There is a problem with bad science that doesn’t end when the science is corrected. Sometimes that is when the real problem begins.

A paper can be methodologically weak. It can be rejected. A preprint can be removed. A journal can withdraw it. Independent researchers can tear its methods apart. The authors can fail to adequately answer those criticisms.

And none of that necessarily puts the claim back in the bottle. Because by then the headline has escaped.

“74% of deaths caused by COVID vaccines.”

“Vaccinated children have far more chronic disease.”

Those are memorable.

“Inappropriate study methodology, uncontrolled confounding, differential healthcare utilization and unequal follow-up” doesn’t exactly fit on a meme.

And therein lies the advantage.

A bad study does not necessarily have to survive scientific scrutiny to accomplish its purpose in the information ecosystem. It only has to survive long enough to generate the headline.

The autopsy paper that wouldn’t die

Consider the COVID-19 vaccine autopsy paper by Nicolas Hulscher, Peter McCullough and colleagues.

In July 2023, the authors uploaded a preprint titled A Systematic Review of Autopsy Findings in Deaths after COVID-19 Vaccination to SSRN’s Preprints with The Lancet.

Its explosive conclusion was that 73.9% of 325 deaths reviewed by the authors were directly caused by, or significantly contributed to by, COVID-19 vaccination.

That is one hell of a headline. There was also a serious problem. The preprint was removed. And let’s be precise about why.

Preprints with The Lancet currently state directly on the paper’s SSRN page:

“This preprint has been removed … because the study’s conclusions are not supported by the study methodology.”

It was not a Lancet paper that had passed peer review and was suddenly censored. It was a preprint, and the platform removed it because the conclusions weren’t adequately supported by the methods.

The paper didn’t disappear.

It was uploaded elsewhere and continued circulating.

Then, in June 2024, a version appeared as an Article-in-Press in Forensic Science International.

Once again, the “74%” claim exploded across social media. And once again, serious methodological concerns followed. This time the journal itself ultimately withdrew the paper.

The withdrawal notice is unusually explicit. Editors cited concerns involving inappropriate citations, inappropriate methodology, errors, misrepresentation, lack of factual support for the conclusions, and failure to recognize or cite disconfirming evidence.

The authors were allowed to respond and submitted a revised manuscript. The journal then sent that revision to two independent peer reviewers.

According to the withdrawal notice, both concluded that the revision did not adequately resolve the concerns and that the manuscript was unsuitable for publication.

The authors disagreed with the withdrawal. That should be noted.

But the scientific history is nevertheless clear: the preprint was removed once because its conclusions weren’t supported by its methodology, and the later journal version was withdrawn after additional scrutiny and independent peer review.

So did the “74%” claim die?

Of course not.

Just find another home

In November 2024, essentially the same central analysis appeared in Science, Public Health Policy and the Law, where it is labeled “Peer Reviewed.”

Its conclusion remained that 73.9% of the deaths reviewed were attributable to COVID-19 vaccination.

And today, the McCullough Foundation still lists the paper in its publication library.

McCullough was still invoking the same 74% figure in May 2025, writing that if autopsies were performed on all VAERS deaths, “~74%” would be due to vaccination—an extrapolation that goes considerably beyond what the autopsy paper itself could establish.

Think about what happened here.

The original claim survived:

Preprint removed → claim survives.

Published version withdrawn → claim survives.

Independent reviewers say the concerns weren’t adequately resolved → claim survives.

Paper appears somewhere else → headline lives again.

Scientifically, the paper encountered repeated problems. Communicatively, it succeeded spectacularly.

Then came An Inconvenient Study

The Henry Ford vaccinated-versus-unvaccinated study provides an even cleaner example.

This wasn’t a retracted study. It wasn’t even a published study. It was a 2020 draft that never made it through Henry Ford Health’s internal review process.

The analysis compared roughly 16,500 children who had received at least one vaccine with about 2,000 classified as completely unvaccinated and reported substantially more diagnosed chronic conditions among the vaccinated group.

That finding sounds devastating. Until you look at how the groups were constructed.

Henry Ford Health says its first internal review identified serious problems, including major demographic differences between the groups, a much smaller unvaccinated cohort and dramatically different periods of observation.

One-quarter of the unvaccinated children were observed only through six months of age, according to Henry Ford, and 75% only through age three—before many chronic childhood conditions are routinely diagnosed.

Independent researchers subsequently identified even more problems.

A peer-reviewed critique published in Clinical Infectious Diseases on August 5, 2026, highlighted uncontrolled confounding, exposure misclassification, unequal follow-up, surveillance bias and multiple-comparison problems.

One number captures the problem nicely:

Vaccinated children averaged about seven healthcare encounters per year. Unvaccinated children averaged about two.

Children who see healthcare providers more frequently have more opportunities to have asthma, eczema, ADHD, speech problems and other conditions discovered and entered into their medical records.

That doesn’t prove the vaccinated children weren’t actually sicker.

It means the study design cannot reliably tell us that vaccination was the reason.

The same critique found median follow-up of approximately 970 days for vaccinated children versus 461 days for unvaccinated children. The vaccinated group was therefore not only seeing doctors more frequently; it was being observed for roughly twice as long.

These aren’t cosmetic criticisms. They go directly to whether the headline result can support a causal interpretation.

But the draft had already escaped

None of that stopped the document from becoming An Inconvenient Study, a documentary produced by Del Bigtree’s Informed Consent Action Network.

The film’s central premise is dramatically more compelling than a discussion of ascertainment bias:

A researcher supposedly conducted the definitive vaccinated-versus-unvaccinated study, discovered an inconvenient truth and then the establishment wouldn’t publish it.

The official film site says:

“That study never saw the light of day... until now.”

It encourages viewers to distribute promotional cards, download and share the trailer, organize screenings and spread the film through social media and group chats. Its website currently claims more than 100 million views worldwide. That number comes from the filmmakers themselves and shouldn’t be treated as independently verified, but it illustrates the claimed scale of the campaign.

To its credit, ICAN doesn’t now claim the study definitively proves vaccines cause chronic disease.

Its response to critics says exactly that. ICAN calls the findings a “signal” warranting additional investigation and argues that sensitivity analyses lessen concerns about unequal follow-up.

That’s a fair position to include.

But it doesn’t resolve the methodological problems identified by other researchers.

And more importantly for this discussion, it doesn’t change what happened next.

The draft became a movie.

The movie became social-media content.

The social-media content became political testimony.

And the draft is still being amplified.

A year later, it was before lawmakers

On September 17, 2026—long after the methodological criticisms were public—the Michigan House Subcommittee on the Child Welfare System held a hearing specifically concerning the Henry Ford material. The official Michigan House website lists testimony from that hearing under “Henry Ford Health Studies.”

A week later, The HighWire featured Hulscher discussing his testimony and again promoted the Henry Ford vaccinated-versus-unvaccinated findings featured in An Inconvenient Study.

Think about that timeline.

The original document failed internal review in 2020. It resurfaced publicly in 2025. A documentary was built around it. Independent statisticians and epidemiologists detailed serious flaws. Henry Ford publicly rejected the conclusions attributed to it.

A peer-reviewed critique appeared in a major infectious-disease journal in August 2026.

And in September 2026, the same draft was being discussed before state lawmakers and amplified again to a large media audience.

Scientifically unfinished. Politically and culturally very much alive.

This is the part we underestimate

There is a widespread assumption that science eventually corrects itself, and therefore misinformation based on bad research will eventually disappear.

The first part is often true.

The second isn’t.

Researchers have actually studied what happens to papers after retraction.

One analysis examined 7,813 retracted biomedical papers and more than 13,000 post-retraction citation contexts.

Only 5.4% of those post-retraction citations acknowledged that the cited paper had been retracted.

In other words, the overwhelming majority continued citing the work without even mentioning its retracted status.

Another study compared more than 3,000 retracted papers with matched controls. Retraction reduced citation rates substantially, but didn’t kill them. The authors concluded simply:

“Retracted papers often live on.”

And that’s inside the scientific literature.

Now imagine the problem on Facebook, X, podcasts, documentaries, Substack and YouTube, where nobody is required to update an old meme because the methodology collapsed six months later.

The asymmetry is enormous

The original claim gets: “Vaccinated kids are 2.5 times sicker!”

The correction gets:

“The hazard ratio may be severely biased because of nonexchangeable cohorts, differential follow-up, ascertainment bias and time-dependent exposure misclassification.”

Guess which one travels farther. The original claim gets a trailer. The rebuttal gets a methods table. The original claim gets a documentary. The correction gets a journal article. The original claim takes ten seconds to understand.

Explaining why it is unreliable can take ten minutes. And once people have emotionally invested in the original story, the correction can actually be folded back into it.

Study rejected? Suppression.

Preprint removed? Censorship.

Paper withdrawn? They were afraid of it.

Institution says the methods were flawed? The institution is covering itself.

Independent reviewers reject the revision? Gatekeeping.

At that point, the normal mechanisms of scientific quality control become evidence for the conspiracy they’re supposed to refute.

It is almost perfectly self-sealing.

Was that the plan?

I don’t know. And importantly, I don’t have to know.

I can’t prove Hulscher, McCullough, Bigtree or anyone else involved sat down and consciously devised a strategy in which scientific rejection would be converted into publicity.

Claiming that as fact would require evidence of intent I don’t have. The more defensible point is also the more troubling one:

The information ecosystem rewards the behavior whether it was planned that way or not.

Make a sensational claim and millions may see it. When the methodological critique arrives, a fraction of them will read it. If the paper is withdrawn, an even smaller fraction will read the withdrawal notice.

And some of the original audience will interpret the withdrawal itself as confirmation that powerful people wanted the information suppressed.

The original claim therefore receives the benefit of both stages:

Publication becomes evidence that the claim is true.

Rejection becomes evidence that the claim is dangerous to powerful interests.

Heads they win. Tails science loses.

That’s convenient propaganda

“Propaganda” doesn’t require every underlying statement to be fabricated.

Effective propaganda frequently contains real numbers, genuine documents and actual events.

What makes it powerful is selection and framing.

Take the most frightening result.

Strip away the methodological uncertainty.

Turn it into a headline.

Repeat it.

Build a narrative around its suppression.

And when the criticism arrives, don’t let that end the story.

Make the criticism part of the story.

That is why arguing about whether a bad study was technically “debunked” sometimes misses the larger issue.

The scientific argument may already be over.

The communications battle isn’t.

The Hulscher autopsy paper could be removed from one platform, withdrawn by a journal, and heavily criticized, yet its central number still circulates.

The Henry Ford analysis can remain an unpublished draft with serious unresolved biases, yet become a documentary, testimony before lawmakers and recurring media content.

They don’t have to win the scientific argument. They just have to keep the claim alive.

And in an ecosystem where the allegation travels faster than the correction, and censorship is the ready-made explanation for every rejection, even losing the scientific argument can be turned into a propaganda victory.

That may be the most inconvenient study of all.

Resources

1. Preprints with The Lancet / SSRN — Hulscher et al. autopsy preprint. The page states that the preprint was removed because its conclusions were not supported by its methodology. SSRN preprint and removal notice

2. Forensic Science International — withdrawal notice. Lists the journal’s concerns and says two independent reviewers found the revised manuscript still unsuitable for publication. Forensic Science International withdrawal notice

3. Science Feedback — analysis of the autopsy paper. Discusses selection bias, causality problems, and why the methodology does not establish that vaccination caused 74% of the deaths reviewed. Science Feedback review

4. Science, Public Health Policy and the Law — later publication of the autopsy analysis. Useful for reading the authors’ own methods, conclusions and limitations rather than relying solely on critics. Later publication of Hulscher et al.

5. McCullough Foundation — Publications. Currently lists the autopsy paper among its publications. McCullough Foundation publication library

6. Henry Ford Health — statement on the unpublished vaccinated-versus-unvaccinated analysis. Describes the institution’s reasons for abandoning the draft after internal review. Henry Ford Health statement

7. Clinical Infectious Diseases — “Vaccines, Bias, and the Perils of Non-peer-reviewed Studies.” Peer-reviewed August 2026 methodological critique addressing confounding, healthcare utilization, exposure classification, follow-up and multiple outcomes. Clinical Infectious Diseases critique

8. ICAN / An Inconvenient Study — Responses to criticism. The filmmakers’ response to Henry Ford, Jake Scott and other critics, including their argument that the draft represents a safety signal rather than definitive proof. ICAN’s responses to criticisms

9. An Inconvenient Study — official site. Shows how the unpublished research has been packaged and promoted, including campaigns for sharing, screenings and promotional materials. Official An Inconvenient Study website

10. Michigan House Subcommittee on the Child Welfare System — September 17, 2026. Official committee page documenting the hearing and Henry Ford Health Studies testimony. Michigan House committee record

11. The HighWire — Episode 495, September 24, 2026. Shows the continued amplification of the Henry Ford analysis after the major methodological criticisms were already public. The HighWire episode

12. Hsiao & Schneider — Continued use of retracted papers. Analysis of 7,813 retracted biomedical papers showing that only 5.4% of examined post-retraction citation contexts acknowledged the retraction. Quantitative Science Studies paper

13. Kühberger et al. — Self-correction in science. Study of more than 3,000 retracted papers finding that retraction reduced citations but did not eliminate them: “retracted papers often live on.” PLOS ONE study on post-retraction citations

Since my original review, other analysts have independently found some of the same numerical problems. A deeper look also raises additional questions about healthcare use, vaccination-status misclassification, practice effects, temporality and multiple testing.

Update: October 7, 2026

When I first went through Comparative Health Outcomes of Vaccinated versus Unvaccinated Children: A Retrospective Cohort Analysis, I came away with a fairly simple conclusion:

There may be signals in this dataset worth investigating.

But the analysis presented in the paper does not establish that vaccines caused vaccinated children to be less healthy.

Since then, I’ve gone back through the study again and looked for responses, corrections, and independent analyses.

Something important has happened.

Other reviewers have now independently identified several of the same problems I raised, including the bizarre Table 4 denominator that somehow contains more vaccinated children than the study says existed.

And the deeper I look, the less this resembles a simple debate over whether the paper has been peer-reviewed.

The bigger issue is whether the published numbers, exposure definitions, and analysis are internally reliable enough to support the claims being made from them.

The paper remains on Zenodo as Version v1, published and last modified September 26. As of October 7, I could not find a revised version correcting the tables or a public response from the authors explaining the discrepancies discussed below. Zenodo

First, the basic design

The researchers analyzed electronic medical records from eight U.S. pediatric practices encompassing 6,239 children. The Zenodo abstract describes vaccination status as at least one vaccine before 13 months versus “zero vaccines ever.” Nine diagnostic categories were examined using ICD codes, with overall, sex-stratified and vaccine-dose analyses. Zenodo

Children’s Health Defense then promoted the findings as evidence that vaccinated children had substantially higher odds — often described as higher “risk” — of several conditions, including recurrent respiratory disease, ear infections and “strep.” Children’s Health Defense

Those associations are reported in the paper. What the study does not establish is that vaccination caused them. And there are some much more basic problems to settle first.

Table 4 still doesn’t add up

This was the first thing that really stopped me when I originally reviewed the paper.

The study says its vaccinated cohort contains:

2,414 children.

But the respiratory-disease line in Table 4 contains:

1,369 vaccinated children with disease

plus

1,324 vaccinated children without disease.

That equals:

2,693 vaccinated children.

That is 279 more vaccinated children than supposedly exist in the cohort.

Since my article was published, this discrepancy has now been independently identified by at least two other analyses.

Vaxopedia/StopAntiVaxPropaganda raised the same 2,693-versus-2,414 discrepancy on October 2. Tech ARP independently reported it again on October 6. Stop Anti-Vax Propaganda

This matters because this isn’t an argument over vaccine policy, confounding or causal philosophy. It’s arithmetic.

And my original observation goes a step further.

When the male and female respiratory counts in Tables 5 and 6 are combined, they do total the stated 2,414 vaccinated children.

Those counts are:

523 + 484 = 1,007 respiratory cases

720 + 687 = 1,407 noncases

Total = 2,414

But those numbers produce an odds ratio of approximately 1.19, rather than the 1.6 reported in Table 4.

Maybe there is an innocent table-construction error. Maybe two analytic populations got mixed together. Maybe a row was copied incorrectly.

But somebody needs to explain it. Until then, one of the paper’s most heavily promoted findings has an unresolved denominator problem.

Hooker has a long history with vaccinated-vs.-unvaccinated claims

Brian Hooker is not a neutral newcomer to this research question. For years, he has argued that vaccinated children have worse health outcomes than unvaccinated children, and he has both published and promoted studies designed to support that conclusion.

In 2020, Hooker and Neil Miller published a vaccinated-versus-unvaccinated analysis in SAGE Open Medicine using records from three pediatric practices. The paper reported higher odds of developmental delay, asthma, ear infections, and gastrointestinal disorders among vaccinated children. Children’s Health Defense promoted the study under the headline that unvaccinated children had “better health outcomes.” ResearchGate

Hooker and Miller followed that with another paper in 2021 examining vaccinated and unvaccinated children while adjusting for breastfeeding and type of birth. That study again reported poorer outcomes among vaccinated children. ResearchGate

Hooker has also spent years promoting Anthony Mawson’s 2017 vaccinated-versus-unvaccinated homeschool survey and similar studies. Children’s Health Defense maintains a page prepared by Hooker compiling multiple “vaxxed vs. unvaxxed” studies, including Mawson, Hooker and Miller, and work by James Lyons-Weiler and Paul Thomas. CHD Pennsylvania

In 2023, Hooker co-authored Vax-Unvax: Let the Science Speak with Robert F. Kennedy Jr., a book built around more than 70 comparisons of vaccinated and unvaccinated groups. Hooker has continued presenting these studies publicly as evidence that unvaccinated children are healthier. Children’s Health Defense

That history does not invalidate the new Yengst et al. study. A researcher’s prior views do not determine whether the data are correct.

But it does make methodological rigor especially important.

Several of the studies Hooker has previously promoted have themselves been heavily criticized for selection bias, inadequate control of confounding, self-reported outcomes, nonrepresentative populations, or other methodological problems. Science Feedback’s 2023 review of Hooker’s vaccinated-versus-unvaccinated claims concluded that the collection presented a one-sided picture and noted that some of the prominently cited studies had been retracted or contained major methodological limitations. Science Feedback

That makes the current paper’s problems especially relevant. Once again we have a highly selected population, inadequate adjustment for healthcare use and confounding, unclear exposure classification, irregular dose-response patterns and, this time, an apparently impossible denominator in a central results table.

The point isn’t that Hooker’s history automatically makes Yengst et al. wrong.

The point is that this is a longstanding advocacy position of one of the paper’s authors, so the appropriate response is not to give the new study less scrutiny. It is to demand enough transparency and reproducibility that the analysis can stand independently of the authors’ prior beliefs.

And right now, some of its most basic numbers still don’t.

Healthcare-use bias isn’t hypothetical

The paper relies heavily on diagnoses appearing in electronic medical records.

For several outcomes, one diagnosis wasn’t enough. Children had to accumulate at least three diagnoses during twelve months. That makes healthcare utilization critically important.

The more often a child sees a pediatrician, the more opportunities that child has to receive a vaccine. The same additional visits create more opportunities for a physician to document an ear infection, rash, respiratory illness, sore throat or fever.

Conversely, a child whose parents use pediatric care less frequently may experience similar minor illnesses without ever generating three medical-record diagnoses.

The authors acknowledge they don’t have complete healthcare-utilization information. And we now have very good reason not to treat that as a hypothetical concern.

A large 2013 JAMA Pediatrics study followed more than 323,000 children and found that undervaccinated children had significantly fewer outpatient visits than age-appropriately vaccinated children. Children undervaccinated specifically because of parental choice also had fewer outpatient and emergency-department encounters. PubMed

That is almost exactly the sort of systematic difference that can distort an electronic-record comparison like this one.

You cannot simply equate fewer diagnoses in the chart with less disease in the child when one group interacts differently with the healthcare system.

There is another EHR problem: are the exposure groups themselves accurate?

This deserves more attention than I gave it originally. The paper wants to distinguish children who received vaccines from children who received none.

But those classifications are being constructed from medical records across eight practices.

How was “zero vaccines ever” established?

Were state immunization registries checked? Were outside pediatricians captured? Health-department vaccination clinics? Pharmacies? Other health systems? Did families change providers?

This is not an obscure concern.

Researchers specifically studying the feasibility of vaccinated-versus-undervaccinated safety studies have examined vaccination-status misclassification in electronic health records.

In a 2017 Vaccine study involving more than 361,000 children, researchers used chart review and parent surveys to validate EHR vaccination classifications. They also found that parents of undervaccinated children were more likely to report using alternative medical providers. PubMed

A subsequent study explicitly examined both confounding and misclassification bias in childhood-immunization-schedule research and concluded that measurement and control of disease risk factors need careful consideration. PubMed

That doesn’t prove Yengst et al. misclassified anybody. It means the paper needs to demonstrate how it ruled that problem out.

And the “unvaccinated” definition is still confusing

The Zenodo abstract says the comparison is:

at least one vaccine before 13 months versus zero vaccines ever. Zenodo

But the manuscript also discusses 469 children whose first vaccination occurred after 13 months and indicates that they were retained in the primary analysis as unvaccinated before 13 months.

Those definitions don’t cleanly match.

Other reviewers have now noticed the same problem. Stop Anti Vax Propaganda

If those 469 children are in the reference group, then this isn’t simply “vaccinated versus never vaccinated.”

If they’re excluded from the never-vaccinated group, the totals and language need to make that clear.

Exposure classification is the foundation of a study like this. It should not require readers to reverse-engineer what happened to hundreds of children.

The sample itself is extraordinarily unusual

The paper’s comparison group contains 3,825 of 6,239 children - about 61% of the entire sample.

Because of the ambiguity surrounding those 469 late-vaccinated children, I would not automatically say all 3,825 remained completely unvaccinated. If the 469 are removed, there would still be approximately 3,356, or about 54% of the entire cohort, left in that category.

Either way, this is an extraordinarily unusual pediatric population.

For context, the large EHR validation study mentioned above found that only 1.3% of more than 361,000 children had received no vaccines by 24 months. PubMed

Those aren’t perfectly comparable definitions or populations, so don’t turn that into “61% versus 1.3%” as though it were an apples-to-apples prevalence comparison.

But the contrast tells us something important:

These eight practices are highly selected. That makes representativeness and confounding even more important.

Tech ARP has now independently raised the same concern about how unusual the study population is compared with ordinary U.S. pediatric populations. Tech ARP

Eight practices introduce another confounder

There is an additional issue I would add to my original critique: practice-level effects.

The children came from eight different pediatric practices.

Suppose one practice has a relatively high vaccination rate and also codes respiratory illnesses aggressively.

Suppose another practice serves many vaccine-refusing families and is less likely to bring children in for minor respiratory illnesses.

Pooling those practices together can create an apparent relationship between vaccination and diagnosis even if the relationship is much smaller — or absent, within individual practices.

This is why site, clinic, or provider can matter in multi-practice observational datasets.

The Vaccine Safety Datalink’s methodological work on studying the childhood schedule has emphasized that these observational comparisons are inherently complex and require careful attention to confounding and bias. PubMed

I would want to see the Yengst data stratified by practice, and preferably analyzed with practice incorporated directly into the statistical model.

Which came first? The disease or the vaccine?

The original article already raised temporality, but the problem is even more fundamental than “the authors didn’t provide enough dates.”

The exposure is determined by vaccination through 13 months. Yet some diagnoses apparently occurred extremely early in life.

If a child develops a respiratory condition at three weeks of age and receives a first vaccine at four months, that child may eventually be classified as vaccinated.

But the disease at three weeks obviously cannot have been caused by a vaccine received months later.

That means a valid causal analysis has to make sure that outcomes counted against the exposed group actually occurred after exposure.

Ideally, vaccination would be treated as a time-varying exposure or the analysis would define appropriate index dates and begin outcome observation afterward.

The authors themselves acknowledge that their analysis did not establish the timing of particular vaccinations relative to diagnoses.

So the problem isn’t simply that chronology wasn’t reported. The design can potentially assign pre-exposure disease to the eventual exposed group.

That seriously weakens causal interpretation.

“Strep throat” still isn’t really a clean strep-throat endpoint

Children’s Health Defense has emphasized the large association with recurrent “strep.”

But the paper’s diagnostic definition isn’t confined to streptococcal pharyngitis.

Yes, J02.0 represents streptococcal pharyngitis.

But the composite also includes much broader diagnostic families. Some cover nonspecific upper-respiratory infections, croup and other infections. The coding list even extends into categories for gonococcal and chlamydial disease.

Official ICD documentation distinguishes these diagnoses rather than treating them as synonymous with streptococcal throat infection. Centers for Medicare & Medicaid Services

StopAntiVaxPropaganda independently noticed this problem after my article appeared. Stop Anti Vax Propaganda

So if the strongest headline is going to be that vaccinated children had dramatically more recurrent strep throat, the endpoint should actually measure recurrent strep throat.

As written, it is a much broader diagnostic bucket.

And R19 is still not a respiratory disease

The paper’s respiratory category reportedly includes R19.

R19 is not a respiratory-disease family. It belongs to signs and symptoms involving the digestive system and abdomen.

Again, there may be an innocent explanation. Perhaps R19 was supposed to be R09. Perhaps it’s only a manuscript typo and wasn’t used in the code.

But that distinction matters.

If it’s merely a typo, correct the table. If R19 actually entered the analysis, rerun the analysis.

Other reviewers independently flagged this after my original article as well. Stop Anti Vax Propaganda

The list of nine outcomes doesn’t even stay consistent

The Methods and diagnostic-code table include nausea/vomiting.

The main outcome table instead contains sinusitis.

Yet no corresponding sinusitis definition appears where you would expect it.

Any single mistake like that could be mundane.

But we’re no longer dealing with one typo.

We now have an impossible vaccinated denominator, conflicting exposure definitions, an unusual “strep” composite, a gastrointestinal code in a respiratory outcome, and an outcome appearing in the result table that doesn’t match the outcome-definition table.

At some point “probably a typo” stops being an adequate response.

The paper needs a corrected version.

The “dose response” remains much less impressive than advertised

The dose-response argument sounds powerful until you actually look at the sequence of odds ratios.

For the paper’s strep endpoint: 1.5 → 2.2 → 3.4 → 2.7

That’s an increase followed by a decrease.

For respiratory disease: 1.2 → 1.6 → 2.0 → 0.50

At the highest exposure category, the association completely reverses.

For developmental delay: 0.82 → 0.66 → 0.74 → 1.4

The middle categories point toward lower odds before the highest group flips toward higher odds. Calling all of that a straightforward harmful dose-response relationship is difficult to justify.

And the interpretive asymmetry remains a problem.

• When a vaccinated group appears healthier, the paper invokes possible selection effects or healthy-vaccinee bias.

• When a vaccinated group appears less healthy, biological vaccine harm receives much more attention.

You can’t use confounding selectively.

If bias can create a protective association, it can also create a harmful one.

Multiple comparisons deserve more than a footnote

The paper examines nine outcomes overall, sex-specific analyses and four vaccine-exposure categories.

That produces dozens of statistical tests.

A rough count is about 63 comparisons if all nine outcomes are considered across the overall, sex-stratified and four dose-category analyses.

The authors state that they did not correct for multiple comparisons. That does not make every significant finding disappear.

But it matters particularly when highlighting one subgroup result because its p-value crossed 0.05 after examining many different combinations.

Take the much-publicized developmental-delay finding in the 10+ category: the Zenodo abstract reports OR 1.4, p = 0.0031. Zenodo

For illustration only, if all 63 comparisons were treated as one statistical family, a simple Bonferroni threshold would be approximately 0.0008.

Under that deliberately conservative approach, p = 0.0031 would not qualify.

I’m not arguing that Bonferroni across exactly 63 tests is necessarily the correct solution. Researchers can reasonably define multiplicity families differently, and less conservative procedures exist.

The point is that “p = 0.0031” does not exist in isolation when it was selected from dozens of analyses.

CHD still confuses odds with risk

This criticism also survives intact. The study reports odds ratios.

CHD and Hooker repeatedly translate them into language such as “60% more likely” or “double the risk.” Children’s Health Defense

For uncommon outcomes, odds ratios and risk ratios can be similar.

For common outcomes they can differ substantially, with odds ratios exaggerating the apparent relative change when interpreted as risk ratios. This is a longstanding statistical issue. BMJ

Using Table 4 as printed - notwithstanding its denominator problem - the respiratory proportions are roughly 50.8% versus 39.6%.

That’s about a 28% relative difference in observed proportions, not “60% higher risk.”

The 60% number comes from treating an OR of 1.6 as though OR and relative risk were interchangeable.

They aren’t.

CHD’s own dose-category reporting remains wrong

This wasn’t something critics imposed on the paper. It’s a discrepancy between the paper and CHD’s own coverage.

CHD reported the developmental-delay increase in the 6–9 dose group.

But Table 7 gives that group an OR of approximately 0.74.

The 1.4 result belongs to the 10+ category.

The same type of error occurs with rhinitis: the OR of 2.3 belongs to the 10+ group rather than the 6–9 group.

When an advocacy organization is using a paper produced by its own staff to make a public case, getting the exposure categories right seems like the minimum requirement.

What has happened since my original article?

This is worth emphasizing.

The criticism has not gone away under scrutiny. It has begun to be independently reproduced.

On October 2, Vaxopedia/StopAntiVaxPropaganda independently identified the oversized Table 4 vaccinated denominator, the inconsistent definition of “unvaccinated,” the broad strep coding, the misplaced R19 category and the lack of healthcare-utilization control. That article also linked back to my original analysis. Stop Anti Vax Propaganda

On October 6, Tech ARP independently identified the same 2,693 versus 2,414 problem and highlighted healthcare ascertainment and the unusually selected eight-practice sample. Tech ARP

Meanwhile, supportive promotion has continued. A post published October 3 presents the paper as evidence that earlier or greater childhood vaccination is not beneficial. Opinyuns

Steve Kirsch has also added the Yengst paper to a compilation whose headline describes the included work as being “all published in the peer-reviewed literature,” even though the Yengst paper is currently a Version v1 Zenodo upload, not a peer-reviewed journal publication. Kirsch Substack

And as of October 7, the Zenodo record still shows Version v1 with no visible corrected version. Zenodo

What would actually strengthen this study?

None of this requires dismissing the dataset. If anything, a cohort containing a large population of minimally vaccinated or unvaccinated children could be scientifically useful.

But usefulness depends on doing the analysis properly.

I would like to see the authors reconcile every Table 4 denominator; explicitly account for the 469 children first vaccinated after 13 months; document how vaccination status was validated outside the originating practices; report healthcare-visit frequency; stratify or adjust for practice; control important demographic, birth, and family confounders; use validated diagnostic definitions; make vaccine exposure time-varying so diagnoses cannot precede the exposure supposedly causing them; address multiplicity; publish analytic code; and provide enough de-identified data for independent reproduction.

Then rerun the study.

If the associations survive all of that, my assessment changes.

That would be genuinely interesting evidence.

The bottom line

I began with the position that this study should not be dismissed merely because it came from Children’s Health Defense or because it hasn’t been peer-reviewed.

I still hold that position. But scrutiny works both ways.

The deeper problem is that some of the paper’s basic numbers do not reconcile, its exposure groups are ambiguous, its disease definitions contain questionable coding choices, healthcare utilization isn’t controlled, important confounders aren’t adjusted, temporal order isn’t reliably established, and dozens of comparisons are interpreted without correction for multiplicity.

Since I first pointed these issues out, other reviewers have independently found several of the same things.

And the central Table 4 question remains unanswered:

Before anyone tells me this paper proves vaccinated children are less healthy, answer that. Then show that the diagnosis happened after the vaccine.

Then show that the result survives healthcare-use adjustment, practice effects, confounding, corrected outcome definitions and a reproducible reanalysis.

That isn’t moving the goalposts. That’s the goalpost.

And if a result is real, it should survive getting the numbers right.

Resources

1. Yengst T, Ray H, Jablonowski K, Hooker BS. Comparative Health Outcomes of Vaccinated versus Unvaccinated Children: A Retrospective Cohort Analysis. Zenodo, September 26, 2026.
   Primary source for the study discussed in this article.
   Read the Zenodo paper

2. Michael Nevradakis. “Vaccinated Kids Have Significantly Higher Odds of Developing Multiple Chronic Health Conditions.” Children’s Health Defense, September 29, 2026.
   CHD’s coverage and interpretation of the study. Children’s Health Defense
   Read the CHD article

3. Vaxopedia / StopAntiVaxPropaganda. “Let’s Talk About Vaccinated vs Unvaccinated Children Again.” October 2, 2026.
   Independent critique identifying the Table 4 denominator discrepancy, inconsistent vaccination definitions, ICD-code problems and healthcare-utilization concerns. Stop Anti-Vax Propaganda
   Read the analysis

4. Adrian Wong. “Fact Check: Children’s Health Defense’s Vaccinated vs Unvaccinated Children Study.” Tech ARP, October 6, 2026.
   Independent review that also identifies the 2,693-versus-2,414 vaccinated-child discrepancy and medical-record ascertainment problems. Tech ARP
   Read the Tech ARP fact-check

5. Glanz JM et al. A population-based cohort study of undervaccination in 8 managed care organizations across the United States. JAMA Pediatrics. 2013;167(3):274–281. DOI: 10.1001/jamapediatrics.2013.502.
   Found different healthcare-utilization patterns among undervaccinated and age-appropriately vaccinated children. PubMed
   PubMed record

6. Daley MF et al. Assessing misclassification of vaccination status: Implications for studies of the safety of the childhood immunization schedule. Vaccine. 2017;35(15):1873–1878. DOI: 10.1016/j.vaccine.2017.02.058.
   Examines vaccination-status misclassification in electronic medical records and outside-provider use among undervaccinated families. PubMed
   PubMed record
   Free full text at PubMed Central

7. Newcomer SR et al. Assessing Potential Confounding and Misclassification Bias When Studying the Safety of the Childhood Immunization Schedule. Academic Pediatrics. 2018.
   Shows why confounding and accurate vaccination classification must be explicitly addressed in schedule-level observational studies. PubMed
   PubMed record

8. Glanz JM et al. White Paper on studying the safety of the childhood immunization schedule in the Vaccine Safety Datalink. Vaccine. 2016;34 Suppl 1:A1–A29. DOI: 10.1016/j.vaccine.2015.10.082.
   Discusses suitable designs, exposure definitions, confounding and bias when comparing different childhood vaccination patterns. PubMed
   PubMed record

9. Institute of Medicine. The Childhood Immunization Schedule and Safety: Stakeholder Concerns, Scientific Evidence, and Future Studies. 2013.
   Major National Academies review of how whole-schedule vaccine-safety questions can be investigated. National Academies Publications
   National Academies report page

10. Davies HTO, Crombie IK, Tavakoli M. “When can odds ratios mislead?” BMJ. 1998;316:989. DOI: 10.1136/bmj.316.7136.989.
    Explains why odds ratios should not automatically be interpreted as relative risks, particularly when outcomes are common. BMJ
    Read the BMJ article

11. Grant RL. “Converting an odds ratio to a range of plausible relative risks for better communication of research findings.” BMJ. 2014;348:f7450. DOI: 10.1136/bmj.f7450.
    Further discussion of the distinction between odds ratios and relative risks. BMJ
    Read the BMJ article

12. Centers for Medicare & Medicaid Services. ICD-10-CM Codes.
    Official ICD-10 coding resources useful for checking the diagnostic categories used in the Yengst paper. Centers for Medicare & Medicaid Services
    CMS ICD-10-CM resources

13. U.S. Food and Drug Administration. “Common Ingredients in FDA-Approved Vaccines.”
    Official reference for vaccine ingredients, including aluminum-containing adjuvants. fda.gov
    FDA vaccine-ingredient information

14. Hooker BS, Miller NZ. “Analysis of health outcomes in vaccinated and unvaccinated children: Developmental delays, asthma, ear infections and gastrointestinal disorders.” SAGE Open Medicine. 2020;8. DOI: 10.1177/2050312120925344.
    Hooker and Miller’s earlier vaccinated-versus-unvaccinated study using records from three U.S. medical practices. PubMed
    Read the paper on PubMed

15. SAGE. “Corrigendum to ‘Analysis of health outcomes in vaccinated and unvaccinated children.’” 2026.
    The correction added a previously omitted acknowledgement that the Pediatric Health Outcomes Initiative referred the physicians from the three practices supplying the EMR data and that one physician was a PHOI board member. The authors stated PHOI did not handle the data, design or financing. Sage Journals
    Read the SAGE corrigendum

16. SAGE. “Publisher’s note to ‘Analysis of health outcomes in vaccinated and unvaccinated children.’” August 28, 2026.
    SAGE says reader concerns led to an Expression of Concern. One concern was addressed by the corrigendum; another concerned possible reuse of data in later articles without disclosure. SAGE said it lacked enough information to resolve the latter and called its investigation inconclusive. The publisher’s note and corrigendum replaced the earlier Expression of Concern. Sage Journals
    Read the SAGE publisher’s note

17. Hooker BS, Miller NZ. “Health effects in vaccinated versus unvaccinated children, with covariates for breastfeeding status and type of birth.” Journal of Translational Science. 2021;7. DOI: 10.15761/JTS.1000459.
    A second Hooker-Miller vaccinated-versus-unvaccinated analysis, this time using survey data associated with three medical practices and models incorporating breastfeeding and birth type. OAText
    Read the 2021 paper

18. Kennedy RF Jr., Hooker BS. Vax-Unvax: Let the Science Speak. Skyhorse Publishing, 2023.
    Hooker co-authored a full book centered on comparisons between vaccinated and unvaccinated populations, demonstrating that this research question and argument long predate Yengst et al. The publisher describes the book as drawing on more than 100 studies involving vaccinated-versus-unvaccinated comparisons. Skyhorse Publishing
    Publisher’s page for Vax-Unvax

19. Science Feedback. “Claim by Brian Hooker that unvaccinated children are healthier compared to vaccinated children relies on flawed studies.” July 28, 2023.
    An independent review of studies Hooker publicly cited in support of the claim that unvaccinated children are healthier, discussing selection bias, confounding and other methodological limitations. Science Feedback
    Read the Science Feedback review

I used to trust Dr. Drew.

Not because I thought he was infallible. Not because I agreed with everything he said. I trusted him because, for decades, Drew Pinsky occupied a pretty specific place in American media: he was the doctor in the room.

On Loveline, while everyone else was joking around, Drew was usually the guy explaining what was medically real and what wasn’t. He spent more than 30 years doing that. Later, HLN gave him a nightly television program and introduced him as a practicing physician who was board-certified in internal medicine and addiction medicine. The network said his job was to provide viewers with useful medical perspective and “crucial information.” CNN Press Site

That’s the Dr. Drew I remember.

Which is why I keep coming back to the same question:

What the hell happened to Dr. Drew?

He got COVID wrong. Then he did something I respected.

Early in 2020, Drew badly underestimated COVID.

He compared it with influenza, called the reaction a “press-induced panic,” and repeatedly suggested the danger was being exaggerated. The Washington Post

Then he admitted he was wrong.

Not six years later. Not after pretending he’d been misunderstood. He publicly apologized.

“I got it wrong.”

He said Anthony Fauci had been his “north star” going back to the AIDS epidemic and said that when Fauci made clear COVID wasn’t ordinary influenza, he changed course. He urged people to listen to the CDC and public-health experts. The Washington Post

I respected that.

Seriously.

Getting something wrong isn’t what destroys my trust in somebody. Refusing to admit you’re wrong does. Science is supposed to work that way. You make your best assessment from the evidence you have. Better evidence comes in. You adjust.

That version of Drew made sense to me. But over the next few years, something changed. And it wasn’t simply that he became more skeptical of government.

I can live with skeptical. I’m skeptical too.

It was the ecosystem he increasingly became part of.

Then he joined The Wellness Company

On November 1, 2023, The Wellness Company announced that Drew Pinsky was joining its Chief Medical Board.

This wasn’t some loose affiliation. TWC said Drew was joining a board that included Peter McCullough, Harvey Risch, and James Thorp, as the company pursued what it openly called a “parallel healthcare system.” Drew himself said TWC’s mission was “completely and totally in line” with his own work. The Wellness Company

Today, TWC lists Peter McCullough as its Chief Scientific Officer and Drew Pinsky as its Chief Patient Officer. Kelly Victory, Harvey Risch, and others hold senior positions on the same medical board. The Wellness Company

And this isn’t just an organization publishing contrarian medical opinions.

It’s a business.

TWC sells an “Ultimate Spike Detox” based on what it calls the McCullough Protocol. It sells medical emergency kits containing prescription drugs, including ivermectin, azithromycin and antibiotics. The Wellness Company

There is even a TWC sales page specifically aimed at Dr. Drew’s audience, offering his fans a discount with the code DREW. The Wellness Company

That matters.

It doesn’t prove Drew is dishonest. It doesn’t prove every claim made by somebody affiliated with TWC is false. And it certainly doesn’t mean ivermectin can’t have legitimate medical uses. It does.

But we have spent years hearing that we should scrutinize the financial interests behind mainstream medicine.

I agree. So let’s do it consistently.

If Pfizer’s commercial interests are relevant when evaluating Pfizer’s messaging, then the commercial interests of an alternative-health company are relevant when evaluating its messaging too.

You don’t get to yell “follow the money” until the money leads to your own side.

Then I found his name on this paper

This was harder for me to dismiss as “Dr. Drew is just interviewing people.”

In September 2026, I came across a paper titled Proteolytic Targets and Clinical Benefits of Oral Nattokinase: Spike Protein and Amyloid Degradation, Fibrinolysis, and Cardiovascular Risk Reduction.

And there was Drew Pinsky’s name.

Not in the acknowledgments.

As a coauthor. Alongside Nicolas Hulscher, Harvey Risch, James Thorp, Kelly Victory and Peter McCullough.

The affiliation beside Drew’s name is The Wellness Company.

Then I read the disclosures.

The paper states that all of its authors are affiliated with and/or receive salary support from The Wellness Company.

That immediately caught my attention because The Wellness Company isn’t merely publishing information about nattokinase.

It sells it.

TWC currently markets an “Ultimate Spike Detox” containing nattokinase as part of what it calls the McCullough Protocol. It even has a webpage promoting this nattokinase paper while selling an 8,000-FU nattokinase-containing Spike Detox product on the same page.

Now here’s the part that matters. The paper does not demonstrate that swallowing nattokinase removes spike protein from people.

The authors themselves acknowledge that the spike-degradation evidence is currently in vitro. They explicitly state that nattokinase-mediated spike clearance has not been demonstrated in animals or humans. They acknowledge that we don’t yet know whether an oral dose produces enough active enzyme in human blood or tissue to reproduce what happens in a laboratory dish.

That’s not some hostile fact-checker’s interpretation.

It’s in their own paper.

So now the distinction becomes incredibly important.

“Nattokinase degrades spike under laboratory conditions” is one claim.

“Take our nattokinase-containing Spike Detox to remove harmful spike protein from your body” is a much bigger one.

The first has experimental evidence behind it.

The second requires the human clinical evidence that the paper itself says we still don’t have.

And once again, I’m back to the same rule:

If Pfizer published a favorable review of an ingredient while Pfizer sold the product containing that ingredient, I’d want the financial relationship disclosed and I’d examine the paper very carefully.

I’m not going to change that standard because the company happens to call itself The Wellness Company.

Follow the money has to work in both directions.

Then I started paying attention to who Drew was putting in front of his audience

This was the bigger turning point for me.

It wasn’t one interview. It wasn’t one guest. It was the pattern.

Take Steve Kirsch.

Kirsch has become one of the loudest promoters of sweeping claims about COVID vaccines. Yet Drew didn’t introduce him to his audience as some controversial figure whose assertions needed to be carefully tested.

In a recent post, Drew called him “our friend Steve Kirsch” and said Kirsch was coming on to help “pull apart the latest media panic cycle.”

Then Drew told his followers to follow Kirsch on X. That’s significant.

No, calling someone a friend doesn’t mean you endorse every sentence they’ve ever spoken. But neither is that a hostile interview.

It’s not 60 Minutes grilling somebody whose claims are under investigation.

You’re giving someone your platform, presenting him favorably and directing your audience toward him. And Drew’s own website went considerably further.

In June 2026, Ask Dr. Drew published an episode headlined:

“2 Million Disabled: Steve Kirsch’s AI Exposes Shocking mRNA Data ‘No Public Health Official Wants To Know.’”

The page presented Kirsch’s estimate that as many as two million Americans had been seriously disabled by COVID vaccination. That particular episode was guest-hosted by Kelly Victory rather than Drew himself, which is worth stating. But it was still produced and promoted under the Ask Dr. Drew banner. Dr. Drew | Official Website

That’s an extraordinary claim. Extraordinary claims should invite extraordinary scrutiny. Not just extraordinary headlines.

And Kirsch wasn’t alone

Then there was Mary Talley Bowden.

In September, Bowden posted about a “medical freedom” coalition pressuring President Trump and HHS Secretary Robert F. Kennedy Jr. to remove mRNA vaccines or face a political price.

Drew reposted it.

His contribution consisted of one word:

“Important.”

That’s not neutral.

It’s still possible for Drew to disagree with particular claims Bowden makes. But when you personally label a political and medical pressure campaign “Important,” you’ve made an editorial judgment about the thing you’re amplifying.

Bowden also appeared on Ask Dr. Drew, where the show’s own description said she was discussing a movement demanding removal of the mRNA vaccine platform. Dr. Drew | Official Website

Then there is Jessica Rose.

Drew has presented Rose to his followers by carefully laying out her scientific credentials - her mathematics degree, master’s in immunology, PhD in computational biology and postdoctoral work—and directing people to her website.

Again, credentials are relevant. There’s nothing wrong with mentioning them.

But look at the broader context.

In August 2026, the Ask Dr. Drew website promoted an episode around Rose’s claim that safety data for Moderna’s newly approved mRNA influenza vaccine showed “higher observed death.” The page prominently featured her interpretation of the trial and linked readers to her work. Dr. Drew | Official Website

So when Drew presents her to his audience, I don’t see that in isolation anymore.

I see the pattern.

Jeffrey Tucker made the pattern harder to miss

Jeffrey Tucker is another name that kept appearing.

Tucker isn’t primarily a medical scientist. He is the founder and president of the Brownstone Institute, which became one of the most prominent organizations attacking lockdowns, mandates and much of the conventional COVID response.

In July, Tucker posted an excerpt from Anthony Fauci’s diary involving a conversation with Xi Jinping.

Tucker’s framing was that Biden had received a lecture from Xi that “democracy is stupid, dictatorship is better.”

Drew reposted it with:

“WTH”

Then, the day before, Drew had promoted a program with Tucker and Peter McCullough reacting to Fauci’s Senate testimony and newly released diary material.

