Myocardial Infarction Classification and Its Different Types

Heart attacks are not all the same event. The international cardiology community divides myocardial infarction into five numbered types based on what triggers the heart muscle damage, and a separate electrocardiogram-based split (STEMI versus NSTEMI) based on how the damage shows up electrically. These distinctions matter because they steer different patients toward fundamentally different treatments: emergency stenting for one person, blood-pressure management for another, and watchful monitoring for a third. Understanding the classification system helps make sense of why two people can both be told they had a “heart attack” yet receive completely different care.

The Five Types at a Glance

The current framework comes from the Fourth Universal Definition of Myocardial Infarction, a consensus document published in 2018 by the European Society of Cardiology, the American College of Cardiology, the American Heart Association, and the World Heart Federation.¹ It establishes five types:

Type 1: caused by a ruptured or eroded plaque in a coronary artery, the classic “heart attack” driven by atherosclerosis.
Type 2: caused by an imbalance between how much oxygen the heart muscle needs and how much it receives, without an acute plaque event.
Type 3: sudden cardiac death with symptoms or ECG evidence suggesting a heart attack, but the patient dies before blood biomarkers can be measured.
Type 4: heart muscle damage related to percutaneous coronary intervention (balloon angioplasty or stenting), subdivided into 4a (the procedure itself), 4b (stent thrombosis), and 4c (restenosis).
Type 5: heart muscle damage related to coronary artery bypass graft surgery.

Each type carries different risk profiles, occurs in different clinical settings, and calls for different management. Type 1 is by far the most common and the most studied, but the others are far from rare, and misclassifying one type as another can lead to the wrong treatment plan.

Type 1 MI and Atherosclerotic Plaque Rupture

When most people picture a heart attack, they are picturing a Type 1 MI. A fatty plaque that has built up inside a coronary artery wall becomes unstable, ruptures or erodes, and triggers a blood clot that partially or completely blocks blood flow to part of the heart. The downstream muscle, starved of oxygen, begins to die. Risk factors include high LDL cholesterol, smoking, chronic kidney disease, diabetes, high blood pressure, and inherited cholesterol disorders.² Treatment centers on reopening the blocked artery as quickly as possible, either with clot-dissolving drugs or emergency catheterization and stenting, followed by long-term medications like blood thinners, statins, and blood-pressure drugs.

Type 2 MI and the Supply-Demand Mismatch

Type 2 MI happens when the heart muscle is damaged not because a plaque blew open, but because something else threw the balance between oxygen supply and demand out of alignment. A severe infection driving a dangerously fast heart rate, a drop in blood pressure during critical illness, profound anemia, respiratory failure, or a hypertensive crisis can all push the heart past what its blood supply can sustain.³ By some estimates, Type 2 MI accounts for anywhere from about 5% to 35% of all diagnosed heart attacks, a wide range that reflects how differently hospitals apply the definition.⁴

Treatment differs sharply from Type 1. Rushing a Type 2 MI patient to the catheterization lab for emergency stenting usually does not help, because the problem is not a blocked artery. Instead, clinicians treat the underlying trigger: transfuse the patient if anemia is severe, manage the infection driving the rapid heart rate, or bring dangerously high blood pressure under control. Patients hospitalized with Type 2 MI received invasive coronary procedures far less often than Type 1 patients in comparative studies, and when they did undergo angiography, roughly two-thirds had normal coronary arteries or only minor disease.⁵ Standard heart-attack medications like dual antiplatelet therapy and statins were prescribed less often at discharge, while blood thinners and diuretics were used more, reflecting the different underlying conditions these patients tend to have.

Type 2 MI in Sepsis and Critical Illness

Sepsis is one of the most common settings for Type 2 MI, and it illustrates how many mechanisms can hit the heart at once. In one study of over 500 patients who had a Type 2 MI during sepsis, roughly two-thirds had hypoxemia, a similar proportion had anemia, and about a third had hypertension, with the vast majority having multiple overlapping causes.⁶ Sepsis-related inflammation has also been linked to heart attacks through mechanisms beyond simple oxygen mismatch, including direct effects on blood vessel function.⁷ This overlap makes classification difficult at the bedside. A critically ill patient with sepsis, dropping blood pressure, and a rising troponin could have a Type 2 MI from oxygen-demand mismatch, or they could have developed an actual plaque rupture because severe illness destabilized a vulnerable plaque. Telling the two apart has real consequences for whether the patient needs cardiac catheterization.

