Sen. Cassidy Told RFK Jr. This Study Proves Vaccines Don’t Cause Autism — Here’s What’s Wrong With the Study

During Thursday’s U.S. Senate confirmation hearing for RFK Jr., Sen. Cassidy cited a meta-analysis published in 2014 in the journal Vaccine as scientific proof that vaccines don’t cause autism.

by Brian Hooker, Ph.D.

During Thursday’s U.S. Senate confirmation hearing for Robert F. Kennedy Jr. — President Donald Trump’s nominee for health secretary — Sen. Bill Cassidy (R-La.) cited a meta-analysis published in May 2014 in the journal Vaccine as scientific proof that vaccines don’t cause autism.

A “meta-analysis” is intended to be an unbiased and scientific review of the literature — in this case, to investigate the link between vaccines and autism. Unfortunately, this meta-analysis, authored by Luke E. Taylor, Amy L. Swerdfeger and Guy D. Eslick, is propaganda poorly masquerading as science.

Using standard criteria, the authors searched the open scientific literature and found 46 eligible studies for their meta-analysis. However, rather than consider the entirety of the 46 studies, the authors cherry-picked only 10 studies — all of which concluded there was no link between vaccines and autism.

After reviewing the list of eligible studies, the only reason I can find for selecting these 10 studies is that they all have one thing in common: They support the authors’ preferred foregone conclusion, which is that vaccines have nothing to do with the autism epidemic we’re facing.

Even worse, the 10 studies included in the analysis considered only the measles-mumps-rubella (MMR) vaccine and the mercury-containing preservative, thimerosal. Yet the authors claim in their title, “Vaccines are not associated with autism.” This is misleading at best and fraudulent at worst, based on the paper’s contents.

It is unscientific to draw a broad-sweeping conclusion about vaccines and autism based on one vaccine and a single vaccine additive when study authors don’t address any of the 16 other vaccines on the Centers for Disease Control and Prevention’s (CDC) childhood vaccine schedule, or the myriad other additives (besides thimerosal) in vaccines.

If you were checking 17 plates of food for poison, would you check one plate, determine it poison-free — and then declare all the plates of food safe to eat? Of course not. Neither can the authors of this meta-analysis presume or conclude that no vaccines cause autism by testing only one vaccine, the MMR, and one vaccine ingredient, thimerosal.

Further, the Institute of Medicine (now the National Academy of Medicine) in 2012 issued a report “Adverse Effects of Vaccines: Evidence and Causality” which concluded that there was insufficient evidence to deny or accept a relationship between the DTaP (diphtheria tetanus acellular pertussis) vaccine and autism.

This conclusion was based on the paucity of research in this area. In other words, this prestigious body concluded that because of insufficient information, “they just don’t know” if DTaP causes autism. This conclusion contradicts the Taylor et al. study’s flawed conclusion that “vaccines are not associated with autism.”

Let’s look further at the cherry-picking of studies by Taylor et al. The study authors chose only four studies that assessed thimerosal (out of the total pool of 46 studies) again, without any explanation of their criteria for narrowing the study pool.

The research papers include Hviid et al. 2003, Andrews et al. 2004, Verstraeten et al. 2003 and Price et al. 2010. A study I authored in 2014 — “Methodological Issues and Evidence of Malfeasance in Research Purporting to Show that Thimerosal in Vaccines is Safe” — thoroughly debunks these four studies in the Taylor et al. meta-analysis.

The Price study also was refuted, by a study published in 2012: “Vaccine Safety Study as an Interesting Case of ‘Over-Matching.’” Even the lead author of the CDC’s own “gold standard” study by Dr. Thomas Verstraeten, in a 2004 letter to the editor of the journal Pediatrics, admitted that his was a “neutral” study meaning that it could not rule out the relationship between thimerosal and autism.

Unfortunately, the meta-analysis cited by Sen. Cassidy also ignored any PubMed-indexed research study showing a statistically significant connection between thimerosal exposure in infants via vaccines and autism. This includes studies focused on thimerosal in the hepatitis B vaccine and cumulative thimerosal exposure from the vaccine schedule, both based on data from the CDC’s Vaccine Safety Datalink.

The same is true for the six studies Taylor et al. cherry-picked for the MMR vaccine: Madsen et al. 2002, Smeeth et al. 2004, Uchiyama et al. 2007, DeStefano et al. 2004, Mrozek-Budzyn et al. 2010 and Uno et al. 2012. Each study denied any relationship between the vaccine and autism.

However, when you look beyond the meta-analysis, the CDC’s own DeStefano et al. 2004 study showed that children receiving the MMR on time were 1.49 times more likely to get an autism diagnosis as compared to those who got their first MMR after their third birthday. When considering boys alone, DeStefano found that the risk increased to 1.67 times — and both results were statistically significant to a greater than 95% certainty.

The same data the CDC used in DeStefano et al. 2004 were reanalyzed and reported in a peer-reviewed study I authored and published in 2018 in the Journal of American Physicians and Surgeons. African American males in the study were 3.86 times as likely to receive an autism diagnosis if they got the vaccine before age 36 months compared to after age 36 months. This result was highly statistically significant.

Finally, none of the studies considered in the Taylor et al. meta-analysis considered a completely unvaccinated control group. Without this type of comparison (vaccinated versus unvaccinated), it is impossible to determine if there is a vaccine-autism link.

Three studies (Mawson et al. 2017, Hooker and Miller 2021 and Mawson and Jacob 2025) — all using an unvaccinated control group — have since been published showing that fully vaccinated children are at least 4.3 times more likely to receive an autism or autism spectrum disorder diagnosis compared to their unvaccinated peers.

In summary, the Taylor et al. 2014 study misses the mark completely regarding the relationship between vaccines and autism. It is unfortunate that this paper made it past the peer-review process as it has most likely deceived and confused many practitioners and researchers. The study gives a false answer to a timely and important medical question.