How the Danish Study on Aluminum in Vaccines Was Designed to Find No Harm
A new Danish study claims aluminum-containing vaccines are safe, but like Hviid’s 2019 MMR-autism study, it was designed to find no health harms.
This article was originally published at JeremyRHammond.com, where I encourage you to sign up for my proper newsletters (plus get a free e-book) and support my work without Substack taking a 10% cut.
Table of Contents
Introduction
The news media are exploding with headlines about a new study out of Denmark purporting to prove that aluminum-containing vaccines do not cause long-term health harms including atopic and allergic diseases, autoimmunity, or neurodevelopmental disorders.
Titled “Aluminum-Adsorbed Vaccines and Chronic Diseases in Childhood: A Nationwide Cohort Study”, the study was written by Niklas Worm Andersson et al. and published in Annals of Internal Medicine on July 15.
The senior author—who is typically listed last in the byline in medical journals—is Anders Hviid, who was the lead author of the study “Measles, Mumps, Rubella Vaccination and Autism: A Nationwide Cohort Study”, published in the same journal in 2019, which the media headlines likewise proclaimed at the time had proven that the MMR vaccine is not associated with autism even in genetically susceptible children.
Dr. Brian Hooker, Dr. Jeet Varia, and I demonstrate, however, why that earlier study’s conclusion does not follow from its findings in our paper “Hviid et al. 2019 Vaccine-Autism Study: Much Ado About Nothing?”, published in the Journal of Biotechnology and Biomedicine on May 7.
As always, with current media reports, an observational study finding no association between vaccines and harms is touted as conclusive without any kind of critical analysis whatsoever.
This contrasts with media reporting on observational studies that do find associations between vaccines and harms, in which case the studies are invariably characterized as having methodological flaws that call the findings into serious question.
For instance, when researchers from the US Centers for Disease Control and Prevention (CDC) published a study in 2017 finding a nearly 8-fold greater risk of miscarriage among women who were vaccinated during pregnancy and who’d also gotten a flu shot the preceding year, the Washington Post curiously described it as finding “a hint of a possible link”. The Post claimed falsely that it was the first study to ever find a link between flu shots and miscarriage, and it dutifully reminded readers that this was just an observational study that cannot “provide a definitive answer” and had “several limitations” indicating that “the results could be biased”.
Similarly, when CDC researchers published a study in 2022 finding an association between aluminum-containing vaccines and asthma, we were told how “experts caution that more work is needed to confirm the findings” because the study had “important shortcomings”.
Somehow, though, studies finding no association never seem to have any methodological weaknesses so can be considered absolutely conclusive! Imagine that!
Here is a sampling of illustrative headlines telling us what we’re supposed to think about the new study out of Denmark:
“Study finds no link between aluminum in vaccines and autism, asthma” — NBC News
“Study Finds No Link Between Vaccine Aluminum and Health Problems” — U.S. News
“Study: No link between vaccine aluminum, health problems” — UPI
“New Study Finds No Links Between Aluminum in Vaccines and Chronic Conditions” — Truthout
“Aluminum in Vaccines Not Linked to Chronic Childhood Disorders: Danish Study” — Medscape
“Aluminum in Vaccines Not Culprit in Kids’ Chronic Diseases, Study Shows” — Medpage Today
“No Risks Seen With Early Childhood Exposure to Aluminum-Adsorbed Vaccines” — Physician’s Weekly
“Aluminum exposure from childhood vaccines not linked to increased risk of certain chronic disorders” — American College of Physicians
“Large study squashes anti-vaccine talking points about aluminum” — Ars Technica
Since the government and media constantly lie to us, though, about everything, let’s exercise independent thought and—at the risk of committing heresy against the vaccine religion—examine the study for ourselves.
A critical analysis of the new aluminum study reveals that it follows essentially the same playbook as Hviid et al. (2019): it is not merely that the study has methodological flaws resulting in questionable findings but that, like the 2019 MMR-autism study, it was rather designed to find no association.
