users were known to be from higher socio-economic groups, slimmer, taller, and to adopt more health-enhancing behaviours than women not on HRT. The WHI trial showed that HRT actually caused harm to hearts, not benefits.

Second, positive effects on surrogate end-points do not always translate into reduction of disease. A number of studies, principally the PEPI study,16 showed that HRT had a positive effect on blood cholesterol levels, and an overall benefit on cardiovascular risk was extrapolated from this. It was on the basis of the PEPI result that the Ministry of Health approved HRT for the prevention of heart disease in 1996, even though it was never approved for this indication by the US Food and Drug Administration. The WHI study also found this positive effect on cholesterol levels, but some other mechanism, possibly the clotting effect of HRT, caused harm to the heart.

Third, diseases have been created to fit the drugs; a trend that has occurred in a number of areas of medicine over the past few decades.17 Heart disease in women was put down to the lower level of oestrogen at menopause, even though there is no steep increase in CHD cases in women at menopause, rather, a gradual rise beginning in the 40s. It is not clear how much oestrogen contributes to bone loss, as opposed to the effects of ageing, and it is unclear how much oestrogen contributes to hip fractures, as opposed to the effects of ageing, polypharmacy and so on. Despite this, heart disease and fractures were redefined as menopausal diseases, in need of treatment.

Fourth, specialists from one specialty should not claim expertise in another realm. Gynaecologists promoted the use of HRT for prevention of heart disease: many cardiologists were highly sceptical. The gynaecologists bypassed other proven interventions, such as exercise, low-fat diet and smoking cessation, and other pharmaceuticals know from randomised controlled trials to be safe and effective.

Fifth, medicine regulatory authorities need to be much more rigorous in approving indications for medications and in monitoring safety. A re-evaluation is needed as to whether trans-Tasman harmonisation of drug regulation will enhance or harm these processes.

Sixth, more distance is needed between pharmaceutical companies and doctors, and between pharmaceutical companies and the public. In particular direct-to-consumer advertising should be halted.

Seventh, the trend to defining well populations as pathological must be rejected. Only a small proportion of menopausal women have severe hot flushes. Many of the other psychological and physiological symptoms ascribed to menopause are found in both sexes and people of all ages. The experience of HRT should provoke re-examination of other well populations that are being labelled with disease.

One of the most disturbing aspects of the current reaction to the WHI study is the culture of menopause that has been revealed. The stereotype of menopause as a disease or deficiency state appears to have been internalised by women and their doctors alike. Many callers to the Women’s Health Action HRT Hotline do not know why they are on HRT but are convinced they cannot cope without it. They fear they will visibly age overnight if they stop their pills; many women have become psychologically addicted to their hormones pills and lack the confidence to go hormone-free.

The situation is providing a bonanza for companies with other pharmaceuticals to promote, both prescription medicines and complementary therapies. We will not move forward if women are simply transferred to bisphosphonates (which lack safety and effectiveness data past seven years) or spend hard-earned money on unproven therapies such as progesterone creams and phyto-oestrogen supplements. We must find good evidence of alternative approaches for women with disruptive menopausal symptoms, but most women simply need affirmation of the normality of the menopausal transition and of the benefits of a healthy, active lifestyle.

Author information: Sandra Coney, Executive Director, Women’s Health Action Trust, Auckland

Acknowledgements: The author was co-driver of the Best practice evidence-based guideline on the appropriate prescribing of hormone replacement therapy.10 However, the views expressed in this article are her own.

Correspondence: Sandra Coney, PO Box 9947, Newmarket, Auckland. Fax (09) 356 7076; email s_coney@xtra.co.nz

Although Ms Coney has made some valid points with regard to the WHI data, I would like to make the following points: