Commentary

The study by Manson et al updates previous WHI data and provides information on additional clinical endpoints. Although CHD rates were higher in women who received hormone therapy (HT) than in women who received placebo, the conservative 95% CIs for CHD and other outcomes each overlapped 1 after adjustment for sequential monitoring and multiple outcomes. In addition, coronary risk factor status was not significantly associated with the risk of CHD for those taking HT. The high rate of discontinuation of HT (42%) and placebo (38%) suggests that the adverse effects of HT may have been underestimated.

Why does this and other randomised trials contradict previous observational studies, which showed a benefit of HT for CHD? Menopausal women who choose to take HT are more likely to be well educated and have lower cardiovascular risk (eg, lower blood pressure, weight, glucose, and cholesterol) than women who are not taking oestrogen.^1,2 Because of baseline differences between users and nonusers of HT, randomisation in the WHI study revealed the potential for an increase in CHD risk attributable to HT that was not apparent in observational studies. As the investigators point out, differences in age and years since menopause, as well as methodological limitations of observational studies, may have perpetuated incorrect impressions. The hypothesis that the timing of HT in these women was too late to halt an already irrevocable atherosclerotic process seems unlikely.

Several important questions remain. Will we find a way to predict which HT users will have increased CHD risk? What are the risks and benefits for women with current menopause symptoms (which was not the focus of WHI)? Considering the current state of evidence, clinicians cannot use coronary risk factor status to predict CHD risk associated with HT use and should integrate the possibility of adverse cardiovascular consequences (although uncommon) into their counselling for women who are starting HT.