Abstract
OBJECTIVE: To evaluate whether post menopausal hormones are associated with atherosclerosis. METHODS: We studied the relation of hormone use to coronary calcification and carotid intima-media thickness in a cross-sectional sample of 3245 post menopausal women, of whom 1620 had used hormones for various periods. Adjusted associations with three measures of hormone use (ever use, duration, and type of hormone) were estimated by multivariable regression. RESULTS: The prevalence of coronary calcification was only 4 percentage points lower in women who had ever used hormones than in women who had not (40% versus 44%), and was not monotonically related to longer use: <2 years: 38%; 2-6 years: 36%; 6-13 years: 41%; >13 years: 48%. Similarly, duration of hormone use did not show a monotonic dose-response relation with the calcium score. Mean differences in carotid intima-media thickness according to categories of years of hormone use and type of hormone ranged from -0.10 mm to +0.08 mm, with no consistent patterns. Most adjusted associations were weak and sometimes contrary to our expectation. CONCLUSIONS: We did not find meaningful associations between hormone use and subclinical atherosclerosis-neither to support benefit or harm, nor to support the prevailing theory of "healthy user" bias (namely, inverse associations due to residual confounding).