A Primer on Hormone Therapy

This excerpt appears in the Harvard Medical School Family Health Guide, paperback edition, by Anthony L. Komaroff, MD, published January 2005 by Free Press. For more information or to order, please click here.)

Reducing menopausal symptoms For nearly a century, women have taken hormone therapy (HT) – either estrogen alone or a combination of estrogen and progesterone/progestin – to relieve menopausal symptoms in the short term. (In many women intense menopausal symptoms last only a few years.) There is little doubt that HT effectively relieves such symptoms, although there are other options for achieving the same goal (see “Alternatives to Hormone Therapy”).

Long-term use of HT Beginning in the 1950s, women began taking HT for longer periods. Several popular theories held that there were many health benefits from continued exposure to “youthful” levels of these hormones. By the mid 1970s, it became clear that women whose uterus had not been removed were at greater risk for cancer of the uterus if they took estrogen alone, and so most women took HT preparations that included a combination of estrogen and progestin.

By the mid-1990s, many large, carefully conducted studies had concluded that HT taken for several years or longer probably did have important health effects, some bad but most good. These studies were largely cohort studies (see p.34). In general, they found that long-term use of HT appeared to reduce the risk of heart disease, osteoporosis and colon cancer, but apparently increased the risk of breast cancer, blood clots, and gallbladder disease. There was weaker evidence that HT might reduce the risk of stroke and dementia. Given the number of women affected by these medical conditions, it seemed as though — for most women — HT was more likely to be helpful than harmful.

In cohort studies, researchers compare the medical records of women who have used hormones with those who have not. Such studies are not regarded as conclusive. It was possible, for example, that the apparent lower risk of heart disease was not caused by taking HT, but rather by something else about the women who chose to take HT—something that the doctors would not have predicted and did not think to measure.

Despite the well-recognized imperfection of cohort studies, the research that had been published involved hundreds of thousands of women followed for several decades, and in general these studies came to similar conclusions. In the mid 1990s this was the best available evidence to guide doctors and patients. And the evidence seemed to say that, for the average woman, the benefits of long-term HT exceeded the risks.

The Women’s Health Initiative Doctors realized, however, that the best available evidence was not the best possible evidence. Experts agreed that the most reliable and conclusive evidence on HT would come from randomized, controlled trials (see p.31). In such a study, researchers randomly assign women to take a medicine (like HT) or a placebo (sugar pill). In such a study, the women who take HT should be identical to the women who take the sugar pill in every respect except one: whether they took HT or the placebo. Therefore, any differences in their medical condition years later are likely to be explained by the one significant difference between them: the kind of pill they took.

The government, through the National Institutes of Health, began just such a trial in the early 1990s. It was called the Women’s Health Initiative (WHI). The WHI was designed to settle once and for all the question of whether long-term hormone therapy could prevent heart disease and other degenerative conditions in healthy postmenopausal women. Women who had a uterus were placed on a specific combination HT pill (Prempro) or placebo. Women who no longer had a uterus were placed on one particular estrogen pill (Premarin) or placebo. Researchers tested Prempro and Premarin because doctors prescribed these HT formulations most often.

In July 2002, the WHI reported the results of the Prempro trial. Just like most of the cohort studies before it, the WHI found that HT users had reduced rates of colorectal cancer, hip fractures, and spinal fractures, and increased rates of invasive breast cancer and blood clots in the veins, in particular, those that traveled to the lungs.

However, the WHI Prempro trial found something strikingly different from—indeed the opposite of—the results of many cohort studies: The risk of heart disease increased. So did the risk of stroke. The increased risks of heart disease and stroke were seen in the first two years of treatment, and the increased risk of breast cancer shortly there-after.

Because heart disease and stroke are common and serious health problems, the WHI concluded that the risks of long-term Prempro outweighed the benefits. As a result, the 16,000 women in the study were told to stop taking the medication, and researchers urged the 6 million women on this medication to reconsider their decision with their doctors.

In 2003, the WHI published another report that Prempro did little to improve overall well-being or cognitive function above and beyond relieving menopausal symptoms. Worse, it doubled the risk of dementia in women over age 65.

