Advances in Understanding Long COVID: Pathophysiological Mechanisms and the Role of Omics Technologies in Biomarker Identification

Abstract

Long coronavirus disease (COVID) is a multisystem condition that affects a significant proportion of individuals following severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, with persistent symptoms ranging from fatigue and cognitive dysfunction to cardiovascular disorders. It is estimated that 30–60% of infected individuals experience symptoms lasting more than 12 weeks. Despite advances in understanding acute infection, the pathophysiological mechanisms underlying long COVID remain unclear. Current hypotheses suggest that viral persistence, immune dysfunction, and metabolic alterations play central roles. Omics approaches, including metabolomics, proteomics, and lipidomics, have played a crucial role in investigating molecular changes, identifying biomarkers, and refining therapeutic strategies. This review discusses recent advances in understanding long COVID, addressing its mechanisms, risk factors, the impact of viral variants, and the role of vaccination, with an emphasis on the importance of omics technologies in elucidating this condition.

Graphical Abstract

Source: Developed using BioRender.

Key Points

Long COVID is a multisystem condition that affects various body systems, with persistent symptoms such as fatigue, cognitive dysfunction, and cardiovascular disorders. These symptoms significantly impact the quality of life for individuals and can last for weeks or even months after the acute phase of SARS-CoV-2 infection.

The underlying mechanisms of long COVID are still unclear, but current hypotheses suggest that viral persistence, immune dysfunction, and metabolic alterations may play central roles in the development and persistence of symptoms. These factors could contribute to long-term disruptions in various physiological systems.

Omics approaches, including metabolomics, proteomics, and lipidomics, have been instrumental in exploring molecular changes associated with long COVID. These technologies are helping to identify potential biomarkers for diagnosis, as well as refining therapeutic strategies aimed at alleviating the condition. Omics research is crucial for a deeper understanding of the pathophysiology of long COVID.

Introduction

Coronavirus disease 2019 (COVID-19) emerged as a global pandemic in December 2019, infecting over 777 million people and causing approximately 7.1 million deaths by January 2025 [1, 2]. Although advances in vaccination and treatment have alleviated the severity of acute infections, many individuals continue to experience persistent symptoms beyond the initial phase of the disease [3]. This clinical condition, known as long COVID, encompasses multisystem manifestations that can last for weeks or months, significantly impacting the quality of life [4].

In the USA, long COVID is estimated to affect between 44.69 and 48.04 million people annually [5]. Moreover, 30–60% of individuals infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) experience prolonged symptoms lasting 12 weeks or more [6]. Given the ongoing circulation of the virus, this number is expected to rise. Long COVID has been associated with over 200 symptoms, which vary widely among individuals [7]. The most common symptoms include fatigue, cognitive dysfunction, headache, myalgia, and arthralgia [8]. The diversity and unpredictability of these symptoms make diagnosis and treatment particularly challenging.

The pathophysiological mechanisms underlying long COVID are not fully understood, but several hypotheses have been proposed. Evidence suggests that viral persistence and immune dysfunction play a central role in sustaining symptoms [9–11]. Viral fragments or residual SARS-CoV-2 proteins may persist in the body, triggering chronic inflammatory responses and immune dysfunctions that affect multiple systems [12]. Moreover, changes in energy metabolism and the gut microbiome have also been associated with the progression of long COVID, underscoring its multifactorial nature [13, 14]. Specific risk factors, such as advanced age, female sex, preexisting comorbidities, and the severity of the initial infection, are associated with a higher chance of developing long COVID [15–17]. However, even young and previously healthy individuals may experience persistent symptoms, suggesting the influence of genetic and immunological factors [18].

The evolution of SARS-CoV-2 variants and vaccination are also critical factors in understanding long COVID [19, 20]. Some variants appear to be associated with a higher incidence of prolonged symptoms, whereas vaccination has shown protective effects, reducing both the frequency and severity of this condition [21].

However, the exact relationship between viral variants, immune response, and symptom persistence remains unclear [19]. From a molecular perspective, the multisystem symptoms of long COVID are characterized by various metabolic and inflammatory changes. Dysfunctions in lipid, protein, and energy metabolism have been implicated in the pathophysiology of the disease, contributing to symptoms such as chronic fatigue, cognitive dysfunction, and cardiovascular disorders [22, 23]. In this context, mass-spectrometry-based “omics” approaches have increasingly been used to investigate the molecular changes associated with long COVID [24–26]. Techniques such as metabolomics, proteomics, and lipidomics enable the identification of biomarkers and metabolic pathways involved in disease progression, deepening the understanding of its mechanisms and aiding in the development of therapeutic strategies [27–29]. However, despite advances in understanding the acute phase of COVID-19, long COVID remains underexplored from this perspective, highlighting the need for further research.

