Risks and benefits of long-term treatment with estrogens

The goal of contemporary hormone replacement is to minimize net predictable lifetime risk; success therefore depends upon quantitative assessments of the net quality of life, of net morbidity and of net mortality. Estrogens ameliorate menopausal symptoms, maintain bone integrity, and produce endometrial cancer; these facts and their quantitative aspects can be stimulated. Other links are gradually becoming clear: estrogens increase biliary disease, but prevent heart disease, and, it now seems, stroke. Conventional wisdom to the contrary notwithstanding, a critical review suggests that breast cancer is probably increased in frequency by long-term estrogen use; the increase is modest as a risk factor but because breast cancer is a common disease, it is substantial in absolute terms. The addition of progestin seems attractive as a method of opposing undesirable estrogenic effects on the endometrium. In fact, there is reason for concern about the effect of an added progestin upon the risks from breast cancer, heart disease, and stroke; even the magnitude of the expected reduction in endometrial cancer may not be great when the hormone is administered sequentially. Using a simple deterministic model of post-menopausal life with hormone replacement, we have inserted available estimates of the regimen-specific relative risk for each outcome, and translated each net impact into common denominators of morbidity and mortality. While estrogens probably increase net morbidity, as measured by either the number of hospitalizations to be anticipated or by the more arbitrary measure of expected days of disability, these negative changes are mostly due to gallbladder disease and endometrial cancer, both largely treatable conditions. Largely because of protection against heart disease, estrogen replacement in modest dose is likely to reduce substantially the number of deaths to be expected in women treated through age 75 and even into more advanced age. Hormone replacement which includes systemic progestin supplementation has not been empirically tested; it may be beneficial, but it is at least as likely to be harmful. Either a modest increment in the number of additional breast cancers produced by estrogens, a modest reduction in the number of cardiac events prevented by estrogens, or a simultaneous minor shift in the same direction for each condition, would have the effect of shifting the net reduction in cumulative deaths into an increase in the number of deaths as a result of treatment.(ABSTRACT TRUNCATED AT 400 WORDS)