A Vaxxed vs. Vaxxed Study
Andersson et al. analyzed over 1.2 million Danish children born between 1997 and 2018 and compared health outcomes between children with different levels of aluminum exposure from the routine vaccination schedule by 2 years of age.
Outcomes they considered include autoimmune diseases, atopic or allergic diseases, and neurodevelopmental disorders, including autism spectrum disorder (ASD) and attention deficit-hyperactivity disorder (ADHD).
While only about 1.2% of study subjects had received no aluminum-containing vaccines by age two, given a study population of over 1.2 million, that still amounted to a substantial population of over 15,000 children whose health outcomes could have been compared with those of children who received a typical exposure level from Denmark’s routine vaccination schedule.
But the authors did not do that vaccinated versus unvaccinated comparison. Instead, they restricted their analysis to comparing incremental differences in aluminum exposure among vaccinated children.
That is, it was a vaccinated versus vaccinated study.
The study is being reported as though it was designed to answer the question: “Does aluminum exposure from routine vaccination increase the risk of health harms compared to no exposure?”
Instead, the question it asked is: “Does increasing the amount of exposure from X mg to X+1 mg increase the risk?”
Assuming the Null Hypothesis
Over the period of observation, as noted by the authors, Denmark’s vaccination schedule expanded, increasing cumulative aluminum exposure in later birth cohorts.
As also noted by the authors, that increase in aluminum exposure correlated to rising rates of diagnoses of the health outcomes being measured.
The media report the study as though it was designed to test the hypothesis that this correlation represents a causal relationship.
In fact, though, the authors treated birth year as a covariate in their statistical model, “adjusting” the data based on the implicit assumption that rising diagnosis rates over time are unrelated to increasing aluminum exposure.
The design of their study thus incorporated a petitio principii fallacy: the authors begged the question by presuming the null hypothesis.
Excluding Children at Highest Risk
Following the playbook from Hviid et al. 2019, the authors of the new study similarly excluded children at higher risk of harm.
The study’s flowchart (Figure 1) shows a series of exclusions that systematically removed high-risk children from the analysis:
Children who died before age 2 were excluded, potentially masking increased infant mortality risk linked to aluminum exposure—despite studies into the “non-specific effects” of non-live, aluminum-containing vaccines finding them to be associated with an increased rate of childhood mortality.
Children diagnosed with respiratory conditions before age 2 were excluded, even though such conditions could have resulted from early aluminum exposure and indicated that these children were particularly vulnerable to its neurotoxic effects.
Children who had “too many vaccinations” were excluded without any clear definition of what constituted “implausible” vaccination counts. This raises concerns about data quality in the registry they relied on.
Children diagnosed with any study outcome before age 2 were excluded. This filters out the very children who may have been most susceptible to harm—those whose conditions manifested earlier.
These exclusions bias the cohort toward healthier survivors, ensuring that any real harms among the most vulnerable children are invisible to the analysis.
Healthy User Bias
In the discussion section of the study, the authors curiously state that “very early prediagnosis symptoms could potentially either reduce or increase the propensity to vaccination, which could bias toward underestimating or overestimating, respectively, the risk in vaccinated children.”
But why would parents who observe early symptoms be more likely to vaccinate?
It’s a peculiar statement since the authors know that studies indicate the opposite.
In fact, the MMR-autism study by Hviid et al. cited Jain et al. (2015), which found a selection bias known as “healthy user bias”.
While the 2015 study, too, was reported at the time as “debunking” a link between the MMR vaccine and autism, it was actually designed to determine whether children who have an older sibling with autism were less likely to receive the MMR shot.
And the key finding of the study is that they were.
It wasn’t that children who received the MMR vaccine were no more likely to develop autism but that children at higher risk of autism were less likely to get the MMR vaccine.
That is, children are higher risk are naturally pooled into the “unvaccinated” or “undervaccinated” cohorts in these types of observational studies.
Researchers aware of this selection bias can utilize it to design studies to find no association by simply failing to control for it.
As noted above, Andersson et al. not only failed to control for healthy user bias but practically ensured it.
