{"paper_id":"d23a8f56-a111-497d-925f-15247ced0d97","body_text":"Shortly after the onset of the COVID-19 pandemic, it became clear that some patients remained unwell for months to years after “recovering” from the acute infection. They suffered from fatigue, cognitive problems, headaches, disrupted sleep, myalgias and arthralgias, post-exertional malaise, orthostatic intolerance, tachyarrhythmias and gastrointestinal complaints, all of which greatly interfered with their ability to function at home and at work. In this post-COVID period, the most severely affected people were housebound and disabled. This was true even of people who had not been severely ill with acute COVID-19. The illness has been given several names, including post-acute SARS-CoV-2 sequelae (PASC), post-COVID-19 condition (PCC), post-acute COVID-19 syndrome (PACS), and “Long COVID”—the name we use in this review. We distinguish it from other post-COVID conditions in  Table 1 , below.\nPost-acute sequelae of SARS-COV-2 infection (PASC).\nIn this review, we compare the symptoms of Long COVID and ME/CFS, noting considerable similarities and some differences. Early in the study of both illnesses, the lack of objective biomarkers led some to question whether the illnesses were “real”—whether people might be imagining or even fabricating their symptoms.\nIn this review we summarize the emerging evidence that, in fact, there are many underlying biological abnormalities reported in both illnesses, documented by multiple laboratories. Moreover, we show that the two illnesses  share  many of these underlying abnormalities, just as they share many symptoms. That is, both illnesses are “real,” and both share similar biological abnormalities.\nUnderstanding the underlying biology of these illnesses is critically important, given the burden they are placing on all societies. The National Academy of Medicine and Centers for Disease Control and Prevention (CDC) estimate that, in the U.S., ME/CFS affects up to 2.5 million people and generates direct and indirect expenses of approximately $17–24 billion annually ( 1 ). It also has been reported in many countries around the world.\nPost-COVID-19 conditions may affect 65 million people, globally ( 2 ). In the U.S., nearly 2% of the total civilian labor force is unable to work because of these illnesses, with foregone wages of $170–230 billion annually ( 3 ). Senior economists have estimated that the aggregate cost of medical care, lost productivity and disability may be $3.7 trillion over the next 5 years ( 4 ,  5 ). The most important, and unanswered, question regarding the economic impact of Long COVID is how long it will last. If, as with most people with ME/CFS, it lasts decades, the burden will be even greater than this projection. Because of the exceptional anticipated burden, the National Institutes of Health has dedicated over $1 billion in support of studies of the underlying biology and natural history of various post-COVID conditions.\nIn  Tables 4 – 7  we try to efficiently summarize the findings from the large body of ME/CFS and post-COVID-19 literature. Our intention is to provide both an overview to general readers and a road map to a very large literature for investigators interested in pursuing the role of specific abnormalities in either or both illnesses. In summarizing what is known about biological abnormalities in both illnesses, we have cited both the “positive” studies that find an abnormality and the “negative” ones that do not. This allows readers to assess the strength (or weakness) of a reported finding.\nAs in any review, the studies we include are of varying size and quality. It is our hope that the road map we provide will help all interested readers judge for themselves how solid the evidence is for each of the underlying biological abnormalities, in each of the two illnesses, and where holes exist in our knowledge that need to be filled.\n\nMost people recover completely from  acute  COVID-19. However, others develop a variety of different post-acute sequelae of SARS-CoV-2 infection (PASC). The terminology for these different endotypes has not been standardized, in part because the detailed longitudinal studies necessary to generate the empirical data by which to define and distinguish different PASC endotypes have not yet been completed.\nIn  Table 1 , we propose three broad categories of PASC. Although we mention the first two categories for completeness, this review discusses  only  the third category—the condition called Long COVID. Some authorities refer to Long COVID as “PASC.” We think that is imprecise, since the tissue injury to multiple organs and the new onset of major diseases listed in  Table 1  surely also are “sequelae” of COVID-19. In any event, the data summarized in  Tables 3 – 7  of this review refer only to Long COVID.