When the researchers on a major American study into the effects of hormone replacement therapy abandoned part of their trial because of fears of its link to cancer, heart attacks and stroke, it made headlines around the world.

Up to 340,000 women in Britain alone are believed to have abruptly halted treatment after learning that a combined oestrogen-progestin pill appeared to increase the risk of breast cancer by 26%, heart attacks by 29% and stroke by 41%.

Now, a year and a half after those alarming statistics emerged, they received a fresh shock when one of the same researchers told them that they had probably done so unnecessarily. Susan Johnson, professor of obstretics and gynaecology at the University of Iowa, said that women had abandoned HRT "prematurely" because of the Women's Health Initiative (WHI) study.

"There are many women for whom HRT continues to be an excellent choice for the treatment of menopause-related symptoms," she told a conference of American scientists on Sunday. "In my clinical practice, I'm putting a lot of women back on HRT."

Women can hardly be blamed for being confused. For years, HRT was promoted as a "wonder drug" that could not only tackle severe symptoms of menopause - such as hot flushes - but could also protect them from osteoporosis, colorectal cancer and even improve their general levels of energy and appearance.

Then, suddenly, they learned that it could be putting their health at risk in other ways. According to the British Menopause Society, many of the women who halted HRT use after seeing the publicity about the WHI study did so without even consulting their GPs.

But many GPs have also had problems wading through the mass of evidence and the claims and counter-claims in the WHI research and other recent studies. "Doctors are having real problems because they're no less confused than their patients. The difference is that they're supposed to have the answers," says Howard Jacobs, emeritus professor of reproductive endocrinology at University College London's medical school.

Where, weighing it all up, do experts stand at the moment? And what is the most sensible guidance for women, and medical practitioners, baffled by the unending contradictory headlines?

What emerges from speaking to those who have studied HRT in depth is that they actually agree on a surprising amount. No one has suggested that the therapy should be abandoned entirely - an important point of agreement to keep in mind when the next unsettling headline comes along - and all agree that it has been proven to help women suffering severe menopausal symptoms. "There is absolutely no doubt that it works for symptoms," says Margaret Rees, consultant medical gynaecologist at the John Radcliffe hospital in Oxford.

They also agree that it is, to date, the only drug proven to prevent osteoporosis, the crippling bone disease that blights many older women's lives.

On the other hand, even HRT's most enthusiastic users accept that the WHI study was a clinically rigorous and well conducted piece of research which has shown that it can have serious side effects.

Professor Jacobs terms the side effects "scary". Yet David Purdie, of the Edinburgh Osteoporosis Centre, says that the sudden withdrawal from the drug was a "catastrophic" reaction. "Many women have been left unprotected against symptoms and bone loss by their quite understandable fear," he adds.

But how applicable is the WHI study to British women? Dr Rees doubts whether one can generalise about risks from the American research: "It was an atypical population of women, on one particular type of HRT, on a higher dosage than we would use in the UK in that age group," she says.

The drug used in the study was a combined oestrogen and progestin therapy which is not available in the UK, although approximately 300,000 of the 1.7m women taking HRT in the UK use a similar product. "Unopposed" oestrogen therapies are only used on women who have had hysterectomies because they increase the risk of endometrial cancer. The American researchers are continuing their trial of unopposed drugs, which appear to present no increased risk of breast cancer.

But even if you accept that the WHI findings are applicable to combined HRT drugs in general, they raise as many questions as they answer. "I don't think there's any doubt in the longer term that there is a small increase in breast cancer, but it's extremely small in absolute terms," says Prof Purdie. "We have to stop using percentages and talking about 'doubling' rates [of risk] and say how many extra cases there are."

The risks for breast cancer actually increase after five years, whereas Department of Health figures suggest that most British women use HRT for only around two to three years.

On top of that, as Dr Rees points out, people wrongly believe breast cancer to be the largest killer of British women - in fact only 4% of deaths each year are caused by it - and forget that many patients recover from it.

The WHI study - which included more than 16,500 women - suggested that for every 10,000 healthy women taking the combined drug, eight more will develop invasive breast cancer, seven more will have a heart attack, 18 more will suffer blood clots and eight more will have a stroke than among 10,000 healthy women not taking the treatment. That prompted the team to stop the research, on the grounds that it would be unethical to continue.