Given this scenario, we review the significant advances in long COVID research, focusing on its pathophysiological mechanisms, viral persistence, immune dysfunction, risk factors, the influence of viral variants, the role of vaccination, and the molecular implications of multisystemic symptoms. Moreover, we emphasize the potential of “omics” approaches in investigating long COVID, highlighting their role in identifying biomarkers and clarifying the biochemical processes underlying the condition.

Pathophysiological Mechanisms of Long COVID

The pathophysiological mechanisms of long COVID have been extensively studied [7, 22, 30]. SARS-CoV-2, a single-stranded RNA coronavirus, causes severe disease, similar to Middle East respiratory syndrome coronavirus (MERS-CoV) and SARS-CoV [31]. Approximately 80% of infected individuals experience mild or asymptomatic infections, while 15% develop severe illness and 5% progress to critical conditions [32]. The acute phase of COVID-19 lasts approximately 4–5 weeks from symptom onset [32]. However, a significant proportion of patients continue to experience one or more symptoms beyond this period, leading to long-term complications classified as long COVID (Fig. 1) [33].

Fig. 1.

Following exposure to SARS-CoV-2, individuals may develop symptoms during the acute phase, which typically spans up to 4 weeks from symptom onset and is characterized by high viral load. The post-acute phase occurs between weeks 4 and 12, during which ongoing or relapsing symptoms may persist despite viral clearance. Symptoms that continue or emerge from week 12 onwards are classified as long COVID, reflecting prolonged or new-onset complications beyond the initial infection.

Created using BioRender.com

During the acute infection phase, SARS-CoV-2 initially binds to the host cell surface via the spike protein, which interacts with the angiotensin-converting enzyme 2 (ACE2) receptor [34]. Following this binding, the virus is internalized through endocytosis [35]. Once inside the host cell, the viral RNA genome is released into the cytoplasm [34]. The virus then hijacks the host cell’s transcriptional and translational machinery to produce large quantities of viral proteins and RNA while simultaneously inhibiting the translation of host mRNA. This process results in the production of infectious viral progeny and ultimately leads to cell death [34, 36]. If the virus is not suppressed, the host immune system begins releasing proinflammatory cytokines, thus triggering hyperinflammation [37–41]. In individuals with long COVID, immune responses have been reported to exhibit elevated levels of nucleocapsid antibodies or increased markers of immune activation and exhaustion. However, these findings are not consistent across all studies, reflecting the complexity and heterogeneity of long COVID. The lack of a universally observed immunological pattern likely arises from differences in study design, patient populations, control groups, and timing of sample collection, among other factors. This variability remains a significant challenge in deciphering the pathophysiology of post-acute sequelae of SARS-CoV-2 infection.

Immune responses in individuals with long COVID exhibit elevated levels of nucleocapsid-specific immunoglobulin G (IgG) and enhanced neutralizing capacity for up to 8 months postinfection. Additionally, persistent alterations in T cell activity, characterized by increased expression of PD-1 and TIM-3 on both CD4⁺ and CD8⁺ T cells, were observed [42]. The pathogenesis of long COVID is also linked to the overexpression of proinflammatory cytokines, such as interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-α), interleukin-1β (IL-1β), and interferon-induced protein 44 (IFI44) [43]. These factors are commonly observed in patients with long COVID, suggesting that chronic inflammation and immune activation are key drivers of persistent symptoms [44–46]

Type I interferon (IFN) signaling, particularly during the initial stages of infection, also plays a critical role in the development of severe clinical manifestations. However, its sustained activation may impair the immune system’s ability to mount an effective antiviral response [47]. Moreover, type II IFN signaling and canonical NF-κB signaling (specifically TNF-α-mediated pathways) are among the most differentially enriched pathways in individuals with long COVID [47]. The presence of IFN-γ, a type II interferon, has been identified as a promising biomarker for long COVID [48].

These inflammatory pathways contribute to a prolonged state of immune activation, which may be exacerbated by the continued presence of viral products in the body (Fig. 2) [48]. In addition to the exacerbated inflammatory response, persistent immune activation is linked to thromboinflammation and dysfunction of the complement system [49]. Evidence indicates that persistent SARS-CoV-2 infection may induce immunosuppression, characterized by reduced expression of ribosomal proteins, which can compromise essential cellular functions and exacerbate long COVID symptoms [50].

Fig. 2.

Pathophysiological mechanisms of long COVID. Scheme of mechanisms contributing to long COVID, including the persistence of a viral reservoir and autoimmunity. The resulting consequences include a cytokine storm, organ damage, chronic inflammation, and an altered immune status, which collectively contribute to the prolonged symptoms observed in long COVID. These factors lead to systemic effects, including endothelial dysfunction, microbiome dysbiosis, mitochondrial dysfunction, metabolic dysregulation, and dysregulation of genes involved in tissue repair and regeneration.

