American Heart Journal
Volume 156, Issue 5 , Pages 964.e1-964.e5, November 2008

How to improve noninvasive coronary artery disease diagnostics in premenopausal women?

The influence of menstrual cycle on ST depression, left ventricle contractility, and chest pain observed during exercise echocardiography in women with angina and normal coronary angiogram

First Department of Cardiology, Medical University of Gdansk, Gdansk, Poland

Received 22 November 2007; accepted 16 August 2008.

Article Outline

Background

The aim was to assess the influence of menstrual cycle on results of exercise echocardiography and electrocardiography.

Methods

Premenopausal women (n = 28) with regular monthly menses, presented typical angina, positive electrocardiogram (ECG) exercise stress test, and normal coronary angiogram were recruited. Exercise supine bicycle echocardiography with simultaneous recording of 12-lead ECG was performed once a week for 4 consecutive weeks. Occurrence of angina, time to angina, time to significant ST deviation, and segmental myocardial contractility were analyzed. Blood samples were drawn to estimate follicle-stimulating hormone, luteinizing hormone, β-estradiol, progesterone concentration and confirm the position in menstrual cycle.

In correlation analysis, linear and logistic regression were used as appropriate. Qualitative variables were categorized into quartiles in logistic regression analysis.

Results

Exercise ST depression was more frequently observed in both luteal phases (early luteal 78%, late luteal 86%) compared to the late follicular phase (50%, P < .05). Time to ST depression was significantly longer in late follicular phase compared to other phases. The rate of segmental exercise left ventricular hypokinesis was low and not significantly related to menstrual cycle. Using linear regression, significant positive correlation was found between estradiol-progesterone ratio and time to ST depression. Using multiple logistic regression, we confirmed that progesterone level is independent factor influencing the presence of ST depression.

Conclusion

In women with typical angina and normal coronary angiogram, the position in menstrual cycle influences the ST depression but not myocardial contractility during exercise echocardiography.

 

Women with typical chest pain suggestive of myocardial ischemia represent a diagnostic challenge. According to American College of Cardiology-National Cardiovascular Data Registry,1 half of all women with chest pain undergoing coronary angiography do not have coronary artery disease (CAD). Premenopausal women are at relatively low risk of CAD. Pretest likelihood of CAD in women aged 40 to 49 years with typical chest pain is 55%.2, 3 The rate of false-positive electrocardiographic (ECG) stress test results in women with angina is high. In a published meta-analysis4 that included 19 exercise ECG studies with 3,721 women, specificity was 70%. On the basis of the aggregate data available in studies of nearly 1,000 women with suspected CAD, stress echocardiography has demonstrated better diagnostic accuracy for detecting or ruling out significant CAD, with mean sensitivity of 81% (89% in women with multivessel disease), specificity of 86%, and overall accuracy of 84%.5, 6 There is a lack of information on the relation between the menstrual cycle and results of ECG and echocardiographic exercise stress tests in premenopausal women with angina and without significant stenosis in coronary arteries.

Back to Article Outline

Methods 

We examined 28 premenopausal women, mean age 44 ± 5 years, with regular monthly menses, symptomatic angina with a positive exercise test, and without significant CAD (<30% diameter stenosis) in coronary angiography. Only women with menstrual cycle duration of 28 ± 3 days were included.

Women taking oral contraceptives, digoxin, with left ventricular hypertrophy, and ST-segment abnormalities in resting ECG (ie, left bundle branch block) were excluded. We performed exercise stress echocardiography during supine bicycle exercise with simultaneous recording of 12-lead ECG. The tests were repeated at the same time of day in the same ambient conditions on 4 consecutive weeks. Treatment was left unchanged throughout the study. Our patients avoided occasional use of other medications during the whole study period (including nonsteroidal antiinflammatory drugs, diuretics, and proton pump inhibitors).

