Nearly 70 weeks after infection, long COVID patients show no detectable inflammation in blood tests

By Tarun Sai LomteReviewed by Susha Cheriyedath, M.Sc.Mar 4 2026

Nearly a year and a half after infection, researchers examined whether long COVID leaves measurable traces of inflammation or neuronal damage in the blood and their findings challenge assumptions about persistent immune activation.

Study: Long-COVID: assessment of circulating markers suggests no cerebral neuronal damage, neuroinflammation or systemic inflammation–a controlled study. Image Credit: p.ill.i / Shutterstock

A recent study published in the journal Scientific Reports analyzed circulating biomarkers of systemic inflammation and neuroinflammation in patients with long COVID (LC).

Background and Long COVID Prevalence

LC has emerged as a global health issue whose prevalence increases with repeated exposure to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The estimated prevalence of LC has increased worldwide from 60 million to 400 million between 2020 and 2024. Some studies with up to two years of observation reveal that lingering symptoms remain essentially unchanged over time, while others suggest that symptoms become less severe.

While acute coronavirus disease 2019 (COVID-19) is considered a multiorgan disorder, the mechanisms underlying the chronic phase are unclear. Some hypotheses suggest that LC might resemble other post-infectious syndromes in which symptoms linger without ongoing organ damage, while others emphasize reactivated viral reservoirs and persistent organ damage. Further, studies have provided evidence of target organ involvement, such as neuronal cell damage.

However, many studies were conducted early in LC, i.e., within a few months after infection, which may reflect ongoing organ damage or viral persistence during healing. Further, early studies described a range of symptoms, which, however, changed over time, with cognitive impairment and fatigue emerging as the main LC features. These observations align with those of post-viral illnesses, in which symptoms often become chronic, whereas inflammatory biomarkers typically normalize.

Study Design and Participant Selection

In the present study, researchers analyzed circulating biomarkers of neuroinflammation and systemic inflammation in LC patients. This case-control study was conducted in a Norwegian hospital between January 1, 2022, and April 1, 2024. Eligible participants, aged 16–80 years, had confirmed SARS-CoV-2 infection. LC cases met the National Institute for Health and Care Excellence (NICE) criteria for LC, i.e., persistent symptoms for > 12 weeks, unexplained by an alternative diagnosis.

Controls were individuals who completely recovered from SARS-CoV-2 infection and did not have persisting symptoms. Individuals with chronic inflammatory diseases, autoimmune diseases, anemia, hypothyroidism, cancer, and untreated comorbidities affecting fatigue, and those using systemic corticosteroids were excluded. Routine biochemical and hematological tests, including CRP, were performed at the hospital.

Biomarker Measurement Methods

Proinflammatory cytokines, i.e., interleukin-6 (IL-6) and tumor necrosis factor α (TNF-α), were measured using the MSD S-Plex electrochemiluminescence immunoassay platform. In addition, glial fibrillary acidic protein (GFAP, a neuroinflammation biomarker), neurofilament light (NfL, a biomarker of neuronal damage), triggering receptor expressed on myeloid cells 2 (TREM2), CRP, IL-6, IL-1β, and TNF-α were measured using the ultrasensitive nucleic acid-linked immuno-sandwich assay (NULISA).

Study Findings on Inflammatory and Neurological Biomarkers

The study recruited 112 individuals, of these, 96 were included in the final analytic sample. They were aged 46.7 years, on average. Most participants were female (85.4%), and the median time since COVID-19 diagnosis was 69 weeks. LC cases and recovered controls were well-matched for sex, age, and the time since COVID-19.

CRP measured in routine hospital analyses was not significantly different between cases and controls. TNF-α and IL-6 levels were marginally elevated in LC cases relative to recovered controls. NULISA revealed nominally (in unadjusted analyses) increased levels of inflammatory markers (CRP, TREM2, TNF-α, and IL-6) in LC cases than in controls. In contrast, GFAP and NfL did not significantly differ between groups.

After false discovery rate (FDR) correction, inflammatory biomarkers were no longer significantly different between LC cases and controls, suggesting that any differences were either absent or too subtle to be detected reliably in this cohort. Finally, Spearman correlation analyses revealed no correlation between inflammatory biomarkers and symptom severity, suggesting that these biomarker levels did not predict symptom burden in the cohort.

Conclusions and Study Limitations

Taken together, the study found no significant differences in neurological and inflammatory biomarkers between LC cases and recovered controls at 69 weeks after SARS-CoV-2 infection. These results do not support evidence of overt immune activation, inflammation, or neuronal injury in the studied cohort at this late stage of infection. This discrepancy with earlier studies may reflect differences in follow-up duration, as the longer follow-up in this study may have allowed sufficient time for the resolution of acute inflammation and viral clearance.

One reason prior studies have found ongoing immune activation and inflammation in LC is that their cohorts included people with preexisting chronic inflammatory or autoimmune conditions, which show many of the same inflammatory markers and clinical symptoms as those in LC. The current cohort was designed to minimize potential confounding from such conditions and, therefore, better isolate LC-related biological signals.

The study’s limitations include its small sample size, cross-sectional design that precludes causal inference, use of a select biomarker panel that may mean other inflammatory pathways are active, and reliance on blood-based biomarkers with no paired cerebrospinal fluid or neuroimaging data, as well as the use of assay outputs reported in normalized protein expression units rather than absolute concentrations, which may limit comparisons with external reference values.

Overall, the findings do not support persistent systemic inflammation, neuroinflammation, or neuronal injury detectable in blood at this stage of long COVID, though the authors note that extremely low-level immune activation, potentially below current biomarker detection thresholds, or other mechanisms could still contribute to symptoms and warrant further investigation in larger cohorts.