Danish Study on Aluminum in Vaccines ‘Should Be Retracted,’ Scientists Say

The authors of a Danish study published July 15 said they found no link between the aluminum in vaccines and autism. However, two scientists who reviewed the study, including corrected data published July 17, said the data don’t support the authors’ conclusion.

by Brownstone Institute

JULY 31, 2025

RSSREPUBLISH 

PRINTSHARE

 1 COMMENTS

By Tomas Fürst

Few issues of contemporary science are more contentious than the relation between childhood vaccines and adverse health outcomes. So, whenever a new study appears, it attracts a lot of attention.

See the new study by Andersson et al., titled “Aluminum-Adsorbed Vaccines and Chronic Diseases in Childhood: A Nationwide Cohort Study,” which appeared in the Annals of Internal Medicine in July.

Conclusion: “Cumulative aluminum exposure from vaccination during the first 2 years of life was not associated with increased rates of any of the 50 disorders assessed.”

We went through the paper, left no stone unturned, and report that we do not like what we found:

1. First, there is the issue of the “evolving” supplement. The original supplement included data on 2,239 neurodevelopmental events (such as autism and attention-deficit/hyperactivity disorder or ADHD), but it is no longer accessible. It has been replaced by a revised version that now reports 5,200 neurodevelopmental events (see Table 11 of the supplement).

This increase in the number of events altered the confidence intervals, and the updated data now show a statistically significant association between certain neurodevelopmental outcomes — particularly autism and ADHD — and aluminum exposure from vaccines.

This finding directly contradicts the paper’s conclusion, which states: “Cumulative aluminum exposure from vaccination during the first 2 years of life was not associated with increased rates of any of the 50 disorders assessed.” (See also the article at Children’s Health Defense.)

2. The study suffers from a major limitation due to its short follow-up period. In Denmark, children are typically diagnosed with autism, autism spectrum disorders and ADHD between the ages of 7 and 12 — or even later, depending on the severity of symptoms. However, the authors only tracked the children up to age 5, virtually ensuring that many relevant outcomes were missed.

Therefore, even if the study had not found an association between aluminum exposure and adverse neurodevelopmental outcomes (which, in fact, it had — see No. 1 above), the authors still would not have been in a position to conclude that no such association exists.

Nevertheless, the corresponding author, Anders Hviid, publicly misrepresented the findings, claiming that the study showed there was no link between aluminum exposure and autism or ADHD. It is also important to notice that the Statens Serum Institut (the institution behind the study) has great economic interests in developing and selling vaccines.

3. The exclusion criteria are suspect:

“To be included in our study, children needed to be alive at age 2 years, not have emigrated from Denmark, not have been diagnosed with certain congenital or preexisting conditions (including congenital rubella syndrome, respiratory conditions, primary immune deficiency, and heart or liver failure).”

However, death may be linked to vaccination. Moreover, many of the listed “preexisting” conditions may be, in fact, adverse events of previous vaccines. Thus, if the authors exclude many children who had been harmed by vaccines, they may miss some effects entirely. We would like to see the analysis with no children excluded.

4. Adjusting for the number of office visits (pre-2 years of age) may mask the effect significantly. Office visits are a “proxy” for the outcome (children who end up with some diagnosis probably visited GPs more often). Thus, there is a risk that the effect will be “adjusted away.”

To give an example, if we want to measure the effect of an intervention on blood flow through the left arm, we should not adjust for blood flow in the right arm. These two quantities are very probably correlated and adjusting for one will probably adjust much of the effect away.

5. According to Figure 1, more than 34,000 children were excluded because they had implausibly many registered vaccines in the first 2 years of life. Why would that be? This casts serious doubt on the integrity of the data.

Also, Figure 1 says that between 0 and 466,000 children were excluded from some analyses because they had the outcome in the first 2 years of life? What does that mean? We do not understand Figure 1, and the supplementary material, to which Figure 1 refers, does not help.

6. Measles-mumps-rubella or MMR vaccines allegedly contain no aluminum. So, recipients of MMR vaccines only are in the control group, together with recipients of no vaccines at all. This makes the control group rather diverse.

Moreover, the control group is very small. Thus, looking for a dose-response relation (by means of the Cox Proportional Hazards model) may not be appropriate because the data for the control group may get “outweighed” by the exposed group. It would be more sensible to directly compare the incidence of the events among the groups.

7. We cannot find raw (unadjusted) values for the incidence of health outcomes in the three groups stratified by different aluminum exposure. The raw numbers are not provided in either the manuscript or the supplement.

Only adjusted hazard ratios are given. Why? Such basic descriptive statistics should be included. We have asked the lead author for the raw data. He has not replied yet.

8. Figure 3 shows statistically significant positive effects of higher doses of Aluminum for many events. Since there is no plausible biological mechanism for this, it is a clear indication of the Healthy Vaccinee Effect, which means that the data were not deconfounded properly.

9. In the supplement Tables 10 and 11, the group with the highest exposure to aluminum is chosen as the reference group. Although this may be correct from the mathematical point of view, it is quite misleading for anyone used to reading forest plots.

All the effects are reversed, so a hazard ratio below one means that higher aluminum exposure is associated with a higher event rate.

In neurodevelopment, and especially in autism, Table 11 shows a statistically significant increase in these outcomes with higher aluminum exposure. This directly contradicts the conclusion of the paper. Moreover, it may be overlooked by the casual reader due to the highly misleading choice of the control group.

In its current state, the data does not support the conclusion of the study. The paper should be retracted.

Originally published by Brownstone Institute. This article was co-authored with Vibeke Manniche, M.D., Ph.D.

Tomas Fürst teaches applied mathematics at Palacky University, the Czech Republic. 

The views and opinions expressed in this article are those of the authors and do not necessarily reflect the views of Children’s Health Defense.

Brownstone Institute

The Brownstone Institute is a nonprofit organization founded May 2021. Its vision is of a society that places the highest value on the voluntary interaction of individuals and groups while minimizing the use of violence and force including that which is exercised by public or private authorities.

Source: Children’s Health Defence