Why Standard Stress Tests Miss Women's Heart Disease

A normal stress test can be a true result to the wrong question. The standard exercise ECG looks for flow-limiting obstruction in major coronary arteries. It does not look for microvascular disease, coronary spasm, or early non-obstructive atherosclerosis. Many women with real ischemia have exactly those things. The diagnostic failure is not in the test itself, which does what it was designed to do; it is in applying a test built for one pattern of disease to a population in which a different pattern is common.

The Mechanism

The standard treadmill stress test works by increasing myocardial oxygen demand through exercise and looking for ST-segment changes on the ECG that indicate the heart muscle is not getting enough blood. When myocardial demand outstrips supply, the most energy-starved layer of the heart, the subendocardium, generates electrical changes that propagate to the skin surface where electrodes can record them. Those ST-segment depressions occur reliably when a major coronary artery has a flow-limiting stenosis of at least 70%, because increased demand cannot be met when supply is physically obstructed upstream.

The test was developed, validated, and refined over decades using populations in which obstructive coronary artery disease was the predominant pathology. The original landmark trials establishing the exercise ECG’s diagnostic performance included predominantly male subjects. The sensitivity and specificity values most physicians use when interpreting a stress test result derive from those populations.

In women, obstructive disease is less common relative to coronary microvascular dysfunction. The myocardium is still ischemic when microvascular dysfunction is the mechanism; the cause is different. The resistance vessels, the arterioles and capillaries that regulate blood flow at the tissue level, fail to dilate appropriately under conditions of increased demand. They may also constrict actively. The epicardial coronary arteries may look unobstructed on angiography while the tissue downstream is underperfused. The exercise ECG has no reliable mechanism for detecting microvascular failure. The large arteries feeding the signal are open; there is no focal obstruction to produce the characteristic subendocardial pattern the test is calibrated to find.

This is not a subtle limitation. It means a substantial category of ischemic heart disease in women produces a normal result on the test most commonly used to evaluate cardiac chest pain.

What the Evidence Shows

The diagnostic accuracy problem in women has been documented across multiple large datasets. The WISE study (Women’s Ischemia Syndrome Evaluation), an NHLBI-sponsored prospective cohort, enrolled women referred for cardiac catheterization for suspected ischemic heart disease. Fewer than half of these women had obstructive coronary artery disease as the catheterization finding. A large proportion had non-obstructive disease with objective evidence of microvascular ischemia. For that group, a standard exercise ECG would have been insensitive not because of technical failure but because the disease being missed operates through a mechanism the test cannot detect.

The WISE-CVD substudy (Bairey Merz and colleagues, JAMA Internal Medicine, 2016) quantified coronary flow reserve in women with non-obstructive coronary arteries and symptoms. Approximately 50% had reduced coronary flow reserve below 2.5, indicating significant microvascular dysfunction. These women had worse cardiovascular outcomes at follow-up than those with preserved coronary flow reserve. They would not have been identified by standard stress testing.

The accuracy problem is also bidirectional. Meta-analyses examining exercise ECG in women, including pooled data analyzed by Kwok and colleagues, have documented sensitivity for obstructive coronary artery disease in the range of 61 to 65% in women, compared with approximately 72 to 77% in men. Specificity is also lower in women, with false-positive rates substantially higher due to resting ST-T wave abnormalities more common in women from hormonal and autonomic factors. A 2020 systematic review in JACC examining sex-specific performance of non-invasive testing confirmed these patterns across exercise ECG, nuclear stress testing, and stress echocardiography.

The ISCHEMIA trial (Maron et al., NEJM, 2020), which enrolled patients with moderate to severe ischemia on standard stress testing and followed them under invasive versus conservative management, found that a substantial proportion of patients in the invasive arm had no obstructive disease that warranted revascularization when the catheterization was performed. Among women enrolled in ISCHEMIA, the discordance between stress test result and catheterization findings was even more pronounced than in men. The test said ischemia; the catheterization did not always find the expected anatomy.

