I. The Entry Point
I encountered a WIRED article while browsing r/Longform. The Comments were overall quite negative.
Official Summary by WIRED in the Comments
Nothing about long Covid adds up.
Consider prevalence rates: How could one study find it affected 3.3 percent of the population of the UK but others an alarming 51 percent of South Americans and 86 percent of Egyptians? Or treatment methods: The BMJ’s systematic review of ways to treat long Covid lists two as supported by moderate evidence, cognitive behavioral therapy and physical exercise. But if you attended the third annual Long Covid International Conference in Boston, you’d think the BMJ was encouraging medical malpractice. During two days of presentations, the world’s leading scientific authorities brought up exercise only to warn against it. Cognitive behavioral therapy received just one mention: “not recommended.”
Then there’s the scientific progress, or lack thereof. Six years since the height of the pandemic, the scientific community remains baffled by long Covid. Researchers still don’t know why some people’s symptoms persist or worsen after the acute phase of SARS-CoV-2 infection has passed. Almost $2 billion and half a decade of international effort have yielded little more than hypotheses about micro blood clots and spike proteins and mitochondrial dysfunction. There isn’t a single approved pharmaceutical treatment, not even a test to verify the presence of the illness.
All of this is very strange. Stranger still are patients’ stories of astonishing recoveries from severe long Covid, achieved entirely outside mainstream medicine. The stories are connected to a growing community of doctors, therapists, and self-styled coaches who insist the riddle of long Covid has been solved. Like so many health gurus, they offer a solution that depends, in part, on your faith in the process. The solution also works for a remarkable variety of ailments—classic red flags for “holistic” pseudoscience and medical charlatanry.
My Follow Up Research
The 3.3 percent UK figure aligns with Office for National Statistics data, which measured prevalence across the entire population of the country at a given time. Meanwhile, the 51 percent (from a recent global meta-analysis) and 86 percent (from African cohort studies) measure prevalence only among people who had a confirmed COVID-19 infection, counting anyone who reported even a single lingering symptom months later.
The BMJ did publish a major systematic review concluding there is moderate certainty evidence that cognitive behavioral therapy (CBT) and physical exercise improve symptoms.Yet, at specialized clinical gatherings like the Long Covid International Conference in Boston, experts actively warn against these very things.
The disconnect comes down to how general medicine evaluates evidence versus how specialized medicine observes the body. Major journals like the BMJ rely on systematic reviews, which average out the results of past clinical trials.
Because exercise and CBT improve chronic fatigue, these interventions are presented as the gold standard. However, researchers who specialize in chronic illnesses recognize that Long Covid is different. Its hallmark symptom is Post-Exertional Malaise (PEM). When a Long Covid patient exerts themselves physically or mentally, the exertion can damage their cells and worsen their condition—a mechanism documented in similar illnesses like Myalgic Encephalomyelitis (ME/CFS).
The almost $2 billion figure closely aligns with the budget of the NIH’s RECOVER initiative. Despite that massive financial investment, we still do not have a clinically validated biomarker test to prove someone has Long Covid, nor a universally approved pharmaceutical treatment. The scientific community genuinely is still stuck at the hypothesis stage, debating microclots, viral persistence, and mitochondrial dysfunction.
Because mainstream medicine has left patients stranded without a pill or a protocol, a vacuum has formed. This space has been filled by brain retraining programs and mind-body coaches. These programs demand absolute psychological compliance (faith in the process) and conveniently claim to cure everything from Long Covid to chronic Lyme and fibromyalgia, textbook red flags for pseudoscience.
II. The Seven Causes
What are the prevailing scientific theories regarding Long COVID? Why have they failed to gain consensus?
When tracing the scientific literature of the past several years, the major hypotheses splinter into several distinct pathways. These mechanisms can be grouped into three overarching categories of dysfunction.
Viral Origins
The most immediate assumption following a viral pandemic is that the virus simply never left.
Viral Persistence: This theory suggests that hidden reservoirs of live SARS-CoV-2, or perhaps highly immunogenic viral fragments (like spike proteins), remain deeply entrenched in tissues—such as the gut lining or the central nervous system—long after the acute infection has passed. The immune system, detecting this debris, remains locked in a perpetual state of alarm.