All 4 exercise tests were performed within 2 months after coronary angiography. Women randomly started the trial protocol irrespective of their position in the menstrual cycle. Exercise testing was performed using Esaote Biomedica SpA (Genova, Italy) formula equipment with automatic signal averaging of ST segments. We used 25 W/2 min protocol to achieve 85% of maximal heart rate or chest pain. ST-segment depression was considered to be significant if there was at least 1 mm of planar or downsloping ST depression in any of 2 consecutive leads lasting for ≥10 beats. Exercise echocardiography was performed using the Image Point HX Agilent echocardiograph (Agilent Technologies, Santa Clara, CA) equipped with tissue harmonic imaging, side-by-side display of cineloops of acquired left ventricle images and digital acquisition and storage. The exercise echocardiography was considered positive when new or worsening of preexisting wall motion abnormality was observed. Two experienced echocardiographers who were unaware of the menstrual and symptomatic status of the subjects interpreted the stress echocardiography. Blood pressure was measured at 2-minute intervals. Before each exercise test, blood samples were drawn for β-estradiol, progesterone, luteinizing hormone, and follicle-stimulating hormone to confirm the position in menstrual cycle (early follicular, late follicular, early luteal, or late luteal phase). The position in menstrual cycle was established based on history of menstruations and confirmed by the levels of aforementioned hormones. In case of any doubts about position in menstrual cycle, the women were excluded from further analysis. The study was approved by the Local Ethics Committee of Medical University of Gdansk (Gdansk, Poland). Informed consent was obtained in all women.

Statistics 

Data were analyzed using the StatSoft Statistica 7.1 statistical package (Statsoft Inc., Tulsa, OK). Two-way comparisons were performed using paired or unpaired t tests as appropriate. Group trends were assessed using analysis of variance for repeated measures. All data fulfilled parametric assumptions as defined by Levine's test. For qualitative variables, linear regression correlation was performed. For quantitative variables, logistic regression analysis was performed. To assess the relation between qualitative variables (ie, hormonal levels) and presence of ST depression by logistic regression, the variables were categorized into quartiles. Probability values of P < .05 were considered significant.

Back to Article Outline

Results 

Thirty-two women were initially recruited. Three women were not qualified to further analysis because of postmenopausal range of hormone concentration and inability in defining the position in menstrual cycle. One woman was excluded due to the poor quality of echocardiographic projections that did not allow valuable analysis. The 28 investigated women were aged 44.3 ± 4.8 years. Seven subjects entered the study during the early follicular phase, 8 in the late follicular phase, 6 in the early luteal phase, and 5 in the late luteal phase. Characteristic of the study group is shown in Table I.

Table I. Clinical characteristic of the investigated group
ParameterValue
No. of investigated subjects28
Mean age44.3 ± 4.8 y
Mean menstrual cycle duration27.7 ± 1.5 d
Child deliveries100%
Mean number of pregnancies2.2 ± 1.2
Angina intensityCCS II, 18 (64.2%)
CCS III, 10 (35.8%)
Diabetes2 (7%)
Smoking0
Hypercholesterolemia14
Hypertension2 (8%)
Mean body weight64.4 ± 5.8 kg
Body mass index24.6 ± 3.1 kg/m2
Waist circumference79.6 ± 4.6 cm
β-Blockers8 (28%)
ACEI10 (35%)
Ca blockers8 (28%)
Statins14 (50%)
Nitrates2 (7%)
Trimetazidine2 (7%)

CCS, Canadian Cardiovascular Society; ACEI, angiotensin-converting enzyme inhibitor.

We observed a significant effect of different phases of menstrual cycle on the occurrence of ST depression, time to ST depression, and time to chest pain. During the period when estrogen concentrations were the highest in late follicular phase, the occurrence of ST depression was at its lowest (50%) and was significantly lower than in early luteal (79%, P < .05) and late luteal phases (86%, P < .01). As far as exercise ST changes are concerned, it was observed that 16 of the investigated women (57%) showed variability during menstrual cycle. In 10 women, ST depression was observed during all 4 tests, in 7 women during 3 of 4 tests, in 9 during 2 tests, and in 2 women, exercise ST depression was not observed during the study.

