The clinical outcome of COVID-19, a viral disease caused by SARS-CoV-2, is highly variable. While most affected individuals return to their baseline health status within weeks after the acute respiratory infection, a substantial proportion of convalescents suffers from persisting post-infective symptoms (1–3). Even children and young adults with predominantly mild COVID-19 infection suffer from prolonged symptoms (4). Commonly reported ailments include fatigue, post-exertional malaise, headache, dyspnea, chest tightness and memory problems (4–6). In the literature, these long-lasting sequelae are referred to as Long COVID, Post-Acute Sequelae of COVID-19 (PASC) or Post-COVID-19 condition (PCC), an umbrella term recently defined by the World Health Organization (WHO) as “the condition that occurs in individuals with a history of probable or confirmed SARS-CoV-2 infection, usually 3 months from the onset of COVID-19, with symptoms that last for at least 2 months and cannot be explained by an alternative diagnosis” (7).

The prevalence of long-lasting COVID-19 related symptoms in young people is disputed, with reports indicating fatigue in a range from 0.52% to 66.5% at 3 months post infection in community managed cases (4, 8–10). In addition to the uncertainty of the true prevalence of persistent post-infective symptoms, several questions regarding the pathophysiology and the natural course of the disease remain unanswered. Similarities may be found with other post-infective fatigue syndromes (PIFS) whose clinical presentations overlap greatly with PCC (11). Sustained low-grade inflammation and immune cell activation have been reported in post-infective fatigue after Epstein-Barr virus infection (12, 13). Similarly, investigations into post-COVID-19 immunology in adults find an immune signature with high plasma cytokine levels and altered immune cell populations in convalescent individuals with persistent symptoms compared to immune profiles in recovered convalescents and unexposed controls (14, 15). However, findings between studies vary due to differences in sample size, lack of appropriate control groups, heterogeneous disease definitions and a multitude of laboratory techniques (16). Evidence in young people is scarce, although pro-inflammatory changes with elevated levels of IL-6 and IL-1β have also been shown in PCC affected children (17). Additionally, considering SARS-CoV-2 primarily infects the respiratory tract, respiratory outcomes are relevant. Persisting dyspnea is common in elderly, while prospective data on lung function and respiratory symptoms after COVID-19 in adolescents and young adults are lacking.

Our aim was to investigate immunological blood markers and pulmonary function tests in a large prospective cohort of adolescents and young adults with mild COVID-19 as well as unexposed controls. Thus, we used cross-sectional data from the six-month follow-up appointment and compared markers for the four following groups: affected COVID-19 positive (COVID+ PCC+), recovered COVID-19 positive (COVID+ PCC-), affected COVID-19 negative (COVID- PCC+) and healthy controls (COVID- PCC-). We elucidated whether immunological aberrations compatible with chronic inflammation could be found in PCC affected participants. Furthermore, we investigated associations between inflammatory aberrations and clinical symptom severity.

At baseline, 516 participants (405 SARS-CoV-2 positive, 111 SARS-CoV-2 negative) consented to participation. At six months follow-up, 448 participants (367 SARS-CoV-2 positive, 81 SARS-CoV-2 negative) met the eligibility criteria and were carried over for further analysis. The participants were evaluated at a median of 212 days (range 179 to 341) after acute COVID-19 symptom onset. An overview of the classification and participant demographics are provided in Table 1 .

This study of a large group of young, non-hospitalized COVID-19 convalescents shows that a biomarker signature remains at six months after mild COVID-19 infection. Elevated levels of eotaxin, MCP-1 and IP-10 were detected in COVID-19 convalescents compared to COVID-19 negative participants. However, these changes in immunological markers did not associate with persistent symptoms or disability experienced by a subset of study participants after mild COVID-19 infection. No alterations in immunological markers were exclusively detected in post-COVID-19 affected participants, regardless of the applied post-COVID-19 definition. This resonates well with the baseline report from our cohort where clinical symptoms were independent of inflammatory markers (18). Additionally, no differences were detected in spirometry measures between COVID-19 positive and negative participants, regardless of post-COVID-19 symptomatology. Thus, complaints of dyspnea in post-COVID-19 affected individuals could not be explained by deterioration of pulmonary function, nor by an inability to repair tissue damage. This is in line with previously reported findings from the sub-acute infection stage of our cohort (18). To the best of our knowledge, this is the largest prospective study assessing post-COVID-19 immune alterations and pulmonary function in adolescents and young adults.