Latent Virus Reactivation: Alternatively, the physiological trauma of the initial COVID-19 infection may exhaust the host’s immune system, dropping its guard enough to allow dormant viruses already residing in the body to wake up. The Epstein-Barr virus (the cause of mononucleosis) is a primary suspect, seizing the opportunity of an immunologically distracted host to trigger chronic illness.
Vascular and Cellular Fallout
If the virus itself is gone, the focus shifts to the wreckage it left behind—specifically, the infrastructure of the blood and the cells.
Microvascular Damage and Microclots: SARS-CoV-2 is notoriously efficient at damaging endothelial cells, the lining of our blood vessels. This damage can trigger the formation of microscopic, amyloid-like blood clots that clog capillaries. These clots starve muscles and organs of oxygen, offering an explanation for the physical fatigue and brain fog patients experience.
Mitochondrial Dysfunction: The virus appears to damage the mitochondria—the organelles responsible for producing ATP. If the body physically loses the capacity to produce and distribute energy efficiently, it answers why even minor physical or cognitive exertion can damage the patient’s health
The Neurological and Systemic Loop
The final grouping moves away from the pathogen and the initial damage, looking instead at how the body’s internal networks begin to misfire in an endless, self-sustaining loop.
Immune Dysregulation and Autoimmunity: The initial viral attack throws the immune system into overdrive, causing it to misfire and produce autoantibodies. Instead of fighting a virus, the immune system begins attacking the body’s own healthy tissues.
Microbiome Alterations (Gut Dysbiosis): The gut is increasingly recognized as an important part of the immune system. Studies consistently show that Long COVID patients show depleted levels of beneficial bacteria and high levels of inflammatory pathogens. This leaky gut dynamic is viewed as a primary driver of chronic, systemic inflammation.
Neuroinflammation and Vagus Nerve Dysfunction: Finally, the virus appears capable of triggering inflammation within the brain and damaging the vagus nerve, the biological highway that controls the autonomic nervous system. When the vagus nerve misfires, autonomic functions like heart rate, blood pressure, and digestion are disrupted. This is the leading explanation for why so many patients develop severe dysautonomia and POTS (Postural Orthostatic Tachycardia Syndrome), where simply standing up causes a violent spike in heart rate.
Despite these theoretically sound and observable theories, no consensus has emerged. Part of the problem is that these categories are not isolated silos. Gut dysbiosis is studied in tandem with neuroinflammation via the gut-brain axis; similarly, endothelial dysfunction is viewed as the bridge between viral persistence and the formation of microclots.
That still isn’t enough to answer why the scientific community failed to reach a consensus. But when you break down how clinical research is actually conducted and funded, the answer emerges in four parts. We possess an abundance of theories but lack the capacity to prove them due to the following factors:
The Umbrella Problem: Long COVID is almost certainly not a single disease. When clinical trials group together a patient suffering from a loss of smell, a patient whose heart races uncontrollably due to POTS, and a patient bedbound by post-exertional malaise, researchers are effectively averaging out the data of different biological mechanisms.
The Cascade Effect: The seven theories outlined above are not mutually exclusive; they are dominos. Viral persistence might trigger immune dysregulation, which in turn causes endothelial damage, which results in microclots. Researchers are unable to solve the biological chicken-and-egg scenario, fighting bitterly over which mechanism is the true root cause and which is merely a downstream symptom.
The Biomarker Void: Without a definitive blood test or imaging scan to definitively prove a patient has Long COVID, researchers must rely on subjective patient surveys and wildly varying diagnostic criteria. This lack of objectivity makes it incredibly difficult, if not impossible, to reliably compare data across different international studies.
Measurement Limitations: Proving a theory like viral persistence in deep tissue requires biopsies from the brain, the heart, or the deep gastrointestinal tract. These are highly invasive, inherently dangerous procedures that are ethically and logistically impossible to perform at scale on a living population.
III. The Limits of Identity
The closest thing we have to a medical consensus on Long COVID is the tacit agreement that it is an umbrella term. This is, effectively, a consensus of ignorance—an admission that the illness is fundamentally ill-defined.
