Secretary of Human Health Services Robert F. Kennedy, Jr. Calls Out Statistical Shamwizardry Published by the Annals of Internal Medicine

The Journal risks now being discredited as an authentic academic journal. Kennedy published this article in Trial Site News.

For decades, vaccine safety research funded or aligned with government agencies (CDC, NIH, EMA) has operated with unquestioned legitimacy in press coverage, despite repeated findings of:

• Causal inference violations (e.g. healthy user bias, differential misclassification)

• Endpoint bundling and label substitutions (e.g., SIDS vs. respiratory failure)

• Unverified exposure estimates (e.g., schedule-assigned aluminum “dose”)

• Systemic over-adjustment for time and access variables, muting real signal.

Kennedy’s critique breaks the performative firewall between public health administration and critique of its own scientific claims. He does not hedge. He does not defer to “peer review.” He calls out the Annals study’s core flaws, including:

• Methodological laundering of exposure;

• Spurious null finding via flawed adjustments;

Suppression of signal through design.

For years, I and others have noted the problem of studies of vaccines as exercises in torturing data to get the results desired by those seeking to defend vaccines. I first pointed out fatal flaws in the particular study Kennedy takes on here on Popular Rationalism (see RELATED, below). The media picked the study up and ran with it as exonerating aluminum as a contributed to chronic illness.

Kennedy goes a step further, and report a significant association between exposure to aluminum-containing vaccines and Asperger’s syndrome, using the author’s own data:

“Compared to the moderate exposure group, for every 10,000 children in the highest aluminum exposure cohort, there were 9.7 more cases of neurodevelopmental disorder, 4.5 more cases of autistic disorder, and 8.7 more cases of the broader category of autism spectrum disorder.”

We have turned a major corner in biomedical research.

Secretary Kennedy’s article can be found in its original state at Trial Site News. Here it is for your convenience:

Flawed Science, Bought Conclusions: The Aluminum Vaccine Study the Media Won’t Question

By Robert F. Kennedy Jr.

On July 15, an intensely ballyhooed study by Andersson et al., published in Annals of Internal Medicine—a journal of the American College of Physicians, claims to find no association between aluminum-adjuvanted vaccines and chronic childhood disorders in Denmark. The slavish, pharma-funded mainstream media, ever eager to defend industry orthodoxies, have triumphantly hailed this study as proof of aluminum’s safety without even a cursory examination of the study’s fatal deficiencies or the financial conflicts of its authors. But a closer look reveals a study so deeply flawed it functions not as science but as a deceitful propaganda stunt by the pharmaceutical industry.

The architects of this study meticulously designed it not to find harm. From the outset, Andersson et al. excluded the very children most likely to reveal injuries associated with high exposures to aluminum adjuvants in childhood vaccines. The exclusion included all children who died before age two, those diagnosed early with respiratory conditions, and an astonishing 34,547 children — 2.8% of the study population — whose vaccination records showed the highest aluminum exposure levels.

These choices suggest an intention to exclude the children at highest risk of harm. The authors, without explanation, deemed these high exposures “implausible,” even though those implausibly high exposures are routine for American children who follow the recommended immunization schedule. At very least, the study findings therefore cannot be generalized to children in the U.S. By systematically stripping the dataset of high-risk individuals, the researchers leave behind a survivor cohort to analyze. The name for this logical fallacy is “healthy subject bias.”

Furthermore, the authors inappropriately treated general practitioner visits before age two as a confounder, without assessing whether these GP visits reflected early aluminum-related illness or were predictive of later diagnoses. This introduced “collider bias” — a distortion that can suppress real associations even to the extent of making aluminum appear protective. It’s like studying whether smoking causes lung cancer while adjusting for coughing or for yellowed fingers — symptoms associated with smoking.

These sleights of hand magnify the potential for allowing the authors to reach their absurd suggestion that higher aluminum exposure is somehow protective against asthma, allergies, and neurodevelopmental disorders, including autism. These findings clash with mountains of contrary literature documenting the neurotoxicity of aluminum and its association with autoimmune and allergic diseases. (Daley, et al. 2023) If the medical establishment truly believed these data, they would be recommending aluminum injections to children as a prophylaxis against neurological and autoimmune diseases.

Andersson and his team initially had a zero-exposure cohort within the study group. But instead of evaluating this non-vaccination group separately and treating these children as the control, they lumped these kids into the least-exposed cohort, diluting any signal of harm. More broadly, their analysis assumed a linear dose-response, ignoring evidence from Crépeaux et al. (2017) demonstrating that low doses of aluminum can produce non-linear neurotoxic effects in animal models.

While adjusting for birth year can be appropriate in some study designs to account for secular trends, it is not a neutral act in this case. Aluminum exposure from vaccines increased over time, as did rates of chronic childhood disorders. Yet the authors failed to explore whether this correlation might reflect a causal relationship. They did not test this hypothesis but conveniently avoided doing so through this statistical artifice.

