Neuroinflammatory Signatures of Long COVID: CNS Profiles and Neurological Symptoms

Long COVID’s neurological symptoms stem from systemic inflammation, BBB disruption, glial activation, and microvascular damage—often diverging into cognitive or autoimmune motor phenotypes.

Section 1: The Pathophysiological Foundations of Neurological Long COVID

The emergence of Post-Acute Sequelae of SARS-CoV-2 infection (PASC), commonly known as long COVID, has presented a formidable challenge to the global medical and scientific communities. Among its most debilitating manifestations are a diverse array of persistent neurological and neuropsychiatric symptoms, including cognitive dysfunction ("brain fog"), profound fatigue, post-exertional malaise, sleep disturbances, headaches, and, in some cases, movement disorders.¹ A central and compelling body of evidence indicates that these neurological sequelae are not the result of a widespread, productive viral infection of the central nervous system (CNS) parenchyma. While SARS-CoV-2 RNA is occasionally detected in the cerebrospinal fluid (CSF) or brain tissue, this is a rare event and does not correlate with the widespread pathology observed.⁵ Instead, the neuropathology of long COVID is increasingly understood as the consequence of a multi-pronged, indirect assault on the CNS. This assault is driven by a complex interplay of sustained systemic inflammation, profound vascular and endothelial injury, and a dysregulated host immune response that ultimately breaches and perturbs the highly protected CNS environment.⁵ This section will establish the mechanistic framework for this indirect pathology, exploring the foundational pillars of blood-brain barrier compromise, chronic glial cell activation, immunothrombotic microvascular disease, and the persistent triggers that sustain this damaging cycle.

1.1 The Indirect Assault: Systemic Inflammation and the Compromised Blood-Brain Barrier (BBB)

The initial host response to SARS-CoV-2 infection is a critical initiating event in the cascade that leads to neurological long COVID. In many individuals, particularly those with severe acute disease, the infection triggers a hyper-inflammatory state characterized by a dysregulated and excessive release of pro-inflammatory cytokines and chemokines.² This systemic inflammatory response, while intended to control the virus, does not remain confined to the periphery; it exerts a direct and deleterious effect on the integrity of the blood-brain barrier (BBB), the specialized endothelial interface that meticulously regulates the passage of molecules and cells into the CNS.⁵ The compromise of this critical barrier is a key step that allows systemic pathology to translate into CNS dysfunction.

Direct in vivo evidence for BBB disruption in long COVID comes from advanced neuroimaging studies. Using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), researchers have demonstrated that patients with long COVID-associated brain fog exhibit a significantly higher volume of "leaky" blood vessels in the brain compared to individuals who recovered from COVID-19 without cognitive symptoms.¹² This BBB permeability is not random; it is particularly evident in the frontal and temporal lobes, cortical regions that are indispensable for the very cognitive functions—such as executive function, attention, and memory—that are impaired in patients with brain fog.¹² This finding provides a compelling anatomical and physiological correlate for a core symptom of long COVID, localizing the pathology to relevant neural circuits.

The molecular mechanisms underlying this BBB disruption are intrinsically linked to the specific cytokine profile of COVID-19. The systemic inflammatory milieu is rich in pro-inflammatory cytokines such as interleukin-6 (IL-6) and interleukin-1β (IL-1β), both of which have been shown to directly compromise the BBB by disrupting the function of GTPases that maintain the integrity of endothelial tight junctions.⁵ Concurrently, the host response to SARS-CoV-2 is often characterized by a muted Type I interferon (IFN) response, which under normal circumstances can have a protective effect on the BBB. This combination of elevated barrier-disrupting cytokines and a deficient barrier-protective response creates a "perfect storm" for increased BBB permeability.⁵

The BBB is not merely a passive structure that is damaged by external factors; it is an active participant in the neuroinflammatory process. This is powerfully demonstrated by in vitro experiments where cultured human brain endothelial cells were exposed to serum from individuals with long COVID. The patient serum was sufficient to induce the expression of key inflammatory molecules in the endothelial cells, including tumor necrosis factor (TNF) and vascular cell adhesion molecule 1 (VCAM1).¹² The upregulation of VCAM-1 is particularly significant. It functions as an adhesion molecule that actively captures circulating leukocytes, facilitating their infiltration into the CNS. This evidence transforms the understanding of the BBB's role from that of a simple, damaged "leak" to that of a pro-inflammatory signaling interface. The brain's own vasculature becomes activated, turning from a protective barrier into a gateway that not only permits the entry of inflammatory mediators but actively recruits peripheral immune cells, thereby amplifying and sustaining the neuroinflammatory state at the neurovascular unit.¹²

1.2 The Brain's Immune Response: Sustained Activation of Microglia and Astrocytes

Once the BBB is breached or receives inflammatory signals from the periphery, the brain's resident immune cells—microglia and astrocytes—are roused into action.¹ In a healthy response to acute injury, this glial activation is a tightly regulated, protective process aimed at clearing pathogens and debris and promoting repair. However, in long COVID, this response appears to become chronic, dysregulated, and maladaptive, shifting from a protective, homeostatic state to a neurotoxic one that perpetuates CNS damage.¹⁴

Direct histological proof of this sustained glial response comes from post-mortem examinations of brain tissue from patients who died with COVID-19. These studies consistently reveal hallmarks of profound glial reactivity, including the formation of microglial nodules (dense clusters of activated microglia) and widespread reactive astrogliosis (the activation and proliferation of astrocytes).⁷ These pathological features are particularly prominent in the brainstem and cerebellum, regions that regulate autonomic function, motor control, and coordination—functions that are often disrupted in long COVID.⁷

The triggers for this glial activation are multifactorial. The influx of peripheral cytokines through a compromised BBB is a primary driver. Additionally, there is evidence that SARS-CoV-2 components can directly stimulate glial cells. For example, the viral accessory protein ORF3a has been shown to activate astrocytes, inducing the production of pro-inflammatory cytokines and ultimately leading to glial cell death.¹⁸ Furthermore, studies have demonstrated that SARS-CoV-2 can directly infect human microglia in vitro, provoking a potent, M1-like pro-inflammatory response that is associated with neurotoxicity.¹⁹

This chronic glial activation can be detected through fluid biomarkers. Glial Fibrillary Acidic Protein (GFAP), a structural protein released by activated or injured astrocytes, has been associated with both neuroinflammation and cognitive symptoms in long COVID.¹³ More sophisticated approaches have combined markers of glial activation with markers of neuronal damage into a plasma-based "Neuro-Glial score," which has been shown to correlate with the severity of anxiety in individuals with long COVID, providing a blood-based biological correlate for neuropsychiatric symptoms.²⁰

The pattern of chronic, low-grade neuroinflammation driven by activated glia in long COVID bears a striking resemblance to the inflammatory signatures observed in the early stages of canonical neurodegenerative diseases. The sustained release of pro-inflammatory mediators like TNF-α and IL-6 from activated microglia and astrocytes is known to promote synaptic dysfunction, inhibit neurogenesis, and cause direct neuronal injury—processes central to the pathology of conditions like Alzheimer's disease (AD).¹³ This connection is more than superficial; transcriptomic analyses of brain tissue have revealed that the gene expression signatures in long COVID significantly overlap with those found in AD, including the dysregulation of pathways involved in tau protein metabolism, oxidative stress, and glial reactivity.¹³ Furthermore, advanced proteomic studies aiming to identify CSF biomarkers of glial activation in AD are investigating many of the same glial-derived proteins that are now being explored in the context of long COVID.²² This convergence of molecular pathways suggests that the neuroinflammatory processes triggered by SARS-CoV-2 are not entirely novel but may engage the same fundamental mechanisms that drive neurodegeneration. This raises the alarming possibility that long COVID could act as an accelerant or a "second hit" that lowers the threshold for the clinical manifestation of neurodegenerative diseases like AD or Parkinson's disease in susceptible individuals.²⁵