From a societal standpoint, consensus brings legitimacy. Patients need a biologically defined, universally recognized diagnosis to secure insurance coverage, negotiate workplace accommodations, or qualify for disability benefits.
Health agencies like the FDA require a rigidly defined disease model, clear entry criteria, and measurable clinical endpoints to approve new treatments. Without a consensus on what the disease actually is, drug development cannot advance from the laboratory to the market. The biological ambiguity is a regulatory dead end.
A landmark systematic review of Long COVID biomarkers (Lai et al.) found that across various studies, researchers have identified well over 100 candidate biomarkers (including inflammatory markers like IL-6 and TNF-alpha, and neurological indicators like neurofilament light chain). Yet, subsequent clinical updates continually arrive at the same regulatory dead end: none of these markers have achieved the sensitivity and specificity required to act as a universal diagnostic tool.
To anyone familiar with etiology, the seven mechanisms are hardly unique to COVID. In fact, they read like the greatest hits of chronic illness.
If you map a condition like Myalgic Encephalomyelitis (ME/CFS) or Post-Treatment Lyme Disease Syndrome against these theories, they easily hit five, six, or even all seven of the categories. The damage pathways we attribute to COVID are the same pathways driving a vast array of established, deeply debilitating human ailments.
When attempting to describe Long COVID in the modern literature, scientists frequently resort to a dense wall of jargon. They use terms like integrated multisystem pathology or cascading, multi-systemic immune-metabolic disorder. These adjectival jumbles point toward a failure in the terminology, and the terminology points toward the limits of the theory.
We are pushing past what we would conventionally understand as a disease. A traditional disease has a linear narrative: pathogen enters, causes specific damage, is eradicated, and the body heals (or fails to). But this is different. And while taking a holistic, systems-level view sounds wonderfully progressive in theory, it presents immediate, fatal problems for clinical medicine.
The array of mechanisms provided captures a net that is simply too wide. It can easily be applied to other malaises. They are a tacit admission that Long COVID, as a novel biological entity, doesn’t really exist.
This is not to say the patients aren’t suffering from something. They are measurably sick. It is saying that Long COVID is better understood without the COVID.
IV. To Classify and Abstract
To claim that Long COVID is better understood without the COVID is provocative. The immediate pushback is obvious: Surely, even if the downstream mechanisms of the illness are shared with Myalgic Encephalomyelitis or Lyme disease, the catalyst still matters?
In a bid to organize the chaos, the scientific community has recently pushed to establish distinct diagnostic criteria for subgroups within the Long COVID umbrella. A wave of 2025 and 2026 literature—spanning cross-continental cohort analyses and computational phenotyping models—has effectively abandoned the singular disease framework. We now see researchers publishing cluster analyses that partition patients into strict neurological categories (frequently branded in the literature as Neuro-PASC), cardiopulmonary subsets, or autonomic cohorts based entirely on their predominant systemic failures.
But this effort is a fatal concession. It is the exact moment clinical science admits that the mechanism matters more than the catalyst.
Once you take this route and deconstruct the illness by its mechanical failures, you run into an immediate logical wall. What do you do when a completely different catalyst—an influenza strain, a bacterial infection, or extreme physiological trauma—triggers the exact same biological cascade? Does a post-Epstein-Barr patient with microvascular clotting fall into a Long COVID subgroup?
Because that outcome is clearly nonsensical, we are forced to reclassify what we are actually looking at. Long COVID, in the vast majority of its presentations, is a viral entry point activation of Post-Acute Infection Syndrome (PAIS).
The only strictly biological consideration left is whether the original catalyst is still actively participating.
Taking a step back from the COVID pandemic, the distinction is historically clear. If a patient harbors active viral reservoirs (like the theories of deep-tissue viral persistence), they do not have a post-viral syndrome; they have a chronic, active SARS-CoV-2 infection. This operates much in the same way a patient can have chronic Hepatitis B or HIV. The pathogen is still the primary antagonist.
But what is far more common in the PAIS umbrella, is the hit-and-run dynamic. This occurs when the initial pathogen enters, inflicts physiological trauma, and is subsequently cleared by the body—but the downstream effects become autonomous causes of their own.