And this wasn’t incidental. Drew’s own site lists Tucker and McCullough together as guests for that July 29 program and directs readers to Brownstone’s coverage of the Fauci diaries. Dr. Drew | Official Website

Again, Tucker has every right to make his argument.

Drew has every right to interview him.

But eventually “I’m just talking to different people” stops explaining what you’re seeing.

• Because who gets invited is an editorial decision.

• Who gets brought back is an editorial decision.

• Who gets promoted is an editorial decision.

• Who gets described as “our friend” is an editorial decision.

• What gets called “Important” is an editorial decision.

• Where you send your audience for more information is an editorial decision.

Those choices don’t prove that every guest is wrong.

But they tell me something about what the platform now considers credible enough to amplify.

Then there’s Kelly Victory.

Victory isn’t just an occasional guest. She has become a recurring presence on Ask Dr. Drew and has even guest-hosted episodes. During the pandemic, she made claims such as “masks don’t work” and that asymptomatic spread of respiratory viruses “simply doesn’t occur”—claims that were contradicted by the available evidence and specifically fact-checked at the time. FactCheck.org

Yet Drew’s platform hasn’t treated her merely as a controversial voice to be challenged. In April 2026, his own website ran the headline:

“Dr. Kelly Victory Was RIGHT.”

The accompanying description said she had warned about the dangers of mRNA vaccines for years and had been unfairly criticized and censored.

Victory is also more than a guest in Drew’s orbit. She is Chief of Emergency & Disaster Medicine at The Wellness Company, and she is a coauthor with Drew, McCullough and Hulscher on the nattokinase paper discussed earlier in this article. Proteolytic Targets and Clinica…

So again, this isn’t simply Drew inviting someone on to hear another side. Victory is a recurring collaborator, colleague, and sometimes host on a platform that presents her as someone who was vindicated.

And then Ed Dowd.

Dowd, a former BlackRock portfolio manager, wrote “Cause Unknown”: The Epidemic of Sudden Deaths in 2021 & 2022, which strongly implies that COVID vaccination helped drive the rise in sudden deaths after 2020.

The problem is that the argument leans heavily on timing. Excess deaths increased after vaccines were introduced, but that does not establish that vaccines caused them. Dowd’s analysis did not directly compare mortality rates in vaccinated and unvaccinated people, and excess deaths during that period also reflected COVID itself and other causes. (factcheck.org)

There was also controversy over the book’s cover: the family of one young man pictured said he died from a congenital brain condition and was never contacted before his image was used. (apnews.com)

That’s the issue in a nutshell: a death occurring after vaccination is not the same thing as a death caused by vaccination. Yet Drew repeatedly gave Dowd a platform to make that case.

The Ecosystem Dr. Drew Chose to Platform

Again, the issue isn’t that Drew is forbidden from interviewing either of them.

The issue is the pattern.

Kirsch. Bowden. Rose. Tucker. Victory. Dowd. McCullough. Hulscher, etc.

Different backgrounds. Different claims. But the same names keep appearing around the same themes: vaccine injury, mRNA danger, hidden safety signals, institutional corruption and distrust of mainstream public health.

It wasn’t one guest that changed my mind about Dr. Drew. It was the network that slowly came into view.

Look at the headlines

This is probably the part that bothers me most. In January 2026, Ask Dr. Drew ran:

“Unsafe Blood: Baby Forms Massive Clot & Dies Suddenly, Parents Say mRNA-Vaxxed Donor Transfusion To Blame.”

To the site’s credit, the headline says these are the parents’ allegations. It does not establish that vaccinated donor blood caused the death.

But that’s still the framing being put in front of the audience. Dr. Drew | Official Website

In June:

“There’s No ‘Bigger Government Scandal’ Than Biden Admin’s mRNA Safety Coverup.” Dr. Drew | Official Website

In July:

“‘Blamed On The Flu’: Engineer Says Death Records Hide Vaccine Deaths.” Dr. Drew | Official Website

In August came the Jessica Rose episode about an mRNA flu vaccine and “Higher Observed Death.” Dr. Drew | Official Website

Read individually, Drew can reasonably say: That’s what the guest alleges. I’m having a conversation.

Fair enough. But read them together.

Then add Kirsch. Bowden. Rose. McCullough. Tucker. The Wellness Company.

At some point the guest list itself starts telling you something.

It wasn’t one guest that changed my mind about Dr. Drew. It was the network that slowly came into view.

Then I started checking the claims for myself

This is where my own thinking changed.

For a while, I was attracted to these voices for the same reason I think a lot of people were.

They questioned institutions.

Good.

• Institutions deserve questions.

• The CDC can screw up.

• The FDA can screw up.

• Drug companies can behave badly.

• Researchers have conflicts.

• Journalists can repeat things without understanding them.

• Public-health officials can overstate certainty.

I don’t want any of those people protected from criticism.

But there is a trap hiding inside skepticism.

You discover that establishment authorities can be wrong, and before you know it you’ve quietly decided that the people opposing the establishment must therefore be right.

That’s just another form of blind trust. So I started doing something very simple.

I started reading the papers. Not the tweet describing the paper. Not the screenshot of an abstract. Not the Substack headline.

The paper.

And I started asking:

• How many people were actually studied?

• What was the control group?

• Was this randomized?

• Was it observational?

• What was the absolute risk?

• Were there confounders?

• Does the study show causation, or only association?

• Was an adverse-event report being treated as proof that the vaccine caused the event?

• Was the conclusion decided before the analysis?

• Did other researchers replicate it?

• Did the headline say something much stronger than the paper?

And once I started doing that regularly, I saw claims fall apart that had sounded devastating on social media.

• “Associated with” became “caused by.”

• A VAERS report became a confirmed vaccine injury.

• A safety signal became proof of mass harm.

• A hypothesis became a conclusion.

• A preprint became “the study they don’t want you to see.”

• A dataset became whatever story somebody wanted it to tell.

That bothered me.

A lot.

And Then the Blockade

Peter McCullough blocked me. Nicolas Hulscher blocked me. Jessica Rose blocked me. And eventually Dr. Drew blocked me too.

Let me say this before somebody inevitably misunderstands the point:

Blocking me cannot prove anything about the science. A person can block me and still be right.

Maybe they thought I was irritating. Maybe they were tired of arguing. Maybe they simply didn’t want me in their mentions.

They’re allowed.

But for me, the blocking became symbolic because of what this whole movement had spent years telling people:

• Question authority.

• Don’t blindly trust experts.

• Do your own research.

• Demand transparency.

• Don’t censor debate.

• Ask questions.

So I did.

I questioned them. And that’s where something clicked for me.

That’s what bothered me. Not losing access to somebody’s X account.

The contradiction.

Dr. Drew bothered me more than the others

McCullough blocking me annoyed me. (Both of his X accounts by the way).

Hulscher blocking me annoyed me.

Rose blocking me annoyed me.

Dr. Drew was different.

Because I remembered him. I remembered the guy on television who seemed to cut through nonsense. I remembered Loveline. I remembered a doctor who had enough credibility that a major television network built a nightly program around him. I remembered the guy who screwed up COVID in 2020 and then publicly said:

I got it wrong.

That’s the Dr. Drew I respected.

His own biography today still says he is dedicated to providing honest, practical medical advice “based on scientific evidence.”

Good.

Then that’s the standard. Not loyalty. Not friendship. Not whether somebody is anti-establishment. Not whether somebody says things I already want to hear.

Evidence.

Maybe I was asking the wrong question

For a while I kept asking:

What happened to Dr. Drew?

Maybe the more important question is what happened to me. Because somewhere along the way I stopped looking for a doctor, politician, institution or movement to tell me which team had the truth.

I don’t need another authority to replace the old authority. I don’t have to choose between Pfizer and Peter McCullough.

Between the CDC and Steve Kirsch.

Between Anthony Fauci and Dr. Drew.

That’s the trap.

They can all be wrong. They can all be right about some things. They can all deserve scrutiny. And somebody being correct about one scandal doesn’t make his next claim true. Somebody being wrong once doesn’t make everything he says false either.

Each claim has to survive on its own evidence.

That’s where I’ve ended up.

• If Pfizer makes the claim: Show me.

• If the CDC makes the claim: Show me.

• If Fauci makes the claim: Show me.

• If McCullough makes the claim: Show me.

• If Hulscher publishes the study: I’ll read it.

• If Jessica Rose presents the analysis: I’ll check it.

• If Steve Kirsch says millions were injured: Show me how you got there.

And if Dr. Drew puts an extraordinary medical claim in front of an audience that has trusted him for decades:

Show me the evidence.

That’s not being pro-vaccine. It’s not being anti-vaccine. It’s not being pro-establishment. It’s not being anti-establishment.

It’s what I thought “question everything” was supposed to mean.

Turns out the final part is the hardest:

Question yourself.

Resources

• The Washington Post — “Dr. Drew’s coronavirus controversy put him back in the spotlight” (April 13, 2020). Documents Drew’s early “press-induced panic” and influenza comments, his subsequent apology, “I got it wrong,” and his description of Anthony Fauci as his “north star.” Read the Washington Post article The Washington Post

• CNN/HLN — “Dr. Drew Debuts on HLN March 28” (January 6, 2011). Contemporary announcement of Drew’s move into a mainstream nightly television role. Read the CNN/HLN announcement CNN Press Site

• The Wellness Company — “Dr. Drew Pinsky Joins the Chief Medical Board” (November 1, 2023). TWC’s own announcement of Drew joining its medical board and its stated goal of building a “parallel healthcare system.” Read the TWC announcement The Wellness Company

• The Wellness Company — Dr. Drew Medical Emergency Kit. TWC’s dedicated Dr. Drew sales page, including the DREW discount code and prescription-kit marketing involving ivermectin and other drugs. View the Dr. Drew TWC page The Wellness Company

• The Wellness Company — Medical Kit / product page. Shows TWC’s Medical Emergency Kit and its Ultimate Spike Detox / McCullough Protocol products. View the TWC Medical Kit page The Wellness Company

• Proteolytic Targets and Clinical Benefits of Oral Nattokinase: Spike Protein and Amyloid Degradation, Fibrinolysis, and Cardiovascular Risk Reduction.

• Ask Dr. Drew — Steve Kirsch, “2 Million Disabled” (June 2, 2026). The program page presenting Kirsch’s estimate that as many as two million Americans were seriously disabled by COVID vaccination. View the Steve Kirsch episode Dr. Drew | Official Website

• Ask Dr. Drew — Mary Talley Bowden (September 28, 2026). The episode page says Bowden discusses a “brewing revolt” demanding removal of the mRNA vaccine platform. View the Mary Talley Bowden episode Dr. Drew | Official Website

• Ask Dr. Drew — Jessica Rose, “Higher Observed Death” (August 14, 2026). Episode built around Rose’s interpretation of safety data for Moderna’s mRNA influenza vaccine. View the Jessica Rose episode Dr. Drew | Official Website

• Ask Dr. Drew — Peter McCullough and Jeffrey Tucker / Fauci hearing (July 29, 2026). Documents the program featuring McCullough and Tucker together discussing Fauci, gain-of-function research and Fauci’s diary. View the McCullough/Tucker episode Dr. Drew | Official Website

• Ask Dr. Drew — “Unsafe Blood” (January 27, 2026). Episode presenting parents’ allegation that blood from an mRNA-vaccinated donor was responsible for their baby’s fatal clot. View the “Unsafe Blood” episode Dr. Drew | Official Website

• Ask Dr. Drew — “mRNA Safety Coverup” (June 23, 2026). Program page built around Sen. Ron Johnson’s allegations concerning federal handling of vaccine-safety signals. View the “mRNA Safety Coverup” episode Dr. Drew | Official Website

• Ask Dr. Drew — “Death Records Hide Vaccine Deaths” (July 8, 2026). Episode featuring John Beaudoin’s claims about death certificates and vaccine injuries. View the death-records episode Dr. Drew | Official Website

• Dr. Drew — Official Biography. Drew’s own biography describes his medical credentials, Loveline, HLN career, Ask Dr. Drew, commitment to advice “based on scientific evidence,” and current Wellness Company role. Read Dr. Drew’s official biography

Dr. Jeff Barke calls himself “America’s REAL Doctor.” The capitalization is his.

He tells patients to question conventional medicine, demand better evidence, and follow the money. His newest posts, podcasts and product promotions raise a fair question:

His current website does not treat the phrase as a casual social-media nickname. Visitors are invited to “BOOK AMERICA’S REAL DOCTOR,” schedule telehealth appointments, buy his vaccine books, and shop in what the site calls the America’s Real Doctor Marketplace. Barke’s public brand now spans medicine, podcasts, books, supplements, wellness products, speaking engagements and a substantial online following. Rx for Liberty

There is nothing inherently suspicious about a physician selling a book or supplement. Nor does disagreeing with public-health authorities make someone wrong. Barke is also right about an important principle: credentials should never end an argument.

A physician’s claim should stand or fall on the evidence behind it. That principle works in both directions.

I first looked closely at Barke because of his vaccine claims and his books Unavoidably Unsafe. Since then, his public platform has expanded. What he is saying in late 2026 gives us a better test than his old COVID arguments alone.

Some of his recent claims are reasonable. Others run well ahead of the evidence. And increasingly, some of those claims sit immediately beside products that Barke sells himself or recommends through personalized commercial links.

I am not going to tell you what label to put on that. I am going to show you what he says, what the evidence says, and where the two diverge.

A September podcast provides a surprisingly useful test

On September 28, Barke devoted an episode of his Informed Dissent podcast to hypertension.

Much of the beginning is sensible.

Barke says that one elevated blood-pressure reading should be confirmed rather than treated reflexively. He discusses sleep, alcohol, weight, stress, exercise and medical conditions that can contribute to hypertension. He explicitly says medication can be essential, tells listeners not to discontinue prescribed blood-pressure medication without consulting their clinician and argues that patients deserve more than a prescription handed across a desk. America Out Loud News

Those aren’t fringe ideas.

The current 2025 ACC/AHA hypertension guideline strongly emphasizes accurate measurement, home monitoring, lifestyle intervention and overall cardiovascular risk. In lower-risk adults with stage 1 hypertension—130–139/80–89—the guideline recommends lifestyle intervention first and medication if pressure remains elevated after three to six months. Treatment is therefore considerably more individualized than the caricature of a doctor seeing 130/83 and immediately reaching for a prescription pad. professional.heart.org

But the podcast then goes much further.

Barke tells listeners that he regards hypertension not as a disease but as a symptom, and says that when he sees newly hypertensive patients whose pressure is not immediately dangerous, he routinely sends them to a chiropractor several times a week for months. He offers a theory in which spinal alignment can effectively “dim” nerve activity to organs and blood vessels, contributing to high blood pressure. He then extends the idea to other medical conditions, citing stories of women becoming pregnant after chiropractic treatment and speculating about effects on thyroid function. America Out Loud News

This is precisely the kind of claim for which Barke normally demands rigorous evidence.

So what does that evidence show?

A randomized clinical trial involving 140 people with high-normal blood pressure or stage 1 hypertension found that adding chiropractic spinal manipulation to a dietary program provided no additional blood-pressure benefit over diet alone. A systematic review subsequently concluded that the low-bias evidence did not support spinal manipulation as a treatment for hypertension. A later sham-controlled trial likewise failed to show a reduction in blood pressure from upper-cervical manipulation. PubMed

An international systematic review examining spinal manipulation for disorders outside the musculoskeletal system went further: acceptable- or high-quality randomized trials did not show benefit for conditions including hypertension, asthma, migraine and infantile colic. PubMed

The infertility claim is thinner still. A review of chiropractic treatment for female infertility found only ten published articles—all case reports, covering just eleven women. Its authors concluded that chiropractic treatment for infertility should be approached cautiously because a robust body of evidence simply did not exist. PubMed

Pregnancy following chiropractic treatment proves that pregnancy followed chiropractic treatment.

It does not prove the adjustment caused the pregnancy.

That is exactly the distinction between after this and because of this that Barke asks people to make when evaluating conventional medicine.

And then the podcast arrives at Barke’s own product

There is another reason the hypertension episode deserves attention.

After discussing magnesium, beetroot, ashwagandha, L-theanine and omega-3 fatty acids, Barke explains that he has recommended those supplements individually for years. He then tells listeners that he combined them into a product of his own: RxForLiberty Blood Pressure Support.

He specifically frames the ingredients as useful for people who want to avoid blood-pressure medication or attempt to get off prescribed medication, while also cautioning listeners not to stop medication on their own and describing his supplement as supportive rather than a treatment.

That qualification matters. We shouldn’t omit it.

But so does what happens next.

The supplement is sold directly through Barke’s store. The current product page calls it a “clinical-strength” formula and markets a “Root-Cause Approach.” It says the blend supports healthy blood pressure, arterial elasticity, circulation, cortisol balance and cardiovascular health. The same page displays testimonials from purchasers reporting lower blood pressure, including one customer saying a family member wanted to get off medication and had begun taking it every other day. The site separately warns people taking prescription medication to consult their healthcare provider. Rx for Liberty

The individual ingredients are not inherently implausible.

A 2025 meta-analysis of 38 randomized trials found that magnesium supplementation produced a modest average blood-pressure reduction—about 2.8 mm Hg systolic and 2.1 mm Hg diastolic—with greater effects in certain hypertensive or magnesium-deficient populations and considerable heterogeneity among studies. A 2024 meta-analysis found that nitrate-rich beetroot juice reduced office systolic blood pressure by about 5 mm Hg in hypertensive patients, although certainty was low and effects on diastolic and 24-hour pressure were not significant. PubMed Central (PMC)

So the scientifically appropriate conclusion isn’t, “Supplements don’t work.”

It is more specific:

Evidence that some ingredients can affect blood pressure is not the same thing as evidence that Barke’s particular five-ingredient finished product produces the benefits implied by its marketing.

I could not locate a published randomized clinical trial of the finished RxForLiberty formulation itself.

Perhaps one exists and I missed it. If Barke can provide product-specific clinical evidence, it belongs in this article.

That seems a particularly fair request from a physician whose public message repeatedly insists that pharmaceutical companies should be required to test the actual products they sell.

The peptide experiment

Another recent Barke post raises the same question even more sharply.

Barke told followers that while recovering from an elbow injury, he had been experimenting with injectable BPC-157 and TB-500. He then said he had begun taking an oral product called Regenerate from The Wellness Company, describing BPC-157 in terms of tissue repair, TB-500 in terms of recovery, and KPV as supporting reduced inflammation.

The commercial product contains BPC-157, KPV and TB-500 and currently sells for $299.99 as a one-time purchase. Its manufacturer describes it as accelerating recovery, improving tissue repair, and promoting cellular regeneration.

Now compare that confidence with what FDA reported when it examined these substances in 2026.

For BPC-157, FDA said it had identified little human safety information for the proposed routes of use and lacked enough information to know whether the substance could cause harm. Its detailed review found limited human studies and no human pharmacokinetic data after oral, subcutaneous, nasal or transdermal administration. FDA also identified safety signals in animal studies that it said required further investigation. U.S. Food and Drug Administration

For KPV, FDA said it had found no human exposure data for drug products containing it.

For TB-500, FDA likewise reported that it could identify no clinical studies or human exposure data using TB-500 by any route, leaving its human safety risks unknown. U.S. Food and Drug Administration

That doesn’t establish that these peptides are dangerous.

It establishes something narrower and more important: the human clinical evidence is extraordinarily limited.

FDA convened an advisory committee in July specifically to consider BPC-157, KPV and TB-500-related substances for pharmacy compounding. That was a regulatory review of uncertain compounds—not an FDA finding that the recovery claims used to market products containing them had been clinically established. U.S. Food and Drug Administration

This is where Barke’s evidentiary philosophy becomes difficult to reconcile.

If insufficient long-term randomized human evidence is a reason to warn strongly about conventional medical products, why does extremely limited human evidence support recommending an oral peptide mixture for tissue repair and recovery?

The molecule doesn’t know whether it came from Pfizer or a wellness company.

The evidentiary standard shouldn’t know either.

Parasites, prescriptions and the discount code

Another recent Barke post turns to parasites.

He tells followers that conventional and holistic medicine approach parasites differently and discusses protocols involving ivermectin and mebendazole. The post then directs readers to The Wellness Company using RXFORLIBERTY for a discount.

Barke’s personalized Wellness Company page currently sells an ivermectin plus mebendazole “Dual-Action Parasite Cleanse” for $599.99. The same page labels a selection of commercial products as “Dr Jeff Barke’s Picks.”

Again, neither medication is imaginary medicine. Ivermectin and mebendazole are legitimate antiparasitic drugs when used for appropriate infections.

The question is what justifies moving from:

These medicines treat particular parasitic infections.

to a broadly marketed parasite cleanse.

Diagnosis matters. The suspected organism matters. Travel and exposure history matter. Dosing matters. Drug interactions matter. In some cases, treating the wrong suspected parasite without appropriate evaluation can create risks of its own.

And the commercial relationship matters.

Barke’s own website has a page explicitly titled “Affiliates and Partners.” It promotes The Wellness Company and numerous other health products, often using RXFORLIBERTY discount codes. Current offerings include nutraceuticals, parasite products, red-light devices, EMF shields, colostrum, detox products, air purification, organ supplements, magnesium products and water filtration. Barke’s site says he hand-selects these products for their clinical integrity, purity and effectiveness.

His legal notice separately acknowledges links to affiliate partners.

What I cannot establish from public records is precisely how Barke is compensated by every partner or how much money any individual sale generates for him.

So I won’t guess.

The documented point is enough: medical recommendations and commercial referral links occupy the same platform.

That isn’t proof that the recommendation is wrong. But it is exactly the sort of financial relationship readers should know about when evaluating medical advice.

One recent Barke headline has a legitimate point buried inside it

A fair audit cannot operate on the assumption that everything Barke says must be false.

A recent example is useful. Barke posted the headline:

“New mRNA COVID Vaccine: Tested Only on Mice, Not Humans.”

The statement sounds outrageous. But there is an important kernel of truth.

For the exact 2026–27 XFG-adapted formulation, regulators did not require a brand-new large human efficacy trial before updating the strain.

FDA says its strain-selection process incorporated human immunogenicity data from previous COVID vaccine formulations, existing vaccine-effectiveness evidence and animal immunogenicity data for newly updated candidate formulations. Pfizer likewise says approval of its XFG-adapted Comirnaty relied on the accumulated clinical, nonclinical and real-world evidence for Comirnaty along with manufacturing, quality and nonclinical data for the new XFG formulation.

So if Barke’s point is:

The exact XFG update did not undergo a fresh large preapproval efficacy trial in humans,

that is a legitimate regulatory criticism to discuss.

But “tested only on mice” can also leave readers with a very different impression: that regulators approved a completely new vaccine technology with nothing but a few animal experiments behind it.

That isn’t what happened.

Years of human clinical and real-world experience with the underlying vaccine platform remained part of the regulatory evidence base. Regulators then treated the strain update more like an update of an established platform than an entirely new biological product.

One can disagree with that regulatory approach. But the complete argument requires explaining it. This is one case where Barke is raising a genuine evidentiary question even if the headline compresses away important context.

“Turbo cancers”: a hypothesis is not yet a diagnosis

Barke recently posted another summary of his COVID-era positions, saying mRNA vaccination is now showing serious safety concerns, including what he describes as “turbo cancers.”

Notice his wording: what I describe as.

That qualification is useful because “turbo cancer” does not describe a defined clinical diagnosis with an established causal mechanism. It is a label being applied to claims that unusually aggressive or rapidly appearing cancers are occurring after COVID vaccination.

There is research worth discussing here rather than dismissing the subject with a sneer.

In 2026, Oncotarget published a case report involving a woman who developed acute lymphoblastic leukemia and lymphoblastic lymphoma after mRNA vaccination. Its authors also reviewed other reports and proposed potential biological mechanisms. Case reports are legitimate tools for generating safety hypotheses. Oncotarget

A large South Korean observational study published in 2025 also reported statistical associations between vaccination and several cancers during one year of follow-up. That study is real and shouldn’t simply be pretended away. PubMed

But association isn’t the end of the analysis.

A 2026 methodological commentary identified substantial problems with the Korean study, including different index dates for vaccinated and unvaccinated participants, residual confounding, possible surveillance bias, multiple comparisons without correction and the difficulty of attributing solid tumors appearing within one year to a new carcinogenic exposure. The authors concluded that the study could not establish causation and rated its evidentiary value as low.

The National Cancer Institute’s current position remains that there is no evidence COVID-19 vaccination causes cancer, recurrence or cancer progression. Cancer.gov

That does not mean researchers should stop looking for rare or unexpected effects.

It means that case reports, biological hypotheses and a disputed observational association do not yet establish a new syndrome in which vaccination causes rapidly accelerating cancers.

Barke is entitled to say the possibility deserves investigation.

Calling the phenomenon an emerging vaccine safety problem requires considerably stronger evidence.

The older claims matter because they show how the method works

The new material isn’t occurring in isolation.

Barke’s earlier article on Enflonsia remains one of the cleanest examples of the problem.

He presented Enflonsia as a new infant RSV vaccine. It isn’t a vaccine. It is clesrovimab, a long-acting monoclonal antibody manufactured by Merck. FDA approved it to protect infants entering their first RSV season.

More importantly, its pivotal infant trial included 3,614 infants in a randomized, double-blind, saline-placebo-controlled study. That is almost exactly the kind of trial design Barke routinely argues vaccine products should undergo.

The larger lesson isn’t merely that Barke got a product category wrong.

It is that the criticism remained highly confident even when the underlying facts did not fit the argument.

The same problem appears in discussions of Vaxelis.

Barke has emphasized that six infants died during trials of the combination childhood vaccine.

That number is true.

But the official prescribing information reports six deaths among Vaxelis recipients and one among active-comparator recipients; none of the deaths was assessed as vaccine-related. Causes included sepsis, SIDS, asphyxia, hydrocephalus, and an unknown cause, occurring from two days to eleven months after vaccination. Nonfatal serious adverse events within 30 days were 2.0% in the Vaxelis group and 2.2% in the comparison group. sotwe

Saying six babies died is not fabricated.

It is incomplete in a way that radically changes what a reader is likely to conclude.

• Denominators matter.

• Comparators matter.

• Timing matters.

• Causes matter.

Barke would rightly demand those things from a pharmaceutical advertisement.

They matter just as much in a warning about the pharmaceutical product.

And then there is Unavoidably Unsafe

The title of Barke’s vaccine books is itself built around an important legal phrase.

In excerpts from Unavoidably Unsafe for Adults that I reviewed, the authors connect the National Childhood Vaccine Injury Act and the Supreme Court’s decision in Bruesewitz v. Wyeth to the concept of vaccines as “unavoidably unsafe products.”

There really is an important liability story here.

Congress created a compensation system and gave manufacturers substantial protection from traditional design-defect litigation. Someone can reasonably argue that those protections are too broad.

But Bruesewitz did not constitute a scientific finding that vaccines as a class were proved unsafe.

Justice Antonin Scalia’s majority opinion specifically distinguished statutory language about vaccine side effects that were “unavoidable” from the Restatement of Torts category of “unavoidably unsafe products.” The majority said there was no reason to conclude that Congress had simply incorporated that product-liability category into the statute. Justice Sotomayor’s dissent interpreted the legal history differently, so there is a genuine legal debate—but not a Supreme Court scientific determination that vaccine risks outweigh their benefits.

The distinction matters because ordinary readers hear unavoidably unsafe very differently from lawyers arguing about design-defect preemption.

Another statement in the book is easier to test.

The excerpts claim, in substance, that MMR is the only vaccine that has been evaluated in relation to autism.

That is incorrect.

A 2014 meta-analysis included cohort data on more than 1.25 million children and separately examined vaccination generally, MMR, thimerosal, and mercury exposure. A 2013 study examined cumulative exposure to vaccine antigens during the first two years of life rather than MMR alone. Neither supported an association with autism.

Barke can reasonably say that scientists haven’t randomized children to every possible version of the complete vaccine schedule and followed them indefinitely for autism.

They haven’t.

That isn’t the same claim as saying researchers looked only at MMR.

This isn’t an argument for trusting institutions blindly

There is a temptation in articles like this to replace one oversimplification with another.

Don’t.

Drug companies have financial conflicts.

Government agencies make mistakes.

Medical journals publish bad papers.

Doctors sometimes follow habits long after the evidence has changed.

Postmarketing surveillance occasionally finds risks that preapproval trials were too small to detect.

And regulators do sometimes accept surrogate endpoints, immunobridging or extrapolation rather than demanding another gigantic efficacy trial.

Those are not conspiracy theories. They’re features, and sometimes weaknesses, of the real medical system.

Barke is also right that absolute risks are often more informative than dramatic relative percentages. He is right that a newly updated vaccine strain can reach the public without a fresh tens-of-thousands-person efficacy trial of that exact formulation. He is right that lifestyle measures are important in hypertension. He is right that patients should be told what we know and what we don’t know.

The question is whether those principles are applied consistently.

If a vaccine is criticized because the exact product hasn’t undergone enough human testing, an oral peptide product with almost no relevant human evidence should face at least as much skepticism.

If anecdotal post-vaccination events don’t prove vaccine injury, pregnancies after chiropractic manipulation don’t prove chiropractic treatment restores fertility.

If conflicts of interest matter at Pfizer, they matter when a physician recommends a $599 parasite product through his own branded discount code.

If regulators shouldn’t be allowed to hide behind “trust us,” neither should America’s REAL Doctor.

This review is not about guessing at Jeff Barke’s motives. It is about whether his public medical claims hold up when checked against the evidence.

Some do. Some raise legitimate questions about regulation, trial design, risk communication and medical conflicts of interest.

Others do not survive the same scrutiny. In several cases, important context is omitted, preliminary evidence is presented with more confidence than it warrants, or the evidentiary standard applied to conventional medicine is noticeably stricter than the standard applied to products and treatments Barke recommends himself.

Those patterns are documented here. Readers can decide what they mean.

Barke currently operates a broad commercial health platform alongside his medical commentary. His own store sells his blood-pressure supplement. His public partner directory contains numerous branded discount codes, and his legal notice acknowledges affiliate partners. The Wellness Company hosts a personalized Barke storefront offering products including peptide formulations, “spike detox” supplements, and an ivermectin-mebendazole parasite cleanse.

His September hypertension podcast contains prudent advice about lifestyle and medication, but also recommends repeated chiropractic treatment for hypertension despite weak or negative higher-quality evidence, invokes anecdotal chiropractic infertility stories and then promotes his own cardiovascular supplement.

His recent peptide promotion uses confident recovery language for compounds for which FDA reported extremely limited or absent human exposure data.

His “mice only” criticism of the newest COVID formulation contains a legitimate point about the absence of a fresh human efficacy trial for the exact strain update, but omits the extensive human evidence regulators relied on from previous formulations of the same platform.

And his use of “turbo cancers” moves ahead of what the current cancer evidence can establish. Hypotheses and observational signals exist; causation has not been demonstrated.

Those are factual observations.

What they add up to is for the reader to decide.

America’s “REAL Doctor”

I keep coming back to the branding because I do object to it.

Calling yourself “America’s REAL Doctor” is not a neutral slogan. It implies that other doctors are somehow less real - less independent, less honest, less willing to tell patients what they supposedly need to hear.

That kind of branding sets a standard.

What makes a doctor REAL?

Surely it cannot simply mean being more skeptical than other doctors.

Real skepticism is not a direction. It does not automatically point away from vaccines, pharmaceuticals, hospitals or regulatory agencies.

It points wherever the evidence requires it to point.

• Sometimes that means challenging Pfizer.

• Sometimes it means challenging the FDA.

• Sometimes it means challenging a paper whose conclusions fit your own beliefs.

• Sometimes it means challenging a supplement company.

• Sometimes it means challenging a chiropractor.

And sometimes it means challenging the doctor you trust precisely because he told you not to trust everybody else.

Barke tells people to question authority.

They should.

He tells them to read studies instead of relying on headlines.

They should do that too.

He tells them to look for financial conflicts and ask who benefits.

Again, fair question.

But those rules do not stop at the boundary of conventional medicine.

• They apply to the peptide bottle.

• They apply to the parasite cleanse.

• They apply to the blood-pressure supplement formulated by the doctor himself.

• They apply to the affiliate relationship.

• They apply to the discount code.

• They apply to the claims that sound reassuring as well as the claims that sound frightening.

And they apply to America’s REAL Doctor.

That is the problem with building your brand around being the doctor who sees through everybody else.

Eventually, the same questions come back to you.

• What is the evidence?

• How strong is it?

• What is missing?

• What would change your mind?

• And who stands to benefit?

No label is necessary.

Follow the evidence far enough, and readers can decide the rest.

Resources and Further Reading

• Jeff Barke’s own platform and commercial relationships
  RxForLiberty’s current site documents the America’s Real Doctor Marketplace, Barke’s Blood Pressure Support supplement, and his “Affiliates and Partners” page. That partner page promotes The Wellness Company and numerous wellness products, often using the code RXFORLIBERTY, while his legal notice acknowledges links to affiliate partners. Rx for Liberty
  RxForLiberty Marketplace
  Blood Pressure Support
  Affiliates and Partners

• The Wellness Company products promoted through Barke’s page
  Barke’s personalized Wellness Company storefront currently lists products including “Ultimate Spike Detox” and an ivermectin-plus-mebendazole “Dual-Action Parasite Cleanse” under “Dr Jeff Barke’s Picks.” The Wellness Company
  Dr. Jeff Barke’s Wellness Company page

• Chiropractic treatment for high blood pressure
  A randomized clinical trial of 140 adults found that adding chiropractic spinal manipulation to dietary treatment provided no additional blood-pressure reduction compared with diet alone. A later sham-controlled trial likewise found no significant blood-pressure benefit from upper-cervical manipulation. PubMed
  Treatment of Hypertension with Alternative Therapies trial
  Upper-cervical manipulation sham-controlled trial

• Chiropractic and infertility
  A review of the literature found only ten articles, all case reports involving just eleven women. The authors concluded that chiropractic treatment for infertility should be approached cautiously because a robust evidence base does not exist. PubMed
  Scoping review of chiropractic management of infertility

• BPC-157, KPV and TB-500
  FDA convened its Pharmacy Compounding Advisory Committee in July 2026 specifically to review these substances. Their consideration for compounding should not be confused with FDA approval of claims that they accelerate healing, repair tissue or improve recovery. U.S. Food and Drug Administration
  FDA 2026 peptide-compounding review

• The 2026–27 COVID vaccine update and the “tested only on mice” claim
  FDA explains that the 2026–27 XFG strain selection relied on a broader evidence base that included human immunogenicity and effectiveness data from existing COVID vaccines, human data from previous formulations, antigenic surveillance, and animal immunogenicity data for newly updated candidate vaccines. That supports a legitimate discussion about the absence of a fresh large human efficacy trial for the exact update, but not the broader impression that regulators relied only on mice. U.S. Food and Drug Administration
  FDA: COVID-19 Vaccines, 2026–2027 Formula

• COVID vaccination and cancer
  The National Cancer Institute states that current evidence does not show COVID-19 vaccines causing cancer, cancer recurrence or disease progression. That does not prevent researchers from investigating case reports or new hypotheses, but it is important context when evaluating claims about “turbo cancers.” Cancer.gov
  National Cancer Institute: COVID-19 Vaccines and People with Cancer

• Current FDA information on the 2026–27 vaccine formulations
  FDA maintains current product pages for the approved 2026–27 COVID vaccines, including Moderna’s mRNA products and Novavax’s protein-based vaccine, with indications, formulation details and regulatory documents. U.S. Food and Drug Administration
  FDA: MNEXSPIKE
  FDA: SPIKEVAX
  FDA: NUVAXOVID

When Brian Hooker and Children’s Health Defense team up, you know you're going to get your fact-checking money's worth.

We read the actual 23-page paper behind the CHD headline. Not just the headline. Not somebody’s thread about the headline. Not a screenshot of a chart. The paper.

Comparative Health Outcomes of Vaccinated versus Unvaccinated Children: A Retrospective Cohort Analysis

The dataset may contain signals worth investigating, but the tables, definitions, statistics, and even CHD’s own reporting raise questions that need answers before anyone claims vaccines caused these outcomes.

And it includes findings that deserve a closer look. I’m not interested in dismissing research because of who funded it, who wrote it, or because the conclusions make one side of a vaccine argument uncomfortable.

If the data show something important, investigate it.

But the reverse is also true: being willing to challenge the medical establishment does not mean we stop challenging papers that tell us something we want to hear.

That’s where this study runs into trouble.

Children’s Health Defense reported that children vaccinated before 13 months had substantially higher odds of several conditions, including recurrent respiratory disease, ear infections, rhinitis and “strep throat.” The underlying preprint analyzed records from 6,239 children at eight pediatric practices.[1][2]

Those associations exist in the authors’ analysis. The study does not establish that vaccination caused them.

And once I started checking the actual tables against both the paper and CHD’s summary, I found problems that go considerably beyond the usual “correlation isn’t causation” objection.

First problem: these are unadjusted comparisons

The paper uses basic 2×2 contingency tables and reports odds ratios. There is no multivariable model controlling for major differences between the vaccinated and unvaccinated families.

The authors acknowledge that they had no usable information on race or ethnicity, socioeconomic status, gestational age, birth weight, breastfeeding, mode of delivery, family medical history, environmental exposures, and other potentially important variables. They explicitly describe their reported odds ratios as unadjusted observational associations.

Later, they acknowledge the problem even more plainly: without multivariable modeling, they cannot disentangle vaccination from confounding variables.

That’s not a minor limitation.

Vaccination isn’t randomly assigned in this population. Parents who decline vaccines can differ from parents who vaccinate in education, income, breastfeeding, daycare use, medical-care utilization, lifestyle and many other ways that can also affect childhood diagnoses.

So when two groups show different rates of disease, you first have to ask whether you’re measuring the effect of vaccination—or differences between the families.

This study cannot adequately answer that question.

Then I found something harder to explain: Table 4 doesn’t match the cohort

According to the paper, there were:

2,414 children vaccinated before 13 months

and

3,825 children in the comparison group.

That’s 6,239 children total.

Now look at the respiratory-disease row in Table 4.

It lists:

1,369 vaccinated children with respiratory disease

and

1,324 vaccinated children without respiratory disease.

Add those together.

That’s 2,693 vaccinated children.

But the paper says there were only 2,414 vaccinated children in the cohort.

Where did the additional 279 come from?

This isn’t limited to respiratory disease. Several rows in Table 4 total 2,693 on the vaccinated side rather than the declared 2,414.

And here’s where it gets even more interesting.

When I add the male and female respiratory-disease counts from Tables 5 and 6, they do reconcile with the stated cohort:

523 + 484 = 1,007 vaccinated respiratory cases

720 + 687 = 1,407 vaccinated noncases

Total: 2,414

The unvaccinated sex-specific counts similarly total 3,825.

But using those sex-table numbers produces an odds ratio of about 1.19, not the 1.6 reported in Table 4.

That’s not a philosophical dispute about vaccines.

That’s an internal numerical discrepancy.

There may be an explanation. There may be an error in the table. There may be a coding or labeling issue.

But until the authors explain it, one of the paper’s headline findings cannot simply be accepted at face value.

The healthcare-use problem may be even bigger

Several of the study’s important outcomes weren’t defined as one diagnosis.

To qualify as recurrent ear infection, respiratory disease, skin rash, strep throat or fever, children needed at least three separate diagnoses within 12 months.

Now ask the obvious question:

Did vaccinated and unvaccinated children visit doctors equally often?

The study doesn’t know.

The authors acknowledge that they did not have total healthcare visit information and that vaccinated and unvaccinated families may interact with the medical system differently.

That matters enormously when your outcome literally requires repeated medical diagnoses. A child who visits the pediatrician frequently has more opportunities to receive vaccines. That same child also has more opportunities to have an earache coded, a sore throat swabbed, a cough entered into the medical record, or a respiratory infection diagnosed.

A child whose parents handle minor illnesses at home may experience the same number of illnesses without accumulating the required diagnostic codes.

This isn’t merely theoretical. A large U.S. cohort study involving more than 323,000 children found that children who were undervaccinated—particularly those undervaccinated because of parental choice—had different healthcare-utilization patterns, including fewer outpatient visits than age-appropriately vaccinated children.[3]

That makes healthcare utilization an especially important confounder in this particular study.

And it wasn’t controlled for.

The advertised “dose-response” isn’t nearly as clean as the headline suggests

This is another part of the paper worth reading for yourself.

For the paper’s “strep” endpoint, the odds ratios by dose category were:

1.5 → 2.2 → 3.4 → 2.7

So the association rises through the first three groups and then falls in the highest-dose group.

The paper nevertheless describes this as a “clear, monotonic dose-response gradient.” But a monotonic increase doesn’t decrease in the final category.

Respiratory disease is stranger:

1.2 → 1.6 → 2.0 → 0.50

At 10+ vaccine administrations, the association reverses completely and becomes strongly protective.

Developmental delay goes:

0.82 → 0.66 → 0.74 → 1.4

So the 4–5 and 6–9 groups actually showed statistically significant lower odds of developmental delay before the result reversed in the 10+ group.

Those patterns may have explanations.

But here’s the problem with the interpretation: when vaccination appears associated with lower disease, the discussion proposes healthy-vaccinee effects, selection effects or other bias. When vaccination appears associated with higher disease, the discussion repeatedly explores biological vaccine harm.

Bias doesn’t get to operate only when the data point in the inconvenient direction.

The same skepticism has to apply both ways.

CHD gets the developmental-delay result wrong

This isn’t an interpretation issue. It’s a numerical reporting error.

CHD’s article says the analysis found 40% higher odds of developmental delay among children receiving six to nine vaccine doses.

Vaccinated Kids Have Significantly Higher Odds of Developing Multiple Chronic Health Conditions

That’s not what Table 7 reports.

The 6–9 group had an OR of 0.74, meaning lower - not higher - odds in that analysis.

The 1.4 OR belongs to the 10+ vaccine group.

That’s a pretty significant distinction when you’re writing an article claiming a dose-response relationship.

CHD appears to put the rhinitis result in the wrong group too

CHD says rhinitis increased from 30% higher odds in the 1–3 group to 130% higher odds in the 6–9 group.

Again, that’s not what Table 7 says.

The 6–9 category has an OR of 1.7, or 70% higher odds.

The OR of 2.3, corresponding to 130% higher odds, belongs to the 10+ category.

Two different outcomes. Same type of reporting mistake.

Odds are not the same thing as risk

CHD also quotes Brian Hooker saying children were “60% more likely” to suffer recurrent respiratory disease and had “double the risk” of recurrent ear and strep infections.

But the paper calculated odds ratios, not relative risks.

Those are not interchangeable when an outcome is common.[4] BMJ has specifically cautioned that interpreting odds ratios as relative risks can overstate the magnitude of an association.

Using Table 4 exactly as printed, even though its denominators themselves need explanation, the respiratory figures correspond to roughly 50.8% versus 39.6%. That’s about a 28% relative increase in observed risk, not 60%.

The 60% figure comes from treating the OR of 1.6 as though it means 60% higher probability.

It doesn’t.

Again, this doesn’t make the association disappear. It means the result should be communicated using the statistic that was actually calculated.

There’s also a problem with who counts as “unvaccinated”

CHD tells readers the study contained 3,825 unvaccinated children.

But the actual paper says 469 children received their first vaccine after 13 months, and that this late-vaccinated subgroup was retained in the primary analysis as “unvaccinated before 13 months.”

Then, confusingly, the next section defines “unvaccinated” as receiving zero vaccines at any point during the observation period. Those statements don’t fit cleanly together.

Were the 469 late-vaccinated children in the reference group or not?

If they were, then this isn’t literally a comparison between vaccinated children and children who remained completely unvaccinated.

If they weren’t, the reported group totals and methods need clarification.

Either way, the paper should make this unambiguous.

What exactly is “strep throat” in this paper?

This may be one of the most important details buried in the tables.

CHD repeatedly describes the strongest association as recurrent strep throat.

But look at the paper’s actual ICD definition.

The “strep throat” outcome includes J02.0, which really is streptococcal pharyngitis.

But it also includes broad wildcard categories such as J03, J04, J05, J06, J36, J39, B27, A54 and A56 - along with a long list of ICD-9 categories.

That’s a problem if the result is being communicated specifically as recurrent streptococcal infection.

Official ICD-10-CM documentation identifies J06 as acute upper-respiratory infections of multiple or unspecified sites—and specifically distinguishes it from streptococcal pharyngitis, J02.0. J05 includes croup and epiglottitis. A54 is the gonococcal-infection family, while A56 covers chlamydial infections.[5]

In other words, as the paper’s coding table is written, this isn’t a clean laboratory-confirmed “recurrent strep throat” endpoint.

It’s a much broader diagnostic composite.

If the strongest headline result is “2.3 times the odds of recurrent strep,” then I want the strep outcome to actually measure strep.

There’s another curious coding choice in “respiratory disease”

The paper defines respiratory disease using J*, R09*, and R19* codes.

J codes include respiratory diseases. R09 includes symptoms involving the circulatory and respiratory systems.

But R19 is not respiratory.

Official ICD documentation classifies R19 as “other symptoms and signs involving the digestive system and abdomen”—including such things as abdominal swelling, bowel sounds, changes in bowel habits and diarrhea.[6]

Maybe R19 is a typographical error in the manuscript. Maybe it was actually included in the analysis. Those possibilities have very different implications. The authors need to clarify which one it is.

Even the list of nine conditions changes

The Methods section and Table 1 include nausea and vomiting among the nine outcomes.

But Table 4 reports sinusitis instead.

There is no sinusitis definition in Table 1.

Again, that may be a table-construction error rather than a problem with the underlying data.

But by this point we’re accumulating enough table and definition discrepancies that asking for a corrected version isn’t nitpicking. It’s basic quality control.

Multiple comparisons matter too

The authors analyzed nine conditions, then broke results down by sex, then analyzed four vaccine-dose categories.

They explicitly state that no correction for multiple comparisons was applied.

The more statistical tests you run, the greater the chance that some will cross a conventional p<0.05 threshold purely by chance.

That doesn’t erase very small p-values or prove the results are false.

It does mean isolated subgroup findings—particularly ones emphasized after looking across dozens of comparisons—need more caution than CHD’s article gives them.

Then there’s the most basic causal question: which came first?

The paper counts vaccine exposure through 13 months of age.

But the authors explicitly say they did not assess temporal relationships between particular vaccinations and diagnoses.

Some first diagnoses occurred very early. The paper reports respiratory diagnoses beginning as early as four days of age and its “strep” diagnoses as early as seven days.

That creates a basic causal problem.

If illness leads to more pediatric visits, and pediatric visits lead to more opportunities both for diagnoses and vaccinations, you can generate an association between vaccine count and illness without the vaccination being the cause of the illness.

