Spike Protein, mRNA Clearance, and a New Autism Study: A Careful Reading

ThomasLee Abshier, ND | 08 October 2026

A commentary on recent articles from Dr. Peter McCullough’s Focal Points newsletter

Two items appeared this week in Dr. Peter McCullough’s newsletter. The first was an interview arguing that persistent spike protein is the hidden cause of a wide range of chronic symptoms, and offering a protocol to remove it. The second announced a new study linking childhood COVID-19 vaccination to autism, ADHD, and related conditions.

At NatureDox we take vaccine injuries seriously. We also hold every claim, from public health agencies and from their critics alike, to the same standard of evidence. Read that way, these articles contain real observations, some overstatements, and a few claims that need correcting. They also contain, once the mechanisms are understood, some reason for hope.

What is real

Post-vaccination injury exists. Myocarditis after mRNA vaccination, especially in young men after the second dose, is acknowledged by every major health agency. A chronic post-vaccination syndrome is now being studied at mainstream institutions.

There is also real evidence that spike protein persists in some people. In Yale’s study of post-vaccination syndrome, spike protein was normally detectable for only a few days after vaccination, but some participants with the syndrome still had detectable spike more than 700 days after their last dose. A Swedish case-control study found spike protein in the endothelial cells of vasculitic skin lesions weeks to months after infection or mRNA vaccination. Patients with these complaints deserve to be heard, not dismissed.

Does the vaccine mRNA break down?

The interview claims that no human enzyme has been shown to break down the vaccine mRNA. This point deserves careful treatment, because McCullough is partly right and the full picture is more hopeful than his framing suggests.

Why the mRNA was modified

Every uridine in the Pfizer and Moderna vaccines was replaced with N1-methylpseudouridine. This substitution comes from the work of Katalin Karikó and Drew Weissman, beginning in 2005, which later earned the Nobel Prize.

Its main purpose was to keep the body from recognizing the injected RNA as a viral invader. Ordinary, unmodified mRNA introduced from outside the cell triggers several innate immune sensors:

TLR7 and TLR8, receptors that detect foreign single-stranded RNA
RIG-I, a cytoplasmic sensor of viral RNA
PKR, a kinase that, when activated, shuts down protein production in the cell
The OAS / RNase L pathway, which, when activated, produces an enzyme (RNase L) that cuts RNA apart

When these alarms go off, the cell stops making protein and actively destroys the foreign RNA. The modified uridine largely hides the RNA from these sensors. The result is that the vaccine mRNA survives longer and produces far more spike protein than unmodified mRNA would.

So McCullough is correct on two points. The modification was intentional, and it makes the mRNA longer-lived than ordinary mRNA. Early public messaging that the mRNA would be “gone in a few days” was too confident.

Why it still breaks down

The cell’s normal mRNA disposal system does not work by reading uridines. It works mainly at the two ends of the molecule, which were not modified. Every mRNA, natural or synthetic, has:

a 5′ cap, a chemical structure at the front end that protects the molecule and helps start protein production, and
a 3′ poly-A tail, a string of adenines at the back end.

Normal mRNA decay proceeds in a well-characterized sequence:

Deadenylation. Enzyme complexes (CCR4–NOT and PAN2–PAN3) shorten the poly-A tail. This is usually the first and rate-limiting step. The vaccine mRNA carries a poly-A tail, and that tail is shortened just as a natural one is.
Decapping. Once the tail is short, the enzyme DCP2 removes the 5′ cap.
5′→3′ degradation. With the cap gone, the exonuclease XRN1 digests the mRNA from the front end.
3′→5′ degradation. Alternatively, the exosome complex digests the mRNA from the back end.

Because the vaccine’s modification is in the bases along the middle of the strand, not in the cap or the tail, this pathway still operates. Some endonucleases may cut modified RNA less efficiently. The innate-immune route of destruction (RNase L) is largely avoided. But the main housekeeping route remains open. The modified mRNA decays more slowly, but it decays.

What the measurements show

Studies have detected vaccine mRNA in blood for days up to about four weeks after injection, at steadily falling levels. By day 28, plasma levels in all measured subjects had dropped below 0.001 ng/mL. In lymph node germinal centers, vaccine mRNA has been found for up to about 60 days (Röltgen et al., Cell, 2022). That is longer than originally advertised. But a steady fall toward undetectable levels is direct evidence that the body is breaking the mRNA down.