Created using BioRender.com

While various factors contribute to long COVID susceptibility, emerging evidence suggests that genetic predisposition, particularly ACE-2 polymorphisms, plays a crucial role in its development [51]. Viral activity during early infection is also associated with the risk of developing long COVID [52, 53]. Higher viral replication correlates with more severe acute cases, which increases the likelihood of long COVID [54]. This risk appears lower with the Omicron variant, possibly due to its reduced viral load compared with earlier variants [55, 56].

Mitochondrial dysfunction is another key molecular mechanism in long COVID. Impaired mitochondrial function reduces cellular energy production, exacerbating oxidative stress and systemic inflammation. This disruption compromises the integrity of endothelial cells and aggravates vascular disorders [14]. Moreover, mitochondrial dysfunction disrupts cellular homeostasis by altering mitochondrial dynamics and impairing tissue recovery [14].

Long COVID may be a result of a virally induced, chronic metabolic imbalance driven by mitochondrial dysfunction. In individuals with pre-existing suboptimal mitochondrial function, the SARS-CoV-2 infection can trigger a persistent inflammatory cycle, leading to symptoms such as brain fog, fatigue, and platelet dysfunction. The virus redirects the host’s metabolism toward replication, and in response, the host’s metabolism tries to control the infection. However, in some individuals, this response does not resolve, and reactive oxygen species continue to drive inflammation and shift toward glycolysis [14, 42, 57, 58].

The gut microbiota also plays a key role in immune function and disease susceptibility, mainly through the gut–lung and gut–brain axes. Dysbiosis can contribute to the development of long COVID [59].

Viral Persistence and Immune Dysfunction

The persistence of SARS-CoV-2 or its fragments in the body is considered a key factor in causing long COVID [60]. The virus uses sophisticated immune evasion strategies, such as antigenic variation, suppression of immune system components, and colonization of immune-privileged sites, making its eradication challenging [61]. The persistence of SARS-CoV-2 in patients with COVID-19 is expected, as coronaviruses are known to establish persistent infections both in vitro and in vivo [62].

Several studies support this hypothesis. Zuo et al. (2024) analyzed tissue samples from 225 patients who had recovered from mild COVID-19 and detected viral RNA in plasma, blood cells, and multiple tissues up to 4 months postinfection [63]. Similarly, an autopsy study conducted in the USA revealed that SARS-CoV-2 can persist in the body for months; although the respiratory tract is the primary site of infection, researchers found viral RNA in the heart and brain for up to 230 days postinfection [64]. Indeed, SARS-CoV-2 can disseminate widely. Another postmortem study confirmed the presence of viral RNA in multiple organs up to 230 days after symptom onset [65]. These findings suggest that SARS-CoV-2 can persist in the body for an extended period, even after the acute phase of the disease.

The gastrointestinal tract is the most studied site for harboring residual viral deposits. This does not imply that other sites are not involved; rather, it indicates that the gastrointestinal tract is more accessible for biopsy and collection of fecal samples. Two primary mechanisms facilitate viral persistence in the gastrointestinal tract: mucociliary clearance and the process of swallowing. In the intestine, epithelial cells, such as enterocytes, express the ACE2 receptor, which facilitates viral entry, binding, and replication [13]. Viral antigens have been detected in feces after SARS-CoV-2 infection [66]. For example, Natarajan et al. [67] reported that viral RNA persisted in the feces of patients with mild to moderate COVID-19 for up to 7 months after acute infection. During this period, these patients continued to experience gastrointestinal symptoms consistent with long COVID.

In addition to the virus’s inherent ability to persist in the body, immunological factors can also contribute to its persistence. Immunosuppression, for example, can prolong the duration of an infection. One study reported the presence of the virus for 154 days in a patient with antiphospholipid syndrome who had received multiple immunosuppressants [68]. T cell dysfunction—critical for controlling viral latency—may also contribute to the persistence of SARS-CoV-2. This dysfunction has been linked to the reactivation of latent viruses, such as Epstein–Barr virus, which was observed in 66.7% of patients with long COVID who were analyzed [69]. Furthermore, Gaebler et al. [70] reported that, between 1.3 and 6.2 months postinfection, levels of anti-spike immunoglobulin M (IgM) and IgG antibodies declined significantly, while levels of receptor binding domain (RBD)-specific memory B cells remained stable. These findings suggest that residual SARS-CoV-2 antigens or incomplete viral clearance sustain continuous immune stimulation [70].

