Growing Insight Into COVID’s Long-Term Brain Effects
When the pandemic began to accelerate in the spring of 2020, Michelle Monje was worried. Early reports showed that SARS-CoV-2 was a highly inflammatory virus, triggering a flood of immune molecules, known as cytokines. “That’s not what normal viruses do,” says Monje, a neurologist at Stanford University who studies neural inflammation and cognitive function. Given that level of inflammation, “I was worried that even in mild cases, we might see cognitive impairment.”
Monje’s concerns were soon borne out — post-COVID patients began streaming into neurologists’ offices complaining of fatigue, memory issues and difficulty concentrating, a constellation of symptoms we now associate with long COVID. For Monje, the pattern looked startlingly like the cognitive symptoms reported by some cancer patients after chemotherapy, known as chemo brain. Research by Monje and others had shown that neuroinflammation is a key culprit underlying cancer therapy’s cognitive effects, and she wondered if something similar was at play with long COVID. “I felt it was important to understand the molecular underpinnings, what the parallels really were,” Monje says. “If many of the same things were happening, it would be useful to know, because we have been working for years on ways to intervene.”
As COVID infections continue to sweep the globe, the number of patients reporting long COVID symptoms is on the rise. Prevalence estimates vary widely, ranging from 5 to 50 percent, due to differences in strains, the severity of infection and inconsistencies in how long COVID is defined. But with more than 600 million documented COVID cases to date, even the lowest estimates represent huge numbers of people. Cognitive symptoms are a common component of long COVID — one systematic review encompassing nearly 10,000 patients estimates that roughly 25 percent of people experience cognitive issues after COVID, even if they had mild infections. “The cognitive issues people report are striking, and we know very little about the long-lasting effects,” says Natalie Tronson, a neuroscientist at the University of Michigan. “Is there a recovery period? Will it continue? Will it get worse?”
Early research on animal models of long COVID point to issues with the brain’s immune system and blood supply, both of which decline with age and neurodegenerative disease. Scientists studying aging are particularly concerned, wondering how these different risk factors will intersect in the brain in old people recovering from COVID today and in younger people in the decades to come. “It’s extraordinarily important to understand how the natural aging process is accelerated or perturbed by having experienced infection, especially the immunological consequences,” says Sarah Lutz, a neuroimmunologist at the University of Illinois at Chicago. “Ten, 20, 30 years from now, the potential ramifications for people who encountered COVID or long COVID as young adults are unknown and potentially huge.”
We still know relatively little about what happens in the human brain after COVID. Autopsy studies of patients who died with or from the infection show dying blood vessels and signs of severe neuroinflammation and disruption to the blood-brain barrier, which regulates which molecules move between the blood and the brain. While illuminating, these studies are few in number and often skewed toward severe cases, making it difficult to assess how broadly the findings apply. But growing evidence suggests that COVID-related neuropathology is more widespread, Lutz says. COVID patients with neurological symptoms have inflammatory cytokines in their cerebrospinal fluid, despite relatively little evidence of infection of brain cells. And a large-scale brain imaging study of participants in the U.K. Biobank found a small but significant decrease in individuals’ cognitive function after infection, as well as decreases in overall brain size and gray-matter thickness in certain regions, compared to controls.
Given SARS-CoV-2’s recent emergence, it’s not yet clear whether neurological symptoms will persist, or for how long. But previous research on other viruses shows that infections can have long-term consequences, altering how the brain’s immune system responds to insults even decades later. A small percentage of people who get the Epstein-Barr virus, for example, which causes mononucleosis, develop multiple sclerosis much later in life. More broadly, some epidemiological studies suggest that the more immune challenges people face in middle age, such as seasonal flu, the higher their risk of later dementia.
Early evidence suggests that’s true for COVID as well. A population study of 3 million people in Denmark found that those who had had COVID or influenza had a higher risk of being diagnosed with Alzheimer’s or Parkinson’s disease. And an analysis of more than 1 million health records, published in The Lancet in August, found that people were at higher risk of developing dementia, brain fog and other issues for up two years post-COVID, even compared to people who had other respiratory illnesses.
As with much epidemiological research, population studies of the long-term cognitive effects of infection can be difficult to interpret. Animal models offer a way to more directly explore the molecular and cellular underpinnings of these chronic issues. Research into better-established viruses is beginning to offer a picture of how peripheral infection can spur cognitive decline. Immune cells release cytokines that circulate in the blood, recruiting immune cells to the site of injury or infection. These signals act on the brain vasculature and can disrupt the blood-brain barrier, sparking an immune response in the brain. This in turn can trigger changes in synaptic plasticity and neurogenesis, all of which affect cognitive function. “More and more data supports the idea that for a lot of infection, even after it clears, there are longer-term sequelae in the brain and nervous system that might be mediated by immune cells,” says Beth Stevens, a neuroimmunologist at Boston Children’s Hospital and an investigator with the Simons Collaboration on Plasticity and the Aging Brain.