When we talk about autoimmune cascades, autonomic nervous system failure, and gut dysbiosis, we are talking with a degree of abstraction. At a certain point, the initial catalyst ceases to matter. The fact that Long COVID’s proposed mechanisms are so widely available to, and documented in, other chronic illnesses severely weakens its case as a uniquely identifiable disease.
We are left watching a conflict between clinical logic and sociological necessity.
The implicit rule of medical nomenclature is that disease names should serve to classify and abstract biological realities, allowing doctors to treat them accurately. But in the modern landscape, the name is a weapon. The sociological necessity of the COVID label fights fiercely against the clinical logic of removing it. Without the cultural gravity of the pandemic, we can look towards the unfunded corners of medicine, where those who suffer from ME/CFS or Fibromyalgia remain as a likely outcome.
Of course, the reason the term Long COVID exists at all has nothing to do with biological taxonomy and everything to do with the historical trauma of the 2020 pandemic. But if we are to seriously consider how disorders that have become abstracted from their initial catalysts should be classified moving forward, the solution is not to invent a novel disease category that defies biological reality.
For a practical compromise between parsimony, scientific rigor and sociological reality, consider flipping the order.
PAIS SARS-CoV-2 Variant
V. Beyond COVID
To consider the matter more broadly, we can step away the novelty of the 2020 pandemic and look at the most mundane pathogens we live with: the influenza virus (the flu) and the rhinovirus (the common cold). If the biological mechanisms driving Long COVID are not unique to SARS-CoV-2, we should be able to map them onto other familiar viruses.
Long Flu absolutely exists. The scientific literature traces many of the same pathways as Long COVID:
Immune Dysregulation: Post-influenza autoimmunity is a rare but well-documented reality, most famously triggering Guillain-Barré Syndrome, where the immune system misfires and attacks the body’s peripheral nerves.
Microvascular Damage: Influenza significantly increases the risk of severe cardiovascular events. The risk of a heart attack is six times higher in the week following a flu infection because the virus triggers systemic inflammation that can destabilize arterial plaques.
Gut Dysbiosis: Respiratory infections like the flu actively alter the gut microbiome through the lung-gut axis, shifting the balance of bacteria and driving systemic inflammation.
Neuroinflammation: The brain fog and neurological fatigue experienced during and after a severe flu are not just psychological; they are the result of inflammatory cytokines successfully crossing the blood-brain barrier.
Mitochondrial Dysfunction: Influenza viruses are known to manipulate mitochondrial dynamics to evade the host’s immune response, directly contributing to post-flu physical exhaustion.
By contrast, Long Cold is exceedingly rare.
The rhinovirus does not cause microvascular clotting, it rarely triggers severe immune dysregulation or autoimmunity, and it does not cause deep neuroinflammation or vagus nerve dysfunction.
The reason for this is that the rhinovirus is highly adapted to replicate in the cooler temperatures of the upper respiratory tract (the nose and throat). It lacks the biological tools to disseminate into the bloodstream or infiltrate deep organs.
So, what can we learn from comparing the flu to the cold?
It proves that the distinct identity of the virus matters far less than its systemic capability. We can abstract away the microscopic specifics of the pathogen (whether it is an influenza strain or a coronavirus) and instead classify viruses using a highly abstracted schema: Tier 1 (Local) versus Tier 2 (Global).
This highlights a distinction between two branches of science that are often conflated. Virology is a science of identity; it asks what a virus looks like under a microscope and how its RNA is folded. Etiology is a systems-engineering science. It asks how a virus breaks the body’s networks.
In the medical scholarship, this concept is tethered to Viral Tropism—the ability of a given virus to infect a specific tissue or cell type. The Tier 1 / Tier 2 schema is much broader. It does not care about which tissues the virus flourishes in; it exists to answer why some diseases are Long while others are short. It suggests that once a pathogen’s tropism becomes wide enough to access the vascular system or the deep organs, it hits an inflection point. The illness ceases to be a compartmentalized infection and becomes a trigger for interactome collapse. Once that threshold is crossed, the body’s interconnected networks take over, and the ensuing damage is ubiquitous and self-sustaining.
There is a strange, cyclical irony to this. Moving away from the microscope and focusing entirely on systemic cascades feels distinctly ancient.
