Thinking Carefully About “Long COVID”

Several months ago, a newspaper columnist reached out to me after noticing a few of my tweets on a condition — “Long COVID” — then, as now, in the news.  I’m a pulmonary and critical care physician, and had tweeted that researchers should show greater caution in making assertions about the still mysterious syndrome; the journalist asked whether I thought that medical professionals more broadly were somehow skeptical of the condition.  “I’m asking as much as a person as a journalist,” they messaged, “because I’m more terrified of this syndrome than I am of death…”

They are not the only one.  Media stories about “Long COVID,” and the patients who call themselves “COVID long-haulers,” have been appearing at an accelerating pace, and what they describe is frightening.  Articles by one of the most respected writers on the pandemic, Ed Yong at the Atlantic, have been particularly influential.  In “Long-Haulers Are Redefining COVID-19,” Yong describes a disease that strikes even those with “mild” COVID-19 — that is to say, those who don’t need hospitalization or oxygen, much less a ventilator, and who recover at home without medical intervention.  Or, rather, in this case, who don’t seem to recover.  One such individual, he notes, described some five months of “extreme fatigue, bulging veins, excessive bruising, an erratic heartbeat, short-term memory loss, gynecological problems, sensitivity to light and sounds, and brain fog.”  For some, Yong noted, “months of illness could turn into years of disability.”  

But can, indeed, a “mild” Sars-Cov-2 infection cause such a debilitating (and potentially lifelong), mysterious, multi-organ syndrome?  Can it, as other news reports have suggested, lead to severe psychosis or a dementia-like condition months after a mild and otherwise resolved bout?  To even begin to answer these questions, we have to recognize that what is being described as “long-COVID” often conflates several distinct entities that are worth differentiating, even if they sometimes overlap.  As important, we need greater acknowledgement that the medical harm of this pandemic goes beyond the direct damage inflicted by the SARS-CoV-2 virus itself.

“Long COVID” has no universal definition, but is often used to describe those whose symptoms last more than a few weeks or months after a bout with COVID-19.  That many individuals have such protracted symptoms after this infection is hardly surprising.  After all, critical illness of any cause can be devastating.  Almost everyone who dies of COVID-19 develops a condition called “acute respiratory distress syndrome” (ARDS), a particularly severe form of pneumonia involving severe inflammation of both lungs that often requires mechanical ventilation. Many studies have shown that ARDS can have long-term effects on the body: one 2003 study found that many survivors of ARDS, at a year out from discharge from the ICU, had significant physical impairment; lung function improved over time but remained abnormal for some.  Meanwhile, as the lungs fail in ARDS, other organs often go down with it, including the kidneys and the heart.  In other words, while many with severe COVID-19 make good recoveries, other survivors of this (and other) critical illnesses can require specialized rehabilitation or care for months or even years.  There is no mystery or surprise here.

At the same time, however, some people have also had protracted symptoms following less severe COVID-19.  An Italian study showed that most patients still had at least one symptom, like fatigue or shortness of breath, two months out from a hospitalization for COVID-19. A more recent Lancet study found that six months following hospitalization for COVID-19, many patients had persistent symptoms, including fatigue and anxiety, and some (particularly those with more severe acute cases) had reductions in lung function.  Again, however, this is not qualitatively specific to COVID-19: recovery from pneumonia from bacteria or other respiratory viruses is often slow. According to one 1997 study, many individuals with pneumonia have symptoms such as cough, fatigue, and shortness of breath that can persist for three months (or longer) after onset.  And even recovery from an upper respiratory tract (URI) infection — a cold — can take a while.  A protracted “post-viral” cough following a cold is a common reason for a referral to a pulmonologist.  The experience of an acute respiratory illness will be different from person to person, and there is no “rule” on when exactly a patient’s respiratory symptoms, fatigue, and general “sick feeling” should resolve by. 

It is also true, however, that we have also seen rare cases of permanent organ damage inflicted by SARS-CoV-2 — including instances of severe myocarditis, inflammation of the heart muscle leading to heart failure and a condition called cardiomyopathy, and strokes — even among those who did not have severe pneumonia, or even any evident respiratory disease.  These can also (rarely) be caused by infections from other respiratory viruses (viruses are a leading cause of acute myocarditis in general).  But in any event, these are specific, diagnosable complications affecting specific organs, not a multi-systemic medical mystery. And outside the critically-ill population, who — as noted, can suffer basically every form of organ failure — they are very uncommon.

Yet if these ailments are sometimes included in reports of “long COVID” in the media, most narratives evoke something entirely different:  a debilitating syndrome seemingly affecting multiple organ systems for months on end — and perhaps indefinitely — but absent any specific diagnosis such as cardiomyopathy or stroke.  As noted, these individuals are often described as having what would be characterized as a mild, flu-like COVID-19 illness, and to have not been hospitalized.  Their plight is what the columnist who contacted me, and many others I’ve spoken to, think of when “Long Covid” is discussed — and it’s the focus of reporting like that of Yong’s.