Risk Factors, Variants, and the Protective Role of Vaccination in Long COVID

Estimating the true prevalence of long COVID is challenging due to the heterogeneity of symptoms, which may fluctuate over time or persist chronically. Although the risk factors for long COVID are still under investigation, genetic predisposition, sex, age, and preexisting comorbidities, as well as environmental, behavioral, and lifestyle factors may influence the manifestation of symptoms [16, 71–75]. These risk factors can be categorized into two main groups: (1) preinfection factors, which include preexisting individual characteristics, and (2) infection-related factors, which encompass disease-specific aspects such as clinical severity, initial symptoms, viral load, and the need for hospitalization [76]. A retrospective cohort study involving approximately 2.1 million patients in the USA supports this classification, demonstrating that advanced age, preexisting comorbidities, and the severity of acute infection significantly increase the risk of developing persistent post-COVID symptoms [72].

In addition to these factors, the SARS-CoV-2 variant and vaccination status also play a significant role in determining the risk of long COVID. The Omicron variant and its subvariants, which are currently predominant, appear to be associated with milder and shorter-lasting persistent symptoms compared with earlier variants, such as Alpha, Beta, and Delta [77]. A study involving two population cohorts found that the risk of developing long COVID is lower after Omicron infection and vaccination [78].

Vaccination against COVID-19 significantly reduces the risk of developing long COVID [79, 80]. Recent studies suggest that protection is more substantial in individuals who have received two doses than in those who have received only one. Moreover, vaccination remains effective even when given after a SARS-CoV-2 infection [81]. In the UK, a 41% reduction in the risk of long COVID was observed among individuals who received two vaccine doses [74]. Additionally, vaccination alleviates specific symptoms, such as brain fog and myalgia, thus improving overall quality of life [82]. Other risk factors are summarized in Table 1.

Table 1.

Risk factors associated with long COVID

Risk factor	Observations	References
Age	The older the age, the higher the risk of long COVID	[16, 83, 84]
Female sex	Women are more likely to experience persistent symptoms	[16, 85]
Hospitalization & history of severe COVID	The risk of sequelae is proportional to the severity of the acute infection, being higher in patients who were hospitalized or admitted to intensive care	[18, 86]
Lack of vaccination	Individuals who are not vaccinated have a higher incidence of long COVID, regardless of the virus variant	[23, 74, 78–81, 87]
Comorbidities	They influence disease progression by affecting pathological mechanisms at various stages of infection.	[15, 16, 76, 88, 89]
Reinfection	Reinfection with SARS-CoV-2 increases the risk of death, hospitalization, and sequelae in various organ systems	[90, 91]
Genotype of SNX9	The specific genotype of the SNX9 gene is associated with an increased risk of severe infection in long COVID, with the relationship mediated by interactions with the KLF15 and RYR3 genes	[92]

Molecular Implications in Multisystemic Symptoms

Long COVID is a challenging syndrome to diagnose due to its wide range of symptoms, which can exceed 200 clinical manifestations [7]. The most common associations are neurological, including chronic fatigue, headache, anxiety, and insomnia [93–95]. Other high-risk conditions, such as depression, pulmonary complications, and cardiac dysfunction, are also characteristic of long COVID [93]. Persistent lung injuries following COVID-19 are linked to an overload of cytotoxic functions, including gamma delta (γδ) T cells and natural killer (NK) cells, as well as an increased number of CD4⁺ and CD8⁺ lymphocytes. Changes in hemoglobin levels due to pulmonary dysfunction have also been observed, along with elevated inflammatory biomarkers, including C-reactive protein (CRP), tumor necrosis factor-alpha (TNF-α), and interferon (IFN). Notably, elevated levels of NK cells have been associated with symptoms such as productive cough [96].

Cardiovascular complications are also common in long COVID. Persistent myocardial inflammation and elevated cardiac troponin levels can be detected for up to 2 months after diagnosis [97]. An observational study found that 32% of survivors of COVID-19 exhibit cardiac damage 3 months after infection onset [73]. These effects likely arise from the abundance of ACE2 receptors in cardiomyocytes, which provide a direct pathway for SARS-CoV-2 infection [98]. Chronic inflammation of cardiomyocytes can lead to myositis and cell death [98]. Prolonged inflammation and cellular damage can trigger fibrotic changes, reducing cell adhesion, which may contribute to arrhythmias and coagulopathy. Moreover, cardiovascular factors can also cause neurological symptoms, such as brain fog [99]. A recent study identified a hypercoagulable state and microclots in patients with long COVID as key contributors to this symptom [100].