A proper causal analysis needs to establish that exposure occurred before the outcome being attributed to it.

This study doesn’t consistently do that.

One CHD claim goes well beyond what this study tested

CHD also quotes Hooker saying the current vaccine schedule exposes infants to “toxic levels of aluminum.”

Whatever one thinks about aluminum-adjuvant research, this particular study did not measure aluminum exposure, blood or tissue aluminum, specific vaccines, or aluminum-related outcomes.

So that claim cannot be derived from this dataset.

The FDA’s current description is different: it states that aluminum-containing vaccine adjuvants have a demonstrated safety profile extending over many decades, with severe local reactions uncommon.[7]

That doesn’t make aluminum research off-limits.

It means a claim of “toxic levels” requires evidence specific to dose, exposure and clinical outcome—not an observational vaccinated-versus-unvaccinated table that never measured aluminum.

What does this study actually tell us?

What I think is fair to say: this dataset contains some sizable associations between vaccination status and several coded diagnoses. They deserve independent analysis and replication. The large number of children with limited or no early vaccination makes the dataset potentially useful.

What I do not think is fair to say is that this paper has demonstrated that vaccines caused vaccinated children to become less healthy. The analysis is unadjusted, healthcare utilization isn’t controlled, temporality isn’t adequately established, some outcome definitions are unusually broad, and several tables or descriptions do not reconcile internally.

What would move this forward: publish a corrected explanation of the Table 4 denominators; clearly identify what happened to the 469 late-vaccinated children; provide visit counts; analyze outcomes only after exposure; stratify by practice; use validated disease definitions; control for major confounders; identify actual vaccine types rather than treating vaccination codes as interchangeable “dose”; correct the outcome tables; and provide enough de-identified data or analytic code for independent reproduction.

What would change my assessment: if the authors produce that reanalysis and the associations remain roughly the same after those corrections, I would consider the findings substantially stronger.

That’s how science is supposed to work. You don’t protect a conclusion from scrutiny because you agree with it. You try to break it. And if it survives, then you’ve learned something.

The bottom line

I’m not dismissing this study. I’m asking for it to be treated seriously enough that its inconsistencies are actually addressed.

There may be something worth investigating in these records. But “we found an association” and “we demonstrated causation” are not the same statement.

And CHD’s article makes the problem worse by misreporting at least two dose categories, repeatedly sliding between odds and risk, calling an internally ambiguous comparison group simply “unvaccinated,” and describing a remarkably broad ICD composite as recurrent strep throat.

If this paper is going to be cited as evidence that vaccinated children are chronically less healthy, the burden isn’t on skeptics to explain away every association.

The burden is on the analysis to survive basic questions about its denominators, definitions, confounders, and chronology.

This study deserves criticism, not deference. Its own tables don’t fully reconcile, key outcomes are poorly defined, major confounders weren’t controlled, and CHD’s write-up misstates some of the results while repeatedly turning odds into “risk.” That isn’t a minor presentation problem. It’s exactly how weak evidence gets dressed up to look stronger than it is.

There may be a useful dataset buried underneath this paper, but the analysis presented here is not strong enough to support the sweeping claims being made from it. When a paper has internal numerical inconsistencies, ambiguous exposure groups, broad diagnostic buckets, uncontrolled confounding, and no clean exposure-before-outcome analysis, the answer is not to promote it as proof. The answer is to fix the analysis.

And when CHD then misreports key quartiles and blurs odds with risk, that stops being a simple scientific disagreement and becomes a credibility problem.

If you want people to trust the conclusion, start by getting the numbers right.

Notes & Resources

1. Yengst T, Ray H, Jablonowski K, Hooker BS. Comparative Health Outcomes of Vaccinated versus Unvaccinated Children: A Retrospective Cohort Analysis. September 26, 2026. The uploaded preprint contains the cohort definitions, statistical methods, ICD definitions and Tables 1–7 discussed above.

2. Michael Nevradakis, “Vaccinated Kids Have Significantly Higher Odds of Developing Multiple Chronic Health Conditions,” Children’s Health Defense, Sept. 29, 2026. Read the CHD article

3. Glanz JM et al. “A population-based cohort study of undervaccination in 8 managed care organizations across the United States.” JAMA Pediatrics. 2013;167(3):274–281. The study reported different healthcare-utilization patterns among undervaccinated and age-appropriately vaccinated children. PubMed record

4. Davies HTO, Crombie IK, Tavakoli M. “When can odds ratios mislead?” BMJ. 1998;316:989. Also see Grant RL, “Converting an odds ratio to a range of plausible relative risks for better communication of research findings,” BMJ 2014. Both discuss why odds ratios should not automatically be communicated as relative risks, particularly for common outcomes. BMJ: When can odds ratios mislead? BMJ: Converting odds ratios to plausible relative risks

5. Centers for Medicare & Medicaid Services, ICD-10-CM Tabular List. The official coding material identifies J02.0 as streptococcal pharyngitis, J06 as acute upper-respiratory infection of multiple or unspecified sites, and includes separate diagnostic families for gonococcal and chlamydial infection. CMS ICD-10-CM tabular list

6. CMS/CDC ICD-10 documentation for R19. R19 is classified under symptoms and signs involving the digestive system and abdomen, not respiratory disease. CMS ICD-10-CM reference

7. U.S. Food and Drug Administration, “Common Ingredients in FDA-Approved Vaccines.” FDA describes why aluminum salts are used as adjuvants and states that aluminum-adjuvanted vaccines have a demonstrated safety profile extending over decades. FDA vaccine-ingredient information

You know the routine.

An account posts something outrageous: a distorted statistic, a conspiracy dressed up as a question, or a claim so obviously wrong that ignoring it feels irresponsible. You reply with evidence. The account dodges it, changes the subject, demands a different source, or adds three new claims. You answer those too.

Twenty replies later, the falsehood is still standing, the account has a larger audience, and you have donated an hour of your life to someone who spent thirty seconds writing the original post.

You thought you were correcting misinformation. The account saw engagement.

That is the rage-bait trap.

Oxford University Press named rage bait its 2025 Word of the Year and defined it as online content deliberately designed to provoke anger or outrage, usually to increase traffic or engagement. The phrase became popular because people recognized the experience immediately. We know we are being baited. We can often see the hook.

We bite anyway.

The question is why, and how we challenge misinformation without becoming unpaid promoters for the people spreading it.

The claim is not always the product. Your reaction is.

A normal disagreement has some path toward resolution. Evidence can narrow the dispute. A clear correction can change the conversation. Someone may admit that a statistic was wrong or that a source did not support the claim.

A rage-bait exchange behaves differently. Every answer creates another demand. Every source becomes evidence that the system is corrupt. Every correction produces a pivot. The discussion never moves toward resolution because resolution would end the content.

At that point, the account’s private thoughts hardly matter. Repeated behavior tells us what the account is doing. When someone recycles claims after they have been corrected, strips away context, ignores every answer, and immediately posts the next provocation, the account is functioning as rage bait. It does not need to issue a signed confession.

The incentive is built into the platform. X says its “For You” system ranks posts using a neural network trained on interactions including likes, reposts, and replies. That does not mean every angry reply automatically boosts every post by a fixed amount. It means your reaction is not invisible. It becomes data inside a system designed to predict what will hold attention.

A quote-post is even more direct: it places the original claim in front of your audience. The caption may say, “Look at this garbage,” but the garbage still gets delivered.

Agreement is useful to a rage-bait account. Disgust can be just as useful.

Outrage works

The effectiveness of outrage is not a guess.

In a large analysis of randomized headline tests from Upworthy. about 105,000 headline variations, 5.7 million clicks and more than 370 million impressions. researchers found that each additional negative word in an average-length headline increased the click-through rate by 2.3 percent.

A 2017 study of more than half a million political posts found that every additional moral-emotional word was associated with roughly a 20 percent increase in sharing. A preregistered replication published in 2025 examined 849,266 posts and found about 17 percent more expected shares for each additional moral-emotional word. In the same paper, a meta-analysis of 27 studies covering more than 4.8 million messages estimated the broader effect at about 13 percent.

Then there is tribal conflict. A 2021 study examining 2.7 million Facebook and Twitter posts from American politicians and news organizations found that each term referring to the political out-group was associated with a 67 percent increase in the odds of a post being shared.

Negativity catches attention. Moral outrage helps it travel. Attacking the other side helps it travel even farther.

Rage-bait accounts did not invent human psychology. They industrialized it.

Why we keep taking the bait

We do not reply only because we are angry. We reply because letting a lie pass feels wrong.

Misinformation creates a sense of unfinished business. If nobody challenges it, readers may assume it cannot be challenged. Silence feels like surrender, and blocking someone feels as if we allowed the person to have the last word.

There is also an audience beyond the person posting. We tell ourselves, often correctly, that the reply is not meant for the account. It is meant for the people quietly reading the exchange.

That is what makes rage bait effective: it borrows our conscience. It turns the desire to protect other people from misinformation into fuel for the account spreading it.

And there is one more reason we stay: we expect arguments to end when one side runs out of evidence. That expectation works only when both sides care about evidence. A person who is performing for engagement never runs out of material. When one claim fails, another appears.

You cannot win a game whose purpose is to keep you playing.

The necessary counterpoint: corrections can work

“Never engage” is too simple. False claims left completely unanswered can spread, and research shows that visible corrections can help the people watching.

In three preregistered experiments involving 6,621 people in the United Kingdom, Italy and Germany, corrective comments from ordinary users reduced both the perceived accuracy of false posts and people’s willingness to engage with them.

The same research also found a warning worth taking seriously: falsely “correcting” accurate information damaged belief in the truth. A correction has power, which is precisely why it should be accurate.

Another experiment found that corrections reduced misperceptions whether the tone was neutral, affirming, or uncivil. But uncivil replies made the surrounding conversation appear less civil. The facts still worked; the hostility contaminated the room.

So the answer is not silence. It is controlled correction.

Correct the claim once. Make the evidence easy for bystanders to see. Then stop. Do not allow one useful correction to become a forty-reply performance that supplies the original account with an afternoon of content.

I got blocked. That’s a good thing, I guess.

The grifter’s advantage

The person spreading misinformation has an enormous time advantage.

It takes seconds to write, “Childhood strokes are a vaccine-era phenomenon.” Properly answering it may require historical records, medical data, and multiple sources. The claim fits on a meme. The correction needs paragraphs.

Then, after you provide the evidence, the account can dismiss all of it with one word: “controlled.” Now you are expected to prove that every hospital, researcher, government database, and medical association is not part of the same conspiracy.

The account makes a claim. You are assigned homework.

This is not an honest burden-of-proof dispute. It is an attention-extraction method. The person making the sensational claim keeps transferring the burden back to anyone who challenges it, while feeling no obligation to establish the original accusation.

By the tenth reply, the thread is no longer educating anyone. You are staffing the account’s engagement desk.

Before replying, stop for thirty seconds. That pause matters. A meta-analysis of 20 experiments involving 26,863 participants found that prompts directing attention toward accuracy improved the quality of what people intended to share, largely by reducing intentions to share false headlines by about 10 percent.

Ask five questions:

1. What is my actual goal? Am I correcting the record for readers, or trying to force an account to admit defeat?

2. Is anyone reachable? Are undecided people seeing this, or is the thread populated entirely by performers and committed partisans?

3. Can I answer the central claim clearly? One strong correction is more useful than chasing five new claims introduced after the first one failed.

4. Will my response inform more people than it exposes? If the post has almost no reach, quote-posting it to a larger audience may solve a problem that did not yet exist.

5. What is my exit point? Decide before replying: one correction, one follow-up if genuinely necessary, then mute or block.

The checkpoint is not a test of whether the falsehood deserves an answer. It is a test of where that answer will do the most good.

Correct the misinformation without promoting the account

The Debunking Handbook 2020, written by a large group of misinformation researchers, recommends a useful structure: lead with the fact, warn briefly that a myth is coming, explain how the myth misleads, and finish by reinforcing the fact.

In practice:

• Lead with what is true. Do not make the false claim the headline of your correction.

• Name the specific error. “This graph changes the denominator halfway through” is stronger than “Everything you post is nonsense.”

• Use one or two strong sources. A link avalanche gives the account ten new exits and gives readers homework.

• Write for the people watching. The original poster may be unreachable. Bystanders are the real audience.

• Avoid unnecessary quote-posting. If the claim is obscure, answer it without delivering the account to your followers.

• Do not chase the pivot. When the subject changes, the original point has been conceded in practice, even if the account refuses to say so.

• End the exchange yourself. Rage bait depends on the belief that the other person is entitled to one more answer.

If the account violates platform rules, report it. If it is merely exhausting, mute it. If it repeatedly targets you or uses your replies as material, block it.

Blocking is not an admission that the account was right. It is the withdrawal of free labor.

The lie deserves an answer. The account does not own your time.

Social media trained us to confuse the last word with victory. Rage-bait accounts thrive on that confusion. They know that if they remain shameless long enough, a responsible person will keep returning with better evidence, clearer explanations, and more links.

We should challenge misinformation. We should also recognize when the challenge has stopped informing the public and started feeding the machine.

Correct what matters. Preserve the evidence. Speak to the people who may still be listening. Then leave.

The lie may deserve an answer.

The person spreading it is not entitled to your afternoon.

Sources and further reading

Oxford University Press, “The Oxford Word of the Year 2025 is rage bait,” December 1, 2025.
X Help Center, “For You Home Timeline Recommendations.”
Robertson, C. E., et al., “Negativity drives online news consumption,” Nature Human Behaviour 7, 812–822 (2023).
Brady, W. J., et al., “Emotion shapes the diffusion of moralized content in social networks,” Proceedings of the National Academy of Sciences 114, 7313–7318 (2017).
Brady, W. J., et al., “Estimating the effect size of moral contagion in online networks: A pre-registered replication and meta-analysis,” PNAS Nexus 4 (2025).
Rathje, S., Van Bavel, J. J., and van der Linden, S., “Out-group animosity drives engagement on social media,” Proceedings of the National Academy of Sciences 118 (2021).
Stoeckel, F., et al., “Social corrections act as a double-edged sword by reducing the perceived accuracy of false and real news in the UK, Germany, and Italy,” Communications Psychology 2 (2024).
Bode, L., Vraga, E. K., and Tully, M., “Do the right thing: Tone may not affect correction of misinformation on social media,” Harvard Kennedy School Misinformation Review (2020).
Pennycook, G., and Rand, D. G., “Accuracy prompts are a replicable and generalizable approach for reducing the spread of misinformation,” Nature Communications 13, 2333 (2022).
Lewandowsky, S., et al., “The Debunking Handbook 2020,” 2020.

By Vincent Mercer

You are sitting somewhere with nothing urgent to do. The television is off. Your phone is out of reach. No one is talking. Within minutes, perhaps seconds, a familiar restlessness begins.

There must be something to look at.

That uncomfortable space is boredom. We usually treat it as a minor irritation, a blank interval that should be filled as quickly as possible. Reach for the phone. Turn on the television. Open the refrigerator, even though we are not hungry. Anything to make the feeling go away.

But boredom is not simply the absence of entertainment. It is a message from the mind. The trouble is that we often silence it before finding out what it is trying to say.

What is boredom?

Psychologists John Eastwood and his colleagues offered a wonderfully plain definition: boredom is the unpleasant experience of wanting to engage in a satisfying activity but being unable to do so.

That last part matters. A person can have plenty to do and still be bored. There may be a hundred television channels, several unfinished chores and the entire internet waiting inside a phone. None of it feels worth entering.

Researcher Erin Westgate describes boredom as a problem involving two basic ingredients: attention and meaning. We tend to become bored when we cannot hold our attention on what we are doing, when the activity feels meaningless, or when both problems occur at once.

A task can be too easy and leave the mind underused. It can also be so difficult that the mind cannot get a grip on it. Both can produce boredom. Then there are activities that hold our attention perfectly well but feel pointless. We can do them. We simply do not care about them.

So boredom is not the mind shutting down. It is the mind refusing to settle.

What does the mind need?

The obvious answer is stimulation, but that is only partly true. If stimulation alone cured boredom, no one holding a smartphone would ever be bored again.

The mind needs a workable match between challenge and ability. Too little challenge and attention slips away. Too much and it gives up trying to enter. The mind also needs some reason to care, even if that reason is small: curiosity, usefulness, connection, progress, pleasure or the satisfaction of finishing something.

It also needs a degree of choice. Boredom becomes especially uncomfortable when we feel trapped in an activity and cannot change it. Think of waiting rooms, long meetings or repetitive work. The experience is not empty. It is crowded with the desire to be doing something else.

What the mind does not need is nonstop novelty. Novelty can capture attention, but capturing attention is not the same as satisfying it.

In a series of experiments published in 2024, people tried to avoid boredom by skipping through videos or switching from one video to another. The strategy often backfired. Digital switching left participants more bored, less satisfied and less engaged than simply watching one video. They kept searching for the thing that would hold them, but the searching itself prevented immersion.

That may be one of the strangest features of modern life. We have more entertainment than any people in history, yet we have built habits that can make sustained engagement harder. We are offered everything, so we commit to nothing.

Did cavemen get bored?

We cannot ask a Paleolithic human what he felt while sitting beside a fire 30,000 years ago. Boredom leaves no fossil. Any confident answer should therefore begin with some humility.

Still, there is no good reason to believe boredom arrived with civilization, television or the smartphone. Early humans had attention, goals, frustration, waiting and repetitive tasks. They experienced periods when nothing immediately useful could be done. It is reasonable to infer that they sometimes felt something recognizable as boredom, even if they had no word for it.

That does not mean prehistoric life was one long empty afternoon. Survival demanded long stretches of work: finding food, making tools, caring for children, preparing shelter and watching for danger. Yet human beings also played, decorated objects, made music and told stories. Those activities were not frivolous additions to human life. They helped people learn, bond, remember and imagine.

Modern hunter-gatherer societies are not living museum pieces, and we should not pretend they are exact replicas of prehistoric people. But they can suggest possibilities. Anthropologist Polly Wiessner compared daytime and nighttime conversation among the Ju/’hoansi of southern Africa. Daytime talk centered more on practical matters and social problems. Around the fire at night, conversation shifted heavily toward stories, along with singing, dancing, and ceremony. Once the day’s work ended, people did not merely wait for sleep. They entered one another’s memories and imaginations.

The first screen may have been the fire. The difference was that the audience also had to supply the stories.

What did people do before screens?

They did many of the things we still do when we put the screens down.

They visited neighbors. They told stories, played music, sang, danced, walked, read, wrote letters, played cards and board games, carved wood, sewed, knitted, cooked, tended gardens and repaired things. Children invented games from whatever happened to be nearby. Adults joined clubs, attended religious services, gathered in pubs and sat on porches talking.

Games are certainly not a modern invention. Archaeologists have found board games in ancient Egypt and the Near East, and possible game boards reach back to the Neolithic period. Human beings have been turning spare time into play for thousands of years.

We should not romanticize the past, however. Many people had far less leisure than we do. Household work took longer. Travel was slower. Jobs could be brutally repetitive. Sailors stood watch, soldiers waited, farm workers repeated the same motions, and the wealthy complained of ennui. People did not conquer boredom. Sometimes they endured it.

And that may be the part we have lost.

Before every quiet second could be eliminated, a person had time to move through boredom instead of escaping it instantly. The mind wandered. A conversation began. A game was invented. A small task became absorbing. An idea had enough silence around it to become visible.

Not every bored moment produced a poem or an invention. Most probably produced nothing remarkable. But boredom created an opening. What came next had to be supplied by the person, not delivered by an algorithm.

Is boredom the same as depression?

No. Ordinary boredom is a temporary state. It usually changes when the situation, task or level of engagement changes. Depression is a mental health condition involving persistent low mood or loss of interest and pleasure, generally lasting most of the day, nearly every day, for at least two weeks, along with other symptoms and difficulty functioning.

The two can overlap. A depressed person may describe life as boring because activities that once mattered no longer feel rewarding. Chronic boredom can also accompany loneliness, low motivation and a loss of meaning.

Can boredom lead to depression? The careful answer is that persistent boredom is associated with depressive symptoms and may contribute to declining mental health, but association is not a simple proof of cause. A 2024 study following 513 young Finnish workers found that greater job boredom predicted later increases in depression and anxiety symptoms, along with lower life satisfaction and positive functioning. The study followed change over time, which makes the finding more informative than a one-time survey, but it was still observational and specific to job boredom in a particular group.

Occasionally having nothing interesting to do is not depression. Feeling unable to enjoy anything, including things you normally value, is different. If that loss of interest persists for two weeks or more, affects daily life or comes with hopelessness, major changes in sleep or appetite, feelings of worthlessness or thoughts of self-harm, it should not be dismissed as boredom.

Would we adapt to life 100 years ago?

Imagine being transported back to 1926. At first, many of us would probably be miserable. Our hands would keep reaching for phones that did not exist. There would be no streaming, social media, video games or instant answers. Entertainment would require planning, participation and often other people.

But the mind adjusts its expectations. After a while, a newspaper might hold our attention longer. A card game might last all evening. We might listen to an entire record without skipping a song or sit through a conversation without checking anything.

We would still get bored. People were bored long before smartphones. But we would probably become less frightened of boredom because there would be no instant escape from it. Eventually, the slower world might stop feeling empty and begin feeling normal.

The real question is not whether we could adapt to 1926. We could. The more interesting question is whether returning to 2026 would then feel unbearably noisy.

Listen before you escape

The next time boredom arrives, it may help to delay the automatic cure for a few minutes and ask a better question.

Is this activity too easy? Too difficult? Does it feel pointless? Am I tired? Do I need company? Am I avoiding something? Or have I become so used to rapid stimulation that ordinary life now feels unbearably slow?

Different forms of boredom need different answers. Make an easy task harder. Break a difficult one into a smaller step. Rest if attention is exhausted. Find a reason the task matters, or admit honestly that it does not. Call someone. Make something. Take a walk without turning it into another delivery system for content.

Most of all, choose one thing and stay with it long enough to enter it.

Boredom is uncomfortable because it sits between what we are doing and what we want to be doing. A screen can cover that distance for a moment. It cannot always tell us where we actually want to go.

If you made it this far, perhaps boredom has already done its job.

Think Critically. Follow the Evidence. Question Everything. Question Yourself.

Sources

• John D. Eastwood et al., “The Unengaged Mind: Defining Boredom in Terms of Attention”, Perspectives on Psychological Science (2012).

• Erin C. Westgate, “Why Boredom Is Interesting”, Current Directions in Psychological Science (2020).

• Katy Y. Y. Tam and Michael Inzlicht, “Fast-Forward to Boredom: How Switching Behavior on Digital Media Makes People More Bored”, Journal of Experimental Psychology: General (2024).

• Polly W. Wiessner, “Embers of Society: Firelight Talk Among the Ju/’hoansi Bushmen”, Proceedings of the National Academy of Sciences (2014).

• The Metropolitan Museum of Art, “Board Games from Ancient Egypt and the Near East” (2018).

• Jie Li, Jari Hakanen and Johanna Huhtala, “Job Boredom as an Antecedent of Four States of Mental Health”, BMC Public Health (2024).

• World Health Organization, “Depressive Disorder (Depression)”.

There is an important distinction between showing that an experimental condition can produce an effect and showing that the condition caused the effect in another experiment.

That distinction is at the center of SARS COV 2 ISOLATION PROTOCOL CONTROL EXPERIMENTS, published by Jamie Andrews on September 27, 2025.

The experiment compares uninfected Vero E6 cells cultured under two serum conditions: 10% fetal bovine serum (FBS) and 2% FBS. The lower-serum cultures grew substantially less and developed morphological changes including rounded cells, granularity, vacuole-like structures, debris and signs of cellular stress.

I am not going to argue that this observation is fake. In fact, doing so would weaken the critique.

Culture conditions can affect Vero-cell growth and survival. In a Vero-cell bioreactor study comparing serum-containing with serum-free medium, the serum-free condition produced a lower maximal viable-cell density and earlier evidence of cell death; apoptosis was detected among detached cells.

That study did not test 10% versus 2% FBS, so it should be treated as background evidence only, not as a replication of Andrews’s experiment.

The legitimate lesson is:

Cell morphology alone can be nonspecific, and culture conditions must be controlled when interpreting cytopathic effects.

That is a useful point. But the Substack article goes much further.
It concludes that serum reduction itself causes the cytopathic effects attributed to viral infection, that cell-culture virus isolation is therefore invalid, and that downstream biochemical assays used to detect viruses are consequently invalid as well.

Those conclusions require evidence that this experiment does not provide. And that is where the argument breaks down.

What the experiment actually shows

The strongest version of the authors’ result is straightforward.

Take otherwise uninfected Vero E6 cultures and alter their serum conditions. Their growth and morphology can change. The contract laboratory’s observations support at least part of that proposition.

The lower-serum cultures showed persistently lower confluence. By later observations, technicians described dense or granular cytoplasm, vacuole-like reflective structures, rounded cells, debris, and cells that might be dying.

That is meaningful.

But there is an important detail that gets lost when the experiment is subsequently summarized as having reproduced viral CPE.

The laboratory’s descriptions are considerably more cautious than the article’s ultimate interpretation.

For several days, even cells in the 2% FBS condition were described as having intact membranes, remaining attached and showing no obvious detachment. On day five, the laboratory characterized the culture as showing some sign of stress, with possible dying cells or debris.

The later classification of these observations into multiple categories of viral CPE is an interpretation by the study authors.

The article itself acknowledges that some claimed morphologies, including syncytia and plaque formation, were not explicitly identified by the laboratory technician. The authors nevertheless interpreted them as being present.

That distinction matters.

The experiment provides evidence of serum-associated differences in cell growth and morphology.

It does not independently establish that every morphology later labeled as CPE was actually present, much less that those morphological changes are equivalent to the complete phenotype seen during a productive SARS-CoV-2 infection.

That is the essential distinction. The control experiment establishes a possible confounder. It does not establish that the confounder explains the phenomenon being challenged.

The missing comparison

Suppose changing the nutrient conditions in a cell culture makes some cells round up, grow slowly, become granular, or die.

That immediately raises a sensible question:

Could culture conditions contribute to morphology that might otherwise be interpreted as CPE?

Yes.

But answering the stronger causal question requires comparing an appropriately matched mock condition with a virus-exposed condition while keeping the relevant culture conditions comparable.

Andrews’s experiment does not do that. It contains no SARS-CoV-2-exposed arm. It compares an uninfected 10% FBS culture with an uninfected 2% FBS culture.

That design is useful for asking what changing the serum concentration does to uninfected Vero E6 cells. It cannot determine how much of the effect in an actual SARS-CoV-2 isolation experiment is attributable to serum concentration and how much is associated with the inoculated clinical material and subsequent viral replication.

Calling the experiment a “true negative control” therefore needs qualification.

It is a useful serum-control experiment.

It is not, by itself, a matched mock-versus-virus experiment in which the relevant handling and culture conditions are held constant while viral exposure is the key difference.

That becomes crucial when the argument moves from:

“Reduced serum can produce some overlapping morphology.”

to:

“Reduced serum is the cause of the CPE attributed to the virus.”

The second proposition requires evidence that the first does not supply.

Published virology guidance does not treat CPE alone as definitive

Andrews’s conclusion states that the American Society for Microbiology treats the observation of “any CPE morphology” as sufficient to indicate viral presence.

The ASM protocol is more qualified than that characterization suggests.

ASM explains that some viruses produce characteristic cytopathic effects and says that the microscopic appearance of CPE from some cytocidal viruses may be sufficiently characteristic to allow provisional identification of an unknown virus.

“Provisional” matters. Morphology can provide evidence and help guide identification without being definitive by itself.

FDA guidance for respiratory viral panel submissions is more explicit still. When viral culture is used, FDA recommends virus-specific identification in addition to CPE and states that CPE alone may not provide accurate viral identification.

Banerjee and colleagues made the same distinction in their SARS-CoV-2 isolation study: “The presence of CPE alone does not indicate successful isolation of a coronavirus.”

So Andrews’s control experiment supports a legitimate caution: nonspecific culture conditions can complicate interpretation of cell morphology.

What it does not establish is that conventional viral identification rests on treating any morphological change as conclusive proof of a specific virus.

What happens when actual SARS-CoV-2 isolation studies are examined?

This is where the serum-only explanation meets evidence that is not reducible to cell morphology.

Consider the work of Harcourt and colleagues on the first U.S. COVID-19 patient, published in Emerging Infectious Diseases in 2020.

The researchers reported mock-infected cultures alongside cultures inoculated with the patient’s nasopharyngeal and oropharyngeal specimens. CPE was not observed in the mock culture and was observed after inoculation with the clinical specimens.

But the identification did not stop there.

They performed confirmatory RT-PCR on culture lysates, tested the isolates for 33 additional respiratory pathogens, and performed whole-genome sequencing.

Most importantly for the present argument, the genomes obtained from the cultured nasopharyngeal and oropharyngeal isolates were identical to each other and showed 100% identity with the corresponding sequence obtained from the patient’s clinical specimen.

The isolate was also passaged and studied with infectious-titer assays and multistep replication measurements. In Vero cells, infectious titers rose rapidly after inoculation and reached high levels.

Those measurements are not all logically independent of one another, but neither are they simply different names for cell rounding. They include morphology, virus-specific nucleic-acid detection, genome sequence, infectious titer, and replication kinetics.

A serum-associated change in cell shape does not, by itself, account for why a SARS-CoV-2 genomic sequence independently detected in the patient specimen was recovered from the cultured isolate while infectious titers increased after inoculation.

That is the problem for a serum-only explanation: it addresses one possible source of nonspecific morphology, but not the rest of the isolation evidence.

What about contagion?

Another objection sometimes raised is that SARS-CoV-2 has never been shown experimentally to pass from an infected host to a previously uninfected host.

That blanket claim is not supported by the experimental literature.

Dowall and colleagues used the SARS-CoV-2 Victoria/01/2020 strain in a hamster transmission model. Donor hamsters were inoculated with the virus, while naïve recipient hamsters were housed in physically separate adjacent cages connected by directed airflow; the recipients themselves were not experimentally inoculated.

SARS-CoV-2 RNA was subsequently detected in throat swabs from 14 of 16 recipient animals. Clinical signs developed in recipient animals, and histopathological examination found pneumonia in 10 of 16 recipients.

There is an important limitation worth stating explicitly: the Dowall study did not report recovery of live virus from recipient throat swabs. Its evidence in the recipients consisted of viral RNA, clinical findings, and histopathology. It should therefore be cited for transmission-associated infection and disease in this model, not as evidence that infectious virus was re-isolated from those particular recipient swabs.

Other animal studies provide complementary evidence. Sia and colleagues reported efficient SARS-CoV-2 transmission from inoculated to naïve hamsters by direct contact and by aerosols; naturally infected animals developed weight loss and neutralizing antibodies.

In ferrets, Richard and colleagues demonstrated transmission by direct contact and through the air, and reported that infectious virus was isolated from all animals that tested positive.

The 2022 Nature Medicine human challenge study addresses a related but different question. It was not a person-to-person transmission experiment. Thirty-six healthy adults aged 18–29 with no evidence of previous infection or vaccination were deliberately inoculated with a defined SARS-CoV-2 preparation; 34 were included in the per-protocol analysis and 18 became infected.

Viral load rose rapidly, viable virus was recoverable from the nose, and most infected participants developed mild-to-moderate symptoms. The study found no clinical or radiological evidence of pulmonary disease in these carefully selected young adults.

So the evidence should be described precisely:

Controlled human infection has been demonstrated, and experimental host-to-host transmission has been demonstrated independently in multiple animal models.

Those are different experimental questions. Neither should be overstated, but neither can fairly be summarized as “there has never been a successful contagion study.”

What about demanding a live video of the virus entering a cell and replicating?

That is not a scientific criterion for establishing the existence or replication of a virus.

Different techniques answer different questions. Conventional transmission electron microscopy provides high-resolution structural information but generally examines fixed specimens, so it is not designed to provide a continuous live movie of an entire infection cycle.

Live-cell systems can visualize infection and spread over time. In a 2024 Nature Communications study, researchers used a fluorescent reporter virus derived from SARS-CoV-2 USA/WA1-2020 to image infected differentiated human bronchial epithelium longitudinally for up to 12 days. The fluorescent signal marked infected cells and allowed infection foci and their movement through the tissue to be followed over time.

Earlier work in the Journal of Virology used a reporter-cell system during SARS-CoV-2 infection and performed live imaging every 10 minutes for 18 hours, observing a time-dependent reporter response in infected cells.

Direct structural observation is also available, although it is not live imaging. Klein and colleagues used in situ cryo-electron tomography of SARS-CoV-2-infected cells to visualize replication-associated double-membrane vesicles, RNA filaments, virion budding sites, spike trimers, viral ribonucleoprotein complexes, intracellular virions, and extracellular virions.

Those studies should not be oversold. They are not a literal camera view of one unmodified virion being followed continuously through attachment, entry, uncoating, genome copying, assembly, and release.

That is not their purpose.

Replication is instead evaluated through converging measurements appropriate to different stages of the process—for example, recovery of infectious material, changes in infectious titer over time, virus-specific nucleic-acid detection, genome sequencing, viral protein expression, replication kinetics, serial passage, and infection of additional susceptible cells.

A demand for one cinematic image sequence does not negate those measurements. The relevant question is whether the individual methods are properly controlled and whether their results converge.

Not every SARS-CoV-2 isolation protocol used the same serum regimen.

Andrews’s article makes a broad claim that documented SARS-CoV-2 isolations reduced infection medium to 2% or even 1% FBS.

The Harcourt protocol does not fit that description neatly.

Harcourt and colleagues report maintaining Vero cultures in medium containing 5% or 10% FBS. During the initial isolation procedure, the Vero cells were resuspended in DMEM containing 10% FBS before being added to the clinical-specimen dilutions.

There is an important precision point here: because that cell suspension was mixed with specimen and serum-free diluent already present in the wells, the final FBS concentration in every well should not be described simply as 10%.

But the published method also does not describe the specific switch to a 2% FBS infection medium tested in Andrews’s control experiment.

That is enough to show that the serum-regimen claim is not universal. SARS-CoV-2 isolation protocols have not all used the same serum regimen, and the effect of a particular 10%-to-2% switch cannot simply be generalized to every isolation protocol.

What about studies that actually did use 2% FBS?

Banerjee and colleagues are particularly relevant because their 2020 SARS-CoV-2 isolation method explicitly states that, after the initial inoculation period, the inoculum was replaced with DMEM containing 2% FBS and TPCK-treated trypsin.

The paper also shows mock-inoculated Vero E6 cultures alongside cultures inoculated with clinical specimens, and the specimen-inoculated cultures developed extensive CPE relative to the mock cultures.

There is, however, a limitation that should not be glossed over: the methods do not separately spell out every reagent and concentration used in the mock arm. For that reason, this paper should not be presented as definitive proof that the mock cultures were matched to the infected cultures in every medium component, including the exact 2% FBS condition.

Fortunately, the argument does not need that claim.

Banerjee and colleagues themselves emphasized that CPE alone was insufficient for identification. They transferred supernatant from the clinical-specimen cultures to fresh Vero E6 cells, confirmed SARS-CoV-2 RNA in culture supernatant by RT-PCR, generated approximately 94%-complete SARS-CoV-2 genome sequences at very high sequencing depth, measured infectious titers, and went on to characterize replication competence.

The paper also reports sequence substitutions shared with direct sequencing of clinical isolates, but that particular comparison refers in part to unpublished data. I therefore would not use that comparison as a central piece of the argument.

The stronger published point is simpler: even in an isolation protocol that used 2% FBS, the investigators did not treat cell morphology as sufficient evidence of viral identity. They combined culture observations with molecular identification, sequencing, infectivity, and replication measurements.

This is where the causal argument fails

The control-study argument effectively proceeds this way:

• Reduced serum can produce cellular abnormalities.

• Some of those abnormalities resemble features that can occur during viral CPE.

• Some virus-isolation protocols use reduced-serum media.

• Therefore, reduced serum causes the CPE attributed to viruses.

• Therefore, the claimed virus was not actually isolated.

• Therefore, molecular tests built around that virus are invalid.

The first three points support a hypothesis worth testing.

They do not establish the remaining conclusions.

To show that serum reduction causes the findings attributed to SARS-CoV-2, the evidence would have to distinguish the effects of the culture condition from the additional effects associated with the clinical inoculum and with recoverable, replicating infectious material.

Andrews’s experiment never performs that comparison.

That is the missing causal link.

The electron-microscopy criticism is partly right

The article also argues that uninfected cells can contain structures resembling particles that have sometimes been identified as coronaviruses under transmission electron microscopy.

That concern is legitimate.

Electron-microscopy specialists Cynthia Goldsmith, Sara Miller, and colleagues cautioned during the pandemic that normal cellular structures—including coated vesicles and multivesicular bodies—can be mistaken for coronavirus particles.

The supported conclusion is therefore narrow but important:

Ambiguous TEM morphology should not be used by itself to identify SARS-CoV-2.

But resemblance is not identity.

Finding a vesicle-like structure in an uninfected culture that resembles a published viral image does not establish that the published structure was also a cellular vesicle, nor does it invalidate virus identification supported by additional molecular, genomic, and infectivity evidence.

This is also why the in situ cryo-electron tomography work matters. In infected cells, investigators have visualized not merely isolated round particles but replication-associated compartments, viral RNA-containing structures, budding events, spike trimers, viral ribonucleoprotein complexes, and assembled virions in their cellular context.

Morphology is strongest when it is interpreted together with the rest of the evidence rather than treated as a stand-alone verdict.

“Isolation” does not mean “chemical purification”

Another major point of disagreement concerns the word isolation.

Andrews argues that unless researchers first obtain a purified, distinct biological particle, subsequent identification lacks a valid benchmark.

That is not the conventional meaning of virus isolation in clinical virology.

CLSI’s own terminology database defines viral culture as the infection of cell cultures by viable infectious viral particles, usually resulting in viral replication, and lists “viral isolation” as an alternate term.

A major review in Clinical Microbiology Reviews likewise repeatedly uses “virus isolation in cell culture” to describe recovery of viable virus in susceptible cells followed by identification and confirmation.

This terminology does not mean purification is useless.

Purification can be essential for particular structural, biochemical, or analytical questions. It is simply a different operation from recovering and propagating a viable virus isolate in susceptible cells.

The distinction matters because otherwise the argument becomes circular: redefine “isolation” to mean “chemically purified virions only,” then declare every conventional cell-culture isolation invalid because it does not satisfy that redefinition.

A fair critique should instead evaluate conventional virus isolation by the criteria the method actually claims to satisfy: recovery of infectious material, reproducible propagation in susceptible systems, appropriate controls, and identification by suitable independent methods.

A note about the CLSI claim

The control article invokes CLSI M41 as support for the proposition that infection media are supposed to be reduced to roughly 1–3% FBS.

This point needs careful wording.

The public CLSI page and publicly available sample confirm that M41 is a 2006 viral-culture guideline covering cell-culture selection and maintenance, quality control, culture-medium preparation, specimen handling, isolate identification, and interpretation. CLSI currently lists it as an archived document that remains technically valid.

The publicly accessible material I could verify does not expose enough of the full standard to establish whether the cited 1–3% FBS recommendation appears in the precise form Andrews describes.

So I would not claim that M41 contains no such recommendation.

But the public material also does not support presenting a specific low-serum concentration as a universal rule governing every virus-isolation protocol.

And even if the full M41 document recommends reduced-serum maintenance medium for particular applications, that would establish a culture practice—not the causal proposition that reduced serum is therefore responsible for the virus-specific findings reported in SARS-CoV-2 isolation studies.

The relevant experimental question remains:

What happens to appropriately matched inoculated and mock-inoculated cultures under comparable conditions, and what independent evidence identifies the recovered agent?

Modern SARS-CoV-2 characterization is explicitly multi-method

This point has become even clearer with time.

ECDC’s updated 2026 handbook for SARS-CoV-2 characterization does not describe identification as “look for CPE and stop.”

It covers cell culture alongside plaque assays, viral titration, antibody-based detection, sequencing and PCR, and specifically emphasizes that protocols vary among institutes and sometimes require adaptation.

That is important because the control article treats one culture condition as though it were the single foundation beneath SARS-CoV-2 identification.

It isn’t.

The evidentiary picture is distributed across independent techniques.

That is exactly what one should want.

The largest unsupported leap: “therefore downstream assays are invalid”

The article’s final step is its broadest.

It argues that if cell-culture isolation is invalid, downstream molecular and biochemical assays must also be invalid because no properly isolated benchmark exists.

But Andrews’s experiment does not test those assays.

• It does not test sequencing specificity.

• It does not test PCR specificity.

• It does not show SARS-CoV-2 sequences arising in its uninfected cultures.

• It does not demonstrate recoverable infectious virus in an uninoculated culture.

• It does not show the patient-associated viral genome appearing reproducibly as a consequence of reduced serum.

Nor does the experiment address the Harcourt result in which the sequence recovered from the cultured isolate matched the corresponding sequence obtained directly from the clinical specimen.

That observation matters because the sequence was documented in the patient specimen independently of the subsequent culture result. The cultured isolate then yielded the corresponding sequence while infectious-virus measurements showed replication in susceptible cells.

One can raise separate questions about PCR design, sequencing, contamination control, or assay validation. Those questions should be evaluated on their own evidence.

A serum-only experiment that measures cell morphology does not, by itself, invalidate them.

What would make this control experiment much more consequential?

This is the scientific question that matters.

Imagine that an appropriately matched mock control, subjected to the same relevant culture and handling conditions, repeatedly and reproducibly produced not merely similar-looking cellular damage but also the same SARS-CoV-2-specific molecular signal, a corresponding viral genome, and serially transferable or quantifiable infectious activity attributed to the supposed isolate.

That would be a profoundly important result.

It would directly challenge the specificity of the isolation process and demand an explanation.

But that is not what this experiment reports.

It reports morphological changes in uninfected Vero E6 cells exposed to different serum conditions.

Interesting?

Yes.

Worth considering when interpreting CPE?

Absolutely.

Evidence that morphology should not stand alone?

Certainly.

Evidence, by itself, that SARS-CoV-2 isolation has been disproven?

No.

The strongest conclusion is also the narrowest one

There is no need to dismiss Andrews’s experiment.

Its strongest observation can simply be accepted:

Reducing serum from 10% to 2% altered the growth and morphology of these uninfected Vero E6 cultures, and several observed features overlap with morphological categories that can also occur during viral CPE.

That supports careful controls.

It supports caution about identifying infection from morphology alone.

And it supports skepticism toward any experiment that treats a nonspecific visual change as sufficient proof of viral identity.

But that is not where the evidence in SARS-CoV-2 isolation studies ends.

Andrews’s experiment never exposes cells to SARS-CoV-2. It therefore does not measure the difference between serum-associated cell stress and the additional effects associated with a SARS-CoV-2-containing clinical inoculum.

Its design also does not test or account for:

• recovery of replication-competent infectious material from clinical specimens;

• rising or quantifiable infectious titers;

• virus-specific molecular detection;

• genome sequencing;

• correspondence between sequences recovered directly from clinical specimens and cultured isolates;

• serial passage into additional susceptible cells; or

• experimental infection and host-to-host transmission evidence from other study designs.

That is the central problem.

The strongest criticism of the study is therefore not that its cell-stress observation must be false. It is that the conclusion extends far beyond what the experiment was designed to test.

When the Conclusion Is Built Into the Standard

Some critics begin by declaring that only one particular method can count as valid evidence. Once that rule is adopted, every other line of evidence is rejected before it is examined. That is the central flaw: the conclusion is built into the standard being imposed.

Disagreeing with the accepted evidentiary framework does not demonstrate that the evidence within that framework is false; it merely shows that the critic refuses to recognize the kinds of converging evidence the field uses. A methodological preference is not, by itself, an experimental refutation.

Resources and Primary Sources

Jamie Andrews — “SARS COV 2 ISOLATION PROTOCOL CONTROL EXPERIMENTS”, September 27, 2025.
The experiment evaluated in this article. It compares uninfected Vero E6 cultures maintained under 10% and 2% FBS conditions and argues that serum reduction can produce morphology interpreted as CPE and invalidate virus isolation.

Harcourt J, Tamin A, Lu X, et al. “Severe Acute Respiratory Syndrome Coronavirus 2 from Patient with Coronavirus Disease, United States.” Emerging Infectious Diseases. 2020;26(6):1266–1273. doi:10.3201/eid2606.200516.
Reports mock cultures, confirmatory RT-PCR, testing for 33 additional respiratory pathogens, whole-genome sequencing, a 100% sequence match between cultured isolates and the corresponding clinical specimen, infectious-virus quantification, and replication studies.

Banerjee A, Nasir JA, Budylowski P, et al. “Isolation, Sequence, Infectivity, and Replication Kinetics of Severe Acute Respiratory Syndrome Coronavirus 2.” Emerging Infectious Diseases. 2020;26(9):2054–2063. doi:10.3201/eid2609.201495.
The isolation method used 2% FBS after inoculation. The paper included mock-inoculated cultures and explicitly stated that CPE alone does not establish successful coronavirus isolation. The methods do not separately itemize every reagent concentration for the mock arm, so this article does not rely on Banerjee as proof of a perfectly matched 2% FBS mock.

American Society for Microbiology — “Cytopathic Effects of Viruses Protocols.”
Describes characteristic CPE as potentially permitting provisional identification of an unknown virus rather than definitive identification from morphology alone.

U.S. Food and Drug Administration — “Respiratory Viral Panel Multiplex Nucleic Acid Assay: Class II Special Controls Guidance.”
States that CPE alone may not provide accurate viral identification and recommends virus-specific identification in addition to CPE when viral culture is used in this regulatory context.

Clinical and Laboratory Standards Institute — “M41: Viral Culture.”
A 2006 guideline covering viral culture, culture-medium preparation and quality control, specimen processing, isolate identification, and interpretation. CLSI lists it as archived but technically valid.

Leland DS, Ginocchio CC. “Role of Cell Culture for Virus Detection in the Age of Technology.” Clinical Microbiology Reviews. 2007;20(1):49–78. doi:10.1128/CMR.00002-06.
A major review describing the conventional use of cell culture for virus isolation and the combination of culture with antigenic and molecular identification methods.

Goldsmith CS, Miller SE, Martines RB, Bullock HA, Zaki SR. “Electron microscopy of SARS-CoV-2: a challenging task.” The Lancet. 2020;395(10238). doi:10.1016/S0140-6736(20)31188-0.
Discusses the difficulty of correctly distinguishing coronavirus particles from normal cellular structures by electron microscopy.

