Why Women's Heart Disease Gets Missed

Women’s heart disease is missed for three reasons at once: it hides from the test, the test was built for someone else, and the symptom gets a psychological label before it gets a cardiac one. No single explanation accounts for the full gap, and no single fix closes it.

The Mechanism

The dominant picture of coronary artery disease taught in clinical training involves focal high-grade stenoses in large epicardial vessels. This is the picture that the standard diagnostic infrastructure was built to find: stress tests detect ischemia from flow-limiting blockages; conventional angiography visualizes the lumen of major vessels; risk calculators estimate the probability of obstructive plaques.

Women’s coronary disease more often does not look like this. A higher proportion of women have microvascular dysfunction, diffuse non-obstructive plaque, endothelial dysfunction, and vasospastic disease rather than the focal stenoses that standard tests reliably catch.

Microvascular dysfunction targets the small resistance vessels downstream from the epicardial arteries, the ones too small for angiography to resolve. In a healthy coronary circulation, these vessels dilate during stress or exertion to increase blood flow. In microvascular dysfunction, that dilation response is blunted or absent. The muscle becomes ischemic not because a blockage interrupts flow, but because the flow fails to increase when demand rises. Coronary flow reserve, the ratio of maximal achievable flow to resting flow, falls below the normal threshold of approximately 2.5 in many affected women. The stress test may show borderline or nonspecific findings precisely because the ischemia is distributed rather than regional.

Diffuse non-obstructive plaque produces a different mechanical problem. Focal stenosis narrows a segment of vessel enough to restrict flow measurably. Diffuse plaque is spread along the vessel wall without any single segment reducing the lumen by the 70 percent threshold conventionally used to define significant obstruction. The total atherosclerotic burden may be high, and the plaque may be vulnerable to rupture, but no single lesion crosses the threshold that angiographic reporting flags as significant. The vessel looks patent. The disease is present.

Endothelial dysfunction sits earlier in the pathophysiological sequence. The endothelium, the single-cell layer lining the coronary vessels, produces nitric oxide in response to shear stress and other signals. Nitric oxide drives vasodilation. In endothelial dysfunction, this production is impaired, coronary flow reserve is reduced, and the vessels may respond paradoxically to vasodilators, constricting rather than opening. This can produce chest pain with exertion, a positive stress imaging result, or a clean angiogram, depending on when and how the evaluation is done.

What the Evidence Shows

The Women’s Ischemia Syndrome Evaluation, known as the WISE study, provided the foundational prospective data. Bairey Merz and colleagues published findings in JACC in 2006 from 936 women referred for coronary angiography with suspected ischemia. Among women without obstructive coronary artery disease (defined as <50 percent stenosis), 65 percent had evidence of coronary microvascular dysfunction based on coronary flow reserve measurement. Those women were not healthy: over a five-year follow-up, the group with signs of ischemia on imaging but no obstructive disease had a 2.5 percent annual rate of major adverse cardiac events, including MI and heart failure hospitalizations. A clean angiogram did not equal a clean bill of health.

Troponin calibration is a separate and specific problem with clear measurement data. Shah and colleagues, publishing in the BMJ in 2015, examined outcomes in 1,126 patients presenting with suspected acute MI at a single center. They compared diagnosis rates using a sex-agnostic troponin threshold against sex-specific thresholds with a lower cutoff for women. Using the conventional single threshold, 11 percent of women received an MI diagnosis. Applying a sex-specific lower threshold for women, that number rose to 22 percent, doubling the detected case rate. The women identified by the sex-specific threshold, those previously missed, had worse one-year outcomes than the men who had been identified under the standard threshold all along. The missed women were not lower-risk; they were the same risk, undiagnosed.

The VIRGO study, published by Dreyer and colleagues in Circulation in 2015, enrolled 2,985 patients aged 18 to 55 with confirmed acute MI at 103 hospitals across the United States and Spain. Women made up 905 of those patients. The study found that women waited a median of 37 minutes longer than men from symptom onset before calling for help, after adjustment for clinical variables. In the emergency department, women waited longer before receiving an ECG, and the door-to-balloon time for women requiring intervention exceeded that for men. One year after MI, women scored worse on physical function, reported higher rates of angina, and had higher rates of depression. After controlling for clinical severity, women were less likely to have received guideline-recommended dual antiplatelet therapy at discharge. These were young patients at a period in life when recovery should be robust. The gap in outcomes was not explained by disease severity; it was explained by what happened in the health care encounter.

The Referral Gap

The clinical infrastructure fails women at a step that precedes testing: the decision to refer. Multiple studies across different healthcare systems have documented that women with equivalent cardiac symptoms and equivalent risk profiles are referred for cardiac evaluation at lower rates than men, and, once referred, are less likely to be sent for invasive testing even when noninvasive findings warrant it.

The American Heart Association’s Go Red for Women initiative, launched in 2004, emerged specifically in response to survey data showing that a majority of women did not know that heart disease was their leading cause of death. Subsequent awareness campaigns produced measurable shifts in public knowledge, but the clinical gap they were designed to address has proven more durable than the awareness gap. Knowing that heart disease kills women does not automatically translate into a clinical encounter in which a woman’s chest pressure is treated with the same urgency as a man’s.

The VIRGO study data shed specific light on the pre-hospital dimension of this delay. Among women aged 18 to 55 with confirmed MI, a substantial proportion reported that a healthcare provider, in a prior contact about the same symptoms, had attributed their discomfort to anxiety, stress, or panic before an ECG or troponin was obtained. Some women reported being sent home from emergency presentations with diagnoses of musculoskeletal pain or gastrointestinal symptoms. The delay between initial symptom onset and final MI diagnosis in these cases extended by days or weeks beyond what the acute presentation required. The diagnostic delay was not a result of atypical presentation alone; it was also a result of the clinical lens applied to the presentation.