Type 3 MI and Sudden Cardiac Death

Type 3 is the most sobering category. It applies when someone dies suddenly with symptoms or electrical findings suggesting a heart attack, but death occurs before blood samples can confirm the diagnosis through biomarkers. In some cases, an autopsy reveals the culprit clot or fresh heart muscle death; in others, the classification rests on the clinical picture alone. Type 3 exists in the framework partly to ensure these deaths are counted in epidemiological data and not lost to an “unknown cause” label. From a treatment standpoint, there is no management strategy for Type 3 because the patient has already died. Its value is in accurate record-keeping and in understanding how many heart attack deaths happen before patients ever reach a hospital.

Procedure-Related Heart Attacks

Types 4 and 5 cover heart muscle damage that occurs as a direct consequence of coronary procedures. Any intervention on the heart’s arteries carries some risk of causing the very injury it aims to prevent.

Type 4a MI refers to damage during or immediately after percutaneous coronary intervention. In one study of nearly 1,400 elective stenting patients, about 7% met the criteria for Type 4a MI, and an additional roughly 22% had measurable heart muscle injury that fell short of the full MI threshold.⁸ Those who had periprocedural MI or injury had nearly four times the rate of cardiovascular events at 30 days compared with those who did not, and the elevated risk persisted at one year. Type 4b (stent thrombosis, where a blood clot forms inside the newly placed stent) was rare in that cohort, occurring in fewer than 1 in 700 patients.

Type 5 MI occurs after coronary artery bypass graft surgery. Because the operation itself involves handling, clamping, and manipulating the heart and its blood supply, some rise in cardiac biomarkers afterward is expected even when everything goes well. Defining what counts as a “real” MI versus routine surgical stress has been a persistent challenge. A meta-analysis found that post-surgical troponin concentrations often surpass the standard cutoff values for Type 5 MI even in uncomplicated surgeries, prompting calls for the thresholds to be reassessed.⁹ Using troponin alone at a well-chosen cutoff, though, appears to work: one cardiac MRI study found that troponin I measured 24 hours after surgery predicted Type 5 MI with about 88% sensitivity and 97% specificity.¹⁰

STEMI Versus NSTEMI

Alongside the five-type system, there is an older and still clinically essential split based on the electrocardiogram. When a heart attack produces a distinctive pattern called ST-segment elevation on the ECG, it is labeled STEMI; when that pattern is absent, it is called NSTEMI. This distinction drives the clock in the emergency department: STEMI signals a total blockage of a coronary artery and triggers an immediate push for reperfusion, either with clot-busting drugs or emergency catheterization.¹¹

The STEMI/NSTEMI split and the five-type system overlap but are not the same thing. A Type 1 MI can present as either STEMI or NSTEMI. A Type 2 MI more commonly shows up as NSTEMI, because the mechanism is rarely a total coronary occlusion. One thing worth knowing is that the relationship between ECG pattern and actual tissue damage is not perfectly tidy. When researchers used cardiac MRI to look at the extent of heart muscle death, STEMI patients had full-thickness (transmural) damage about 63% of the time, while NSTEMI patients showed transmural damage about 27% of the time.¹² So the ECG pattern is strongly associated with worse damage, but it is not a clean binary: some NSTEMI patients have extensive damage, and some STEMI patients do not.