Where the open question lies

Reports of persistence for years come from a small number of people. Most are case reports or small cohorts, and some come from laboratories whose methods have not yet been independently replicated. Several measure spike protein rather than mRNA. Persistent protein could come from long-lived cells, from tissue reservoirs, or from other sources, not necessarily from mRNA surviving indefinitely.

Since the chemical modification is identical in everyone who was vaccinated, it cannot by itself explain why a minority appear to retain spike much longer than everyone else. Why some people clear these products slowly is a legitimate research question, and an important one.

The hopeful conclusion

The human body has machinery that clears vaccine mRNA, and in most people it does so within weeks. For someone worried that the mRNA in his body is permanent, the evidence says otherwise. Where symptoms persist, the question is why a particular person’s clearance or immune response has gone wrong. That is a question medicine can investigate and, in time, treat.

Other claims in the interview

“Vaccine spike doesn’t bind ACE2.” The “2P” proline substitutions lock spike in its pre-fusion shape. That is the shape that binds ACE2; the receptor-binding domain is unchanged. The proposed reason that vaccine spike escapes clearance therefore rests on a mistaken premise.

The antibody test is not a spike test. The LabCorp anti-spike assay measures a person’s antibodies to spike, not the spike protein in his body. Antibody levels are highest in those with the most doses and infections, whether or not they have symptoms. No validated threshold, at 1,000 units or anywhere else, separates “free and clear” from “spike burden.” A falling number after treatment is also what time alone would produce. The Yale investigators noted that some participants with post-vaccination syndrome had no measurable spike protein at all, and they could not say whether spike levels cause the symptoms.

Batch variability. A 2023 Danish analysis found that adverse-event reports varied across Pfizer lots. But lots also differed in size, timing, and who received them, all of which affect reporting rates. The analysis does not show that 75% of batches were effectively placebo, and one person’s low antibody level does not show her doses were inert.

One cause for twelve diagnoses. Attributing EBV reactivation, mold illness, SIBO, tick-borne disease, and POTS to a single upstream cause is attractive but risky. Some of these are distinct, testable, and treatable conditions, and a patient told that everything is “spike” may miss one that matters.

Conflict of interest. The supplement bundle recommended in the interview is sold under Dr. McCullough’s name. That does not make it ineffective, but readers should know it when they weigh the recommendation.

“A healthy child remains healthier with no vaccines whatsoever.” This goes far beyond anything the COVID-19 data can show. It sweeps in vaccines against measles, pertussis, tetanus, and Hib, where the evidence of benefit is extensive.

The detox protocol: clinical experience versus proof

Dr. McCullough reports that patients on his protocol of nattokinase, bromelain, and curcumin have improved, and he cites his own antibody level falling from about 2,300 to about 800. These reports deserve respect. Clinicians see things before trials confirm them, and some of medicine’s best treatments began as clinical observations.

But clinical improvement, by itself, cannot tell us whether the protocol caused the improvement. Three other explanations must be ruled out:

Natural recovery. Many post-viral and post-vaccination syndromes improve over months without treatment.
Regression to the mean. Patients usually seek treatment when symptoms are at their worst, so improvement often follows whatever is done next.
Placebo response. Expectation of benefit measurably improves symptoms such as fatigue, brain fog, and pain.

A randomized, placebo-controlled trial is the tool designed to separate these explanations from a real effect, which is why McCullough himself calls for one. Until it is done, the protocol is best described as plausible and unproven.

The laboratory evidence is suggestive: nattokinase has been shown to degrade spike protein in a test tube. That does not establish that oral nattokinase reaches tissues at effective levels or relieves symptoms in people.

It also carries real risk. Nattokinase is fibrinolytic, and bromelain and curcumin both have antiplatelet effects. At high doses together they increase bleeding risk, especially for anyone taking aspirin, anticoagulants, antiplatelet drugs, or SSRIs, or anyone facing surgery or dental work. Anyone considering this protocol should do so under the care of a physician who knows his other medications.

Finally, there is no evidence that sweating eliminates spike protein. Detection of vaccine mRNA in breast milk does not establish sweat as a clearance route; breast milk is not “modified sweat” in any physiologically useful sense.

The NHIS autism study

The study by Hulscher, McCullough, and colleagues analyzed 21,990 children in the CDC’s National Health Interview Survey. It reports higher odds of autism, ADHD, anxiety, asthma, and mental-health treatment among COVID-vaccinated children.