1.3 Vascular Pathology: Endothelial Dysfunction, Immunothrombosis, and Cerebral Hypoperfusion

The pathology within the CNS of long COVID patients extends beyond BBB leakage and glial activation to encompass a profound and persistent disease of the brain's microvasculature. The neurovascular unit is subjected to a sustained pro-thrombotic state, a phenomenon termed "immunothrombosis," where inflammation and coagulation become pathologically intertwined.² This process involves direct endothelial injury, aberrant platelet activation, and the formation of persistent microclots that can occlude small vessels, leading to localized ischemia and hypoxia that significantly contribute to neuronal dysfunction and cell death.¹³

Autopsy studies provide the most stark and compelling evidence of this neurovascular pathology. Post-mortem examinations reveal multifocal vascular damage characterized by widespread activation of endothelial cells, the adherence of platelet aggregates and microthrombi to vessel walls, and the leakage of plasma proteins into the surrounding brain parenchyma.⁷ The initiating event for this vascular assault appears to be an autoimmune attack. The deposition of immune complexes, composed of IgG and IgM antibodies, and components of the classical complement pathway, such as C1q and C4d, has been observed directly on the surface of endothelial cells. This suggests an antibody-mediated cytotoxic process targeting the brain's own vasculature, which triggers a downstream cascade of platelet aggregation, thrombus formation, and inflammation.⁷

This process is not limited to the acute, fatal phase of the disease. A key feature identified in the peripheral blood of living long COVID patients is the presence of persistent, amyloid-containing microclots that are resistant to the body's natural fibrinolytic (clot-dissolving) processes.²⁷ These anomalous microclots are hypothesized to circulate and lodge in the body's smallest capillaries, including those in the brain. By physically obstructing these vessels, they can severely limit oxygen and nutrient exchange, causing significant tissue hypoxia.² This provides a direct mechanistic link between a systemic hematological abnormality and a primary driver of brain injury in long COVID.

The functional consequence of this microvascular pathology is borne out by in vivo neuroimaging. Studies using arterial spin labeling functional MRI have found evidence of decreased neurovascular perfusion in long COVID patients who report persistent cognitive problems.¹ Given that the brain is a highly metabolic organ, consuming up to 20% of the body's total blood supply, it is uniquely vulnerable to disruptions in perfusion.¹ This evidence collectively suggests that neurological long COVID can be conceptualized, in large part, as a microvascular disease of the brain. Symptoms like "brain fog" may not arise solely from cytokine-mediated disruption of neuronal signaling but could be a direct consequence of localized hypoxia and a chronic state of energy failure within critical brain networks—such as the frontotemporal circuits responsible for executive function and memory—caused by widespread microvascular occlusion. This perspective significantly broadens the potential therapeutic landscape, suggesting that strategies targeting endothelial stabilization, platelet activation, and coagulation may be as important as purely anti-inflammatory approaches.

1.4 Persistent Triggers: The Roles of Viral Reservoirs and Autoimmunity

The chronicity of long COVID, with symptoms persisting for months or even years, strongly implies the presence of a persistent trigger that sustains the cycle of immune dysregulation and inflammation. Two leading, non-mutually exclusive hypotheses have emerged to explain this chronicity: the persistence of SARS-CoV-2 viral components in tissue reservoirs and the development of a post-infectious autoimmune response where the body's immune system mistakenly attacks its own tissues.⁴

Mounting evidence supports the hypothesis of viral persistence. Early autopsy studies detected SARS-CoV-2 RNA and proteins in a wide array of non-respiratory tissues, including the brain, as long as 230 days after the initial onset of symptoms.²⁹ This finding directly established the potential for long-term antigen presence within the body. More recent and sensitive studies have extended these findings dramatically, detecting SARS-CoV-2 antigens (such as the spike protein) in the blood of individuals up to 14 months post-infection and viral RNA in tissue samples for more than two years.³¹ Crucially, the presence of these circulating viral antigens correlates with a higher likelihood of having multi-system long COVID symptoms.³² These tissue reservoirs, even if they harbor low levels of virus or viral components, could act as a chronic source of pathogen-associated molecular patterns (PAMPs). The continuous presentation of these PAMPs to the innate immune system would provide a constant stimulus, driving the state of low-grade, systemic inflammation that underpins many features of the disease.¹

The second major hypothesis posits that the initial infection triggers a break in self-tolerance, leading to autoimmunity.⁴ This can occur through mechanisms such as molecular mimicry, where viral proteins resemble host proteins, causing the immune response to cross-react, or through bystander activation, where intense inflammation during the acute infection leads to the exposure of self-antigens and an aberrant immune response against them. Evidence for this process has been found within the CNS itself. Studies of CSF from long COVID patients with cognitive symptoms have identified the presence of unexpected autoantibodies, some of which were unique to the CSF and not found in the blood. This suggests an intrathecal, compartmentalized production of "turncoat" antibodies that may be targeting CNS antigens.³⁴ The evidence becomes even more specific when examining distinct neurological phenotypes. In case reports of patients who developed post-COVID myoclonus (a movement disorder characterized by involuntary muscle jerks), researchers have identified autoantibodies in the CSF that specifically target astrocytes, providing a direct link between a targeted autoimmune response and a specific neurological syndrome.³⁵

The distinction between viral persistence and autoimmunity may be somewhat artificial, as they likely represent two interconnected points on a continuous pathogenic spectrum. A plausible sequence of events begins with the establishment of persistent viral reservoirs in various tissues. The chronic immune stimulation resulting from this persistence, particularly in a genetically predisposed individual, could lead to a dysregulated immune response, epitope spreading (where the immune response expands from targeting viral proteins to targeting host proteins), and the eventual loss of self-tolerance. This model would explain the heterogeneity of long COVID. The specific tissue in which the virus persists and the specific autoantibodies that are subsequently generated could determine the ultimate clinical phenotype. For example, a patient with viral persistence in the gut might develop gastrointestinal symptoms, while another who develops anti-astrocyte antibodies might present with myoclonus. This integrated view suggests that viral persistence may be the initial driver that, over time, ignites a secondary, self-sustaining autoimmune process in a subset of patients.

Section 2: Inflammatory Profiles in the Cerebrospinal Fluid: A Window into the CNS

The cerebrospinal fluid (CSF), which bathes the brain and spinal cord, offers a unique and invaluable window into the biochemical and immunological environment of the central nervous system. Analysis of the CSF allows for the direct measurement of inflammatory mediators, immune cells, and markers of neuronal injury, providing critical evidence of neuroinflammatory processes in living individuals. However, the study of CSF in long COVID has yielded a complex and often contradictory set of findings. While some studies have identified clear signs of an active immune response, others have found largely normal profiles, even in patients with severe neurological symptoms. This section will critically examine the evidence from CSF analysis, cataloging the specific inflammatory signatures that have been identified, dissecting the controversy surrounding the conflicting results, and exploring how advanced proteomic techniques are providing a more nuanced understanding of the pathology.