Time to ST depression in women with ECG-positive tests was significantly longer in the late follicular phase (334.0 ± 92.7 seconds) compared to early follicular (230.1 ± 108.6 seconds), early luteal (221.0 ± 97.0 seconds), and late luteal phases (218.5 ± 108.6 seconds); P < .05. Only nonsignificant trends were observed in the maximum depth of ST-segment depression. In women with at least one ECG-positive test during the study period, the maximum depth of ST-segment depression was similar in early follicular, late follicular, early luteal, and late luteal phases as follows: 1.38 ± 0.39 mm, 1.42 ± 0.45 mm, 1.5 ± 0.69 mm, and 1.39 ± 0.62 mm, respectively. We did not observe significant ST changes at rest in the investigated group during the study.

Chest pain during exertion tended to occur more frequently in late luteal phase (75%) and early follicular phase (71%) than in late follicular (61%) and early luteal phases (54%), but the differences were not significant. Time to chest pain was longer in the late follicular phase (281.8 ± 128.2 seconds) compared to early follicular (236.7 ± 101.5 seconds), P < .05; early luteal (203.7 ± 86.0 seconds), P < .05; and late luteal phases (240.4 ± 107.7 seconds), P < .05 (Table II). As far as chest pain is concerned, 14 women (50%) showed variability during menstrual cycle. In 12 women, chest pain occurred during all 4 tests, in 3 women during 3 of 4 tests, in 5 during 2 tests, in 6 during 1 test, and in 2 women, chest pain did not occur at all.

Table II. Estradiol and progesterone levels, estradiol-progesterone ratio, segmental wall motion abnormalities, ST depression, time of exercise, time to chest pain, time to ST depression, and average ST depression during exercise stress test for the 4 phases of the menstrual cycle
Early follicularLate follicularEarly lutealLate luteal
Estradiol level (pmol/L)370.3870.1529.5446.9
Progesterone level (nmol/L)3.641.4327.431.9
Estradiol-progesterone ratio0.1 ± 0.080.61 ± 0.640.02 ± 0.0130.01 ± 0.007
Exercise hypokinesis4 (14%)4 (14%)2 (7%)2 (7%)
ST depression19 (68%)14 (50%)⁎22 (79%)24 (86%)
ECG vs echocardiogramP < .001P < .001P < .001P < .001
Exercise time (s)308.0 ± 108.6324.9 ± 123.1345.5 ± 173.7296.5 ± 102.8†
Time to chest pain (s)236.7 ± 101.5281.8 ± 128.2‡203.7 ± 86.0240.4 ± 107.7
Time to ST depression (s)230.1 ± 108.6334.0 ± 92.7§221.0 ± 97.0218.5 ± 108.6
Average ST depression|| (mm)1.45 ± 0.391.52 ± 0.451.50 ± 0.691.42 ± 0.62

⁎P < .05 as compared to early and late luteal phases.

†P < .05 as compared to late follicular and early luteal phases.

‡P < .05 as compared to early follicular, early, and late luteal phases.

§P < .05 as compared to early follicular, early, and late luteal phases.

||Average for women with at least one ECG-positive test during the study period.

Echocardiographic loops recorded at rest revealed normokinetic or hyperkinetic left ventricle in all performed tests. We did not find any valvular pathologic condition at rest and during exercise as well. The rate of segmental left ventricular hypokinesis observed during exercise echocardiography was low and not significantly related to menstrual cycle. Left ventricular hypokinesis was observed only in 4 women in early follicular phase, in the same 4 women in late follicular, in 2 (of the 4) in early luteal, and in the same 2 in late luteal phase. Segmental wall motion abnormalities were observed in 4 of 28 investigated women (14%). Only 2 of 28 investigated women (7%) presented wall motion variability during menstrual cycle.

The ST depression occurred much more frequently than segmental wall motion abnormalities in each of the menstrual phases, P < .001 (Table II). Using linear regression analysis, we analyzed the relationship between ovarian hormone levels and time to ST depression as well as maximal depth of ST depression. The results of the analysis were shown in Table III.

Table III. Correlation between level of ovarian hormones and time to ST depression and maximal depth of ST depression during exercise stress tests
Time to ST depressionMaximum depth of ST depression
Estradiol levelr = 0.11r = −0.27
P = .25P < .05
Progesterone levelr = −0.11r = 0.11
P = .21P = .53
Estradiol-progesterone ratior = 0.29r = −0.07
P < .05P = .55

r, Correlation index.