You can draw philosophical parallels between this modern systems approach and ancient frameworks like Humoral Theory or Traditional Chinese Medicine (TCM). We have replaced the mystical flow of Qi or Bile with the scientifically grounded flow of cytokines, T-cells, and endothelial signaling. But the underlying philosophy is identical. Ancient physicians did not possess microscopes; they could not observe the specific pathogen or obsess over its structural identity. Out of necessity, they practiced systems medicine. They developed diagnostic vocabularies based entirely on observing network imbalances and downstream physiological failures.
Modern medicine, armed with highly advanced diagnostics, has largely rejected this holistic network approach in favor of specialization. And not without reason. Network Medicine remains highly controversial in the modern clinical landscape because its failures directly mirror the historical failures of TCM. While it provides a beautiful, sweeping framework for understanding how diseases cascade, it struggles immensely with verifiability and intervention. Just as ancient physicians could not surgically fix a blockage of Qi, modern doctors cannot write a prescription to fix an interactome collapse. Abstract theories about systemic cascades do not readily produce the actionable predictions, targeted pills, or isolated biomarkers that modern clinical trials demand.
VI. The Credibility of Systems
We can glance at the intellectual history of another discipline—one where the credibility of systems has been a central, bruising point of debate for half a century.
In economics, the battle over how to model complex systems followed a distinct, three-act structure:
The 1960s (Big Systems): Post-WWII macroeconomics dominated. Economists attempted to model entire national economies from the top down, focusing on sweeping, systemic forces.
The 1990s (Micro-foundations): The discipline fractured and pivoted toward reductionism. The rise of DSGE (Dynamic Stochastic General Equilibrium) models insisted that you could only understand the system by modeling the rational behavior of the smallest individual agents.
The 2010s (Post-Financial Crisis): The 2008 crash shattered the credibility of reductionist models. The system collapsed in ways the micro-foundations completely failed to predict, forcing a reluctant return to the messy, increasing complexity of systemic models.
In medicine, the timeline begins much earlier, but the trajectory of the pendulum is identical.
Ancient medicine was entirely systemic. From Humoral Theory to metaphysical models, early physicians were forced to look at the body as an interconnected whole. Following the invention of the microscope, medicine swung violently toward reductionism. Things got smaller and smaller—from organs, to tissues, to cells, to pathogens.
But in both economics and medicine, the absolute victory of reductionism ultimately set the stage for its own failure.
In biology, this failure began to materialize with the sequencing of the human genome. We mapped the genetic code, expecting to find the blueprint for human being, only to discover that we knew less and less about gene expression, necessitating the creation of computational genomics.
A decade later, the Human Microbiome Project of the 2010s delivered another blow to reductionist logic, revealing that what we previously thought of as the biological boundaries of the human body were quite illusory. We are, at a fundamental level, walking ecosystems.
Depending on how the historical dust settles, COVID-19 might be remembered as the 2008 Financial Crisis of etiology—a catastrophic shock forcing an overly reductionist discipline to recognize what it already knew.
Having set the stage, we can finally draw back the curtain and look at the underlying philosophical divide: Medical Realism versus Medical Nominalism
Medical Realism is the intuitive idea that diseases exist independently in nature, like undiscovered islands, waiting for a scientist to find and name them. Medical Constructivism argues the opposite: diseases are human-made categories, taxonomic boxes created by societies to organize biological failures.
Constructivism has some heavyweight thinkers behind it, and when applied to the chaotic landscape of Post-Acute Infection Syndromes, their critiques sound remarkably incisive:
Michel Foucault (The Birth of the Clinic, 1963) traced how medicine shifted in the 18th and 19th centuries from observing the whole patient to searching for the specific, localized cause of suffering. He argued this shift trained doctors to believe the catalyst (the lesion, the germ) was the disease itself.
Ludwik Fleck (Genesis and Development of a Scientific Fact, 1935) used the history of syphilis to show that a disease does not possess a single, eternal identity. Syphilis evolved in the human consciousness from an astrological curse, to a cluster of skin symptoms, to a blood disease, and finally to an infection by the Treponema pallidum bacterium.
H. Tristram Engelhardt Jr., a modern bioethicist, bluntly argued that diseases do not exist in nature the way rocks or trees do. Instead, they are explanatory models designed to achieve specific clinical, social, and economic goals.