Miller SE, Goldsmith CS. “Caution in Identifying Coronaviruses by Electron Microscopy.” Journal of the American Society of Nephrology. 2020;31(9):2223–2224. doi:10.1681/ASN.2020050755.
Documents the risk of mistaking ordinary cellular structures, including coated vesicles and multivesicular bodies, for coronavirus particles.

Quesney S, Marvel J, Marc A, Gerdil C, Meignier B. “Characterization of Vero cell growth and death in bioreactor with serum-containing and serum-free media.” Cytotechnology. 2001;35(2):115–125. doi:10.1023/A:1017589526145.
Background evidence that serum conditions can affect Vero-cell growth and cell death. This study compared serum-containing with serum-free conditions, not 10% versus 2% FBS.

European Centre for Disease Prevention and Control — “Update of standard laboratory protocols for SARS-CoV-2 characterisation – 2026.”
Describes multiple methods used in SARS-CoV-2 characterization, including cell culture, plaque assays, viral titration, antibody-based methods, neutralization, PCR, and sequencing, and notes protocol variability between institutes.

Killingley B, Mann AJ, Kalinova M, et al. “Safety, tolerability and viral kinetics during SARS-CoV-2 human challenge in young adults.” Nature Medicine. 2022;28:1031–1041.
A controlled human infection study inoculated 36 young adults with a characterized SARS-CoV-2 preparation; 34 were included in the per-protocol analysis, and 18 developed infection. In those participants, viral levels rose over time and viable virus was recoverable, demonstrating productive infection rather than simple persistence of the inoculum. The study was not designed to test person-to-person transmission, and no pulmonary disease was observed in these carefully selected volunteers. Its purpose was narrower: to determine whether SARS-CoV-2 could produce a time-dependent, replicating infection in susceptible humans, and it did.

Sia SF, Yan L-M, Chin AWH, et al. “Pathogenesis and transmission of SARS-CoV-2 in golden hamsters.” Nature. 2020;583:834–838.
Demonstrated efficient transmission from inoculated to naïve hamsters by direct contact and aerosols. Naturally infected animals developed weight loss and neutralizing antibodies.

Richard M, Kok A, de Meulder D, et al. “SARS-CoV-2 is transmitted via contact and via the air between ferrets.” Nature Communications. 2020;11:3496.
Demonstrated direct-contact and airborne transmission between ferrets and reported infectious virus isolation from all positive animals.

Dowall S, Salguero FJ, Wiblin N, et al. “Development of a Hamster Natural Transmission Model of SARS-CoV-2 Infection.” Viruses. 2021;13(11):2251. doi:10.3390/v13112251.
Used SARS-CoV-2 Victoria/01/2020 in a physically separated airflow transmission system. Viral RNA was detected in 14 of 16 recipients and pneumonia in 10 of 16. The study did not report live-virus recovery from recipient throat swabs.

Becker ME, Martin-Sancho L, Simons LM, et al. “Live imaging of airway epithelium reveals that mucociliary clearance modulates SARS-CoV-2 spread.” Nature Communications. 2024;15:9480. doi:10.1038/s41467-024-53791-4.
Used a fluorescent reporter virus derived from SARS-CoV-2 USA/WA1-2020 and longitudinal live imaging of differentiated primary human bronchial epithelium for up to 12 days, allowing infection foci and viral spread to be followed over time.

Klein S, Cortese M, Winter SL, et al. “SARS-CoV-2 structure and replication characterized by in situ cryo-electron tomography.” Nature Communications. 2020;11:5885. doi:10.1038/s41467-020-19619-7.
Used in situ cryo-electron tomography to directly visualize SARS-CoV-2 replication-associated structures, viral ribonucleoprotein complexes, budding sites, and assembled virions in infected cells. These are cryopreserved structural observations, not live-cell movies.

Pahmeier F, Neufeldt CJ, Cerikan B, et al. “A Versatile Reporter System To Monitor Virus-Infected Cells and Its Application to Dengue Virus and SARS-CoV-2.” Journal of Virology. 2021;95. doi:10.1128/JVI.01715-20.
Used a reporter-cell system for live-cell monitoring of SARS-CoV-2 infection, including imaging at 10-minute intervals over an 18-hour period.

Spend a few days following Nicolas Hulscher, and a pattern begins to emerge: a preliminary association becomes a warning, a warning becomes a crisis, and by the time the finding reaches social media, much of the scientific uncertainty has disappeared.

BREAKING. BOMBSHELL. SHOCKING. STUNNING. EXPLOSIVE. SMOKING GUN. FRAUD. COVER-UP. LETHAL.

Those words were not collected by searching years of archived posts.

The examples examined here came from Hulscher posts dated largely between September 12 and September 18, 2026. This is not presented as a quantitative sample of everything he has ever published. It is a close examination of claims encountered within a remarkably short recent window, followed back to the papers, preprints, and datasets underneath them.

That distinction matters.

This article is not an argument that Nicolas Hulscher has no scientific credentials. He does. He earned an MPH in epidemiology from the University of Michigan in 2024. The McCullough Foundation identifies him as its epidemiologist and administrator and says he plays a leading role in scientific research, public communications and policy engagement, manages its social-media platforms, authors publications and presents research to public audiences.

Nor is the argument that every study he cites is worthless, every safety signal imaginary, or every public-health institution beyond criticism.

Some of the papers Hulscher highlights contain findings worth investigating.

The question is what happens between the scientific evidence and the headline.

• Again and again, the language changes.

• An association begins to sound like an effect.

• An adverse-event report begins to sound like a confirmed vaccine death.

• A hypothesis becomes a smoking gun.

• A possible determinant becomes undeniable proof.

• A dispute over pharmacovigilance becomes a cover-up.

• Separate controversies in different countries become a coordinated global operation.

• And a prototype for recording vaccination history becomes a future system determining who may shop, dine or travel.

The evidence is not merely summarized. It is amplified.

I also gave Nicolas Hulscher a direct opportunity to respond before publication. I sent him the central factual questions raised in this article, including the evidentiary basis for his broader claims, the way he characterizes preliminary or observational findings, and the specific sources supporting statements presented as established fact. As of publication, he had not responded. That silence does not prove the claims are false, but it does mean he declined an opportunity to clarify, correct, or defend them on the record.

How this review was done

For each example, I took Hulscher’s public claim and went back to the source, the paper, preprint, database, or institutional record, to see what the evidence actually showed. I compared his wording with the study design, results, limitations, and authors’ conclusion. Where passive-surveillance data such as VAERS were involved, I checked the interpretation guidance issued by the agencies that operate the system. I checked commercial relationships against published financial disclosures and current company product pages.

That is important because the standard applied here is simple:

If a headline is stronger than the evidence beneath it, the difference should be visible.

Exhibit 1: His scientific paper says “potential.” His social media says “undeniable.”

In September 2026, Hulscher and six coauthors uploaded a preprint describing a 35-year-old man with an aggressive cardiac intimal sarcoma following COVID-19 vaccination.

The authors reported detecting DNA sequences they attributed to Pfizer vaccine-manufacturing plasmid material in archived tumor tissue. The case is unusual. The finding is worth investigating.

It is also one patient.

Zenodo identifies the paper as a preprint. The manuscript describes primary cardiac intimal sarcoma itself as an exceedingly rare and aggressive cancer and discusses vaccination as a potential determinant, rather than demonstrating that vaccination caused the malignancy.

Hulscher’s public wording was markedly stronger.

In the posts reviewed for this article, he called the case the “TURBO CANCER smoking gun,” said it made vaccine-related turbo cancer “undeniable,” and said the study “obliterates” the claim that turbo cancer is a myth.

But the evidence had not changed between manuscript and social-media post.

The sample was still one patient. There was no control group.

There was no population comparison demonstrating that vaccinated people develop cardiac intimal sarcoma at a greater rate.

And detecting reported sequences in extracted tumor tissue does not, by itself, establish genomic integration into malignant cells or demonstrate that those sequences caused or accelerated the tumor.

That difference can be expressed in three words:

potential → smoking gun → undeniable

That is not simply simplification for a general audience.

It is an increase in certainty.

Exhibit 2: A draft Henry Ford says failed internal review becomes one of the “most POWERFUL” vaccine studies ever conducted

Hulscher also promoted a vaccinated-versus-unvaccinated childhood analysis originating at Henry Ford Health.

His headline statistic was memorable:

57% of vaccinated children developed at least one chronic condition, compared with 17% of unvaccinated children.

He called the birth-cohort analysis one of the “most POWERFUL vaccine safety studies ever conducted.”

Henry Ford Health describes the document very differently.

According to the health system, the analysis was a 2020 draft that was shelved after its first internal review because of serious problems with its data and methodology. Henry Ford says the unvaccinated cohort was substantially smaller and differed demographically from the vaccinated cohort. Most importantly, follow-up was dramatically different: about one quarter of the unvaccinated children were observed only through six months of age, and 75% only through age three.

That is a serious problem when the outcome being counted is the accumulation of chronic diagnoses throughout childhood.

Children who remain in a healthcare system longer have more opportunity to receive diagnoses than children who disappear from observation earlier.

Hulscher subsequently promoted crude proportional comparisons across numerous categories, including enormous relative increases and several disorders described as appearing “ONLY IN VACCINATED.”

Large relative percentages are emotionally powerful. They can also become unstable when numbers are small, and they are especially difficult to interpret when the populations being compared contribute substantially different amounts of follow-up time.

The same problem appears in Hulscher’s cancer framing. He highlighted a claimed 54% relative increase in cancer and then discussed possible mechanisms involving DNA fragments and formaldehyde.

But the Henry Ford dataset did not test whether DNA fragments caused insertional mutagenesis in those children.

It did not establish formaldehyde as a cancer mechanism in the cohort.

Those are proposed mechanisms layered onto an observational comparison already affected by serious follow-up and cohort differences.

The progression is recognizable:

problematic comparison → dramatic percentage → causal implication → proposed molecular mechanism

That is considerably more persuasive to an audience than the actual state of the evidence.

Exhibit 3: A mathematical model becomes a public-health fraud

Another Hulscher headline announced:

“BREAKING: 86% of PCR-Positive ‘COVID Cases’ Were Not Real Infections.”

The underlying paper is real and deserves to be described accurately.

Researchers published a 2025 analysis in Frontiers in Epidemiology comparing aggregate German PCR positivity with aggregate IgG seroprevalence. Using a least-squares model, the authors estimated a scaling factor of approximately 0.14. They interpreted this value as indicating that roughly one in seven PCR-positive individuals represented a true infection.

So the provocative 14% interpretation did not originate with Hulscher.

That needs to be acknowledged.

But the study was not a patient-by-patient validation experiment in which researchers independently evaluated thousands of PCR-positive people and discovered that 86% were false-positive cases. It modeled aggregated population-level time series. The authors themselves list the aggregated nature of the data among the study’s limitations.

Hulscher then went considerably beyond even the paper’s provocative interpretation.

His presentation characterized PCR testing as a “FRAUDULENT PCR testing SCAM” and alleged that health authorities used it to frighten populations into accepting vaccination.

His longer commentary described the episode as potentially one of the greatest public-health frauds ever committed.

But the mathematical analysis did not investigate whether governments knowingly committed fraud.

It did not investigate policymakers’ motives. It did not demonstrate that health authorities intentionally used false testing to create fear.

That is the crucial distinction.

The authors made a controversial epidemiological inference.

Hulscher added intent, fraud and coercion.

The progression becomes:

model → contested inference → fraud → deliberate coercion

The final steps are not results produced by the model.

Exhibit 4: VAERS death reports become evidence that MMR is “quite lethal”

One of the clearest examples involves the Vaccine Adverse Event Reporting System.

Hulscher coauthored a 2026 preprint examining VAERS reports after MMR and MMRV vaccination.

The paper identified 299 U.S. death reports in its analytical cohort. Of those, 193 reports with identifiable dates occurred since 1995. The authors compared those 193 reports with seven measles-associated deaths and described the vaccine-report count as 2,657% higher.

Hulscher promoted the work by saying MMR vaccines were “quite lethal” and referring to “HUNDREDS of sudden infant deaths.”

There are two major problems.

First, VAERS is not a database of events already determined to have been caused by vaccination.

CDC states plainly that VAERS accepts reports even when the reporter is unsure whether the vaccine caused the event, and that a report alone does not establish causation. CDC also warns that report counts cannot generally be used by themselves to calculate risk or frequency.

The CDC’s VAERS interpretation guide goes further: presenting VAERS reports as verified vaccine deaths or injuries misrepresents the nature of the system.

So comparing:

193 deaths reported after vaccination

with

seven deaths attributed to measles

does not create a valid vaccine-versus-disease mortality comparison.

The numerators represent different things.

There is also a problem with the phrase “hundreds of sudden infant deaths.”

The preprint reports 299 U.S. death reports overall, but says SIDS or sudden unexplained death represented 24% of the reports, not hundreds of them.

In other words, even the preprint’s own category breakdown does not support describing all or most of those reports as sudden infant deaths.

The paper itself ultimately acknowledges the central unresolved issue. Its authors call for record-linked datasets capable of assessing background mortality and determining causal relationships.

That is exactly what VAERS alone cannot do.

Yet the public message becomes:

report after vaccination → vaccine death → MMR is “quite lethal”

The noun changes.

And with it, the meaning.

What has actually been established?

At this point, the pattern does not prove that every Hulscher claim is wrong. It does not prove that he intentionally deceives his audience. It does not tell us his private motives.

What it does establish is narrower and more defensible.

In multiple examples collected within a short period, Hulscher’s public language becomes more causal, more certain or more dramatic than the evidence underneath it.

A one-patient preprint becomes “undeniable.”

A draft rejected by its originating health system becomes “powerful” evidence.

An aggregate statistical model becomes deliberate fraud.

VAERS reports become vaccine deaths.

The issue is not that scientific uncertainty exists.

The issue is what happens to that uncertainty when the story reaches the public.

Exhibit 5: A legitimate pharmacovigilance controversy becomes a “coordinated GLOBAL OPERATION”

This example requires balance because there is a legitimate underlying story.

A September 2026 paper examining Israeli pharmacovigilance records reported an anonymized dataset containing 531 hospitalization records and 67 unique deaths reported to Pfizer between March 2021 and May 2022. It found substantial numbers of myocarditis and pericarditis reports and argued that the emerging cardiac safety signal was communicated publicly only after approximately three to four months.

That deserves scrutiny.

Adverse-event surveillance should be transparent.

Safety signals should be investigated quickly.

Risk communication should be accurate.

But Hulscher moves from those propositions to something much larger.

He described the records as evidence of vaccine deaths being covered up and linked the Israeli controversy to disputes over U.S. surveillance, saying the concealment appeared to constitute a “coordinated GLOBAL OPERATION.”

Those are different evidentiary claims.

The Israeli dataset documents deaths reported to Pfizer. It does not establish that all 67 were caused by vaccination.

And showing serious pharmacovigilance problems in two countries does not, by itself, demonstrate coordination between those countries.

A coordinated international operation requires evidence of coordination: communications, directives, shared planning, common participants or some comparable connective mechanism.

Without that evidence:

two controversies remain two controversies.

The progression here is:

safety signal → communication failure → cover-up → global coordination

Each step requires additional evidence.

The rhetoric arrives before that evidence does.

Exhibit 6: “No longer met diagnostic criteria” becomes “82% clinically recovered”

The amplification can run in the opposite direction when the subject is a treatment.

A 2025 study examined 28 patients who developed ME/CFS following COVID-19 vaccination. Twenty-seven had vitamin D deficiency or insufficiency. The intervention included dietary counseling, sun-exposure recommendations and oral vitamin D supplementation.

After treatment, 23 of 28 patients—82%—no longer met the study’s ME/CFS diagnostic criteria.

That is an interesting finding.

But it was a retrospective uncontrolled study.

There was no untreated comparison group.

The authors concluded that vitamin-D replacement “may lead to symptomatic relief” and said they were preparing a randomized controlled trial to evaluate efficacy.

Hulscher’s version was:

“82% of COVID vaccine-injured patients CLINICALLY RECOVERED.”

Those statements are not identical.

Falling below a diagnostic threshold does not necessarily mean complete clinical recovery.

Twenty-eight patients selected for a specific post-vaccination ME/CFS syndrome are not equivalent to the much broader category of all “vaccine-injured patients.”

And an uncontrolled before-and-after study cannot isolate the treatment effect from spontaneous improvement, regression to the mean, concurrent care or other influences.

This is revealing because the certainty works both ways.

When preliminary evidence suggests harm, the public language may become very certain.

When preliminary evidence supports a favored intervention, certainty can appear there too.

The deeper pattern is therefore not simply pessimism.

It is asymmetric certainty.

Exhibit 7: A serious paracetamol finding becomes a broader fertility implication

This is an important example because the underlying research should not be dismissed.

The 2026 COPANA study published in Human Reproduction Open was a prospective observational cohort. Researchers enrolled pregnant women, assessed paracetamol exposure using maternal reports and urinary measurements, and examined 302 infant girls. They reported associations involving ovarian volume, follicle count, uterine volume and reproductive markers. They also found related outcomes in an independent confirmatory cohort followed into adolescence. Some analyses suggested dose-response patterns.

This is substantive research.

It deserves follow-up.

Hulscher highlighted findings including approximately:

41% smaller ovarian volume,
23% fewer follicles,
13% smaller uterine volume.

Those numbers come from the study.

The problem is not that the numbers were fabricated.

It is the implication created around them.

The paper does not establish that girls exposed prenatally to paracetamol will be infertile.

The researchers explicitly say that the observational design cannot establish causality and that the long-term clinical significance, including effects on reproductive lifespan, remains uncertain.

That qualification is central.

A difference in ovarian morphology is not the same outcome as infertility.

A smaller uterine volume at one developmental stage is not proof that a girl will later be unable to conceive.

This example demonstrates a subtler form of amplification.

Nothing needs to be literally fabricated.

The framing can do the work.

Put the largest percentages in red.

Put the widespread use of Tylenol next to them.

Leave the uncertainty about actual future fertility in smaller print—or outside the headline entirely.

Readers can be led toward a conclusion the study itself has not yet established.

Exhibit 8: A medical-record prototype becomes the “MARK OF THE BEAST”

In 2019, researchers published an experimental system for recording vaccination history using biocompatible near-infrared quantum dots delivered into the skin by microneedle patches.

The purpose was straightforward: maintaining vaccination records can be difficult in low-resource settings without reliable centralized medical databases. The quantum-dot patterns were invisible to the naked eye and readable using near-infrared imaging. The paper reported resistance to simulated sunlight exposure equivalent to approximately five years.

Hulscher presented the project as:

“MARK OF THE BEAST” mRNA implants

that would “PERMANENTLY” mark the body with quantum-dot QR codes and were planned to become biological vaccine passports controlling who could shop, dine or travel during a future pandemic.

The original research establishes none of those social-control claims.

The technology was not a conventional consumer QR code.

The experiment did not establish permanent lifetime marking.

Five-year-equivalent photostability testing is not permanence.

And the research paper did not describe a government program for controlling entry to restaurants, stores or transportation.

This is a different type of amplification:

prototype → hypothetical future application → presumed plan → dystopian certainty

Or, more simply:

could becomes will.

The vocabulary is part of the message

Now return to the words appearing across these examples:

BREAKING.
BOMBSHELL.
SHOCKING.
STUNNING.
EXPLOSIVE.
SMOKING GUN.
UNDENIABLE.
FRAUD.
SCAM.
COVER-UP.
LETHAL.
GLOBAL OPERATION.
MARK OF THE BEAST.

Scientific papers often use less exciting language:

associated with.

may.

possible.

requires confirmation.

limitations.

cannot establish causality.

Those qualifications are not defects.

They are the mechanisms by which researchers separate what they observed from what they think might explain it.

In the examples examined here, that machinery repeatedly runs in the opposite direction as the findings move into Hulscher’s public communication.

Uncertainty shrinks.

The verbs strengthen.

Relative differences become headlines.

Possible mechanisms become explanations.

Signals become conclusions.

By the end, something requiring investigation can sound like something already proven.

The McCullough commercial context

This portion requires particular care because association is not proof of motive.

There is no evidence presented here that Nicolas Hulscher personally receives revenue from The Wellness Company’s supplement sales, and this article does not allege that he does.

But there is a relevant institutional and commercial context.

Hulscher works for the McCullough Foundation and, according to the foundation, plays a major role in both its research and its communications.

Peter McCullough, the foundation’s president, has publicly disclosed another relationship.

In a published article, McCullough disclosed that he receives partial salary support and holds an equity position in The Wellness Company, which markets dietary supplements.

The Wellness Company currently markets an Ultimate Spike Detox built around the “McCullough Protocol.” Its product page says the formula is intended to remove harmful spike proteins, markets it to people experiencing problems after COVID or vaccination, and lists a one-time price of $89.99 or $80.99 under subscription pricing at the time this article was checked.

The company also sells a six-month Ultimate Spike Detox + Testing System for $539.99, including before-and-after antibody testing. Notably, the company’s own page cautions that a change between the two tests does not, by itself, prove that the product caused the change.

That caveat is revealing in the context of this article.

The commercial relationship does not establish why Hulscher communicates as he does.

It does, however, create a financial context readers should know about when vaccine-injury and persistent-spike narratives circulate within a professional ecosystem whose most prominent figure has an equity interest in a business marketing products to people concerned about those problems.

That is not guilt by association.

It is disclosure.

Readers can decide what significance to attach to it.

Some of the signals are real

Any fair examination of Hulscher’s work has to acknowledge something else.

Real vaccine adverse effects exist.

Real pharmacovigilance failures can occur.

Government agencies make mistakes.

Drug companies deserve scrutiny.

Researchers should challenge assumptions.

Unpopular hypotheses should not be dismissed merely because they are unpopular.

And a paper does not become wrong because Nicolas Hulscher shares it.

The problem arises when the distinctions necessary for evaluating evidence collapse:

signal versus causation;

report versus confirmed injury;

association versus mechanism;

hypothesis versus established fact;

institutional failure versus coordinated conspiracy.

Skepticism works only when it is applied consistently.

Including to the people selling skepticism.

Less than a week

Perhaps the most striking thing about this examination is how little archival digging it required.

The examples above came from posts appearing within roughly a September 12–18, 2026 window, along with the papers and articles those posts referenced.

They are not presented as a statistical analysis of Hulscher’s entire output.

But within that short span we encountered claims about:

vaccination and childhood chronic disease;

vaccination and “turbo cancer”;

PCR testing and public-health fraud;

MMR and supposedly massive vaccine mortality;

an alleged coordinated international vaccine-death cover-up;

vitamin D and an 82% “clinical recovery” rate;

prenatal Tylenol and reproductive development;

and quantum-dot technology portrayed as the “Mark of the Beast.”

At some point the individual topic stops being the whole story.

The recurring transformation becomes the story.

What happens when you challenge it?

During this review, I challenged Hulscher’s claims on X.

He blocked my account.

That screenshot establishes only that fact. It does not establish that Hulscher blocks every critic, that he systematically removes dissent, or that the block had any particular motive.

It should not be stretched beyond what it proves.

But scientific claims themselves should be capable of surviving scrutiny whether or not their authors choose to engage with critics.

If a claim is supported by strong evidence, disagreement does not weaken that evidence.

It creates an opportunity to show it.

Read the paper. Then read the headline.

The most useful response to this kind of scientific communication is surprisingly simple.

Open the study.

Find out who was actually studied.

Find the control group.

Look at the raw numbers.

Look at the absolute difference as well as the relative difference.

Check whether follow-up was comparable.

Ask whether the study was randomized, prospective, retrospective or purely observational.

If a surveillance database is being used, determine whether its reports have been causally adjudicated.

Find the limitations section.

Ask whether the proposed mechanism was measured or merely hypothesized.

Then read the authors’ conclusion.

Only after that should you return to the headline.

And ask:

How much did the certainty grow between the paper and Nicolas Hulscher’s post?

In the examples examined here, the answer is repeatedly:

a great deal.

That—not any single vaccine, paper, drug, statistic or controversy—is the central finding.

The signal may be scientific.

The fear is added later.

Resources and primary sources

• Nicolas Hulscher / McCullough Foundation: The foundation’s own biography describes Hulscher’s MPH, his epidemiology role and his responsibilities for scientific research, public communications and social media. McCullough Foundation — Nicolas Hulscher biography

• Henry Ford vaccinated-vs.-unvaccinated draft: Henry Ford Health explains why it says the 2020 draft failed internal review, including cohort imbalance and dramatically shorter follow-up among unvaccinated children. Henry Ford Health — Vaccine Study Fact Check

• Cardiac intimal sarcoma / “turbo cancer”: Hulscher and colleagues’ September 2026 Zenodo preprint involving one 35-year-old patient. Zenodo — Fatal Cardiac Intimal Sarcoma case report

• PCR/IgG analysis: The original 2025 Frontiers in Epidemiology paper behind the 14%/86% claim. Frontiers — Calibration of PCR by IgG tests in Germany

• MMR/MMRV death-report preprint: The 2026 Zenodo preprint based on VAERS reports, including its 299 U.S. reports, 24% SIDS/sudden-unexplained-death category and 2,657% comparison. Zenodo — Deaths Following MMR and MMRV Vaccination in the United States

• How VAERS should be interpreted: CDC’s current guidance explains that reports are accepted regardless of whether vaccination caused the event and cannot alone establish causality or rates. CDC — About VAERS

• CDC VAERS data warning: CDC WONDER explicitly cautions that treating VAERS reports as verified vaccine deaths or injuries misrepresents the database. CDC WONDER — VAERS interpretation guidance

• Israeli pharmacovigilance: The September 2026 EXCLI paper analyzing 531 hospitalization records and 67 unique deaths reported to Pfizer and discussing the myocarditis communication timeline. EXCLI Journal — Israeli pharmacovigilance study

• Vitamin D / post-vaccination ME/CFS: The PubMed record for the retrospective 28-patient study in which 23 patients no longer met ME/CFS diagnostic criteria after vitamin-D replacement therapy. PubMed — Vitamin D replacement study

• Paracetamol / COPANA: The peer-reviewed 2026 prospective cohort examining prenatal paracetamol exposure and ovarian and uterine development. Human Reproduction Open — COPANA study

• Quantum-dot vaccination records: The original 2019 Science Translational Medicine research describing near-infrared quantum dots delivered through microneedle patches for vaccination recordkeeping. PMC — Biocompatible near-infrared quantum-dot vaccination record study

• McCullough financial disclosure: McCullough’s published disclosure states that he receives partial salary support and holds an equity position in The Wellness Company. Published McCullough disclosure

• The Wellness Company — Ultimate Spike Detox: Current marketing, pricing and claims surrounding the McCullough Protocol product. Ultimate Spike Detox

• Six-month detox/testing program: Current $539.99 program combining supplements and before-and-after spike-antibody testing; the page itself cautions that changes between tests do not prove the supplement caused them. Ultimate Spike Detox + Testing System

The reasonable temptation is to answer every post. It is also a trap.

One says ivermectin cut COVID deaths by 47 percent. Another says a new mRNA flu vaccine harmed three-quarters of the people in its trial. Another compares the mercury concentration in a multidose flu-vaccine vial with the threshold for hazardous waste. Then come the vaccinated-versus-unvaccinated study, the graph claiming vaccines never saved anyone, the microneedle “tattoo,” the graphene cloud-seeding patent, the injected 5G antenna and the nurse who says COVID patients were murdered for money.

Each claim could consume an article. By the time the tenth rebuttal is finished, twenty more posts have appeared.

That is why Valerie Anne Smith’s X account, @ValerieAnne1970, is more useful as a study of misinformation flow than as a checklist of false statements. The account does not ask followers to accept one isolated claim. It surrounds them with a continuous story: Smith says she recovered after conventional medicine failed her; official medicine cannot be trusted; every alarming fragment confirms that distrust; and the treatments, services and products she accepts can be found through the next link.

The public evidence does not tell us whether Smith believes every claim she posts, what she earns from any particular link or what ultimately motivates her. It does show what the account publishes, how the claims are constructed and where the audience is sent. That is enough to examine the system without pretending to read her mind.

Ten posts captured on August 31, 2026. Their subjects vary, but the pattern repeats: start with something real, remove the context that limits it, add a catastrophic interpretation, assign a villain and move on before scrutiny can catch up.

The account begins with its strongest evidence: her own life

Smith has publicly described a long struggle with anorexia and mental illness, followed by a dramatic recovery after adopting an animal-based ketogenic diet. Her public materials connect that experience to a 2023 case series of three patients with severe anorexia. All three gained substantial weight and reported sustained improvement after adopting high-fat, animal-based ketogenic diets.

Nothing in this article disputes the reality of Smith’s suffering or the improvement she reports. A person is entitled to tell the story of her own recovery. The criticism begins only when that experience is extended into broad medical claims, or used to supply authority on subjects the experience cannot answer.

That is evidence worth taking seriously. It is also limited evidence, as the paper’s authors acknowledged. The patients selected the diets themselves. The report was retrospective. There was no control group, and the researchers could not determine whether the improvements resulted from ketosis, nutritional replenishment, psychological factors or some combination. They concluded that ketogenic diets might help some patients and deserved further study.

The scientific paper says possible treatment deserving study. Smith’s public bio and interview titles repeatedly say cured, completely healed and fixed. Her Linktree contains titles such as “How She Used Carnivore to COMPLETELY HEAL,” “40 Years of Mental Illness Healed in ONE Year,” and “How the Carnivore Diet FIXES Depression.”

That change in language matters. A three-person case series can generate a hypothesis. It cannot establish that an animal-based ketogenic diet cures eating disorders or psychiatric conditions generally. Nor can participation in that case series confer expertise in vaccine epidemiology, antiviral drugs, toxicology or radio-frequency engineering.

But social media does not assign authority by field. It assigns authority to a person who has earned attention and trust.

Researchers have repeatedly found that anecdotes can exert more influence than their evidentiary value warrants. In four experiments involving more than 4,100 participants, even a single negative anecdote changed medical judgments despite participants also receiving stronger statistical evidence or a decision aid. The story does not have to disprove the data. It can simply become the part people remember.

In the structure of the account, Smith’s recovery story therefore does two jobs. It tells people what she says happened to her, and it creates a transferable sense of credibility: I found what doctors missed, so I can help you see what they are hiding elsewhere.

The first proposition may be true. The second does not follow from it.

Six fact-checks—and the technique behind each one

The point of sampling these posts is not to declare every sentence false. Several contain real facts. The problem is what happens between the fact and the conclusion.

1. Ivermectin: a study count is made to impersonate a scientific consensus

Two screenshots repeat figures from c19early.org: 106 controlled studies, 54 randomized trials and a 62 percent lower risk with early ivermectin treatment. The second turns the same numbers into an AI-made graphic and adds a larger allegation: Anthony Fauci supposedly lied about ivermectin so the government could authorize mRNA vaccines, and “people died.”

The number of studies is not the same thing as the weight of the evidence. The site pools unlike material: randomized and nonrandomized designs, different patient populations, different treatment stages, different doses, different co-treatments and different outcomes. A small observational study of viral clearance and a large placebo-controlled trial of hospitalization or recovery do not become equivalent because both can be plotted as dots. Nor does a pooled association establish a plot to suppress treatment.

When reviewers restricted the evidence according to study reliability, the dramatic benefit disappeared. The 2022 Cochrane review found no evidence supporting ivermectin for COVID-19 treatment or prevention. In the double-blind ACTIV-6 randomized trial, 1,591 participants completed follow-up and ivermectin did not significantly improve recovery; a later trial of a higher dose for a longer period also found no benefit.

The move here is aggregation without commensurability. “There are many studies” becomes “the treatment works,” and “the treatment works” becomes “officials knowingly killed people.” The final accusation is not a finding of any of the studies in the graphic.

2. MFLUSIVA: a side-effect tally is presented as an injury rate

The mRNA flu-vaccine graphic has a stronger factual core. MFLUSIVA was approved by the FDA on August 5, 2026 for adults 50 and older. The approval was traditional for ages 50–64 and accelerated for ages 65 and older, with a confirmatory trial required for the older group. Its pivotal study did report more short-term reactions than the standard-dose flu-vaccine comparator.

But “75%+ of participants had at least one adverse reaction” does not mean three-quarters were seriously harmed. In the FDA’s clinical review, 75.7 percent of a 3,015-person MFLUSIVA solicited-safety subset reported at least one solicited reaction during the first seven days. That category included injection-site pain, redness, fatigue, headache, muscle pain, joint pain, chills, nausea and fever. Most were mild or moderate. Grade 3 systemic reactions were higher—5.5 percent versus 0.9 percent for the standard flu vaccine—which is a legitimate reactogenicity difference to discuss. It is not the same claim as “75 percent injured.”

Calling the comparator a fake placebo is also misleading. This was an active-comparator trial asking how the new product performed against an existing flu vaccine, not whether any flu vaccine beats saline. That comparison produced useful information: in roughly 40,700 participants, laboratory-confirmed influenza-like illness occurred in 2.0 percent of the MFLUSIVA group and 2.8 percent of the standard-vaccine group, a relative vaccine efficacy of 26.6 percent.

The death claim loses its denominator too. There were 40 deaths among 20,350 MFLUSIVA recipients and 34 among 20,353 comparator recipients—0.2 percent in each group—and investigators assessed none as vaccine-related. After reviewing the case narratives, FDA said causation was unlikely. In its pooled safety analysis, however, FDA found more deaths recorded without a specified cause, or as sudden death, in the mRNA-vaccine groups. All-cause mortality remained balanced, but the 29 affected mRNA recipients had no autopsies and there was no systematic cause-of-death adjudication. FDA therefore said residual uncertainty remained. Serious adverse events in the pivotal trial were 2.2 percent and 1.9 percent, respectively. FDA required confirmatory efficacy work for the accelerated approval and accepted a separate postmarketing safety-surveillance commitment. That is a reason for continued monitoring, not evidence that trial deaths were concealed.

The move is denominator removal. A common short-term reaction becomes “an adverse reaction”; six additional deaths among more than 20,000 people become “there were deaths in the mRNA group”; the comparator, the equal rounded death rate and the efficacy result disappear.

3. Mercury: two valid concentrations are used in an invalid comparison

The flu-shot post claims that a vaccine containing 50,000 parts per billion of mercury is 250 times more toxic than material the EPA classifies as hazardous waste at 200 ppb.

The arithmetic can be reconstructed. Some multidose influenza vaccines contain 25 micrograms of mercury from thimerosal in a 0.5 mL dose, which can be expressed as 50,000 micrograms per litre. Federal hazardous-waste rules list mercury at 0.2 milligrams per litre, or 200 micrograms per litre.

What the post hides is that the EPA value is not a human injection-safety limit. It is the threshold in a leaching test used to decide whether discarded solid waste could release contaminants into groundwater. The vaccine figure is the concentration of a measured dose of ethylmercury used as a preservative. Toxicology depends on the chemical species, total dose, route and how the body processes it—not concentration alone. The FDA distinguishes ethylmercury from methylmercury, and the CDC notes that most single-dose flu shots and prefilled syringes contain no thimerosal.

The move is unit laundering. Two numbers share a unit, so the post treats them as if they answer the same safety question. They do not.

4. “22X”: the headline misstates even the paper it is promoting

One supplied post declares: “Vaccinated Children Suffer 22X Higher Chronic Disease Rates.” Its own text then describes a “peer-reviewed reanalysis” of a large vaccinated-versus-unvaccinated birth cohort.

The underlying 2025 paper does not report that the overall chronic-disease rate was 22 times higher. It says all 22 categories in the source table had higher proportions in the vaccinated group. Elsewhere it highlights a 5.49-fold figure for a cluster of neurodevelopmental diagnoses and a roughly threefold difference in the draft’s 10-year chronic-disease-free analysis. Neither is 22-fold. The account has turned the number of categories into a multiplier.

The deeper problem is methodological. The reanalysis relies heavily on proportional comparisons between 16,511 vaccinated children and 1,957 unvaccinated children even though the vaccinated children were followed longer and had far more medical encounters. More observation creates more opportunities for diagnosis. The unpublished Henry Ford draft entered into the Senate hearing record itself reported roughly seven annual encounters in the vaccinated group and two in the unvaccinated group, a major risk of ascertainment bias. Its authors used patient-time rates and survival models in an attempt to address unequal follow-up. The reanalysis argues that the greater number of visits reflects vaccine-caused illness, but that assumption cannot be used to prove the conclusion it is supposed to test. Henry Ford Health says the draft failed internal review because of serious data and methodological problems.

Its publication label deserves context too. One of the paper’s three authors is also one of the journal’s editors in chief. The paper says that another editor temporarily replaced him for its review and obtained two reviewers. That does not settle whether the analysis is right or wrong. It does make “peer reviewed” a poor substitute for examining the data and method.

The move is credential compression. A paper published in a journal becomes “the largest study ever”; a contested secondary analysis becomes a clean conclusion; 22 measured categories become “22X.”

5. “Vaccines never saved us”: mortality is substituted for disease incidence

The historical graph says mortality from scarlet fever, whooping cough, measles, diphtheria and smallpox had already fallen by more than 98 percent before the relevant vaccines, thanks to sanitation, nutrition, clean water and better living conditions.

Part of that history is real. Better nutrition, less crowding, antibiotics and improved supportive care reduced the chance that an infected person would die. But a falling death rate does not show that vaccines failed to prevent infections, outbreaks, hospitalizations, disability or later deaths. It is possible for case fatality to decline while a disease continues circulating widely.

Measles makes the substitution easy to see. Before the vaccine was introduced in England and Wales in 1968, there were 160,000 to 800,000 notifications a year and roughly 100 acute deaths. In the United States, the decade before the vaccine brought an estimated 3 million to 4 million infections, 48,000 hospitalizations, 1,000 cases of encephalitis and 400 to 500 deaths each year. The vaccine-era collapse in incidence cannot be read from a chart designed to show mortality alone. Smallpox is more decisive still: the disease was eradicated through coordinated surveillance and vaccination, with the WHO declaring eradication in 1980.

The move is outcome substitution. The chart answers “Why did infected people become less likely to die?” and is presented as if it answered “Did vaccines prevent disease?”

6. Graphene and 5G: a real paper and a real patent become a secret bodily network

Three posts form a miniature misinformation chain.

The first announces that a Bill Gates microneedle mRNA patch will permanently tattoo people with modified mRNA and “quantum-dot QR codes.” This appears to splice together separate technologies. A real 2019 Science Translational Medicine paper tested near-infrared quantum-dot patterns delivered by dissolvable microneedles as a possible vaccination record. The researchers used a polio vaccine in rats and tested pattern visibility in cadaver skin. It was not an approved product, not an mRNA vaccine, not a QR code connected to a network and not evidence that such a patch “is here.” The Gates Foundation helped fund the research; funding is not ownership of a secret deployed device.

The second post displays US patent 2022/0002159 A1 and says “they’re patenting the skies” and pumping graphene oxide into clouds. The patent is real, and the application later became an active patent. It describes a reduced-graphene-oxide and silica composite intended to promote ice nucleation. A patent establishes a claim to an invention. It does not show where, whether or at what scale the invention has been manufactured or deployed, much less prove a covert program.

The third post assumes the missing steps: graphene oxide is already “inside you,” 5G switches it on, it amplifies signals, links a person to nanotechnology and turns the body into a node on the Internet of Things. No evidence is supplied for any of those transitions. UK regulators have stated that authorized COVID-19 vaccines contain no graphene oxide, and vaccine ingredients and batch specifications are controlled as part of authorization.

The move is narrative chaining. Real research establishes that a technology has been studied. A real patent establishes that somebody sought legal rights to an invention. The next post treats secret deployment as settled. The final post places the material inside the follower’s body and gives it a science-fiction function. Each unsupported step borrows credibility from the true fact before it.

Testimony is repeatedly promoted into proof

The final screenshot quotes a nurse who says she saw no patients die from COVID, only from remdesivir and ventilators, and calls hospital care “murder for money.” Whatever that individual believes she witnessed, one person’s recollection cannot establish the cause of death for a country, the efficacy of a drug or a coordinated murder scheme.

The randomized evidence on remdesivir is more complicated than either “miracle” or “murder.” The placebo-controlled ACTT-1 trial found faster recovery and fewer serious adverse events with remdesivir, without a statistically conclusive mortality reduction. The larger WHO Solidarity trial found no significant overall mortality benefit, though its updated meta-analysis suggested a modest benefit among patients not already ventilated. Mechanical ventilation was used for people in severe respiratory failure; the fact that many ventilated patients died does not show the machine caused the illness that required it.

The post replaces those questions with a witness and a villain. “I saw it” is emotionally potent. It is not a controlled comparison.

This is why Smith’s own recovery story matters beyond diet. The account’s structure encourages followers to treat testimony as a superior way of knowing: the patient knows what doctors missed; the nurse knows what hospitals hid; the whistleblower knows what regulators concealed. Formal evidence is not merely judged inadequate. It is recast as a tool of the suspected conspiracy.

The recurring pattern is more important than the topics

Ivermectin, flu vaccines, mercury regulation, childhood health records, mortality graphs, microneedles, patents, cloud seeding and 5G look like unrelated subjects. They are held together by a small set of moves:

• Borrow authority from a true fragment. Start with a personal recovery, a real study, an FDA document, a concentration, a historical graph or a patent number.

• Remove the limiting context. Drop the denominator, study quality, active comparator, chemical form, route of exposure, time at risk or distinction between mortality and incidence.

• Convert uncertainty into intent. A disputed result is not merely wrong; somebody lied, hid the cure or killed patients for money.

• Make disagreement self-incriminating. Doctors, regulators and researchers who reject the conclusion are treated as members of the system under accusation.

• Repeat the conclusion in new wrappers. A website chart becomes an AI image, a quote card, a video clip and a fresh all-caps warning.

• Move before the claim can settle. The volume makes careful checking feel permanently behind.

The account’s recurring 2,094-percent autism statistic follows the same route. The number can be traced to a non-peer-reviewed 2016 document by Richard Lear called “The Root Cause in the Dramatic Rise of Chronic Disease”, which built a custom “Generation Growth” measure by mixing incidence and prevalence, using data from different countries and extrapolating shorter trends across a 25-year period. The author acknowledged that the results were over-precise. On X, the method disappears and the giant percentage survives.

This is content laundering: paper to chart, chart to screenshot, screenshot to conclusion, conclusion to memory.

Repetition turns alarm into background knowledge

The two ivermectin posts in the collage are not redundant. They show the mechanism in real time. The first presents the graph. The second repeats the same source numbers, adds an AI-generated visual and attaches the moral verdict: officials lied and people died.

Repetition changes how information feels. In a 2024 field experiment with professional content moderators, a single prior exposure made repeated headlines more likely to be judged true, including false ones. Prompting people to consider accuracy when they first encountered a claim eliminated the effect.

Habit matters too. A 2023 study in the Proceedings of the National Academy of Sciences found that the 15 percent most habitual sharers accounted for roughly 30 to 40 percent of the false news shared in the researchers’ experiments. Continuous misinformation does not require every follower to become a committed believer. It needs enough people to become accustomed to reacting and reposting.

Outrage supplies the energy. A large study of Twitter rumor cascades found that false news was 70 percent more likely to be retweeted than true news. Later work in Science found that misinformation evokes moral outrage and that outrage increases sharing, including sharing without opening the linked material.

The language in the screenshots is built for that response: “They lied. People died.” “Murder for money.” “They may have injected the antenna.” The follower is offered a role—protector of children, defender of patients, opponent of corruption—before being offered time to examine the source.

Once the world is divided into victims and villains, a correction from a doctor, regulator or public-health agency arrives pre-discredited. Its disagreement is treated as evidence that the institution belongs to the corrupt side.

That is how the stream protects itself. It does not have to rebut every future correction. It gives followers a ready-made reason that corrections should not count.

The bio completes the flow

Smith’s Linktree, captured on August 31, 2026, places commercial and health-related destinations at the top: RightMD, Primal Kitchen, Frontier Pharmaceutical’s DioxiCare chlorine-dioxide products and “Daily Detox Solutions.” The page also contains a Shop tab, a payment link, product discount codes and a long list of interviews repeating her recovery story. Her connected platforms include X, Instagram, Facebook and YouTube.

RightMD advertises telehealth services involving cancer, COVID-19, weight loss and other conditions, along with ivermectin and natural supplements. Frontier Pharmaceutical offers chlorine-dioxide products and advertises an affiliate program. Other links promote food, supplements, salt, skin products and sleep products. Some use personalized discount codes.

The destinations complicate the account’s simple opposition between “nature” and the pharmaceutical industry. Its X bio says nature provides cures and vaccines cause harm, while its link page directs followers to a telehealth medication service and a company called Frontier Pharmaceutical. The dividing line is not natural versus pharmaceutical. It is between products and providers accepted by the account and institutions the account has taught followers to distrust.

The commercial pathway becomes especially important when a post makes a safety claim. In an August thread, Smith said the FDA “no longer claims that chlorine dioxide is unsafe” because warning pages had disappeared from its website. The thread described drinking, nebulizing and injecting chlorine dioxide, then linked to DioxiCare products.

Removing or relocating a webpage is not a regulatory finding. FDA enforcement material still describes ingested sodium-chlorite products that generate chlorine dioxide as unapproved and capable of severe harm. Federal authorities described “Miracle Mineral Solution” as producing chlorine dioxide, a powerful bleach used in industrial treatment. Chlorine dioxide has legitimate antimicrobial uses at controlled concentrations, including certain food-processing applications with strict residual limits. That does not make it an approved treatment for cancer, autism or COVID-19.

The product URL in Smith’s thread contains a referral identifier. That does not establish how she is compensated or whether a purchase pays her. It does establish that a broad medical claim, an attack on the regulator and a trackable product link appear in the same path.

The public flow is visible:

Personal recovery → borrowed authority → alarming claim → repeated exposure → outrage → institutional distrust → alternative treatment → product or service link

The product link is not sitting separately from the message. It is where the message has prepared the follower to go.

A 2026 systematic scoping review in JAMA Network Open examined 12 articles about influencer promotion of prescription drugs. It found recurring problems with misinformation, inconsistent disclosure and audiences’ difficulty recognizing promotional intent when marketing was embedded in personal narratives. The authors also described the evidence base as small and fragmented. The finding does not establish Smith’s motives or compensation; it explains why clear disclosure matters when personal health stories and product pathways overlap.

Are we making the opposite mistake?

This is where it helps to pause the argument and question our own reasoning. A critical article can repeat the same error it condemns if it turns fragments into a verdict about a person.

Smith’s recovery may be exactly as she describes it. Early or unconventional research can identify ideas that established medicine has missed. Regulators and researchers can be wrong. An active-comparator trial does not answer every question a saline-controlled trial would answer. A real patent can eventually lead to a real product. A referral code may be used for tracking, a discount or compensation, and the public URL alone does not tell us which.