After acknowledging the possibility of healthy user bias, they attempt to downplay the risk of it in their study by adding, “However, as mentioned above, our analyses are primarily based on the random variation in aluminum exposure through vaccination, in contrast to a vaccinated versus unvaccinated comparison.”
Of course, they also admit elsewhere that “our study lacked true randomization of exposure.”
But more importantly, the remark ignores how excluding children at higher risk of the outcomes biased the study in favor of finding no association.
The “random variation in aluminum exposure” over time also brings us right back to how the null hypothesis was assumed in their methodology by treating birth year as a “covariate” an “adjusting” the data accordingly.
The data needs to be “corrected”, you see, so that it doesn’t show an association.
Of course, the null hypothesis is that there is no association between aluminum-containing vaccines and various health outcomes.
However, like both the 2015 and 2019 MMR-autism studies, Andersson et al. actually found negative associations that with many outcomes were statistically significant—despite the biological implausibility that injecting children with a known neurotoxin is protective against chronic diseases and disorders.
This screams healthy user bias.
Overadjustment and Collider Bias
Another inappropriate covariate that the authors “adjusted” for was the number of visits to a general practitioner before age 2.
They thus treated office visits as a confounder, which falsely assumes it is a factor that is independent of aluminum exposure and health outcomes.
Children who visited their doctor more often may have been more likely to have higher aluminum exposure from vaccinations.
The more frequent visits might also have indicated early health problems, with symptoms potentially being explainable by aluminum exposure from vaccines.
Consequently, the number of doctor visits could very well have been a risk factor, and inappropriately treating it as a “confounder” to statistically “control” for it is yet another example of how the study was biased in favor of finding no association.
This is known as “collider bias”—a distortion that can suppress real associations or even make them appear protective.
This also helps to explain the implausible negative associations found in the study.
Treating Mitkus et al. (2011) as Good Science
Andersson et al. note how many parents are concerned about health risks from early aluminum exposure, then seek to reassure readers by stating, “However, a previous study suggests that the aluminum levels in infants from vaccinations are well below established minimal risk level thresholds.”
The reference is to a 2011 study by Robert J. Mitkus et al. in the journal Vaccine titled “Updated aluminum pharmacokinetics following infant exposures through diet and vaccination”.
Done by researchers from the US Food and Drug Administration (FDA), this is the key study that the CDC cites to assure parents that it is “safe” to inject children with aluminum—a claim that, given the actual nature of that study, is tantamount to scientific fraud.
As I detailed in my 2022 article “How the CDC Lies about the Safety of Aluminum in Vaccines”:
The FDA used a “minimal risk level” of aluminum exposure that had already been falsified by other research showing it to be at least several times too high.
That “safe” level of aluminum was also based on a study examining ingestion of a soluble form of aluminum in adult mice—not injection of insoluble aluminum particles in immature rodents.
While public vaccine policy apologists are fond of arguing that children get more aluminum exposure from food than vaccines, only 0.1% to 0.3% of ingested aluminum is absorbed through the gut.
The FDA researchers considered only the amount of aluminum in the blood as contributing to the body burden of aluminum toxicity, ignoring the particulate aluminum that remains elsewhere in the body—including aluminum that is taken up by macrophages and transported into the brain, where it accumulates.
The fact that Andersson et al. uncritically cite Mitkus et al. as though demonstrative of the safety of injecting infants with aluminum-containing vaccines reveals the depths of their own confirmation bias and readiness to accept whatever evidence, not matter how weak or scientifically invalid, to support their predetermined conclusion.
Likely Misclassification of “Unvaccinated” Children
Another problem with the 2019 MMR-autism study by Hviid et al. that is shared by the new aluminum study he coauthored is the use of a national vaccination registration database that has been shown to misclassify vaccinated children as “unvaccinated”.
In fact, as Hooker, Varia, and I discuss in our journal paper, an earlier study had shown that 55% of children indicated as “unvaccinated” in the database were in fact vaccinated, according to medical records.
The likely misclassification of many aluminum-exposed children as “unvaccinated” further biases the study in favor of finding no association between the exposure and health harms.