When Foucault warns that the clinic equates the catalyst with the disease, it perfectly describes the trap of Long COVID.
And yet, consider how incredibly well Medical Realism works.
The primary pushback against these philosophical critiques is that they do not sufficiently explain why it is so deeply attractive to view the catalyst as the disease itself. To do that, you need to engage with the actual, physical nature of life.
Notice that these critiques are made by philosophers who are applying their broadly constructivist worldviews to the problem at hand. Foucault was obsessed with power and institutions; naturally, he viewed the clinic as a theater of power. Fleck was a sociologist of science; he saw the influence of thought collectives. Engelhardt was a pragmatist; he saw utilitarian tools.
Because they approached medicine from the humanities, they tended to view human biology as a text to be interpreted rather than a physical reality asserting itself.
Let’s be honest here. Medical Realism didn’t become the dominant paradigm just because of a sociological shift in the clinical gaze, as Foucault argued.
It won because of penicillin.
VII. The Normal and the Pathological
In 1943, the French physician and philosopher Georges Canguilhem published The Normal and the Pathological. In it, he argued against the idea that a disease is a foreign entity invading a healthy body, or a quantitative deviation from a statistical norm—like a high white blood cell count or a spiked fever.
Instead, Canguilhem believed that pathology is an organism establishing a new, albeit restricted, biological norm to survive in a compromised state.
When you look at the patients suffering from Post-Acute Infection Syndromes, Canguilhem’s definition feels close to the truth of the matter. The fatigue, the autonomic misfiring, the forced pacing—this is the body surviving by shrinking its world. It is a restricted norm.
This philosophical middle ground resolves the Constructivist critique we explored earlier. But we are still missing a piece of the puzzle. If diseases are fundamentally systemic failures, why do they so often arrive in such small, self-contained packages?
Why does reductionism work so incredibly well biologically?
Viruses and bacteria are driven by a single imperative: maximum replication at all costs. To achieve this, they survive by living right on the edge of the error threshold—the maximum mutation rate they can sustain without their genome collapsing into nonsense. To maintain that blazing-fast mutation rate, their genetic code must be exceptionally tiny. They cannot afford dead weight. They shed any biological mechanism that does not serve their immediate need for replication.
By contrast, the human body is like a vastly intricate mechanical watch. It requires immense energy, overlapping redundancies, and constant genetic regulation just to maintain its baseline homeostatic ticking. The pathogen is a grain of sand. It doesn’t need to be complex; it just needs to be hard enough to strip a vital gear.
Reductionism works in medicine because removing a grain of sand is vastly easier than redesigning the watch. When a disease is triggered by a parsimonious, self-contained package—a bacterium, a parasite, a localized tumor—hunting down and destroying that package is the most efficient strategy. The Realist approach of targeting the invader saves millions of lives.
The Constructivists were right: the grain of sand isn’t the disease. The disease, as Canguilhem pointed out, is the whole entity—the watch steadily grinding its own gears down in an inflammatory attempt to chew through the sand.
Their theory holds. But this is an example of where the truth obscures without proper context.
We conclude this essay by noting that it is time to face the truth. The theoretical weaknesses of reductionism are no longer theoretical. The abstract criticisms of the Constructivists are quite tangible. The current reductionist paradigm is failing Long COVID patients and driving them toward pseudoscience.
But, what does a systems-level clinical approach actually look like in practice?
This is the fatal problem. Because a systems-level clinical approach, in an environment where the mechanisms are unclear, looks suspiciously like general wellness advice. How do you clinically support a vastly complex, cascading biological system when you don’t fully understand the feedback loops? You recommend rest, pacing, anti-inflammatory diets, stress reduction, and gentle rehabilitation.
It is theoretically consistent and a practical disaster.
We are fighting an uphill battle against human psychology. Patients do not want to be told to rest and breathe. They want penicillin. They want a specific, targeted cure that validates their suffering as a discrete, eradicable enemy. They want virology as their identity, because etiology as a system is an incredibly lonely place to be.
So while Long COVID may be a misnomer, an umbrella term masking a much older, broader failure of medicine, the Long Illness persists.