Introduction

The COVID-19 has emerged as a global health crisis affecting billions of lives and posing unprecedented challenges to healthcare systems worldwide. While the acute phase of the disease has been the primary focus, there are increasing concerns about the persistence of post-COVID-19 sequelae, often referred to as Long COVID (LC), which continue to impact individuals long after the initial infection.

According to the World Health Organization (WHO), LC generally occurs three months after the onset of COVID-19, characterized by symptoms that persist for at least two months and cannot be explained by another diagnosis. Symptoms such as fatigue, shortness of breath, and cognitive problems can vary in intensity over time, significantly affecting the quality of life and the ability to perform daily activities (1,2). Although the acute phase of the pandemic has passed, managing LC remains a significant challenge, imposing substantial physical, emotional, and financial burdens on patients, families, and healthcare systems3.

The development of LC is influenced by a series of interconnected factors, including viral mutations and host characteristics such as genetic polymorphisms, sociodemographic aspects, and underlying medical conditions. Among these, chronic inflammation, metabolic and endocrine dysregulation, immunological imbalance, and potential autoimmunity are particularly significant1,2. Preliminary investigations have indicated that a hyperinflammatory response, characterized by cytokine overproduction, including tumor necrosis factor (TNF)-ɑ and interleukin (IL)-6, plays a critical role in the severity of COVID-19 and the development of sequelae4,5. Moreover, processes such as platelet activation, cytokine and neutrophil interaction, expression of tissue factor by monocytes, and activation of the complement system are involved in immunothrombosis and complications that can lead to LC4,5,6.

Studies have also pointed to a correlation between prolonged neurological manifestations, such as fatigue and dyspnea, and increased complement activation, thromboinflammation, and elevated levels of biomarkers such as C-reactive protein (CRP), IL-6, Interleukin (IL)-1β, and TNF-ɑ7,11,12. Persistent microvascular endotheliopathy, possibly associated with cryptic tissue reservoirs of SARS-CoV-2, may further contribute to the pathophysiology of LC, with a particular impact on the vascular, gastrointestinal, cardiorespiratory, and nervous systems7.

This study aims to investigate the role of sociodemographic conditions, comorbidities, and hemostatic/inflammatory biomarkers in individuals recovering from COVID-19, with a focus on evaluating the persistent effects of the disease on the musculoskeletal, cardiorespiratory, and vascular systems. Significant findings have demonstrated links between comorbidities, the clinical spectrum of the disease, and specific biomarkers associated with prolonged COVID-19 symptoms. Additionally, the study examines the risk of post-COVID-19 sequelae and their impact on the quality of life among participants in a Brazilian cohort, offering valuable insights into the long-term repercussions of the disease.

Results

Impact of sociodemographic factors, comorbidities on disease Severity.

Considering only the sociodemographic aspects described in Table 1, a substantial relationship was found between disease severity (moderate or severe) and the following parameters: age group (P < 0.001), lower level of education (P < 0.001), ethnicity (P = 0.003), and BMI classification (P < 0.001). As for comorbidities described in this table, disease severity was positively associated with chronic cardiovascular disease (P = 0.002) and diabetes mellitus (DM) (P = 0.006).

Table 1 Sociodemographic, comorbidities and genetic analysis association data with disease severity.
Full size table

Genetic predisposition in post-COVID sequelae

The presence of post-COVID-19 sequelae also was significantly associated with severity of the disease (P < 0,001). Genetic predisposition was assessed in relation to blood type, main polymorphisms for hereditary thrombophilia and IL-6/TNFA. Only The MTHFR gene rs1801133 polymorphism was related to COVID-19 severity (P = 0.035) (Table 1).

Association between biomarkers and COVID-19 severity

This study analyzed 16 hemostatic, inflammatory, and immunological biomarkers potentially involved in continuous endothelial activation, as shown in Table 2. These samples were collected at least 30 days after the resolution of acute symptoms or hospital discharge. Increased plasma levels of FVIII (P = 0.046), VWF (P = 0.035), DD (P = 0.002), and HbA1c (P < 0.001) were significantly associated with participants who developed severe forms of the infection, even during the convalescent phase (Fig. 1). Other biomarkers, such as thrombin time, free protein S, and prothrombin time, also showed substantial associations; however, they varied within the local laboratory reference range established by Brazilian laboratories, and in accordance with international consensus. Other common inflammatory and immunological plasma biomarkers in the acute phase, such as IL-6, TNF-ɑ, antiphospholipid antibodies, and lupus anticoagulant, did not show a significant difference between the groups (mild, moderate and severe) during the convalescent phase of the disease.

