Estrogens, lipoproteins, and cardiovascular risk factors: an update following the randomized placebo-controlled trials of hormone-replacement therapy.

Abstract

The effects of hormone-replacement therapy on cardiovascular risk factors are examined. In an attempt to explain the results of recent randomized controlled trials in which no benefit of hormone-replacement therapy for postmenopausal women has been observed, Changes in lipoproteins in response to hormone-replacement therapy have now been analysed for both primary and secondary prevention studies. In none of the large randomized controlled trials was there any effect of hormone-induced changes in low-density lipoprotein, high-density lipoprotein, or triglyceride on clinical outcome. Further detailed studies of lipoprotein metabolism have not revealed any adverse effect of hormone-replacement therapy. Recent analysis of the Heart Estrogen/Progestin-Replacement Study data suggests hormone-replacement therapy reduces the risk of developing diabetes. The effect of hormone-replacement therapy on inflammatory markers and on flow-mediated dilatation is largely beneficial, although the effect on flow-mediated dilatation is modulated according to endothelial function, which is adversely affected by known risk factors, including age and presence of atherosclerosis. In this respect the work on polymorphisms of estrogen receptor-alpha may in due course help to define those women who would benefit most from use of estrogen. Crucially, oral but not transdermal hormone-replacement therapy increases activated protein C resistance independently of the presence of factor V Leiden. This effect increases the risk of venous thromboembolic events, which is reflected in the results of a hospital case control study of thromboembolism. Despite the outcome of the hormone-replacement therapy trials, recent work has confirmed the putative antiatherogenic effects of hormone-replacement therapy on lipoprotein metabolism. Metabolic differences of route of administration of estrogen, particularly on haemostatic variables, may explain this clinical paradox, which continues to be an important research area.