Debunking Hooker et al. (2026)

Why the “Aluminium Adjuvants Cause Autism” Manuscript Fails Scientific Scrutiny

On 31 January 2026, a manuscript titled “Aluminum Adjuvants, Autoimmunity, and Autism Spectrum Disorder” was circulated online and promoted as evidence that childhood vaccines cause autism. The document has been widely shared on social media and framed as a “comprehensive scientific analysis”.

It is not.

This article explains, clearly and dispassionately, why the Hooker et al. manuscript fails to meet basic scientific, epidemiological, toxicological, and evidentiary standards, and why it should not be relied upon by parents, clinicians, courts, or policymakers. It has 7 fatal flaws.

Fatal Flaw 1 - This Is Not a Peer-Reviewed Scientific Paper

The manuscript was not published in a recognised medical or scientific journal. It was released by Children’s Health Defense, an American political advocacy organisation with an explicit anti-vaccination agenda.

Reputable journals require:

• independent peer review

• conflict-of-interest disclosures

• funding declarations

• adherence to strict formatting and methodological standards

None of these safeguards are present.

The absence of a funding or conflict-of-interest statement alone would disqualify the manuscript from serious consideration in any reputable journal.

Fatal Flaw 2 - Authorship Does Not Substitute for Evidence

Science is not decided by credentials, but credibility does require relevant expertise and transparency.

The manuscript’s senior authors do not collectively demonstrate:

• formal training (or even experience) in vaccine pharmacology or regulatory toxicology, or

• clinical immunology or paediatric neurology experience, or

• authorship of high-quality, recent, peer-reviewed vaccine safety research

More importantly, the paper does not present original data. It is a narrative review that selectively assembles existing literature to support a predetermined conclusion.

That makes methodological rigor, not author reputation, the decisive issue. On that basis, the paper fails.

Fatal Flaw 3 - The Core Epidemiological Claim Is Invalid

The manuscript repeatedly asserts an “80-fold increase” in autism prevalence since the 1960s and treats this as evidence of environmental causation linked to vaccines.

This argument is scientifically indefensible.

• Autism did not have modern diagnostic criteria in the 1960s

• ASD as a unified diagnosis did not exist until DSM-IV (1994), where it was released as a diagnosis of "Asperger's".

• DSM-5 (2013) substantially broadened the diagnostic umbrella

Comparing modern ASD prevalence figures to an era in which the diagnosis effectively did not exist is not epidemiology.

Crucially, the authors rely on ecological correlation (vaccine schedule expansion vs ASD prevalence) while dismissing or ignoring individual-level cohort studies involving millions of children that directly test vaccination exposure and autism risk, and consistently find no association.

Accepting weak correlations while rejecting stronger contradictory evidence is methodologically unsound.

Fatal Flaw 4 - The Bradford Hill Criteria Are Misused

The manuscript claims that all nine Bradford Hill criteria for causation are “satisfied”.

They are not.

The Bradford Hill criteria are interpretive guidelines, not a checklist. They were never intended to convert biological plausibility into proof of causation.

The paper’s central failure is this:

It does not demonstrate that vaccine-relevant aluminium exposure produces neurotoxic effects in vivo in humans at real-world doses.

Biological plausibility without dose-appropriate human evidence is speculation.

Invoking “genetic susceptibility” does not strengthen causation; it weakens specificity and renders the hypothesis unfalsifiable.

Fatal Flaw 5 - Toxicology and Pharmacokinetics Are Treated Superficially

The manuscript correctly notes that aluminium adjuvants persist longer than soluble aluminium. This is not controversial.

What it fails to show is everything that matters in toxicology:

• persistence ≠ toxicity

• presence ≠ pathological concentration

• mechanism ≠ demonstrated harm

There is no demonstrated dose–response relationship showing that aluminium from vaccines reaches neurotoxic concentrations in human brain tissue.

The paper cites cumulative aluminium exposure figures (~4.9 mg by 18 months) without:

• comparing dietary aluminium exposure

• accounting for gastrointestinal absorption differences

• modelling renal clearance in healthy infants

• demonstrating CNS bioavailability at toxic thresholds

Toxicology without quantitative dose modelling is not evidence. It is narrative.

Fatal Flaw 6 - Neuropathology Is Misattributed

The manuscript cites real findings regarding neuroinflammation, microglial activation, and immune markers in some individuals with autism.

What it does not show is that vaccines (or aluminium adjuvants) cause those findings.

Neuroinflammation in ASD is non-specific. Similar patterns are observed in:

• maternal immune activation

• prematurity

• genetic synaptopathies

• epilepsy syndromes

• perinatal hypoxia

The manuscript repeatedly re-labels non-specific pathology as vaccine-induced without direct evidence. This is post hoc attribution, not inference.

Fatal Flaw 7 - Legal Framing Undermines Scientific Credibility

The paper repeatedly blends scientific discussion with arguments about liability shields, tort law, and regulatory immunity. This manuscript is not a law journal, a bench book, or a legal commentary, and it does not purport to be one.

In both Australia and comparable common-law jurisdictions, authoritative legal analysis is ordinarily undertaken by legal academics, judges, or senior practitioners (for example, a senior barrister such as a KC/SC depending on jurisdiction). None of the authors identify themselves as legal practitioners or legal scholars, and no legal peer review is disclosed.

Scientific causation and legal causation are distinct concepts governed by different standards. Scientific evidence seeks to establish biological mechanisms and empirical associations, whereas legal causation is a normative construct shaped by policy considerations, evidentiary thresholds, and statutory frameworks. The two are not interchangeable.

Borrowing the language and framing of toxic tort litigation does not strengthen biomedical evidence. On the contrary, it signals an advocacy-driven approach that substitutes legal rhetoric for scientific proof.

This conflation of scientific analysis with legal strategy is characteristic of advocacy documents, not of rigorous scientific research. It is also worth noting that whilst the Children's Health Defense files a number of lawsuits relating to vaccines (primarily in the US), they are yet to win any lawsuits (all finalised lawsuits initiated by them have been dismissed, often with prejudice). In Australia this would put them into the category of 'vexatious litigant' so any legal analysis from this entity and its affiliates is unlikely to hold true when tested.

What the Paper Gets Right (And Why It Still Fails)

It is reasonable to argue that:

• aluminium adjuvants were approved under older regulatory frameworks

• immune-mediated neurodevelopmental vulnerability warrants further research

• regulatory processes should evolve with new science

  
  

None of these points demonstrate that childhood vaccines cause autism.

Calls for further research are not evidence of harm.

Conclusion

The Hooker et al. manuscript does not establish that aluminium adjuvants cause autism.

It relies on:

• invalid epidemiological comparisons

• selective evidence use

• speculative toxicology

• misapplication of causation criteria

• and advocacy-driven framing

It would not survive peer review in a mainstream medical, epidemiological, or toxicological journal, and it should not be relied upon for medical decision-making.

Parents deserve evidence, not rhetoric.

The original manuscript:

Brian Hooker et al., Aluminum Adjuvants, Autoimmunity, and Autism Spectrum Disorder: A Comprehensive Mechanistic, Neuropathological, and Legal Analysis, Zenodo, 2026, https://doi.org/10.5281/zenodo.18442127 (accessed 5 February 2026).