Table 2 Relationship between quantitative hemostatic/inflammatory parameters and the clinical spectrum of covid-19.
Full size table
Fig. 1
figure 1

Biomarkers associated with the severity of COVID-19 during the convalescent phase. Plasma levels of biomarkers associated with participants with different clinical spectrums (n = 60 mild group, n = 65 moderate group, and n = 101 severe group). (a) Factor VIII, reference value 50–150%; (b) Von Willebrand factor, reference value 50–150%; (c) dimer D, reference value < 500 µg/L FEU; (d) Glycated hemoglobin, reference values: normoglycemia (< 5.7%), pre-diabetes or increased risk for DM (≥ 5.7% and < 6.5%), and DM (≥ 6.5%). Groups were compared using the Kruskal–Wallis test. If P < 0.05, paired comparisons were conducted with the Dunn-Bonferroni multiple comparison post-hoc test. *P < 0.05, **P < 0.01, ***P < 0.001.

Full size image

Altered hemostatic markers follow-up

Six months after the initial blood draw, a follow-up was conducted for participants in the moderate and severe groups with altered hemostatic markers. Approximately 45% of the participants complied with the recall. Significant variations over time were observed in three biomarkers: FVIII (N = 27, P < 0.001), VWF (N = 28, P = 0.002), and DD (N = 17, P < 0.001). While there was a significant decrease in biomarker levels over time, FVIII and VWF levels remained elevated in participants (Fig. 2). Only DD returned to normal levels six months after the first collection.

Fig. 2
figure 2

Biomarker’s follow-up associated with COVID-19 severity (a) FVIII, reference value 50–150%, (b) VWF, reference value 50–150%, (c) DD, reference value < 500 µg/L FEU. Showing altered plasma levels of key biomarkers after a six month follow-up. (N = 27 FVIII, N = 28 VWF and N = 17 DD). Wilcoxon test; Significant if P < 0.05. *P < 0.05, **P < 0.01, ***P < 0.001.

Full size image

Among the patients with these altered biomarkers, 26 individuals with persistently altered FVIII, 23.1% (N = 6) reported joint pain, 46.2% (N = 13) experienced fatigue, 34.4% (N = 9) had myalgia, 30.8% (N = 8) suffered from dyspnea, 23.1% (N = 6) had a dry cough, 15.4% (N = 4) experienced chest pain, 7.7% (N = 2) reported palpitations, 19.2% (N = 5) had phlebitis, 26.9% (N = 7) experienced swelling, and 42.3% (N = 11) reported cramps.

For those with altered VWF-Ag after six months, 28.6% (N = 8) reported joint pain, 42.9% (N = 12) experienced fatigue, 32.1% (N = 9) had myalgia, 32.1% (N = 9) suffered from dyspnea, 17.9% (N = 5) had a dry cough, 17.9% (N = 5) experienced chest pain, 10.7% (N = 3) reported palpitations, 17.9% (N = 5) had phlebitis, 32.1% (N = 9) experienced swelling, and 39.3% (N = 11) reported cramps.

Regarding DD, 17 participants were reassessed, with 47.1% (N = 8) reporting joint pain, 58.8% (N = 10) experiencing fatigue, 41.2% (N = 7) having myalgia, 35.3% (N = 6) suffering from dyspnea, 17.6% (N = 3) having a dry cough, 11.8% (N = 2) experiencing chest pain, 35.3% (N = 6) reporting palpitations, 11.8% (N = 2) having phlebitis, 41.2% (N = 7) experiencing swelling, and 29.4% (N = 5) reporting cramps.

Association of sociodemographic aspects, comorbidities, and biomarkers regarding long COVID development

The sociodemographic profile, comorbidities, and biomarkers of 206 participants who self-reported experiencing prolonged symptoms after COVID-19 infection was evaluated. The association was analyzed using the Chi2 test and the risk predicted of robust logistic regression test.