What are we to make of this entity?  

Reasons to Slow Down on the Long-COVID Narrative

Media reports exploring this type of Long COVID have been appearing at a steady clip.  The symptoms of this condition are often, if not mostly, non-respiratory in nature, and the demographics of those affected are basically the opposite of those most at risk from dying from COVID in some cohorts.  For instance, whereas severe COVID-19 disproportionately strikes Black, Hispanic, low-income, and older individuals, “long-haulers” appear to have distinct demographics: a recent preprint study, for instance, identified a long-COVID cohort that is predominantly white, high-income, and non-elderly.  The headline of a story in Buzzfeed in August captures this demographic divergence.  “COVID Is Making Younger, Healthy People Debilitatingly Sick For Months. Now They’re Fighting for Recognition.” More than a hundred of the “long-haulers” described to that reporter a complex of unremitting symptoms suggestive of a mostly-non-respiratory, multi-organ chronic process: palpitations, headaches, severe fatigue, difficulty sleeping, hair loss, and, again, brain fog. 

Some news reports, meanwhile, have described even more alarming symptoms.  In October, a story in the New York Times described a dementia-like illness following a mild infection:  “It’s becoming known as Covid brain fog: troubling cognitive symptoms that can include memory loss, confusion, difficulty focusing, dizziness and grasping for everyday words.” And last month, another Times article went further still, asserting a linkage between mild coronavirus infection and “severe psychotic symptoms.”  One individual described in the article had a minor coronavirus infection in the spring that had completely resolved.  But then, “months later, she heard a voice that first told her to kill herself and then told her to kill her children.” Such stories, no doubt, chill readers to the bone.  

But, for several reasons, such reporting needs to be more cautious. 

First, a cause-and-effect relationship is typically unestablished in these articles.  The Times article contending that mild and resolved COVID-19 infections can lead to extreme psychosis months later left out some important context.  According to one international study, the incidence of psychotic disorders is around 27 per 100,000 persons per year, which would suggest that in the US, there are tens of thousands of new diagnoses of psychosis every year.  In 2020, a solid proportion of those new diagnoses will have occurred in individuals with a prior coronavirus infection.  Obviously, although the temporal link will no doubt feel powerfully suggestive to patients and their doctors, this does not establish causality.

Another reason to question the causal link between the virus and some “Long COVID” symptoms stems from the fact that some, and perhaps many labelled with long COVID appear to never have been infected with the SARS-CoV-2 virus.  For instance, in his August Atlantic article, Yong cites a survey of COVID “long-haulers” that found that some two-thirds of these individuals had negative coronavirus antibody tests, which are blood tests that reveal prior infection.  Meanwhile, the aforementioned study published on a pre-print server, organized by a group of Long COVID patients named Body Politic that recruited participants from online long COVID support groups, similarly found that some two-thirds of the long-hauler study participants who had undergone serological testing reported negative results.

Now it’s no doubt true that while serological tests are reported as highly sensitive and specific, they are imperfect; there are both false positives and false negatives.  Hence, some of those classified as negative may have indeed a prior infection.  Moreover, there is growing evidence that antibodies can wane over time — to an extent.  For instance, the aforementioned Lancet study that examined long-term symptoms among individuals hospitalized with documented COVID-19 found that more than 90% of those tested at baseline still had positive IgG antibodies (the long-lasting type) six-months after infection.  A CDC study of healthcare workers, on the other hand, found that among a group of healthcare workers who had been screened for antibodies and found positive, 72% had positive antibodies around two months later, although the proportion was much higher (98%) among the subgroup with previous confirmed COVID-19.  Another preprint study, published by researchers from the Rockefeller University, examined the performance of four commonly used antibody tests for outpatients with documented SARS-CoV-2 infections.  About a month after a positive COVID-19 test, antibodies were positive in 95-100% individuals with three of four tests, and 85% with one of the four tests.  After 81 days (or more), sensitivity remained >90% for three out of four tests, but worsened to 71% for one test. 

So again, these tests are not perfect, but they do appear to remain positive for a majority of individuals with documented infections for months in most studies.  Hence, even accounting for antibody waning, it is highly improbable that 100% of a large group of seronegative individuals without a prior diagnosed infection are false negatives; some — and it’s hard to know how many — will be true negatives.  But why does this matter?  Well, for one thing, it makes clear that it is possible for at least some individuals to incorrectly attribute chronic symptoms to SARS-COV-2.  That’s hardly surprising, but there is more to be said of the issue. 