Grant all of that. The central findings still hold because they do not depend on diagnosing Smith, denying her recovery or proving a hidden motive. They depend on public, checkable acts: a 22-category result is posted as “22X”; a waste-leaching threshold is presented as an injection limit; solicited reactions are framed as mass injury; a research prototype becomes a deployed control system; and a regulator’s missing webpage becomes proof that an unapproved therapy is safe.

Four questions keep the standard consistent:

• Would we accept this reasoning if it supported the opposite conclusion? A patent, testimony or raw proportion would still need the missing steps.

• Are we claiming more than the evidence can establish? This sample shows a repeated misinformation pattern. It does not establish what Smith privately believes, what every post says or what she earns.

• What evidence would change the conclusion? Reliable trials supporting the medical claims, proof of the alleged deployment, corrections that preserve denominators and study limits, or clear disclosures separating testimony from promotion would all matter.

• Are we criticizing the claim or the person’s vulnerability? The proper target is the public leap from experience to authority, not the suffering that preceded the experience.

Uncertainty deserves room. It does not make unsupported conclusions and supported findings interchangeable.

Why correcting one post is not enough

A conventional fact-check treats misinformation as a defective statement: identify the claim, present the evidence, render a verdict. That work is necessary. The examples above show why it is insufficient on its own.

When followers have an established relationship with an influencer, a correction arrives from outside that relationship. A 2026 study of Community Notes on X found that community fact-checking reduced reposting but was less effective among people who had previously interacted with the misleading post’s author. It was least effective with health and political misinformation. Trust in the person can blunt the note attached to the post.

The better question is not only, Is this claim wrong? It is, What keeps carrying me to the next one?

Before sharing health content, follow it backward and forward:

• What is the original source—not the screenshot, video or person repeating it?

• Did that source measure what the post says it measured?

• What denominator, comparison group, follow-up period or definition has been left out?

• Has a possibility, association, case report, patent or early trial been converted into proof of deployment or a cure?

• Is a personal recovery being used to claim authority in an unrelated field?

• Can the account’s central belief ever be proved wrong, or is every correction absorbed into the conspiracy?

• Where does the bio send the audience, and are commercial relationships clearly disclosed?

The value of examining Smith’s account is not that it contains the worst claim on the internet. It is that the pattern is unusually visible.

The story supplies trust. The real fragment supplies plausibility. Context removal supplies the alarming conclusion. Repetition supplies familiarity. Outrage supplies reach. Institutional suspicion blocks the exits. The bio supplies somewhere to act—and sometimes somewhere to spend.

The individual post is only a frame. The influence is in the flow.

Method note: This article is based on publicly accessible profile material, indexed X posts, Smith’s Linktree, the supplied Linktree capture and ten supplied X screenshots captured on August 31, 2026, together with the linked scientific and regulatory sources. Clinical claims were checked primarily against regulator reviews, trial reports, systematic reviews, and original publications.

Advocacy sites, social posts, and commercial pages are used to document what was claimed or sold, not as proof that the claims are medically correct. The sample was selected to represent recurring claim types, not to quantify every post on the account. The article analyzes public claims and content structure; it does not infer private medical facts, undisclosed income, or intent.

Resources

The sources below are grouped by how they were used. Regulatory reviews, original studies and systematic reviews carry the factual analysis. Social, advocacy and commercial pages document what the account claimed, repeated or promoted; they are not treated as proof that the medical claims are correct.

Medical, scientific and regulatory sources

• Ketogenic diet treatment of anorexia nervosa: a three-patient case series (2023)

• Cochrane: Ivermectin for preventing and treating COVID-19

• JAMA: ACTIV-6 randomized trial of ivermectin for mild-to-moderate COVID-19

• JAMA: Higher-dose, longer-duration ivermectin trial

• FDA: MFLUSIVA approval letter, August 5, 2026

• FDA: MFLUSIVA clinical review

• Electronic Code of Federal Regulations: Toxicity-characteristic limits for hazardous waste

• FDA: Thimerosal and vaccines

• CDC: Thimerosal in flu vaccines

• IJVTPR: Vaccinated-versus-unvaccinated child-health reanalysis (2025)

• U.S. Senate hearing record: Henry Ford vaccinated-versus-unvaccinated draft and testimony

• Henry Ford Health: Institutional response to the vaccinated-versus-unvaccinated study

• UK Health Security Agency: Measles, the Green Book chapter

• CDC: History of measles

• World Health Organization: Smallpox

• Science Translational Medicine: Near-infrared quantum-dot patterns delivered by microneedles

• U.S. patent record: Reduced-graphene-oxide/silica composite for ice nucleation

• UK Medicines and Healthcare products Regulatory Agency: COVID-19 vaccines and graphene oxide

• New England Journal of Medicine: ACTT-1 final report on remdesivir

• WHO Solidarity trial: Final results and updated meta-analysis of remdesivir

• FDA enforcement record: Chlorine-dioxide “Miracle Mineral Solution” case

• FDA food-contact notification: Controlled chlorine-dioxide use and residual limits

Research on misinformation and influence

• Four experiments on the influence of anecdotes in medical decisions

• PNAS Nexus: Repetition and belief among professional content moderators

• PNAS: Habitual sharing and false news

• Science: The spread of true and false news online

• Science: Moral outrage and the sharing of misinformation

• JAMA Network Open: Systematic scoping review of influencer promotion of prescription drugs

• Nature Communications: Community Notes, prior interaction and misinformation sharing

Account, claim and commercial materials

• Valerie Anne Smith’s X account

• Valerie Anne Smith’s Linktree

• C19Early ivermectin meta-analysis page cited in the posts

• IJVTPR editorial team

• Richard Lear’s non-peer-reviewed “Generation Growth” document

• Archived thread presenting chlorine dioxide as a treatment and linking to products

• RightMD

• Frontier Pharmaceutical

The Joe Tippens story: How one remission became a cancer protocol

Joe Tippens entered remission after adding a dog dewormer to his cancer treatment. His story is remarkable. It is not proof of a cure—or evidence for the protocols and sales pitches built around it.

The Joe Tippens story has everything a medical miracle story needs: an aggressive cancer, a devastating prognosis, a chance conversation with a veterinarian, a cheap drug intended for dogs, and scans that went from riddled with disease to “all clear.”

It is easy to understand why the story spread. Small-cell lung cancer is fast-growing and often deadly. The American Cancer Society reports a five-year relative survival rate of 4 percent for people diagnosed after it has spread to distant parts of the body. Those population figures cannot predict an individual outcome, and they do not verify Tippens’s own estimate that he had only months to live. They do show why his reported survival is extraordinary.

But the viral version leaves out the fact that matters most. Tippens did not take fenbendazole in isolation. Before the remission he had received chemotherapy, chest radiation and preventive brain radiation. When his cancer reportedly spread, he entered a clinical trial. He began fenbendazole and several supplements during that same period.

That is not a minor complication. It is the difference between a powerful personal story and evidence that a drug works.

The timeline in Tippens’s own words

The detailed chronology comes from Tippens’s website. It is his account, not an independently published case report, and the site does not provide the complete medical record, pathology, scan files, trial identifier or treatment protocol needed to audit it. That does not mean the story is false. It means every medical claim drawn from it has to be described for what it is: self-reported.

Tippens says the veterinarian told him about an unnamed Merck Animal Health scientist who had supposedly cleared stage IV brain cancer after taking a canine medicine. The scientist is not identified on his site, and no medical documentation is linked. That story is therefore another anecdote nested inside the better-known anecdote.

The part we can say plainly is that Tippens lived far beyond the prognosis he says he received. The part we cannot honestly say is that fenbendazole cured him.

Checkpoint

Tippens’s remission can be genuine and fenbendazole can still be unproven. A striking outcome establishes what happened after several treatments; it does not identify which treatment caused it.

---

The treatment that disappears from the miracle retelling

Tippens’s own timeline places fenbendazole beside a clinical-trial drug, not before it or after it. He says he agreed to the trial in January 2017 and began his alternative regimen in the third week of that month. By May, his scan was clear.

His site does not name the trial drug. A joint American Cancer Society and American Society of Clinical Oncology explainer describes the well-known lung-cancer story as involving an immunotherapy treatment that was approved for that type of cancer. Without the trial number and medical record, it is safer not to guess which drug he received.

What matters is that immunotherapy can sometimes produce unusual and durable responses in small-cell lung cancer. In pooled results from two pembrolizumab studies, 16 of 83 previously treated patients responded, including two complete responses. An estimated 61 percent of responders had a response lasting at least 18 months. That study does not prove Tippens received pembrolizumab, and it does not prove immunotherapy caused his remission. It demonstrates why dismissing the clinical-trial treatment is unjustified.

Tippens reached the opposite conclusion. After his first clear scan, he wrote that he was certain the trial drug was not responsible. He later described himself as the trial’s exceptional outlier and took that as confirmation that his alternative regimen had done the work. But an outlier is not an explanation. Every effective cancer treatment has nonresponders, typical responders and rare exceptional responders. Finding the exceptional case does not tell us why it occurred.

There is another problem. Tippens says he concealed the fenbendazole and supplements from his oncologist until the trial ended and he had already had multiple clear scans. That decision may be understandable from the perspective of a frightened patient worried about losing access to a trial, but it destroyed the chance to document the experiment prospectively. His care team could not assess interactions, attribute adverse effects, measure exposure or incorporate the products into the trial record while the response was happening.

The result is permanent uncertainty. Fenbendazole may have helped. It may have contributed alongside the trial drug. It may have done nothing. The supplements may have mattered, or not. The prior chemotherapy and radiation may have played a role in later disease control. The available record cannot separate those possibilities.

How a personal experiment became “the protocol”

The regimen now called the Joe Tippens Protocol is not even a stable description of what Tippens took before his first clear scan.

His original 2017 routine was one gram of Panacur C granules—approximately 222 mg of fenbendazole—on three consecutive days each week, followed by four days off. He also took CBD, a curcumin product and vitamin E. That already creates a basic attribution problem: four nonstandard products were added during a cancer clinical trial, so the effect of any one ingredient cannot be isolated.

The instructions then changed. In a July 2020 update, Tippens said daily fenbendazole was now prudent for virtually everyone and removed vitamin E as a required ingredient. His current protocol page, updated in December 2025, specifies 222 mg every day and adds four branded “ONCO ADJUNCT” pathways containing combinations of CBD, frankincense, C60, curcumin, quercetin, berberine, green-tea compounds, resveratrol, fisetin and beta-glucans.

None of those versions has been tested as a complete regimen in a controlled human cancer trial. No study establishes that 222 mg is an effective human cancer dose, that daily use is better than three days on and four days off, that lifelong use improves survival, or that the growing supplement stack adds benefit. There are no validated cancer-specific adjustments, treatment-duration rules, laboratory-monitoring requirements or stopping criteria.

This evolution matters. If the original story is offered as evidence for the protocol, only the original combination can even claim to have coincided with the remission. A later daily schedule and newly added products cannot inherit evidence from an earlier anecdote.

Tippens repeatedly says he is not a doctor and is only telling his story. That disclaimer deserves to be quoted accurately. So does the rest of the page. On the same site he recommends taking the regimen with chemotherapy, radiation or immunotherapy, advocates daily use for virtually everyone, and says he plans to continue for life. A disclaimer does not turn dosing instructions and broad recommendations back into autobiography.

A protocol is more than a recipe

In medicine, a protocol defines who should receive an intervention, the exact formulation and dose, how long it is used, what else can be given with it, how benefit and harm are measured, and when treatment must be changed or stopped. Those decisions are built from pharmacology and clinical evidence.

The Joe Tippens Protocol has a list of ingredients, but not that evidentiary foundation. It is better described as a personal regimen that became a public recipe.

One statement on Tippens’s blog illustrates the problem. He wrote that fenbendazole had already undergone human clinical trials as a dewormer and had been deemed safe for human consumption. That is incorrect for fenbendazole. The American Cancer Society states that the FDA has never approved fenbendazole for any human use. A 2024 scientific review likewise noted that its human pharmacokinetics and safety are poorly documented and called for trials to determine anticancer effects, optimal doses, schedules and tolerability.

Other members of the benzimidazole family, including albendazole and mebendazole, are used in human medicine. That does not make fenbendazole approved or establish an equivalent dose. Drugs in the same class can differ in absorption, metabolism, toxicity and activity.

Checkpoint

“It worked for me” is a reason to investigate. “Here is the dose for everyone” requires evidence that the story does not provide.

---

What the laboratory science actually shows

There is real preclinical research behind the interest in fenbendazole. It should neither be exaggerated nor waved away.

In a widely shared 2018 study, researchers found that fenbendazole disrupted microtubules, interfered with glucose uptake and killed human cancer cells at micromolar concentrations. It also inhibited human-tumor xenograft growth in mice. The authors concluded that it deserved evaluation as a potential anticancer agent. The study was published in Scientific Reports.

That is a plausible starting point for drug development. It is not a clinical result. Cancer cells in a dish do not model a whole patient, and a dose that reaches a mouse tumor may not produce a safe or useful exposure in humans. Absorption is a particular concern with fenbendazole because it is poorly water-soluble.

Animal findings are also not uniformly positive. A Yale study of mice with mammary tumors found that a therapeutic fenbendazole diet did not alter tumor growth, invasion, metastasis or the response to radiation, although higher concentrations inhibited cells in culture. Mixed preclinical results are normal. They are one reason human testing is necessary before a mechanism becomes a treatment claim.

The human evidence does not validate the protocol

The human literature remains a collection of anecdotes, small case reports and uncontrolled observations—not a demonstration that fenbendazole treats cancer.

A 2021 three-patient case series described favorable responses in genitourinary cancers. Yet the patients also received treatments including nivolumab, surgery, chemotherapy or other alternative products. The fenbendazole dose was one gram three times weekly, not Tippens’s current 222 mg daily regimen. The report itself shows the concurrent treatments. It cannot establish that fenbendazole caused the responses, and it did not test Tippens’s own case.

A more recent paper by William Makis and colleagues reported three people with advanced cancers who took fenbendazole alongside other treatments. The paper was retracted in January 2026. The publisher said the first author had offered services related to the paper’s subject when it was submitted, had not declared that potential conflict, and gave an unsatisfactory response when questioned. The editor concluded that the issue affected interpretation and recommendation.

The retraction does not prove that the patients were invented or that their outcomes were fabricated; the notice does not allege either. It does mean the paper should not be presented as dependable clinical support, especially without disclosing why the journal withdrew it. Even before retraction, three selectively reported successes with concurrent treatments could not establish efficacy or a dose.

As of this review, no randomized or adequately controlled human trial has reported evidence that fenbendazole improves tumor response, progression-free survival, overall survival or quality of life in cancer patients. No trial has validated the Joe Tippens combination. In May 2026, ASCO issued a clinical notice strongly cautioning against using fenbendazole or ivermectin to treat cancer, or as an adjunct, outside well-designed clinical trials.

That conclusion is compatible with continued research. Promising laboratory results and dramatic case histories can justify carefully designed studies. They cannot substitute for them.

What about the “84.4 percent clinical benefit” study?

A 2026 study of ivermectin and mebendazole is now being promoted as stronger human evidence. The headline number sounds impressive: “84.4 percent clinical benefit.” But it is not a remission rate, a survival rate or proof that either drug treated cancer.

The study began with 197 cancer patients who were prescribed compounded capsules containing 25 mg of ivermectin and 250 mg of mebendazole through a U.S. telemedicine platform. Only 122—61.9 percent—completed the six-month follow-up survey. Among those who responded, 32.8 percent reported no evidence of disease, 15.6 percent reported tumor regression and 36.1 percent reported stable disease. The researchers combined all three groups to produce the 84.4 percent figure.

Those categories do not tell us what the drugs caused. The study had no untreated comparison group, mixed many different cancers, relied on patients to report their own cancer status and had no follow-up data for 75 participants. Many respondents also reported chemotherapy, radiation, surgery, supplements or diet changes. Someone who was already free of detectable cancer after surgery, or whose disease remained stable during chemotherapy, could still be counted as a “clinical benefit.” The paper itself called the findings hypothesis-generating rather than proof of effectiveness.

The study was also not independent of the seller. Its conflict-of-interest statement says every author was affiliated with and/or received salary support from The Wellness Company, which operated the telemedicine service that prescribed the capsules and offers compounded ivermectin and mebendazole. The company now uses the 84.4 percent figure on a page promoting the regimen. That financial relationship does not automatically make the findings false. It does make transparent methods, independent verification and cautious marketing especially important.

The journal, Anticancer Research, issued a formal Expression of Concern in June after questions were raised about the data’s verifiability, statistical reliability and ethical oversight. A September 1 update narrowed that concern. The journal said it was no longer pursuing the ethical-oversight question, would draw no adverse inference from materials that were not produced and was not accusing the authors of fabrication, falsification, plagiarism or other misconduct.

But the update did not clear the study’s methods. The journal said its review was continuing on the meaning of the combined “clinical benefit” measure, the self-reported survey data, when patients were assessed, the dose and side-effect analysis, whether the 122 respondents represented the full group, and how missing follow-up data and conventional cancer treatments were handled. As of this review, the paper has not been retracted, corrected or otherwise finally resolved.

Even if those questions are eventually answered, this was a different regimen. It combined ivermectin with mebendazole—not fenbendazole—and did not test the Joe Tippens supplement combination or the Makis dosing chart. It cannot pass its headline number on to either protocol.

A separate registered phase 2 study at the University of Florida is designed to test ivermectin alongside standard immune-checkpoint treatment. It uses 0.2 or 0.4 mg/kg on three days each week for four weeks and is designed primarily to measure an immune-system marker and safety. The registry’s latest posted update, dated July 15, listed it as not yet recruiting, and it has not reported results. Its existence shows that researchers can ask the question in a structured trial; it does not show that ivermectin works against cancer. Its doses and schedule also bear little resemblance to the chart’s highest-tier anecdote of 2.5 mg/kg every day.

Checkpoint

A large percentage from an uncontrolled survey is not a treatment effect. Without a comparison group and verified outcomes, the study cannot tell us what the drugs caused.

---

The ivermectin chart is not the Joe Tippens story

The supplied dose chart adds another drug—and another layer of unsupported certainty. It is headed as a William Makis cancer protocol and carries Ivermectinkart branding. The image alone does not establish who designed it, whether Makis approved that wording, or what relationship exists between the named protocol and the seller whose brand appears on it.

It is also not Tippens’s original regimen. Tippens’s story centers on fenbendazole, CBD, curcumin and, initially, vitamin E. An ivermectin chart circulating beside his name should not be treated as evidence from his remission.

The chart places people into four groups based on their cancer risk or how advanced their cancer is. It recommends higher doses of ivermectin for each group. It even recommends ivermectin for people who do not have cancer but have a family history or genetic risk.

The dose labels are confusing. They say “less than” 0.5, 1.0, 2.0 and 2.5 milligrams for each kilogram of body weight. Read literally, those ranges overlap. A dose below 0.5 is also below 1.0, 2.0 and 2.5. The highest-dose column then describes someone taking exactly 2.5 mg/kg every day, even though its heading says less than 2.5. The first three columns also make the unsupported promise of “No side effects.”

No controlled human evidence validates those cancer categories, the dose escalations or preventive use. The chart provides no pharmacokinetic rationale, interaction rules, monitoring plan or stopping criteria. Putting mg/kg beside a recommendation makes it look medical; it does not make it evidence-based.

One of its own testimonials changes when traced to the source. The chart describes a woman with ovarian cancer taking 12 mg of ivermectin daily with chemotherapy. The underlying account republished by Tess Lawrie says she took it every other day, not daily, while also receiving paclitaxel and carboplatin followed by surgery. The source itself acknowledges that ivermectin’s contribution is unknown. A response during established ovarian-cancer treatment cannot validate ivermectin dosing, much less prevention.

There is preliminary human cancer research on ivermectin. A 2025 early-phase study combined it with the investigational immunotherapy balstilimab in metastatic triple-negative breast cancer. Nine patients enrolled; among eight evaluable patients, one had a partial response, one had stable disease and six had progressive disease. Median progression-free survival was 2.5 months. The small, uncontrolled combination study cannot isolate ivermectin’s effect and used an intermittent schedule, not the chart’s broad daily recommendations.

Checkpoint

The existence of early research means the question is worth studying. It does not mean an online chart has already answered it.

---

“No side effects” is demonstrably false

No drug can honestly be guaranteed to cause no side effects in everyone. The assurance is especially indefensible here because serious harm has been reported in the same setting being promoted: cancer patients self-treating after reading online claims.

Fenbendazole has been linked to clinically significant liver injury. In a 2021 report, an 80-year-old woman with lung cancer took one gram a day for three days followed by four days off—the schedule popularized online—while receiving pembrolizumab. After about a month, her AST reached 386 U/L and ALT 487 U/L. Her liver function improved after she stopped fenbendazole, and her tumors did not shrink during its use.

A 2024 case report described severe biopsy-confirmed liver injury in a 67-year-old woman who had been taking fenbendazole. Her tests normalized over three months after stopping it.

A January 2026 case is especially relevant to Tippens’s current daily schedule. A 47-year-old woman with metastatic colon cancer took 222 mg of fenbendazole three times a week for six weeks, then increased it to 222 mg every day. One week after the increase, she developed symptoms of severe liver injury. Her AST reached 2,435 U/L and ALT 2,407 U/L. She was also receiving nivolumab and relatlimab, but the authors assessed fenbendazole as the probable cause. Her liver tests improved rapidly after she stopped it and remained normal after immunotherapy was restarted. The full case report documents the timing and causality assessment.

Another 2026 report described a 65-year-old man with prostate cancer who developed jaundice and severe liver injury after alternating veterinary fenbendazole and ivermectin for about three months based on online advice. His liver enzymes improved after both drugs were stopped; the authors assessed the drug relationship as highly probable. The report provides the exposure and laboratory details.

Ivermectin has an established role in treating human parasitic infections, but its label does not promise zero risk. The official STROMECTOL prescribing information lists adverse reactions and postmarketing reports including seizures, hepatitis and neurotoxicity with altered consciousness, confusion, coma and death.

In a 2026 cancer case, a 73-year-old woman with metastatic breast cancer took high-dose ivermectin after finding information online. She developed altered mental status, generalized seizures and respiratory failure, requiring intubation and mechanical ventilation. She recovered neurologically within 48 hours. The published abstract does not state the numerical dose or connect her exposure to this chart, so it would be wrong to claim that the chart caused her hospitalization. It does establish that life-threatening toxicity can occur during high-dose cancer self-treatment.

Case reports cannot tell us how often these events happen, and they do not prove that every user will be harmed. They are more than enough to disprove “No side effects.”

The chart’s highest anecdotal exposure also deserves context. The labeled ivermectin dose for strongyloidiasis is approximately 0.2 mg/kg once. Each 2.5 mg/kg dose is 12.5 times that labeled amount—and the chart’s anecdote repeats it every day. That comparison is not a toxicity threshold, but it shows why safety data from a standard single antiparasitic dose cannot be casually extended to months of high-dose cancer use.

Then comes the sales pitch

The account question needs precision because two profiles were initially mistaken for Tippens himself.

The first, @JoeTippensl, displays X’s “Commentary account” label. According to X’s own explanation, parody, fan and commentary labels identify profiles that depict another person or organization while remaining unaffiliated. It is therefore not evidence of Tippens’s personal presence on X. Its blue check does not change that: X says a blue check now primarily indicates an active Premium subscription and does not carry the old review for notability and authenticity.

The second, @JoeTippens_, calls itself “Joe Tippens Protocol.” In the supplied screenshot its bio says “My name is Joe Tippens,” claims “MD. Oncologist,” and advertises an “Ivermectin & Fenben online-shop” with worldwide shipping. That biography conflicts directly with Tippens’s own site, where he describes a career in accounting, finance and business and states emphatically that he is not a doctor or scientist.

As of September 15, 2026, Tippens’s official website links readers to a Facebook community and YouTube videos but does not list a Twitter or X profile. That cannot prove he has never opened an X account, but it gives no basis for treating either reviewed profile as official.

The responsible conclusion is narrow and strong: someone is using Tippens’s name, photograph and survival story to offer the promoted drugs for sale. The screenshot establishes a commercial sales pitch, not completed sales, revenue or profit. It does not establish that Joe Tippens operates, endorses or benefits from the account.

In fact, his own history points the other way. In November 2019, Tippens wrote that a seller was using his name to market fenbendazole capsules and said he had sent a cease-and-desist demand. In 2020 he again warned followers about what he called entrepreneurial companies selling questionable products. Those warnings remain on his blog.

There is a separate commercial issue on Tippens’s current protocol page. It links directly to purchasable branded supplements and praises their claimed properties. That is product promotion, regardless of intent. The same page explicitly says Tippens receives no financial remuneration and owns no stock in the company. No evidence reviewed for this article contradicts that disclosure, so it would be unfair to claim he profits from those links.

The online-shop account is different. Its sales interest is explicit. The Ivermectinkart branding on the ivermectin chart likewise puts a retailer’s name on dosing claims. A commercial interest does not automatically make a medical claim false, but it creates an obvious reason to demand stronger evidence and full disclosure. Here, the sales pitch is far more developed than the clinical evidence.

What the story can, and cannot, support

The most defensible reading of Joe Tippens’s story is neither “fenbendazole cured him” nor “fenbendazole definitely had nothing to do with it.” The evidence cannot support either conclusion.

What it can support is a hypothesis. A patient with a grim cancer history reported an exceptional, durable remission while taking a clinical-trial treatment, fenbendazole and several supplements after chemotherapy and radiation. Laboratory research gives fenbendazole plausible anticancer mechanisms. That combination of facts can justify serious study.

What it cannot support is the leap from one confounded outcome to a universal protocol. It cannot establish the active ingredient, effective dose, best schedule, duration, compatible treatments, target cancers, survival benefit or risk. It provides no evidence for cancer prevention. Adding more testimonials does not repair those missing controls, particularly when unsuccessful users are not systematically counted.

There is a human cost to pretending those answers already exist. People with cancer may spend money on unproven products, conceal them from clinicians, suffer interactions or toxicity, or delay treatment with evidence of benefit. Patients deserve curiosity without credulity and hope without a sales funnel.

Final checkpoint

Joe Tippens is an exceptional survivor. His story is a reason to ask a research question, not an answer to one. No clinical evidence validates the Joe Tippens Protocol, its dose, its changing supplement combinations, the ivermectin chart or the preventive claims. And the accounts selling these drugs in his name should not be mistaken for him.

Anyone considering fenbendazole, ivermectin or supplements during cancer treatment should tell their oncology team exactly what they are taking. That is not an instruction to abandon research or dismiss repurposed medicines. It is the minimum needed to manage interactions, monitor harm and keep a difficult medical decision anchored to evidence.

Research checked September 15, 2026. Tippens’s diagnosis, treatment sequence, scans and conversations with clinicians are described from his own website because independent medical records were not available for this review. The X-profile and chart descriptions come from supplied screenshots; their operators and the chart’s creator were not independently authenticated. This article is an evidence review, not medical advice.

Resources and further reading

• Joe Tippens, “The Blog” — his first-person chronology, original regimen, later updates and past warnings about sellers using his name.

• Joe Tippens, current protocol page — the daily fenbendazole schedule and four branded supplement pathways, including his no-remuneration disclosure.

• Joe Tippens, 2026 posts and Q&As — recent official-site activity; it is not an independent medical record.

• American Cancer Society and ASCO, “What to Know About Fenbendazole” — patient-facing review of evidence, human approval status, interactions and liver risk.

• National Cancer Institute, small-cell lung cancer treatment — staging, prognosis factors and established treatment categories.

• Dogra et al., Scientific Reports (2018) — the frequently cited cell and mouse study supporting further investigation.

• Duan et al., Journal of the American Association for Laboratory Animal Science (2012) — a mouse study that found no effect on mammary-tumor growth, spread or radiation response at the tested exposure.

• Nguyen et al., Anticancer Research (2024) — review of preclinical rationale, bioavailability, unanswered human dosing questions and the need for clinical trials.

• Hulscher et al., Anticancer Research (2026) — the uncontrolled ivermectin–mebendazole cohort that reported an 84.4 percent self-reported “clinical benefit” among follow-up respondents.

• The Wellness Company’s study-promotion page — the company’s use of the 84.4 percent figure to promote the regimen tested through its telemedicine service.

• Expression of Concern and September update — the journal’s initial concerns, later narrowing of the audit and still-pending methodological review.

• ICONIC phase 2 trial, NCT07487805 — a registered study of ivermectin with immune-checkpoint therapy; no results have been reported.

• Yamaguchi et al., Case Reports in Oncology (2021) — liver injury after a Tippens-style fenbendazole schedule in a patient receiving pembrolizumab.

• Krishnan et al., World Journal of Clinical Cases (2026) — severe liver injury after a patient increased fenbendazole from 222 mg three times weekly to 222 mg daily.

• Powderly et al., Cureus (2026) — severe liver injury after veterinary fenbendazole and ivermectin self-treatment.

• ASCO clinical notice (2026) — professional guidance against use for cancer outside well-designed trials.

• Retraction notice for Makis et al. (2026) — the publisher’s stated reason for withdrawing the three-patient fenbendazole paper.

• STROMECTOL prescribing information — approved human uses, standard dosing and documented adverse reactions for ivermectin.

• X Help, profile labels and blue checks — X’s explanation of commentary accounts and what a blue check does and does not establish.

Of all the improbable things that have ever happened, the strangest is not simply that I exist.

It is that I exist as me.

There are ordinary answers to how I got here.

• Two people met.

• One particular sperm fertilized one particular egg.

• A body developed, a brain organized itself, and a child entered a world already crowded with language, history, and other people’s hopes.

• Change almost any detail in that chain and this person might never have existed.

But that is not quite the question.

I am not asking only why this body exists. I am asking why I am the one looking through its eyes. Why is its pain felt here? Why are its memories mine? Why did this life, rather than any other life, or no life at all, become the place from which the universe is present?

The question sounds circular: Why am I me? Who else could I be?

Yet the word me is doing two different jobs. It names a person whom others can observe, and it points toward the private subject to whom the world appears.

Science has learned a great deal about the person. It is still struggling with the subject.

Two questions hiding inside one

The question contains two mysteries. The first is how a brain and body produce a sense of self: a body that feels like mine, actions that feel like mine, and a story that connects the child I was to the person reading this sentence now.

The second is harder to get hold of: Why is any of that accompanied by experience? Why does activity in a brain feel like something from the inside?

Philosopher David Chalmers famously distinguished between the functional questions of consciousness, how a system pays attention, integrates information, or reports what it sees, and what he called the “hard problem”: why physical processes are accompanied by subjective experience at all. The “easy” problems are not easy in practice. They are called easy because, at least in principle, they can be approached by explaining mechanisms.

We might explain how a hand detects heat and pulls away from a flame. The harder question is why the burn hurts.

That distinction matters. A theory of how the brain models a person is not automatically a theory of consciousness. A brain may build an extraordinarily detailed model of its body, its past, and its surroundings, while leaving us with the question of why there is anything it is like for that model to be running.

How a brain builds a self

The self does not seem to arrive all at once. Philosopher and cognitive scientist Shaun Gallagher distinguishes between a “minimal self” and a “narrative self”. The minimal self is immediate and embodied: the basic sense that experience is happening here, that this is my body, and that I am initiating this movement. The narrative self reaches across time. It is the remembered and continually revised story of who I have been, who I am, and who I might become.

Long before a child can tell an autobiography, the beginnings of an embodied point of view are already in place. Later come names, mirrors, memories, social roles, promises, embarrassments, and plans. Other people help write the story. They tell us what we were like before we could remember, reward some versions of us, resist others, and give us the language with which we learn to describe ourselves.

Even apparently simple self-recognition is not a universal switch that flips at one exact age. In many urban Western studies, children begin passing the mirror-mark test during the second year of life. Results vary across cultural settings, and researchers debate how familiarity with mirrors and the social demands of the test affect performance. A 2026 study of 56 toddlers in rural Vanuatu found that 29 percent passed the face-mark test. Success increased with age, and a majority passed by 36 months—later than commonly reported in urban Western samples. That does not mean the other children lacked a self. The test captures one narrow, visually mediated form of explicit self-recognition.

Adults are not as fixed as we feel, either. In the classic rubber-hand experiment, synchronized touches to a hidden real hand and a visible fake hand can create the uncanny feeling that the rubber hand belongs to the participant. Virtual-reality experiments have similarly shifted people’s experienced location toward an artificial body seen from a distance.

These illusions suggest that bodily ownership and self-location are not read from an infallible inner label. The brain appears to assemble them by combining vision, touch, movement, and signals from within the body.

Calling the self “constructed” does not mean it is imaginary. A melody is constructed from notes; that does not make the music unreal. It means the self may be less like a hidden object lodged somewhere in the skull and more like an activity the living brain and body must keep performing.

That activity is remarkably stable. Most mornings, we wake without having to determine whose life has resumed. But it is not perfectly stable. Memory can fail. The sense of ownership can be manipulated. Injury and illness can alter personality, agency, or the boundary between self and world. The continuity we take for granted is an achievement, not a given.

And yet: who experiences the construction?

Neuroscience is not empty-handed. Researchers can study which patterns of activity accompany seeing a face, feeling pain, recognizing a body, or becoming aware of a stimulus. Competing theories offer different accounts of what makes information conscious. There is, however, no settled theory.

A review of 412 experiments found that which theory a study appeared to support could be predicted from its methodological choices, regardless of the findings. In 2025, a large preregistered collaboration tested predictions from two leading accounts—global neuronal workspace theory and integrated information theory—using fMRI, MEG, and intracranial EEG across 256 participants. The results supported some predictions from each theory while substantially challenging key claims of both. There was no decisive winner.

This does not show that consciousness is beyond science. It shows that declarations of victory are premature.

Some philosophers reject the hard problem itself. Daniel Dennett argued that once we explain attention, memory, discrimination, report, and the brain’s beliefs about its own inner life, no additional mystery may remain. On this view, asking what turns brain activity into an extra ingredient called “experience” may be a mistake. There is no extra ingredient; there are only the complex things a conscious organism does.

Others think that leaves out the very fact needing explanation. Thomas Nagel put the point simply: an organism is conscious if there is something it is like to be that organism. A complete description of a bat’s sonar, behavior, and nervous system might still fail to tell a human what echolocation is like for the bat.

The disagreement is not merely about missing data. It is about what an explanation of experience would have to explain.

The question that makes you dizzy

Then there is the most intimate part of the puzzle: not only Why is there experience? but Why is this experience mine?

Philosopher Benj Hellie calls this the “vertiginous question”. Imagine possessing a perfect, objective inventory of the universe. It describes every person, every brain, every thought, and every point of view. The inventory may still seem to omit one fact: Why is this particular stream of experience the one that is live for me?

The most economical naturalistic answer is that nothing was omitted. There was no bare observer waiting outside the universe to be assigned a body. This organism came to exist, and as it developed a conscious point of view, that point of view was the organism’s. I could not have awakened as someone else because there was no detachable “I” available for reassignment.

On this account, asking why I am this person resembles asking why here is this place. Every location can be described from the outside, but here refers to the place from which the word is spoken. In the same way, every conscious subject refers to itself as I. The word does not select one privileged observer from a cosmic list. It marks the perspective doing the speaking.

That answer dissolves part of the puzzle. It explains why no soul lottery is required. But for many of us, it does not touch the deepest source of the dizziness: Why is there a first-person presence anywhere? Why is the universe not merely unfolding, but unfolding for someone at countless local points within it?

Religious and metaphysical traditions offer other answers. The subject may be an immortal soul, a fragment of a universal mind, or a stream that extends beyond one biological life. These ideas can carry profound meaning, but they are not conclusions current evidence allows science to establish. A gap in one explanation is not positive evidence for whichever alternative we prefer.

A mystery is not a blank check.

The honest position is narrower and less satisfying. The natural sciences can increasingly describe the conditions under which human selves and conscious reports arise. They do not yet provide an agreed explanation for why those conditions are accompanied by felt experience—or why reality is present from any particular point of view.

Why joy and heartache come together

Whatever consciousness ultimately is, ours is not detached observation. Human experience is organized around what matters to us.

Love provides a clear example. In a small 2005 brain-imaging study, 17 people who described themselves as intensely in love viewed photographs of their partners. The images engaged parts of the brain associated with reward and motivation, including the ventral tegmental area and caudate. The study did not discover love inside a handful of brain regions. It showed that romantic attachment recruits systems involved in wanting, learning, and directed pursuit.

Separation reaches into some of the same architecture of significance. In the well-known Cyberball experiment, being excluded from a virtual ball-tossing game increased activity in the anterior cingulate cortex, a region also implicated in the distressing component of physical pain. That activity tracked participants’ reported distress. Rejection is not identical to physical injury, but social loss is not merely an abstract opinion, either. The body treats belonging as consequential.

Grief makes the connection clearer. In one fMRI study of 23 women who had lost a mother or sister to breast cancer, reminders of the deceased engaged regions previously associated with emotional pain in both groups. Only the 11 women with complicated grief also showed greater activity in the nucleus accumbens, a reward-related region, and that activity correlated with yearning. The study was small and narrowly sampled; it cannot explain grief as a whole. But it fits an experience many bereaved people recognize: someone can be painfully absent while the mind and body continue to reach for them.

These findings do not reduce joy, love, or sorrow to colored patches on a brain scan. They reveal something more modest. The systems that allow another person to become rewarding, expected, and woven into our future are also the systems that make separation consequential.

Love teaches a nervous system that someone matters. Loss can change the world faster than that learning can change.

This is why joy and heartache are not unrelated accidents that happen to share a life. Both become possible when a conscious creature can care—when another person, place, hope, or purpose becomes part of how the future is imagined. The capacity that opens us to delight also exposes us to grief.

That does not make suffering good. It does not mean every loss was “worth it,” or that pain must be romanticized into a lesson. It means vulnerability is built into attachment. A life protected from every possible heartache would also have to be protected from much of what makes anything matter.

Did I receive this life?

To say I received a life suggests that an “I” existed before the life arrived. The naturalistic account reverses the order: A life did not happen to a preexisting self; a living process gradually became capable of saying I. The receiver was made by what it received.

That answer may describe the sequence without exhausting its meaning. Gift is not always a scientific claim about a giver. Sometimes it is a way of acknowledging that none of us authored the conditions of our arrival. We did not choose our century, body, first language, early caregivers, or the accidents that placed certain people in our path. By the time we became able to choose, much of the chooser had already been formed.

We inherit a life before we can evaluate it. Then, within limits we did not select, we participate in making it.

How can I cease to be me?

Before going further, I should be clear about the scope of this essay. It is not meant to promote a religious, atheistic, agnostic, or strictly secular interpretation of death. I will not be discussing heaven or hell, nor trying to prove or disprove an afterlife. Those questions matter deeply to many people, but they are not the question I am asking here. I am interested in a narrower and stranger puzzle: why it is so difficult for a conscious mind to imagine its own absence, and why love makes that imagined absence feel like missing the people who remain.

There is one point at which the question stops being an intellectual puzzle.

I can understand the sentence One day I will die. I know what the words mean. What I cannot genuinely picture is this point of view—the only one from which I have known anything—simply ending. I have never experienced a moment in which I was not present. Every darkness I imagine is still a darkness I see. Every silence still has me listening to it.

When I try to imagine my family after my death, I can see them continuing their lives. I imagine conversations I will not join, birthdays I will not attend, troubles I will not be there to help with. Then comes the thought that feels completely natural: I will miss them.

But who, exactly, would be doing the missing?

The imagination has played a quiet trick. To picture my own absence, it has placed me back inside the scene as an invisible observer. My body is gone, but my point of view remains, watching everyone else go on without me. Perhaps that is as close as a conscious mind can come to picturing nonexistence. Every mental picture still has someone looking.

This is not abstract for me. My mother passed away recently, and I apply this question to her every day. Does she miss us? Can someone who was so completely herself simply cease to be? I know my mind may be placing her back inside the world as an unseen observer, still loving us and still aware that we are here. Perhaps those questions tell me more about my grief than about what comes after death. But knowing that does not make them disappear.

On a naturalistic account, if consciousness ends when the brain can no longer sustain it, I will not experience an endless absence. There will be no future me waiting somewhere in the dark, longing for the people I love. Darkness is still an experience. Missing someone is still something a conscious person does.

Epicurus argued more than two thousand years ago that when we exist, death is not present, and when death is present, we no longer exist. That addresses the imagined experience of being dead. It does not make every fear surrounding death irrational. Dying can involve suffering. Our absence can wound the people who remain. And death may still harm us by depriving us of a future we otherwise could have lived, even if we never experience that deprivation.

On that account, I will miss them is logically misplaced but emotionally exact. The missing is happening now. It is love in the present trying to imagine a future from which I have been removed. I am not really picturing myself dead. I am picturing myself alive but unable to reach the people who made life matter.

If some form of personal consciousness survives death, as many religions hold, then the question takes a different shape. Science has not established that it does, so honesty requires leaving the possibility unresolved. The feeling itself cannot settle that question. It reveals something certain about this life: I am deeply attached to it.

That need not be morose. The thought of death does not tell me what comes next. It tells me what I do not want to leave: these people, these voices, these ordinary days, and this particular place in their lives. The ache is not evidence about an afterlife. It is evidence of how much there is here to love.

The most honest answer

So why am I me? If the standard naturalistic picture is right, there was no separate “me” who might have been placed somewhere else. This organism’s embodied, remembered, socially shaped point of view is not a container I occupy. It is what I am.

Why did this organism exist? Because of an immense chain of causes, much of it accidental.

How did it become a self? Through a developing brain and body that learned to locate experience here, distinguish action from event, remember a past, anticipate a future, and build a story with other people.

Why is any of that felt from within? We do not know. We do not even agree on whether the question identifies a missing mechanism, a conceptual confusion, or a fundamental feature of reality.

And how did I come to experience a life with joy and heartache? Because this conscious human life became capable of attachment. People and possibilities ceased to be neutral facts. They mattered. Joy was the felt presence of what mattered; heartache was what happened when reality tore away from it.

That answer does not reveal why existence produced you. But it explains something about why existence could matter once you were here.

I do not know why I am me. I know only that, for a brief interval, this life is not merely an item in the universe’s inventory. It is present. It is felt. It can love, lose, remember, create something that did not exist before, and ask what any of it means.

The mystery is not proof of a hidden answer. But neither is it nothing.

The universe is happening, and from here, for a while, it feels like me.

Resources

• David J. Chalmers, “Facing Up to the Problem of Consciousness” (1995) — The influential distinction between functional questions about cognition and the “hard problem” of subjective experience.

• Thomas Nagel, “What Is It Like to Be a Bat?” (1974) — A foundational argument about the subjective character of consciousness.

• Daniel C. Dennett, “Facing up to the hard question of consciousness” (2018) — A prominent challenge to the idea that consciousness presents a separate hard problem.

• Benj Hellie, “Against Egalitarianism” (2013) — Introduces the “vertiginous question” of why one subject’s perspective is the one that is live for that subject.

• Shaun Gallagher, “Philosophical conceptions of the self: implications for cognitive science” (2000) — Develops the distinction between minimal and narrative forms of selfhood.

• Petrović et al., “Mirror self-recognition in ni-Vanuatu toddlers: developmental trajectory and cultural validity” (2026) — Examines the timing and cultural validity of the mirror-mark test in a rural ni-Vanuatu community.

• Botvinick and Cohen, “Rubber hands ‘feel’ touch that eyes see” (1998) — The classic demonstration that multisensory cues can alter experienced body ownership.

• Lenggenhager et al., “Video ergo sum: manipulating bodily self-consciousness” (2007) — Shows that conflicting visual and tactile signals can alter self-location and identification with a body.

• Yaron et al., “The ConTraSt database for analysing and comparing empirical studies of consciousness theories” (2022) — Examines how experimental methods relate to reported support for competing theories of consciousness.

• Cogitate Consortium et al., “Adversarial testing of global neuronal workspace and integrated information theories of consciousness” (2025) — A preregistered theory-comparison study that challenged important predictions of both accounts.

• Aron et al., “Reward, motivation, and emotion systems associated with early-stage intense romantic love” (2005) — An early fMRI study of the motivational and reward systems recruited by romantic love.

• Eisenberger, Lieberman, and Williams, “Does rejection hurt? An fMRI study of social exclusion” (2003) — Evidence that social exclusion engages some neural systems associated with distress and pain.

• O’Connor et al., “Craving love? Enduring grief activates brain’s reward center” (2008) — A small imaging study examining pain, reward, and yearning in bereavement.

• Steven Luper, “Death,” The Stanford Encyclopedia of Philosophy (Winter 2024) — Reviews Epicurus’s argument that being dead is not an experience and the competing view that death can still harm us by depriving us of a future we otherwise would have had.

The BBC headline was too neat: “Cervical cancer deaths fall to zero in young women given vaccine.”

The study behind it did not link each woman’s vaccination record to her death record. It examined cervical-cancer mortality across England and related the pattern to the age and coverage of the national HPV vaccination programme. So the most exact headline would have been something like this:

No cervical-cancer deaths were recorded among women aged 20–24 in England from 2020 through 2024; 23.1 would have been expected if the earlier mortality rate had continued.

That caveat matters. It does not, however, turn the result into its opposite.

In a series of posts, the pseudonymous account Jikkyleaks claimed that deaths had fallen to zero “irrespective of vaccination status,” that screening and a sudden collapse in teenage sexual activity explained the result, and—eventually—that “it was the iPhone, not the HPV vaccine.”

A set of tiny country graphs was offered as evidence of a “secret rise” in cervical cancer. A disclosed competing interest became an insinuation of Merck ghostwriting. When challenged with the study’s actual design and findings, the substantive answer was simply that the rebuttal was “incorrect.”

There is a fair criticism here. The new mortality study is observational, so it cannot tell us what happened to each vaccinated or unvaccinated woman. But that limitation does not make the competing explanation true. The dates, age groups, randomized trials, and linked cancer registries all point away from it.

A limitation should narrow a claim, not reverse it.

What the mortality study actually found

England introduced routine HPV vaccination in September 2008 for girls aged 12–13, with a catch-up programme for girls aged 13 to under 18. Before the pandemic, programme-level vaccination coverage in the routine cohorts was generally around 80–90 percent. These were population estimates, not individually linked vaccination records.

The 2026 Lancet study by Peter Sasieni and Milena Falcaro analysed national cervical-cancer deaths from 2001 through 2024 among women aged 20–34. It used official coverage estimates by birth cohort and modelled how many deaths would have been expected without vaccination, based on earlier mortality patterns.

Its headline result was striking. From 2020 through 2024, no women aged 20–24 died from cervical cancer in England. The model expected 23.1 deaths. That corresponds to a population-level reduction of 100 percent, with a 95 percent confidence interval of 84–100 percent.

The result was not confined to that single age-period group. Other groups that included women eligible for routine or catch-up vaccination also showed reductions:

• Among women aged 20–24 in 2015–19, mortality was 80 percent lower than expected (95% CI 51–94).