To the authors’ credit, they state the main limitation plainly. The survey is cross-sectional and records no diagnosis dates, so it cannot show that vaccination came before diagnosis. Several further problems weigh against a causal reading:

It is a preprint. It was posted to Zenodo in October 2026 and has not yet passed peer review.
Confounding by indication. During the pandemic, the CDC listed asthma and neurodevelopmental and developmental conditions among those that raised a child’s risk from COVID-19. Parents of such children were urged to vaccinate them, and more of them did. That alone would produce higher rates of asthma and autism among vaccinated children with no causal effect. It would also produce a “dose-response” pattern, since higher-risk children were more likely to receive boosters.
The negative control failed. The authors used flu vaccination as a comparison exposure, and it too was associated with autism, ADHD, and asthma. That is the signature of families who use more healthcare and therefore receive more diagnoses. The authors reply that children who received a COVID shot but no flu shot had 45% higher odds of autism than children who received a flu shot but no COVID shot. But families who made those opposite choices in 2022–2024 likely differ in ways the survey does not capture.
Healthcare engagement appears throughout. The largest associations were with mental-health therapy (+60%) and mental-health medication (+57%), outcomes that depend heavily on access to and use of care.
The headline subgroup is small. The 154% figure for children aged 5–7 with three or more doses comes from a small group. Most autism is diagnosable by age 2–3, before many of these children could have received three doses.
Other evidence points the other way. In a study from an NIH clinical trials network, children whose mothers received an mRNA COVID-19 vaccine during or shortly before pregnancy showed no difference in neurodevelopmental outcomes at 18 to 30 months compared with matched controls.

The authors’ own recommendation is the right one: a linked birth-cohort study with complete vaccination and medical records and clinically confirmed diagnoses. Until that is done, this paper is a weak signal at best. Its conclusion that childhood COVID vaccination “should cease immediately” goes further than the authors’ own description of their finding as an association, not proof.

Our view

A person can reject the autism claim and still conclude that healthy children have little need for COVID-19 mRNA vaccination. Severe disease in healthy children has been rare, myocarditis risk is real, and several countries stopped recommending routine vaccination for healthy children years ago. That case rests on risk-benefit reasoning, not on this study.

For adults with persistent symptoms after vaccination or infection:

Your symptoms are real and worth investigating. Seek a careful workup that looks for treatable conditions rather than assuming a single cause.
Your body can clear vaccine mRNA. The decay machinery described above works on it, and in most people it is gone within weeks.
Treat supplement protocols as promising but unproven, and do not start high-dose nattokinase or similar combinations without a physician who knows your other medications.
Read a spike antibody number for what it is: a record of immune exposure, not a measure of spike protein in your body.

Sources

Yale School of Medicine, “Immune markers of post-vaccination syndrome indicate future research directions” (2025): https://news.yale.edu/2025/02/19/immune-markers-post-vaccination-syndrome-indicate-future-research-directions
“Persistent SARS-CoV-2 Spike Protein in Vasculitic Skin Lesions after Infection or mRNA Vaccination,” Acta Dermato-Venereologica: https://medicaljournalssweden.se/actadv/article/view/46594
Review of decay kinetics of vaccine mRNA and spike protein, Preprints.org (2025): https://www.preprints.org/manuscript/202507.1359/v1
Röltgen K, et al. “Immune imprinting, breadth of variant recognition, and germinal center response in human SARS-CoV-2 infection and vaccination.” Cell 2022;185:1025–1040.
Karikó K, Buckstein M, Ni H, Weissman D. “Suppression of RNA recognition by Toll-like receptors: the impact of nucleoside modification and the evolutionary origin of RNA.” Immunity 2005;23:165–175.
Hulscher N, et al. “Neurodevelopmental Outcomes in COVID-19-Vaccinated Versus Unvaccinated U.S. Children: Analysis of the National Health Interview Survey, 2022–2024” (preprint): https://zenodo.org/records/23165256
US Pharmacist, “Study Finds No Autism Link with mRNA COVID-19 Vaccination in Pregnancy” (2026): https://www.uspharmacist.com/article/study-finds-no-autism-link-with-mrna-covid19-vaccination-in-pregnancy
McCullough P. “The Un-Spiking of America” and Hulscher N. “Breaking Study: COVID-19 Vaccination of Children…,” Focal Points (2026): https://www.thefocalpoints.com

Essays

261008 – Spike Protein and mRNA Clearance