2.1 Cytokine and Chemokine Signatures: Identifying Key Mediators

Multiple investigations, particularly those focusing on patients during the acute or subacute phases of COVID-19-associated neurological complications, have successfully identified elevated levels of pro-inflammatory cytokines and chemokines in the CSF. These findings provide direct evidence of an active immune response occurring within the CNS compartment.

A systematic review that aggregated data from 28 publications on acute COVID-19 patients with CSF analysis found evidence of elevations in 37 different cytokines.⁶ Among these, the pro-inflammatory cytokine Interleukin-6 (IL-6) and the neutrophil-attracting chemokine Interleukin-8 (IL-8) were the most frequently elevated, found in 60% and 51% of patients tested, respectively. These elevations were often associated with poor clinical outcomes, underscoring their potential pathological significance.⁶ Other studies focusing on patients with acute neuro-COVID syndromes, such as encephalopathy or encephalitis, have corroborated these findings, reporting elevated CSF levels of IL-6, IL-8, Interleukin-18 (IL-18), and Monocyte Chemoattractant Protein-1 (MCP-1) compared to healthy controls.³⁶

Distinct inflammatory signatures have also been described. For instance, a study of cancer patients who developed neurologic sequelae of COVID-19 identified a prominent Interferon-gamma (IFN-γ)-driven signature in the CSF.³⁷ This was characterized by elevated levels of IFN-γ itself and its downstream effector chemokines, CXCL9, CXCL10 (also known as IP-10), and CXCL11. This profile was detectable nearly two months after the initial SARS-CoV-2 infection, suggesting a persistent, specific type of immune activation within the CNS.³⁷

An important consideration is the origin of these inflammatory mediators. While their presence in the CSF confirms a neuroinflammatory state, it does not definitively prove that they are being produced within the CNS (intrathecally). Some research suggests that at least a portion of these molecules may enter the CSF from the systemic circulation via a compromised BBB. Calculations of the CSF/serum ratio (or index) for molecules like IL-6 and CXCL10 have indicated that their concentration gradient is consistent with passage from the blood into the CSF, rather than intrathecal synthesis.³⁸ This finding lends further support to the model in which a primary systemic inflammatory process drives secondary CNS pathology.

The following table synthesizes the findings on key inflammatory mediators and biomarkers identified in the CNS of individuals with long COVID, providing a consolidated reference.

Table 1: Key Inflammatory Mediators and Biomarkers in the CNS of Long COVID Patients

2.2 The Controversy of CSF Findings: A Critical Appraisal of Studies Showing Inflammation vs. Normalcy

While the evidence for CSF inflammation in acute neuro-COVID is relatively robust, the picture in chronic long COVID is far more ambiguous and is defined by a significant scientific controversy. A compelling body of evidence from well-designed studies has reported largely normal CSF profiles in individuals with persistent and debilitating neuropsychiatric symptoms, directly challenging the hypothesis that ongoing, measurable neuroinflammation is the primary driver of their condition.

A pivotal case-control study conducted at Yale University represents the strongest evidence for this "normal" CSF profile.⁴¹ The investigation involved 37 individuals with neuro-PCC, most of whom reported brain fog and fatigue, and compared their CSF to that of 22 pre-pandemic controls. The results were striking: the long COVID group showed no significant elevation in routine markers of inflammation, including CSF white blood cell counts or total protein levels. Furthermore, the CSF-to-blood albumin ratio, a sensitive measure of BBB integrity, was not elevated, suggesting the absence of a widespread, ongoing "leaky" barrier in this chronic phase.⁴¹ A comprehensive analysis of a 15-plex cytokine panel in both CSF and plasma from this cohort also failed to reveal any significant differences between the long COVID patients and controls after appropriate statistical correction for multiple comparisons. The marker of microglial and macrophage activation, neopterin, was also found to be at normal levels in the CSF.⁴¹ Based on these comprehensive negative findings, the researchers concluded that persistent, high-grade CNS immune activation and BBB dysfunction are unlikely to be the primary drivers of neuropsychiatric symptoms in this cohort of long COVID patients.⁴³

This stands in stark contrast to findings from other research groups. For example, a study from the University of California, San Francisco (UCSF) examined individuals with post-COVID cognitive symptoms and found that a majority (10 out of 13) had demonstrable abnormalities in their CSF.³⁴ These abnormalities included elevated protein levels, indicative of inflammation or BBB disruption, and, perhaps most importantly, the presence of unexpected oligoclonal antibodies that were not present in the CSF of control participants who had recovered from COVID without cognitive sequelae. The presence of these antibodies points towards a compartmentalized, intrathecal humoral immune response, a classic sign of CNS-specific inflammation.³⁴

Reconciling these conflicting results is crucial for advancing the field. The discrepancy does not necessarily invalidate the neuroinflammation hypothesis but rather suggests that the nature of inflammation in long COVID is more complex and subtle than in classic neuroinflammatory diseases like multiple sclerosis or viral encephalitis. Several factors could explain these divergent findings. First, there may be significant cohort differences. The clinical presentation of long COVID is notoriously heterogeneous, and different underlying pathologies may drive similar symptom clusters in different patient subgroups. Factors such as the severity of the initial infection, vaccination status, time elapsed since infection, and underlying genetic predispositions could all influence the inflammatory profile. The Yale cohort, for instance, was studied at a median of over a year post-infection, a time point where an initial inflammatory phase might have resolved, leaving behind persistent neuronal damage or dysfunction.

Second, there is the "window" problem. A lumbar puncture provides only a single snapshot in time of a dynamic, chronic disease process. It is plausible that a critical neuroinflammatory event occurs in the subacute phase (weeks to a few months post-infection) that sets in motion a cascade of neuronal injury, synaptic pruning, or microvascular damage. This damage could then persist and cause symptoms long after the inflammatory mediators in the bulk CSF have returned to baseline levels. In this "hit-and-run" model, a CSF sample taken a year later would miss the causative inflammatory event.

Third, and perhaps most importantly, is the "location" problem. CSF analysis primarily samples the fluid within the leptomeningeal space and ventricles. It may not accurately reflect pathological processes that are deeply embedded within the brain parenchyma. As autopsy studies have shown, the pathology of COVID-19 can include discrete microglial nodules and perivascular inflammation.⁷ It is conceivable that this type of smoldering, localized inflammation does not shed a sufficient quantity of cytokines into the bulk CSF to be detected above the noise of standard assays. This is strongly supported by PET imaging studies that successfully visualize localized patches of glial activation in the brains of living long COVID patients.¹ Therefore, the "negative" CSF studies do not definitively rule out neuroinflammation. Instead, they critically refine the hypothesis, pushing the focus away from a diffuse, high-grade meningitis or encephalitis model toward one of a low-grade, chronic, and potentially compartmentalized parenchymal or neurovascular pathology. This has profound implications for diagnostics, suggesting that CSF analysis alone may be insufficient and must be integrated with more sensitive tools like advanced neuroimaging and proteomics to fully capture the disease process.

2.3 Insights from Proteomics: Beyond Cytokines to Complement, Coagulation, and Neuroaxonal Injury Markers

To move beyond the limitations of pre-selected cytokine panels and gain a more comprehensive, unbiased view of the pathological processes within the CNS, researchers have begun to apply advanced proteomic techniques to CSF analysis. This approach allows for the simultaneous measurement of hundreds or thousands of proteins, revealing alterations in entire biological pathways and identifying novel biomarker candidates.