\nThe symptoms of Long COVID limit an individual’s ability to function at home or at work. The CDC states that in the “post-COVID condition” (which we take as equivalent to “Long COVID”) symptoms can be present four or more weeks after infection with SARS-CoV-2. The World Health Organization defines the condition as the continuation or development of new symptoms 3 months after the initial SARS-CoV-2 infection and lasting for at least 2 months with no other explanation.\nSome evidence indicates that Long COVID is more likely to develop in: people who are sickest and have the greatest evidence of inflammation with acute COVID-19; were PCR positive; are female; and have a premorbid history of asthma, chronic obstructive pulmonary disease and depression. However, many people with Long COVID have none of these risk factors ( 6 – 8 ).\nHow frequently does Long COVID follow acute COVID-19? A wide range of values for the incidence of Long COVID following infection with SARS-CoV-2 has been reported, depending on how the investigators have defined Long COVID and how rigorously they have pursued alternative explanations for the persisting symptoms.\nA meta-analysis of 57 studies involving over 250,000 people found that ongoing symptoms impairing functional mobility persisted in 43% of people with acute COVID-19 for at least  6 months  after acute infection ( 511 ). A study of nearly 100,000 cases and matched never-infected controls found that over 40% of cases remained with persistent symptoms and impairment across all daily activities at  18 months  ( 8 ). A study of six thousand hospitalized patients compared to uninfected matched controls found persistent symptoms in approximately 30% of cases—much more often than in uninfected controls—at least  2 years  after acute infection ( 512 ). Among people with persistent, debilitating symptoms following acute COVID-19, an estimated 13–45% meet the National Academy of Medicine case definition for ME/CFS ( 447 ,  513 – 515 ).\nInterest in ME/CFS (originally called “chronic fatigue syndrome”) surged in the mid-1980s, although a very similar illness has been described in the medical literature for several centuries. In the 19th and early 20th centuries the name “neurasthenia” was used to refer to a similar illness ( 516 ,  517 ). Since the 1980s, the names “chronic fatigue syndrome” and “systemic exertional intolerance disease” have been used. The illness also has several case definitions. We prefer the one proposed by the U.S. National Academy of Medicine ( 1 ), that is summarized in  Table 2 .\nU.S. National Academy of Medicine case definition for ME/CFS.\nMyalgic encephalomyelitis/CFS often, but not always, follows in the wake of an apparent “infectious-like” illness characterized by respiratory and gastrointestinal symptoms, fatigue, myalgias and other symptoms as well as fever and lymphadenopathy ( 518 ). This “infectious-like” illness often is little different, initially, from the common, transient infectious illnesses that most people experience throughout life. It is not standard medical practice to test for the responsible infectious agent in people with common and transient infectious illnesses. Thus, typically no testing has been done to determine the cause of the initial “infectious-like” illness that then becomes a chronic illness in subsequent months and years.\nAlthough in this review we are describing cases of ME/CFS that occur sporadically, there also have been apparent epidemics of a similar illness described in communities ( 141 ,  519 – 523 ) and in large hospitals over the past century ( 524 ). In one report, 6% of a community of 2,500 people—most young or middle-aged adults—became suddenly ill with an infectious-like illness followed by at least 5 months of a cyclic, debilitating illness characterized by fatigue, cognitive impairment (confirmed on objective testing) and pain ( 520 ). Details on symptoms, physical examination findings and laboratory test results in these studies are not sufficient to determine how similar these illnesses were to ME/CFS.\nThe severity of the symptoms, and the functional impairment, can range widely from one person to another. The functional impairment in people with ME/CFS may be even greater than in those with congestive heart failure and major depression ( 525 ,  526 ), and greater than in those with Long COVID ( 527 ). Some people remain able to fulfill their main responsibilities at work and at home, although hobbled. Others are bed-ridden or housebound, and unable to work. For most individuals with ME/CFS the symptoms are cyclic, with some relatively “good” days and frequent “bad” days. Several stressors—exercise, prolonged upright position, cognitive and emotional upset—typically produce a worsening of all of the symptoms of the illness. This condition, called post-exertional malaise, is a cardinal feature of the illness ( 1 ).