• Among women aged 25–29 in 2020–24, mortality was 69 percent lower than expected (95% CI 55–79).

• Across the women aged 20–34 included in the analysis through the end of 2024, the model associated vaccination with an estimated 199.6 fewer cervical-cancer deaths (95% CI 125.0–274.2).

“Associated with” is the right phrase. The study compared population cohorts rather than individually matched vaccinated and unvaccinated women, and the number of expected deaths came from a modelled counterfactual. It also relied on five-year age groups and assumed no herd-protection effect. Attributing the decline to vaccination therefore requires assumptions about the relationship between vaccine uptake and underlying cervical-cancer risk. The authors correctly described the evidence as observational.

Because the study was observational and used aggregate coverage data, it could not calculate individual mortality risk by vaccination status. The rebuttal goes wrong when it treats that absence of individual stratification as evidence that vaccination status was irrelevant.

An aggregate zero does not show that vaccination status was irrelevant

Because no one aged 20–24 died from cervical cancer during this period, the observed death count was necessarily zero among both vaccinated and unvaccinated members of the cohort. In that narrow, literal sense, saying the zero applied “irrespective of vaccination status” is true.

The study did not match individual vaccination records to mortality records, so it could not calculate separate death rates for vaccinated and unvaccinated women. It therefore cannot tell us whether the two groups had the same underlying risk or whether vaccination status made no difference. There is another complication: the unvaccinated minority lived in a highly vaccinated population and may have received some indirect protection.

So yes, the zero included everyone. But it is a mistake to turn that population-wide observation into evidence that vaccination was irrelevant. The study was not designed to make that individual-level comparison. Instead, it used earlier mortality patterns to estimate what might have happened without the national vaccination programme.

It did, however, find an age-and-coverage pattern. The largest declines occurred in the youngest cohorts, where vaccination coverage was highest. The declines were smaller in cohorts offered vaccination at older ages or with lower coverage, while the estimate for women aged 30–34 was much less certain. That pattern does not prove causation on its own, but it is what we would expect from a vaccine that works best before HPV exposure.

Routine screening is unlikely to explain the zero in 20–24-year-olds

Screening is essential to cervical-cancer prevention. It detects precancer so it can be treated, and vaccinated women should still attend when invited. But it cannot plausibly be the main explanation for the new mortality result.

The reason is simple: routine cervical screening in England begins at age 25. The first invitation is sent at 24½, and screening is not recommended below 25 for people who have not been invited. The group with zero deaths was aged 20–24.

Changes in screening can influence cancer incidence and mortality in older groups. They do not explain why the strongest decline appeared before routine screening began, nor why the effect followed eligibility for vaccination and age at vaccination.

The earlier incidence data make that pattern harder to wave away. In a 2021 English registry study covering 13.7 million person-years, cervical-cancer incidence was estimated to be 34 percent lower in the cohort offered vaccination at ages 16–18, 62 percent lower when offered at 14–16, and 87 percent lower when offered at 12–13, compared with an earlier birth cohort that was not eligible for the vaccination programme.

The same pattern appeared in CIN3, the most serious conventional grade of cervical precancer. The reductions were 39, 75 and 97 percent. By mid-2019, the researchers estimated that there had been 448 fewer cervical cancers and 17,235 fewer CIN3 cases than expected.

The age at vaccination matters. The HPV vaccine prevents new infections; it does not clear an infection that is already present. You would therefore expect it to work best when given before HPV exposure and less well when given later. That is exactly the pattern the researchers found. Better screening or cancer treatment might lower rates across the population, but it would not readily explain why the reductions closely followed age at vaccination and vaccine eligibility.

The claim that women offered vaccination in their late teens “did not see a drop” is simply wrong. They did see one. It was smaller, as expected when more of them may already have encountered HPV before being vaccinated.

The teenage-pregnancy theory fails three basic tests

The next claim is that teenage pregnancy “fell off a cliff in 2007,” proving that adolescents abruptly stopped having sex and therefore stopped acquiring HPV.

This is the argument’s starting point. But England’s programme began in September 2008, and a conception rate is not a direct measure of sexual activity.

In a later post, the iPhone—launched in 2007—is floated as the real cancer-prevention intervention.

The proposed alternative explanation relies on a conception trend and a chart of clinic contacts, neither of which directly measures sexual activity.

This story fails on chronology, measurement and biological specificity.

First, the English HPV programme did not roll out in 2007. It began in September 2008. The first routine cohort received the bivalent Cervarix vaccine; England switched to quadrivalent Gardasil in 2012 and later to Gardasil 9. An arrow placed at 2007 on a pregnancy chart does not move the programme back a year.

Second, a conception rate is not a direct measure of sexual activity. It reflects exposure to pregnancy together with contraceptive use and effectiveness. ONS conception estimates count pregnancies ending in a maternity or legal abortion and exclude miscarriages. They therefore cannot reveal, by themselves, how frequently teenagers were having sex.

The trend was also not “constant for nearly 30 years” before a vaccine-year cliff. ONS data show a longer decline, with rates falling from the late 1990s and then falling faster after 2007.

Research comparing Britain’s national sexual-attitudes surveys found no statistically significant change in recent sexual activity among women aged 16–19 between 2000–01 and 2010–12. The researchers attributed most of the modelled reduction in pregnancy risk to increased use of more effective contraception. The study did not measure sexual behaviour among girls younger than 16, so it cannot settle that part of the argument.

The NHS graph used to “prove” contraception fell measures contacts with community contraceptive clinics. It is not a census of all contraceptive use. Young people can obtain contraception from general practice, pharmacies, retail outlets and other services; methods also differ greatly in effectiveness. Fewer clinic contacts cannot be converted into an 80 percent fall in sexual activity.

Third, English surveillance measured vaccine-type HPV prevalence among young females attending chlamydia-screening services. The UKHSA’s Green Book reports that HPV16/18 prevalence among 16–18-year-old females fell from about 15 percent in 2008 to 8.2 percent in 2010–11 and 1.6 percent in 2016. Among 19–21-year-olds it fell from 14 percent in 2010–11 to 0.7 percent in 2018; by 2018–20 it was below 1 percent among females aged 16–24.

A broad decline in sexual activity might reduce several sexually transmitted infections. It does not readily explain why the HPV types targeted by the vaccine collapsed in step with vaccine coverage, while surveillance also detected indirect protection in less-vaccinated groups and no evidence of type replacement.

The iPhone hypothesis is not a competing causal analysis. It is a date coincidence attached to an outcome that does not measure the proposed mechanism.

The “secret rise” graphs do not establish that vaccination increased cancer rates

Another post presents small, poorly labelled line charts and claims that countries introducing HPV vaccination around 2007 suffered a spike in cervical cancer among women aged 25–34 eight to ten years later.

England’s 2008 rollout illustrates the cohort problem. The age arithmetic alone shows that most women in the displayed 25–34 age band were not members of the routine school-vaccination cohort.

A girl vaccinated at 12–13 in England in 2008 was about 20–21 eight years later. She was not 25–34. Most women aged 25–34 in 2016 were born between roughly 1982 and 1991 and were too old for the routine school programme; only the younger edge of that band could have entered the catch-up programme. A broad period-rate graph for 25–34-year-olds therefore cannot be labelled a post-vaccine effect in the routine adolescent cohort.

The cropped screenshots do not provide enough source information to reproduce and verify every plotted series, so the underlying rates should not be declared false on the basis of these images alone. What the graphs clearly lack is individual vaccination status, confidence intervals, and adjustment for screening or registration changes. Even if every plotted rate is accurate, an upward segment in an ecological age-period graph does not establish that vaccination caused the increase.

The appropriate question is not, “Can I draw an arrow near a bump?” It is, “Do women with documented vaccination have less cervical cancer than comparable women without it, and does protection increase when vaccination occurs earlier?”

Several national linkages answer that question.

• In Sweden, researchers followed 1,672,983 girls and women and found 19 invasive cervical cancers among vaccine recipients and 538 among those not vaccinated. After adjustment, the incidence-rate ratio was 0.37 overall and 0.12 for those vaccinated before age 17.

• In Denmark, linked national records likewise showed the largest reduction among women vaccinated at age 16 or younger.

• A 2025 Dutch national linkage study is especially useful because it compared women from the same 1993 birth cohort and adjusted for differences in screening participation. There were 5 cervical cancers among 47,130 fully vaccinated women and 42 among 50,831 unvaccinated women. The adjusted cumulative risk ratio was 0.085 (95% CI 0.025–0.24).

Replace the Scotland bullet and the paragraph following the country list with this section:

Scotland’s “zero cancers” finding was oversold—but not erased

A separate criticism from Demasi concerns the 2024 Scottish linked-registry study. It reported no invasive cervical cancers among women vaccinated at ages 12–13, producing the BBC headline “No cervical cancer cases in HPV-vaccinated women.”

That zero deserved more caution. Most women in the routinely vaccinated group were still under 25 when follow-up ended, an age at which cervical cancer is rare. The absence of cases in that young subgroup was therefore encouraging, but it could not provide a precise estimate of long-term protection by itself.

In 2025, Samir Saidi and Mark Jones published a reanalysis and critique using data supplied by the original researchers. They argued that the vaccinated and unvaccinated groups differed substantially in screening participation and identified population-count anomalies that required explanation. Those are legitimate concerns. Screening can prevent invasive cancer by finding and treating precancerous changes, so higher screening participation among vaccinated women could account for part of the observed difference.

But Demasi’s conclusion that the Scottish result was “nothing more than a statistical illusion” goes beyond what that critique establishes.

The original study did not rely solely on the zero cases among the youngest women. Among women vaccinated at ages 14–22 who received three doses, invasive-cancer incidence was 3.2 per 100,000 person-years, compared with 8.4 among unvaccinated women. Screening differences weaken the certainty of that comparison; they do not prove that the vaccine’s true effect was zero.

The Scottish finding should therefore be treated as contested supporting evidence, not definitive proof. It also does not stand alone. Linked studies from Sweden, Denmark and the Netherlands found lower invasive-cancer incidence among vaccinated women, with the greatest reductions following vaccination at younger ages. The Dutch study compared women from the same birth cohort and adjusted for screening participation.

Demasi is right that vaccination must never be presented as a reason to abandon cervical screening. Both remain important. But showing that the Scottish headline overstated one young cohort’s zero-case finding does not erase the wider evidence that HPV vaccination reduces oncogenic infection, high-grade precancer and invasive cervical cancer.

“No RCT showed fewer cancer deaths” is true—and profoundly incomplete

The original HPV-vaccine trials were never set up to count cervical-cancer deaths decades later. That wasn’t an oversight. Cervical cancer is rare in young women and usually takes years to develop after a persistent HPV infection. A mortality trial would have needed an enormous number of participants and decades of follow-up.

There was also an ethical problem. Trial participants received cervical screening, and any high-grade precancer that was found was treated. Researchers could not leave CIN3 or adenocarcinoma in situ untreated simply to see which women eventually developed invasive cancer or died.

That is why the trials measured earlier stages in the disease process: persistent infection with cancer-causing HPV types and biopsy-confirmed precancer. Those outcomes are not the same as cancer deaths, but they are the stages the vaccine was designed to prevent.

Those are not arbitrary laboratory outcomes. Persistent high-risk HPV causes the overwhelming majority of cervical cancers; a landmark worldwide analysis detected HPV DNA in 99.7 percent of adequate cervical-cancer specimens. Modern classifications nevertheless recognize rare HPV-independent cervical cancers. CIN3 and adenocarcinoma in situ are established high-grade precursors, although not every lesion progresses to invasive cancer. Preventing these outcomes is therefore powerful evidence relevant to cancer prevention, but it is not the same as directly measuring invasive cancer or death in a randomized trial.

The randomized evidence is strong. A 2018 Cochrane review of HPV-vaccine trials found that among 15–26-year-olds negative for high-risk HPV at baseline, vaccination reduced HPV16/18-related CIN2 or worse from 164 to 2 per 10,000 and CIN3 or worse from 70 to 0 per 10,000. It also reduced high-grade disease from any HPV type, though by less, because the early vaccines did not target every oncogenic type.

Professor Jeffrey Morris has set out the causal argument: randomized trials show that vaccination prevents persistent oncogenic infection and high-grade precancer, while national registries show the predicted reduction in invasive disease. That reasoning depends on biological assumptions and does not produce a randomized estimate of cancer mortality. It does, however, show why the absence of a randomized death endpoint is not equivalent to an absence of causal evidence.

The honest formulation is therefore not “an RCT proved the vaccine prevents every cervical-cancer death.” Nor is it “there is no causal evidence.”

It is this: randomized trials show prevention of oncogenic HPV infection and high-grade precancer; individually linked registry studies show lower invasive-cancer incidence among vaccinated women; and the new English population study finds a mortality decline consistent with that evidence. Only the first part is randomized, and the mortality result is not individually linked.

The “14.0 versus 14.0” row is not a placebo comparison

In the latest challenge, Jikkyleaks explicitly says he is comparing two RCTs conducted at different times. He highlights identical rates of 14.0 cases per 1,000 person-years in a later trial and argues that the Gardasil recipients therefore appeared to receive no benefit against “actual disease.” The comparison looks alarming, but it combines results from trials with different comparators, populations and endpoints.

It comes from the 2015 trial of the nine-valent vaccine. One arm received Gardasil 9, labelled 9vHPV. The other received quadrivalent Gardasil, labelled qHPV. In other words, the table compares a nine-type HPV vaccine with a four-type HPV vaccine. There is no unvaccinated or placebo group in that comparison.

The identical 14.0 rates refer to all high-grade cervical, vulvar and vaginal disease, irrespective of HPV type, in a modified intention-to-treat population that included women already infected when vaccination began. Within this trial, the result means Gardasil 9 was not superior to original Gardasil on that broad all-type endpoint during the reported follow-up. It does not estimate original Gardasil’s benefit versus placebo. Both vaccines targeted the four HPV types they shared, neither could treat infections already present, and lesions caused by non-vaccine types were included in the total.

The more specific rows show why the broad 14.0-versus-14.0 endpoint does not isolate the added protection from Gardasil 9. Among women uninfected on day one, the rate of all high-grade disease was 2.4 with Gardasil 9 versus 4.2 with original Gardasil. For disease related to the nine vaccine types, there were 0 cases versus 13. In the susceptible per-protocol population, Gardasil 9’s efficacy against high-grade disease caused by its five additional oncogenic types—31, 33, 45, 52 and 58—was 96.7 percent (95% CI 80.9–99.8).

FUTURE II was a different trial, published in 2007. It randomized 12,167 women aged 15–26 to quadrivalent Gardasil or placebo and used actual high-grade cervical disease—CIN2, CIN3, or adenocarcinoma in situ related to HPV16/18—as its primary endpoint. Efficacy was 98 percent in women who were susceptible to those types and followed the protocol. It was 44 percent in the unrestricted intention-to-treat population, which included women already infected before vaccination. Against all high-grade lesions regardless of HPV type, efficacy in that broad population was 17 percent.

Those lower broad-population estimates are not evidence of concealment. They demonstrate two basic properties of the product: it is prophylactic, not therapeutic, and the original vaccine covered two major cancer-causing types, not every possible cause of a cervical lesion. Vaccinating before exposure is precisely why programmes target early adolescents.

Original Gardasil was supported by randomized, controlled endpoints measuring HPV-related high-grade cervical disease. Gardasil 9 was later compared with that active vaccine: efficacy was assessed for the five additional HPV types, while immune responses to the four shared types were tested for non-inferiority. The correct conclusion from the 14.0-versus-14.0 row is narrow: Gardasil 9 did not outperform original Gardasil on the broad all-type modified intention-to-treat endpoint. It does not show that original Gardasil provided no benefit versus placebo, and a cross-trial comparison cannot replace the randomized comparison within each trial.

Cochrane’s press language deserved scrutiny—but its data did not reverse direction

Editor’s note: This article addresses the central claim presented in the publicly available portion of Maryanne Demasi’s article that Cochrane’s own data contradict its conclusion. It does not purport to rebut every argument contained in the paywalled article.

Maryanne Demasi made a more serious methodological criticism of Cochrane’s 2025 reviews. She noted that the cervical-cancer estimates came from observational studies with important risks of confounding and argued that Cochrane’s public claim of “strong and consistent evidence” sounded more definitive than the technical review.

That is a fair issue to raise.

The Cochrane observational review included 225 studies covering more than 132 million people. Of the 20 studies reporting cervical cancer, nine were rated at critical risk of bias, seven at serious risk and four at moderate risk, chiefly because of confounding.

Five adjusted cohort studies including 4.39 million females produced a pooled risk ratio of 0.37 (95% CI 0.25–0.56), with substantial heterogeneity. For vaccination at age 16 or younger, the pooled risk ratio was 0.20 (95% CI 0.09–0.44). Cochrane graded the cancer evidence moderate certainty and concluded that vaccination “probably” reduces cervical-cancer incidence.

Public communications should have retained those words. “An approximately 80 percent lower observed risk when vaccination occurred by 16, based on moderate-certainty observational evidence” is more faithful than an unqualified promise that any girl vaccinated before 16 becomes 80 percent less likely to develop cancer.

But that does not mean Cochrane’s data pointed in the opposite direction. The review reported the limitations and graded the cervical-cancer evidence as moderate certainty. Its five adjusted cohort studies produced a pooled risk ratio of 0.37, but heterogeneity was substantial and none of the cervical-cancer studies was judged to be at low overall risk of bias. The result is observational evidence consistent with substantial protection—not bias-free proof.

Cochrane’s separate 2025 randomized-trial review included 60 studies and 157,414 participants, but it had no data showing an effect on cervical cancer itself. It found randomized evidence that vaccination reduces HPV infection and precancerous outcomes such as CIN2+. Cochrane’s public claim about preventing cancer therefore came from combining those trial findings with observational cancer evidence—not from randomized cancer endpoints.

The observational review does not stand alone. It sits beside randomized evidence on persistent HPV infection and high-grade precancer, established cervical-cancer biology, individually linked registry studies of invasive cancer and the new English mortality signal. The convergence of those different forms of evidence strengthens the causal inference, but it does not turn Cochrane’s observational cancer estimate into a randomized result.

A disclosed interest is a reason to inspect the work, not permission to invent its authorship

The mortality paper disclosed that Sasieni was lead investigator on a Gardasil 9 trial partly supported by a Merck investigator-initiated grant, had served on the data-safety board of an HPV-vaccine trial in India, and that both authors had received funding for an intervention intended to increase vaccine uptake. Readers are entitled to know that.

The mortality study itself was funded by Cancer Research UK. The disclosed Merck support concerned a separate Gardasil 9 trial on which Sasieni was an investigator, not the funding of this mortality analysis. The authors also disclosed using ChatGPT to search for relevant papers and Copilot to produce a first draft of the Summary, which they reviewed and edited.

That is relevant authorship context and should not be omitted, but it is not evidence that Merck wrote or controlled the study. No documentary evidence of Merck ghostwriting was presented in the posts reproduced here.

There is another awkward fact for that insinuation: the first English routine cohorts—the ones showing the greatest early reductions—were given Cervarix, made by GSK, not Merck’s Gardasil. Scotland’s zero-cancer result also followed a bivalent-vaccine programme, and the Dutch study used the bivalent vaccine. The signal appears across products, researchers and national systems.

Conflicts should sharpen scrutiny of the model, assumptions, outcome definitions and reproducibility. They are not a substitute for finding an error in any of them.

What the evidence actually shows

Before accepting either the BBC headline or the rebuttal, stop and separate four stages.

1. Observation

No cervical-cancer deaths were recorded in England among women aged 20–24 from 2020 through 2024. Based on earlier mortality patterns, the study estimated that 23.1 deaths would otherwise have been expected.

2. What the mortality study measured

The researchers analysed mortality by age, birth cohort and HPV-vaccination programme coverage. They did not link each woman’s vaccination record to her death record. The result is therefore a population-level association based partly on a modelled counterfactual.

3. Supporting causal evidence

Randomized trials show that HPV vaccination prevents persistent infection with cancer-causing vaccine types and prevents high-grade cervical precancer. Individual-level registry studies then show substantially less invasive cervical cancer among vaccinated women, with the largest reductions when vaccination occurs before likely HPV exposure.

4. Remaining uncertainty

The mortality paper cannot establish the exact share of the decline caused by vaccination, eliminate every possible confounder, or tell us the vaccination status of women in a cohort with zero deaths. Longer follow-up will also refine the size and durability of the mortality effect.

Evidence checkpoint

Did the mortality paper prove that every woman in the zero-death cohort was vaccinated? No.

Did the zero include women regardless of vaccination status? Yes. No deaths occurred anywhere in the age group, so the observed count was necessarily zero among both vaccinated and unvaccinated women. But the study did not calculate separate mortality rates by vaccination status and therefore cannot show that their underlying risks were equal or that vaccination made no difference.

Can routine screening explain the zero among 20–24-year-olds? Not plausibly as the main explanation. Routine cervical screening in England begins at age 25.

Does a fall in teenage conception prove an 80 percent fall in sexual activity? No. Conception rates reflect both exposure to pregnancy and contraceptive use and effectiveness. The cited clinic-contact data neither measure sexual activity nor capture every source or method of contraception.

Do the country graphs show that HPV vaccination caused a rise in cervical cancer? No. The cropped graphs do not provide enough source information to verify every plotted series, and they contain no individual vaccination status or adequate causal comparison. Even accurate population trends cannot, by themselves, establish that vaccination caused an increase.

Does the highlighted “14.0 versus 14.0” row show that original Gardasil failed? No. Within the later trial, it shows that Gardasil 9 did not outperform original Gardasil on a broad all-type endpoint. It does not compare Gardasil with placebo, and its event rates cannot be directly substituted for the within-trial results from FUTURE II.

Does the absence of a mortality RCT mean there is no causal evidence? No. Randomized trials show prevention of oncogenic HPV infection and high-grade precancer, while individual-level registry studies show lower invasive-cancer incidence among vaccinated women. The new mortality study adds population-level evidence consistent with that causal chain, but it is not a randomized or individually linked mortality comparison.

Does the declaration of interests establish Merck ghostwriting? No. It discloses Merck support for a separate trial, along with the authors’ use of ChatGPT for literature searching and Copilot for an initial draft of the Summary. Those disclosures warrant transparency and scrutiny, but neither demonstrates that Merck wrote, funded, or manipulated the mortality study.

Are we making the opposite mistake? We would be if we treated the mortality study as individual-level proof or claimed that vaccination alone explains every avoided death. The BBC headline was too causal-sounding, and the estimate of roughly 200 fewer deaths remains observational and model-dependent. Those limitations do not make the alternative explanations true.

What would make us reconsider? Credible linked evidence showing no reduction in vaccine-type precancer or invasive cancer among vaccinated women; a reanalysis showing that the age-at-vaccination gradient disappears after proper control for screening, behaviour and calendar trends; or evidence that type replacement cancels the reduction in disease caused by vaccine-targeted HPV types. I found no credible evidence meeting those tests in the sources reviewed for this article.

What can responsibly be said

The BBC compressed a nuanced study into a causal-sounding headline. The mortality paper itself cannot tell us that every woman in the zero-death cohort was vaccinated, or that vaccination alone accounts for every change in cervical-cancer mortality. Its estimate of roughly 200 fewer deaths depends on a modelled counterfactual and should be described that way.

Those are real limitations.

The rebuttal goes much further than the evidence allows. It takes a population-wide zero and uses it to imply that vaccination status made no difference, something the study could not test. It invokes routine screening as the explanation for a group younger than England’s routine screening age, treats conception rates as a direct measure of sexual activity, and places the vaccine rollout in the wrong year.

It presents cropped ecological graphs as evidence of vaccine-caused increases without individual vaccination data or adequate cohort comparisons. It turns a disclosed relationship with Merck into an allegation of Merck authorship without evidence. And it treats an all-type modified intention-to-treat result from the later Gardasil 9-versus-Gardasil trial as though it were directly comparable with FUTURE II’s within-trial efficacy estimates. None of those inferences follows from the evidence presented.

The strongest conclusion does not rest on the BBC headline, or even on the mortality paper by itself. It rests on the whole evidentiary chain:

• randomized trials show that HPV vaccination prevents persistent oncogenic HPV infection and high-grade cervical precancer;

• pathology and epidemiology show that persistent high-risk HPV causes the overwhelming majority of cervical cancers, while CIN3 and adenocarcinoma in situ are established high-grade precursor lesions; rare HPV-independent cervical cancers nevertheless exist;

• individual-level national registries show substantially less invasive cervical cancer among vaccinated women, especially when vaccination occurs before likely exposure;

• population studies find the largest incidence and mortality reductions in the earliest, highest-coverage cohorts, with smaller reductions after later vaccination; these patterns are consistent with a vaccine effect but remain observational.

One study can be confounded. One graph can mislead. One headline can overstate. But an alternative explanation has to account for all of the evidence, not merely point to something else that changed around the same time.

The iPhone does not do that.

Evidence reviewed through 9 September 2026. This article concerns vaccine effectiveness against cervical cancer; individual medical decisions should be discussed with an appropriately qualified clinician.

Key sources

• Sasieni P, Falcaro M. Cervical cancer mortality trends following HPV vaccination in England, 2001–24. The Lancet. 2026.

• Falcaro M, et al. Effects of England’s national HPV vaccination programme on cervical cancer and CIN3 incidence. The Lancet. 2021.

• Arbyn M, et al. Prophylactic HPV vaccination to prevent cervical cancer and its precursors. Cochrane Database of Systematic Reviews. 2018.

• Bergman H, et al. HPV vaccination for the prevention of cervical cancer and other HPV-related diseases: a network meta-analysis. Cochrane Database of Systematic Reviews. 2025.

• Henschke N, et al. Effects of HPV vaccination programmes on community rates of HPV-related disease and harms. Cochrane Database of Systematic Reviews. 2025.

• Lei J, et al. HPV vaccination and the risk of invasive cervical cancer. New England Journal of Medicine. 2020.

• Middeldorp M, et al. Effect of bivalent HPV vaccination against invasive cervical cancer and CIN3+ in the Netherlands. The Lancet Regional Health – Europe. 2025.

• Palmer TJ, et al. Invasive cervical-cancer incidence following bivalent HPV vaccination in Scotland. Journal of the National Cancer Institute. 2024.

• FUTURE II Study Group. Quadrivalent vaccine against HPV to prevent high-grade cervical lesions. New England Journal of Medicine. 2007.

• Joura EA, et al. A nine-valent HPV vaccine against infection and intraepithelial neoplasia. New England Journal of Medicine. 2015.

• UK Health Security Agency. Human papillomavirus: the Green Book, chapter 18a.

On September 3, 2026, John Catanzaro, Nicolas Hulscher, Raphael Stricker, Jamie Waselenko, and Peter McCullough uploaded a new paper titled Potential Oncogenicity of Synthetic mRNA Vaccines: Convergent Mechanistic, Clinical, and Population Evidence for a Concurrent-Hit Model of Accelerated Malignancy. The accompanying publicity was considerably less tentative.

Hulscher announced that the question had now been “fully answered,” declaring:

“TURBO CANCER IS REAL.”

The numbers are certainly attention-grabbing: 35 proposed cancer-promoting mechanisms, 333 reported cases across 27 countries, as many as 197,000 excess U.S. cancer deaths, rising early-onset cancers, two large observational studies, and a recently terminated mRNA cancer trial. Put together, it sounds overwhelming.

But before treating all of that as one body of proof, there is a simpler question to ask:

What does each piece actually show?

A plausible biological mechanism is not the same thing as a clinical outcome. A case report is not the same thing as a population-level risk estimate. An association is not the same thing as causation.

That does not make those findings meaningless. It means they have to be weighed according to what they can actually establish.

And that is where the paper’s argument starts to weaken.

First, this is a preprint

As of September 6, Zenodo labels the paper as a preprint. It has a DOI, but a DOI is a permanent identifier, not a certificate of peer review.

That does not make the paper worthless. Plenty of important research appears first as a preprint.

It does mean the paper has not yet gone through the usual peer-review process, where independent reviewers challenge the methods, assumptions, source selection, and whether the conclusions actually match the evidence.

For a paper making claims this consequential, that extra scrutiny matters.

That is particularly relevant here because the paper’s public messaging goes substantially further than much of its underlying literature.

What do “35 mechanisms” actually mean?

The centerpiece is an impressive-looking diagram entitled “35 Mechanisms by Which mRNA-LNP Vaccines May Induce or Accelerate Cancer.” The operative word is may. The authors group molecular phenomena into four broad routes involving oncogenic signaling, mutation pressure, protein-interaction networks, and cancer stem-cell behavior. Some of the underlying biological findings are real and deserve attention.

But “35 mechanisms” can easily be heard as 35 independently demonstrated ways in which vaccination causes cancer. That is not what has been shown. Consider ribosomal frameshifting.

A 2024 Nature paper genuinely demonstrated that N1-methylpseudouridine-containing mRNA can cause +1 ribosomal frameshifting. The researchers also detected immune responses against some resulting frameshifted products in mice and humans following BNT162b2 vaccination.

That is an interesting finding. But the researchers explicitly reported no evidence that those products were associated with adverse outcomes in vaccinated humans. They described the work as important for improving the design and translational fidelity of future mRNA therapies. They did not report cancer, malignant transformation or genomic instability caused by the phenomenon.

The new preprint effectively extends the chain:

The first portion has experimental evidence. The complete chain does not. The same problem appears with LINE-1 reverse transcription.

A widely cited 2022 experiment exposed Huh7 liver cancer cells in culture to BNT162b2 and detected DNA corresponding to vaccine mRNA. The investigators reported evidence of reverse transcription under those laboratory conditions. But reverse transcription is not genomic integration.

Integration is not insertional mutagenesis. And insertional mutagenesis is not cancer. The authors themselves acknowledged that they had not established whether the reverse-transcribed DNA integrated into the cell genome and said additional studies would be necessary.

For the cancer claim to be established, researchers would need to demonstrate something closer to:

The experiment did not do that. This distinction repeats throughout the “35 mechanisms.” Real molecular findings are frequently followed by one or more hypothetical biological steps and eventually connected to carcinogenesis.

That is a research hypothesis. It is not 35 demonstrations that vaccination produces cancer. The number itself is another problem. Cancer signaling pathways are highly interconnected. STAT3, MYC, NF-κB, inflammatory cytokines, immune surveillance, stem-cell behavior and tumor-microenvironment signaling are not 35 isolated biological systems.

Dividing a network into numerous molecular nodes can produce an impressive count without producing 35 independent causal observations.

Counting pathways is not the same as measuring cancer risk.

Then there are the “333 turbo cancer cases”

This may be the easiest claim to misunderstand.

The 333 patients come primarily from a 2026 review by cancer researchers Charlotte Kuperwasser and Wafik El-Deiry published in Oncotarget. The review identified 69 publications. Of those, 66 article-level reports described 333 patients, while the remaining publications included population-level or longitudinal analyses.

Fifty-five of the 69 publications — about 80% — were single-patient case reports or small case series. Importantly, those 333 patients were not 333 cancers proven to have been caused by vaccination. The review included malignancies reported after COVID-19 vaccination, after SARS-CoV-2 infection, and in some patients with both exposures.

And the paper contains an unusually clear warning about what those reports mean. The review states that it:

“was not designed to estimate cancer risk or incidence, nor to draw causal inferences”

The authors characterize the literature as early-stage, hypothesis-generating safety-signal detection requiring much more rigorous investigation. That is very different from saying 333 people developed cancer because of vaccination. Case reports are useful. Medicine has discovered important adverse effects because clinicians noticed unusual events and reported them.

But case reports lack the critical denominator. Suppose 333 cancers were diagnosed following vaccination. Were 33 expected? Three hundred? Three thousand? Without knowing how frequently the same cancers would have occurred in comparable people who were not vaccinated, the number alone cannot answer that question.

Billions of vaccine doses have been administered. In a population that large, cancers, recurrences, heart attacks, strokes, miscarriages and deaths will inevitably occur after vaccination by coincidence. That does not prove none are caused by vaccination. It means “after” cannot by itself establish “because of.”

And that is not my objection imposed on the Kuperwasser-El-Deiry review. It is the limitation stated by the review’s own authors. The new paper takes literature explicitly described as incapable of establishing causation and then uses it as one of the converging pillars supporting the declaration that the causal question has effectively been answered. That is an evidentiary upgrade the source itself does not support.

South Korea: a real signal, but not a causal answer

The South Korean study deserves more attention because, unlike case reports, it actually included a comparison group.

Using a national insurance database covering 8.4 million people, the researchers reported higher one-year rates of several cancers among vaccinated individuals, including thyroid, gastric, colorectal, lung, breast, and prostate cancer.

Those findings are real and should not be dismissed.

But they are still associations. The harder question is whether vaccination caused those differences, and that requires ruling out other explanations such as screening patterns, healthcare use, prior infection, and other sources of bias.

A peer-reviewed methodological critique published in 2026 identified several major problems. Most importantly, vaccinated and unvaccinated people were assigned different index dates: the vaccination date for vaccinated participants versus January 1, 2022 for the unvaccinated. That can create calendar-time bias because the two populations are not entering follow-up under equivalent conditions.

The critique also noted that more than 77% had previous SARS-CoV-2 infection, yet the study did not perform the sensitivity analysis necessary to separate that exposure adequately. Important variables such as smoking, alcohol use, family history, genetic susceptibility and screening frequency were unavailable.

South Korea also has extensive national cancer-screening programs. If vaccinated people interact with the healthcare system differently from unvaccinated people, previously existing cancers can be detected at different rates. That is surveillance bias.

Thirty cancer-site and subgroup comparisons were also conducted without adjustment for multiple testing, increasing the likelihood that some statistically significant findings appear by chance. The new preprint attempts to address the short one-year latency by arguing that vaccination may accelerate an already existing microscopic cancer, rather than create a solid tumor from nothing.

That hypothesis cannot simply be dismissed on latency grounds. But it makes the epidemiological problem even harder, because a one-year increase in detection of previously occult tumors is exactly the situation in which differences in healthcare utilization and cancer screening become especially important.

The Korean study is therefore worth replicating with better causal methods. It is not proof that vaccination caused those cancers.

Italy: the missing half of the result

The Italian study involved 296,015 residents of Pescara province. Its headline finding is accurately quoted by Hulscher: people receiving at least one vaccine dose had an adjusted hazard ratio of 1.23 for cancer hospitalization compared with unvaccinated residents. But there is an important result missing from the viral summary.

When the researchers required at least 12 months between vaccination and cancer hospitalization, the association reversed direction. It was also significant only among people without documented previous SARS-CoV-2 infection. The study’s own authors describe their cancer findings as preliminary, citing healthy-vaccinee bias and unmeasured confounding that they could not quantify. Cancer hospitalization is also not the same thing as newly developed cancer. A hospitalization database can capture diagnosis, treatment, recurrence, and healthcare-use differences.

None of this means the original HR of 1.23 should be ignored. But presenting the 1.23 while omitting the reversal under a longer latency analysis creates a considerably stronger impression than the study itself warrants.

The “197,000 excess cancer deaths” claim has a much bigger problem

This is where the paper’s population argument becomes most vulnerable.

Look carefully at its U.S. mortality figure. The vertical axis is “cancer deaths per day.”

Those are raw counts. They are not age-adjusted cancer mortality rates.

That distinction matters because the U.S. population is both growing and aging. Since cancer risk rises steeply with age, the total number of cancer deaths can increase even while the age-adjusted cancer mortality rate declines.

Official CDC final mortality statistics show:

2021: 146.6 deaths per 100,000
2022: 142.3 deaths per 100,000
2023: 141.8 deaths per 100,000
2024: 139.4 deaths per 100,000

The age-adjusted cancer mortality rate fell 2.9% from 2021 to 2022, was essentially unchanged from 2022 to 2023, and then fell another 1.7% in 2024.

Here is why raw counts can be misleading.

Cancer deaths increased from 613,352 in 2023 to 619,876 in 2024.

Yet during the same period, the age-adjusted cancer mortality rate fell from 141.8 to 139.4 per 100,000.

So both statements are true at the same time:

More people died of cancer in absolute numbers.

The age-adjusted cancer death rate declined.

This does not, by itself, invalidate every counterfactual excess-death model. A model could still argue that cancer mortality declined less than would have been expected in the absence of some additional exposure.

But the official age-adjusted data establish something narrower and important:

U.S. cancer mortality did not surge after 2021. It continued declining overall.

A rising raw-death curve therefore cannot, by itself, be interpreted as evidence of an emerging cancer hazard unless population size, age structure, and the expected baseline trend are handled properly.

And even if an excess above a statistical forecast were established, another enormous step remains:

What caused the excess?

A time-series change occurring after vaccination cannot independently distinguish vaccination from SARS-CoV-2 infection, delayed diagnosis, disrupted cancer treatment, demographic shifts or other pandemic-era effects. The National Cancer Institute’s 2025 Annual Report to the Nation, produced jointly with CDC, ACS and NAACCR, found that overall cancer mortality continued declining through 2022. That is extraordinarily difficult to reconcile with the public characterization of a massive nationwide vaccine-driven cancer mortality catastrophe.

It does not exclude a signal in a particular cancer, age group or susceptible subgroup. It does mean that 119,000 to 197,000 “vaccine-caused cancer deaths” has not been demonstrated.

What about the increase in early-onset cancers?

Here again, there is a genuine problem wrapped in an unjustified causal interpretation. Overall early-onset cancer incidence, and several important cancer types, really are increasing. But those trends were already present before anyone received a COVID-19 vaccine.

A large SEER analysis published in JAMA Network Open examined Americans younger than 50 from 2010 through 2019. Overall early-onset cancer incidence increased significantly during that period, with particularly rapid growth in gastrointestinal cancers. The pattern was not uniform: incidence declined among men overall and for several individual cancer sites. That matters when a graph places a bright red line at 2021 labeled “mRNA INJECTIONS BEGIN.” A vertical line is not a causal model.

There was also a major pandemic disruption in cancer diagnosis. NCI reports that cancer incidence fell sharply in 2020 as screening and routine medical care were disrupted, then largely returned toward pre-pandemic levels in 2021. Importantly, NCI researchers found little evidence of a compensatory rebound large enough to account for all of the diagnoses missed in 2020. So pandemic disruption should not be offered as a complete explanation for subsequent cancer trends. It does, however, make comparisons anchored around 2020 and 2021 unusually difficult to interpret.

SEER contains extraordinarily useful information about cancer trends. But placing the vaccine rollout date onto an ecological incidence graph does not convert SEER into a vaccinated-versus-unvaccinated study. The registry does not establish that the individuals responsible for the increase were vaccinated, when they were vaccinated, which vaccine they received or whether the trend differed appropriately according to dose. The graph is therefore a hypothesis generator, not a causal test.

One finding I would not dismiss: the LNP mouse study

There is, however, one new piece of evidence in this discussion that deserves serious follow-up. A 2026 study in Nano Today reported that lipid nanoparticles used experimentally in mice could create inflammatory conditions favoring pulmonary metastasis.

The researchers did not administer a licensed Pfizer or Moderna COVID-19 vaccine. They prepared experimental LNP formulations containing ALC-0315, an ionizable lipid used in Pfizer’s vaccine formulation, along with other lipid components, and tested them in mouse tumor models.

The researchers found that LNP injection caused local muscle injury and release of mitochondrial DNA, followed by neutrophil activation and formation of neutrophil extracellular traps. In their experimental mouse metastasis model, this environment facilitated establishment of tumor cells in the lungs. That is a legitimate preclinical finding that warrants replication and further safety investigation. It should be replicated.

It should be tested with different LNP formulations, clinically relevant exposures and doses, different cancer models, animals with spontaneous or preexisting tumors, and ultimately compared against carefully designed human epidemiology. But notice what the experiment did not establish.

It did not show that COVID vaccination initiates human cancers. It did not demonstrate increased metastatic disease among vaccinated people. It did not establish that every LNP formulation behaves identically.

And a mouse experimentally challenged with tumor cells is not equivalent to a person receiving an intramuscular vaccine. Calling the finding irrelevant would be unreasonable. Calling it proof of a human “turbo cancer” epidemic would be equally unreasonable.

Science lives in the distance between those two statements.

The failed BioNTech cancer trial

Finally, there is the recently terminated BNT122-01 trial. Unlike most of the material being discussed, this actually was randomized. BioNTech was testing autogene cevumeran, an individualized mRNA neoantigen treatment, in patients with high-risk colorectal cancer who had undergone surgery but still had circulating tumor DNA.

The trial crossed its prespecified futility boundary and was ultimately terminated. Its futility boundary was crossed in October 2025. At that point the independent Data Safety Monitoring Board concluded that the efficacy data were immature and follow-up insufficient.

With no identified safety concerns and no objection from the DSMB, BioNTech chose to continue the trial. In August 2026, the DSMB subsequently observed a numerical imbalance in overall survival, concluded that continued enrollment was unlikely to change the efficacy outcome, and recommended termination.

But BioNTech’s regulatory disclosure also states:

“No new safety signals were identified by the DSMB regarding autogene cevumeran.”

That doesn’t make the survival imbalance unimportant. I want to see the complete data: deaths by treatment arm, causes of death, recurrence, disease progression, confidence intervals, baseline characteristics, and duration of follow-up.

Until those results are available, however, there is a major difference between:

“An mRNA cancer treatment failed, and a numerical survival imbalance emerged.”

and

“The mRNA treatment accelerated cancer and killed patients.”

Only the first statement is presently established. And an individualized cancer immunotherapy encoding patient-specific tumor antigens is not biologically interchangeable with a COVID vaccine merely because both employ mRNA and lipid delivery.

What evidence would actually settle this?

If the concurrent-hit hypothesis is correct, it is testable. The most informative next steps would be:

• Large exposure-linked cohorts comparing vaccinated and unvaccinated people with equivalent calendar-time entry, healthcare utilization, screening history, SARS-CoV-2 infection history and established cancer risk factors.

• Prespecified analyses of cancer recurrence and progression in people who already had cancer, where an “acceleration” hypothesis can be tested more directly than new cancer incidence.

• Dose-response and timing analyses using exact vaccination dates and products rather than ecological state-level vaccination rates.

• Replication of the LNP metastasis experiments across formulations and clinically relevant dosing, followed by prospective translational studies.

• Molecular evidence in tumors demonstrating the proposed pathway linking a particular exposure to the malignant clone, rather than simply showing that a cancer appeared sometime afterward.

If those studies produce a reproducible signal, it should be taken seriously regardless of how politically inconvenient the result might be. If they do not, the hypothesis should weaken accordingly. That is how a falsifiable scientific claim is supposed to work.

Are We Making the Opposite Mistake?

Rejecting the claim that this paper proves an mRNA-driven cancer epidemic does not prove that mRNA vaccines can never influence cancer biology.

Some findings deserve serious follow-up. The Korean study reported associations that should be replicated with better controls. The Italian study produced an initial cancer-hospitalization signal.

The LNP mouse experiment raised a legitimate question about metastatic environments. And the BNT122 survival imbalance deserves examination when the full trial data are released.

So the correct response is not:

“There is absolutely nothing here.”

It is:

“There is not enough here to justify the causal conclusion being claimed.”

That distinction matters. If future well-designed studies show reproducible increases in cancer incidence, recurrence, or progression tied to vaccination, especially with dose-response, appropriate latency, individual exposure data, and biological corroboration, the conclusion should change. But scientific skepticism has to work both ways.

We should not dismiss a signal because we dislike its implications. And we should not promote a signal into proof because we like its implications.

The question is not whether the hypothesis is uncomfortable. The question is whether the evidence has earned the conclusion.

What the paper actually establishes

The paper succeeds at one thing: it assembles a large number of questions that deserve investigation. It does not establish the answer claimed in the publicity surrounding it. Some molecular phenomena are real.

The cancer case reports are real. The Korean statistical associations are real. The initial Italian HR is real.

Overall early-onset cancer incidence and several important cancer types really are increasing. The LNP mouse metastasis finding is real. And the BioNTech colorectal trial crossed its futility boundary and was terminated.

None of those facts requires denial or minimization. But the causal chain connecting them remains unproven.

The recurring transformation is:

The first three can justify investigating the fourth. They cannot simply substitute for it. This distinction is especially important because the authors call for immediate market withdrawal of nucleoside-modified mRNA-LNP products from broad preventive use and a halt to their expansion into adjuvant and neoadjuvant cancer treatment. That recommendation requires evidence commensurate with its magnitude. The preprint does not provide it.

There is a perfectly defensible scientific position between “mRNA technology could never affect cancer biology” and “turbo cancer has now been proven.”

It is this:

There are specific mechanistic and epidemiological signals worth testing rigorously, but the available evidence does not establish that COVID-19 mRNA vaccines are causing or accelerating a population cancer epidemic.

Thirty-five proposed routes are not 35 proven cancer mechanisms.

Three hundred thirty-three temporally associated cases are not 333 causally attributed cases. A rising raw cancer-death count is not the same thing as a rising age-adjusted cancer mortality risk. And a failed mRNA cancer trial does not retrospectively prove that COVID vaccines cause cancer.

The new paper has assembled a research agenda.

It has not “fully answered” the question.

Resources & Further Reading

For readers who want to examine the evidence directly, these are the primary papers, datasets, and methodological critiques discussed in this article.

