But there has been a lingering belief that for younger women in the early stages of menopause, hormone risks may be negligible, at least for a while. So those who are really suffering from hot flashes, insomnia and other symptoms are often told that it is probably all right to take the drugs, as long as they use the lowest possible dose for the shortest possible time. Some researchers are even testing an idea, called the timing hypothesis, that starting hormone treatment early in menopause may help protect women from heart disease.
Now, information from a vast study in Britain suggests that the women thought to be at the lowest risk from hormones may actually be at the highest risk, at least when it comes to breast cancer. The study found that women with the greatest risk of breast cancer from hormones were those who took them earliest — before or soon after menopause began.
The new findings, published on Friday in The Journal of the National Cancer Institute, are not the strongest type of evidence: they do not come from a randomized trial, an experiment in which people are picked at random to take either a drug or a placebo and are then studied and compared over time. Instead, the British study was observational, meaning that the women being studied had made their own decisions about whether and when to take hormones.
There is always a chance, in observational studies, that there is some underlying difference between people who choose to take drugs and those who do not, and that the difference — not just the drugs they have taken — may help account for different health outcomes. Observational findings are sometimes disproved by randomized trials.
But this particular observational study also has a unique strength — it included more than a million postmenopausal women, one in four British women who were aged 50 to 64 during the enrollment period, from May 1996 to December 2001.
The research, called the Million Women Study, found that in women ages 50 to 59 who had never taken hormones, 0.3 percent a year developed breast cancer. The rate was higher, 0.46 percent a year, in women who started taking the most commonly used hormones — estrogen combined with progestin — five or more years after menopause began. But it was highest of all — 0.61 percent a year — in women who started taking the drugs before or less than five years after menopause began. And the risk was increased even in women who took the drugs for less than five years.
The lead investigator of the study, Dame Valerie Beral, a professor of epidemiology at the University of Oxford, said her research group had decided to look at the interval between the start of menopause and the start of treatment because other studies had found evidence of higher risk among women who started earlier. But in the other studies, the number of women who had started early was relatively small, and so the evidence was not statistically significant. In the Million Women Study, 90 percent of the women taking hormones had begun them before or within five years of the start of menopause, so there was a better chance of finding an answer, Dame Valerie said.
One of the studies that had detected a possible but inconclusive link between earlier hormone use and increased cancer risk was the Women’s Health Initiative, the randomized trial that in 2002 found that combined hormones were causing small but significant increases in the risk of breast cancer, heart disease, strokes and blood clots in the lungs.
Dr. Rowan Chlebowski, an investigator in the Women’s Health Initiative and a medical oncologist at the Los Angeles Biomedical Research Institute, wrote in an editorial in the cancer journal that the Million Women Study provided “substantial support” for the findings from other studies.
He said the new findings made it harder than ever to define any safe timing or duration for hormone use. Women who believe they need the hormones for severe symptoms should try to stop taking them after a year or so, he said.
But Dr. JoAnn E. Manson, also an investigator on the Women’s Health Initiative and a professor of medicine at Harvard Medical School and Brigham and Women’s Hospital, cautioned that the Million Women findings were not conclusive because the study was observational.
“The jury is still out,” Dr. Manson said.
Women should not take hormones unless they really need them, she said, but those who do need them should not be afraid to take a low dose for a limited time.
“It may have a favorable benefit-risk profile in many cases,” Dr. Manson said.