The Chi2 analysis showed that in the musculoskeletal system, the joint pain sequelae (N = 42) was significant to clinical spectrum (P < 0.001), schooling (P = 0.01), ethnic group (P = 0,04), chronic lung disease (P = 0.018), DM (P < 0.001), obesity (0.046) and IL-6 (pg/mL) (P = 0.018). The fatigue (N = 96) was significant only in DM (P = 0.044). The myalgia (N = 32) was significant to clinical spectrum (P = 0.001), schooling (P = 0.023), practice physical activities (P = 0.038), chronic lung disease (P = 0.001), DM (P = 0.002) and obesity (P = 0.014). In the cardiorespiratory system, dyspnea (N = 42) was significantly associated with clinical spectrum (P < 0.001), schooling (P < 0.001) and DM (P = 0.001). Dry cough (N = 22) only associated with schooling (P = 0,003). Chest pain (N = 22) significant to clinical spectrum (P = 0.040), schooling (P = 0.047) and DM (P = 0.005). The palpitations were associated with the clinical spectrum (P = 0.026), ethnic group (P < 0.001), chronic lung disease (P = 0.001) and DD (P < 0.001). The analysis of vascular system outcomes in LC patients revealed that severe cases of COVID-19 were significantly associated with the development of cramps (N = 38, P = 0.024), with a higher incidence observed in those with elevated FVIII (P < 0.001) and VWF levels (P = 0.009). Additionally, individuals with high DD levels were more likely to experience cramps (P = 0.039), while chronic lung disease showed a significant association with swelling (N = 39, P = 0.040). Male sex was also significantly associated with phlebitis (N = 19, P = 0.016). The remaining biomarkers and sociodemographic factors did not show significant associations with the vascular outcomes. Statistical analyses of other factors not significantly associated with the post-COVID-19 sequelae studied are available in Supplementary Tables (ST) S1, S2, and S3.

Subsequently, robust logistic regression was employed to further analyze the odds (odds ratio—OR) of associations between these factors and LC (musculoskeletal, cardiorespiratory, and vascular systems) in the cohort that responded to the second questionnaire on prolonged symptoms (Table 3). Considering the musculoskeletal system, participants with a history of lung disease and DM had higher chances of developing joint pain sequelae (OR: 3.08, CI 1.10–8.43 and OR: 3.78, CI 1.42–9.99 respectively) and myalgia (OR: 4.33, CI 1.48–12.48 and OR: 3.42, CI 1.17–9.79 respectively). IL-6 was also a determinant of joint pain (OR: 2.33, CI 1.06–5.06), as described in Table 3A.

Table 3 Sequelae risk.
Full size table

Chronic lung disease increased the risk of dyspnea and palpitations by more than three times (OR: 3.37, CI 1.17–9.64/OR: 3.95 CI 1.20–12.38). The group of participants who had severe COVID-19 also had more than 5 times the chance of developing dyspnea as a sequela (OR: 5.24, CI 2.20–13.64). Non-white participants (Brown/Black) had a significantly higher risk of prolonged palpitations, with an OR of 4.073 (CI 1.44–12.83), as did those with elevated DD levels during the convalescent phase (OR: 4.22, CI 1.44–12.83). The level of knowledge related to COVID-19 symptoms, sequelae, and treatment were evaluated in relation to participants' education level, showing that participants with only high school and elementary education complained more about persistent dry cough after infection (OR: 3.385, CI 1.354–8.968) as shown in Table 3B.

Female individuals were more likely to develop phlebitis or painful varicose veins in the long term (OR: 3.87, CI 1.34–13.97), while regular physical activity before infection was protective against the sequela of lower limb swelling (OR: 0.34, CI 0.15–0.75). Lastly, elevated plasma levels of FVIII during the convalescent phase of the disease increased the risk of developing cramps by more than four times (OR: 4.42, CI 1.65–12.37), as shown in Table 3C.

