"What we are talking about is still-fertile women, finished with having children, who have some event that leads to surgery to remove their uterus, and the surgeon removes one or both ovaries to prevent cancer," study leader Walter Rocca, MD, MPH, tells WebMD.
The 2,511 women in the study had these operations between 1950 and 1987. The women were then matched with women who did not undergo ovary removal. Rocca's team determined whether the women developed dementia by interviewing a family member or by giving the women a test by telephone.
"What we showed is that for the surgery to be a risk factor, it has to be done before menopause," Rocca says. "If before age 46 you remove two ovaries, you get a 70% increased risk of dementia. And we discovered that women who have only one ovary removed before age 38 -- this is a surgery more often done in younger women -- we see a 260% increase in dementia. That is quite a dramatic and somewhat unexpected finding."
Rocca reported the findings at the annual meeting of the American Academy of Neurology in San Diego.
Rocca expects his findings to be controversial. And they are. One skeptic is Paul Norris, MD, director of reproductive health at the University of Miami School of Medicine. Norris did not attend the AAN meeting but reviewed the written abstract of Rocca's study and a Mayo Clinic news release.
"I am raising both eyebrows as I read this," Norris tells WebMD. "There are a lot of reasons to keep the ovaries in, ranging from heart health to vagina health. But there is just no indication that it makes a difference in dementia."
There is, however, support for Rocca's theory. It comes from rat studies by Johns Hopkins neuropathologist Alena V. Savonenko, MD, PhD, and colleague Alicja Markowska, PhD, of the National Institute on Aging.
Savonenko and Markowska recently found that removing the ovaries of middle-aged rats sped the animals' age-related loss of mental function. Estrogen replacement -- if done on the proper schedule -- prevented this effect.
"I was surprised to see that there was so much similarity between Dr. Rocca's epidemiology study and our basic-science study in rats," Savonenko tells WebMD. "Our rats were absolutely doing the same thing as these women."
Still, Savonenko would be cautious in advising women not to undergo doctor-recommended ovary removal.
"I think there are very dramatic reasons why women undergo this surgery," she says. "I don't think the possible side effects on [mental function] should hold women from necessary ovary removal. But what I would suggest is consideration of estrogen replacement therapy for this cohort of women, particularly since there now is this evidence of cognitive decline."
One reason Norris is skeptical of the Rocca study is that removal of one ovary had a much larger effect than removing both ovaries.
"Why are women with one ovary taken out doing worse than those with both taken out? You have to raise an eyebrow about that," he says.
Rocca suggests that the situation is more complicated than most researchers realize. He suggests that women who have the kinds of problems that result in hysterectomy -- such as uterine fibroids or endometriosis -- have some underlying defect in the way their bodies use estrogen.
In another study he presented at the AAN meeting, Rocca found that women with Parkinson's disease tend to have defects in certain estrogen-related genes. He suggests that women who carry these unusual genes may be at higher risk of dementia -- and ovary removal simply magnifies this risk.
"So the game is more complex than just the issue of how much estrogen a woman's ovaries are making," Rocca says. "We will at some point be better able to advise the woman facing the surgeon and predict final balance between the benefit of preventing ovarian and breast cancer vs. the risk of dementia, Parkinson's disease, stroke, or heart attack. It is very complicated -- but if what we are seeing gets confirmed by others, in the future we could offer to the individual woman a more intelligent and informed basis to make a decision on ovary removal."
Norris, however, says the Rocca study offers far too little evidence for such predictions.
"I think he is out on a limb here," he says.