\nAttempts to identify a single and possibly novel infectious agent as the cause of most cases of ME/CFS have been unsuccessful. For example, claims that murine leukemia viruses cause ME/CFS have been refuted ( 528 ,  529 ), as have similar claims for Borna disease virus ( 530 ).\nSeveral other conditions often are present in some people with ME/CFS: autonomic dysfunction (particularly postural orthostatic tachycardia syndrome), various rheumatologic conditions (particularly fibromyalgia), neurologic deficits (such as sensory hypersensitivity and small fiber neuropathy), secondary depression and anxiety, new or worsened allergic disorders, and endometriosis ( 531 ,  532 ).\n\nThe symptoms reported by people with both ME/CFS and Long COVID have been integrated in a recent meta-analysis of 21 studies ( 17 ). In  Table 3 , we summarize the long list of symptoms shared by these two illnesses (and shared with other illnesses). The cardinal symptom of ME/CFS, post-exertional malaise, also is reported by the vast majority of people with Long COVID ( 533 ,  534 ). As seen in  Table 3 , most symptoms in the two illnesses are similar, although decreased smell and taste, rash and hair loss are frequent in Long COVID but only rarely reported by people with ME/CFS, and painful lymph nodes, chemical sensitivities and tinnitus are frequent in ME/CFS but rarely reported in Long COVID.\nComparison of symptoms, ME/CFS, and Long COVID.\nME/CFS, myalgic encephalomyelitis/chronic fatigue syndrome. Adapted from: Wong DJ ( 17 ).\n\nA wide range of objective abnormalities have been reported in both Long COVID and ME/CFS. Multi-organ magnetic resonance imaging (MRI) imaging of many Long COVID patients and non-symptomatic post-COVID control subjects has revealed abnormalities involving many organs up to 1 year following acute COVID-19, although it is unclear if the  anatomic  MRI abnormalities reflect defective organ  function : correlation of MRI abnormalities with persistent symptoms was poor ( 535 ).\nProbably with both illnesses, and certainly with ME/CFS, underlying biological abnormalities appear to change over time. Initially, there is activation of neuroendocrine axes and cytokine production, followed after several years by an apparent “exhaustion” of neuroendocrine and immunologic activity ( 260 ,  536 ).\nThe most compelling findings thus far have been described in the central and autonomic nervous system, immune system, infectious agents, metabolism, and the cardiopulmonary system. We will focus on those abnormalities in this review.\nTable 4  summarizes the different abnormalities that have been identified in the central nervous system (CNS) and autonomic nervous system (ANS) in ME/CFS and in Long COVID. Most of them have been reported by multiple laboratories and are common to both illnesses. Often, the measured abnormality has been shown to be correlated with the presence and severity of symptoms. We comment in the text only on those abnormalities where the evidence is most extensive.\nNeurologic/neuromuscular abnormalities.\nPsychometric testing has revealed cognitive deficits in both illnesses. In people with ME/CFS, testing typically has been performed years after the illness began: there have been few longitudinal studies of the trajectory of cognitive deficits. In contrast, because the inciting infectious agent is clear in Long COVID, and because longitudinal studies were launched shortly after the onset of the pandemic, it is clear that the initial cognitive deficits can persist for many months. One study found there may be some improvement by the end of the first year ( 537 ).\nWith both illnesses, the most consistently observed abnormalities have been impaired attention and information processing speed. Long-lasting objective cognitive impairment has been demonstrated even in people with mild acute COVID-19 ( 538 ,  539 ). One study estimated that the measured cognitive deficit was equivalent to 10 years of aging ( 540 ). Measured cognitive impairment correlated significantly with abnormalities seen with EEG and MRI studies ( 50 ,  51 ).\nThe most persuasive evidence that the persistent cognitive deficits in Long COVID reflect damage to the CNS comes from a large population-based longitudinal study in which MRI scans and cognitive testing were performed both before the pandemic and after. Compared to a matched group that did not develop COVID-19, those who developed COVID-19 (even those not requiring hospitalization) had greater reduction in gray matter thickness and in global brain size, and greater cognitive decline ( 541 ).\nIn both illnesses, there is a reduction in cerebral blood flow. This likely results from autonomic dysfunction ( Table 4 ) and possibly from reduced blood volume ( Table 7 ). It may contribute to some of the symptoms of both illnesses, including fatigue and impaired cognition; it may also reflect reduced physical activity and reduced oxygen consumption (hypometabolism).