MINOCA: When the Arteries Look Clear

About 5% to 15% of patients who meet all the standard criteria for a heart attack turn out to have no significant blockage when their coronary arteries are imaged. This is called MINOCA, myocardial infarction with nonobstructive coronary arteries, and it challenges the assumption that a heart attack always means a clogged artery. The underlying causes are varied: a plaque that ruptured or eroded but did not create an obvious blockage, spasm of a coronary artery, dysfunction in the tiny blood vessels of the heart (microvascular disease), or coronary artery embolism.¹³

Diagnosing MINOCA requires looking beyond the standard angiogram. Cardiac MRI has become a key tool because it can distinguish between true infarction, inflammation (myocarditis), and stress-related wall-motion abnormalities. Intravascular imaging techniques like optical coherence tomography (OCT) can reveal plaque rupture or dissection that standard angiography misses. In one study, coupling OCT with cardiac MRI provided a clear diagnosis in all 40 MINOCA patients studied, whereas neither technique alone reached that.¹⁴ This matters because treatment depends entirely on which underlying cause is found. Vasospasm responds to calcium channel blockers, not to stents. Microvascular disease calls for a different drug strategy entirely. Treating all MINOCA patients the same way would leave many of them on medications that do not address their actual problem.¹⁵

Prognosis Varies Sharply by Type

You might assume that Type 2 MI, being less dramatic-sounding than a full plaque rupture, carries a better outlook. The opposite tends to be true. A meta-analysis of observational studies found that inpatient mortality for Type 2 MI was about 15%, compared with roughly 5% for Type 1. At one year, mortality was about 27% for Type 2 versus 13% for Type 1.¹⁶ This seemingly paradoxical finding reflects the fact that Type 2 MI often strikes people who are already seriously ill from other conditions. Sepsis, respiratory failure, major surgery, and severe anemia are common backdrops. A study focusing on younger adults found that over roughly a decade of follow-up, Type 2 MI carried about 2.7 times the risk of cardiovascular death compared with Type 1, even after adjusting for other health differences.¹⁷

The higher mortality is not necessarily because the heart damage itself is worse. It is because the population that gets Type 2 MI tends to be sicker at baseline. The heart attack is often one problem layered on top of many others. This has practical implications: improving outcomes for Type 2 MI likely depends more on managing the underlying illness effectively than on cardiac-specific interventions.

Spontaneous Coronary Artery Dissection

Spontaneous coronary artery dissection, or SCAD, is a tear in the wall of a coronary artery that creates a pocket of blood (a hematoma) compressing the vessel from within. It accounts for roughly 1% to 4% of acute coronary syndrome cases and is a recognized cause of heart attacks, particularly in younger women and people who lack the usual risk factors for atherosclerosis.¹⁸ SCAD has been linked to pregnancy, intense physical or emotional stress, and an underlying blood-vessel condition called fibromuscular dysplasia.¹⁹

SCAD fits awkwardly into the five-type classification. It is not a plaque rupture (Type 1), not a supply-demand mismatch in the usual sense (Type 2), and not procedure-related. It can be classified as a Type 1 MI when the dissection triggers thrombosis, or it may fall under MINOCA if the angiogram does not show an obvious obstruction. Because the vessel wall is already torn, the standard heart-attack intervention of stenting can sometimes make things worse by extending the dissection. Many SCAD patients are managed conservatively, with medications alone, unless blood flow is severely compromised.²⁰ Recurrence rates are high enough to warrant long-term follow-up, which again sets SCAD apart from ordinary atherosclerotic heart attacks.

Takotsubo Cardiomyopathy as a Mimic

Takotsubo syndrome, sometimes called “broken heart syndrome” or stress cardiomyopathy, can look almost identical to a heart attack on initial presentation. Patients arrive with chest pain, ECG changes, and elevated troponin, yet their coronary arteries are unblocked. What distinguishes takotsubo is a characteristic ballooning of the heart’s left ventricle, usually affecting the apex and the mid-ventricular segments in a pattern that crosses the territory of any single coronary artery.²¹ In a true heart attack, the area of reduced function typically maps to a single artery’s territory.

ECG patterns can help distinguish the two. Research comparing admission ECGs in 200 takotsubo patients and 200 MI patients found that certain subtle features, like ST-elevation in a specific lead configuration, were highly specific for one diagnosis or the other, though sensitivity was often low.²² In practice, many takotsubo patients end up in the catheterization lab before the distinction becomes clear. The importance of recognizing takotsubo is that the treatment is supportive rather than interventional, and the condition is usually reversible within weeks, unlike the permanent muscle loss from an actual MI.

