Validation of Estrogen Status as an Independent Predictor of Coronary Artery Disease Presence and Extent in Women

Abstract

To validate the independent nature and incremental value of estrogen status concerning the diagnostic evaluation of women with suspected coronary artery disease. The study comprised a total of 1,168 patients, the first 915 of whom served as a training or derivation population. Data from the subsequent 253 patients were prospectively collected and used to validate observations from the derivation group. We performed logistic regression analysis of age, sex, symptoms, smoking habits, diabetes, hypertension, and other variables with and without estrogen status (defined according to menopausal and estrogen replacement status). Diagnostic outcomes were assessed by receiver-operating characteristic curve area analysis. Within the derivation group, estrogen status was an independent predictor of both significant (> or = 1 lesion > or = 50% stenosis) and serve (at least two vessels with > or = 70% stenosis) coronary disease on angiography. Receiver-operating characteristic curve areas for models applied to the validation group were as follows: women without estrogen status, 77 +/- 4 for significant disease and 76 +/- 5 for severe disease; women with estrogen status, 80 +/- 4 for significant disease and 80 +/- 5 for severe disease; men, 72 +/- 4 for significant disease and 74 +/- 4 for severe disease (women with versus women without estrogen status; P < 0.05). Estrogen status has significant incremental value for enhancing the diagnostic discrimination of significant and severe coronary disease in women. Estrogen status, as defined in this study, is an important and independent diagnostic clinical variable in women with suspected coronary disease.