Different hypotheses have been proposed for PCC pathogenesis, including antigen persistence (31), latent virus reactivation (32), chronic inflammation and auto-immunity (33). Our data suggest an immune signature specific to COVID-19 but unrelated to post-COVID-19 symptomatology. Multiplex plasma cytokine analysis revealed elevated levels of three chemokines (monocyte chemoattractant protein 1 (MCP-1), eotaxin, and interferon-γ induced protein (IP-10)) in COVID-19 positive participants. IP-10 and eotaxin were also elevated in the COVID-19 group during the sub-acute infection stage (18). Chemokines are a superfamily of proteins that mediate a range of pro-inflammatory effects on leukocytes, such as chemotaxis, degranulation, and integrin activation. Eotaxin, an eosinophil-specific chemoattractant belonging to the CC family of chemokines, is implicated in the accumulation and activation of eosinophils at sites of allergic inflammation (34). MCP-1 is structurally closely related to eotaxin, also belonging to the CC family, but principally attracts monocytes, activated T cells and basophils (35). MCP-1, together with IP-10, a CXC family member, has been brought forward as part of the “cytokine storm” in COVID-19, a type of hyperinflammatory state associated with worse outcome in acute SARS-CoV-2 infection (6). Upon viral infection, IP-10 is upregulated by locally produced IFN-γ and becomes available for the recruitment of monocytes, T lymphocytes and NK cells as part of the anti-viral response (35). Taken together, while chemokines are essential in the initial phases of antiviral response, sustained elevated levels can produce hyperinflammation and tissue damage.

One hypothesized explanation for the persistent elevation of these chemokines is viral persistence in COVID-19 convalescents, where the pathogen chronically stimulates the innate and adaptive immune system (31). In our cohort, SARS-CoV-2 anti-nucleocapsid titers declined over time and anti-SARS-CoV-2 specific antibodies were similar between unvaccinated convalescents with or without persistent post-infectious symptoms, decreasing the likelihood of antigen persistence as a trigger of PCC. This contrasts with findings by Klein et al. showing enhanced anti-Spike immune response among PCC+ participants (36). However, conventional antibody screening in peripheral blood might not be sufficient to detect viral persistence in “anatomical sanctuaries”. Previously, SARS-CoV-2 nucleocapsid protein and viral RNA have been detected in gut tissue after COVID-19 recovery (37).

Recent studies have found immunological markers and immune cell populations that remain altered following COVID-19 infection, some changes being specific to PCC. Bergamaschi et al. identified persistent abnormal levels of IL-6 and TNF after clinical recovery from COVID-19 infection (38). Phetsouphanh et al. identified six cytokines (IFN-β, IFN-λ1, IFN-γ, CXCl9, CXCL10, IL-8, sTIM-3) that were elevated in the COVID19+ PCC+ and PCC- versus controls at four months after COVID-19 infection, pointing to an immune signature specific to COVID-19 (14). Interestingly, at eight months IFN-β and IFN-λ levels were significantly elevated in PCC+ compared to all other groups, indicating that PCC related immune alterations might become more apparent with time. These findings align with a recent study assessing PCC immune profiles at one year after infection: marked alterations in immune cell population frequencies are accompanied by higher intracellular IL-4 and IL-6 production in peripheral blood mononuclear cells (PBMC) from PCC+ patients, together with elevated IL-8 and CCL4 plasma levels (36). While the exact immunological markers differ between studies, all indicate an ongoing inflammatory response and immune dysregulation in COVID-19 convalescents weeks-to-months after the initial SARS-CoV-2 infection. Variations in the immune markers on the cellular and transcriptional levels are attributed to persistent symptoms in PCC+ convalescent individuals (33, 36, 39).

