What evidence links persistent spike protein to long COVID symptoms in peer-reviewed studies?
This fact-check may be outdated. Consider refreshing it to get the most current information.
Executive summary
Multiple peer‑reviewed studies and high‑profile preprints report detectable SARS‑CoV‑2 spike protein persisting in blood or tissues months to years after infection and sometimes after vaccination; several link this persistence to inflammation or neurological markers but others find no clear correlation with symptoms [1] [2] [3] [4]. Key recent papers report spike protein in skull marrow, meninges and brain border regions up to four years post‑infection with accompanying CSF markers, while cohort studies show spike detectable in a minority of convalescent or long‑COVID patients and mixed associations with clinical syndrome [1] [5] [3] [6].
1. Why spike persistence matters: the reservoir hypothesis
Researchers propose that pockets of persistent virus or viral components — “reservoirs” in gut, marrow or border tissues — continuously release spike protein that could drive ongoing immune activation and symptoms; several reviews and early studies put viral persistence–mediated inflammation at the center of long COVID pathophysiology [7] [8] [9]. This framework appears across laboratory, animal and human tissue studies cited by Helmholtz Munich and in systematic reviews that identify spike protein as a plausible effector of endothelial injury, cytokine release, autoantibody generation and thromboinflammation [7] [10] [1].
2. Tissue studies: spike in skull, meninges and brain borders
A prominent study published in Cell Host & Microbe and summarized by Helmholtz Munich reports spike protein persisting in the skull bone marrow and meninges — the brain’s border tissues — for up to four years post‑infection and associates that persistence with elevated CSF biomarkers (tau, NfL, GFAP) that indicate neuroaxonal injury and gliosis [1] [5]. The same group showed that vaccinated mice had lower tissue accumulation and that injected spike protein produced CNS inflammation in mouse models, supporting a mechanistic link between persistent spike and local inflammatory damage in animal experiments [1] [11].
3. Blood and serum detection: who has circulating spike, and how often?
Clinical cohort work finds circulating spike protein in a subset — not the majority — of recovered or long‑COVID patients. Preprint and peer‑reviewed cohorts report spike detectable in roughly single‑digit to low‑teens percentages of convalescent groups (examples: 2–14% in one exploratory study; other reports describe spikes in some long‑COVID cohorts) and some studies show persistence months to over a year post‑infection [3] [6] [2]. A peer‑reviewed J. Clin. Med. study concluded serum spike persistence after COVID is not associated with ME/CFS in their cohort, underscoring heterogeneity in findings [4].
4. Association with symptoms: evidence is mixed
Some investigators find correlations between persistent spike and specific biomarkers or symptoms — for example, spike in brain border tissues co‑occurring with CSF markers linked to neurological injury and reports tying circulating spike to subsets of long‑COVID patients — but other cohort studies find no statistically significant association between serum spike levels and symptom severity or ME/CFS diagnosis [1] [5] [3] [4]. Commentators emphasize that detectable spike may be a marker of an underlying reservoir rather than direct proof it causes symptoms [6].
5. Mechanistic plausibility: how spike could cause harm
Reviews and lab work document plausible mechanisms: spike interacts with endothelial cells, pericytes and immune receptors, can provoke pro‑inflammatory cytokines and complement imbalance, and in animal models causes CNS inflammation and worse outcomes after injury — all pathways that could plausibly contribute to multisystem long‑COVID manifestations [7] [8] [11] [10]. These mechanistic data strengthen causal plausibility but do not, by themselves, prove that persistent spike causes long COVID in humans [7] [10].
6. Vaccination, vaccinal spike and interpretation caveats
Some studies detect spike protein sequences from both infection and vaccine sources; distinguishing origin matters because vaccine‑derived spike sequences and viral spike differ and typically clear quickly, although isolated reports and preprints describe prolonged detection after vaccination in rare cases [7] [12]. Available sources do not mention definitive population‑level evidence that vaccinal spike persistence drives large‑scale long‑COVID burden; instead Yale and others discuss persistent spike as one of several possible drivers and propose targeted trials [12].
7. What the current literature still does not settle
Peer‑reviewed and preprint evidence establishes that spike protein can persist in tissues or plasma in at least some people and that spike is biologically active in models [1] [11] [5]. However, available sources show inconsistent clinical correlations: some cohorts find no link between serum spike and symptom severity or ME/CFS [3] [4], while tissue studies and mechanistic work argue for potential causation in neurological sequelae [1] [10]. Large, longitudinal, case‑control and interventional studies testing whether removing spike or clearing reservoirs reduces symptoms are not described in the current reporting (not found in current reporting).
8. Takeaway for clinicians and patients
The balance of peer‑reviewed reports and expert commentary says persistent spike protein is a credible biologic clue to one pathway of long COVID and justifies further targeted research and trials, but it is not yet a proven, universal cause of long COVID across patients; multiple studies find persistence only in subsets and clinical correlations remain mixed [1] [3] [4] [6]. Researchers recommend treating spike persistence as a testable hypothesis and pursuing biomarker‑guided trials and reservoir‑targeted therapies to move from correlation to causation [12] [6].