The etiologic factor that is most often associated with postmenopausal osteoporosis is estrogen deficiency. The results of clinical trials and case-control studies have supported the beneficial effect of estrogen replacement therapy in reducing the rate of bone loss and the incidence of fractures. However, these reports had many methodologic problems, particularly in the description of and the adjustment for confounding variables and in the definition of estrogen replacement therapy. Few of the reports gave confidence limits to accompany the statistical analyses; therefore, the absolute levels of risk or protection are difficult to determine.
If the use of estrogen were without risk of adverse effects, the accrued evidence might provide weak support for the widespread use of this agent to prevent or retard bone loss in postmenopausal women. The extent of risk related to estrogen use, however, is not clearly defined. Certainly increased risk of endometrial cancer has been associated with estrogen use. However, the addition of progesterone to estrogen replacement therapy may reduce this risk, and the overall risk may be low. Progesterone may also have adverse effects, particularly in relation to ischemic heart disease. However, low doses of progesterone have had little effect on serum lipid concentrations.
Because neither the absolute benefit nor the absolute risk of estrogen therapy is quantified, a recommendation advocating the widespread use of estrogen to prevent osteoporosis appears to be premature. A clinical trial of the effect of estrogen therapy on osteoporosis and related fractures is being considered. Perhaps a more definitive answer will result from this project. Better means of identifying people at risk and of quarantining risk should be developed. Risk factors should be carefully assessed through history taking and physical examination of all postmenopausal women, and clinical judgement should be used in prescribing estrogen replacement therapy for those thought to be at increased risk. The risk-benefit ratio for estrogen therapy in the high-risk group will differ from the ratio in the average-risk population and in the general population.