In this meta-analysis, HT reduced CHD events in younger postmenopausal women by 32% compared with placebo or no treatment, and by 34% compared with treatment in older women. It should be noted that the risk for CHD events in younger postmenopausal women is low. This meta-analysis pooled over 70,000 patient-years of treatment in younger women, thus providing the power needed to observe significant results in this age group. In order to estimate the absolute risk reduction attributed to HT, those trials without reported events were included in the analysis, thus contributing to the denominator. The event rate over 4.6 years was estimated to be 0.29% in the treatment group and 0.65% in the control group, with a number needed to treat of 256 over 5 years and an absolute risk reduction of approximately 1 event per 1,000 patient-years.
In younger women, the reduction in CHD events seen with HT is similar to the reduction in total mortality that has been seen in pooled trial data (OR 0.6 [CI, 0.4 to 0.95]).4
This reduction in cardiac morbidity and mortality is similar to that found in the observational Nurses' Health Study, which followed a cohort of 120,000 women below the age of 55 years.1
After adjusting for potential confounding variables, such as age, cardiovascular risk factors, and socioeconomic status, HT use was associated with a 40% reduction in CHD events and total mortality.1,31
Two different cutoff criteria were used to divide the groups into younger and older women. Years from menopause is thought to be a stronger risk factor for CHD events than age, and was used first as the cutoff.32
In the WHI estrogen–progestin trial, women less than 10 years from menopause had a trend toward reduced CHD events, while women below the age of 60 years had a trend toward increased risk.2
This discrepancy could be explained if some women in the younger age group had been postmenopausal for greater than 10 years. Similar results are seen when this trial is excluded from the analysis (OR 0.53 [CI, 0.33 to 0.88]). In the WHI estrogen-alone trial, women with hysterectomies under the age of 60 years had a trend toward reduced CHD events.33
It is biologically plausible that HT, if started early after the development of menopause, can slow the progression of coronary atherosclerosis.34
Hormone therapy has been shown to reduce cardiovascular risk factors such as lipids, blood pressure, cell adhesion molecules, procoagulant factors, abdominal obesity, and insulin resistance, as well as to decrease the incidence of new-onset diabetes and improve glycemic control in those with diabetes.35
In older women, HT was associated with an increase in events in the first year, followed by a reduction after 2 years. In contrast, no initial increase in cardiac events was seen in younger women. Hormone replacement has been shown to increase C-reactive protein levels, which is a significant risk factor for myocardial infarction and death, especially in those with underlying atherosclerosis.36
It is possible that HRT has an initial prothrombotic effect in older women with underlying atherosclerosis, but that this effect is not apparent in those without significant disease.
This meta-analysis has several limitations. There was marked variability in the trial characteristics, with a wide range in study size, medication used, and method of administration. The age groups used in this study were defined according to the trials as a whole and not based on individual patients. However, similar results were found when only younger women were evaluated. Many of the trials did not include CHD events as a primary outcome, but rather as an adverse event, so it is not clear how ascertainment of events was made in each trial or whether there were any unreported events. The analysis was based only on published literature and therefore is subject to publication bias. However, funnel plots of effect size versus standard error for the trials in this analysis showed no evidence of bias.
In summary, HT reduces CHD events in younger postmenopausal women, indicating that it may have a primary preventive effect if started before atherosclerosis develops. Large randomized trials of younger women are needed to evaluate clinical CHD endpoints, as well as other outcomes such as cerebrovascular events and cancer. In the interim, clinicians should individualize decisions about the use of HT, taking into account the woman's age, menopausal symptoms, and any underlying risk factors.