• The paper being evaluated: Catanzaro JA, Hulscher N, Stricker RB, Waselenko JK, McCullough PA. Potential Oncogenicity of Synthetic mRNA Vaccines: Convergent Mechanistic, Clinical, and Population Evidence for a Concurrent-Hit Model of Accelerated Malignancy. Zenodo, 2026. Read the Zenodo preprint

• Cancer case-report literature: Kuperwasser C, El-Deiry WS. Review of reported cancers following COVID-19 vaccination and infection. Oncotarget, 2026. Particularly important because the authors explicitly caution that their review cannot estimate cancer incidence or establish causation. Read the Oncotarget review

• mRNA frameshifting: Mulroney TE et al. Research demonstrating that N1-methylpseudouridylation can produce unintended ribosomal frameshifting. Nature, 2024. The study demonstrated altered translation but did not demonstrate cancer or adverse clinical outcomes from the phenomenon. Read the Nature study

• LINE-1 / reverse-transcription experiment: Aldén M et al. Intracellular Reverse Transcription of Pfizer BioNTech COVID-19 mRNA Vaccine BNT162b2 In Vitro in Human Liver Cell Line. Current Issues in Molecular Biology, 2022. This was an in-vitro experiment in Huh7 cancer cells and did not establish genomic integration in vaccinated humans. Read the study

• South Korean cancer study: Kim H et al. 1-year risks of cancers associated with COVID-19 vaccination: a large population-based cohort study in South Korea. Biomarker Research, 2025. The study reported positive associations for several cancers but cannot by itself establish causation. Read the study on PubMed Central

• Methodological critique of the Korean study: A subsequent analysis discusses index-date differences, surveillance bias, residual confounding, prior infection, multiple comparisons, and the difficulty of interpreting cancers diagnosed within a short latency period. Read the methodological critique

• Italian population study: The Pescara cohort reported an initial HR of 1.23 for cancer hospitalization among vaccinated residents, while longer-latency analyses changed the result and the authors emphasized substantial limitations and possible confounding. Read the Italian study

• Early-onset cancer trends before COVID vaccination: Koh B et al. analyzed SEER data showing that several early-onset cancers were already increasing during 2010–2019. JAMA Network Open, 2023. Read the SEER analysis

• U.S. cancer mortality: CDC/NCHS final mortality statistics provide age-adjusted cancer death rates alongside raw death counts. CDC 2024 mortality statistics

• Pandemic-era cancer diagnosis: NCI analyses document the sharp decline in new diagnoses during 2020 and the return toward pre-pandemic levels in 2021, while finding little evidence of a compensatory rebound large enough to recover all missed diagnoses. NCI pandemic diagnosis analysis

• National cancer trends: The National Cancer Institute’s Annual Report to the Nation on the Status of Cancer tracks incidence and mortality using population-standardized methods and documents the continuing long-term decline in overall U.S. cancer mortality. NCI Annual Report to the Nation

• LNP and metastasis experiment: The 2026 Nano Today study reporting increased metastatic establishment in a mouse model following particular lipid-nanoparticle exposures is one of the more important preclinical findings discussed in the article and warrants independent replication. Read the Nano Today study

• BNT122/autogene cevumeran trial: BioNTech’s August 2026 company disclosure describes termination of the Phase 2 colorectal-cancer trial after futility and a numerical overall-survival imbalance, while also reporting that the independent DSMB identified no new safety signal related to the treatment. Full peer-reviewed trial results were not yet available when this article was written. Read BioNTech’s trial update

A note on how to read these sources

Readers should distinguish between mechanistic plausibility, cell experiments, animal studies, case reports, observational associations, and causal evidence in humans. Each can contribute useful information, but they answer different questions and carry very different evidentiary weight. That distinction is the central issue examined in this article.

At some point in human history, someone cracked open an oyster, stared directly into what appeared to be a small bag of seawater and phlegm, and thought:

Dinner.

There was no menu.

No Yelp review.

No guy standing nearby saying, “They’re actually really good with lemon.”

Just a wet rock containing something that looked very much as though nature had gone out of its way to tell us not to put it in our mouths.

And yet somebody did.

Then, presumably, somebody else watched him.

And instead of saying, “Frank has finally lost his mind,” he said, “Give me one.”

This may be one of the most underappreciated forces in human history.

Nearly everything we eat required, at some point, a first volunteer.

We talk about human beings as toolmakers, explorers, inventors and problem solvers. All true.

But we are also the species that wandered around the planet pointing at suspicious objects and asking:

Can I eat that?

The answer was occasionally no.

Very, very no.

But every so often the answer was cheese.

And that kept us going.

The Oyster Problem

Oysters are a good place to begin because they require almost no imagination to understand the problem.

Fruit advertises itself.

Fruit develops bright colors, sweetness and aromas specifically because plants benefit when animals eat it and spread the seeds.

An oyster does none of this. An oyster hides inside a rock. You have to find it, pry it open and then make a conscious decision to swallow the contents.

And humans have been doing exactly that for an astonishingly long time.

Archaeological evidence shows Indigenous peoples harvesting oysters for thousands of years. Ancient coastal settlements left behind enormous piles of discarded shells, known as shell middens.

Which means oysters weren’t some desperate experiment attempted once during a particularly bad Tuesday.

People liked them.

Imagine the first sales pitch.

“Okay, hear me out.”

“No.”

“You haven’t even seen what’s inside.”

“I have seen what’s inside.”

“But you haven’t tasted it.”

“That is because I possess survival instincts.”

Somehow Team Oyster won.

Thousands of years later, human beings put them on silver platters, charge $4 apiece for them and argue about which coastal inlet produces the finest flavor.

That is quite a promotion for something that spends its entire life glued to the bottom of the sea.

And Then Someone Let Milk Go Bad

Cheese requires an even stranger leap.

Milk makes sense immediately.

A mammal produces it. Babies drink it. Humans figured out that the milk of domesticated animals was nutritious.

Fine.

But cheese?

Cheese begins with someone looking at milk that is no longer behaving like milk and refusing to throw it away.

It curdled.

It separated.

It smelled different.

Today we call this cheesemaking.

Without several thousand years of cultural context, it would have looked suspiciously similar to a kitchen accident.

Archaeological evidence indicates that humans were making cheese thousands of years ago. Residues found in ancient pottery strainers provide evidence of cheesemaking in prehistoric Europe roughly 7,500 years ago, while broader archaeological evidence places dairy fermentation even earlier in parts of the Near East.

Someone discovered that milk could be transformed into something that lasted longer, traveled better and tasted completely different.

That discovery eventually gave humanity Parmesan, mozzarella, cheddar, brie, provolone and the ability to charge an extra $2.50 to put cheese on a hamburger.

Progress.

But imagine getting there without knowing what bacteria were.

There was no microbiology.

No fermentation science.

No refrigeration.

Just:

“The milk became a lump.”

“Throw it away.”

“Wait.”

“No.”

“I think I can do something with this.”

Human cuisine owes an enormous debt to people who refused to throw things away when any reasonable person would have.

Now Let’s Eat Something That Hurts

At least cheese tastes like food.

Chili peppers present a different question.

Why would you eat something that causes your nervous system to report an emergency?

Capsaicin, the compound responsible for the heat in chili peppers, activates some of the same sensory pathways involved in detecting high temperatures. Your mouth isn’t literally burning, but your nervous system is responding to a chemical signal that resembles one.

The rational response would seem obvious:

Stop eating it.

Humanity’s response was:

Can you make it hotter?

People in the Americas were eating wild chilies thousands of years ago and cultivating them more than 6,000 years ago. After peppers traveled around the world, they became so deeply embedded in Asian cuisines that it is easy to forget that chili peppers did not arrive in places such as India, Thailand, Korea or China until after contact between the Americas and the Old World.

Think about what happened psychologically.

Person One eats pepper.

Person One begins sweating.

Eyes water.

Nose runs.

Mouth feels as though it has been placed against the surface of the sun.

Person Two observes all of this.

Person Two says:

“I want one.”

And here we encounter something distinctly human.

We don’t merely tolerate discomfort.

Under the right circumstances, we enjoy it.

Researchers have proposed several explanations for our attraction to spicy foods, including cultural learning and the pleasure some people get from controlled risk or intense sensation.

A roller coaster frightens you without actually throwing you off a mountain.

A horror movie terrifies you while you remain safely on the couch.

A hot pepper convinces your nervous system that something terrible is happening while you calmly reach for another chicken wing.

Humans apparently discovered recreational suffering long before CrossFit.

Coffee Makes Even Less Sense

Then there is coffee.

Today coffee is so normal that we rarely consider what a bizarre chain of decisions was required to produce it.

Coffee fruit grows on a plant.

Inside that fruit are seeds.

We take those seeds.

Dry them.

Roast them until brown.

Grind them into powder.

Pour hot water through the powder.

Throw the powder away.

Drink the brown water.

And somehow this process became so important to civilization that millions of people refuse to engage in conversation before completing it.

The exact origin of coffee drinking is murky, and the famous story of Kaldi, the Ethiopian goat herder who supposedly noticed his goats becoming energetic after eating coffee berries, is legend rather than established history. Coffee itself has deep roots in Ethiopia, while Yemen became crucial to its early cultivation and development as the beverage we recognize today.

I prefer the goat story anyway.

Partly because it involves goats.

But mostly because it perfectly captures the recurring principle of human culinary discovery.

An animal eats something strange.

The animal behaves differently.

A human watches.

And instead of thinking, “I should keep the goats away from that,” the human thinks:

I wonder what happens if I eat it.

Science, ladies and gentlemen.

Early-stage science, admittedly.

But science.

The Person Who Figured Out Cassava Deserves a Statue

Some foods took more than courage.

They required genuine ingenuity.

Cassava is one of the world’s important staple crops, but the plant contains naturally occurring cyanogenic compounds that can release cyanide. Improperly processed cassava can cause serious poisoning. Traditional processing methods developed by Indigenous peoples made the food safer to eat.

That is remarkable.

Imagine encountering a plant that isn’t safely edible in its natural state and deciding not simply to abandon it, but to experiment until you discover a way to turn it into food.

This is where the joke about our ancestors eating random things stops being entirely a joke.

They were observant.

They experimented.

They remembered.

Most importantly, they passed knowledge forward.

One person learned something.

Another person improved it.

A community preserved the method.

Eventually something that could harm you became a dependable source of calories.

That process happened again and again throughout human history.

Fermentation.

Cooking.

Soaking.

Drying.

Grinding.

Roasting.

Smoking.

Salting.

Pickling.

We did not simply discover food.

We invented ways of making food out of things that did not initially cooperate.

There Must Have Been a Lot of Bad Days

Of course, we are seeing only the winners.

History remembers bread.

History doesn’t remember whatever Kevin ate in 14,000 BC that caused everybody else in the village to agree never to touch the purple berries again.

That knowledge was just as important.

Human beings learned food partly through success and partly through what must have been an enormous, multigenerational game of:

Don’t eat that.

The accumulated knowledge became culture.

These mushrooms are safe.

Those mushrooms aren’t.

Cook this first.

Don’t eat that part.

Harvest this when it turns this color.

Soak that.

Dry this.

Ferment this one.

If Uncle Grog starts hallucinating after eating those leaves, make a note.

Long before written language, people were building enormous databases of practical information inside traditions, stories, habits, and recipes.

Your grandmother telling you exactly how long something has to cook is the descendant of a very old system.

Except somewhere near the beginning of that chain, the consequences of getting the recipe wrong were considerably worse than dry chicken.

The First Person Is Still With Us

We tend to imagine ancient humans as fundamentally different from us.

They weren’t.

The same impulse still exists.

Walk through any supermarket and look at how many products exist solely because somebody asked:

“What happens if we combine these?”

Hot honey.

Chocolate-covered bacon.

Pickle-flavored potato chips.

Cronuts.

Sushi burritos.

Deep-fried Oreos.

Pizza with cheeseburgers embedded in the crust.

We have laboratories, food scientists, health departments and ingredient labels now, but somewhere deep inside the human brain, the original experiment continues.

Can I eat that?

Sometimes the question becomes:

Should I eat that?

And occasionally:

How much will you pay me to eat that?

TikTok has merely accelerated a research program that began in prehistory.

Curiosity Was Probably the Secret Ingredient

There is a temptation to look backward and imagine that everything humans discovered was inevitable.

Of course people discovered cheese.

Of course somebody roasted coffee.

Of course people learned to ferment grapes.

Of course somebody ate the oyster.

But none of it was inevitable to the person standing there before it had been done.

Every familiar food hides thousands of forgotten experiments.

Someone had to notice that heat changed food.

Someone had to discover that fermentation could preserve it.

Someone had to learn which plants were dangerous.

Someone had to discover that the terrifying red pepper was actually delicious.

Someone had to open the oyster.

And behind all of those discoveries was one of the strangest characteristics of our species.

Curiosity frequently overrides common sense.

That sounds like a defect.

Sometimes it is.

It is also why we crossed oceans, climbed mountains, built telescopes, split atoms, walked on the Moon and created several thousand varieties of cheese.

Human progress has always depended on somebody being willing to ask a question whose answer wasn’t obvious.

Sometimes the question was profound.

What are the stars?

Sometimes it changed civilization.

Can we grow this ourselves?

And sometimes it was simply:

What happens if I put this in my mouth?

Fortunately for the rest of us, somebody was usually willing to find out.

So the next time you’re sitting in a restaurant looking at a plate of oysters, a wheel of cheese, a bowl of painfully hot peppers and a cup of coffee, consider what you’re actually looking at.

It isn’t just lunch.

It’s a record of thousands of years of human curiosity, experimentation, occasional recklessness and the accumulated wisdom of people whose names we will never know.

Raise your coffee to them.

Especially the oyster guy.

That man took one for the team.

A Mind Less Wasted

Question everything. Especially the first person who looked at an oyster and said, “Pass me a fork.”

Resources and Further Reading

Smithsonian Institution — “Indigenous Peoples Have Shucked Billions of Oysters Around the World Sustainably.”
Archaeological evidence documenting Indigenous oyster harvesting over hundreds and thousands of years, including extensive shell middens.
Read at Smithsonian

University of Bristol — “Chemical Analysis of Sieve Vessels Reveals First Cheese Making in Northern Europe in the 6th Millennium BC.”
Research published in Nature identifying milk residues in prehistoric pottery strainers and providing direct evidence of cheesemaking more than 7,000 years ago.
Read at the University of Bristol

Smithsonian Magazine — “What’s So Hot About Chili Peppers?”
A look at the archaeology and biology of chili peppers, including evidence that humans were eating wild chilies at least 8,000 years ago and cultivating them more than 6,000 years ago.
Read at Smithsonian Magazine

National Library of Medicine — “A Molecular Perspective on Identifying TRPV1 Thermosensitive Regions and Disentangling Polymodal Activation.”
A scientific review of TRPV1, the sensory receptor activated by both heat and capsaicin that helps explain why chili peppers produce a burning sensation.
Read at PubMed Central

Oxford Research Encyclopedia of Food Studies — “The Ethiopian and Yemeni Roots of Coffee.”
An academic history of coffee’s origins in Ethiopia, its development in Yemen, and the evidence separating coffee history from legends such as the story of Kaldi and his goats.
Read at Oxford Academic

World Health Organization — “Natural Toxins in Food.”
An overview of naturally occurring food toxins, including the cyanogenic glycosides found in cassava and the potential for cyanide poisoning when cyanogenic foods are improperly prepared.
Read at the World Health Organization

Food and Agriculture Organization of the United Nations — “An Overview of Traditional Processing and Utilization of Cassava in Africa.”
A detailed review of traditional cassava processing techniques, including fermentation, drying and other methods used to reduce cyanide content and make cassava suitable for consumption.
Read at the FAO

A man in an X discussion challenged me to produce a vaccine-safety study.

I pointed him to a nationwide Danish cohort study of 657,461 children. The researchers used linked health registries and more than five million person-years of follow-up to examine a specific question: Did MMR vaccination increase the risk of an autism diagnosis?

The study did not answer every conceivable question about every vaccine. It was not a randomized trial, and its follow-up came from registries rather than repeated contact with each family. Those are fair distinctions. A serious argument should never pretend that one study establishes more than it does.

But it should not pretend that a study establishes nothing because it does not establish everything.

The paper reported its main estimate, confidence interval, methods, covariates, crude cumulative-incidence graph, subgroup analyses, sensitivity analyses, and limitations. The supplementary material even included the SAS input and output for the main model.

His first response was that he could not find the results.

I pointed him to the Results section.

Then the objection became the use of adjusted analysis.

I pointed him to the unadjusted cumulative-incidence graph and explained why age and other differences between groups matter in an observational study.

Then he said he wanted “the raw data.” That phrase can mean two different things: unadjusted summary results, which the paper and supplement provide, or the underlying individual-level registry records, which are restricted by privacy rules. Those are not the same request.

Every question in that sequence can be legitimate. The issue is what happens after an answer is supplied. Does the answer change our confidence, even slightly? Or does a new requirement immediately replace the old one?

That is where skepticism can stop functioning as a method of inquiry and start functioning as a shield.

I cannot know another person’s private motives from a social-media exchange. Neither can anyone else. What we can examine is the visible reasoning pattern: Were the standards defined before the evidence appeared? Did they remain stable afterward? Were they applied just as strictly to evidence supporting the preferred conclusion?

Those questions get us much closer to the difference between critical thinking and conclusion-protection.

A hypothesis must risk something

The broad claim in this discussion was that children who receive more vaccines experience “more harm than benefit on balance.”

That sounds testable until we try to define the test.

What counts as “more vaccines”: injections, doses, antigens, products or visits? Which benefits count: infections prevented, hospitalizations avoided, disability prevented, deaths prevented or reduced transmission? Which harms count, over what period, and how should events of radically different severity be compared? Who belongs in the comparison group? Most important, what result would count against the claim?

Those are not evasions. They are the work required to turn a conviction into a hypothesis.

No single study can settle the net balance of every vaccine, every outcome and every population. A study can ask a narrower question and answer it well or badly. The Danish study asked about MMR and autism. Its value depends on how well it answered that question—not on whether it resolved the entire childhood schedule in one paper.

Until a broad claim is broken into measurable questions, it can absorb almost any result. If one outcome shows no association, another can be substituted. If one time window shows nothing, a different window can be chosen. If a large cohort is unfavorable, it can be rejected for using adjustments. If crude results are supplied, the demand can shift to individual records.

A hypothesis that cannot lose is not strong. It is protected.

A National Research Council report on scientific inquiry describes empirically testable and refutable questions, methods capable of challenging competing explanations, explicit reasoning, professional scrutiny and replication as guiding principles. It also makes an important qualification: there is no single mechanical recipe called “the scientific method.” Observation, measurement, exploration, model building and hypothesis testing may occur in different sequences.

What holds those activities together is exposure to correction. Scientific reasoning makes it possible, at least in principle, to discover that an explanation is incomplete or wrong.

Two different jobs for reasoning

Psychologist Ziva Kunda distinguished reasoning aimed at accuracy from reasoning aimed at a preferred conclusion.

Under an accuracy goal, we try to use the rules and evidence most likely to produce a correct answer. Under a directional goal, we search memory, evidence and argument for a defensible route to the answer we want. That does not necessarily feel dishonest. We still believe we are reasoning. We still require an explanation that sounds plausible to us.

That is what makes the problem difficult to see from the inside.

Confirmation bias is often described too simply as looking only for information that agrees with us. The more revealing form is unequal scrutiny. Supporting evidence is accepted at face value. Contrary evidence is inspected for every conceivable weakness.

In a classic 1979 experiment, Charles Lord, Lee Ross and Mark Lepper showed people with opposing views two studies that appeared to support different conclusions about capital punishment. Participants judged the study supporting their existing position more favorably and found more faults in the study pointing the other way.

Later researchers examined related patterns under labels such as disconfirmation bias and motivated skepticism. We do not always stop thinking when evidence threatens a belief. Sometimes we think harder—but mainly to defeat it.

This does not mean every disagreement is psychological defensiveness. People can have different background knowledge, assumptions or legitimate concerns about source credibility. Researchers have warned against inferring a person’s motivation merely from a directionally biased judgment. The safer test is behavioral:

• Are the standards stable?

• Are they applied symmetrically?

• Does contrary evidence ever reduce confidence?

• Does answering an objection move the discussion forward?

  ---

The moving standard

Consider the sequence in the MMR discussion:

1. Show me a safety study.

2. Where are the results?

3. Those results are adjusted.

4. Show me the raw data.

This sequence alone cannot prove anyone’s motive. Nor is asking a follow-up question the same as moving the goalposts. Good scrutiny often produces new questions.

The warning sign is more specific: a stated requirement is met, but meeting it never changes the person’s confidence because another requirement instantly takes its place. The conclusion remains fixed while the standard needed to challenge it keeps changing.

There is also an important fairness point here. The Danish paper used registry follow-up, not active clinical contact. If active follow-up is an essential criterion, that rule should be stated clearly and applied consistently to every registry study—whether its result supports or challenges the preferred position. Rejecting this paper on that ground would still not make it irrelevant to the narrower MMR-autism question. It would identify a limitation to weigh.

The asymmetry becomes easier to see when favorable evidence receives gentler treatment. In this discussion, a self-selected survey promoted by Steve Kirsch was defended as support for a vaccine-autism hypothesis. Such a survey may generate questions, but without representative recruitment, verified records and a suitable comparison group, it cannot provide a reliable population estimate of comparative risk.

Meanwhile, a nationwide registry study was treated as unusable even though it reported crude incidence patterns, detailed methods, sensitivity analyses and the model input and output. If “raw data” meant the individual registry records, public release was not a realistic requirement under Danish privacy rules.

That is a result-dependent hierarchy of evidence:

• A weak method supports the conclusion, so its weaknesses are treated as tolerable.

• A stronger method challenges the conclusion, so every limitation is treated as fatal.

This is the point at which “question everything” becomes “question only what threatens my position.”

Adjustment is not evidence tampering

The objection to adjusted analysis deserves attention because it sounds reasonable: Why not look at the raw comparison first?

We should. The Danish study did. Figure 2 shows the crude cumulative incidence of autism by age and MMR-vaccination status. The supplement provides additional unadjusted incidence graphs and crude associations for the variables used in the analysis.

But crude does not mean pure, and adjusted does not mean corrupted.

In an observational study, the groups being compared may differ in ways that affect the outcome. Here, MMR status changed over time: a child contributed unvaccinated follow-up before receiving MMR and vaccinated follow-up afterward. Autism diagnosis also varies strongly with age. Ignoring age would not remove researcher judgment; it would build a known distortion into the comparison.

The researchers therefore used Cox regression with attained age as the underlying time scale. In the fully adjusted model, they stratified the baseline hazard by birth year, sex, other early-childhood vaccinations, sibling history of autism and decile of an autism-risk score based on measured factors.

The main estimate comparing MMR-vaccinated with MMR-unvaccinated follow-up was an adjusted hazard ratio of 0.93, with a 95% confidence interval from 0.85 to 1.02. In plain language, the study did not find an increased autism hazard associated with MMR. Because the confidence interval includes 1.00, it also does not establish that MMR reduced autism risk.

The investigators tested the result several ways. They changed the age at which follow-up ended, required two autism registrations instead of one, examined autism categories separately, accounted for the second MMR dose, replaced stratification with covariate adjustment and substituted the individual risk factors for the combined risk score. The estimates remained similar.

The study was not flawless. It was observational, residual confounding remained possible, and the researchers did not review individual medical charts to validate every diagnosis. Those limits belong in the assessment. They do not erase the design, the data, or the result.

Could adjustment be done badly? Certainly. Researchers can choose inappropriate variables, omit important confounders or select models after seeing the outcome. That is why we inspect the rationale for the variables, the stated analysis, crude patterns, alternative models, and sensitivity analyses.

“Adjusted” is not a synonym for “manipulated.” It is an instruction to inspect the adjustment.

The table repeatedly posted in the exchange did not contain the MMR estimate. It was Supplement Table 2, titled “Decile cutpoints and hazard ratios from the autism risk score model.” The crude hazard ratios in that table describe the autism-risk score used in the analysis—not the association between MMR and autism.

That is the danger of searching a paper for something that looks suspicious before establishing what the table actually measures.

Raw data matter, but they are not a universal veto

Requests for data can be entirely appropriate. Reanalysis can uncover errors, test alternative assumptions, and increase confidence. The National Academies distinguishes reproducibility—obtaining consistent computational results with the same data and methods—from replication, in which a new study addresses the same question with new data.

Open data make reproduction easier. Medical records, however, cannot always be posted for public download.

Statistics Denmark provides authorized institutions with secure access to pseudonymized microdata. Its transfer rules prohibit removing individual-level microdata and require disclosure control before analysis results leave its research systems. Those protections limit what an ordinary reader can reproduce at home. They do not turn the underlying records into nonexistent evidence.

The Hviid supplement did provide the full SAS input and output for the main Cox model. That is not the same as releasing the microdata, but it makes the model specification and reported output inspectable.

Restricted data are a real limitation. The proper response is to ask whether qualified researchers can obtain controlled access, whether the variables and methods are described, whether code or model specifications are available, whether sensitivity analyses are reported and whether independent studies using other populations reach compatible results.

Demanding transparency is scientific. Declaring every study that protects confidential medical data invalid is not.

Exploration is not the enemy

There is nothing wrong with noticing a pattern and developing a hypothesis afterward. Much of science begins that way.

The problem comes when we forget which stage we are in.

Brian Nosek and colleagues describe preregistration as one way to preserve the distinction between prediction and postdiction. Researchers specify the primary question, measures, exclusions and analysis before seeing the relevant results. Preregistration does not guarantee a good study. It makes later changes visible.

Exploratory analysis asks, What might be happening here? Confirmatory analysis asks whether a specific prediction survives a test capable of failing. Both matter. Trouble begins when an exploratory pattern is presented as though it had been predicted in advance, or when outcomes, time windows and evidentiary standards are repeatedly revised after the results are known.

This is not a problem unique to vaccine critics or to one political side. Researchers are vulnerable too. Joseph Simmons, Leif Nelson and Uri Simonsohn showed how undisclosed flexibility in sample size, outcomes, exclusions and statistical models can make false-positive findings easier to produce. Scientific safeguards exist because scientists are human.

The same habits should guide public reasoning:

• Define the claim before selecting the evidence.

• State what observation would weaken it.

• Separate exploration from confirmation.

• Apply the same evidentiary rules regardless of the result.

• Confront the strongest contrary evidence, not merely the evidence easiest to dismiss.

• Revise confidence instead of switching standards.

  ---

The mirror test

There is a simple way to check whether skepticism has become selective:

If this study had produced the opposite result, would I make the same methodological criticism with the same intensity?

Charles Lord, Mark Lepper and Elizabeth Preston tested a version of this strategy in 1984. Simply telling people to be objective did little. Asking them to consider how they would evaluate the same evidence if it had produced the opposite result reduced biased assimilation.

The question works because it separates the method from the result. It forces us to decide whether a design is unacceptable or merely inconvenient.

Apply that test here. If a nationwide registry analysis of 657,461 children had reported a credible increase in autism after MMR, would the same critic dismiss it because the main estimate was adjusted? Would restricted access to the individual records make the finding unmentionable?

Now reverse the test. If a self-selected parent survey had found no relationship between vaccination and reported autism onset, would he still defend its recruitment, verification and comparison methods?

The answers tell us whether the methodological standard is real or result-dependent.

Can your belief lose?

Before claiming that the evidence supports what you already believe, ask:

1. What exactly is my hypothesis?

2. What result would make me reduce my confidence in it?

3. Did I choose that standard before or after seeing the result?

4. Would I accept this method if its conclusion favored the other side?

5. Am I treating a limitation as something to weigh—or as permission to discard the entire study?

6. Have I demanded stronger evidence from opponents than from my own side?

7. Am I describing an exploratory clue as though it were a confirmatory test?

8. When an objection is answered, does my confidence change at all?

If no possible answer changes the conclusion, the discussion is no longer about evidence.

What critical thinking actually requires

Critical thinking is not the ability to find a flaw. Every study has limitations. Every dataset has boundaries. With enough effort, an intelligent person can construct an objection to almost anything.

The harder skill is deciding how much a flaw matters—and applying that judgment consistently.

Real skepticism does not mean distrusting institutions by default or trusting them by default. It does not require accepting one large study as final. It asks whether the study addressed the question it claimed to address, whether its methods handled plausible alternative explanations, and whether its result should change our confidence when considered with the rest of the evidence.

Science advances because its conclusions remain exposed. Better measurements, stronger designs, failed replications and new evidence can challenge them. A conclusion protected by endlessly changing requirements may survive every argument, but that survival tells us nothing about whether it is true.

The most important question in critical thinking is not, Can I find a reason to reject this?

It is this:

What evidence would I allow to change my mind?

If the honest answer is “none,” skepticism is no longer serving the search for truth. It is guarding the door.

Follow the evidence with us

Critical thinking is easy when it challenges someone else. The harder test is applying it to our own conclusions.

If you value careful, evidence-based examinations of the claims spreading across social media, subscribe to A Mind Less Wasted. It’s free, and every new article will be delivered directly to your inbox.

Subscribe now

Resources and further reading

• Ziva Kunda, “The Case for Motivated Reasoning”, Psychological Bulletin (1990).

• Charles G. Lord, Lee Ross and Mark R. Lepper, “Biased Assimilation and Attitude Polarization”, Journal of Personality and Social Psychology (1979).

• Kari Edwards and Edward E. Smith, “A Disconfirmation Bias in the Evaluation of Arguments”, Journal of Personality and Social Psychology (1996).

• Charles S. Taber and Milton Lodge, “Motivated Skepticism in the Evaluation of Political Beliefs”, American Journal of Political Science (2006).

• Ben M. Tappin, Gordon Pennycook and David G. Rand, “Thinking Clearly About Causal Inferences of Politically Motivated Reasoning”, Current Opinion in Behavioral Sciences (2020).

• Brian A. Nosek and colleagues, “The Preregistration Revolution”, Proceedings of the National Academy of Sciences (2018).

• Joseph P. Simmons, Leif D. Nelson and Uri Simonsohn, “False-Positive Psychology”, Psychological Science (2011).

• National Research Council, Scientific Research in Education: Guiding Principles for Scientific Inquiry (2002).

• National Academies of Sciences, Engineering, and Medicine, Reproducibility and Replicability in Science (2019).

• Anders Hviid and colleagues, “Measles, Mumps, Rubella Vaccination and Autism: A Nationwide Cohort Study”, Annals of Internal Medicine (2019).

• Statistics Denmark, Data for research and analysis and Rules on transfer of analysis results.

• Charles G. Lord, Mark R. Lepper and Elizabeth Preston, “Considering the Opposite: A Corrective Strategy for Social Judgment” (1984).

Subscribe now

On August 31, President Donald Trump posted an image on Truth Social making a striking claim:

“330 years later… No published case has explicitly documented an unvaccinated Amish child with autism.”

The graphic cited a 2010 autism study of Amish children. Researchers screened 1,899 children, confirmed seven cases of autism spectrum disorder, and reported that they had collected vaccination histories. What they did not report was the vaccination status of those seven children.

That omission has become the centerpiece of a much larger argument.

Within hours, the post was being treated in some corners of social media as something close to a natural experiment: the Amish vaccinate less, autism appears less common among them, and nobody can point to a published case explicitly documenting autism in a completely unvaccinated Amish child. Therefore, the reasoning goes, vaccination must be the missing variable.

Steve Kirsch, a technology entrepreneur and prominent vaccine-safety activist, pushed the argument even further, calling the Amish a “perfect control group” and asserting that the only significant difference between autistic and non-autistic Amish children is vaccination.

That conclusion is far ahead of the evidence.

But there is an uncomfortable detail underneath the rhetoric that deserves more than a dismissive fact-check.

The researchers really did collect vaccination histories.

And as far as I can determine, those histories were never published.

That does not establish that completely unvaccinated Amish children never develop autism. It does not establish that vaccination caused the autism cases that were identified. And it certainly does not turn the Amish into a randomized control group.

But sixteen years later, it leaves a fair question sitting in plain sight:

What did those vaccination histories show?

That is where this story becomes interesting.

The 2010 Study Behind Trump’s Post

The study cited in Trump’s graphic was presented at the 2010 International Meeting for Autism Research under the title Prevalence Rates of Autism Spectrum Disorders Among the Old Order Amish.

One qualification matters immediately: this was a conference abstract, not a full peer-reviewed journal article.

Researchers conducted door-to-door screening in Amish communities in Holmes County, Ohio, and Elkhart-Lagrange County, Indiana. They screened 1,899 children using two autism screening instruments.

Twenty-five children screened positive on at least one measure. Fourteen subsequently received more extensive evaluations. Seven were ultimately classified as having autism spectrum disorder.

Then comes the sentence that has received renewed attention:

“A Vaccination History and a brief family history including questions specific to the ASD phenotype were also taken.”

So Trump’s graphic is correct about something important: vaccination histories were collected during the study.

What the abstract does not tell us is whether the seven children diagnosed with ASD were fully vaccinated, partially vaccinated, or never vaccinated.

Nor does it provide vaccination information for the broader screened population in a form that would allow us to calculate autism prevalence by vaccination status.

I have also been unable to locate a later peer-reviewed publication from this project reporting those data.

That missing information is worth discussing.

What it does not do is tell us what the missing information would have shown.

“1 in 271” Is Preliminary, Not Definitive

The seven identified ASD cases among 1,899 screened children produce the frequently quoted figure of approximately 1 in 271.

But that number needs context.

Twenty-five children screened positive. Only 14 of those 25 received the more extensive evaluation, and seven of those fourteen were classified as having ASD.

We therefore do not know how many of the eleven remaining screen-positive children might have met diagnostic criteria had they also completed the evaluation.

Screening itself is another issue. A screening instrument is designed to identify children who may need further evaluation. It is not a perfect census of every autistic child in a population.

A large real-world study published in Pediatrics involving nearly 26,000 children found that a commonly used toddler autism screening process identified substantially fewer eventual ASD cases than might be assumed from controlled validation studies. That does not tell us how well the specific tools used in the Amish study performed, but it illustrates an important point: measured prevalence depends partly on how cases are sought and confirmed.

The Amish investigators themselves noted another complication. Results differed somewhat depending on the diagnostic instrument used, and they raised the possibility that Amish caregiver reporting styles and cultural factors could influence the assessment.

The authors therefore described their findings appropriately:

preliminary data.

The approximately 1-in-271 estimate is real.

It simply should not be treated as a definitive measurement of the true prevalence of autism throughout the Amish population.

That distinction becomes even more important when people compare it directly with the current CDC estimate of roughly 1 in 31 among 8-year-old children.

The CDC figure comes from active surveillance across 16 U.S. sites using medical, educational, diagnostic and administrative records. It is a prevalence estimate for a very different population, period and surveillance system.

Putting 1 in 271 beside 1 in 31 makes for a striking meme.

It does not create an apples-to-apples epidemiological comparison.

But They Collected Vaccination Histories

This is the part that should not be brushed aside.

According to the study’s own methods, vaccination histories were collected.

Yet the abstract does not report them.

Sayer Ji, founder of the natural-health website GreenMedInfo and a longtime critic of vaccination policy, recently highlighted the same omission in an article arguing that the forgotten vaccination data are central to the renewed Amish-autism controversy. Ji reports searching for a subsequent publication and being unable to find one that disclosed the vaccination status of the identified ASD cases.

He reports searching for a subsequent publication and being unable to find one that disclosed the vaccination status of the identified ASD cases.

On that narrow point, he has identified something worth asking about.

The abstract also indicated that additional work was underway.

Sixteen years later, a vaccination-stratified autism analysis from the project does not appear to have been published.

Why?

I don’t know.

Neither does Ji.

That difference matters.

“Unpublished” Does Not Mean “Suppressed”

Research projects routinely collect variables that never become published analyses.

Funding disappears. Researchers move. Follow-up recruitment fails. Data turn out to be incomplete. Consent agreements restrict later use. Samples become too small to answer the intended question. Projects simply end.

And yes, sometimes potentially useful data sit unpublished when they probably should have been made available.

Without hearing from the investigators, we do not know which explanation applies here.

That means the claim that the study was deliberately “buried” goes beyond the evidence currently available.

We can demonstrate nonpublication.

We cannot demonstrate suppression simply from nonpublication.

Still, I would like to know what happened.

If usable vaccination-history data remain and ethical and privacy requirements permit further analysis, publishing them could help resolve at least part of a controversy that has lingered for years.

What Does “No Published Case” Actually Mean?

This is where the argument frequently changes shape.

There is an enormous difference between these statements:

No published paper I can find explicitly identifies an Amish autistic child as completely unvaccinated.

and:

No completely unvaccinated Amish child has autism.

The first is a statement about the published literature. The second is a statement about reality. One does not prove the other.

Medical researchers also generally do not publish the identities of individual children alongside sensitive medical histories. The scientifically useful evidence would therefore not be somebody producing the name of a particular child.

What we would need is a study that systematically identifies autism and reliably determines vaccination history in the same population.

That specific Amish vaccinated-versus-never-vaccinated comparison apparently has not been published.

So the proper conclusion is not:

Vaccines therefore cause autism.

Nor should it be:

There is nothing here worth investigating.

It is:

We do not currently have the data necessary to answer that specific Amish vaccinated-versus-never-vaccinated question.

But That Does Not Mean Vaccine-Autism Research Is Absent

This narrow uncertainty should not be confused with an absence of vaccine-autism research generally.

Large population studies have examined specific vaccine-autism hypotheses.

A 2019 nationwide Danish cohort followed 657,461 children and found no increased autism risk associated with MMR vaccination, including in the susceptible subgroups and post-vaccination periods examined.

A U.S. study of 95,727 children likewise found no increased ASD risk following MMR vaccination, including among children with an older sibling with autism.

And a 2014 meta-analysis pooling cohort and case-control evidence involving more than a million children found no association between vaccination and autism.

Those studies do not answer every conceivable question about every vaccine, every combination of vaccines, the entire childhood schedule, or Amish children specifically.

They should not be represented as though they do.

But the reverse is equally important.

An unpublished Amish vaccination analysis cannot reasonably be treated as though it outweighs the much larger body of evidence that actually has been published.

The uncertainty should remain exactly where the uncertainty exists.

The Amish Aren’t a “Perfect Control Group”

The Amish are potentially an interesting population for research.

They are not a naturally randomized vaccine experiment.

For one thing, the Amish do vaccinate.

Rates vary enormously between settlements, affiliations, families, and time periods.

A 2011 survey of Amish parents in Holmes County found that 68% said all of their children had received at least one immunization, another 17% said some of their children had, and 14% reported that none had.

A survey roughly a decade later found much lower vaccination uptake: 59% of respondents reported not vaccinating their children.

That alone tells us there is no single “Amish vaccination rate.”

The later study also reported something particularly important for causal inference: children with special needs were more likely to have been vaccinated than healthy children.

That does not tell us why.

But it is an immediate warning about confounding.

Families who use healthcare differently may also vaccinate differently. Families with children who have developmental or medical problems may interact with physicians more frequently. Conversely, some parents may stop vaccinating younger children after attributing a problem in an older child to vaccination.

Health status can therefore become correlated with vaccination status even when vaccination did not cause the health difference being measured.

This is exactly why merely dividing children into “vaccinated” and “unvaccinated” groups does not automatically produce a valid causal comparison.

The Amish are interesting precisely because their vaccination patterns vary.

But they are not a ready-made randomized control group.

Finding Autism Depends Partly on How Hard You Look

Another complication is almost completely absent from viral memes about Amish autism.

Autism prevalence is not measured identically everywhere.

A major systematic review of autism prevalence around the world found striking variation between populations and concluded that measured prevalence is influenced by diagnostic definitions, awareness, service availability, help-seeking behavior and methods of identifying cases.

More recent research on diagnostic inequities has likewise found that autism identification can be delayed or missed because of differences in healthcare access, cultural interpretation, provider practices, and availability of appropriate diagnostic services.

That does not prove autism is underdiagnosed among the Amish.

It means underascertainment is a plausible competing explanation that must be investigated rather than assumed away.

Consider how autism is identified in much of mainstream America: pediatric developmental screening, specialist referrals, school evaluations, speech therapy, occupational therapy, insurance claims, special-education records and repeated interactions with medical systems.

Now compare that with a culturally distinct population whose relationships with formal education, medicine and disability services may differ.

A lower rate of recorded autism could therefore reflect some combination of true biological difference, case ascertainment, healthcare use, cultural recognition, genetics, environment, family structure, pregnancy factors, lifestyle, and many other variables.

Vaccination is a legitimate variable to include.

It is not the only difference between Amish and non-Amish children.

Sayer Ji Gets Something Important Right

Ji’s argument is more interesting than the simplistic statement that “the Amish prove vaccines cause autism.”

He acknowledges that the lack of an explicitly documented, completely unvaccinated Amish autism case is an argument from absence, not direct proof.

He also correctly calls attention to something that deserves attention: the 2010 researchers say they collected vaccination histories, yet those histories never appeared in the published abstract.

That is worth asking about.

Where Ji becomes less convincing is when an unanswered question begins turning into evidence for a much larger conclusion.

What the Institute of Medicine Actually Said

Ji points to a 2013 Institute of Medicine report on the childhood vaccination schedule.

Some of his description is correct.

The IOM recommended against conducting a randomized trial in which children would deliberately be assigned to receive the full vaccination schedule versus no vaccines or an alternative schedule. The committee did not believe such a trial was ethically justified.

It also considered prospectively studying naturally unvaccinated groups such as Amish communities.

And it did not recommend that approach as its preferred design, citing two major problems: potentially inadequate numbers of completely unvaccinated children and serious confounding because culturally distinct populations differ from more highly vaccinated populations in many ways besides vaccination.

That deserves to be stated plainly.

But the IOM did not conclude that comparisons between vaccinated and unvaccinated children should simply never be studied.

It discussed observational alternatives, including research using large healthcare databases such as the Vaccine Safety Datalink, and specifically considered comparing health outcomes among children with different vaccination patterns.

So the IOM position was not:

Don’t ask the question.

It was closer to:

Don’t answer it with a randomized trial or assume a small culturally distinct population automatically provides a clean causal experiment.

Whether researchers have since done enough with the observational alternatives is a separate question.

And that question is fair.

Are the Amish Being “Erased” as a Control Group?

Ji goes further, arguing that efforts to increase vaccination among Amish communities are effectively eliminating a valuable natural control group.

There is a narrow statistical point here worth conceding.

If the number of completely unvaccinated children declines, some vaccinated-versus-unvaccinated comparisons become harder to conduct with adequate statistical power.

That is true.

But that does not demonstrate the much larger claim that public-health efforts are intentionally trying to eliminate an unvaccinated population to prevent vaccine-safety research.

Showing that a policy has an effect is not the same as showing that producing that effect was the purpose of the policy.

Amish communities have experienced outbreaks of vaccine-preventable diseases. Public-health agencies therefore encourage vaccination for the same stated reason they encourage it elsewhere: disease prevention.

Researchers also do not get to preserve a community in a particular exposure state merely because it would create a convenient comparison group.

And increasing vaccination does not make observational research impossible. Amish vaccination patterns remain heterogeneous enough that meaningful comparisons may still be possible if researchers obtain reliable individual exposure data and account carefully for confounding.

The study would be difficult.

That does not mean it cannot be done.

So Why Hasn’t Someone Done the Obvious Study?

This is the part of the controversy that interests me most.

Why, after years of argument, don’t we have a large, carefully designed study of Amish and other Plain communities that includes standardized autism assessment and verified vaccination history?

There are legitimate obstacles.

• The proportion of completely unvaccinated children varies sharply by community and affiliation.

• Families accepting vaccines may differ systematically from families declining them.

• Medical records may be incomplete.

• Researchers would need diagnostic methods that work reliably in the cultural setting.

• They would need enough participants to produce meaningful estimates and enough community trust to obtain accurate information.

Amish populations also have distinctive founder genetics. That makes them scientifically interesting, but it creates another variable that must be considered rather than ignored.

None of those problems means the study should not be attempted.

A strong study would define its population and hypotheses before looking at outcomes. Vaccination histories would be verified where possible. Autism diagnoses would be established independently of vaccination status. Researchers would account for age, sex, genetics, healthcare use, pregnancy and birth factors, socioeconomic differences, and other plausible confounders.

And the results would be published whether they showed an association, no association, or something nobody expected.

That would move the discussion forward.

The Question Steve Kirsch Should Be Asking

There is a revealing difference between these two questions:

Can anyone name an unvaccinated Amish child with autism?

and:

What is autism prevalence among Amish children stratified by verified vaccination history?

The first makes for an effective social-media challenge.

The second is a scientific question.

And it is the second one we actually need answered.

If you begin with the conclusion that vaccination causes autism, every missing dataset starts looking suspicious. Every unexplained case becomes supporting evidence. And every failure to find contrary evidence becomes confirmation.

Science cannot work that way. Absence can generate a hypothesis. It cannot become the evidence that proves it.

The 2010 Amish autism study was real, although it was presented as a conference abstract rather than a full journal paper.

Researchers screened 1,899 children and reported seven confirmed ASD cases among the 14 screen-positive children who underwent the more extensive evaluation.

The resulting approximately 1-in-271 figure was explicitly preliminary and should not be treated as a definitive estimate of autism prevalence throughout Amish communities.

The researchers also reported collecting vaccination histories.

Those histories were not presented in the abstract, and I have not located a subsequent peer-reviewed publication reporting the vaccination status of the seven identified ASD cases.

That is a legitimate unanswered question.

But there is currently no evidence demonstrating that those data were intentionally suppressed because they showed vaccination caused autism.

Amish communities are also not uniformly unvaccinated. Vaccination practices vary substantially between communities, families, and time periods.

And even if no published case currently explicitly documents both complete nonvaccination and autism in an Amish child, that does not establish that such children do not exist.

There is something here worth investigating. Just not the conclusion being advertised.

One Question I Would Like Answered

I would genuinely like to hear from the investigators responsible for the 2010 project.

Do the vaccination histories collected during that 1,899-child screening project still exist?

If they do, were they complete? Were they based on medical records or parent reports? What was the vaccination status of the seven children ultimately classified as having ASD? Why was that information not reported? Were the eleven additional screen-positive children ever evaluated? And did the “further studies” mentioned in the abstract ever take place?

• Those are not anti-vaccine questions.

• They are not pro-vaccine questions either.

• They are simply questions the available record leaves unanswered.

• Maybe the answers would strengthen the vaccine-autism hypothesis.

• Maybe they would weaken it.

• Maybe the data are incomplete or unusable.

• Maybe they no longer exist.

Bottom line: The most interesting fact here is not that nobody can produce the name of an unvaccinated Amish autistic child. It is that researchers once collected vaccination-history information that might have helped answer the question — and sixteen years later, we still do not know what those data showed or whether they were sufficient to answer it.

If you found this useful, subscribe to A Mind Less Wasted for more evidence-first investigations that separate what the data actually show from what people want them to show.