In contrast, our data did not reveal immunological aberrations specific to PCC, nor did we find any relationship between immunological markers and symptom severity. Our study is based on a prospective cohort of non-hospitalized adolescents who have been rigorously assessed with operationalized PCC and PIFS criteria. Given the high prevalence of PCC-associated symptoms in the general population (40), our study might have captured PCC cases too liberally and therefore lacked power to detect differences in immune markers in true PCC cases. The broad PCC case definition by the WHO can make it difficult to distinguish between affected individuals due to the impact of the pandemic-associated stress or due to the impact of the infection itself. However, the analysis has been repeated with case assessment based on post-infective fatigue syndrome (PIFS) criteria and the classification was done blinded to COVID-19 status, increasing the validity of our findings. Phetsouphanh et al. have based their PCC classification on the presence of few hallmark symptoms (cough, fatigue, dyspnea) (14), while Klein et al. have recruited their PCC participants from Long COVID clinics (36), increasing the risk that observed differences between PCC, convalescents and controls are due to unrecognized factors (age, co-morbidities, initial disease severity) rather than actual differences in immunological markers. Self-selection, for instance, can lead to inclusion of PCC patients representing the extremes of the symptom spectrum related to post-COVID-19 symptomatology and produce biased findings. All findings, including our own, should be validated in a comparable cohort.

There are several limitations to our study. Our analysis is based on a small number of COVID-19 negative participants, many of which were recruited after SARS-CoV-2 testing for infectious symptoms. They might have contracted a different viral disease and therefore may not be true controls. However, it is assumed that common colds do not provoke post-infectious fatigue states (41) and participants with recent Epstein-Barr virus infection, a known precipitator of post-infective fatigue (42), were excluded from the analysis. In addition, there is a constantly evolving SARS-CoV-2 variant landscape, and B1.1.7. (alpha) was the dominant SARS-CoV-2 virus variant throughout the recruitment period for this study. The impact of the delta and omicron variants cannot be assessed here as the disease burden and pathophysiology might be variant-dependent. Finally, only seven out of 27 cytokines assessed by the multiplexed analysis yielded levels sufficient for conclusive analysis, and only three showed statistically significant differences between the groups. This is not surprising as peripheral blood cytokine levels are often undetectable with standard assays in healthy individuals (43). Reservations should be made regarding (over)interpretation of these results as the differences in cytokine levels were modest and fell within the physiological range for healthy individuals (44). Most participants received a COVID-19 vaccine during the follow-up period, potentially influencing the peripheral immune markers to a greater extent than the mild COVID-19 infection itself. Hence, the biological and clinical relevance of our findings is not clear, especially given the lack of association with post-COVID-19 symptomatology. Cytokine levels in plasma might not reliably reflect those in inflamed tissues and therefore, fail to show an association between tissue damage, inflammation, and Post-COVID-19 symptomatology. Other techniques, ultra-sensitive cytokine assays and single cell transcriptomics might still shed light on immune cell populations or biomarkers associated with PCC.

In the future, more prospective studies are needed to investigate the drivers of immune activation in both SARS-CoV-2 convalescence and Post-COVID-19 condition. The focus should lie on both high-resolution immune profiling and appropriate study design, including comparisons with normative data on post-COVID-19 symptomatology and adequate post-COVID-19 condition case ascertainment.

SARS-CoV-2 virus exerted a modest, prolonged effect on the expression of chemokines eotaxin, MCP-1 and IP-10 at 6 months after convalescence. These elevated immunological markers were, however, unrelated to post-COVID-19 symptomatology.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Regionale Komiteer for Medisinsk og Helsefaglig Forskningsetikk (REK). Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin.

The authors confirm contribution to the paper as follows: Study conception and design: VW. Data collection: SS, LB, JS, LH, TS-S. Analysis and interpretation of results: SS, LB, VW, LH, JS, FL-J, TM, AM, GE, TL, TT. Draft manuscript preparation: SS, LB, VW. All authors contributed to the article and approved the submitted version.