Using multiple logistic regression analysis, we confirmed that progesterone level is independent factor influencing the presence of ST depression after adjustment for exercise duration and estradiol level (odds ratio 2.23, CI 1.3-.9, P < .01).

Exercise duration in early follicular, late follicular, early luteal, and late luteal phases was 308.0 ± 108.6 seconds, 324.9 ± 123.1 seconds, 345.5 ± 173.7 seconds, and 296.5 ± 102.8 seconds, respectively. Exercise duration in late luteal phase turned out to be significantly shorter compared to late follicular, P < .05, and early luteal phases, P < .05.

Back to Article Outline

Discussion 

It has been established that estrogen and other female hormones have important effects on many different organs, including cardiovascular system.7, 8 The net result of estrogen action is to cause vasodilatation in both the coronary and the peripheral arteries, although this is antagonized to some extent by the progestin or progesterone.9 Various mechanisms have been proposed for the estrogen-induced vasodilatation, including restoration of impaired vasodilatator function, calcium antagonism, alterations in sympathetic tone, and converting enzyme inhibition.10, 11, 12, 13

A greater prevalence of false-positive, exercise-induced ST-segment depression in women compared with men has been described, and gender-related hormones have been implicated empirically.14 Wu et al15 reported that false-positive response was significantly higher in young women than in young men, whereas the reverse was the case in older age groups. The causative role of estrogens is supported by the lower prevalence of false-positive responses after 45 years of age, when there is a decrease in plasma concentration of these hormones with an increase in prevalence of coronary disease. Clark et al16 suggested that progesterone fluxes are implicated in the ECG abnormalities observed in women. It is also possible that both estradiol and progesterone as well as their ratio is important in the development of false-positive ECG stress tests.

The investigated group in our study was composed of women with normal coronary angiogram so each occurrence of ST depression, wall motion abnormalities, or chest pain during exertion has to be treated as false-positive result. The present study showed direct relation between menstrual phase and occurrence of ST-segment depression, time to ST depression, and time to chest pain during supine exercise bicycle echocardiography in premenopausal women with angina and normal coronary angiogram. During the late follicular phase (maximal β-estradiol concentration and minimal progesterone concentration), the occurrence of ST depression was the lowest and time to ST depression and time to chest pain were the longest. Perhaps premenopausal women with angina should be referred to ECG exercise tests during their late follicular phase to increase test specificity.

The chemical structure of β-estradiol is similar to digitalis, which is known to cause false-positive ST-segment changes. Our results did not confirm the occurrence of “digoxin-like effect” of β-estradiol. In late follicular phase (maximal β-estradiol concentration), the frequency of exercise ST depression was the lowest and time to ST depression was the longest.

In our study, the exercise wall motion abnormalities were not related to menstrual phases. Low frequency of exercise hypokinesis during each of the phases may be the reason why differences did not reach significance. During each of the phases, the frequency of exercise hypokinesis was significantly lower compared to the frequency of ST-segment depression. The results seem to indirectly confirm that exercise echocardiography provides higher specificity than does standard exercise ECG testing in premenopausal women.5

The low frequency of exercise wall motion abnormalities during exercise tests in our study may suggest that the reason of chest pain and ST depression in most women may not be ischemia. To prove or rule out the occurrence of ischemia, the other more ischemia-specific tests (ie, phosphorus-31 nuclear magnetic resonance spectroscopy) should be performed. Buchthal et al17 reported that about 20% of women with chest pain but without obstructive CAD had stress-induced reduction in myocardial phosphocreatine-adenosine triphosphate ratio by phosphorus-31 nuclear magnetic resonance spectroscopy, consistent with myocardial ischemia.