\nThe first neurological abnormalities to be documented in ME/CFS were a group of different neuroendocrine abnormalities. In particular, there appears to be down-regulation of activity in several hypothalamic-pituitary axes ( Table 4 ) in ME/CFS, with some similar reports in Long COVID ( 542 ). These neuroendocrine changes could have bidirectional interactions with various immunologic and vascular abnormalities that also are seen in both illnesses ( 542 ).\nMany of the common symptoms seen in people with both illnesses could reflect autonomic dysfunction. In addition, autonomic dysfunction could explain other underlying biological disturbances, such as the reported neurovascular abnormalities. The autonomic dysfunction, in turn, could be caused by other biological abnormalities seen in both illnesses, particularly antibodies directed at autonomic receptors ( 129 ) and biopsy-documented small-fiber neuropathy ( 72 ,  180 – 182 ).\nAs summarized in  Table 4 , autonomic abnormalities have been well documented in both illnesses. Abnormalities of both the sympathetic and parasympathetic arms of the autonomic nervous system reflect “dyshomeostasis” ( 130 ): poor modulation of the balance between the two systems, with the imbalance favoring expression of the sympathetic system ( 72 ,  73 ,  131 – 134 ). Objective autonomic dysfunction is common in the first 6–12 months following acute COVID-19, becoming less common by 24 months following infection, in comparison to matched comparison groups without COVID-19 ( 73 ).\nMagnetic resonance abnormalities involving both gray matter and white matter have been found in both illnesses and been shown to correlate positively with measured objective cognitive impairment ( 50 ,  51 ,  149 ,  150 ). Impaired responses to a variety of challenges have been revealed by functional MRI.\nMyalgic encephalomyelitis/CFS often follows in the wake of an “infectious-like” illness. Long COVID (by definition) follows in the wake of acute infection with SARS-CoV-2. As summarized in  Table 5 , a variety of immunological parameters distinguish people with ME/CFS from healthy control subjects of the same age and gender. The same is true of Long COVID: immune parameters distinguish patients from post-COVID patients who no longer have symptoms, and from non-infected controls subjects. Moreover, many of the  same  immune parameters distinguish patients with ME/CFS and with Long COVID from comparison/control groups.\nImmunologic findings and infectious agents.\nIn addition, as summarized in  Table 5 , differences in the frequency of several reactivated herpesviruses, and differences in the gut microbiome, distinguish patients with these two illnesses from comparison/control groups.\nBeginning in the 1990s, evidence began to accumulate indicating that in people with ME/CFS who experienced a lack of “energy,” a contributing factor might be a cellular failure to generate and utilize adenosine triphosphate (ATP). As summarized in  Table 6 , considerable evidence has since emerged in support of that hypothesis. The ability to generate energy from multiple sources is impaired: from fatty acids, amino acids, glucose and oxygen. Moreover, similar evidence is emerging in people with Long COVID.\nMetabolic abnormalities.\nIn addition to energy metabolic deficits, evidence of a systemic hypometabolic state (also manifest in the brain,  Table 4 ), abnormalities in redox balance, and abnormalities in the kynurenine pathway have emerged in both illnesses, as summarized in  Table 6 . Later, we discuss briefly how abnormalities in redox balance may have bidirectional connections to abnormalities in the immune response and energy generation.\nAlthough less extensively studied than neurological, immunological, infectious or metabolic abnormalities, a growing number of cardiopulmonary abnormalities have been identified, as summarized in  Table 7 . The most well documented abnormalities are diminished exercise capacity on exercise testing, particularly when a second exercise test is performed 24 h after the first; reduced ventilatory efficiency; and endothelial dysfunction (particularly in Long COVID but also in ME/CFS). Some of the metabolic abnormalities noted in  Table 6  are provoked by exercise, and thus there is some overlap between  Tables 6 ,  7 .\nCardiopulmonary and vascular abnormalities.\n\nAs summarized in  Table 3 , most of the symptoms reported with ME/CFS and Long COVID are similar. Decreased smell and taste, rash and hair loss are more likely in Long COVID than ME/CFS; this may reflect pathology induced specifically by SARS-CoV-2 ( 543 ).