Widely publicized results of the Women's Health Initiative (WHI) and the Heart Estrogen/Progestin Replacement Study (HERS), showing an increased risk for cardiac events with HRT, caused many to draw the conclusion that "hormones are bad for everybody," Salpeter told heartwire. However, the mean age of women enrolled in these two trials was about 63 and 67 years, respectively. In practice, HRT is not generally started in women in their 60s, she noted, but rather in those who have recently begun menopause. "We needed to look at that group," she said.

Estimation of mortality effects in younger women would require a large, long trial, she noted. As a surrogate, they undertook a meta-analysis, pooling 30 randomized, controlled trials of HRT vs placebo or no treatment, each longer than six months in duration and reporting at least one death. Together, the pooled trials, including both the WHI estrogen-plus-progestin trial and HERS, provided data on 26708 subjects.

When the results for all ages were combined, there was no significant difference in total mortality between women in the treatment and control groups. However, when trials were divided into those with participants of mean age less than and greater than 60 years at baseline, there was a 40% reduction in mortality for the younger age group. There was no significant increase or decrease in mortality seen in the older age group.

For all ages combined, treatment did not significantly affect the risk for cardiovascular or cancer mortality but was associated with reduced mortality from other causes (OR 0.67 [95% CI 0.51-0.88]). This may be due to a reduction in hip fractures and diabetes mellitus that is seen with hormone replacement, Salpeter noted. The subgroups for cause-specific mortality were too small to produce statistically significant results, she added, but there was a nonsignificant trend to reduced mortality from cardiovascular disease in the younger, but not older, women.

Salpeter pointed out that a paper published by Manson et al in 2003 included a graph dividing the WHI data not by age but by years since menopause at baseline.2 There was a nonsignificant trend toward a reduced relative risk for cardiac events in women 0 to 10 years from menopause, changing to a trend toward increasing risk with increasing years from menopause. "Unfortunately, we started making conclusions about younger women from what we learned about the WHI and HERS trials, and with hormone replacement, we can't do that," she said.

Contacted for comment by heartwire, WHI investigator Dr Marian C Limacher (University of Florida, Gainesville) was not convinced by the findings of Salpeter et al's study. "The selection of a highly diverse group of mostly short-duration studies in women with a variety of clinical conditions (with and without coronary disease), use of different formulations of hormone treatment, and investigating a broad spectrum of end points makes the results of this meta-analysis somewhat less than rigorous," Limacher said.

As the authors acknowledge, Limacher notes, a major limitation of this study is the use of mean age to categorize each study, rather than analyzing the mortality rates by age as a continuous variable within each study. "By this method, for example, the thousands of women under 60 in the WHI estrogen-plus-progestin study were not considered in the under-60 analysis."

"Overall," she concludes, "their report does not provide much additional information in the ongoing efforts to fully determine the risks of hormone therapy in younger postmenopausal women."

Salpeter acknowledged that pooling smaller studies is less rigorous than evaluating a larger trial but pointed out that such a trial in younger women is not yet available. "It would have been helpful to have included the WHI data on women less than 60 years in the younger age group, but the WHI investigators refused to provide us with the data. We do know, however, that there was a nonsignificant trend toward reduced cardiac events in the younger women in the WHI trial, so that is consistent with our findings," she said.

"The most important conclusion we can make from this meta-analysis is that there is a significant difference in the way younger and older women respond to hormone replacement and that there is promising evidence of a beneficial effect in younger women," she concluded.

Salpeter pointed out that a new trial has just begun, called the Kronos Early Estrogen Prevention Study (KEEPS). KEEPS is a multicenter trial conducted by a not-for-profit research institute called the Kronos Longevity Research Institute in Phoenix, AZ, studying 900 perimenopausal women aged 45 to 54 with menopausal symptoms, to investigate the hypothesis that protection may be seen from treatment in these younger women. End points are carotid intimal medial thickness by ultrasound and the progression of coronary calcium by electron beam tomography.