Effects of Hormone Therapy May Hinge on Timing, Genes


by Sandra A. Swanson

March, 2009

Like a pendulum, the prevailing scientific opinion of hormone therapy has swung in recent years. Is it a vital preventive treatment for menopausal women, or a serious health risk? Mounting evidence suggests that hormone therapy’s effects on mood and memory depend on the timing of treatment.

In 2003, enthusiasm for estrogen treatment dwindled after the Women’s Health Initiative Memory Study (WHIMS) linked it to a heightened dementia risk in women 65 and older. But subsequent research shows that problem does not apply to all uses of estrogen therapy—instead, many cognitive health issues may stem from treatment that begins well after menopause ends. What’s more, current research may identify those who would  benefit most from hormone therapy and prompt treatments that address patients’ genetic vulnerabilities.

Conflicting Findings about Memory

In a study published in the Jan. 13 issue of Neurology, researchers analyzed the magnetic resonance imaging scans of more than 1,400 women who participated in WHIMS. The women, ages 71 to 89, had received either hormone therapy (estrogen alone or combined with progestin) or a placebo. Compared with the placebo group, participants who received hormone therapy had, on average, a smaller hippocampus and frontal lobe, which play essential roles in memory function.

The implied risks echo the findings of a May 2003 study—also part of the WHIMS—which suggested that estrogen combined with progestin could double the likelihood of dementia in women age 65 and older. But John Morrison, dean of basic sciences and the graduate school of biological sciences at Mount Sinai School of Medicine in New York City, notes that hormone treatments may have a different effect in women who experienced menopause more recently.

“The women in those studies were at least 15 years post menopause and didn’t have hormone treatments for at least 10 years” after the onset of menopause, Morrison says. “We don’t know what’s happened to estrogen receptors during that time.”

The “healthy cell bias” hypothesis may explain why hormone treatment appears to have different effects depending on timing. The hypothesis is based on the idea that estrogen treatment can exacerbate problems in damaged neurons, such as those degenerating in Alzheimer’s disease, but that it has a positive effect on healthy neurons. 

Morrison’s work underscores the distinction between “older” and “diseased.” He recently conducted a study in monkeys that suggests estrogen can help preserve neuronal health—even in aged brains. In the study, researchers induced menopause by removing the ovaries of both young and old monkeys (with sufficiently healthy neurons), then provided estrogen treatment to some animals in each group.

In behavior tests to assess the monkeys’ working memory, the older monkeys with no estrogen treatment performed poorly. Researchers then took a closer look at estrogen’s effect by examining the dendritic spines in the animals’ prefrontal cortex neurons under a high-powered microscope.

Using a computer program, the researchers assessed the spines’ density and shape and discovered that the monkeys had sufficient synaptic health to perform well—except for the older animals that did not receive estrogen.

“They had a tremendous loss of synapses,” says Morrison, particularly dendritic spines critical for brain plasticity. “The small spines are most useful for acquiring new information, and estrogen can help retain them.”

The monkeys in the study received cyclical treatment (once every three weeks) instead of a daily hormone dose (via pill or patch), which is how most women receive estrogen therapy. Morrison’s current research will focus on timing by comparing cyclical and chronic treatment in monkeys.

“Hormones in the body, in normal circumstances, virtually never stay at the same level,” he says. “Intuitively, it seems like you want variable levels of hormones in cycles. But there’s not really any solid scientific data to suggest that.”

Timing and Mood

Unanswered questions shroud hormone therapy’s effect on mood, too. But recent research has helped clarify that, as with memory, the timing of hormone treatment makes a difference.

Peter Schmidt, head of behavioral endocrinology at the National Institute of Mental Health, has spent more than two decades studying the effect of hormones on mood. “For depression, the beneficial effects of estrogen on the brain seemed to be defined by giving it early in the menopausal transition, compared to late in post-menopause,” he says.

