As a pulmonary specialist, I spend most of my clinical time in the hospital—which, during pandemic surges, has meant many long days treating critically ill COVID-19 patients in the ICU. But I also work in an outpatient clinic, where I also treat those same sorts of patients after they’re discharged: people who survived weeks-long hospitalizations but have been dealing ever since with lung damage. Such patients often face the same social and economic factors that made them vulnerable to COVID-19 to begin with, and they require attentive care.
Patients like these undoubtedly suffer what researchers have been calling post-acute sequelae of SARS-CoV-2, or PASC—which, according to one highly publicized recent CDC study, afflicts some 20 percent of COVID-19 survivors ages 18 to 64. Other studies have yielded lower estimates of the condition also called long COVID, and while differences in study methodology account for some of this variability, there’s a more fundamental issue eluding efforts to uncover the one “true” estimate of the likelihood of this condition. Quite simply, long COVID isn’t any one thing.
The wide spectrum of conditions that fall under the umbrella of long COVID impedes researchers’ ability to interpret estimates of national prevalence based on surveys of symptoms, which conflate different problems with different causes. More importantly, however, an incomplete and constrained perspective on what long COVID is or isn’t limits Americans’ understanding of who is suffering and why, and of what we can do to improve patients’ lives today.
The cases of long COVID that turn up in news reports, the medical literature, and in the offices of doctors like me fall into a few rough (and sometimes overlapping) categories. The first seems most readily explainable: the combination of organ damage, often profound physical debilitation, and poor mental health inflicted by severe pneumonia and resultant critical illness. This serious long-term COVID-19 complication gets relatively little media attention despite its severity. The coronavirus can cause acute respiratory distress syndrome, the gravest form of pneumonia, which can in turn provoke a spiral of inflammation and injury that can end up taking down virtually every organ. I have seen many such complications in the ICU: failing hearts, collapsed lungs, failed kidneys, brain hemorrhages, limbs cut off from blood flow, and more. More than 7 million COVID-19 hospitalizations occurred in the United States before the Omicron wave, suggesting that millions could be left with damaged lungs or complications of critical illness. Whether these patients’ needs for care and rehabilitation are being adequately (and equitably) met is unclear: Ensuring that they are is an urgent priority.
Recently, a second category of long COVID has made headlines. It includes the new onset of recognized medical conditions—like heart disease, a stroke, or a blood clot—after a mild COVID-19 infection. It might seem odd that an upper respiratory tract infection could trigger a heart attack. Yet this pattern has been well described after other common respiratory-virus infections, particularly influenza. Similarly, various types of infections can lead to blood clots in the legs, which can travel (dangerously) to the lungs. Respiratory infections are not hermetically sealed from the rest of the body; acute inflammation arising in one location can sometimes have consequences elsewhere.
But mild COVID-19 is so common that measuring the prevalence of such complications—which also regularly occur in people without COVID-19—can be tricky. Well-controlled investigations are needed to disentangle causation and correlation, particularly because social disadvantage is associated both with COVID exposure and illnesses of basically every organ system. Some such studies, which analyzed giant electronic-health-record databases, have suggested that even mild COVID-19 is at least correlated with a startlingly wide spectrum of seemingly every illness, including diabetes, asthma, and kidney failure; basically every type of heart disease; alcohol-, benzodiazepine-, and opioid-use disorders; and much more.
To be clear, this research generally suggests that such complications occur far less often after mild COVID-19 cases than severe ones, and the extent to which the coronavirus causes each such complication remains unclear. In other words, we can surmise that at least some of these complications (particularly vascular complications, which have been well-described in many studies) are likely a consequence of COVID-19, but we can’t say with certainty how many. And more importantly, we don’t yet understand why some people with mild COVID recover easily while others go on to experience such complications. However, an estimated 81 percent of Americans have now been infected at least once, so the public-health ramifications are large even if COVID causes only some of the aforementioned recognized diseases, and even if our individual risk of complications after a mild infection is modest. Regardless of cause, patients who do develop any such chronic diseases require attentive, ongoing medical care—a challenge in a nation where 30 million are uninsured and even more underinsured.
Another category of long COVID is something rather more quotidian, if still very distressing for those experiencing it: respiratory symptoms that last longer than expected after an acute upper-respiratory infection caused by the coronavirus, but that are not associated with lung damage, critical illness, or a new diagnosis like a heart attack or diabetes. Symptoms such as shortness of breath and chest pain are common months after run-of-the-mill pneumonia unconnected to the coronavirus, for instance, while many patients who contract non-COVID-related upper respiratory infections subsequently report a protracted cough or a lingering loss of their sense of smell. That a COVID-related airway infection sometimes has similar consequences only stands to reason.
However, none of these may be what most people think of when long COVID is invoked. Some may even argue that such syndromes are not, in fact, long COVID at all, even if they cause long-term suffering. “Long Covid is not a condition for which there are currently accepted objective diagnostic tests or biomarkers,” wrote Steven Phillips and Michelle Williams in the New England Journal of Medicine. “It is not blood clots, myocarditis, multisystem inflammatory disease, pneumonia, or any number of well-characterized conditions caused by Covid-19.” Instead, for some the term may invoke a chronic illness—a complex of numerous unexplained, potentially debilitating symptoms—even among those who may barely have felt sick with COVID in the acute phase. Symptoms may vary widely, and include severe fatigue, cognitive issues often described as brain fog, shortness of breath, “internal tremors,” gastrointestinal problems, palpitations, dizziness, and many other issues around the body—all typically following a mild acute respiratory infection. If the other forms of long COVID seem more easily explainable, this type is often characterized as a medical mystery.