Impact on quality of life

Quality of life of 206 participants who responded the LC questionnaire was assessed according to COVID-19 clinical spectrum1 using the EURO-QOL-5 Dimension questionnaire8, which serves as a tool to evaluate well-being and health of different populations. A five-point scale was used accordingly: no problems (> 80%); few problems (60–80%); moderate problems (40–60%); severe problems (20–40%); extreme problems < 20% in maintaining abilities/functions. The perception of "how is your health today" after a 2-year period was measured according to the stratification: excellent (> 80%); very good (60–80%); good (40–60%); poor (20–40%); very poor < 20%. In all evaluated questions, greater impairment was observed in participants classified as moderate and severe. The rate of less than 40% maintained abilities or functions were found in the group with the broadest disease clinical spectrum: personal care (2.2%), mobility (8.79%), pain and discomfort (16.49%), usual activities (5.50%), and anxiety and depression (20.88%). Sensation of pain and discomfort were described by 16.67% of people in the mild group, Fig. 3. In the severe COVID-19 group (letter f, Fig. 3), there was a lower perception of current health, where almost 16.5% of respondents rated their post-COVID-19 life condition as poor or very poor. In the moderate and mild groups, this perception was around 7%.

Fig. 3
figure 3

Quality of life impact. Clinical spectrum total percentage, in columns. 0–10%: very low (might be represented with a light color, such as pale yellow). 10.01–25%: low (could be a slightly more intense color, like light orange). 25.01–50%: moderate (a more saturated color, such as deep orange). 50.01–75%: high (an even more intense color, such as red–orange). 75.01–100%: Very High (the most saturated color, such as red). (a) anxiety or depression; (b) usual activities; (c) pain/discomfort; (d) personal cares; (e) mobility and (f) health condition ratins by severity (0–100%). The frequency test was done on SPSS and data analyst heatmaps/radar plot graph create.

Full size image

Discussion

The difficulties in understanding LC disease origin and progression are not only due to the lack of knowledge about inflammatory processes, but also the scarcity of research in underdeveloped countries9. Thus, discussing the environmental and pathophysiological components that impact on COVID-19 outcome and LC in a Brazilian cohort become crucial for better epidemiological assessments and treatment of the world population in the post-pandemic period9,10.

Older age accounted for the largest number of individuals infected with SARS-Cov-2, which is in line with the work of Zsichla and Müller 10. Older age is a determining factor in the severity of the disease, given that aging is linked to worsening of immune activity and hemostatic dysregulation10.

The worsening of the clinical spectrum of the disease can be attributed to sociodemographic differences, impacting the outcome of COVID-19 and long-term clinical findings, possibly due to health care neglect, especially in low-income populations11,12,13. Factors such as comorbidities and restrictions on health care have been associated with worsening SARS-CoV-2 infection, as highlighted by Leeuw and Yashadhara 14. Our findings reinforce the connection between COVID-19 severity and sequelae, associated with elements such as access to education and lack of information for vulnerable populations14. Although the Brazilian population is mixed, this cohort corresponds to a majority of white individuals, who showed an association with severity. In contrast, the cross-sectional study by Silva et al.15 reported around 78% of deaths among black and indigenous people hospitalized for COVID-19.

As for host genetics, the initial objective was to examine genes linked to thrombophilia and inflammation in individuals who had severe forms of the disease. We have found that only the 677C>T variant (rs1801133) of the Methylene Tetrahydrofolate Reductase (MTHFR) gene showed a significant relationship (P = 0.035) with severity. The MTHFR enzyme is involved in folate metabolism and participates in the conversion of homocysteine (Hcy) into methionine. Folate is essential for the metabolism of DNA and RNA16 and expression variations can reduce its activity by around 50%17. This genetic variation is considered important for cardiovascular and neurological complications18, which are frequent in COVID-19 patients19. In addition, hyperhomocysteinemia (HHE) is a predictor of endothelial dysfunction20, which in this case is due to the unbalanced release of reactive oxygen species, triggered by pro-inflammatory cytokines21. In the current cohort, comorbidities such as obesity, cardiovascular disease and diabetes were associated with COVID-19 severity, corroborating Thakur et al.22.

Although the importance of inflammatory and coagulation markers in the outcome of COVID-19 is well-established23, to date, there are scares studies that investigate these markers after the disease acute phase. We have found a significant difference between patients with mild and severe COVID-19 regarding levels of FVIII, VWF, DD and HbA1c (Fig. 1), suggesting that patients who progressed to the more severe stages tended to have higher levels of these biomarkers even after the acute phase when compared to the mild group.