Consider that the Body Politic study of long-haulers not only reported that most individuals who had antibody testing were seronegative, but also that there was virtually no difference in symptoms whatsoever between those with a diagnosed SARS-COV-2 infection and those who were seronegative.  A finding of no association between a reasonably reliable serological marker of a prior acute infection and a particular complex of chronic symptoms has one rather likely explanation: that the acute infection may not be the predominant driver of the chronic symptoms in that cohort.

But if true, how could we explain the fact that such individuals might perceive a direct causal link? Quite easily: our ability, as individuals, to link our particular chronic symptoms with a past infection from a particular microbial agent — in this case, SARS-CoV-2 — is fallible.  After all, the symptoms reported as distinct to “Long COVID” are nonspecific, and could stem from any number of factors and diseases.  Before COVID-19, the majority of Americans, for instance, had one or more unpleasant gastrointestinal symptoms.  More than one in ten of us felt “very tired or exhausted” every or most days. Hair loss — which has been noted in several articles about long-COVID — can be caused by multiple medical issues and stressors.  Headache has been described as an “almost universal human experience”, while “chronic daily headaches” affect 4% of adults, or millions of people a year.  “Brain fog” is not rare, and certainly not specific to any particular condition.  For these reasons, it’s important that “long COVID” studies have adequate controls of some sort, which to date has typically not been the case.  But the bigger point is that physical suffering is, sadly, very common in our society.  That is the case today, but was also the case before the outbreak.  Many people develop new physical symptoms (which often come in groups) each year — and this past year will be no exception.  But even beyond the direct effects of the SARS-CoV-2 virus, there is reason to think that 2020-21 will be even worse than the typical year in this respect.  

The Role of Anguish

This brings me to the elephant in the room.  Could mental distress, anguish, and illness be one contributing factor behind new physical suffering for some individuals this past year?  It’s a dicey question to ask — the invocation of depression and anxiety as causes of physical symptoms is sometimes seen as a form of “medical gaslighting.”  It is, however, nothing of the sort.  There is no question that mental suffering can produce physical suffering that is often undistinguishable from any other.  Those who have had a severe panic attack know it can cause chest pain and shortness of breath — indeed, those are among the cardinal signs.  Such symptoms as weight change, sleeping problems, “psychomotor retardation”, persistent fatigue, and concentration problems (what one might call “brain fog”) are actual DSM-V criteria for major depression.  The physical manifestations of mental suffering can also be chronic and insidious.  A study in the New England Journal found that across multiple continents, about half of those with depression have unexplained physical symptoms, and for more than half, that physical symptoms actually predominated over the mental ones.  For some, in other words, the experience of psychological suffering may be mostly experienced physically.

And on that note, let’s acknowledge that this is a time of incredible trauma, sorrow, and hardship.  The loved ones of more than 400,000 Americans are in mourning.  Tens of millions have lost their jobs.  This has been a period of prolonged social isolation with no obvious parallel in history.  One study, published in early September, suggested that the proportion of Americans suffering from symptoms of depression has skyrocketed, from 8.5% in 2018 to 27.8% 2020.  Whatever the long-term effects of COVID-19 may or may not be for those with mild infections — and certainly there is much to still be learned — we should expect a surge in both mental anguish and physical suffering that, while connected to the epidemic, will not always be directly connected to viral infection itself.  

Needless to say, all such symptoms — all of this suffering — is in every sense real.  There will be, as is often the case, lack of clarity about the distinct cause of suffering in each individual; often, there may be multiple causes, stretching from the virologic to the psychosocial.  Every such patient of course deserves careful, empathetic evaluation and appropriate treatment and referrals.  And more broadly, policies to stem psychosocial suffering stemming from the pandemic and the recession and social isolation it has generated are urgently required. 

What we know is bad enough

But fear-instilling assertions about illness that go beyond the evidence at hand are never helpful: we should not be positing that a mild COVID-19 infection can have frightening chronic sequelae like severe psychosis or profound cognitive impairment, or lead to a potentially lifelong, debilitating and mysterious syndrome, without firm scientific evidence.  Unfounded linkages can also be medically harmful, leading to the provision of unproven and hazardous medical therapies, as we’ve seen in the case of patients with chronic symptoms and a prior diagnosis of Lyme disease.  Some of these patients have suffered serious consequences from long-term antibiotic treatments, and others have been the victim of unproven alternative therapies ranging from colloidal silver to stem cell transplantation by unscrupulous providers.  It is far from unlikely that we could see something similar for those labelled as having “long COVID,” again to these patients’ detriment.

Make no mistake, what we know about COVID-19 today is more than enough to justify doing what we can to quash it: it is the respiratory pandemic of a century.  And of course, rigorous research on its long-term effects, including among those with mild disease, is no doubt needed. At the same time, we need to start thinking more critically — and speaking more cautiously — about long-COVID.  

Note: A shorter, revised version of this post was published in STAT on March 22, 2021.