Clearing the fog
Monje’s recent work, published in Cell in July, is helping to uncover the neuroimmune and other processes that may accompany long COVID. The study was guided by Monje’s previous efforts to unpick the molecular mechanisms underlying chemo brain, which showed that chemotherapy can have a host of effects on the brain’s immune system, including resident immune cells called microglia, which can in turn alter neural circuits. In the healthy brain, microglia help tend to neural circuits. But when they sense infection or other challenges, microglia enter a reactive state designed to stop the spread of pathogens. Sometimes reactive microglia can veer out of control, sparking harmful downstream effects — impairing neurogenesis, triggering astrocyte reactivity and harming oligodendrocytes, all of which are essential for healthy neural circuit function.
To understand whether similar factors are at play in long COVID, Monje and collaborators studied mice engineered to carry a human version of the receptor that SARS-CoV-2 uses to infect cells. The receptor is expressed only in respiratory system tissue, so researchers could assess the neurological effects of mild respiratory COVID infections — the animals had mild symptoms and did not lose weight, clearing the infection in about a week. But seven weeks later, the mice had signs of inflammation in the brain, including reactive microglia in the hippocampus, a brain region essential for learning and memory. Post-COVID animals had lower levels of neurogenesis in the hippocampus, as well as fewer oligodendrocytes, the cells that provide axons with their myelin sheath. About 10 percent of axons had lost their myelin, Monje says, which could have profound cognitive effects. (Human imaging studies also suggest that people with COVID infections have irregularities in white matter.) The animals also had higher levels of a cytokine called CCL11, which increases with age. “I think this neuroinfIammation is really central to what’s going wrong,” Monje says. “Peripheral respiratory infection, even mild, induces a neuroinflammatory response that impairs myelin homeostasis and likely myelin plasticity, as well as hippocampal neurogenesis, which can lead to profound impairments in cognitive function.”
One major question is whether COVID’s behavior is typical of viral infections, or if it causes more chronic neurological issues. Few direct comparisons between SARS-CoV-2 and other viruses exist, so it’s not yet clear if it triggers a fundamentally different long-term neuroimmune reaction. “Based on the data we have now, we don’t know whether specific pathogens elicit different signals in the brain or converge on some common inflammatory cytokine pathway,” Stevens says. “The only way to know will be comparative studies.”
Monje’s study starts to get at this question by comparing coronavirus infection with H1N1 influenza, a relatively mild version of the flu for mice. Her work reveals some similarities and some differences. Post-flu mice also showed microglia reactivity and loss of oligodendrocytes in subcortical regions of the brain, but those effects were temporary. In the hippocampus, however, microglia reactivity persisted, as did loss of neurogenesis and higher CCL11 levels, similar to the pattern in COVID mice.
Monje suspects that much of what they are discovering about COVID’s long-term neurological effects will be relevant to aging. Microglia, for example, can become more reactive with age as well as more senescent, and both of these changes can increase inflammation, making the brain more vulnerable. (For more, see “In Senescent Cells, a Promising Route to Slowing Brain Aging.”) Monje and collaborators next plan to look at how age influences microglia reactivity and other inflammatory signals, as well as how immune insults early in life might predispose animals to later issues. Researchers will also explore how COVID infection might raise the risk of dementia and cognitive decline by assessing how it alters the onset of neuropathology and cognitive issues in a mouse model of Alzheimer’s disease.
Aging and infection
Multiple sclerosis offers a striking example of how age and prior infection can interact. “The way that MS patients age is different than the way other people age,” Lutz says. “There is superimposition of chronic inflammation that happens at the same time as the normal processes of aging.” Lutz has previously worked on MS, and COVID’s apparent similarities inspired her to explore how age affects COVID’s neurological consequences. “In particular, we were struck by the fact there were immune cells getting into the brain in COVID patients in a way reminiscent of acute neuroinflammatory disease,” she says.
Lutz’s study focused in part on the blood-brain barrier, which has been implicated in both aging and infection. Previous research shows that the blood-brain barrier can break down with age, allowing immune cells to enter the brain and offering a foothold for neuroinflammation. (For more, see “Searching for the Secret of How Young Blood Rejuvenates the Brain.”) Early pathology studies of COVID also point to the blood-brain barrier, noting signs of inflammation in the blood vessels of the brain.
Lutz and collaborators found that mice infected with a mouse-adapted version of SARS-CoV-2 showed highly age-dependent effects. Young animals recovered after a few days, exhibiting some cerebrovascular damage and memory impairment. In middle-aged mice, however, “everything is worse — and different,” Lutz says. According to research posted on the bioRxiv in June, older COVID-infected mice had severe cerebrovascular inflammation and signs of vascular regression, which happens when blood vessels lose their accompanying endothelial cells, blocking the vessel’s ability to ferry helpful molecules to the brain.
