Current issues of ACP Journal Club are published in Annals of Internal Medicine


Estrogen status information improved the accuracy of electrocardiographic exercise testing for diagnosing coronary artery disease in women

ACP J Club. 1993 Sept-Oct;119:48. doi:10.7326/ACPJC-1993-119-2-048

Source Citation

Morise AP, Dalal JN, Duval RD. Value of a simple measure of estrogen status for improving the diagnosis of coronary artery disease in women. Am J Med. 1993 May;94:491-6.



To determine if estrogen status information can improve the diagnostic accuracy of electrocardiographic (ECG) exercise testing to detect coronary artery disease (CAD) in women.


Blinded comparison of ECG exercise testing with coronary angiography using multivariate analyses that included estrogen status information.


Exercise laboratory of a university hospital.


326 men (mean age 54 y) and 234 women (mean age 56 y) referred for exercise testing because of suspected CAD who had coronary angiography within 3 months. Patients with known CAD, abnormal resting ECGs, or unknown estrogen status were excluded.

Description of test and diagnostic standard

Clinical data were collected before testing. Resting ECGs were done and classified as normal, equivocal, or abnormal. Women were estrogen-status positive (n = 112) if they were premenopausal or took any oral medication containing estrogen (n = 63). Estrogen data were taken from medical records (50%) and patient contact (50%). Standard treadmill exercise testing was done with a positive test defined as ST-segment depression ≥ 1 mm (horizontal or downsloping 0.08 s after the J-point) in the lead with maximal ST-segment change. Angiograms and ECGs were read blinded to other test results. Angiogram-defined CAD was ≥ 1 vessel with ≥ 50% luminal diameter narrowing estimated visually in 2 views.

Main outcome measure

Area-under-the-curve (receiver operator characteristic [ROC] curves) comparing exercise testing with coronary angiographic results, with and without inclusion of estrogen status.

Main results

134 men (45%) and 75 women (36%) had CAD (P < 0.02). Pretest ROC curve areas were 0.79 for women without estrogen status, 0.85 for women with estrogen status, and 0.78 for men (P < 0.001 for women with estrogen status vs other groups). Post-test ROC curve areas were 0.83 for women without estrogen status 0.87 for women with estrogen status, and 0.88 for men (P < 0.001 for women without estrogen status vs other groups).


Estrogen status information improved the pretest clinical diagnosis of coronary artery disease in women. Using estrogen status information, the diagnostic accuracy for combined clinical and exercise test data for coronary artery disease was improved so that it was similar for men and women.

Source of funding: American Heart Association.

For article reprint: Dr. A.P. Morise, Section of Cardiology, HSC-South, West Virginia University, Morgantown, WV 26506, USA. FAX 304-293-7828.


Current estrogen use is a strong predictor of CAD in women (1). CAD accounts for one third of all deaths in women, approximately 250 000 deaths in the United States annually. Despite suboptimal representation in most studies, it is clear that women have unique clinical features. Symptoms tend to be atypical compared with those in men so that, as noted in this study, it is more difficult to estimate the pretest likelihood of major CAD. In addition, the prevalence of CAD is lower in women than men having coronary angiography for suspected CAD, with significant narrowing present in only 35% to 65% of women with typical angina and < 20% of women with atypical chest pain.

The study by Morise and colleagues addresses clinical diagnosis of CAD in women. CAD prevalence was only 13% in women who were premenopausal or taking estrogen replacement therapy compared with 58% in postmenopausal women without estrogen replacement. Thus, including estrogen status in the pretest assessment of likelihood of CAD in women greatly increases the diagnostic accuracy of treadmill testing. Although the sensitivity and specificity of the treadmill test is unchanged, improving pretest accuracy results in higher post-test accuracy for diagnosing CAD in women. Without estrogen status, exercise testing is much less useful in women than in men. A clinical strength of this study is that assessment of estrogen status was a simple "yes or no" without consideration of dose or other hormonal therapy.

Clinically it is clear that estrogen status should be included in estimates of the pretest likelihood of disease in women with suspected CAD. The sensitivity and specificity of exercise testing remain, however, suboptimal in women. Other noninvasive methods for diagnosis of CAD in women (such as exercise echocardiography (2, 3) or dobutamine stress-echocardiography) have been shown to improve diagnostic accuracy. Further studies of the effect of estrogen replacement therapy on development of CAD are currently underway (4, 5).

Catherine M. Otto, MD
University of WashingtonSeattle, Washington, USA


1. Stampfer MJ, Colditz GA, Willett, WC, et al. Postmenopausal estrogen therapy and cardiovascular disease. Ten-year follow-up from the Nurses' Health Study. N Engl J Med. 1991;325:756-62.

2. Heupler S, Mehta R, Lobo A, Leung D, Marwick TH. Prognostic implications of exercise echocardiography in women with known or suspected coronary artery disease. J Am Coll Cardiol. 1997;30:414-20.

3. Marwick TH, Anderson T, Williams MJ, et al. Exercise echocardiography is an accurate and cost-efficient technique for detection of coronary artery disease in women. J Am Coll Cardiol. 1995;26:335-41.

4. Wenger NK. Coronary heart disease: an older woman's major health risk. BMJ. 1997;315:1085-90.

5. Design of the Women's Health Initiative clinical trial and observational study. The Women's Health Initiative Study Group. Control Clin Trials. 1998;19:61-109.