At the same time, a study of nearly 1 million women in the United Kingdom showed that those who took hormones after menopause were 20 percent more likely to develop ovarian cancer or die from it than women who never took the pills. That study was published online by the London-based journal The Lancet.
For consumers, the new research doesn't change the advice to use the lowest dose for the shortest time possible for hot flashes and other menopause symptoms that can't otherwise be controlled.
For cautious scientists, the new breast cancer numbers were more evidence of the hormone-breast cancer link.
"The story has gotten stronger," said Dr. Peter Ravdin, a biostatistician at the University of Texas M.D. Anderson Cancer Center in Houston who led the research.
Some were skeptical several months ago when Ravdin andresearchers first reported the 2003 drop in the breast cancer rate. The new numbers, which add 2004, prove this was no fluke, said Dr. Julie Gralow, a spokeswoman for the American Society of Clinical Oncology and cancer expert at the University of Washington in Seattle.
"Because it didn't bump back up again," it supports the idea that the rate has stabilized at a new lower level, said Gralow, who had no role in the study.
Brenda Edwards, one of the journal authors who is a National Cancer Institute researcher, agreed. "Now we have a statistically significant decline" over three years and clear proof of a trend, she said.
Although some recent analyses suggest heart risks from menopause hormones are not as great as had been believed for younger, newly menopausal women, the statistics out this week add to the worries about cancer.
After rising steadily through the 1990s, the breast cancer rate dipped from 2001 to 2002, from 138 cases to 135 cases per 100,000 women. After the federal Women's Health Initiative study reported in July 2002 on the health risks of hormones, use of the pills plunged.
So did the breast cancer rate the following year to 126 cases per 100,000 women. It was the steepest fall since the government started keeping records in the 1970s.
The drop was seen in all of the cancer statistics registries reviewed in the study, and no other cancer rate changed as dramatically strong signs that hormones were playing a role, specialists said.
The 2004 rate held steady at about 126 cases per 100,000.
Stopping hormone use may have stopped some cancers from growing and caused them to disappear, scientists speculate. Or it may have just slowed them down so that they won't appear until years later, said Ahmedin Jemal, anresearcher. Only time will tell which is true, he said.
Wyeth Pharmaceuticals, which makes top-selling hormone pills Prempro and Premarin, criticized the study as overly speculative. Company spokesman Dr. Joseph Camardo said hormone prescriptions continued to fall in 2004 but breast cancer rates did not decline proportionately.
Ravdin said the company's criticism does not invalidate the cancer trends.
Breast cancer is the most common major cancer in American women and the second leading cause of cancer deaths in women. About 180,000 new cases are expected to occur in the United States this year and more than 1 million worldwide.
Ovarian cancer is far less common. The British study found that even with the 20 percent greater risk from hormones, the actual risk was very low: 2.6 of every 1,000 hormone users developed ovarian cancer over five years compared to 2.2 in 1,000 non-hormone users.
Still, that means about 1,000 extra ovarian cancer deaths from 1991 through 2005, said study leaders at the Cancer Research UK Epidemiology Unit in Oxford.
Hormone use has declined already, and the new report should cause it to fall further, Dr. Steven Narod of the University of Toronto wrote in an editorial accompanying the study in The Lancet.
"We hope that the number of women dying of ovarian cancer will decline as well," he wrote.
Camardo, Wyeth's spokesman, said hormone labels already warn about an elevated risk of ovarian cancer.
On the Net:
New England Journal: http://www.nejm.org
Women's Health Initiative study: http://www.nhlbi.nih.gov/whi/index.html
American Cancer Society: http://www.cancer.org
National Cancer Institute: http://www.cancer.gov
The Lancet: http://www.thelancet.com