Three months after the unexpected end of a huge study of postmenopausal hormone use, a consensus is emerging that there is essentially no use for the drugs in the prevention of chronic ailments that come with age.
Although the hormones have both good and bad effects -- raising the risk for heart attack, breast cancer and blood clots while lowering it for osteoporosis and colon cancer -- their net effect is harmful in terms of disease prevention. They still have a role in the treatment of symptoms of menopause. But how large a role is a matter of dispute.
Those were among the conclusions that emerged from a two-day meeting held this week at the National Institutes of Health. The gathering was called to assess the immediate consequences of the hormone study results and talk about what new research may be needed.
"For the community of practitioners, the clear message is: If you're using hormones, try to limit it to short-term treatment for symptoms. It's not a prescription for life -- and that's a big, big change," said Florence Comite, a physician and founder of the women's clinic at Yale University, and one of about 500 medical researchers, clinicians and regulatory officials who attended.
When the results from the Women's Health Initiative hormone trial were released in early July, Comite had more than 100 patients taking combinations of estrogen and progestin, the two hormones that were studied. About one-third have stopped. "I think the dust has settled to some degree," she said.
Ronald K. Ross, a preventive medicine physician at the University of Southern California said he believes there is less confusion in the scientific community about the study results than the public might think. "Combination treatment [for disease prevention] is certainly out of the question," he said, summing it up.
Agreement on the study's implications, however, is far from universal.
Some people at the NIH workshop argued that hormone preparations different from those used in the study might have given different results. Some believe that the long-imagined cardiovascular benefit would have been seen if younger women had been enrolled, and if they'd started taking hormones as soon as they entered menopause. Some people think lower doses of the medicines won't cause harm. There was much talk about "individualizing" therapy for each woman -- a concept that appears to leave much room for the liberal prescription of hormones for long-term use.
"Is there a role for hormone therapy in prevention? Absolutely," said Frederick Naftolin, an obstetrician-gynecologist who is also affiliated with Yale. "Preventing bone loss, preserving skin, possibly in dementia. And I'm still not convinced there is no role in [preventing] cardiovascular disease."
This view appeared to be a distinct minority. A number of experts -- and authority figures -- went out of their way to say it was also an incorrect interpretation of study's results.
"It is clear that this combination hormone therapy should not be generally used for prevention purposes of chronic diseases," NIH's director, Elias A. Zerhouni, said in a statement published at the end of the workshop. Asked if this means that prescribing estrogen and progestin to a woman without symptoms is bad practice, he said: "Yes, it is bad practice."
Marcia L. Stefanick, a physiologist at Stanford University and one of the chief architects of the study, said she believes there is a real hazard in trying to find exceptions to the main findings.
"To say that the risks don't exceed the benefits -- unless you are focusing on menopausal symptoms -- is simply wrong," she said.
In what is probably the closest to an official position on the issue, the United States Preventive Services Task Force -- which provides advice to the Department of Health and Human Services -- next month will publish a guideline that "recommends against" combination hormone use for the prevention of disease in women.
In the mid-1980s, about 40 percent of postmenopausal women in America took hormone replacements at least for a while. The fraction who do today is unknown, although it is in the millions.
The main reason estrogen and progestin combinations are prescribed is to relieve hot flashes and other symptoms of menopause. The Women's Health Initiative study didn't examine that use -- or the benefit gained from the medicines' near-certain ability to relieve those symptoms. Instead, it looked at the claims that hormones, taken for years or decades, prevent disease, heart disease in particular.
The study randomly assigned nearly 17,000 women to take either a placebo or an estrogen-and-progestin combination. The participants' average age was 63, meaning that most were more than a decade beyond menopause, which occurs on average at age 51 in the United States.
Although designed to last eight years, the study was stopped after a little more than five because it was clear that the major anticipated benefit -- fewer heart attacks in hormone takers -- wasn't occurring. In fact, there were more heart attacks in women assigned to the drugs.
(A study of about 11,000 women who have had a hysterectomy -- the surgical removal of the uterus -- is continuing. They are assigned to take estrogen alone or a placebo. None is taking progestin.)
In the entire group, the number of bad events was small, and the absolute risk for any individual woman very low.
For example, at the rates seen over the course of the study, hormone use will cause an extra heart attack each year in about 1 in every 1,100 women taking the medicines; an extra stroke in 1 in 1,200; an extra blood clot of a serious nature in 1 in 600; and an extra case of breast cancer in 1 in 1,300. At the same time, hormone use will protect 1 in about every 2,000 women from a hip fracture she would otherwise have suffered; and 1 in 1,700 from a case of colon cancer.
A calculation presented at the meeting by Deborah Grady, a physician and epidemiologist at the University of California in San Francisco, suggested that given those numbers, essentially nobody is likely to benefit from preventive use of hormones.
For example, if women with family histories of colon cancer (which doubles their own risk) take hormones, the net effect still tips toward harm, with 1 in 700 users suffering an additional one of the bad events. For women with osteoporosis, it's 1 in 650. Only in women with osteoporosis who have already suffered a fracture do the risks and benefits balance out, according to Grady's calculation. But for them, she argued, there are many lower-risk interventions, such as exercise, smoking cessation, calcium and vitamin D supplements, and the anti-osteoporosis drugs called bisphosphonates.
Further analysis of the data is underway. But the investigators said the main trends were seen in all age groups and races, and that no subpopulations appear to benefit.
For example, for cardiovascular events (heart attack, stroke, clots) the risk was increased 67 percent in the youngest group of hormone users, those age 50 to 59 at the start of the study. For women 60 to 69, it was up 26 percent, and for the oldest women, those 70 to 79, it was up 18 percent. Breast cancer risk was up 23 percent, 22 percent and 42 percent, respectively, in those three age groups.
The potential wild card in hormone use is dementia.
Studies have shown that women with Alzheimer's disease are less likely to have taken hormones than women without Alzheimer's. That observation, however, doesn't mean hormones protected them. Only a randomized, controlled trial could determine that.
A subgroup of elderly women in the Women's Health Initiative study is being observed for the development of dementia. There is also a trial of two estrogenlike compounds (tamoxifen and raloxifene) underway.
A recent trial of estrogen in women who already have Alzheimer's found the hormone didn't help. However, if it turned out there was a preventive effect, that benefit might swamp the harms.
For example, out of 10,000 women who are older than 65 and have a close relative with Alzheimer's, 500 every year will develop dementia. That's far, far more than the 20 additional bad events per 10,000 women per year seen in the hormone users in the Women's Health Initiative study.