Several molecular alterations have been associated with the neurological symptoms of long COVID. For example, elevated levels of exosomes containing SARS-CoV-2 proteins have been detected in extracellular vesicles derived from neurons and astrocytes [57]. The virus can cross the blood–brain barrier and induce neuroinflammation in the brain parenchyma, whereas hyperinflammation in the brainstem may result in autonomic dysfunction. Neuroinflammatory responses in the central nervous system (CNS) have been implicated in neuropsychiatric manifestations, including chronic malaise, fatigue, sleep disorders, ageusia, anosmia (loss of taste and smell), cognitive impairment, and even stroke [93]. Neurological manifestations are among the most prevalent in long COVID, accounting for up to 70% of reported symptoms [101]. In the UK, one in three individuals reported experiencing neuropsychiatric symptoms 6 months after SARS-CoV-2 infection [102]. Imaging studies have identified brain abnormalities in patients with cognitive impairment during the subacute phase of COVID-19 compared with healthy controls [103].

In the urinary system, hyperinflammation of renal tissues can activate the complement system, contributing to glomerulosclerosis [104]. Moreover, acute kidney injury has been observed in hospitalized patients with COVID-19, with 35% of recovered individuals experiencing reduced renal function [105].

SARS-CoV-2 can also affect the pancreas, triggering pancreatitis during the acute phase of the disease. In cases of long COVID, pancreatic damage may arise from both direct viral invasion and systemic inflammation. ACE2, the primary receptor for SARS-CoV-2, functions as an endocrine regulator in the renin–angiotensin–aldosterone system. Another receptor, TMPRSS2, is expressed in pancreatic β-cells. Viral infection directly impairs ACE2- and TMPRSS2-expressing cells, disrupting the renin–angiotensin–aldosterone system and long-term metabolic homeostasis [106].

The gut microbiota plays a crucial role in immune response, and its dysbiosis has been linked to long COVID, with effects persisting for up to a year after hospital discharge [107]. In patients with long COVID, an increase in CD8⁺/CD4⁺ cytotoxic T cell receptor clonotypes was observed approximately 2.5 months after the onset of symptoms. These molecular changes are associated with symptoms such as disrupted bowel movements, heartburn, and nausea [63, 108].

Omics Approaches in Long COVID

Mass spectrometry (MS)-based omics approaches have played a crucial role in uncovering the molecular alterations associated with prolonged SARS-CoV-2 infection. Among these, proteomics, lipidomics, and metabolomics have been extensively used to investigate the biological pathways affected in individuals with post-acute sequelae of COVID-19 (PASC), commonly referred to as long COVID. These high-throughput analytical techniques enable the comprehensive profiling of proteins, lipids, and metabolites in biological samples, providing insights into persistent inflammation, immune dysregulation, metabolic dysfunction, and tissue-specific damage.

Metabolomic

According to a recent study, patients with long COVID exhibit significant metabolic alterations, with 53 metabolites differing between acute and long COVID stages, and 27 remaining dysregulated even after 2 years. Key metabolites, such as lactic acid, arginine, and the lactate/pyruvate and ornithine/citrulline ratios, were identified as markers for more severe cases, suggesting mitochondrial dysfunction, altered energy metabolism, and persistent inflammation as central features of the syndrome [109]. An integrated analysis of serum/plasma in patients with LTCS revealed significant differences in lactate and pyruvate compared with healthy controls and patients with long COVID. Correlation analysis showed that histidine and glutamine were primarily linked with proinflammatory cytokines, while alterations in triglycerides and apolipoproteins Apo-A1 and A2 were observed. LTCS and acute COVID-19 samples were distinguished by phenylalanine, 3-hydroxybutyrate, and glucose concentrations, indicating imbalanced energy metabolism [110]. The plasma metabolomics analysis revealed significant alterations in metabolomic pathways in patients with long COVID compared with recovery and healthy control groups. Patients with long COVID (LC) showed elevated proinflammatory biomarkers (e.g., IL-1α, IL-6, and TNF-α) and reduced ATP levels, indicating persistent metabolomic abnormalities 12 months after acute COVID-19 (Table 2). Additionally, a notable reduction in sarcosine and serine concentrations in patients with LC correlated inversely with depression, anxiety, and cognitive dysfunction scores, suggesting long-lasting metabolic recovery challenges [111].

Table 2.

