The Heart of the Matter

A recent review gathers the evidence that Long COVID leaves lasting marks on the heart and blood vessels — myocarditis, arrhythmias, heart failure, clots — and argues that immune dysregulation and inflammation are the "central drivers" of that injury. It calls the cardiovascular manifestations "core characteristics" of Long COVID and points toward treatments that target inflammation. The underlying cardiovascular epidemiology it draws on is some of the most carefully done work in the entire field, and none of what follows disputes it. Post-COVID heart disease is real, and the people who have it deserve care built around it.

But a review like this circulates far beyond cardiology, and a phrase like "central driver" travels. The question worth asking is whether this cardiovascular story explains the patients who actually fill a Long COVID clinic — the ones whose illness is unrelenting fatigue, crashes after minor exertion, unrefreshing sleep, and cognitive fog. My reading of the same data is that it mostly does not, and that treating it as if it does harms both groups.

A big risk ratio can sit on a very small risk

The most-cited cardiovascular study in Long COVID built a cohort of more than 150,000 infected veterans against some eleven million controls and followed them for a year. It is rigorously constructed, and it found genuinely elevated risks — even in people who were never hospitalized. But there is a distinction that gets lost when these numbers are summarized, and it changes the clinical meaning entirely: the difference between how much a risk is multiplied and how many people it actually affects.

Myocarditis is the clearest example. It carried the single highest relative risk in the study — a more than five-fold increase. That sounds alarming. But the actual number of extra cases was about three per ten thousand people over the year. A large multiple of a very small number is still a very small number. The broadest possible category — "any cardiovascular event," which lumps everything together — reached about 4.5 percent, but the specific diagnoses that would actually prompt a cardiac work-up are much smaller: heart failure around 1.2 percent, atrial fibrillation around 1.1 percent, pulmonary embolism around 0.5 percent, stroke around 0.4 percent. And every one of these risks climbed steeply with how sick the person was to begin with — lowest in those never hospitalized, higher in the hospitalized, highest after intensive care.

Long COVID is not one population

When researchers sort large groups of Long COVID patients by their symptoms, a fairly consistent split appears: a cardiorespiratory cluster (chest pain, breathlessness, palpitations) and a separate fatigue-dominant cluster. The two overlap at the edges — palpitations and shortness of breath are nonspecific and show up in both — but they are not the same people. One review of Long COVID subtypes went so far as to exclude studies that had blended ME/CFS and Long COVID symptoms into a single analysis, an implicit acknowledgment that the two shouldn't simply be merged.

Here is the gap that matters. No cardiovascular-outcomes study — including the large veteran cohort — has separated its findings by the features that define the ME/CFS-like phenotype: post-exertional malaise, unrefreshing sleep. The exposure is just "Long COVID," by diagnostic code or symptom persistence, which is far too broad to tell us whether the cardiac risk concentrates in the fatigue-and-crash subgroup or in a different, sicker one. The review under discussion never engages this at all — across its length, it does not mention ME/CFS or post-exertional malaise once. Fatigue and exercise intolerance appear, but they are quietly absorbed into a cardiovascular-inflammatory story rather than treated as their own problem.

And when the fatigue phenotype is actually studied on its own terms, it looks different. Careful invasive exercise testing of ME/CFS and Long COVID patients points not to structural heart disease but to a functional problem — the heart fills poorly on standing and during effort ("preload failure"), with impaired oxygen extraction by the tissues, in people whose hearts are structurally normal. That is a categorically different thing from a scarred myocardium, a blocked coronary artery, or heart failure defined by imaging and blood markers. It is exercise intolerance with a normal echocardiogram — which is precisely the patient the cardiovascular literature was not built to describe.

What the numbers can and can't be generalized to

Two cautions about applying these figures to a clinic like mine. First, severity is a real confounder: the cardiovascular excess rises sharply with hospitalization, and while COVID does appear to carry a genuine, virus-specific cardiac and clotting risk even compared with other severe infections like influenza, that comparison has only ever been made in hospitalized, severely ill cohorts. It has not been tested in the mild, outpatient infection that most fatigue-phenotype patients actually had.

Second, the veteran cohort is older and mostly male — the authors say so themselves. The relative risks probably carry over reasonably to other groups, but the absolute numbers were generated in a population with a much higher baseline rate of heart disease. Applied to a younger, predominantly female clinic population, those absolute risks almost certainly overstate the true burden.

Old damage or a new process? Honestly unresolved

There is also a genuine scientific gap about what these findings even are. Most long-term cardiac data pairs a blood marker of heart injury taken at hospitalization with later follow-up, and shows a mixed picture — most of the acute elevation resolves within months, but some persists. Cardiac MRI, the scan that generates the most dramatic prevalence figures, is essentially never done during the acute infection, so there is no "before" picture to say whether a later abnormal scan is leftover acute damage or something that developed afterward. One study that tested this indirectly found no significant difference in scar or active-myocarditis rates between previously hospitalized and non-hospitalized patients (though heart function was somewhat lower in the hospitalized group) — which argues against the simplest "unresolved acute damage" story without proving a distinct later process. The honest answer is that we don't know.

What this means at the bedside

The practical payoff is reassuring, and it cuts against over-testing rather than for it. A basic, symptom-triggered cardiac work-up — an ECG, a troponin, a standard echocardiogram — is proportionate, inexpensive, and, given how small these absolute numbers are, genuinely reassuring when it comes back normal. In a patient whose presentation is fatigue and post-exertional crashes without focal cardiac findings, that work-up does not need to be escalated to cardiac MRI or routine cardiology referral to chase an occult myocarditis that, in absolute terms, is rare.

That is a real gift, because it stands in sharp contrast to the rest of the syndrome. A fear of hidden heart damage can be answered with a normal echocardiogram. The cognitive dysfunction, dysautonomia, exercise intolerance, and unrefreshing sleep that define most of these patients have no comparably simple test to "rule out" — they don't resolve to a clean scan. The danger in treating Long COVID as a single entity, unified by an immune-inflammation narrative shared across the cardiology and ME/CFS literatures, is that it drives cardiac anxiety and specialist referral in exactly the patients whose real illness sits somewhere else entirely.