Have long COVID patients been shown to have spike protein or spike fragments correlated with symptoms?
Executive summary
Multiple studies report detection of SARS‑CoV‑2 spike protein or spike fragments months to years after infection and, in several cases, a statistical association with long COVID symptoms; notable findings include spike detected in plasma of ~60% of PASC patients in a Boston research cohort (preprint) and spike accumulation at the skull‑meninges‑brain axis lasting up to four years in human tissue and mice (Cell Host & Microbe/Helmholtz reporting) [1] [2] [3]. Other cohort and serum studies find no clear correlation between serum spike levels and symptom severity, so the evidence is mixed and may depend on tissue, detection method, timing, and patient selection [4] [5].
1. What researchers have actually measured: proteins in blood, brain borders, or neither
Teams have used different assays and tissues. Ultra‑sensitive single molecule array (Simoa) tests detected S1 subunit, full spike, or nucleocapsid in plasma of PASC patients — with spike detected in 60% of that cohort and antigens present in 65% overall — while separate proteomic and imaging work visualized spike protein persisting in the skull bone marrow and meninges of deceased COVID patients and in mouse models [1] [3] [2]. A mass‑spectrometry study of 81 long‑COVID sera identified spike in one post‑infection patient and detected vaccine‑derived spike in two patients after vaccination, showing lower detection frequency by that method [5].
2. Correlation with symptoms: supportive findings
Some teams report associations consistent with a pathogenic role. The Boston/MGH–Brigham group argued that plasma spike was a marker distinguishing PASC patients from fully recovered controls, suggesting an active viral reservoir could explain persistent symptoms [1]. The skull‑meninges study linked spike presence in the brain border tissues to elevated CSF biomarkers of neurodegeneration and to neuroinflammatory proteomic signatures; in mice, injection of spike alone produced neuroinflammation and behavior changes, tying spike persistence to plausible mechanisms for neurological long COVID [3] [6].
3. Contradictory and null findings
Not all investigations find a symptom link. An exploratory Berlin cohort using serum quantification reported persistent serum spike but found no correlation between serum spike concentration and post‑COVID symptom severity or laboratory biomarkers across 121 patients with post‑COVID fatigue syndromes [4]. The differences in results mirror methodological divergence: blood versus tissue sampling, Simoa versus mass spectrometry, cross‑sectional versus tissue clearance/imaging, and variable patient populations [1] [4] [5].
4. Mechanistic plausibility and experimental evidence
Laboratory and animal studies show spike protein can cause endothelial damage, disrupt the blood‑brain barrier, activate inflammatory pathways, and alter neuronal/vascular signaling — mechanisms compatible with long‑term symptoms [7] [8] [9]. The skull‑meninges paper further showed that spike accumulation correlated with dysregulated inflammatory pathways and proteomic overlap with Alzheimer’s disease signatures, strengthening biological plausibility for neurological sequelae [3] [6].
5. Limits, caveats and hidden variables in the literature
Key limitations shape interpretation: assays differ in sensitivity and specificity (Simoa can detect low‑level antigen; mass spectrometry may miss low abundance), tissues sampled vary (blood vs skull marrow vs meninges), many studies are observational or preprints, and sample sizes and selection bias differ across cohorts [1] [4] [5]. Vaccination can introduce vaccine‑derived spike antigen transiently and complicate attribution to infection versus vaccine exposure, a point raised by proteomic detection of vaccine spike in a few patients [5] [2].
6. Competing interpretations and implications for patients
One interpretation: persistent spike (or fragments) reflects a hidden viral reservoir driving inflammation and symptoms — supported by plasma antigen detection in a majority of one PASC cohort and by tissue persistence at brain borders [1] [3]. A competing interpretation: detectable spike is epiphenomenal, not causative — supported by a study that found no correlation between serum spike levels and symptom severity [4]. Both views acknowledge mechanistic studies showing spike can injure cells, but they disagree on whether measured spike in patients is the proximate cause of long COVID [7] [8] [9].
7. What investigators recommend next
Authors across studies call for standardized, assay‑validated, longitudinal sampling across tissues; larger cohorts with matched controls; and mechanistic work to test whether removing antigen or eradicating reservoirs alters symptoms. The skull‑meninges group also reported vaccination reduced but did not eliminate spike accumulation in mice, pointing to potential modifiers of persistence that need human study [3] [2].
Conclusion: Multiple lines of evidence show spike protein or fragments can persist in blood or at brain borders and associate with pathophysiologic signals; however, results are inconsistent across studies and dependent on methods and sample sets. Available sources do not provide a definitive, universally replicated causal link between spike persistence and symptom severity for all long COVID patients [1] [3] [4] [5].