The typical profile of disseminated intravascular coagulation (DIC) observed in critically ill COVID-19 patients acute phase shows increased fibrinogen, VWF and DD levels, whereas prothrombin time (PT), activated partial thromboplastin time (aPTT), and platelet count, remain normal24. These levels were also altered in our study, except for fibrinogen, which is a marker of the acute phase, although it varied within normal limits in our findings.

In addition, SARS-CoV-2 infection can promote endothelial chronic oxidative stress, culminating in the release of VWF multimers and, consequently, in a state of hypercoagulability25. The increase in FVIII can be explained by the VWF/FVIII interaction, since VWF is responsible for keeping FVIII stable in the bloodstream. In other cases, thrombocytopenia, characteristic of the initial phase of infection, can cause a predisposition to pro-thrombosis, evidenced by increased plasma levels of thrombin and DD4.

Weibel-Palade bodies are organelles found in endothelial cells, responsible for storing and releasing important molecules involved in hemostasis and inflammatory responses, such as VWF and P-selectin4. During SARS-CoV-2 infection, the chronic oxidative stress induced in the endothelium can trigger the release of VWF from these Weibel-Palade bodies4. This release contributes to the hypercoagulable state observed in patients, exacerbating thrombus formation and increasing the risk of vascular complications25,26. The observed increase in FVIII levels can be attributed to the interaction between VWF and FVIII, as VWF is responsible for stabilizing FVIII in the bloodstream4. Additionally, thrombocytopenia, often seen in the early stages of infection, can increase the risk of thrombosis, as indicated by elevated plasma levels of thrombin and DD4,27. This pro-thrombotic environment is driven by several mechanisms initiated by the virus and the body's immune response, including the NF-κB/NLRP3 inflammasome pathway, vasoactive peptides, cytokine storm, NETosis, and complement system activation26,27. These processes lead to changes in coagulation mediators such as factor VIII, fibrin, tissue factor, the VWF:ADAMTS-13 ratio, and the kallikrein-kinin or plasminogen-plasmin systems. Moreover, an imbalance between pro-thrombotic and thrombolytic factors, such as tPA, PAI-I, and fibrinogen, can leads to a state of hypercoagulation and hypofibrinolysis. In the presence of comorbidities like diabetes mellitus, these mechanisms may further disrupt hemostasis26.

In the cohort analyzed, severe disease IL-6 and TNF-ɑ serum levels did not differ significantly when compared to the mild disease group. Cytokines and chemokines are considered fundamental in the regulation of immunity and pathogenesis during viral infections28. In individuals affected by COVID-19 who have severe symptoms and an unfavorable prognosis, there is a rapid rise in IL-6 and TNF-ɑ levels. On the other hand, in patients with milder symptoms, there is a decrease in these cytokines to lower levels28. In addition, it is common to observe variations in plasma levels of IL-6 and TNF-ɑ throughout the infection, which generally decrease during the convalescence phase28, a characteristic which may explain our results.

With regard to HbA1c, according to Wang et al.28, it is common to observe a correlation between COVID-19 patients who exhibit elevated HbA1c levels and more severe disease, as demonstrated in this study. The infection itself may contribute to increased levels of this metabolic parameter. The authors also suggest that COVID-19 may cause an imbalance in glucose metabolism, caused by high levels of cortisol released during the infection, contributing to glucose-related metabolic disorders observed during infection28.

Participants with varying levels of severity in hemostatic and inflammatory biomarkers exhibited abnormal and persistently elevated levels of VWF and FVIII, indicating ongoing microvascular damage30. These alterations are likely contributors to the onset of LC sequelae by promoting prothrombotic changes that facilitate pulmonary (micro)thrombosis and sustained endothelial activation. The vasculature, being one of the main systems affected by SARS-CoV-2, plays a crucial role in the development of these sequelae31. Viral tropism leads to endotheliitis, which disrupts endothelial homeostasis, triggers cytokine release, and creates a pro-coagulant microenvironment. This continuous endothelial activation, as highlighted by Fogarty et al.32, may be marked by the ongoing exocytosis of Weibel-Palade bodies (WPB), favoring a sustained imbalance in the VWF:ADAMTS-13 axis. This study's findings correlate with these observations, where FVIII and VWF levels remain altered even after six months, possibly due to the chronic release of viral spike proteins and abnormal angiogenesis, particularly in individuals with comorbidities and older age30,32.