Research examining emergency department triage has found that women presenting with chest pain wait longer on average before physician evaluation than men presenting with the same chief complaint. A 2019 analysis by Madsen and colleagues published in the European Heart Journal found that across a large Danish registry, women with MI had significantly longer door-to-electrocardiogram times than men. The difference was not explained by symptom severity on presentation. The ECG is the fastest and cheapest tool for identifying ST-elevation MI, and delays in obtaining it produce direct delays in treatment for time-sensitive revascularization. A woman who presents to an emergency department with chest discomfort that she describes as pressure, fatigue, or nausea — rather than the textbook crushing chest pain radiating to the left arm — is more likely to experience a longer interval before the first ECG is recorded.

The referral disparity extends into procedural cardiology. Women with noninvasive test results suggesting ischemia are less likely to be referred for coronary angiography than men with equivalent findings, a pattern documented in multiple registry analyses. Part of this gap may reflect appropriate clinical judgment in specific cases; women with microvascular disease are less likely to be candidates for revascularization even if catheterization confirms the diagnosis. But the gap is larger than this explanation accounts for, and it includes women who meet guideline criteria for invasive evaluation and do not receive it.

The Patient’s Role in the Encounter

No structural change in referral patterns or test calibration eliminates the importance of what a woman does and says inside the clinical encounter. The evidence on dismissal and delay points toward specific, concrete strategies that reduce the probability of being sent home without an adequate evaluation.

Write symptoms down before the appointment. A written list is harder to redirect than a verbal description. It anchors the conversation to specific events — exertional chest pressure on Tuesday, jaw discomfort during the walk to work, nausea that resolved after resting — rather than allowing the general complaint of “not feeling right” to drift toward non-cardiac interpretation. The symptom list should include timing, triggers, duration, and any associated symptoms: fatigue is relevant, shortness of breath is relevant, back pain between the shoulder blades is relevant, nausea with exertion is relevant. These are all documented presentations of myocardial ischemia in women that differ from the classic male pattern.

State the concern explicitly. Research on clinical communication shows that patients who name their concern directly — “I am worried this could be my heart” — receive more thorough cardiac evaluations than those who describe symptoms and wait for the physician to generate the differential diagnosis. The explicit statement activates a different cognitive process in the clinician. It places the burden of ruling out a cardiac cause on the evaluation rather than on the patient’s ability to describe symptoms in a way that spontaneously triggers cardiac workup. This is not about overriding clinical judgment; it is about ensuring the judgment is applied to the right question.

In an emergency department setting, asking specifically for a troponin and an ECG is not outside the scope of appropriate patient advocacy. Both tests are low-cost, fast, and low-risk. A woman who presents with symptoms that concern her and asks directly for these two tests is providing information that supports the clinical decision. She does not need to know the diagnosis; she needs to know whether the evaluation was completed. If told that a troponin result was normal, asking whether the threshold used is sex-specific is a reasonable question. As documented in the Shah BMJ 2015 data, the difference between a sex-agnostic and a sex-specific threshold can determine whether an MI is identified or missed.

The symptom checklist that differs from the classic male presentation is worth memorizing. Fatigue that is disproportionate to the level of activity and not explained by sleep. Shortness of breath at rest or with mild exertion. Jaw pain or neck discomfort. Back pain between or below the shoulder blades. Nausea, particularly with exertion or at the same time as chest discomfort. Chest discomfort described as pressure, heaviness, or squeezing rather than sharp pain. None of these symptoms is definitively cardiac, and all of them have non-cardiac causes. But all of them are documented presentations of MI in women, and all of them can be present in the absence of the textbook crushing chest pain that anchors most public awareness of heart attack symptoms.

If a prior evaluation concluded that symptoms were non-cardiac before a cardiac evaluation was completed, that history is clinically relevant and should be stated. “I was told this was anxiety in [month/year] and my symptoms have continued” is information that changes the weight of a subsequent evaluation. A clinician who knows that a patient’s symptoms were previously attributed to a non-cardiac cause and the symptoms persisted is in a different position than one evaluating a first presentation.

What to Do This Week

1. Bring your full history to every cardiac evaluation: pregnancy complications including preeclampsia and preterm birth, autoimmune diagnoses such as lupus or rheumatoid arthritis, family history through female relatives, and the specific character and timing of your symptoms.

2. When told a test is normal, ask what specifically the test ruled out, not whether your heart is generally fine. A normal stress test rules out flow-limiting epicardial stenosis under stress conditions. It does not rule out microvascular dysfunction or endothelial dysfunction.

3. If symptoms persist after a normal stress test or normal angiogram, ask whether non-obstructive coronary disease has been considered and whether coronary function testing, including coronary flow reserve measurement with adenosine, has been discussed. Ask your cardiologist about INOCA by name.

4. If cardiac symptoms were attributed to anxiety, stress, or hormones before a cardiac evaluation was completed, ask for the cardiac evaluation to be completed first. A cardiac evaluation should precede, not follow, a non-cardiac explanation for cardiac-sounding symptoms.

5. If you are being evaluated in an emergency setting, ask whether the troponin threshold being used is sex-specific. Many labs now use lower thresholds for women, but not all do. Asking the question puts the issue on record and may prompt a physician to verify the result against a sex-appropriate cutoff.

The three mechanisms described here are identifiable, and identification is the starting point. A woman who knows that her disease may not look like the textbook picture, that the tests may not be calibrated for her, and that symptom attribution patterns work against her, is better positioned to ask the right questions and to recognize when a negative result is incomplete rather than reassuring. The miss is not inevitable. It is the output of nameable mechanisms, and naming them is a concrete step.