These datasets converge on the same conclusion: the standard exercise ECG systematically underperforms in the population most likely to have microvascular ischemia, and also generates more false-positive signals from baseline ECG differences. The net effect is that women are more likely to have real ischemia missed and also more likely to be told they have a problem when they do not.

Why Baseline ECG Differences Matter

Women have higher resting rates of ST-T wave changes, including ST segment flattening and T-wave inversions, compared with men, independent of structural heart disease. These differences arise from hormonal influences on repolarization, smaller cardiac dimensions, and autonomic differences. When the resting ECG already shows ST-T changes, exercise-induced changes become harder to interpret: any further deviation from baseline is ambiguous, and the false-positive rate rises accordingly.

This matters clinically because a positive result on a standard stress test in a woman with resting ST-T abnormalities has substantially lower positive predictive value than the same result in a man with a normal resting ECG. A positive stress test in that context should lead to further evaluation, not directly to an intervention assumption. Yet practice patterns do not always reflect this; the positive result can carry inappropriate weight.

What Imaging Modalities Add

Stress echocardiography combines exercise or pharmacological stress with ultrasound imaging of wall motion. It detects regional dysfunction that occurs when territory supplied by an obstructed artery becomes ischemic. It is more specific than exercise ECG alone and less prone to the false-positive problem from baseline ECG abnormalities, but it still primarily detects regional ischemia from focal obstruction rather than diffuse microvascular disease. A normal stress echo in a woman with microvascular dysfunction is still a test that answered the wrong question.

Cardiac MRI with stress perfusion uses a vasodilator stress agent and gadolinium contrast to map blood flow through the myocardium directly. It can detect subendocardial ischemia and reduced myocardial perfusion reserve at the tissue level, including in microvascular disease where there is no focal wall motion abnormality. The MR-INFORM trial (Richards et al., NEJM, 2019) demonstrated that CMR perfusion imaging performs as well as invasive angiography-guided management for guiding treatment decisions in stable coronary disease in a trial of approximately 918 patients. The resolution of stress perfusion CMR extends to the microvasculature in a way that neither exercise ECG nor stress echo can match.

PET myocardial perfusion imaging with flow quantification measures absolute coronary blood flow and coronary flow reserve, providing a quantitative measure of the microvascular response to stress. It is the most direct noninvasive test for microvascular dysfunction currently available. A coronary flow reserve below 2.0 on PET has been associated with adverse outcomes in multiple cohorts. Work from Taqueti and colleagues at Harvard (published in Circulation and JACC) has specifically linked reduced PET coronary flow reserve in women with non-obstructive coronary arteries to adverse cardiovascular events including heart failure hospitalization. PET flow quantification is particularly suited to cases where microvascular coronary disease is the primary diagnostic hypothesis.

Invasive physiologic testing during catheterization, measuring coronary flow reserve (CFR) and the index of microcirculatory resistance (IMR) with a pressure-temperature sensing wire, is the gold standard for microvascular assessment. The CorMicA trial (Ford et al., JACC, 2018) showed that catheterization-based endotype stratification followed by endotype-specific therapy improved angina burden and quality of life at six months compared with standard care. This is the most direct evidence that identifying the right mechanism changes outcomes.

Who Benefits from Advanced Stress Imaging

Guidelines from the AHA and ESC identify stress imaging as appropriate for intermediate pre-test probability of coronary artery disease, for patients with baseline ECG abnormalities, and for those with ongoing symptoms after a normal exercise ECG. The 2021 ESC guidelines on cardiovascular disease prevention explicitly address the sex-specific limitations of the exercise ECG and recommend imaging-based testing in women with a presentation suggesting microvascular disease.

For women with suspected microvascular disease, PET with flow quantification or stress CMR offers more specific information than exercise ECG. These are not interchangeable options; the choice depends on what question is being asked and what is available at the referring center.

The Duke Treadmill Score, which incorporates ST changes, exercise capacity, and chest pain during testing, partially addresses the sensitivity limitation by using multiple variables, but it was also developed in predominantly male cohorts. Applying it to women with microvascular disease does not resolve the fundamental problem that the test is looking for obstructive disease in a population whose disease is predominantly microvascular.