In 2004, the WHI published results from those women that had taken estrogen (Premarin) only. While estrogen did lower the risk of hip fracture, surprisingly it did not affect the risk of heart disease, breast cancer, colorectal cancer, or blood clots in the lungs. It did, however, increase the risk of stroke, and for that reason the women in this study, too, were told to stop taking Premarin.

Putting the risks in perspective Although the WHI demonstrated conclusively that for study volunteers taking Prempro, the risks outweighed the benefits, the magnitude of the risks was quite small (see Risks Associated with Taking Prempro).

Were the women in the WHI too old? The WHI trial used the most powerful study design, involved a large number of participants, and was carefully conducted. Yet, the results apply to the type of women who were in the study and to the forms of estrogen and progesterone/progestin tested. It is uncertain if the results apply to women who are different in some important respect, or to alternative hormone preparations.

The average age for women in the WHI was 63. The average age at which women enter menopause in the U.S. is 51. Some critics argue that had the typical woman in the study been in her early 50s the results with regard to heart disease and stroke might have been different. They cite studies in animals showing that HT has beneficial effects in younger animals, with less atherosclerosis, whereas it has harmful effects in older animals with more atherosclerosis.

Other doctors counter that there were over 5,000 women (33%) younger than 60 in the WHI, and the results for them were not statistically different from those in the older women. Furthermore, none of the previous cohort studies demonstrated any clear difference in HT effect according to the woman’s age or years since menopause.

So does age matter? It would take a study like the WHI enrolling younger women to demonstrate for sure that HT has a different effect in younger women than in older women. Such a study would have to be many times the size of the WHI, and therefore very expensive, in order to provide definitive answers, because the number of heart attacks and strokes that occur in younger women is very small, under any circumstances.

No such large studies have been started. However, a randomized, controlled study of HT in younger women is under way. Instead of measuring the rates of heart disease, stroke, and other diseases, this study is measuring markers of atherosclerosis such as the thickness of the wall of the main arteries sending blood to the brain. If this smaller study shows a positive effect on markers of atherosclerosis, it will not prove that HT in younger women actually reduces the risk of atherosclerosis.

Did WHI use the right forms of estrogen and progesterone? Some argue that the results might have been different had an estrogen and progesterone/progestin preparation other than Prempro and Premarin been used. Others argue that virtually all of the earlier cohort studies that suggested HT prevented heart disease were based on Prempro and Premarin, not on other preparations.

Nevertheless, there are several reasons to think that if the WHI had used different estrogen and progesterone/progestin preparations, the results might have been different. Prempro and Premarin use multiple estrogens collected from the urine of pregnant mares: These are chemically different from the two primary forms of estrogen circulating in a woman’s body. Different forms of estrogen can have quite different effects on tissues, such as the lining of blood vessels in the heart.

Prempro also uses a synthetic form of progesterone that is chemically different from that made by a woman’s ovaries. As with estrogens, different forms of this hormone can have different effects on tissues.

Finally, most women take HT in pill form. The medication is absorbed in the gut. High concentrations of these hormones travel first from the gut through the liver. Thereafter, much lower concentrations travel through the blood to reach the rest of the body. This is quite different from what happens naturally: the estrogen and progesterone from the ovaries go immediately to all parts of the body. High concentrations do not pass through the liver.

High concentrations of estrogen passing through the liver cause the liver to make increased amounts of certain molecules that may promote atherosclerosis. For this reason, even before the results of the WHI were reported, some doctors recommended that women use an estrogen skin patch rather than an estrogen-containing pill to reduce the risks of blood clots.

Weighing Individual Factors

Tips for Tapering Off HT

THE TWO-WEEK PLAN

• Take a full dose every other day (instead of daily) for one week
• Cut back to half doses on alternate days the second week
• Then discontinue completely

F=FULL DOSE ½ =HALF DOSE

THE SIX-WEEK PLAN

• Skip one dose in the first week
• Skip one additional dose in each following week
• Discontinue use completely in sixth week