Like people who have had COVID infections, the mice showed blood proteins in the brain and loss of brain endothelial cells, both of which were significantly worse in middle-aged mice than in young mice, Lutz says. In addition to memory problems, older COVID-infected mice showed signs of anxiety and repetitive behavior, such as excessive grooming. Lutz speculates that the middle-aged animals fared worse because their blood vessels were less resilient. “When there is infection, the aged brain is less equipped to respond and has a worse outcome,” she says.
What happens in old mice is still an open question. Lutz’s study compared young and middle-aged mice, focusing on middle age in part because of the clinical observation that women in their 30s and 40s are most likely to suffer brain fog post-COVID. “It’s still wildly unknown how these things change in advanced age,” Lutz says. “We speculate that whatever we see exacerbated in middle age will be even more dramatic in aged animals.” Her team hopes to look at older animals, as well as how animals infected at different ages fare over time.
Lutz’s study and others highlight how age can worsen COVID outcomes. But what about decades after infection? Unlike peripheral immune cells, microglia can live for months or years, meaning any changes they experience early on can persist over time. Previous research has shown, for example, that microglia can have long-lasting memory. In research published in Nature in 2018, Jonas Neher, a neuroscientist at the German Center for Neurodegenerative Diseases in Tübingen, and collaborators found that mice treated with a substance that induces inflammation exhibited epigenetic changes to their brains’ microglia more than six months later, the equivalent of roughly 20 to 25 human years. These epigenetic changes altered both gene expression and cell behavior, and in turn influenced how the microglia responded to pathology developing in the brain. In a mouse model of Alzheimer’s disease, midlife inflammation boosted levels of beta-amyloid in the brain. “Normally microglia eat anything that’s not supposed to be there, such as amyloid-beta aggregates,” Neher says. “But the early reprogramming switched them from this phagocytic to a more inflammatory activity.”
Tronson and collaborators use a similar model to study the longer-term cognitive impacts of inflammation in mice. Her team has found that at midlife, two months after an immune challenge, the animals began exhibiting memory impairments. They also showed changes in the expression of genes involved in synaptic plasticity and neuroimmune function in the hippocampus.
When Tronson started hearing reports of long COVID, particularly memory impairment, she began adapting their protocol to more closely resemble the pattern of inflammation that follows COVID. “This isn’t COVID. It’s not a viral infection,” Tronson says. “That’s an advantage because we can start to separate the effect of the virus itself and the inflammatory consequences of the virus.” Tronson and collaborators are now looking at the effect of that immune challenge on the brain several months later by assessing changes in gene and protein expression, synaptic function, neuroimmune function, and memory over several months.
Tronson, Lutz, Monje and others hope that their efforts to pick apart COVID’s long-term neuroinflammatory effects will inspire ways to prevent and treat it. Dampening rampant inflammation offers one possible route to repair. “On a very basic level, it’s possible to suppress the enhanced immune response,” Neher says. “Signaling molecules that dampen the immune system might trigger changes that prevent immune cells from overreacting.” Monje’s work on chemo brain has shown that rescuing microglia, by depleting or resetting harmful forms of the cells, restores cellular and cognitive function.
Lutz and collaborators have identified another potential target — the WNT/β-catenin pathway, which plays a role in cell proliferation and other processes. “We and others have shown that if you enhance WNT/β-catenin signaling in brain endothelial cells, it can rescue or improve cerebrovascular inflammation and disease,” Lutz says. “If you block this pathway, the pathology is much worse.” In their COVID study, Lutz’s team found that young mice seemed to boost this pathway in response to infection, but middle-aged mice did not. Researchers don’t yet know whether a similar pattern is taking place in people with COVID.
Despite the enormous progress researchers have made in the last two and a half years, many important questions remain. It’s not yet clear how long COVID rates differ with different strains of the virus, or to what extent vaccination protects against chronic issues or future disease risk. Estimates of how much vaccination reduces the risk of long COVID vary widely and are not as high as many had hoped.
It’s also unclear why only some people are susceptible to long COVID, and whether only a subset might suffer longer-term consequences. “What makes some people more vulnerable?” Stevens says. “Are there windows of time, either in development or the aging brain, where impact is pronounced?” As with Epstein-Barr and MS, it may turn out that only a small percentage of people who get COVID are at risk of long-term cognitive decline. “I don’t think everyone who got COVID will end up with Alzheimer’s disease,” Tronson says. “What promotes resilience and what are potential risk factors?”
With so much up in the air, it’s difficult to know how these potential risks should influence public health efforts. “We’re in a tricky situation,” Stevens says. “Lots of people are not taking COVID as seriously, but long COVID is not going away. We need to think about the chronic impact that can take a toll on people’s future.”