\nNot only is a common core of symptoms shared by ME/CFS and Long COVID: these same symptoms also are also reported following multiple infectious illnesses ( 544 ) and major  non-infectious  injury such as post-critical illness syndrome or post-intensive care unit syndrome ( 545 ,  546 ), including heat stroke ( 547 ).\nIn this review we have compared what is known about the underlying biology of ME/CFS and Long COVID. Both clearly are systemic illnesses involving multiple organs and physiological systems. Long COVID is triggered by infection with SARS-CoV-2. ME/CFS often is triggered by an “infectious-like” illness. We think it is unlikely that ME/CFS is triggered by a single, novel infectious agent: more likely, it represents a dysfunctional response to infection with any of multiple agents, as recently described ( 544 ).\nAs summarized in this review, both illnesses share abnormalities involving the central and autonomic nervous systems, the immune system, reactivation of latent infectious agents (primarily herpesviruses), the gut microbiome, energy metabolism, a hypometabolic state, redox imbalance, and various cardiac, pulmonary and vascular abnormalities.\nMany of these abnormalities bidirectionally influence each other. This creates the potential for multiple, self-reinforcing “vicious” pathophysiological cycles that could lead to persisting illness ( 548 ,  549 ). It also means that the precipitating event, which sets in motion those vicious cycles, may be different in one person with the illness from the precipitating event in another person.\nIn summarizing what is known about biological abnormalities in both illnesses, we have cited both the “positive” studies that find an abnormality and the “negative” ones that do not. These citations allow readers to assess the strength (or weakness) of a reported finding.\nAs shown by the tables, some reported findings are supported by robust evidence—confirmatory reports from multiple laboratories: even though there may be some dissenting reports, the preponderance of the evidence supports the finding. The tables also identify abnormalities that are not supported at this time by robust evidence and require future investigation.\nThe similar symptoms and pathology of ME/CFS and Long COVID raise the question of whether these disorders represent just two examples of a broader illness in which symptoms occur because they are generated by a carefully orchestrated, stereotyped, multi-system response to infection and injury.\nWhy might such a carefully orchestrated group of symptoms be generated following infection or injury? We speculate, as have others ( 550 – 552 ), that these symptoms are generated because they lead to metabolic reprogramming ( 553 ) as well as to behavioral changes that reduce non-essential, energy-consuming activities—thereby maximizing the amount of energy available to facilitate recovery. Fatigue, myalgia and orthostatic intolerance, for example, lead to reduced physical activity, redirecting energy stores to eradicate infection and heal tissue injury.\nSuch a carefully orchestrated response to vital threats exists throughout the animal kingdom. The best studied examples are hibernation and torpor in “higher” animals ( 554 – 556 ), and the larval state of dauer in the worm  C. elegans  ( 557 – 559 ).\nDauer, hibernation and torpor all involve “abnormal” innate immune responses, redox imbalance, increased glycolysis, decreased aerobic respiration, and possibly even alteration in the organisms’ microbiome. They also involve orchestration by the autonomic nervous system (and its counterparts in the nervous systems of more primitive organisms). Thus, these well-recognized “hunkering down” mechanisms all exhibit similarities to the emerging pathophysiology of ME/CFS and Long COVID, as summarized in this review.\n\nThe goal of this report is to provide a road map to the state of knowledge about the underlying biology of ME/CFS and Long COVID. The often-similar findings suggest that insights into each disorder will have implications for the other. They may also enhance our understanding of evolutionarily preserved biological responses that fight infection and heal injury. We urge that investigators studying the underlying biology of Long COVID take note of the robust findings in ME/CFS that have not yet been investigated in Long COVID: given the many similarities in the underlying biology of the two illnesses, it is likely that pursuing such abnormalities in Long COVID will prove instructive.\nResearch into the pathophysiology of these responses has the potential to lead to new strategies for reducing the morbidity of ME/CFS and Long COVID, and of similar illnesses that can follow a variety of infections and non-infectious traumatic injury.\n\nAK and WL conceptualized and wrote the manuscript. Both authors contributed to the article and approved the submitted version.","source_license":"CC-BY-4.0","license_restricted":false}