Assuming a Linear Dose-Response
In our paper, we observe that Hviid et al. (2019) was not faithfully designed to test whether vaccines can contribute to the development of autism in genetically susceptible children and therefore cannot possibly have falsified that hypothesis.
Similarly, Andersson et al. (2025) was not faithfully designed to test whether aluminum-containing vaccines can contribute to the development of allergic and atopic diseases, autoimmunity, and neurodevelopmental disorders.
It seems rather designed to find no association.
The question the public wants an answer to is: “Does aluminum exposure from vaccines harm children?”
But this new study answers a much narrower question: “Among healthy children who survived infancy without early diagnoses, do small differences in aluminum exposure predict later diagnoses?”
The answer to that narrow question might be “no,” but it tells us nothing about the broader safety of aluminum-adjuvanted vaccines.
Their model assumes that each additional milligram of exposure—e.g., going from X mg to X+1 mg, or from X+1 mg to X+2 mg, etc.—has the same proportional effect on risk across the entire exposure range.
That is, they assumed a linear dose-response for aluminum toxicity.
This ignores the finding of Crépeaux et al. (2017) that low-dose aluminum exposure caused neurobehavioral changes in mice, while higher doses did not, indicating a possible threshold effect where additional exposure did not result in additionally increased risk.
The authors of that study, titled “Non-linear dose-response of aluminium hydroxide adjuvant particles: Selective low dose neurotoxicity”, concluded that the assumption that aluminum neurotoxicity “obeys ‘the dose makes the poison’ rule of classical chemical toxicity appears overly simplistic.”
Conclusion
The media headlines, as ever, are blaring that science has repeatedly proven that vaccines are safe and effective.
There is no association between vaccinations and chronic illnesses and disorders, we are told.
We are supposed to believe that the new study out of Denmark “squashes anti-vaccine talking points about aluminum”.
The reality is that it is just more junk science—garbage in, garbage out—aimed at shielding vaccines from proper scientific inquiry.
This is easily enough explained by the authors’ conflicts of interest, including association with the Statens Serum Institute (SSI), a government agency that develops vaccines and is responsible for purchasing and supplying vaccines for Denmark’s national vaccination programs.
Hooker, Varia, and I detailed the similar conflicts of interest of the 2019 MMR-autism study, providing this chart illustrating the problem:
What emerges from critical analyses of these studies is a clear pattern of bias and data manipulation that is arguably tantamount to scientific fraud.
Paid propagandists masquerading as scientists have learned how to design studies to fail to find associations between vaccines and harms, and public vaccine policy advocates masquerading as journalists then report the findings of those junk studies as though conclusively proving that vaccines are “safe and effective”.
It’s also important to note that the CDC’s routine childhood vaccine schedule is much more aggressive than Denmark’s, which excludes, for instance, the three-dose series of the aluminum-containing Hepatitis B (HepB) vaccine injected into infants in the US starting on the very first day of their lives.
What we need are properly designed studies comparing long-term health outcomes—including all-cause mortality—between children who are fully vaccinated according to the CDC’s schedule and children who’ve never received a single vaccine.
As I show in my book The War on Informed Consent: The Persecution of Dr. Paul Thomas by the Oregon Medical Board, which features a Foreword by the current Secretary of Health and Human Service, Robert F. Kennedy, Jr., studies by independent researchers have found that unvaccinated children are healthier.
There’s a very good explanation for why the CDC won’t do the type of vaxxed-unvaxxed studies that parents have long demanded: you cannot find what you refuse to look for.
That’s once again the lesson to be learned from reading past the latest round of news headlines preposterously proclaiming to tell us what science says about vaccines.
There is only one study that is worth doing and that is vaccinated vs unvaccinated! Of course the “powers that shouldn’t be “ already know that,that will prove the horrific damage done by vaccines to our children! Dr.Paul Thomas long time paediatrician has already proved this using his own patient data.He was absolutely vilified and hounded by the medical mafia,but many good people have proof of this.People need to protect their young and research vaccines as much as they research buying a new car or appliance….