Proteomic studies of CSF from critically ill patients with acute COVID-19 have provided powerful confirmation of the immunothrombosis hypothesis within the CNS. One such study identified 76 differentially expressed proteins, with a significant number of the upregulated proteins being key components of the complement and coagulation cascades. These included haptoglobin, prothrombin (F2), and the fibrinogen alpha and beta chains (FGA, FGB).⁴⁹ The coordinated upregulation of these pathways in the CSF provides strong molecular evidence for a state of hypercoagulability and complement-mediated inflammation occurring directly within the CNS environment.⁴⁹

Another critical contribution of CSF proteomics is the ability to quantify markers of direct neuronal damage. Neurofilament light chain (NfL), a structural protein of the neuronal axon, is released into the CSF and bloodstream upon axonal injury. Multiple studies have shown that CSF NfL levels are elevated in patients with COVID-19-associated neurological complications and that these elevated levels are associated with poor clinical outcomes.⁶ NfL serves as a direct biochemical correlate of neuronal damage, providing a quantitative measure of the downstream consequences of the neuroinflammatory and vascular insults.

Data-driven analytical approaches, such as principal component analysis (PCA), applied to proteomic datasets can reveal co-regulated networks of proteins that define the pathological state. In a study of PASC patients with cognitive impairment, PCA of CSF inflammatory proteins identified distinct clusters.⁴⁴ One cluster was driven by markers of general inflammation (soluble TNF receptor 1 and 2, or sTNFR1/sTNFR2) and axonal injury (NfL). A second, distinct cluster was driven by a marker of microglial activation (soluble Triggering Receptor Expressed on Myeloid cells 2, or sTREM2) and proteins related to IL-6 signaling. This type of analysis demonstrates that the neuroinflammatory state is not monolithic but is composed of multiple, potentially independent biological processes—such as general inflammation, axonal injury, and specific microglial responses—that can be tracked simultaneously.⁴⁴

2.4 Cellular Analysis: Evidence of Immune Cell Trafficking and Intrathecal Activation

While routine CSF analysis in long COVID often reveals normal or only mildly elevated white blood cell counts (pleocytosis), more detailed and specific immunological assays can uncover subtle but important signs of immune cell trafficking and activation within the CNS.³⁸ In large cohorts of patients with COVID-19 and neurological symptoms, CSF cell counts are typically within the normal range.⁵¹ However, pleocytosis is more common in patients with specific, severe syndromes like acute encephalitis, where it is a diagnostic feature.⁵⁰

The absence of significant pleocytosis in many long COVID patients does not mean the absence of a cellular immune response. More sophisticated analyses have identified CNS-specific T and B cell activation in the CSF of individuals with neurological symptoms during acute COVID-19.³³ Furthermore, CSF-derived IgG antibodies from these patients have been shown to be reactive against neuronal antigens, suggesting the presence of an autoimmune cellular response within the CNS.³³

One of the most specific indicators of a compartmentalized CNS immune response is the presence of oligoclonal bands (OCBs). OCBs are distinct bands of immunoglobulins seen on electrophoresis of CSF that are not present in a matched serum sample, indicating that they are being produced by clones of B cells residing within the CNS. As noted previously, the UCSF study on patients with cognitive symptoms identified the presence of these unexpected antibodies in the CSF, providing strong evidence for an intrathecal, localized humoral immune response in a subset of patients with cognitive long COVID.³⁴ Together, these findings indicate that even when the overall number of immune cells in the CSF is not elevated, a pathogenic, compartmentalized cellular and humoral immune response may still be occurring.

Section 3: Neuropathological Evidence from Brain Tissue and Advanced Neuroimaging

While analysis of the CSF provides an invaluable, albeit indirect, view of the CNS environment, the "ground truth" of long COVID neuropathology is found through direct examination of the brain itself. Evidence from post-mortem histopathological studies and advanced in vivo neuroimaging techniques provides a powerful and complementary perspective, visualizing the cellular and molecular consequences of the disease process. These approaches have been instrumental in confirming the central roles of glial activation and neurovascular injury and have firmly established that the primary pathology is one of inflammation and vascular damage, rather than direct viral encephalitis.

3.1 Histopathological Correlates: Cellular Infiltrates, Glial Reactivity, and Vascular Damage in Post-Mortem Studies

Autopsy studies of patients who died with or from COVID-19 have provided unequivocal and consistent evidence of a profound neuroinflammatory and neurovascular pathology, even in cases where SARS-CoV-2 is not detected within the brain parenchyma.⁵ These findings are critical because they define the cellular landscape of the disease.

A cardinal feature observed in these studies is the presence of perivascular infiltrates, where immune cells are found clustered in the space surrounding the brain's small blood vessels.⁷ Detailed immunohistochemical analysis of these infiltrates has revealed a specific cellular composition. The predominant cell types are CD68+ macrophages and CD8+ cytotoxic T-cells. In contrast, CD4+ helper T-cells and CD20+ B-cells are rare.⁷ This specific immune signature points towards a pathology driven primarily by the innate immune system (macrophages) and the cytotoxic arm of the adaptive immune system (CD8+ T-cells), rather than a B-cell-driven, antibody-producing process within the infiltrates themselves.

Beyond the perivascular space, the brain parenchyma shows widespread signs of a reactive glial response. Microglial activation is a prominent feature, with microglia shedding their ramified, resting morphology and adopting an activated, amoeboid state. In many cases, these activated microglia aggregate to form dense clusters known as microglial nodules, a classic hallmark of viral or inflammatory CNS disease.⁷ These nodules are particularly concentrated in the hindbrain, including the medulla oblongata and cerebellum.⁷ Critically, these activated microglia are often observed in the act of neuronophagia—surrounding and engulfing damaged or dying neurons. This provides direct visual evidence of active neuronal injury and removal being carried out by the brain's own immune cells.⁷ Alongside microgliosis, prominent reactive astrogliosis is also observed, with astrocytes increasing in size and upregulating their characteristic protein, GFAP. This astrocytic activation is most intense in the perivascular regions and in areas of neuronal loss, indicating a response to both vascular injury and parenchymal damage.⁷

As detailed in Section 1.3, the vasculature itself is a primary site of injury. Autopsy confirms widespread endothelial cell activation, the deposition of immune complexes and complement on the vessel walls, and the presence of platelet-rich microthrombi occluding or lining the microvessels.⁷ The combination of these findings—the consistent failure to detect significant viral load in the brain parenchyma, coupled with the overwhelming evidence of an immune-mediated attack centered on the neurovascular unit—leads to a crucial conclusion. The neuropathology of fatal COVID-19 is not a classic viral encephalitis, which would be characterized by widespread viral infection of neurons and glia. Instead, it is best described as a neuroinflammatory and neurovascular endotheliitis. In this model, a systemic immune response, likely triggered by the virus, mistakenly targets the endothelial cells of the brain's vasculature. This initial attack on the blood vessels then precipitates a secondary cascade of BBB breakdown, perivascular immune cell infiltration, parenchymal glial activation, and, ultimately, neuronal injury and death. This model elegantly explains how a systemic respiratory infection can cause profound and widespread neurological damage without requiring extensive viral neuroinvasion.

3.2 In Vivo Evidence: PET and MRI Findings of Glial Activation and Neuroinflammation

Advanced neuroimaging techniques provide a critical bridge between the static, end-stage findings of autopsy studies and the clinical symptoms experienced by living individuals with long COVID. In particular, Positron Emission Tomography (PET) has emerged as a powerful tool for visualizing and quantifying neuroinflammation in vivo.