Furthermore, impaired fibrinolysis—resulting from the ongoing imbalance between coagulation and fibrin breakdown—can exacerbate this prothrombotic state. The reduced ability to effectively break down clots, combined with tissue hypoxia and compromised oxygen exchange during the acute phase, may persist for months, further contributing to the perpetuation of LC symptoms such as muscle fatigue, dyspnea, cognitive impairment, insomnia33.

Sequelae involving the musculoskeletal system are frequently observed in people with DM34. In the present study, DM emerged as a significant risk factor for the development of joint pain, myalgia and fatigue. Mittal et al.35 demonstrated that diabetic individuals who had COVID-19 showed a higher incidence of post-COVID-19 fatigue compared to the disease-free diabetic group. However, the risk was not evident. Besides the presence of DM, the coexistence of other comorbidities such as obesity may predispose to the onset of persistent symptoms. Moreover, SARS-CoV-2 and DM share endothelial and immune dysfunction36,37. Furthermore, elevation of reactive oxygen species levels in hyperglycemic contexts triggers inflammatory processes, which results in cellular damage mediated by a wide range of cytokines and growth factors34. This inflammatory imbalance observed in diabetic patients may be identified as one of the possible underlying causes of musculoskeletal sequelae development after COVID-19. Altered levels of IL-6 have been identified as a risk factor for the onset of joint pain. IL-6 is a critical mediator of joint pain and acts directly on the nociceptive system to induce hyperalgesia and sensitization. Studies indicate that elevated expression of IL-6 and its receptors in the dorsal root ganglia is associated with pathological pain models. Administration of IL-6 in experimental joints promoted increased responses to mechanical stimuli, corroborating its association with nociceptive sensitization38.

Considered a risk factor for COVID-19 severity, individuals with pulmonary diseases naturally experience greater lung function impairment. Cecchetto et al.39 found a higher correlation between persistence of post-COVID-19 dyspnea and lung capacity reduction. Additionally, this sequelae was also associated with decreased lung diffusion capacity and respiratory muscle strength40. Furthermore, it is expected that COVID-19 infected patients who progress to severe disease present dyspnea and cough, symptoms reflecting damage to the respiratory and cardiovascular systems2. Additionally, the study by Carfi et al.47 in Italy revealed that, on average, 60 days after symptom onset, 42% of hospitalized patients experienced dyspnea. Moreover, the need for mechanical support may increase chances of muscle atrophy, lung damage, and persistent sequelae41. However, underlying diseases must be ruled out before these sequelae can be considered as originating from the infection. Additionally, the present study demonstrated that sociodemographic factors may influence the emergence of cardiorespiratory sequelae (palpitations and chest pain) in non-white individuals, groups with greater vulnerability and less access to healthcare42,43. Considering the findings of this study, the persistence of post-COVID-19 palpitations seems to have a direct correlation with persistent DD values, suggesting a continuous thromboinflammatory process even after discharge from the acute phase, according to Kalaivani and Dinakar44. Here we describe an increased risk of developing phlebitis or symptoms of painful varicose veins in women in post-COVID-19. Exposure to estrogen and progesterone, either individually or combined, increases susceptibility to varicose vein development in individuals at high risk of vascular disorders, including those with a family history, obesity, and sedentary lifestyle. These hormones can also exacerbate superficial venous complications in these patients45.

Physical activity has been shown to be protective against swelling in the present cohort, especially in individuals who were exercising before the pandemic. This finding can be explained by the role of physical activity in improving mitochondrial function and oxygen uptake, maintaining energy production during cellular respiration, and avoiding tissue hypoxia and consequent microvascular endotheliopathy33,46.

FVIII has been linked in previous studies both to greater disease severity and as an independent predictor of COVID-19 associated mortality4,27,30. However, sustained elevation of FVIII in LC is associated with thromboinflammatory manifestations and vascular dysfunction32,33. This study found an increased risk of developing cramps, a vascular-related sequela, in participants who maintained elevated levels of this marker during the convalescent phase.

Regarding quality of life, the presented results indicated that participants who developed moderate and severe forms of COVID-19 have a worse quality of life compared to mild cases. The persistence of pain and discomfort is evident in a considerable proportion of patients, regardless of the initial severity of the disease, as well as anxiety and depression. Additionally, those facing the most severe forms of the infection tend to evaluate their current health more negatively. These findings underscore the need for targeted interventions to improve the well-being of patients after the initial recovery from COVID-19, especially for the most severe cases, in alignment with other research investigating quality of life in the pandemic and after it47,48,49,50,51.

