9 Oestrogen therapy and cardiovascular disease: Do the benefits outweigh the risks?

Abstract

The clear majority of studies which examined the effect of non-contraceptive oestrogen on CVD found moderate to marked reductions in CVD among users. In fact, of the 31 studies evaluated (one clinical trial, 14 prospective, 13 case-control and three population-controlled), 28 are consistent with a 30% or greater reduction in the risk of CVD among users, and 25 are consistent with a 50% or greater risk reduction. A meta-analysis of these 31 reports, weightin each risk estimate for size and type of study, finds an overall relative risk of 0.45. Clearly, at this time, the evidence strongly supports the contention that unopposed oestrogen use is associated with a marked reduction in CVD occurrence. Unfortunately, only three of these studies address the effect of oestrogen/progestin use on cardiovascular risk (Nachtigall et al, 1979; Hunt et al, 1987; Thompson et al, 1989). The study by Nachtigall et al (1979) was very small, but the result were reassuring, as women assigned to the oestrogen/progestin therapy had a marked reduction in CVD occurrence. The other two studies (Hunt et al, 1987; Thompson et al, 1989) were conducted in Great Britain and, like the clinical trial, are limited by small numbers of users and various progestin formulation and dosages. However, these reports are also generally reassuring in that risk estimates for CVD among oestrogen/progestin users were not elevated. In the population-controlled study by Hunt et al (1987), use of oestrogen and progestin was associated with a moderate (relative risk of 0.5) reduction in CVD occurrence, while Thompson et al (1989) found a more moderate (relative risk of 0.8) reduction in risk. Thompson et al (1989) also evaluated the effect of progestin only on CVD, and found that unopposed progestin use was associated with a significantly elevated risk of CVD (relative risk of 1.9; confidence limit (CL)=1.1-3.2). One concern, given these three studies, is that the effect of progestin on CVD risk may diminish or cancel the beneficial effects of unopposed oestrogen. Clearly, while the studies are in general agreement that unopposed oestrogen is beneficial, additional studies of oestrogen/progestin therapy and CVD risk are needed. Given the problems of selection bias inherent in observational studies, a clinical trial of oestrogen and oestrogen/progestin therapy is necessary.