Teasing apart which kind of long COVID a person has is important, both to advance our understanding of the illness and to best care for people. Yet lumping and splitting varieties of long COVID into categories is not easy. A given patient’s case might have features of more than one of the types that I’ve described here. Some patient advocates and researchers have tended to exclude patients in the first category—that is, survivors of protracted critical illness—from their conception of COVID long-haulers. I would argue that, insofar as we define long COVID as lasting damage and symptoms imposed by SARS-CoV-2, the full variety of severe long-term manifestations should be included in its scope. “Clinical phenotyping” studies now under way may eventually help scientists and doctors better understand the needs of different types of patients, but patients in all categories deserve better care today.
The biological mechanisms by which an acute coronavirus upper respiratory infection might lead to a bewildering range of chronic, burdensome symptoms even in the aftermath of mild infections are debated. Some scientists, for instance, believe that the virus causes an autoimmune disease akin to lupus. Meanwhile, one group of researchers has argued that even a mild respiratory infection from SARS-CoV-2 causes tiny clots to block tiny blood vessels all over the body, depriving tissues of oxygen throughout the body. Still others believe that the coronavirus causes a chronic infection, as such viruses as HIV or hepatitis C do. Meanwhile, some have emphasized the possibility of structural brain damage. While some published studies have provided support for each theory, none has been adequately validated as a central unifying thesis. Each is, however, worth continuing to explore.
A recently published investigation, conducted at the National Institutes of Health, suggests that clinicians and scientists should consider additional possibilities as potential drivers of symptoms for at least some patients. The researchers found far higher levels of physical symptoms and mental distress among subjects who had had COVID (many with long COVID) than among those who had not. Yet symptoms could not be explained by basically any test results: Researchers found effectively no substantive differences in markers of inflammation or immune activation, in objective neurocognitive testing, or in heart, lung, liver, or kidney function. And yet these patients were suffering from such symptoms as fatigue, shortness of breath, concentration and memory problems, chest pain, and more. Notably, researchers did not identify viral persistence in the bodies of patients reporting troublesome symptoms.
What this means in practice is that there are some people suffering from long COVID symptoms without evidence of structural damage to the body, autoimmunity, or chronic infection. Psychosocial strain and suffering, moreover, appears common in this population. Even pointing this out is sensitive territory—it leads some people to wrongly suggest that long COVID is less severe or concerning than those suffering from it describe, or even to question the reality of the illness. And, understandably, the invocation of psychosocial factors as potential contributing factors to suffering for some individuals may make patients feel as though they are being second-guessed. The reality, though, is that psychosocial strain is an important driver of physical symptoms and suffering—one that clinicians should treat with empathy. All suffering, after all, is ultimately produced and perceived in one place: our brain.
Severe depression, for instance, can inflict debilitating and severe physical symptoms of every sort, including crushing fatigue and withering brain fog, and is itself linked to having had COVID-19. And notably, a recent study in JAMA Psychiatry found that pre-infection psychosocial distress—e.g. depression, anxiety, or loneliness—was associated with a 30–50 percent increase in the risk of long COVID among those infected, even after adjustment for various factors. A false separation of brain and body has long plagued medicine, but it does not reflect biological reality: After all, diverse neuropsychiatric processes are associated with numerous “physical” changes, ranging from reduced blood flow to the brain to high (or low) levels of the stress hormone cortisol.
Illnesses of any cause that result in protracted time off one’s feet can also instigate (likely in conjunction with other factors) reversible cardiovascular deconditioning, wherein the blood volume contracts and the amount of blood ejected by the heart with each squeeze falls—changes that can lead to a racing heart rate or faintness when standing, as decades of studies have shown. Diverse neurological symptoms can also be produced by a glitch in the function rather than the structure of the brain—or what has been described as problems of brain “software” rather than “hardware”—resulting in conditions known as functional neurological disorders. Similar glitches, known as functional respiratory disorders, can disturb our breathing patterns or cause shortness of breath, even when our lungs are structurally normal. My point is not to speculate on some overarching hypothesis to explain all symptoms among all patients with long COVID. The whole point is that there’s unlikely to be just one. And there is still much to learn.
Research is underway to better understand this spectrum of illnesses, and their causes. But whichever diverse factors might be contributing to patients’ symptoms, we can take steps—both among clinicians and as a society—to improve lives now. Social supports can be as important as medical interventions: For those unable to work, qualification for disability assistance should not depend on a particular lab or lung-function test result. All patients with long-COVID symptoms deserve and require high-quality medical care without onerous cost barriers that may bankrupt them, which further compounds suffering. Universal healthcare is, that is to say, desperately needed to respond to this pandemic and its aftermath.
Additionally, while no specific long-COVID medications have emerged, some treatments may be helpful for improving certain symptoms regardless of the specific type of illness, such as physical rehabilitative treatments for those with shortness of breath or reduced exercise tolerance. Ensuring universal access to such specialized rehabilitative care is essential as we enter the next stage of this pandemic. So is helping patients avoid the emerging cottage industry of dodgy providers hawking unproven long-COVID therapies. Health-care professionals also need more education about the broad spectrum of COVID-19-related issues, both to improve care and reduce stigmatization of patients with all types of this illness.
Doctors and scientists still have much to learn about symptoms that continue—or first turn up—months or weeks after an initial COVID infection. What’s clear today is that long COVID can be many different things. That may confound our efforts to categorize it and discuss its implications, but the sheer variety should not get in the way of care for all who are suffering.