A Reduction in Cardiovascular Mortality with Postmenopausal Estrogen Therapy
Abstract & Commentaries
Synopsis: Current ERT use was associated with decreased cardiovascular mortality and a decrease in sudden cardiac death. Breast cancer risk did not increase with current ERT use.
Source: Sourander L, et al. Lancet 1998;352: 1965-1969.
Sourander and associates performed a community-based prospective cohort study of estrogen replacement therapy in a Finnish population by inviting all women born between 1923 and 1930 to participate in a free mammography screening program. Of the 8164 invited, 7944 women enrolled. Women were followed at two-year intervals for a total of four follow-up visits. Four hundred five women died by the last follow-up visit. The mean oral estradiol dose was 1.46 mg daily and the mean duration of use at enrollment was 8.2 years. The total follow-up consisted of 53,305 person-years. Nine hundred eighty-eight women were current users and 757 were former users. Of the current users, 514 took unopposed ERT and 139 used a progestin. Hysterectomized women were excluded from the analysis for endometrial cancer. Women were classified as never, current, or former users based on information provided at enrollment and verified. Women who began hormone use during the study were excluded from the analysis. Current users had a clear survival advantage. Only endometrial cancer increased, with a relative risk (RR) of 5.06 (confidence interval [CI], 2.47-10.41). The RR for cardiovascular mortality in current users was 0.21 (0.08-0.59) and 0.75 in former users (0.41-1.37). Current users had a reduced risk of coronary artery disease, stroke, and sudden cardiac death. The RR for breast cancer was 0.57 (0.27-1.20) in current users.
COMMENT BY SARAH L. BERGA, MD
This study is interesting for several reasons. First, the estrogen most commonly used was estradiol and the dose was relatively high. Apparently, most women received either 1 mg or 2 mg daily of oral estradiol. Nonetheless, the incidence of endometrial cancer during follow-up was only 2.1%. There is a prevailing folklore in the United States that estradiol is the "strongest" or "most risky" estrogen. While this study did not afford a direct comparison of different estrogen preparations, there is no suggestion of undue risk associated with estradiol use. Breast cancer rates did not vary by estrogen use, but women in the highest socioeconomic class had the highest rates of breast cancer. Possible factors identified by Sourander et al that might explain the increased risk of breast cancer in women in the highest social class were late menopause, early menarche, short breastfeeding, or increased alcohol intake. Smoking might be a factor, but less than 3% in all groups smoked. Diet or weight might be factors, but body mass index was comparable in all groups. Time since menopause also did not explain rates of breast cancer in this study. The prevalence of diabetes and hypertension was highest in never users, but predicted survival curves showed the greatest benefit of ERT use accrued to those with hypertension and diabetes. The main limitation of the study was that few women used combined estrogen-progestin regimens. Also, bias must have existed with regard to which women were prescribed estrogen. Interestingly, most of the benefits of ERT use were lost when use was discontinued.
Overall, the results of this study are reassuring and supportive of the concept that extended ERT use after menopause is safe and beneficial. ERT appears to be a sound investment from both a personal and population perspective. Despite the evident benefits, however, only 10% of Finnish women were using ERT. We can speculate as to why rates of use are so low on both sides of the Atlantic, but I don’t think anyone really knows. It is imperative that the barriers to HRT use be better defined. It is a truism that risk and risk perception are rarely aligned. Perhaps it is time to reimburse physicians for spending time with patients so that issues of risk perception can be addressed to the patient’s rather than the insurer’s satisfaction.
Comment by Leon Speroff, MD
The conclusion of this study does not have the strength of a randomized trial, but Sourander et al, being very aware of the major criticism against observational postmenopausal hormone studies, carried out some meaningful analyses. The common and favorite critism of the observational studies is selection bias, specifically, that healthier women choose to use postmenopausal hormone therapy and, therefore, develop less coronary heart disease. Indeed, at the beginning of this follow-up study, the hormone users had less hypertension, less diabetes, and were more prevalent of smoking or previous occurrences of cardiac problems. Therefore, this study analyzed differences according to cardiovascular risk factors and socioeconomic class. The effect of estrogen therapy was the same in the lower social classes as compared with the higher social classes. When adjusted for the absence or presence of hypertension, coronary artery disease, cardiac failure, and diabetes, the protective effect of estrogen remained prominent. These results argue against a healthy user effect in observational studies, but, of course, this issue will not be settled definitively until data are available from the ongoing, long-term, randomized, clinical trials of postmenopausal hormone therapy.
Another important observation in this study is the lack of effect in former users of estrogen. It is increasingly apparent that the beneficial effect of estrogen on bone and the cardiovascular system is rapidly lost after discontinuation of treatment. For example, in the Nurses’ Health Study, reduced risk of mortality (largely cardiovascular) was lost by the fifth year after discontinuing treatment.1
There was an observation in this study that is hard to understand. The study failed to observe a decrease in the incidence of cardiovascular disease. Sourander et al suggested that the results reflected a positive effect on existing disease but a failure to prevent diease. This, in fact, would be contrary to the HERS trial and the large number of observational studies dealing with primary prevention of coronary heart disease. The authors further suggest that their findings could represent their specific method of identifying women with coronary heart disease (in this study, patients were identified through a diagnostic registry rather than by only fatal cardiac events).
Overall, this study provides further support for the contention that postmenopausal estrogen therapy protects against cardiovascular disease, and it addresses the "healthy user" criticism as well as can be done without a randomized trial.
References
1. Grodstein F, et al. N Engl J Med 1997;336:1769-1775.
2. Register TC, et al. Arterioscler Thromb Vasc Biol 1998; 18:1164-1171.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content