These studies typically employ PET radiotracers, such as those targeting the translocator protein (TSPO). TSPO is a protein located on the outer mitochondrial membrane that is significantly upregulated in activated microglia and astrocytes, making it an excellent biomarker of glial reactivity and neuroinflammation.¹⁶ Multiple studies using TSPO-PET have demonstrated significantly increased tracer binding in the brains of long COVID patients compared to healthy controls, providing direct in vivo confirmation of the glial activation seen in post-mortem tissue.¹

Importantly, this neuroinflammation is not uniformly distributed throughout the brain. The increased TSPO signal is most prominent in specific brain regions and networks, including the midcingulate and anterior cingulate cortex, corpus callosum, thalamus, and basal ganglia.¹ The involvement of these structures is clinically relevant; the cingulate cortex is a key hub for regulating emotion, pain, and executive function, while the thalamus is a critical relay for sensory information, and the basal ganglia are central to motor control and motivation. Inflammation in these specific areas could plausibly underlie symptoms such as mood disturbances, chronic pain, cognitive fog, and fatigue.

Perhaps the most significant finding from these in vivo imaging studies is the direct link they establish between vascular pathology and CNS inflammation. One study performed both TSPO-PET and analysis of peripheral blood in the same long COVID patients.¹ The results showed a significant positive correlation between the degree of neuroinflammation measured by PET in the brain and the levels of circulating plasma markers of vascular dysfunction and coagulation, such as fibrinogen. This finding is of paramount importance as it provides an in vivo connection between the two central hypotheses of long COVID neuropathology: the systemic vascular/immunothrombotic hypothesis and the CNS glial activation hypothesis. It suggests that the sustained vascular injury and pro-thrombotic state observed in the periphery are directly related to, and predictive of, the intensity of the inflammatory response within the brain itself, uniting these processes into a single, cohesive pathogenic axis.

Section 4: Delineating Inflammatory Signatures by Neurological Phenotype

The clinical presentation of neurological long COVID is remarkably heterogeneous, ranging from the pervasive "brain fog" that affects cognitive function to more discrete and rare syndromes such as movement disorders. A critical question, and the central focus of this report, is whether these distinct clinical phenotypes are driven by different underlying inflammatory pathologies. A synthesis of the available evidence suggests that this is indeed the case. The inflammatory signature associated with cognitive dysfunction appears to be one of widespread, low-grade neuroinflammation and microvascular compromise, whereas the signature of post-COVID movement disorders points toward a more targeted, and often autoimmune, assault on specific neuronal circuits. This section will delineate and contrast these phenotype-specific profiles.

The following table provides a comparative summary of the key pathophysiological and inflammatory features associated with the two major neurological phenotypes of long COVID: cognitive dysfunction and movement disorders.

Table 2: Comparative Summary of Inflammatory Signatures for Major Neurological Phenotypes

4.1 The "Brain Fog" Enigma: Correlates of Cognitive Dysfunction

The constellation of symptoms colloquially known as "brain fog"—encompassing deficits in memory, attention, processing speed, and executive function—is one of the most common and debilitating features of long COVID.¹³ The emerging inflammatory signature for this phenotype is not one of acute, high-grade encephalitis but rather a more insidious combination of chronic neurovascular compromise, a specific inflammatory milieu that impairs neuronal plasticity, and widespread, low-grade glial activation.

The vascular component appears to be central to the pathology of brain fog. As discussed, the strongest and most direct evidence links cognitive symptoms to sustained systemic inflammation that drives localized BBB dysfunction, particularly in the frontal and temporal cortices—the hubs of higher-order cognition.¹² This vascular compromise leads to downstream consequences of hypoperfusion and microthrombosis, which can create a state of chronic, localized hypoxia. This effectively starves these highly metabolic brain regions of the oxygen and nutrients required to perform complex cognitive tasks, providing a direct physiological explanation for slowed processing and difficulty with concentration.¹³

Within this context of vascular compromise, a specific chemokine signature has been identified that may directly impair the brain's capacity for learning and memory. The chemokine CCL11 (also known as eotaxin-1) has emerged as a key molecular player. In human studies, serum levels of CCL11 have been found to positively correlate with the severity of post-acute cognitive deficits.⁵ This clinical correlation is supported by compelling mechanistic evidence from animal models. In mice, systemic administration of CCL11 is sufficient to impair the generation of new neurons (neurogenesis) in the hippocampus, a brain region absolutely critical for the formation of new memories.⁵ Furthermore, mice subjected to a mild respiratory-only SARS-CoV-2 infection—a model that closely mimics the initial illness of many long COVID patients—exhibit elevated levels of CCL11 in their CSF, which is accompanied by persistently impaired hippocampal neurogenesis and myelin loss for weeks after the infection has cleared.⁴² This provides a powerful and direct molecular link between a peripheral inflammatory signal (which can cross the compromised BBB) and the impairment of fundamental brain plasticity mechanisms.

Finally, this phenotype is associated with the broad, low-grade glial activation visible on TSPO-PET scans across multiple brain regions.¹ At a molecular level, this chronic activation involves the release of pro-inflammatory cytokines like TNF-α and IL-6, which are known to promote the stripping of synapses (synaptic pruning) and disrupt neuronal function.¹³ Peripheral blood biomarkers also offer clues that align with this theme of impaired plasticity. Some studies have linked brain fog to higher serum levels of IL-10 (a complex cytokine that can be a marker of a dysregulated inflammatory response) and, importantly, lower levels of Nerve Growth Factor (NGF), a key molecule for neuronal survival and adaptability.⁵⁸ The combination of a pro-inflammatory environment and a deficit in neurotrophic support creates a state that is hostile to optimal cognitive function. The seemingly contradictory "normal" CSF findings from the Yale study, which focused on this neuropsychiatric/cognitive cohort, can be contextualized within this model. The absence of high levels of cytokines in the bulk CSF does not negate the presence of a more subtle, parenchymal, or microvascular pathology that is driving the symptoms.⁴¹

4.2 The Spectrum of Movement Disorders: Ataxia, Myoclonus, and Parkinsonism

In contrast to the relatively diffuse inflammatory picture associated with brain fog, the movement disorders that can emerge as a sequela of COVID-19—such as ataxia (impaired coordination), myoclonus (involuntary muscle jerks), and parkinsonism—appear to be driven by more specific and often autoimmune-mediated attacks on the precise neuronal circuits that control movement, namely the cerebellum, brainstem, and basal ganglia.³³ The evidence base for these rarer conditions is composed primarily of case reports and small case series, which makes establishing a single, consistent inflammatory profile challenging. However, a clear pattern is emerging that distinguishes them from the cognitive phenotype.

