Bethesda, MD - There is new evidence that timing of initiation of hormone replacement therapy (HRT) may determine the extent of its cardiovascular effects, this time from a new analysis of the Women's Health Initiative (WHI) randomized controlled trials [1]. The study suggests that coronary heart disease (CHD) risk associated with HRT is not significantly increased in women who take HRT within 10 years of the onset of menopause and even showed a trend toward reduced CHD risk but indicates that this risk increases if women start taking HRT after a longer gap. Risk of stroke, however, was increased at any time postmenopause.

Lead author Dr Jacques E Rossouw (National Heart, Lung, and Blood Institute, Bethesda, MD) emphasized to heartwire that the findings should be viewed as "reassuring" for younger women with moderate to severe menopause symptoms who have worried about the risks of starting HRT. The study should not, however, be viewed as supporting any role for HRT to reduce cardiovascular events.

"This analysis should be appreciated mainly for the somewhat-positive message for the short-term use of HRT for menopausal symptoms," he said. "For those women who were afraid to use it before, there's some encouragement here that it might be a reasonable thing to do in the short term. But it doesn't change anything about our overall recommendations—that is, you don't use HRT for the prevention of heart disease at any age."

Rossouw and colleagues report the results of the WHI analysis in the April 4, 2007 issue of the Journal of the American Medical Association. While subgroup analyses of both the estrogen-only WHI trial as well as the estrogen-plus-progesterone WHI trial have suggested a trend toward reduced CHD deaths in women age 50 to 59, this new analysis, combining the two studies, carries greater statistical power for assessing the association, authors say.

For their study, Rossouw et al evaluated CHD outcomes and "other" outcomes among the 10 739 postmenopausal women who had undergone a hysterectomy, randomized to either placebo or estrogen in the estrogen-only WHI study and among the 16 608 postmenopausal women, without hysterectomy, randomized to placebo or estrogen plus progesterone in the combination-therapy study. All women were between the ages of 50 and 79 and recruited from one of 40 US centers during the 1990s.

When analyzed by time since menopause began, risk of CHD and risk of total mortality was nonsignificantly reduced in patients taking HRT within 10 years of menopause but rose as time since menopause increased. Risk of stroke, however, was elevated across all time points. In an age-based analysis, the number of events increased with age, but there was no statistically significant added effect of HRT for any outcome by age, although risk of CHD events trended higher in women without prior cardiovascular disease.

"What the findings suggest is that CAD risk from hormones is not increased close to the menopause; however, there is increasing risk for women from hormone therapy as they move further from the menopause," Rossouw commented. "We also find that the stroke risk from hormones is increased irrespective of years since menopause. Finally, we found that the increased risk in the older women is really confined to those who have persistent hot flashes and night sweats, and in part that seems to be because those women have more of the conventional risk factors like high blood pressure, high cholesterol, diabetes, overweight, and so forth, and that's a new finding, too."

The recommendation that HRT not be used for preventing heart disease stands.

Rossouw cautioned that the finding of a trend toward reduced CHD events in women who were closer to menopause should not be overblown. "We know for sure that older women who start [HRT] are at higher risk, so even if you have lower risk when you started, we have no way of being sure that if you start early and continue for 10 or 20 years that any benefit or lack of harm will persist. As women get older, their arteries get older, and so at what point does any potential benefit switch to risk? That's really unknown, and unknowable, because you could never do such long-term trials."

As such, he says, "The recommendation that HRT not be used for preventing heart disease stands. This supports that. But it also supports the approved indication that it can be used for women with moderate to severe menopausal symptoms, and those are symptoms that interfere with daily activities."

The stroke findings, however, suggest that any woman considering HRT soon after the start of menopause should be screened for stroke risk factors, primarily high blood pressure, Rossouw said. The risk of breast cancer also needs to be taken into account, the authors note, pointing out that in this study, 72 cases of breast cancer occurred in women less than 10 years postmenopause who were taking estrogen plus progesterone, compared with 57 cases among women taking placebo.

To heartwire, Rossouw reiterated that the study findings are in keeping with current recommendations that specify HRT should be used only short-term to treat moderate to severe vasomotor symptoms and primarily in younger women close to menopause.

While other studies have indicated that HRT may have CHD benefits—or at least less risk—in women younger than 50 or in women who are even closer to last menstrual period, the current analysis was not powered to look at this interaction. Researchers are exploring this possibility in other studies and are also investigating whether it is younger age or time since menopause that is more important. Preliminary statistical tests for strength of the trend suggests it is the delay postmenopause, not age, that may be more important, but this is only hypothesis-generating at this stage, Rossouw commented.

"It seems that it's the lowering of estrogen levels, rather than age, which is somewhat more important, but it's hard to be sure, because they are so highly correlated," he said. "For what it's worth, we think that years since menopause, that is, years since reproductive hormone levels were decreased either naturally or surgically, seems to be the more important factor. It does support a hormone hypothesis."

Commenting on the study to heartwire, Dr Marian C Limacher (University of Florida, Gainesville), who was an original investigator for the WHI but not an author on the current study, said the analysis "lends hard data to some of the speculation" on the HRT timing hypothesis.

"It's not a perfect confirmation of the theory that beginning hormone therapy before vascular injury is protective, but it is somewhat reassuring that women who are closer to menopausal years did not have an increase in CV end points."

She also warned of the hazards of overinterpreting the stroke end point, since the absolute number of events in both the HRT and placebo groups was very small, particularly in the youngest age group (44 out of 4476 patients in the HRT group and 37 out of 4356 in placebo-treated patients aged 50-59).

During a Wyeth-sponsored teleconference discussing the WHI paper, Dr Lila Nachtigall (New York University School of Medicine, NY) suggested that the WHI analysis should be reassuring not only for younger women "who have really been suffering from hot flashes" but worried about starting HRT, but also their cardiologists.

"Their cardiologists often take them off [HRT] if we put them on because they're so afraid that we're increasing the risk of heart disease, which clearly we're not doing," Nachtigall stated.

But Limacher pointed out to heartwire that most of the women seeing a cardiologist would be over 60. "Most of the patients being seen by the typical cardiologist will be older, and there is no evidence that hormone therapy should be continued; in fact, the longer one is away from the age of menopause, the higher the risk. I think it is reasonable for cardiologists treating patients largely with or at high risk for coronary disease to recommend discontinuation of these forms of hormone therapy."