Metabolomics studies on long COVID using mass spectrometry

Metabolites/markers	Biomarkers/other observations	References
Lactic acid, arginine, lactate/pyruvate ratio, and ornithine/citrulline ratio	Mitochondrial dysfunction, altered energy metabolism, and persistent inflammation	[109]
Histidine, glutamine, triglycerides, apolipoproteins Apo-A1, & Apo-A2	Proinflammatory cytokines (e.g., IL-1α, IL-6, and TNF-α); imbalanced energy metabolism	[110]
Sarcosine, serine, phenylalanine, 3-hydroxybutyrate, & glucose	Elevated proinflammatory biomarkers (IL-1α, IL-6, and TNF-α), reduced ATP	[111]

Lipidomic

Garrido et al. (2024) investigated lipidomic changes in patients with long COVID using untargeted lipidomics to analyze plasma samples from 147 patients. The analysis revealed significant alterations in various lipid subclasses, including elevated levels of lysophosphatidylglycerols (LPGs) and phosphatidylethanolamines (PEs) and reduced levels of lysophosphatidylcholines (LPCs), suggesting these lipids as potential biomarkers for long COVID [112]. Lopez-Hernandez et al. (2023) evaluated the clinical and lipidomic profiles of patients who recovered from COVID-19 who had experienced mild or severe infections during the acute phase of the first epidemic wave, assessing them 2 years after recovery. Nontargeted lipidomics were used to analyze lipidomic alterations. The study found that fatigue (59%) and musculoskeletal symptoms (50%) were the most persistent. Functional analyses revealed that sterols, bile acids, isoprenoids, and fatty esters were the predicted metabolic pathways affected in both patients with COVID-19 and long COVID (Table 3). Several phosphatidylcholines and sphingomyelins were found in higher concentrations in patients with long COVID compared with controls [113].

Table 3.

Summary of key lipidomic findings in long COVID studies

Cohort/design	Method	Main lipidomic findings	References
147 patients with long COVID	Un-targeted lipidomics (plasma)	↑LPGs, ↑PEs, ↓LPCs—identified as potential lipid biomarkers for long COVID	[112]
Recovered patients (mild or severe acute COVID), assessed 2 years postinfection	Non-targeted lipidomics	↑Phosphatidylcholines, ↑sphingomyelins; altered pathways: sterols, bile acids, isoprenoids, fatty esters	[113]

Proteomic

Captur et al. investigated the proteomic response in the plasma of healthcare workers infected with SARS-CoV-2 during the first wave of the pandemic in the UK [114]. Using multiple reaction monitoring (MRM) mass spectrometry to analyze 91 preselected proteins, the authors identified persistent proteomic changes correlated with symptom severity and immune response. These changes involved pathways related to lipid metabolism, complement activation, coagulation, autophagy, and lysosomal function, suggesting a prolonged impact of infection on molecular homeostasis (Table 4). Li et al. [115] investigated the systemic metabolic signatures of survivors of nonsevere COVID-19 6 months postdischarge, revealing persistent metabolic abnormalities through metabolomics analysis. Despite some recovery, such as normalization of the kynurenine pathway and itaconic acid levels, survivors exhibited ongoing dysregulation in serum amino acids, organic acids, purine, fatty acids, and lipid metabolism. These metabolic disturbances were found to be associated with liver injury, altered mental health, impaired energy production, and heightened inflammatory responses [115]. Machado et al. [116] reported that 19 salivary metabolites showed significantly different levels between patients with long COVID and healthy individuals. Patients with long COVID showed reduced levels of calenduloside G methyl ether, Gly-Pro-Lys, and creatine, with the latter associated with fatigue and decreased physical capacity [116].

Table 4.

Key proteomic studies and findings in long COVID

Description	Key findings	References
Investigated the proteomic response in healthcare workers infected with SARS-CoV-2 during the first wave in the UK	Identified persistent proteomic changes correlated with symptom severity and immune response; changes involved lipid metabolism, complement activation, coagulation, autophagy, and lysosomal function	[114]
Analyzed systemic metabolic signatures of nonsevere COVID-19 survivors 6 months postdischarge	Found persistent metabolic abnormalities in serum amino acids, organic acids, purine, fatty acids, and lipid metabolism, associated with liver injury, altered mental health, and increased inflammation	[115]
Reported salivary metabolite differences between patients with long COVID and healthy individuals	Found 19 salivary metabolites with significant differences, with reduced levels of calenduloside G methyl ether, Gly-Pro-Lys, and creatine in patients with long COVID, the latter linked to fatigue and decreased physical capacity	[116]
Studied the proteomic changes in individuals with PV/PIS	Revealed elevated levels of serum amyloid A1 and A2, attractin, and coagulation factors X and XI, indicating prolonged activation of inflammatory response and coagulation system	[117]
Investigated immune responses in patients with long COVID	Found lower neutralizing antibody titers and subtle increases in co-inhibitory receptors (PD-1, TIM-3) on SARS-CoV-2 non-spike-specific CD8⁺ T cells; identified a plasma biomarker signature linked to breathlessness	[118]
Explored sustained complement activation in long COVID	Observed persistently elevated soluble C5bC6, reduced membrane-bound TCCs, and markers of thromboinflammation, highlighting ongoing endothelial damage and immune dysregulation	[49]
Conducted multiomic profiling of plasma from patients with COVID-19 6 months postinfection	Identified altered metabolic pathways, with a 20-molecule signature predicting long COVID outcomes with high accuracy (AUC 0.96)	[25]
Studied proteomic changes in COVID-19 survivors over time	Revealed that, while most immune, complement, coagulation, and metabolic pathways normalized within 2 years, Fc receptor signaling and neurogenesis pathways remained altered	[119]
Used machine learning to identify protein panels distinguishing patients with long COVID	Identified key proteins reflecting multisystem involvement, including leukocytes and platelets, supporting the role of these cells in long COVID pathophysiology	[120]
Investigated immune cell phenotypes in outpatients with long COVID	Found altered immune cell phenotypes, with NK cells in a resting state and neutrophils forming extracellular traps; linked vascular inflammation to ANGPT1, VEGFA, and TGF-β1 signaling	[121]