A key observation across multiple reports of post-COVID myoclonus and ataxia is the frequent finding of a completely unremarkable routine CSF analysis. Patients often present with normal CSF cell counts, protein levels, and glucose, with no evidence of oligoclonal bands.⁵³ This lack of classic, overt inflammation in the CSF strongly suggests that the pathology is not a general meningoencephalitis. Instead, it points toward a more subtle and targeted pathological process, with the leading hypothesis being a post-infectious autoimmune phenomenon.⁵⁴

Crucial evidence supporting this autoimmune hypothesis has come from more specialized testing. In a case series of patients with COVID-19-associated myoclonus, researchers were able to identify anti-astrocyte IgG autoantibodies in the CSF.³⁵ This is a highly specific finding, indicating a targeted autoimmune response against a key glial cell type within the CNS. Similarly, in a case series of patients with COVID-19-associated cerebellar ataxia, specific anti-cerebellar autoantibodies (anti-Homer-3 and anti-Yo antibodies) were detected in two of the eleven patients.⁵⁶ While not found in all patients, the identification of these well-characterized pathogenic autoantibodies in even a subset of cases provides proof-of-concept that a targeted autoimmune attack is the underlying cause for at least some individuals. The proposed mechanism is one of molecular mimicry, where antibodies generated against SARS-CoV-2 cross-react with structurally similar self-antigens in the cerebellum or other motor structures.⁵⁵

The development of parkinsonism post-COVID is also hypothesized to have a neuroinflammatory basis, although the specific signature is less well-defined. The prevailing theory is that the neuroinflammation induced by SARS-CoV-2 can facilitate or accelerate the pathological aggregation of the protein alpha-synuclein in the basal ganglia, which is the hallmark of Parkinson's disease.²⁵ In this scenario, the viral infection may act as a "second hit" that either triggers the onset of the disease or unmasks a pre-existing, pre-symptomatic neurodegenerative process.²⁵

The divergence in inflammatory signatures between cognitive dysfunction and movement disorders represents a critical finding. It strongly suggests that these are not simply different manifestations of the same underlying pathology but are, in fact, distinct immunopathological entities. "Brain fog" appears to result from a diffuse, low-grade neuroinflammatory state driven by systemic inflammation and widespread microvascular injury. In contrast, post-COVID movement disorders often represent a more classic, targeted post-viral autoimmune disease, where a highly specific, aberrant immune response is directed against discrete components of the motor system. This distinction has profound implications for both diagnosis and treatment. The diagnostic workup for a patient with post-COVID brain fog might focus on markers of vascular inflammation and neuroimaging, whereas the workup for a patient with new-onset ataxia or myoclonus should include a comprehensive panel for anti-neural autoantibodies. Similarly, therapeutic approaches might diverge, with vascular-protective and broad anti-inflammatory agents being considered for cognitive symptoms, while targeted immunotherapies like intravenous immunoglobulins (IVIG) or corticosteroids may be more appropriate for autoimmune-mediated movement disorders.³⁵

Section 5: Synthesis, Clinical Implications, and Future Trajectories

The extensive and multifaceted research into the neurological sequelae of long COVID has painted a complex picture of a disease driven not by direct viral assault on the brain, but by the downstream consequences of a dysregulated host immune and inflammatory response. By synthesizing the evidence from cerebrospinal fluid analysis, advanced neuroimaging, and post-mortem neuropathology, a cohesive, albeit intricate, model of pathogenesis is emerging. This integrated understanding has significant implications for clinical practice and illuminates a clear path forward for future research aimed at developing effective diagnostics and targeted therapeutics.

5.1 An Integrated Model of Neuroinflammatory Pathogenesis in Long COVID

A comprehensive model of neurological long COVID must account for the initial trigger, the key pathological mechanisms, and the divergence into distinct clinical phenotypes. Such a model can be conceptualized as a multi-stage process:

1. Initiation: The process begins with the acute SARS-CoV-2 infection, which triggers a potent systemic inflammatory response. This response is often dysregulated, with high levels of pro-inflammatory cytokines (e.g., IL-6, IL-1β, TNF-α). In a subset of individuals, this inflammatory state is sustained long-term, potentially driven by the persistence of viral antigens in tissue reservoirs, which provide a chronic source of immune stimulation.

2. Vascular Compromise: This sustained systemic inflammation leads to chronic activation and injury of the endothelial cells lining the blood vessels throughout the body, including the highly specialized vasculature of the brain. This results in the compromise of the blood-brain barrier, increasing its permeability. This "leaky" BBB allows peripheral inflammatory mediators, activated immune cells, and potentially harmful circulating factors (like autoantibodies or microclots) to gain access to the CNS.

3. CNS Immune Activation: The influx of these peripheral factors, combined with signals from the activated endothelium, triggers a state of chronic activation in the brain's resident immune cells, the microglia and astrocytes. This sustained glial reactivity establishes a pro-inflammatory environment within the CNS parenchyma, characterized by the local production of cytokines, chemokines, and reactive oxygen species, which can lead to synaptic dysfunction and neuronal injury.

4. Phenotypic Divergence: From this common foundation of systemic inflammation, BBB compromise, and glial activation, the pathology can diverge to produce different clinical syndromes, likely dictated by individual host factors (e.g., genetics, prior immune history) and the specific nature of the immune response:

• The Cognitive Phenotype Pathway ("Brain Fog"): In this pathway, the dominant pathology is widespread microvascular injury, immunothrombosis, and the resulting cerebral hypoperfusion. This is coupled with a specific inflammatory milieu, potentially characterized by high levels of chemokines like CCL11, that directly impairs synaptic plasticity and neurogenesis in vulnerable, high-metabolism cognitive networks, such as the hippocampus and prefrontal cortex. The result is a diffuse deficit in cognitive processing.

• The Movement Disorder Pathway: In this pathway, the dysregulated immune response leads to a critical loss of self-tolerance and the generation of highly specific autoantibodies. These autoantibodies target distinct neuronal or glial antigens located within the precise motor circuits of the cerebellum, brainstem, or basal ganglia. This targeted autoimmune attack causes focal neurological damage, resulting in specific clinical deficits such as ataxia, myoclonus, or parkinsonism.

This integrated model accounts for the key evidence—from systemic inflammation to phenotype-specific autoantibodies—and provides a framework for understanding the heterogeneity of neurological long COVID.

5.2 Toward Phenotype-Specific Biomarkers: Current Limitations and a Roadmap for Future Research

Despite significant progress, the field faces substantial limitations that hinder the translation of these pathogenic insights into clinical tools. Current research is often hampered by the use of small, heterogeneous patient cohorts, the lack of standardized diagnostic criteria for long COVID and its sub-phenotypes, and a reliance on cross-sectional study designs that cannot capture the dynamic evolution of the disease. The profound controversy surrounding CSF findings underscores the limitations of relying on a single diagnostic modality and highlights the need for a more integrated, multi-modal approach.

A clear roadmap for future research can be defined to address these challenges:

• Establishment of Longitudinal, Multi-modal Cohorts: The highest priority is the establishment of large, well-phenotyped patient cohorts that are followed longitudinally over time. These studies must collect a comprehensive set of data at multiple time points, integrating detailed clinical phenotyping with biospecimen collection (peripheral blood and CSF) and advanced neuroimaging (e.g., DCE-MRI for BBB integrity, TSPO-PET for glial activation). This is the only way to map the temporal evolution of the disease and definitively link specific biological changes to clinical outcomes.

• Deployment of Advanced CSF and Blood Analysis: Research must move beyond standard, pre-selected cytokine panels. The use of unbiased, high-throughput techniques such as mass spectrometry-based proteomics and single-cell RNA sequencing of CSF immune cells is essential to discover novel pathological pathways and identify new, more sensitive biomarker candidates.²²

• Systematic Autoantibody Discovery: For patients presenting with specific neurological syndromes, particularly movement disorders or suspected encephalitis, a systematic and unbiased search for novel anti-neural autoantibodies is critical. This will allow for the definitive classification of these syndromes as autoimmune and will provide highly specific diagnostic markers.

• Validation of Peripheral Biomarkers: Given the invasive nature of lumbar puncture, a major goal should be the identification and validation of peripheral blood biomarkers that can reliably reflect CNS pathology. Further research into plasma-based markers of glial activation and neuronal injury (like the Neuro-Glial score), specific chemokines (like CCL11), and neurotrophic factors (like NGF) is needed to develop accessible, scalable diagnostic and prognostic tools.⁵

5.3 Therapeutic Implications

The heterogeneity of the underlying pathology strongly implies that a "one-size-fits-all" therapeutic approach to neurological long COVID is destined to fail. Instead, the emerging pathogenic models point toward a future of phenotype-directed, personalized medicine.