Sex Differences in Diagnosis and Timing

Heart attack classification does not happen in a vacuum. The speed and accuracy of diagnosis are influenced by how a patient presents, and that presentation varies between sexes in ways that affect care. Women with heart attacks are more likely to have atypical symptoms like shortness of breath, nausea, or fatigue rather than classic crushing chest pain. This can delay recognition. One large study found that male STEMI and NSTEMI patients received their first ECG about three minutes faster on average, and men had roughly 36% lower odds of an ECG delay, a gap that persisted even when both sexes presented with a chief complaint of chest pain.²³ For STEMI patients needing emergency catheterization, men had shorter wait times to the procedure as well, though once the time-to-ECG delay was accounted for, sex alone was no longer a significant predictor of catheterization delay. That finding suggests the bottleneck is in initial recognition, not in what happens after the diagnosis is made.

Whether these timing gaps translate to worse outcomes remains debated. A registry from Kerala, India, found that after adjusting for confounders like age and comorbidities, there was no significant difference in death rates between men and women with acute coronary syndromes.²⁴ Other studies have reached different conclusions, and the inconsistency across populations has been acknowledged as an ongoing problem in the field.²⁵ What is consistent is that women are more likely to have MINOCA-type presentations and SCAD, conditions that fall outside the standard atherosclerotic heart-attack playbook and can be missed if clinicians are not looking for them.

The Role of Troponin in Sorting Out Types

Troponin, a protein released by damaged heart muscle cells, is the central biomarker for diagnosing any type of MI. But a single troponin measurement does not tell you what type of MI occurred. Serial measurements, taken hours apart, help. The pattern of rise and fall in troponin levels provides clues about whether the damage is acute and evolving (suggesting a fresh event) or chronic and stable (suggesting ongoing low-level injury).

For early diagnosis, the absolute change in troponin between two measurements appears to outperform the relative (percentage) change. A multicenter trial found that the absolute change in troponin I was significantly better at identifying MI in the first few hours, with the diagnostic gap between the two approaches being widest in the earliest measurements.²⁶ When baseline troponin was very low and the absolute change remained small, the ability to rule out MI was excellent, with a negative predictive value above 99%. This kind of granularity matters in emergency departments, where deciding whether someone is having a heart attack and what kind of heart attack it is determines everything from which unit they go to, to whether they need an immediate procedure.

Where Artificial Intelligence Fits In

AI-powered ECG interpretation is an active area of research, particularly for catching heart attacks that the standard ECG criteria miss. One deep-learning algorithm trained to detect MI from 12-lead ECGs achieved an area under the curve of about 0.90 in both internal and external validation, meaning it performed well at separating MI from non-MI cases. A version that worked from just six ECG leads (relevant for portable or wearable devices) still reached about 0.85 to 0.88, outperforming the rule-based interpretation built into conventional ECG machines.²⁷

The most clinically interesting application may be in NSTEMI patients, where the standard ECG often fails to flag a complete coronary occlusion that actually needs emergency intervention. One study testing an AI-augmented triage approach in high-risk NSTEMI patients found that it cut false positives from about 42% to 22%, though sensitivity for ruling in occlusive MI was moderate. Serial ECGs reduced false negatives somewhat. A scoping review of the broader AI-ECG literature struck a cautionary note, finding that while many studies reported very high accuracy figures, those numbers often came from small datasets with validation methods that tend to overestimate real-world performance. Accuracy frequently dropped when algorithms were tested on truly independent patient groups.²⁸ The technology is promising but not yet at the point where it changes bedside decisions on its own.

Anatomical Location and the Limits of the Standard ECG

Beyond type and electrical pattern, heart attacks are also classified by where in the heart muscle the damage occurs: anterior (front wall), inferior (bottom), lateral (side), or posterior (back). Location matters for prognosis and for identifying complications. Inferior-wall heart attacks, for instance, sometimes extend into the right ventricle or the posterior wall, and these extensions change both treatment and outlook. The standard 12-lead ECG has blind spots, particularly for posterior and right ventricular involvement, and additional electrode placements on the right side of the chest or the back improve detection.²⁹ This is one of those details that mostly matters for clinicians, but it helps explain why patients sometimes hear that their heart attack was “missed” on an initial ECG. The machine was not looking in the right place.