This study considers the interaction between genetic factors, biomarkers, comorbidities and sociodemographic determinants essential for understanding LC. We identified significant correlations between COVID-19 severity, comorbidities such as cardiovascular diseases, diabetes, and chronic lung conditions, and the development of persistent symptoms, which compromise quality of life. Sociodemographic factors, including lower education levels and belonging to non-white ethnic groups, also increase vulnerability to post-COVID-19 sequelae, emphasizing the need for targeted interventions. The persistence of elevated hemostatic and inflammatory markers, like FVIII and VWF, suggests ongoing microvascular damage, contributing to LC. Physical activity showed a protective effect against swelling, underlining the role of lifestyle in recovery. This study highlights the complexity of LC and the need for an integrated approach that considers both modifiable and immutable factors to develop more effective therapies, accelerate recovery, and reduce the global impact of COVID-19.

Methods

Patients and control group

Volunteers who had COVID-19 in its asymptomatic or mild form (n = 60), moderate (n = 65) and severe (n = 101), recruited from reference hospitals in the metropolitan region of Greater Vitória—ES, Brazil, between November 2020 and December 2021, following WHO guidelines1. The selection criteria to participate in the research were that volunteers had a positive RT-PCR test for COVID-19, aged between 18 and 65 years, without acute symptoms for at least 30 days after infection and not vaccinated. (considering the 2 minimum doses)8. As a standard procedure, all research volunteers signed in person an Informed Consent Form (ICF) authorizing molecular and laboratory tests. After reading and signing the ICF, participants filled out their clinical, epidemiological and laboratory data using an online questionnaire prepared by researchers available at https://redcap.saude.es.gov.br/. After the first visit to the Research Center, participants in the Moderate and Severe groups who had altered laboratory test results (> 1 standard deviation/SD of the normal range for the test) were recalled again after the 6-month period to repeat the altered test. Reports of all trials were sent by email to each participant. Between March and July 2023, participants were contacted again by telephone to answer a questionnaire of information about post-COVID-19 consequences. Of the total number of initial participants, 206 (91.15%) people responded to this questionnaire dedicated to LC. The race/skin color was self-declared based on the categories used by the Brazilian Institute of Geography and Statistics (IBGE). Participants who self-identified as brown, black, yellow, and indigenous were grouped into the non-white category. Race/skin color was understood as a sociocultural construct52.

The second questionnaire had LC symptoms related questions and participants' quality of life in general. This questionnaire was based on WHO validated tools (2021 and 2023)1,53, and COVID-19 treatment guidelines from the American Institute of Health54 addressing symptoms that affect vascular, cardiorespiratory, musculoskeletal, reproductive, neurological and gastrointestinal systems2. The methods summary is shown in Fig. 4.

Fig. 4
figure 4

Study design for Long COVID questionnaire. Methods: 1. Recruitment of outpatient volunteers and after hospital discharge for a period of 1 year. 2. Stages of venous blood collection, sample processing, storage at – 80 °C, questionnaire application, and informed consent form signing. 3. DNA extraction stage, laboratory analyses, and molecular assays. 4. Recall of participants with altered results after 6 months from the first collection. 45% return rate. 5. Remote application of the questionnaire dedicated to LC. Long COVID Questionnaire: 6. The musculoskeletal system (joint pain, myalgia, and fatigue) was evaluated. 7. Cardiovascular system (chest pain, palpitations, dyspnea, and dry cough) 8. Vascular system (phlebitis, swelling, and cramps). 9. Evaluation of the EURO-QOL-5 Dimension questionnaire.

Full size image

“Joint pain”, “myalgia (muscle pain)” and “prolonged fatigue (weakness/tiredness)” within the musculoskeletal system sequelae were considered in this study. “Dyspnea (shortness of breath)”, “dry cough”, “chest pain” and “palpitations (sensation of increased heart rate)” were studied within the cardiorespiratory system and “phlebitis (painful varicose veins/vasculitis)”, “swelling of the lower limbs” and “cramps” within the vascular system. Prolonged symptom intensity and duration of SARS-CoV-2 infection were assessed2.