• For Cognitive Dysfunction ("Brain Fog"): Therapeutic strategies for this phenotype should be multi-pronged, targeting the key mechanisms of neurovascular injury and low-grade inflammation. This could include therapies aimed at restoring BBB integrity and endothelial health, anti-platelet or anti-coagulant agents to disrupt microclot formation and improve cerebral microcirculation, and novel drugs that specifically target microglial activation, such as NLRP3 inflammasome inhibitors.¹³ Furthermore, agents that can neutralize specific pathogenic chemokines like CCL11 could represent a highly targeted approach to restoring neurogenesis and cognitive function.

• For Autoimmune-Mediated Movement Disorders: If a specific autoimmune basis is confirmed through the detection of pathogenic autoantibodies, the therapeutic approach should be guided by established principles of neuroimmunology. Treatments such as high-dose corticosteroids, intravenous immunoglobulins (IVIG), plasma exchange, or more targeted B-cell depleting therapies (e.g., rituximab) may be highly effective, as has been suggested in several case reports.³⁵

• Addressing Upstream Triggers: Finally, a crucial area of ongoing investigation is the possibility of treating the upstream triggers that sustain the chronic inflammatory state. If viral persistence is confirmed as a key driver, clinical trials of extended courses of antiviral medications, such as Paxlovid, will be essential to determine if eliminating the viral reservoir can lead to clinical improvement and resolution of the downstream inflammatory cascade.³⁰

In conclusion, the neurological sequelae of long COVID are rooted in a complex, indirect pathology driven by sustained inflammation and immune dysregulation. While the specific inflammatory signatures are still being fully elucidated, a clear divergence is apparent between the diffuse neurovascular and glial pathology underlying cognitive dysfunction and the targeted autoimmune processes that appear to drive movement disorders. Continued, rigorous, multi-modal research is essential to refine these models, validate phenotype-specific biomarkers, and ultimately develop the targeted therapeutics that are desperately needed by the millions of individuals affected by this debilitating condition.

Acknowledgement

I acknowledge the use of Gemini AI in the preparation of this report. Specifically, it was used to: (1) brainstorm and refine the initial research questions; (2) assist in writing and debugging Python scripts for statistical analysis; and (3) help draft, paraphrase, and proofread sections of the final manuscript. I reviewed, edited, and assume full responsibility for all content.

Works cited

1. Neuroinflammation in post-acute sequelae of COVID-19 (PASC) as assessed by [11C]PBR28 PET correlates with vascular disease measures - PMC - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC11225883/

2. A review of cytokine-based pathophysiology of Long COVID symptoms - Frontiers, accessed August 31, 2025, https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2023.1011936/full

3. Long COVID: neurological manifestations - an updated narrative review - PMC, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC10901563/

4. Long COVID and Its Impact on Multi-system Inflammatory Disorders: A Meta Analysis of Emerging Evidence and Unanswered Questions | Biores Scientia, accessed August 31, 2025, https://bioresscientia.com/article/long-covid-and-its-impact-on-multi-system-inflammatory-disorders-a-meta-analysis-of-emerging-evidence-and-unanswered-questions

5. Neuroinflammation and COVID-19 - PMC, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC9239981/

6. Cerebrospinal fluid cytokines in COVID-19: a review and meta-analysis - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC10591843/

7. Neurovascular injury with complement activation and inflammation ..., accessed August 31, 2025, https://academic.oup.com/brain/article/145/7/2555/6621999

8. Neurological sequelae of long COVID: a comprehensive review of diagnostic imaging, underlying mechanisms, and potential therapeutics - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC11881597/

9. (PDF) A review of cytokine-based pathophysiology of Long COVID symptoms, accessed August 31, 2025, https://www.researchgate.net/publication/369721058_A_review_of_cytokine-based_pathophysiology_of_Long_COVID_symptoms

10. Role of Inflammatory Cytokines in COVID-19 Patients: A Review on Molecular Mechanisms, Immune Functions, Immunopathology and Immunomodulatory Drugs to Counter Cytokine Storm - MDPI, accessed August 31, 2025, https://www.mdpi.com/2076-393X/9/5/436

11. Inflammation at the crossroads of COVID-19, cognitive deficits and depression - PMC, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC8894741/

12. Blood-brain barrier disruption and sustained systemic inflammation ..., accessed August 31, 2025, https://pubmed.ncbi.nlm.nih.gov/38388736/

13. The Molecular Mechanisms of Cognitive Dysfunction in Long COVID: A Narrative Review, accessed August 31, 2025, https://www.mdpi.com/1422-0067/26/11/5102

14. The Influence of Microglia on Neuroplasticity and Long-Term Cognitive Sequelae in Long COVID: Impacts on Brain Development and Beyond - MDPI, accessed August 31, 2025, https://www.mdpi.com/1422-0067/25/7/3819

15. Astrocytes and the Psychiatric Sequelae of COVID-19: What We Learned from the Pandemic - PMC - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC9362636/

16. Understanding Neuroinflammation in Post-COVID-19 Syndrome: Biological Mechanisms, Diagnostic Biomarkers, and Therapeutic Prospects - Preprints.org, accessed August 31, 2025, https://www.preprints.org/frontend/manuscript/fc261d84e5b0e4f2a9bbf440c873e39d/download_pub

17. The link between SARS-CoV-2 related microglial reactivity and astrocyte pathology in the inferior olivary nucleus - Frontiers, accessed August 31, 2025, https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2023.1198219/full

18. COVID-19-associated neuroinflammation and astrocyte death in the brain linked to ORF3a-induced activation of Sur1-mediated ion channels | mBio - ASM Journals, accessed August 31, 2025, https://journals.asm.org/doi/10.1128/mbio.02012-25

19. SARS-CoV-2 Infection of Microglia Elicits Proinflammatory Activation and Apoptotic Cell Death | Microbiology Spectrum - ASM Journals, accessed August 31, 2025, https://journals.asm.org/doi/10.1128/spectrum.01091-22

20. New Study Finds Plasma Biomarkers of Neuropathogenesis in Hospitalized COVID-19 Patients and Those With Long COVID - Northwestern Medicine ​Breakthroughs for Physicians, accessed August 31, 2025, https://breakthroughsforphysicians.nm.org/plasma-biomarkers-of-neuropathogenesis-in-hospitalized-covid-19-patients-and-those-with-long-covid

21. Linking brain inflammation to long-COVID and other neurological conditions, accessed August 31, 2025, https://pls.llnl.gov/article/52661/linking-brain-inflammation-long-covid-other-neurological-conditions

22. Neuroinflammation and Alzheimer's Disease: A Machine Learning Approach to CSF Proteomics - Olink, accessed August 31, 2025, https://olink.com/publication/neuroinflammation-and-alzheimers-disease-a-machine-learning-approach-to-csf-proteomics

23. Signatures of glial activity can be detected in the CSF proteome - PNAS, accessed August 31, 2025, https://www.pnas.org/doi/abs/10.1073/pnas.2119804119

24. Proteomic signatures of neuroinflammation in Alzheimer's disease, multiple sclerosis and ischemic stroke, accessed August 31, 2025, https://findresearcher.sdu.dk/ws/files/153662496/Proteomic_signatures_of_neuroinflammation_in_Alzheimer_s_disease_multiple_sclerosis_and_ischemic_stroke.pdf

