Hormone-replacement therapy and cardiovascular diseases.

Eighteen years ago, the Journal published two observational studies1,2 of the relations between postmenopausal hormone-replacement therapy and cardiovascular diseases. One found that women who used such therapy had about half the risk that nonusers had; the other found that they had almost twice the risk. In each study, the relative risk was statistically different from a relative risk of 1.0, or no effect. These results seemed to be in serious conflict, although each study taken alone looked sound. In an accompanying editorial,3 I noted differences in the drugs used, the groups of patients studied, the outcomes ascertained, and other . . .