Laboratory studies

The laboratory tests were carried out in the Hemostasis, Hematology and Immunohematology laboratories of the Blood Bank of Espírito Santo (HEMOES) and in the Human and Molecular Genetics Center (NGHM), of the Department of Biological Sciences of the Federal University of Espírito Santo (UFES).

Processing and storage of biological samples

Samples were processed in a refrigerated centrifuge, plasma from citrated samples and serum were aliquoted into 1.5 ml microtubes and frozen at -80ºC for dosing at appropriate times. Whole blood samples were analyzed on the same day for platelet counts.

Complementary laboratory tests

Using the ACL TOP 550Ⓡ coagulometer from Instrumental LaboratoryⓇ (WerfenⓇ), tests for prothrombin time (PT), activated partial thromboplastin time (aPTT), thrombin time (TT), and FVIII levels of coagulation were performed and lupus anticoagulant (LA) using nephelometry. Using immunoturbidimetry, FIB, DD, VWF and resistance to activated protein C (APCR) assays were performed. Free protein S (PS) was measured using a colorimetric method.

Platelet counting was performed using impedance measurement using the Sysmex XN-1000Ⓡ equipment. Glycated hemoglobin (HbA1c) and hemoglobin electrophoresis (EHB) were analyzed by high-performance liquid chromatography (HPLC) dosed on BIORAD HPLC equipment D100Ⓡ and D1000Ⓡ respectively. Blood typing (BT) was performed by automated agglutination on the GriffolsⓇ Erytra Eflexis equipment. Anticardiolipin (aCL) IgG/IgM and antiβ2 glycoprotein I (aβ2GPI) IgG/IgM antibodies (OrgentecⓇ), IL-6 and TNF-ɑ (SIGMAⓇ) were measured by enzyme-linked immunosorbent assay (ELISA), using the RT 6000Ⓡ microplate reader and RT 3000Ⓡ microplate washer.

Genotyping

Five genetic polymorphisms were studied: F5 1691G > A (rs6025); F2 20210G > A (rs1799963); MTHFR 677C > T (rs1801133); IL-6 174G > C (rs1800795) and TNFA 308G > A (rs1800629). Genotypes were mostly determined by Real-Time PCR with TaqmanⓇ probes (fluorescent hybridization probes) on 7500 Fast Real-Time PCRⓇ devices. Only the MTHFR SNP rs1801133 was performed by polymerase chain reaction (PCR), followed by restriction fragment length polymorphism (RFLP) technique.

Statistical analysis

Initially, the Shapiro–Wilk tests were carried out to analyze normal distribution of data. Analyzes of categorical variables were performed using proportions and, in some cases, differences between proportions were assessed using the Chi2 test (dichotomous variable and percentage above five) or Fisher's exact test (when the value was less than five in some hut). Non-parametric tests were used on the data, such as the Dunn test with Bonferroni correction, when there were groups with different sizes, the Wilcoxon test, for analysis of paired samples, and the Kruskal–Wallis test for comparing three independent samples or more. Statistical tests were two-tailed, and the level of significance used was 0.05. For the correlation with LC sequelae cutting-edge tests related to multivariate classification were used. Crude and adjusted Odds Ratio and their respective 95% confidence intervals were estimated using logistic regression, to minimize the precepts that would be broken (outliers) if another logistic regression test were used55. Graphs were generated using GraphPad Prism version 8.0.1 and Python version 3.9.13. All analyzes were performed using the statistical software IBM SPSS (Statistical Package for the Social Sciences (SPSS), version 20.0 and RStudio version 4.3.1.

Approval of standard protocol, patient records and consents

This research was approved by the Human Research Ethics Committee of the Health Sciences Center/UFES, under number CAAE: 37094020.6.0000.5060 and is a retrospective cohort study to collect data from medical records and apply a questionnaire and of an experimental study for genotyping genes of interest and measuring inflammatory biomarkers in patients who developed COVID-19. All volunteers signed in person an Informed Consent Form (ICF) authorizing molecular and laboratory tests. All experiments were performed in accordance with relevant guidelines and regulations.

Use of artificial intelligence

LLM language models were used only to create Fig. 3 in the topic on quality of life. The data used comes exclusively from the research protocol.