25. Investigating neurological symptoms of infectious diseases like COVID-19 leading to a deeper understanding of neurodegenerative disorders such - ScienceOpen, accessed August 31, 2025, https://www.scienceopen.com/document_file/79964a3b-80ff-4618-9bfb-f86fc1610514/PubMedCentral/79964a3b-80ff-4618-9bfb-f86fc1610514.pdf

26. Long COVID and its association with neurodegenerative diseases: pathogenesis, neuroimaging, and treatment - PMC - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC11024438/

27. A review of cytokine-based pathophysiology of Long COVID symptoms - PMC, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC10103649/

28. The Potential Role of Viral Persistence in the Post-Acute Sequelae of SARS-CoV-2 Infection (PASC) - MDPI, accessed August 31, 2025, https://www.mdpi.com/2076-0817/13/5/388

29. Autopsies show COVID-19 virus in brain, elsewhere in body - CIDRAP, accessed August 31, 2025, https://www.cidrap.umn.edu/covid-19/autopsies-show-covid-19-virus-brain-elsewhere-body

30. NIAID Pandemic Autopsy Study Fosters Long COVID Treatment Trial, accessed August 31, 2025, https://www.niaid.nih.gov/news-events/covid-autopsy-study

31. COVID-19 Virus Can Stay in the Body More Than a Year after Infection | UC San Francisco, accessed August 31, 2025, https://www.ucsf.edu/news/2024/03/427241/covid-19-virus-can-stay-body-more-year-after-infection

32. Study Finds Persistent Infection Could Explain Long COVID in Some People, accessed August 31, 2025, https://www.massgeneralbrigham.org/en/about/newsroom/press-releases/study-finds-persistent-infection-could-explain-long-covid-in-some-people

33. Mechanisms, Effects, and Management of Neurological ... - MDPI, accessed August 31, 2025, https://www.mdpi.com/2227-9059/11/2/377

34. Cerebrospinal Fluid Offers Clues to Post-COVID 'Brain Fog' | UC San Francisco, accessed August 31, 2025, https://www.ucsf.edu/news/2022/01/422156/cerebrospinal-fluid-offers-clues-post-covid-brain-fog

35. Myoclonus in patients with COVID‐19: Findings of autoantibodies against brain structures in cerebrospinal fluid - DiVA portal, accessed August 31, 2025, https://www.diva-portal.org/smash/get/diva2:1817936/FULLTEXT01.pdf

36. CSF Biomarkers in COVID-19 Associated Encephalopathy and Encephalitis Predict Long-Term Outcome - Frontiers, accessed August 31, 2025, https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2022.866153/full

37. Inflammatory Leptomeningeal Cytokines Mediate COVID-19 Neurologic Symptoms in Cancer Patients - PMC - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC7833316/

38. The brain reacting to COVID-19: analysis of the cerebrospinal fluid proteome, RNA and inflammation - PMC - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC9909638/

39. COVID-19-Associated Neurological Disorders: The Potential Route of CNS Invasion and Blood-Brain Barrier Relevance - MDPI, accessed August 31, 2025, https://www.mdpi.com/2073-4409/9/11/2360

40. Blood Markers Show Neural Consequences of LongCOVID-19 - PMC - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC10969290/

41. Study reveals no neuroinflammation in long COVID patients with neuropsychiatric symptoms, accessed August 31, 2025, https://www.news-medical.net/news/20231113/Study-reveals-no-neuroinflammation-in-long-COVID-patients-with-neuropsychiatric-symptoms.aspx

42. Mild respiratory COVID can cause multi-lineage neural cell and myelin dysregulation - PMC, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC9189143/

43. Brain Fog: New Study Examines Causes of This Long COVID Symptom, accessed August 31, 2025, https://medicine.yale.edu/news-article/brain-fog-new-study-examines-causes-of-this-long-covid-symptom/

44. CSF inflammatory proteomic profiles in PASC-CI according to prognosis... - ResearchGate, accessed August 31, 2025, https://www.researchgate.net/figure/CSF-inflammatory-proteomic-profiles-in-PASC-CI-according-to-prognosis-PCA-in-118-CSF_fig2_380557512

45. Self-Reported Neuropsychiatric Post–COVID-19 Condition and CSF ..., accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC10638645/

46. Long COVID Immune Profiles Appear Largely Normal | The Transmission, accessed August 31, 2025, https://www.unmc.edu/healthsecurity/transmission/2023/11/14/long-covid-immune-profiles-appear-largely-normal/

47. Yale study sheds light on new findings on the causes of long Covid symptoms, accessed August 31, 2025, https://yaledailynews.com/blog/2023/11/30/yale-study-sheds-light-on-new-findings-on-the-causes-of-long-covid-symptoms-2/

48. (PDF) Neuroinflammation in post-acute sequelae of COVID-19 (PASC) as assessed by [11C]PBR28 PET correlates with vascular disease measures - ResearchGate, accessed August 31, 2025, https://www.researchgate.net/publication/379937979_Neuroinflammation_in_post-acute_sequelae_of_COVID-19_PASC_as_assessed_by_11CPBR28_PET_correlates_with_vascular_disease_measures

49. Cerebrospinal Fluid Proteomic Profiling Reveals Proteins Associated with Neuroinflammatory Response in COVID-19 Patients - PMC - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC12199065/

50. Cerebrospinal fluid analysis in patients with COVID-19-associated central nervous system manifestations: a systematic review, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC9648929/

51. Cerebrospinal fluid findings in COVID-19 patients with neurological symptoms - PMC, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC7417278/

52. Cerebrospinal fluid analysis in patients with COVID-19-associated central nervous system manifestations: a systematic review - SciELO, accessed August 31, 2025, https://www.scielo.br/j/anp/a/Ny43FWPQ6Tmy7fQ4rBvYpcC/

53. Myoclonus-Ataxia Syndrome Associated with COVID-19 - Journal of Movement Disorders, accessed August 31, 2025, https://www.e-jmd.org/journal/view.php?number=333

54. Myoclonus and Cerebellar Ataxia Following COVID‐19 - PMC - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC7436437/

55. Acute Post‐Infectious Cerebellar Ataxia Due to COVID‐19 - PMC - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC8088087/

56. COVID-19-Associated Cerebellar Ataxia: A Case Series | Request PDF - ResearchGate, accessed August 31, 2025, https://www.researchgate.net/publication/382143609_COVID-19-Associated_Cerebellar_Ataxia_A_Case_Series

57. Neurocognitive Impairment in Long COVID: A Systematic Review - Oxford Academic, accessed August 31, 2025, https://academic.oup.com/acn/article/40/1/125/7689909

58. Study links long COVID brain fog to inflammation and stress markers - News-Medical, accessed August 31, 2025, https://www.news-medical.net/news/20250515/Study-links-long-COVID-brain-fog-to-inflammation-and-stress-markers.aspx

59. Neuroscience: Long Covid Brain Fog Linked to Inflammation and Stress Markers, accessed August 31, 2025, https://meassociation.org.uk/2025/05/neuroscience-long-covid-brain-fog-linked-to-inflammation-and-stress-markers/

60. Scientists find two brain biomarkers in long COVID sufferers may be what's causing their brain fog, other cognitive issues | EurekAlert!, accessed August 31, 2025, https://www.eurekalert.org/news-releases/1083190

61. Long COVID: systemic inflammation and obesity as therapeutic targets - PubMed Central, accessed August 31, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC9034853/