However, it also showed that six fewer women will develop colon cancer and five fewer will suffer hip fractures. But concern was amplified by the British "Million Women Study" which reported in August last year that women on some combined types of HRT were twice as likely to develop breast cancer as those who had not used it.

The study, which looked at the medical histories of almost 1.1m British women who were cancer-free as they entered the national breast screening programme, also indicated that oestrogen-only therapies might increase the risk of breast cancer by 30%.

And it suggested that the increased risk started after between one and two years of HRT use, and continued to rise the longer the treatment continued - although it disappeared within five years of halting treatment.

The researchers - funded by Cancer Research UK, the National Health Service breast screening programme and the Medical Research Council - calculated that the last decade had seen an extra 2,000 women between 50 and 64 developing breast cancer each year due to HRT, on top of the 15,000 in this age group who would usually be diagnosed with the disease.

Prof Jacobs argues that, even if one relies on the WHI's statistics, women are right to be concerned by the risks associated with long-term use. "The increased risk of breast cancer is measurable after five years (of use); cause for concern after 10," he says.

He suggests that scientists are unwilling to accept the data because the use of hormones - the "natural" option - seems intuitively correct, and points out that there was shock when researchers first showed that oestrogen alone could cause cancer of the uterus.

"We are going through exactly the same process but even more so, because the outcome of the present study is even more bleak."

"There's a terrible tendency to downplay risks," agrees Dr Klim McPherson, visiting professor of public health and epidemiology at the department of obstretics and gynaecology at Oxford University, who says that for years HRT was portrayed as a universal panacea.

"[HRT] will do wonders for your bones, no question - but it thereby increases your risk of a [heart attack], breast cancer and a stroke.

"It has a dramatic effect on breast cancer: if you are 50, only two in 1,000 women get breast cancer in a year. It goes to four in 1,000 if they are taking HRT for five years."

The problem is that to prevent osteoporosis, HRT must be taken over a long period, because its beneficial effects last only two to three years after its use is halted. That means a woman of 50 might well have to take it for 10 or 15 years, even if she later switched to bisphosphates, which treat rather than prevent brittle bones.

That, suggests Prof Jacobs, means that it might be better to encourage healthy women to think about other ways of preventing the disease - such as taking weight-bearing exercise, giving up smoking and ensuring there is plenty of calcium in their diet.

The UK's Committee for the Safety of Medicines stated that doctors should use HRT only for dealing with menopausal symptoms after looking at the WHI data. But Prof Purdie points out that the WHI study may actually offer good reason for using it to prevent osteoporosis; it clearly demonstrates that it reduces the chance of fractures in healthy women.

"It should not be forgotten that oestrogen is the natural bone protector in woman. It would be extremely surprising if such a hormone, evolved over aeons, was to kill her the moment she turned 50," he says. "Its rational and cautious use should remain an option for our daughters."

Prof Purdie also argues that doctors should be wary of assuming that bisphosphates are a safer way of preventing osteoporosis, because they are a new and relatively untested option, which have been in use for only around 10 years.

But the specialists agree on one final matter: the only person who can say whether a woman "should" be using HRT is the patient herself - after a full discussion with her GP.

Medical experts are notoriously reluctant to generalise; and with good reason. A woman who has been anorexic and whose mother has suffered severe osteoporosis may have good reason to use HRT to prevent bone fractures; another, with a family history of breast cancer but not brittle bones, may see it as an unnecessary risk.

It is not just that the risks and benefits of HRT depend on the seriousness and nature of a woman's symptoms and on her health and lifestyle and her medical history and that of her family; but that she needs to make her own decisions about how important each factor is to her.

"Women are individuals; not statistics," says Prof Purdie. "And the day we treat patients as numbers is the death of medicine as I understand it."

Working it through

In contrast to government health websites in Britain - where you will search hard to find detailed guidance to help you navigate through the scientific studies and assess treatment options involving hormone replacement therapy - the website of the network making up America's Centers for Disease Control and Prevention is a mine of information.

These centres comprise a national network, and the key to the usefulness of their site (www.cdc.gov) lies not only in its authoritative and clearly written content - but also in the workbooks that accompany each section on osteoporsis, heart disease and breast cancer.

The workbooks set out questionnaires, prompting women to think about and assess their individual situations, before seeing their GP or specialist to discuss options. The full workbook can be found at: www.cdc.gov/nccdphp/pdf/HRT_Workbook.pdf