Furthermore, the authors almost exclusively relied on diagnoses from hospital inpatient registers. This gimmick allowed the authors to exclude the vast majority of affected children whose autism and food allergies would most likely be diagnosed and managed outside of hospital settings. This potentially biases the data toward undercounting true cases, particularly among children with milder presentations or less frequent hospital contact.

There were additional problems with this data source. A 2017 analysis by Holt et al. identified substantial misclassification in the Danish National Health Service Registry — the same source used by Andersson et al. to assess vaccine exposure — finding that children’s medical records often documented vaccinations that were absent from the registry. This casts further doubt on the accuracy of exposure classification in the study. In other words, it’s highly likely that many of the children that the authors classified as not receiving aluminum-containing vaccines actually did.

The CDC’s routine childhood vaccine schedule is also considerably more aggressive than Denmark’s. For example, while Denmark recommends the aluminum-containing hepatitis B (HepB) vaccine only for infants deemed at risk, the CDC recommends a three-dose series for all newborns starting on the first day of life.

Andersson et al. also neglects to consider susceptible subpopulations of children based on known genetic risk factors, such as mitochondrial dysfunction, or environmental co-exposures, such as the synergistic toxicity of mercury and aluminum. This suite of deceptive devices and strategic exclusions seems calculated as an additional flimflam for deliberately excluding or diluting out the most vulnerable children.

Despite all the deceptive devices the authors used to conceal the signals of harm, Andersson et al.’s own supplementary data are a devastating indictment of aluminum-containing vaccines. These data, which they were forced to publish because of public criticism of their analysis, directly contradict the study’s conclusions. The data show a statistically significant 67% increased risk of Asperger’s syndrome per 1 mg increase in aluminum exposure among children born between 2007 and 2018. Compared to the moderate exposure group, for every 10,000 children in the highest aluminum exposure cohort, there were 9.7 more cases of neurodevelopmental disorder, 4.5 more cases of autistic disorder, and 8.7 more cases of the broader category of autism spectrum disorder. Yet the authors gloss over these harms to children by claiming they “did not find evidence” for an increased risk.

To reassure readers that infants’ vaccine aluminum exposure is “well below” an established “minimal risk level,” Andersson et al. cite Mitkus et al. (2011). But that FDA analysis was based on ingested soluble aluminum in adult rodents, making its findings irrelevant to injected particulate aluminum in human infants. To cite this study as proof of safety is scientifically indefensible.

Three of the study’s authors are affiliated with Denmark’s Statens Serum Institut (SSI), a government-owned vaccine company that develops a number of aluminum-containing vaccines. SSI also procures and supplies vaccines for the Danish national vaccination program — a clear institutional conflict given its role in supporting vaccine manufacturing and promoting vaccine uptake. Yet two of the three authors nevertheless declared no conflicts of interest. The senior author, Anders Hviid, reported funding from sources including the Novo Nordisk Foundation, which is directly linked to the pharmaceutical giant Novo Nordisk and maintains a substantial investment footprint in the industry. Such affiliations call into question the study’s independence and underscore the need for raising international standards of gold standard science outlined in the Trump administration’s recent executive order requiring transparency, reproducibility, and data sharing. These are standards that will determine with whom the United States will collaborate or do business going forward.

The study, in fact, offers the opposite of transparency. For example, there is no table showing how many of the children in each aluminum exposure cohort were diagnosed with each outcome. Consequently, there is no way to independently evaluate the calculations made, leaving readers to place faith in the authors’ opaque modeling decisions at the expense of scientific reproducibility.

Finally, the Annals of Internal Medicine failed to share raw data that is essential for transparency and reproducibility. The disclosures on the Andersson study say “owing to data privacy regulations in Denmark, the raw data cannot be shared.”

Public health policy should not rest on non-reproducible observational studies that are not merely inconclusive but appear to be intentionally designed to find no association between vaccines and health harms. If the authors are truly committed to science, they should ask the Danish government to waive the requirements of the law and allow full access to their raw data for scientists across the world to verify their findings.

This study does not just suffer from mere methodological limitations. Its design flaws are defining. The only thing this study proves is the thorough corruption of the scientific journals that publish such garbage-in, garbage-out exercises in statistical manipulation.

The Annals of Internal Medicine should immediately retract this badly flawed study.

For years, American parents have been calling for rigorous, transparent, and independently conducted science comparing the long-term health outcomes of children vaccinated according to the CDC schedule with those of completely unvaccinated children. Yet studies like Andersson et al. showed they had the data to make this comparison between vaccinated and unvaccinated children, but instead excluded and lumped data that made their insights opaque. These authors squandered an important opportunity to restore trust by animating an international scientific process to develop safer vaccines. By excluding unvaccinated children from meaningful analysis, obscuring raw data, and relying on hidden statistical assumptions, this study exemplifies the kind of institutional obfuscation that continues to erode public trust. What’s needed is not more statistical modeling designed to bury signals of harm, but independent research grounded in full transparency, methodological integrity, and the courage to confront inconvenient truths.

Robert F. Kennedy Jr. is the 26th Secretary of the U.S. Department of Health and Human Services.