Provocative Vasospasm Testing: When the Exercise Test Tests the Wrong Mechanism

For a subset of women with chest pain and normal or near-normal coronary anatomy, the exercise ECG fails to provide useful diagnostic information not because of its sensitivity limitations for obstructive disease but because the underlying mechanism is coronary vasospasm — a fundamentally different process that exercise does not reliably provoke. Understanding when vasospasm is the likely diagnosis changes which testing strategy is appropriate.

Coronary vasospasm involves abrupt, transient, reversible constriction of an epicardial coronary artery in the absence of fixed obstruction. It typically produces ST elevation rather than the ST depression associated with subendocardial ischemia from fixed obstruction, and is triggered by cold exposure, emotional stress, smoking, or pharmacological agents such as acetylcholine — rather than by the exertion that drives demand-related ischemia. An exercise stress test increases catecholamines and coronary blood flow demand, but the net vasomotor effect of beta-adrenergic stimulation during exercise is vasodilatory. For vasospasm-prone arteries, exercise may paradoxically suppress, rather than provoke, the critical event. Most women with vasospastic angina have completely normal exercise ECGs.

The WISE study found that approximately 20 to 30 percent of women with INOCA and non-obstructive anatomy on angiography had documented vasospasm on provocative acetylcholine testing during catheterization. In this group, the conventional exercise ECG not only failed to diagnose the problem — it generated false reassurance that contributed to diagnostic delays of years in many patients. 4 / Promising

Intracoronary acetylcholine provocation during catheterization is the diagnostic gold standard. Acetylcholine normally produces vasodilation through endothelial nitric oxide release; in vessels with endothelial dysfunction and vasospastic tendency, it paradoxically induces vasoconstriction and reproduces the patient’s symptoms with simultaneous ECG changes. The COVADIS criteria (Beltrame, European Heart Journal 2017) define vasospastic angina based on provocation testing results and clinical symptom pattern.

For women with chest pain that is predominantly at rest, nocturnal, cold-triggered, or reproducibly associated with emotional rather than physical stress, and who have had a normal exercise ECG, vasospasm should be in the diagnostic differential. Ambulatory Holter monitoring over 48 to 72 hours, looking for spontaneous ST changes during symptomatic episodes, is a less invasive initial step. A normal exercise ECG in this clinical context is not reassurance — it is an absence of testing for the relevant mechanism.

What to Do This Week

1. If you have had a normal stress test but continue to have exertional symptoms, ask specifically whether microvascular coronary disease was in the differential and whether a different stress imaging modality, such as stress CMR or stress PET, might answer the question more precisely. A normal exercise ECG does not make these tests redundant.

2. If you are being referred for stress testing, ask whether the test ordered is appropriate for your risk profile and symptom pattern, not simply the standard first step. Symptoms that are rest-predominant, nocturnal, or cold-triggered are more consistent with vasospasm than with exertional demand-related ischemia; the exercise ECG is not the right tool for that clinical question.

3. Know that a normal exercise ECG does not mean your heart is fine in all possible respects. It means obstructive large-vessel disease is less likely. Microvascular ischemia, vasospasm, and early non-obstructive atherosclerosis are not excluded.

4. If your baseline resting ECG has ST-T wave abnormalities and a stress test was interpreted as positive, ask whether the result should be clarified with imaging before any further decisions are made based on that result.

5. Bring documentation of your symptom pattern to every cardiology appointment: when symptoms occur, what triggers them, how long they last, and whether rest or nitroglycerin relieves them. This information changes the probability estimate going into any stress test and determines which test is most informative.

The standard exercise ECG is a useful test for the disease it detects. The clinical problem is the assumption that it answers the question of whether a woman has significant ischemic heart disease, which it does not, particularly when the more likely pathology is microvascular. A normal result warrants interpretation in context, not automatic reassurance. The test answered one question. It may not have been the right one.