AUC area under the curve, PV/PIS postvaccination/postinfection syndrome, NK natural killers, TCC terminal complement complexes

In line with these persistent molecular alterations, a recent study used mass spectrometry to investigate the plasma of 30 individuals with postvaccination/postinfection syndrome (PV/PIS), revealing distinct proteomic changes compared with healthy controls. The authors observed elevated levels of serum amyloid A1 and A2, attractin, and coagulation factors X and XI, suggesting prolonged activation of the inflammatory response and coagulation system. Additionally, downregulation of immunoregulatory proteins was identified, indicating a state of persistent immune dysregulation. Although there is some overlap with long COVID, the proteomic profile of PV/PIS showed only partial overlap with data from patients infected before the availability of vaccines, highlighting a distinct molecular signature associated with endothelial dysfunction, immune activation, and sustained coagulopathy [117]. A multicohort study involving individuals from Sweden and the UK revealed that patients with long COVID exhibited lower neutralizing antibody titers compared with healthy convalescents. Although no consistent alterations in immune cell composition or antiviral T cell immunity were observed, individuals with long COVID showed a subtle increase in co-inhibitory receptors, particularly PD-1 and TIM-3, on SARS-CoV-2 non-spike-specific CD8⁺ T cells. Proteomic analysis identified a plasma biomarker signature associated with breathlessness, involving proteins such as CCL3, CD40, IKBKG, IL-18, and IRAK1. Additionally, dysregulated pathways linked to cell cycle progression, lung injury, and platelet activation were observed, suggesting persistent inflammatory and apoptotic processes that may contribute to symptomatology and offer potential diagnostic or therapeutic targets in long COVID [118].

Recent findings further support a role for sustained complement activation in long COVID. Persistently elevated soluble C5bC6, reduced membrane-bound terminal complement complexes (TCCs), and markers of thromboinflammation—including increased von Willebrand factor (vWF) and platelet–monocyte aggregates—highlight ongoing endothelial damage and immune dysregulation even 6 months postinfection [49]. Multiomic profiling of plasma from patients with COVID-19 revealed persistent inflammation and metabolic dysregulation 6 months postinfection. Altered pathways included arginine, methionine, taurine, and tricarboxylic acid (TCA) cycle metabolism, and a 20-molecule signature predicted long COVID outcomes with high accuracy (area under the curve (AUC) = 0.96) [25]. Longitudinal proteomic profiling of COVID-19 survivors revealed four distinct recovery trajectories across biological pathways. While most immune, complement, coagulation, and metabolic pathways normalized within 2 years, Fc receptor signaling and neurogenesis pathways remained altered. Specific proteins were associated with lung function and persistent anosmia, indicating prolonged molecular disruption [119]. Machine learning identified protein panels with perfect accuracy (AUC = 1.00) in distinguishing patients with long COVID. Key proteins reflected multisystem involvement, with leukocytes and platelets emerging as central players in pathophysiology, as supported by natural language processing of symptom data [120]. Outpatients with long COVID showed altered immune cell phenotypes, with NK cells in a resting state and neutrophils forming extracellular traps. Vascular inflammation was linked to ANGPT1, VEGFA, and TGF-β1 signaling, suggesting a vasculo-proliferative state that could drive neurologic and cardiometabolic dysfunction in long COVID. Several markers were validated, indicating potential prognostic value [121].

Future Directions

Long COVID remains a multifaceted condition, impacting multiple organ systems and manifesting through a wide range of symptoms that span from mild, flu-like manifestations to severe complications involving cardiovascular and neurological systems. Its pathogenesis involves complex interactions across numerous cellular and molecular pathways, highlighting its unique impact on physiological processes. As a result, personalized therapeutic interventions, such as immune response modulation, the restoration of cellular metabolism, and therapies targeting mitochondrial function, are emerging as potential strategies to address the diverse clinical manifestations of long COVID. The condition’s multifaceted nature necessitates an interdisciplinary approach, integrating clinical, immunological, and molecular assessments alongside consideration of genetic predisposition, preexisting conditions, and the interplay between viral persistence and the host immune response.