The alternative explanation for time to chest pain differences across the menstrual cycle may be the influence of ovarian hormones on pain sensitivity. Estrogens in particular have been linked to pain processing through modulatory effects on γ-aminobutyric acid receptors, μ-opioid receptors, and nerve growth factor receptors in the dorsal root ganglion.18 Many studies demonstrated a link between higher level of estrogen and decreased pain sensitivity. In the 14 human studies summarized by Fillingim and Ness,19 it was shown that women are less sensitive to pain before than after menstruation.

It is known that estrogens affect ST-segment response to exercise in healthy women, through a nonischemic mechanisms.20 However, the precise mechanisms of ST-segment changes during exercise in women without obstructive CAD are still unknown. There are data21, 22, 23, 24, 25, 26, 27, 28 suggesting that estradiol can modulate expression of cardiac calcium and potassium currents and influence both early and late repolarizations. Estrogen receptors were found in atrial and ventricular myocardial cells.29 That is why another possible explanation for ST depression variability across the menstrual cycle may be the influence of estrogens on ventricle myocardial cells directly or by estrogen receptors affecting the repolarization phase.

Correlation analyses, described in result section, suggest that both estradiol and progesterone as well as the estradiol-progesterone ratio may influence the exercise ST response. Special notice should be taken to progesterone level that seems to be independent factor influencing the occurrence of ST depression.

In our study, the exercise duration in late luteal phase was significantly shorter compared to late follicular and early luteal phases. However, it should be mentioned here that the tests were arbitrarily stopped upon occurrence of chest pain so the reason of the difference may be more frequent occurrence of chest pain in late luteal phase.

Back to Article Outline

Conclusion 

In women with typical angina, without significant stenosis in coronary arteries, exercise ST depression was more easily induced in early luteal and late luteal phases than in late follicular phase. In the late follicular phase, estradiol concentration was the highest and progesterone concentration the lowest. In our study, progesterone level independently influenced occurrence of exercise ST depression. Time to angina and time to ST depression were the longest in the late follicular phase. Significant positive correlation was found between estradiol-progesterone ratio and time to ST depression. These observations may be important for timing the ECG stress test in women with chest pain. During exercise echocardiography in each of the menstrual phases, ST depression occurred more frequently than segmental wall motion abnormalities. The results of the study seem to confirm that in premenopausal women the rate of false-positive results of ECG stress test is significantly higher compared to exercise stress echocardiography.