In 2001, Schmidt led a study that suggested women going through menopause who received estradiol (via a skin patch) experienced short-term, anti-depressant benefits from it. Mary Morrison (unrelated to John Morrison), a professor of psychiatry at University of Pennsylvania, followed with a similar study—with one critical difference. The research, published in February 2004 in Biological Psychiatry, focused on depressed women who were five to 10 years past their final menstrual period— and found that estrogen was no more beneficial than a placebo.

Although the timing of hormone treatment can play a critical role in its efficacy, it isn’t the only critical factor. “The window of time for treatment is very important, but it’s also important to realize that changes in the hormones are not something that exerts a universal effect,” says Schmidt. “There’s some other vulnerability there, whether it’s genetic, or based on past experience, that will define the line between these hormones that change the brain” and also change behavior.

The existence of two estrogen receptors in the brain, an alpha and a beta, could steer hormone-based treatment approaches, Schmidt says.

“In animal studies, beta seems to be more involved with reducing anxiety and stress,” Schmidt says. He is currently testing the anti-depressive effect of a type of compound known as a selective estrogen receptor modulator, or SERM, that targets the beta receptors.

The emphasis on these different estrogen receptors “will probably be the most wonderful development in our understanding of hormone-based therapies over the next 10 to 15 years,” Schmidt says.

Meanwhile, whole-genome studies could help identify certain patients as ideal candidates for hormone therapy treatment. Schmidt would like to study CREB1. “Variations in that gene are present in women with depression,” he says. “And that gene is also regulated by estrogen.”

Jennifer Payne, research director at the Women’s Mood Disorders Center at John Hopkins University, mainly studies postpartum depression—but her work may also reveal the genetic underpinnings that could make some women better candidates for hormone therapy. Payne examines the biological vulnerabilities that cause hormone fluctuations to trigger mood disorders in some women.

Her work has shown that women who experience mood disorders during menstrual cycles or postpartum

are more likely to have those problems during perimenopause, the time leading up to menopause. Currently, Payne is studying pregnant women with a history of mood disorders. When completed, the study will have included 60 participants.

Payne follows up with them one week, one month, and three months after childbirth to determine if any of the women have developed depression. In addition to measuring hormone levels, she takes a blood sample of each participant for genetic testing. Payne also plans to create a control group with women who adopt infants—they experience the same environmental changes as pregnant women, but not hormonal ones.

“If we can eventually identify the genes underlying postpartum depression, they may actually have a direct bearing on who’s going to have trouble [with depression] during [perimenopause] as well,” she says.

Payne notes that research into hormones and mood is not very advanced. “We don’t even totally understand how the mood system is tied into the hormone system,” she says. “We know that at times of hormonal fluctuation, many women, if not most, have some mood changes. But we don’t know how it happens.

“We don’t understand that cascade of events. But it we can identify the genetic underpinnings of postpartum depression, we might have a clue about that cascade of events— and develop better treatments, more preventative strategies.”

A Lack of Long-term Research

Claudia Kawas, a professor of neurology, neurobiology and behavior at the University of California, Irvine, has experienced both sides of the ethical debate that swirls around hormone therapy. In 2000, Mary Sano, then a professor of clinical neuropsychology at Columbia University, Kawas and several other researchers proposed a study into whether estrogen might help stave off Alzheimer’s disease in women with a family history of the disease.

Initially, reviewers questioned the ethics of denying some women the benefits of estrogen, Kawas says. But 20 months later, the researchers were forced to halt the study when new studies suggested that estrogen treatment could carry significant health risks, including dementia. “All of a sudden, everybody [was] concerned about the ethics of us randomizing women to estrogen,” Kawas says.

Today, Kawas continues to study the effects of estrogen, primarily in the “90+ Study,” which analyzes the health of participants age 90 and older. Some of her research has indicated that estrogen helps increase longevity, but after more than two decades of studying estrogen, Kawas still is not convinced that it protects cognitive health.

Creating a comprehensive study that compares different timing of hormone therapy and analyzes the effects years later is the problem, Kawas says: “Nobody, especially at this point, is willing to fund a 20-year study … that will take women at the appropriate age and randomize them, and then wait 20 years.”