The role of omics technologies—particularly proteomics, lipidomics, and metabolomics—has been central to advancing our understanding of long COVID. These high-throughput techniques provide invaluable insights into the molecular alterations associated with the condition, facilitating the identification of biomarkers to predict disease severity, monitor progression, and inform treatment strategies. Integration with advanced computational methods enables a more comprehensive view of these molecular dynamics over time. Longitudinal studies are particularly valuable for capturing biomarker evolution from the acute phase through recovery and for assessing the effects of interventions such as rehabilitation programs or emerging therapies.

However, significant challenges remain. The studies often face limited sample sizes and must contend with the profound heterogeneity of long COVID—both in symptom presentation and in underlying biological mechanisms. The wide range of manifestations, including respiratory, neurological, and metabolic, suggests multiple distinct pathophysiological pathways. This diversity complicates data interpretation, as integrating findings across heterogeneous populations may dilute individual-specific signals or obscure relevant associations. Additionally, discrepancies across studies frequently stem from variations in design, cohort characteristics, and symptom classification, highlighting the urgent need for standardized protocols and stratified analyses to disentangle the complex biology of long COVID, particularly given the heterogeneity of symptoms, patient populations, and potential underlying mechanisms.

Further investigation into the role of the microbiota in long COVID could yield new therapeutic avenues, especially concerning gut–brain interactions, immune system modulation, and gastrointestinal symptoms. Exploring dysbiosis of the gut microbiome could help identify microbiota-targeted therapies, such as prebiotics, probiotics, or dietary interventions, to alleviate persistent symptoms. Additionally, the influence of various SARS-CoV-2 variants on long COVID, particularly in terms of immune evasion, viral persistence, and symptomatology, remains an important area of research.

The impact of comorbidities, such as diabetes, hypertension, and cardiovascular disease, on the progression and severity of long COVID is another significant research avenue. Understanding how these conditions influence disease pathophysiology will be crucial in developing targeted treatments for individuals with underlying health issues. The bidirectional relationship between long COVID and mental health conditions, such as depression, anxiety, and cognitive impairment, should also be carefully considered, as these factors can significantly hinder recovery and quality of life.

In parallel, examining the role of COVID-19 vaccinations, including booster doses and their effectiveness against emerging variants, remains crucial in understanding their potential to prevent or mitigate long COVID. Assessing the long-term benefits of vaccination on long COVID incidence could have important public health implications. Research into post-COVID rehabilitation programs, focusing on both physical and mental recovery, is also essential for improving patient outcomes.

The future of long COVID research lies in the continued refinement of omics technologies, integrating multiomics approaches to identify new metabolic and genetic pathways that underlie disease progression. By combining proteomics, metabolomics, lipidomics, and genomic analyses, we can gain a deeper understanding of the complex molecular mechanisms driving long COVID. This comprehensive approach will facilitate the development of novel diagnostic tools, biomarkers, and therapies, thereby advancing personalized medicine for patients with long COVID. As we continue to explore the intricate connections between viral infection, host response, and disease persistence, we move closer to identifying effective, targeted treatments that could offer relief to the millions affected by this debilitating condition.

Declarations

Conflict of interest

The authors (M.D.d.S., T.S.d.S., C.G.M., M.C.M.V., A.K.B.T., L.F.L., S.M.B., R.D., and M.A.M) declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Author contributions

Conceptualization, M.D.d.S., T.S.d.S., C.G.M., M.C.M.V., A.K.B.T., L.F.L., S.M.B., R.D., and M.A.M.; methodology, M.D.d.S., T.S.d.S., C.G.M., M.C.M.V., A.K.B.T., L.F.L., S.M.B., R.D., and M.A.M.; writing—original draft preparation, M.D.d.S., T.S.d.S., C.G.M., M.C.M.V., A.K.B.T., L.F.L., S.M.B., R.D. and M.A.M.; writing—review and editing, M.D.d.S., T.S.d.S., C.G.M., M.C.M.V., A.K.B.T., L.F.L., S.M.B., R.D., and M.A.M. All authors have read and agreed to the published version of the manuscript.

Funding

Open access funding provided by FCT|FCCN (b-on). This work was supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (Process No. 200177/2022-2).

Ethics approval

Not applicable.

Consent (participation and publication)

Not applicable.

Data availability

Data sharing is not applicable to this article, as no datasets were generated or analyzed during the current study.

Code availability

Not applicable.