Back to Article Outline

References 

  1. Shaw RE, Anderson HV, Brindis RG, et al. Development of a risk adjustment mortality model using the American College of Cardiology-National Cardiovascular Data Registry (ACC-NCDR) experience: 1998-2000. J Am Coll Cardiol. 2002;39:1104–1112
  2. Diamond GA, Forrester JS. Analysis of probability as an aid in the clinical diagnosis of coronary artery disease. N Engl J Med. 1979;300:1350–1358
  3. Chaitman BR, Bourassa MG, Davis K, et al. Angiographic prevalence of high-risk coronary artery disease in patient subsets (CASS). Circulation. 1981;64:360–367
  4. Kwok Y, Kim C, Grady D, et al. Meta-analysis of exercise testing to detect coronary artery disease in women. Am J Cardiol. 1999;83:660–666
  5. Mieres JH, Shaw LJ, Arai A, et al. Role of noninvasive testing in the clinical evaluation of women with suspected coronary artery disease: Consensus Statement From the Cardiac Imaging Committee, Council on Clinical Cardiology, and the Cardiovascular Imaging and Intervention Committee, Council on Cardiovascular Radiology and Intervention, American Heart Association. Circulation. 2005;111:682–696
  6. Roger VL, Pellikka PA, Bell MR, et al. Sex and test verification bias: impact of the diagnostic value of exercise echocardiography. Circulation. 1997;95:405–410
  7. Collins P, Rosano GM, Sarrel PM, et al. 17-B-estradiol attenuates acetylcholine-induced coronary arterial constriction in women but not in men with coronary heart disease. Circulation. 1995;92:24–28
  8. Liebermann E, Gerhard M, Uehata A, et al. Estrogen improves endothelium-dependent flow mediated vasodilatation in postmenopausal women. Ann Intern Med. 1994;121:936–941
  9. Rosano GM, Sarrel PM, Chierchia SL. Medoxyprogesteron but not natural progesterone reverses the beneficial effect of estradiol 17-beta on exercise induced myocardial ischaemia. Circullation. 1996;91:1–18
  10. Collins P, Shay J, Jiang C, et al. Nitric oxide accounts for dose dependent oestrogen-mediated coronary relaxation following acute estrogen withdrawal. Circulation. 1994;90:1964–1968
  11. Zhang F, Ram JL, Standley PL, et al. 17-B oestradiol attenuates voltage-dependent Ca currents in A7r5 vascular smooth muscle cell line. Am J Phisiol. 1994;266:975–980
  12. Rosano G, Patrizi R, Leonardo F, et al. Effects of estrogen replacement therapy on heart rate variability and heart rate in healthy postmenopausal women. Am J Cardiol. 1997;80:815–817
  13. Proudler A, Ahmed A, Crook D, et al. Hormone replacement therapy and serum angiotensin converting enzyme activity in postmenopausal women. Lancet. 1993;342:89–90
  14. Morise AP, Diamond GA. Comparison of the sensitivity and specificity of exercise electrocardiography in biased and unbiased populations of men and women. Am Heart J. 1995;130:741–747
  15. Wu SC, Secchi MB, Radice M, et al. Sex differences in the prevalence of ischemic heart disease and in the response to a stress test in working population. Eur Heart J. 1981;2:461–465
  16. Clark PI, Glasser SP, Lyman GH, et al. Relation of results of exercise stress test in young women to phases of the menstrual cycle. Am J Cardiol. 1988;61:197–199
  17. Buchthal SD, den Hollander JA, Bairey Merz CNB, et al. Abnormal myocardial phosphorus-31 nuclear magnetic resonance spectroscopy in women with chest pain but normal coronary angiograms. N Engl J Med. 2000;342:829–835
  18. Aloisi AM. Gonadal hormones and sex differences in pain reactivity. Clin J Pain. 2003;19:168–174
  19. Fillingim RB, Ness TJ. Sex-related hormonal influences on pain and analgesic responses. Neurosci Biobehav Rev. 2000;24:485–501
  20. Henzlova MJ, Croft LB, Diamond JA. Effect of hormone replacement therapy on the electrocardiographic response to exercise. J Nucl Cardiol. 2002;9:385–387
  21. Merri M, Benhorin J, Alberti M, et al. Electrocardiographic quantitation of ventricular repolarization. Circulation. 1989;80:1301–1308
  22. Rautaharju PM, Zhou SH, Wong S, et al. Sex differences in the in the cardiovascular evolution of the electrocardiographic QT interval with age. Can J Cardiol. 1992;8:690–695
  23. Yang H, Elko P, Fromm BS, et al. Maximal ascending and descending slopes of the T wave in men and women. J Electrocardiol. 1997;30:267–276
  24. Pham TV, Rosen MR. Sex, hormones, and repolarisation. Cardiovasc Res. 2002;53:740–751
  25. Harder DR, Coulson PB. Estrogen receptors and effects of estrogen on membrane electrical properties of coronary vascular smooth muscle. J Cell Physiol. 1979;100:375–382
  26. Boyle MB, Azhderian EM, MacLusky NJ, et al. Xenopus oocytes injected with rat uterine RNA express very slowly activating potassium currents. Science. 1987;235:1221–1224
  27. Pragnell M, Snay KJ, Trimmer JS, et al. Estrogen induction of a small, putative K channel mRNA in rat uterus. Neuron. 1990;4:807–812
  28. Ishii K, Kano T, Ando J. Sex differences in [3H]nitrendipine binding and effects of sex steroid hormones in rat cardiac and cerebral membranes. Jpn J Pharmacol. 1988;46:117–125
  29. Grohe C, Kahlert S, Lobbert K, et al. Cardiac myocytes and fibroblasts contain functional estrogen receptors. FEBS Lett. 1997;416:107–112

PII: S0002-8703(08)00729-1

doi:10.1016/j.ahj.2008.08.014

American Heart Journal
Volume 156, Issue 5 , Pages 964.e1-964.e5, November 2008

Article Tools