Subscribe now

Resources and Further Reading

Robinson JL, Nations L, Suslowitz N, Cuccaro ML, Haines J, Pericak-Vance M. Prevalence Rates of Autism Spectrum Disorders Among the Old Order Amish. International Meeting for Autism Research, 2010. Preliminary 1,899-child screening project reporting seven confirmed ASD cases and collection of vaccination histories. View the conference abstract record

Wenger OK, McManus MD, Bower JR, Langkamp DL. Underimmunization in Ohio’s Amish: Parental Fears Are a Greater Obstacle Than Access to Care. Pediatrics. 2011;128(1):79–85. DOI: 10.1542/peds.2009-2599. PubMed Full article

Scott EM, Stein R, Brown MF, et al. Vaccination Patterns of the Northeast Ohio Amish Revisited. Vaccine. 2021;39(7):1058–1063. DOI: 10.1016/j.vaccine.2021.01.022. PubMed

Guthrie W, Wallis KE, Bennett AE, et al. Accuracy of Autism Screening in a Large Pediatric Network. Pediatrics. 2019;144(4):e20183963. DOI: 10.1542/peds.2018-3963. PubMed

Zeidan J, Fombonne E, Scorah J, et al. Global Prevalence of Autism: A Systematic Review Update. Autism Research. 2022;15(5):778–790. DOI: 10.1002/aur.2696. Full article

Srivarathan A, Bradford A, Shearkhani S, et al. Bridging Diagnostic Safety and Mental Health: A Systematic Review Highlighting Inequities in Autism Spectrum Disorder Diagnosis. BMJ Quality & Safety. 2026;35(7):487–497. DOI: 10.1136/bmjqs-2025-018723. Full article

Hviid A, Hansen JV, Frisch M, Melbye M. Measles, Mumps, Rubella Vaccination and Autism: A Nationwide Cohort Study. Annals of Internal Medicine. 2019;170(8):513–520. DOI: 10.7326/M18-2101. PubMed

Taylor LE, Swerdfeger AL, Eslick GD. Vaccines Are Not Associated With Autism: An Evidence-Based Meta-analysis of Case-Control and Cohort Studies. Vaccine. 2014;32(29):3623–3629. DOI: 10.1016/j.vaccine.2014.04.085. Journal article

Institute of Medicine. The Childhood Immunization Schedule and Safety: Stakeholder Concerns, Scientific Evidence, and Future Studies. National Academies Press, 2013. DOI: 10.17226/13563. Read the full National Academies report

Sayer Ji. Trump Just Pointed at the Low Autism Rates in Unvaccinated Amish. Here’s the Study They Buried and the Cohort They’re Erasing. August 31, 2026. Read Ji’s article

There is an easy way to write about Dr. Vladimir Zelenko.

Pick a side.

Either portray him as the courageous doctor who discovered a cheap COVID treatment that the medical establishment refused to acknowledge, or dismiss him as a physician who became convinced of something the evidence ultimately failed to support.

Neither version tells the whole story.

March 2020 was different from August 2026. COVID was new. Doctors had sick patients in front of them. There were no proven outpatient antivirals. Hospitals were filling, clinical guidance was changing rapidly, and physicians were trying to figure out what might keep people from getting worse.

Zelenko was a family doctor trying to keep his patients out of the hospital.

I have no reason to doubt that he believed deeply in what he was doing.

But sincerity and scientific validity are different questions.

Six years later, we don’t have to judge the Zelenko Protocol only by what was known in March 2020. We now have randomized trials, systematic reviews, additional laboratory research and years of clinical experience.

That later evidence matters.

Which brings us to a new document published by Zelenko Labs in March 2026: COVID-19, Viral Replication, and the Zelenko Protocol.

The paper presents a scientific explanation for the protocol centered largely on zinc, zinc ionophores and viral replication. It describes itself as a Zelenko Labs scientific whitepaper prepared for educational and research purposes.

But there is something immediately striking about its bibliography.

It contains six references.

The newest is from 2020.

A scientific defense of a controversial COVID treatment published in 2026 stops citing evidence at almost exactly the point when the large body of subsequent clinical evidence begins.

That deserves a closer look.

Start with what Zelenko got right

The biological idea behind the original protocol wasn’t invented out of nowhere.

Zelenko focused initially on hydroxychloroquine, zinc and azithromycin. One part of his reasoning was that zinc might interfere with coronavirus replication inside infected cells, while another substance could help increase intracellular zinc.

There was legitimate laboratory science behind the first part of that hypothesis.

In 2010, Aartjan te Velthuis and colleagues reported that zinc ions could inhibit the RNA-dependent RNA polymerase, or RdRp, used by SARS coronavirus. In cell culture, increasing intracellular zinc with an ionophore also interfered with viral replication.

That is a real and interesting finding.

But the details matter.

The virus was SARS-CoV, not SARS-CoV-2.

The ionophore used in the viral cell-culture experiment was pyrithione, not hydroxychloroquine.

And this was laboratory research, not a trial showing that zinc plus an ionophore treated coronavirus disease in human patients.

Yet a 2026 Zelenko Labs essay describing how the protocol was born says Zelenko found a 2010 paper showing that zinc ionophores “like hydroxychloroquine,” taken with zinc, were effective at treating coronavirus.

That goes beyond what the study showed.

The study supplied a good reason to investigate the idea.

It didn’t establish the treatment.

That distinction matters throughout this story.

From a promising idea to certainty

Frank Zelenko’s account of the protocol’s beginnings gives us a revealing picture of how his brother became convinced.

Patients were being treated.

His office followed up with them.

Many were recovering.

Frank describes calling elderly and vulnerable patients who were alive, well and back to their lives.

It is easy to understand why that would have been persuasive in March 2020.

The problem is that recovery after treatment does not tell us whether recovery happened because of treatment.

If 97 out of 100 treated patients recover, the result sounds extraordinary. But if 96 comparable untreated patients would also have recovered, the treatment contributed very little.

Without a comparison group, we don’t know.

That is why controlled trials exist.

Frank’s account says that once Zelenko became convinced he had found the treatment for COVID, he was not going to wait for blinded clinical studies.

Given the emergency confronting doctors in early 2020, trying a treatment on the basis of incomplete evidence is understandable.

But there is a difference between saying:

The evidence isn’t complete, but I believe the potential benefit outweighs the risk.

and saying:

I have established that this treatment works.

The first is a medical judgment made under uncertainty.

The second requires evidence showing that treated patients do better than comparable patients who don’t receive the treatment.

That evidence eventually arrived.

There actually was peer review

This part needs to be stated accurately.

It would be wrong to say that the Zelenko treatment approach was never examined in peer-reviewed literature.

In 2020, Roland Derwand, Martin Scholz, and Vladimir Zelenko published a retrospective outpatient study in the International Journal of Antimicrobial Agents.

The study reported outcomes for 141 patients treated with zinc, low-dose hydroxychloroquine and azithromycin and compared them with 377 untreated patients drawn from public community data.

Hospitalization occurred in 2.8% of the treated patients compared with 15.4% in the comparison population.

Deaths were 0.7% versus 3.4%. That mortality difference was not statistically significant.

Those results were worth investigating.

But the study wasn’t randomized.

The treated patients were selected according to Zelenko’s own risk-stratification criteria, while the comparison group came from public data. That makes it difficult to know how comparable the two groups really were or how much of the apparent difference was caused by treatment.

Peer review means that a journal subjected a paper to outside scientific review before publication.

It does not convert an observational study into a randomized controlled trial.

The appropriate conclusion from the 2020 study was that there was an interesting signal worth testing more rigorously.

And then it was.

The controlled studies eventually arrived

Hydroxychloroquine was studied extensively.

One objection often raised is that the drug failed only because researchers waited until patients were severely ill.

That isn’t the whole story.

A randomized double-blind trial enrolled 491 symptomatic nonhospitalized adults with confirmed or probable early COVID. Of those, 423 contributed primary-endpoint data. Hydroxychloroquine did not substantially reduce symptom severity compared with placebo.

Researchers also tested hydroxychloroquine as prevention.

In a randomized post-exposure prophylaxis trial involving 821 participants after moderate- or high-risk exposure, hydroxychloroquine did not significantly prevent COVID-compatible illness compared with placebo. Side effects were more common in the hydroxychloroquine group.

In Brazil, 667 hospitalized patients with mild-to-moderate disease were randomized to hydroxychloroquine, hydroxychloroquine plus azithromycin, or standard care. The primary modified intention-to-treat analysis included 504 patients with confirmed COVID. Neither hydroxychloroquine regimen improved clinical status at 15 days. QT prolongation and liver-enzyme abnormalities occurred more often in patients receiving hydroxychloroquine.

These were not the only studies.

By April 2025, the World Health Organization said its recommendation against hydroxychloroquine for treatment rested on 30 trials involving more than 10,000 COVID patients. Its prevention recommendation drew on six trials involving more than 6,000 participants.

That doesn’t prove that every possible combination, timing and dose could never produce any effect.

But it makes the claim that hydroxychloroquine is an established treatment for COVID very difficult to defend.

There is also a problem with the zinc-ionophore explanation

A central part of the Zelenko story is that hydroxychloroquine carries zinc across cellular membranes.

Later laboratory work challenged that explanation directly.

In 2022, researchers tested hydroxychloroquine’s interaction with zinc using chemical-binding experiments, artificial membranes and A549 human lung epithelial cells.

They found that hydroxychloroquine did not behave as a direct zinc ionophore.

The researchers did observe changes in intracellular zinc associated with hydroxychloroquine, so this doesn’t mean the drug has no effect whatsoever on zinc handling inside cells.

But the simple “HCQ carries zinc through the membrane” explanation did not hold up under direct testing.

That is especially relevant because the March 2026 Zelenko Labs whitepaper continues to describe hydroxychloroquine as one of the zinc ionophore options.

A review written in 2026 should address the later evidence.

This one doesn’t.

Zinc itself is more interesting

None of this means zinc is biologically irrelevant.

It isn’t.

Zinc is essential for normal immune function, and laboratory research showing that zinc can interfere with viral replication deserves attention.

Quercetin and EGCG have also demonstrated zinc-ionophore activity experimentally. The Zelenko Labs whitepaper cites the 2014 work of Dabbagh-Bazarbachi and colleagues showing this activity in cells and artificial membranes.

But again, the question changes once we leave the laboratory.

Can swallowing zinc and quercetin or EGCG reliably produce enough free intracellular zinc in infected human respiratory cells to inhibit viral replication?

At what dose?

For how long?

With what clinical effect?

And with what risks?

Those questions require human evidence.

Clinical studies of zinc have produced mixed results. Quercetin has produced some encouraging but still limited evidence. Vitamin D has remained an active area of study, particularly in people who are deficient.

That makes these legitimate areas for research.

It doesn’t prove the complete Zelenko Protocol.

The protocol eventually became much broader

The historical protocol currently presented by the Zelenko Protocol Educational Center contains considerably more than zinc and hydroxychloroquine.

For higher-risk patients it lists zinc, vitamin C, high-dose vitamin D, azithromycin or doxycycline, hydroxychloroquine and/or ivermectin.

Additional options include dexamethasone or prednisone, inhaled budesonide, anticoagulants, colchicine, monoclonal antibodies, oxygen and IV fluids.

That leads to an important point:

Some treatments associated with the broader protocol really did turn out to be useful.

Dexamethasone became an important treatment for patients with severe or critical COVID who required oxygen.

But timing matters. Corticosteroids were not beneficial as a blanket early treatment for otherwise stable patients and could be harmful when used at the wrong stage.

Inhaled budesonide produced a more modest signal. In the open-label PRINCIPLE trial, higher-risk community patients who received budesonide reported recovery roughly three days sooner. The hospitalization/death result did not meet the trial’s prespecified threshold for superiority.

Oxygen is obviously appropriate when a patient’s oxygen level is dangerously low.

Anticoagulation has legitimate roles in certain hospitalized patients.

Some monoclonal antibodies were genuinely effective against susceptible SARS-CoV-2 variants until newer variants escaped them.

Other treatments fared considerably worse. Larger and better randomized trials substantially weakened the case for ivermectin as a COVID treatment.

So the record isn’t:

Everything associated with Zelenko failed.

It didn’t.

But neither can we say:

Some components worked, therefore the Zelenko Protocol worked.

A treatment has to be judged according to the patient, dose, timing and outcome for which it was tested.

Then we get to the 2026 whitepaper

The new paper defines the “Zelenko Protocol” differently from the sprawling historical treatment regimen.

It says that for purposes of the paper, the term refers primarily to the strategy of increasing intracellular zinc through a zinc ionophore rather than to one fixed medication regimen.

That is the authors’ right.

But it also has an effect.

It centers the discussion on one of the most scientifically interesting parts of the story—the zinc mechanism—while much of the later clinical evidence surrounding hydroxychloroquine, ivermectin, azithromycin and other components largely disappears.

Then there is that reference list.

Six papers.

The publication years are:

2010
2020
2020
2020
2020
2014

Nothing from 2021.

Nothing from 2022.

Nothing from 2023.

Nothing from 2024.

Nothing from 2025.

Nothing from 2026.

That isn’t a minor omission.

Those missing years contain precisely the evidence needed to evaluate whether the early hypothesis survived clinical testing.

What about peer review?

The March 2026 PDF displays the identifier 10.5281/zenodo.19170511.

However, I was unable to independently verify a live Zenodo record for that identifier, and the Zelenko Protocol Educational Center currently lists the whitepaper’s DOI as “Pending Registration.”

That discrepancy may have a simple explanation. But until it is clarified, the safest thing is to report exactly what is visible rather than assume the status of the DOI.

More important is the peer-review question.

I can find no evidence that this March 2026 Zelenko Labs whitepaper has been published in a peer-reviewed journal, and I found no peer-reviewed journal version under its title.

That does not mean nobody has ever read or informally reviewed it.

It means I can find no conventional journal peer review of the document.

And that matters because the whitepaper makes strong claims.

For example, it says the protocol “works” for prevention and treatment and describes intracellular zinc as stopping viral replication.
A formal review would almost certainly have to deal with some obvious questions:

• Why does a 2026 literature review end in 2020?

• Why isn’t the 2022 research challenging hydroxychloroquine’s direct zinc-ionophore mechanism discussed?

• Where are the randomized outpatient hydroxychloroquine studies?

• Where are the prophylaxis trials?

• Where are the later ivermectin trials?

• What evidence shows that the proposed oral doses create antiviral concentrations of intracellular zinc in infected human respiratory tissue?

And most importantly:

• Where is the controlled clinical evidence showing that this mechanism produces the claimed patient benefit?

Those questions remain unanswered in the whitepaper.

From advising Honduras to “saving” Honduras

Another Zelenko Labs essay is titled:

“When Dr. Zelenko Saved Honduras.”

Frank Zelenko describes how his brother made contact with Honduran officials and says President Juan Orlando Hernández arranged for him to train more than 1,000 Honduran physicians.

There may well be an important historical story there.

Honduras did pursue aggressive early multidrug treatment.

But contemporary Honduran sources describe the country’s national programs, known as MAÍZ and CATRACHO, as protocols developed by Honduran physicians, including Miguel Sierra-Hoffman, Omar Videa and colleagues.

MAÍZ included Microdacyn, azithromycin, ivermectin and zinc.

That was not Zelenko’s original hydroxychloroquine, azithromycin and zinc regimen.

An analysis of Honduras later found that implementation of its multidrug programs was associated with declining case-fatality rates.

But an association across an entire country cannot tell us that the medications caused the decline.

Testing changed. Medical capacity changed. Who was being diagnosed changed. The epidemic changed.

A review in the Revista Médica Hondureña concluded that the MAÍZ regimen lacked sufficient evidence demonstrating effectiveness and called for clinical trials.

That doesn’t mean Zelenko played no role.

He may have influenced Honduran officials or physicians.

But there is a large evidentiary gap between:

Dr. Zelenko advised people involved in Honduras’s early-treatment effort.

and:

Dr. Zelenko saved Honduras.

The second makes a causal claim the available evidence cannot establish.

The marketing angle cannot simply be ignored

Selling supplements doesn’t invalidate someone’s science.

A physician can develop a product.

A researcher can own a company.

A company can fund scientific work.

None of those things automatically means the research is wrong or the people involved are dishonest.

But commercial context matters.

And Zelenko Labs does not keep the medical claims, the stories about Zelenko’s legacy and sales of Zelenko-branded supplements in separate worlds.

Consider the article Five Times Dr Zelenko Was Right.

It opens by telling readers that virtually every conspiracy theory from the pandemic has now become fact.

It presents five Zelenko claims as capital-T “TRUTH.”

Those include claims involving COVID vaccine lots, Bill Gates, vaccine injury, masks and influenza vaccination.

The article then tells readers that these truths helped motivate Zelenko to create an over-the-counter supplement line.

From there, it moves into elderberry, promotes Z-Flu Elderberry Gummies, describes purported health benefits and advertises its extract as having four times greater bioavailability than competing formulas.

The article eventually tells readers to stock up on Z-Stack and Z-Flu, promotes free shipping and offers a discount for new customers.

Then, at the bottom, comes the disclaimer:

The claims have not been evaluated by the FDA, and the products are not intended to diagnose, treat, cure or prevent disease.

That sequence deserves attention.

Not because it proves fraud.

It doesn’t.

But the reader is being asked to move through a very specific story:

Zelenko was right.

The establishment was wrong.

His understanding of viruses was vindicated.

His formulations were based on that understanding.

And here are the products carrying his name.

When medical authority and scientific certainty help sell a product, the evidence supporting those claims deserves especially careful scrutiny.

CHECKPOINT: What do we actually know?

This is where it helps to separate what the evidence supports from what people may believe about it.

The zinc/RdRp hypothesis had a legitimate experimental basis.

Zelenko had reasonable grounds to investigate it in the extraordinary uncertainty of early 2020.

His initial outpatient results were interesting enough to publish and study further.

His 2020 observational study was peer-reviewed. But it wasn’t randomized and couldn’t establish that the treatment caused the better outcomes that were observed.

Subsequent randomized hydroxychloroquine trials, including early outpatient and prophylaxis studies, failed to show the clinical benefit predicted by the original theory.

Later laboratory research also challenged the simple claim that hydroxychloroquine directly carries zinc through cell membranes.

Some treatments that later became part of broader Zelenko-style regimens proved useful for particular patients at particular stages of disease. Others did not.

Zinc and compounds such as quercetin remain scientifically interesting. That is not the same thing as having established the complete protocol.

The March 2026 Zelenko Labs whitepaper has not been shown to have undergone conventional journal peer review, and I can find no peer-reviewed journal version of it.

Its six references end in 2020.

Zelenko Labs also sells branded supplements, and some of its medical and historical articles lead directly into promotion of those products.

None of that tells us what Vladimir Zelenko privately intended.

It does tell us to keep five things separate:

belief, hypothesis, observational evidence, demonstrated clinical effect and marketing.

They aren’t interchangeable.

The question isn’t whether Zelenko was a good man

Criticism of a medical claim does not require criticism of the physician’s character.

Zelenko died in 2022.

The 2026 whitepaper, the current website, the recent Honduras story and today’s supplement marketing are being produced by people carrying his work forward.

That distinction matters.

Maybe Zelenko sincerely believed he had discovered something extraordinary.

The accounts from his brother certainly suggest that he did.

Maybe he would have disputed the interpretation of the later trials.

We can’t ask him.

But we can examine the evidence that exists now.

Science doesn’t freeze at the moment somebody becomes convinced.

A hypothesis becomes stronger when it survives serious attempts to prove it wrong.

A treatment becomes credible when properly designed studies show that patients receiving it do better than comparable patients who don’t.

And when later evidence changes the picture, the conclusion should change with it.

Six years later

My problem isn’t that Zelenko looked for an early treatment. Doctors should look for treatments.

My problem isn’t zinc. Zinc biology is worth studying.

My problem isn’t even that Zelenko became convinced before randomized trials were available. March 2020 forced doctors to make decisions under extraordinary uncertainty.

My problem is what happened after the evidence changed.

There was a plausible idea.

There were encouraging patient stories.

There was an observational study suggesting benefit.

Then came controlled trials.

More laboratory research.

Systematic reviews.

Updated clinical recommendations.

Six years later, Zelenko Labs published a new scientific whitepaper.

And its references stop in 2020.

At the same time, the website preserving and promoting Zelenko’s legacy sells supplements based in part on the health philosophy associated with his name.

While that does not prove a sinister motive, it does leave us with a question worth asking:

Are we being shown the current scientific case for the Zelenko Protocol, or are we being shown the version of the story that best preserves the Zelenko legacy and the products built around it?

Perhaps there is some of both.

Dr. Zelenko deserved to have his hypothesis tested. It was.

People who continue to believe in his work deserve to see the results of those tests too.

If a scientific case for the Zelenko Protocol is going to be made in 2026, then it should use the evidence available in 2026.

Positive evidence. Negative evidence. Evidence that complicates the story.

Anything less isn’t following the evidence.

It’s stopping where you liked what the evidence was saying.

If you found this article useful, please subscribe so you don’t miss the next one.

Subscribe now

And if you know someone who would appreciate a careful look at the evidence instead of another round of slogans, share it with them.

Share

The more people who read the sources for themselves, the better.

Selected research and resources

Zelenko Labs. COVID-19, Viral Replication, and the Zelenko Protocol. March 2026. The paper identifies itself as a Zelenko Labs scientific whitepaper, and its bibliography contains six references ranging from 2010 through 2020.
te Velthuis AJW, van den Worm SHE, Sims AC, Baric RS, Snijder EJ, van Hemert MJ. “Zn²⁺ Inhibits Coronavirus and Arterivirus RNA Polymerase Activity In Vitro and Zinc Ionophores Block the Replication of These Viruses in Cell Culture.” PLOS Pathogens. 2010;6(11):e1001176. DOI: 10.1371/journal.ppat.1001176.

Derwand R, Scholz M, Zelenko V. “COVID-19 outpatients: early risk-stratified treatment with zinc plus low-dose hydroxychloroquine and azithromycin: a retrospective case series study.” International Journal of Antimicrobial Agents. 2020;56(6):106214. DOI: 10.1016/j.ijantimicag.2020.106214.

Skipper CP, Pastick KA, Engen NW, et al. “Hydroxychloroquine in Nonhospitalized Adults With Early COVID-19.” Annals of Internal Medicine. 2020.

Boulware DR, Pullen MF, Bangdiwala AS, et al. “A Randomized Trial of Hydroxychloroquine as Postexposure Prophylaxis for Covid-19.” New England Journal of Medicine. 2020.

Cavalcanti AB, Zampieri FG, Rosa RG, et al. “Hydroxychloroquine with or without Azithromycin in Mild-to-Moderate Covid-19.” New England Journal of Medicine. 2020.

Kavanagh O, Marie Healy A, Dayton F, Robinson S, O’Reilly NJ. “Hydroxychloroquine Does Not Function as a Direct Zinc Ionophore.” Pharmaceutics. 2022.

Dabbagh-Bazarbachi H, Clergeaud G, Quesada IM, et al. “Zinc ionophore activity of quercetin and epigallocatechin-gallate: from Hepa 1-6 cells to a liposome model.” Journal of Agricultural and Food Chemistry. 2014;62(32):8085–8093. DOI: 10.1021/jf5014633.

World Health Organization. Hydroxychloroquine evidence and recommendations for COVID-19 treatment and prevention, updated April 2025.

World Health Organization/BMJ living guideline. Therapeutics and COVID-19, including updated recommendations concerning ivermectin and corticosteroids.

PRINCIPLE Collaborative Group. Randomized evaluation of inhaled budesonide for higher-risk community patients with COVID-19. The Lancet. 2021.

Revista Médica Hondureña. Review of the Honduran MAÍZ COVID-19 regimen and the need for clinical evidence.

Zelenko Labs. How The Zelenko Protocol Was Born. Primary source for the account of how Zelenko developed and interpreted his early treatment experience.

Zelenko Labs. When Dr. Zelenko Saved Honduras. Primary source for claims concerning Zelenko’s role in Honduras.

Zelenko Labs. Five Times Dr Zelenko Was Right. Primary source for the current combination of Zelenko legacy claims, health claims and branded supplement promotion.

Ask a dog to choose between two cups, and he will study your face, follow your hand, and make a decision. Ask a cat the same question and she may solve the problem, inspect the researcher, wash one paw and leave.

The obvious conclusion is that the dog is smarter.

The less obvious conclusion is that we may have designed an intelligence test that rewards behaving like a dog.

This is one of the problems scientists face when they study animal minds. Dogs and cats cannot explain how they reached an answer. They cannot tell a researcher, “I followed your finger because I understood that you were directing my attention,” or, “I understood perfectly, but your snack was inadequate.” Researchers have to infer what the animal knows from what it does.

That makes the original question, Can dogs and cats think critically?, harder than it sounds.

First, what do we mean by critical thinking?

When we use the term for people, we usually mean more than solving a puzzle. Critical thinking includes weighing evidence, comparing explanations, noticing contradictions, reconsidering assumptions, and recognizing when our information is incomplete.

There is no good evidence that dogs or cats can examine an argument, question the reliability of a source, or consciously audit their own beliefs. Your cat may look judgmental, but that is not the same as peer review.

What pets do show are some of the smaller abilities from which more complex thinking is built: memory, inference, expectation, flexible learning, impulse control, information seeking, and sensitivity to cause and effect. Scientists study those pieces one at a time.

And once they do, the familiar household stereotypes become considerably less reliable.

The dog is reading the room

Dogs are unusually attentive to human beings. They follow our gaze and gestures, distinguish some of our emotional expressions, and notice where our attention is directed. Point toward the cup containing food and many dogs understand that the gesture refers to something beyond the finger itself. Anyone who has watched a person point at an object while a cat stares directly at the finger will appreciate that this is not a trivial accomplishment.

Some of that sensitivity appears remarkably early. In a 2021 study of 375 eight-week-old puppies, the puppies successfully used human gestures despite having had limited experience with people. Within that group, genetic factors accounted for more than 40 percent of the variation in point-following and attention to human faces.

Dogs did not merely learn to tolerate us. Over many generations, they became specialists in us.

That specialization can look a lot like reasoning because sometimes it is. In one experiment, dogs watched a reward being placed behind one of two barriers. On other trials, they were prevented from seeing where it went. When they had not seen the hiding place, the dogs were more likely to seek additional visual information before choosing. The researchers cautiously described this as evidence consistent with a basic form of metacognition: the dogs behaved differently when they lacked information.

That does not prove a dog sits there thinking, I should reserve judgment until I have reviewed the evidence. Simpler learned strategies may produce some of the same behavior. Still, the dog recognized uncertainty well enough to investigate before committing. Plenty of internet arguments would improve immediately if humans did the same.

Some dogs also display startling powers of inference. Rico, a border collie studied in 2004, could retrieve more than 200 objects by name. When researchers placed a new object among familiar ones and used an unfamiliar word, Rico often selected the new object, apparently reasoning by exclusion: I know the names of those seven, so the new sound probably means this one. He could still retrieve some of the newly named objects four weeks later. It was an exceptional case, not a description of the average dog, but it showed that at least some dogs can do more than memorize commands. They can use what they already know to narrow down an answer. The Rico study appeared in Science.

Dogs are also so good at reading people that we sometimes mistake their skill for something else.

Consider the famous guilty look: lowered head, tucked tail, averted eyes, perhaps a slow retreat from the scene of a murdered sofa cushion. It certainly looks like remorse. In Alexandra Horowitz’s 2009 experiment, however, dogs showed more “guilty” behavior when their owners scolded them, including dogs that had obeyed and had not eaten the forbidden treat. The expression tracked the owner’s reaction better than the dog’s actual offense.

The dog may not have been confessing. It may have been reading the prosecutor.

The cat received your message

Cats have a public-relations problem. A dog’s intelligence is easy to see because dogs often demonstrate it in cooperation with us. A cat may possess the information and see no reason to provide a live demonstration.

In a 2013 study, cats heard recordings of three strangers followed by their owner, all calling the cat’s name while out of sight. The cats became less responsive as the unfamiliar voices continued, then showed a renewed reaction to the owner’s voice. Their responses were usually small movements of the ears or head, not an enthusiastic reunion.

The finding was not that cats come when called. Let’s keep this within the boundaries of credible science. It was that they could recognize who was calling.

Research published in 2019 found that cats could also distinguish their own names from other words, even when an unfamiliar person spoke. This does not show that a cat understands a name as a human understands personal identity. The name may simply predict food, affection, play, a carrier appearing from the closet or some other event worth monitoring. But the cat is not necessarily ignoring you because the sound failed to register. Sometimes the message has been received and the request has been denied.

Cats also appear to learn more about household conversations than we deliberately teach them. In a 2022 study, household cats heard the name of a familiar companion cat and then saw a photograph. They looked longer when the name and face did not match, suggesting that they had learned which name belonged to which cat through ordinary life, without formal training.

Then there is the matter of surveillance.

Researchers placed cats in a room with speakers in different locations. The owner’s voice first came from one place and then suddenly from another. The cats appeared more surprised when their owner seemed to have teleported across the room, suggesting that they had been mentally tracking the unseen owner’s location.

Your cat may appear to be asleep. Your cat may also know exactly where you are.

Feline reasoning is not limited to keeping tabs on the staff. In a 2016 experiment, researchers shook a container that either rattled or remained silent, then turned it over. Cats looked longer when events violated the expected relationship between sound and object, such as a silent container suddenly producing an object. The results suggested that cats could use sound to predict the presence of something they could not see.

No, this does not mean cats understand Newtonian physics. It does mean they notice when the physical world fails to behave as expected. For an animal that often hunts prey it cannot see, reasoning from faint sounds is a useful talent.

More recently, a small 2024 experiment found that cats quickly associated two invented words with two animated pictures. When the pairings were switched, they looked longer, as if the change violated what they had just learned. The study involved only 31 cats and does not give them human language, but it adds to the evidence that cats are quietly collecting associations while we assume they are thinking about curtains.

They may be thinking about the curtains too. Both things can be true.

Are we giving them the wrong exam?

Comparing dogs and cats is tempting, but cognition is not one race with one finish line. A species can be excellent at social communication and less impressive at physical problem-solving. Another can be an expert at spatial tracking but uninterested in following instructions from a stranger holding a clipboard.

In a 2005 comparison, both dogs and cats used human pointing to locate hidden food, with no significant difference between them. When the food was made inaccessible, however, dogs were more likely to look toward a person and solicit help. Cats kept working on the problem more independently.

What does that tell us? Perhaps dogs are better communicators. Perhaps cats are more persistent. Perhaps dogs treat humans as partners while cats treat us as furniture capable of opening cans. The behavior is real, but the label we attach to it requires care.

A 2023 study illustrates the problem neatly. Dogs outperformed cats at following a distant pointing gesture, but cats also made far fewer choices and became less willing to participate as the laboratory trials continued. When tested at home, their willingness did not decline in the same way. The researchers discussed stress, motivation, restraint and the unfamiliar setting as possible influences.

So a laboratory result may measure several things at once: comprehension, attention, confidence, hunger, comfort and willingness to humor a graduate student.

This does not mean every failed test hides a secret feline genius. It means that a score is not pure intelligence poured into a measuring cup. Reviews of animal-cognition methods warn that attention, motivation, personality, sensory abilities and test design can all influence performance. A fair test has to fit the animal well enough to measure the ability it claims to measure.

If intelligence were judged by following a pointed finger, dogs would do very well. If the test involved locating a nearly silent movement behind a wall, the cat might demand a recount. If it required identifying the contents of a trash can from across the yard, the dog would regain the lead.

Humans would be eliminated in the first round.

So, can they think critically?

Not in the full human sense, as far as the evidence shows. Dogs and cats do not appear to evaluate formal arguments, compare published sources or deliberately challenge an ideology. They do, however, remember, infer, predict, seek information, and revise their behavior when circumstances change. Those are genuine cognitive abilities, and some are ingredients of critical thinking.

The most accurate answer is that pets possess forms of practical reasoning suited to their lives.

A dog’s world has been shaped by cooperation with people. Reading our faces, hands, voices, and attention can be more useful than solving an abstract mechanical puzzle. A cat’s world rewards spatial awareness, patience, sound-based inference and independent action. Neither mind is a smaller copy of ours, and the two are not smaller copies of each other.

The interesting question, then, is not whether a pet can think like a person. It is whether we can design a question that allows the animal to think like itself.

Your dog is reading you. Your cat heard you the first time. And both may be revealing as much about the intelligence of the examiner as the intelligence of the pet.

The dog reads the room. The cat reads the room and decides whether the room deserves a response.

Selected research and sources

• Emily E. Bray et al., “Early-emerging and highly heritable sensitivity to

• human communication in dogs”, Current Biology, 2021.

• Julia Belger and Juliane Bräuer, “Metacognition in dogs: Do dogs know they could be wrong?”, Learning & Behavior, 2018.

• Juliane Kaminski, Josep Call and Julia Fischer, “Word learning in a domestic dog: Evidence for ‘fast mapping’”, Science, 2004.

• Alexandra Horowitz, “Disambiguating the ‘guilty look’”, Behavioural Processes, 2009.

• Ádám Miklósi et al., comparison of dogs’ and cats’ use of human visual signals, Journal of Comparative Psychology, 2005.

• Atsuko Saito and Kazutaka Shinozuka, “Vocal recognition of owners by domestic cats”, Animal Cognition, 2013.

• Atsuko Saito et al., “Domestic cats discriminate their names from other words”, Scientific Reports, 2019.

• Saho Takagi et al., “There’s no ball without noise: Cats’ prediction of an object from noise”, Animal Cognition, 2016.

• Saho Takagi et al., “Socio-spatial cognition in cats: Mentally mapping owner’s location from voice”, PLOS ONE, 2021.

• Saho Takagi et al., “Cats learn the names of their friend cats in their daily lives”, Scientific Reports, 2022.

• Saho Takagi et al., “Rapid formation of picture-word association in cats”, Scientific Reports, 2024.

• Attila Salamon et al., direct comparison of dogs and cats using human pointing gestures, Scientific Reports, 2023.

• Michèle N. Schubiger, Claudia Fichtel and Judith M. Burkart, review of validity and confounding factors in animal-cognition tests, Frontiers in Psychology, 2020.

There is a reason this chart keeps circulating. It makes its argument before you have read a single study.

Row after row ends with a bright red NO. Beside those red boxes are safety review periods such as 4 days, 5 days, 30 days, and 42 days.

Then comes the claim:

“Not a SINGLE routine injected childhood vaccine was licensed based on a placebo controlled trial.”

Another version adds that most ordinary drugs are licensed after multi-year placebo-controlled trials, while most childhood vaccines receive only days or weeks of safety review.

Put those statements together, and the conclusion seems obvious:

But that is not actually what the chart proves. And this is where the story gets interesting.

The ICAN chart is not simply fabricated. Many of its numbers come from real FDA package inserts and clinical reviews. Some of the limitations it points to are real. Some prelicensure monitoring periods were short. Some currently marketed vaccines were compared with other vaccines rather than an inert saline injection.

That is exactly why this deserves something better than a quick TRUE or FALSE stamp.

The question is not whether the chart contains facts.

It does.

The question is whether those facts support the conclusion people are being encouraged to draw from them.

This article is not arguing that every vaccine trial was ideally designed, that every follow-up period was long enough, or that an inert placebo is never preferable. We are asking a narrower question: Does the evidence in this chart justify the much broader claim that childhood vaccines were never properly safety tested?

Start With the Wording

Read the headline carefully:

“None of the vaccine doses the CDC recommends for routine injection into children were licensed by the FDA based on a long-term placebo-controlled trial.”

There are several important qualifiers packed into that sentence:

• routine

• injection

• licensed based on

• long-term

• placebo-controlled

Notice that “long-term” is not a self-defining scientific threshold. Does it mean six months? One year? Several years? Before accepting an absolute claim built around that phrase, we should know what duration is being required and why.

Every qualifier narrows the claim.

That matters.

Consider these three statements:

• This exact currently marketed vaccine formulation was not licensed based on a long-term trial against an inert saline control.

• This vaccine was never tested against a placebo.

• Therefore, we do not know whether this vaccine is safe.

They sound related.

They are not equivalent.

The first statement can be true without establishing either of the other two.

What Counts as a Placebo?

A lot of this argument revolves around one word.

Placebo.

In online vaccine debates, placebo is frequently treated as another word for saline.

Saline can certainly be used as an inert placebo. But clinical trials are not required to use the same type of control in every situation.

One distinction is important: another vaccine is an active comparator, not a placebo. The point is not that every control should be called a placebo. It is that a randomized controlled trial can use different kinds of control groups, and each design answers a somewhat different question.

FDA regulations recognize several kinds of controlled drug trials, including placebo controls, dose-comparison controls, no-treatment controls, active-treatment controls, and historical controls. The FDA specifically recognizes that different controls may be appropriate in different clinical settings. (U.S. Food and Drug Administration)

FDA’s own description of vaccine development similarly explains that vaccine study control groups may receive an approved vaccine, a placebo or another substance, depending on the trial. Phase 3 trials generally involve thousands of participants and gather both effectiveness and additional safety information. (U.S. Food and Drug Administration)

Why would researchers use an active control instead of saline?

Sometimes the question is whether a new vaccine performs at least as well as an existing vaccine. Sometimes withholding an established preventive intervention would expose participants to avoidable risk.

But there is a limitation here that shouldn’t be brushed aside.

If you want to determine the absolute difference in adverse events between receiving a vaccine and receiving an inert injection, an active comparator cannot answer that question by itself.

That is a legitimate criticism.

What does not follow is that an active-control study provides no useful safety information.

If Vaccine A is compared with Vaccine B, researchers can still determine whether one group experiences substantially more of the measured outcomes than the other.

That is comparative safety information. What it does not establish by itself is the absolute adverse-event rate compared with receiving no vaccine or an inert injection.

Even the word “placebo” does not always mean a solution containing nothing but saline literally. In blinded trials, a placebo may contain inactive formulation ingredients so that it resembles the study product, while an active comparator is a different category altogether because it contains another active intervention.

The better question is not simply:

Was there saline in the other syringe?

It is:

What question was the trial designed to answer, and was the comparator capable of answering it?

We Checked the Red Boxes

Once you leave the graphic and open the underlying records, the picture gets more complicated very quickly.

Hepatitis B: The Short Follow-Up Criticism Is Real

This is one area where critics have a legitimate point.

The Recombivax HB prescribing information says that three clinical studies administered 434 doses to 147 healthy infants and children up to age 10. They were monitored for five days after each dose for the adverse reactions described in that section. (DailyMed)

The Engerix-B label says that 13,495 doses were administered to 5,071 healthy adults and children, including healthy neonates, across 36 studies. Participants were monitored for four days after administration. (DailyMed)

Those really are short monitoring windows. There is no need to pretend otherwise.

Could five days of observation in 147 infants and children, by itself, exclude an adverse event that appears several weeks later?

No.

Could it reliably detect an adverse event occurring once in 50,000 or 100,000 recipients?

Of course not. The sample isn’t remotely large enough. That is a fair limitation of that particular prelicensure safety dataset.

But another mistake occurs when that observation turns into:

“Hepatitis B vaccines were never tested against placebo.”

They were.

A 1984 randomized, double-blind, placebo-controlled trial studied hepatitis B vaccination in high-risk newborns born to mothers who were HBeAg-positive hepatitis B carriers.

Infants were assigned to several vaccine and hepatitis B immunoglobulin regimens or placebo. The first injections were administered within an hour of birth, and comparisons were reported among infants followed during a trial lasting 495 days. (ScienceDirect)

There is an important qualification.

That was an earlier plasma-derived hepatitis B vaccine, not the exact recombinant Recombivax HB or Engerix-B formulations used today.

We shouldn’t erase that distinction. But we shouldn’t erase the trial either. So the accurate version is considerably less dramatic:

Both statements can be true. That is often where critical thinking gets interesting.

“Seven Days of Safety Testing” Can Be Technically True and Deeply Misleading

One of the most persuasive columns in the ICAN chart is labeled Safety Review After Injection.

You see:

4 days

5 days

30 days

42 days

A reasonable reader can come away believing researchers simply stopped looking after that number of days. Sometimes the cited monitoring period really is short, as we just saw with the hepatitis B labels. Other times it represents only one layer of the safety monitoring.

Daptacel

For Daptacel, parents recorded solicited local and systemic adverse events during days 0 through 7.

Stop reading there, and you might reasonably say:

Seven days of safety testing.

Keep reading.

The same prescribing information says investigators made periodic telephone calls to inquire about adverse events and that serious adverse events were monitored through six months after the last dose in the three U.S. studies. (DailyMed)

Seven days was a monitoring period.

It was not the entire safety follow-up.

Prevnar 13

Prevnar 13 makes the problem even clearer.

Parents recorded specified local and systemic reactions for seven days after each vaccination.

Unsolicited adverse events were monitored over longer periods.

Information concerning serious adverse events, newly diagnosed chronic medical conditions, and hospitalizations was collected during later visits and again through a scripted telephone interview six months after the fourth dose. Serious adverse events were collected throughout the study period. (DailyMed)

Reducing all of that to a box that says 7 days changes the impression of what happened.

The number is not necessarily false.

The interpretation is incomplete.

Gardasil 9

FDA’s current Gardasil 9 prescribing information makes the distinction even clearer. Unsolicited systemic adverse reactions were collected during the short post-vaccination window, but serious adverse events were collected throughout the entire study period across seven clinical studies, with follow-up ranging from one month to 48 months after the last dose.

That does not mean every possible adverse event was actively solicited for four years. It does mean that describing Gardasil 9 as having only a few days or weeks of “safety review” leaves out a substantial part of the actual follow-up.

Ask “Days of What?”

Someone tells you:

“This vaccine was safety tested for only seven days.”

Before arguing, ask:

Seven days of what?

Was that seven days of recording fever and injection-site pain?

• Seven days for all unsolicited medical events?

• Seven days for hospitalizations?

• Seven days for serious adverse events?

• Seven days of total participant follow-up?

Until you know what the clock was measuring, the number alone doesn’t tell you very much.

Then We Got to Varivax

This is where our audit uncovered something much more interesting.

Look at the Varivax row in the ICAN chart.

The control is described as containing approximately 45 milligrams of neomycin per milliliter.

That number wasn’t invented by ICAN.

It came from the scientific literature. And it sat there for more than forty years.

There was just one problem.

It was wrong.

In December 2025, the New England Journal of Medicine formally corrected its 1984 varicella vaccine study. The original paper said the vaccine contained 45 mg of neomycin per milliliter.

The corrected amount is:

no more than 0.64 micrograms per milliliter.

The journal also corrected the description of the placebo. Rather than containing approximately 45 mg/mL, it consisted of lyophilized stabilizer containing approximately the same trace concentrations of neomycin as the vaccine. (New England Journal of Medicine)

That is not a rounding error.

Forty-five milligrams is roughly 70,000 times 0.64 micrograms.

There is an important fairness point here.

The ICAN chart shown above is dated October 2023.

The NEJM correction was not published until December 2025.

ICAN could not have cited a correction that didn’t exist.

But the graphic continues to circulate today.

And that row is now outdated.

The underlying study is even more interesting.

The 1984 trial was double-blind and placebo-controlled. It enrolled 956 children between ages 1 and 14. Among the 914 confirmed to be susceptible to chickenpox, 468 received vaccine and 446 received placebo.

During nine months of surveillance, 39 clinically diagnosed chickenpox cases occurred.

All 39 were in the placebo group. (New England Journal of Medicine)

The current FDA Varivax insert explicitly says that its protective efficacy was established in part by a placebo-controlled, double-blind clinical trial. (U.S. Food and Drug Administration)

That trial was primarily designed to establish efficacy, not to rule out very rare or long-latency adverse events. Its relevance here is narrower: it shows that a placebo-controlled trial of the varicella vaccine did, in fact, exist.

But here comes another important qualification.

The FDA also says that the placebo trial used an earlier formulation containing 17,000 PFU per dose, rather than the exact current vaccine formulation. (U.S. Food and Drug Administration)

Now compare these two statements:

“The exact current Varivax formulation was not tested in that placebo-controlled trial.”

True.

“There was no placebo-controlled trial of the chickenpox vaccine.”

False.

Same evidence. Very different claim.

Following a Citation Is Not the End of Fact-Checking

A citation can be real and still be outdated. A respected journal can publish an error and correct it decades later. A chart can accurately quote a source and still become wrong when that source changes.

The presence of a citation proves that somebody cited something.

It does not prove the conclusion.

Sometimes #FollowTheEvidence means following the citation one step farther than the person who posted it.

Rotarix Disappears Because of One Word

Consider Rotarix.

The FDA licensing review describes a Phase 3 study that was randomized, double-blind, placebo-controlled, multinational, and involved roughly 60,000 infants.

One of its primary safety objectives specifically evaluated intussusception, a rare but serious bowel complication, during the defined post-vaccination risk period. (U.S. Food and Drug Administration)

That is a very large randomized, placebo-controlled vaccine trial in infants.

But there is an important limitation to what this example proves. It does not settle Siri’s narrower argument about whether a control was completely inert or saline. Its relevance here is simpler: a large randomized, double-blind, placebo-controlled childhood vaccine trial clearly existed.

So why doesn’t it count in the viral claim?

Because you swallow it.

Rotarix is an oral vaccine. It isn’t injected.

That one word, injected, removes a roughly 60,000-infant randomized placebo-controlled trial from the discussion.

This is why qualifiers matter.

The broad social-media claim sounds like:

“Childhood vaccines were never placebo tested.”

The defensible version becomes something closer to:

“The exact currently marketed standalone routine injected childhood vaccines were not licensed based on the particular kind of long-term inert-control trial we are defining as acceptable.”

That is not nearly as memorable.

But it is much closer to the actual argument.

Fluzone Creates Another Problem for the Sweeping Claim

The current Fluzone prescribing information contains another inconvenient example.

A study in children ages 6 through 24 months was randomized, double-blind and placebo-controlled.

There were 786 children in the intent-to-treat population.

525 received Fluzone.

261 received placebo. (DailyMed)

And here the control really was saline.

The label identifies the placebo as 0.4% NaCl. It also reports 12 months of follow-up in the first influenza season and six months in the second. (DailyMed)

That does not mean every annual flu formulation has subsequently undergone a brand-new saline-controlled trial. Influenza strains change, and updated formulations introduce a separate regulatory question.

That does not refute Siri’s narrowly worded claim about what a particular formulation was “licensed based on.” It does refute the much broader claim people routinely take from it: that injected childhood vaccines simply have not been tested against saline controls.

What About “Most Drugs Get Multi-Year Placebo Trials”?

This may be the least examined part of the viral claim.

The post says:

“Most drugs are licensed based on multi year placebo controlled trials.”

That creates a powerful contrast.

Ordinary drugs supposedly face years of placebo testing.

Vaccines supposedly get four days.

Except there is no general FDA rule requiring ordinary drugs to undergo multi-year placebo-controlled trials before approval.

FDA recognizes multiple legitimate study designs, including active-treatment controls. (U.S. Food and Drug Administration)

And it takes very little digging through FDA records to find ordinary drug trials measured in weeks, not years.

Latuda: Its pediatric bipolar depression indication was established in a six-week randomized, double-blind, placebo-controlled study. (FDA Access Data)

Seysara: Its acne program included two 12-week randomized, double-blind, placebo-controlled trials involving about 2,000 participants. (FDA Access Data)

Pristiq: Two pediatric depression trials were randomized, double-blind and placebo-controlled for eight weeks. Longer-term risks were examined separately in six-month open-label extension studies. (FDA Access Data)

Those examples do not prove what “most drugs” receive, and we should not pretend they do. To establish that, we would need a defined sample of drug approvals and a clear rule for what counts as a “multi-year placebo-controlled trial.”

What they do establish is narrower: there is no universal FDA requirement that ordinary drugs undergo multi-year placebo-controlled trials before approval.

So when somebody says:

“Most drugs...”

stop.

Ask for the denominator.

Most of what?

All drugs?

Chronic medications?

Antibiotics?

Cancer treatments?

Psychiatric medications?