Correlation Doesn't Equal Causation: On the Acetaminophen Controversy

If your pregnant, just talk to your doctor

Professional and personal obligations have prevented me from completing much of the writing I’ve wanted to do. However, a recent exchange with a family member has compelled me to touch on the Trump administration’s linking of acetaminophen to autism. As the uncle of two autistic kids, as well as having friends with autistic children,¹ I understand how deeply personal this issue is for people. I’ll also note that I was diagnosed with generalized anxiety disorder more than ten years ago and that I’m in the process of being evaluated by my doctor for ADHD.² Both disorders are considered to be part of the neurodivergent spectrum.

The exchange I mentioned was about a meme that I shared on Instagram, which led to a brief back-and-forth. The family member wrote, “Ignore the science all you like, but he was just reporting a Harvard done study, and you don’t get more left leaning than Harvard.” After my response casting serious doubt about the linkage, he wrote, “According to the Harvard study there is evidence to suggest that there is a correlation in some people.”

To be clear, there’s no credible evidence or study finding that acetaminophen use during pregnancy causes autism.

The study was a meta-analysis coauthored by Dr. Andrea Baccarelli, the Dean of the Harvard T.H. Chan School of Public Health. The study only concludes that there is an association between acetaminophen use during pregnancy and increased incidence of neurodivergent disorders (NDDs) like autism.

Notably, the study did not conclude that acetaminophen causes NDDs, nor does it say that women who are pregnant should stop taking acetaminophen. The Food and Drug Administration news release on the topic only mentions the “association” and “correlation” while also stating “acetaminophen is the only over-the-counter drug approved for use to treat fevers during pregnancy, and high fevers in pregnant women can pose a risk to their children” and that “aspirin and ibuprofen have well-documented adverse impacts on the fetus.”

In its conclusions, the study states, “While this association warrants caution, untreated maternal fever and pain pose risks such as neural tube defects and preterm birth, necessitating a balanced approach. We recommend judicious acetaminophen use—lowest effective dose, shortest duration—under medical guidance, tailored to individual risk–benefit assessments, rather than a broad limitation.”

Separately, the Harvard School of Public Health’s statement on the release of the study includes a quote from Baccarelli, who states that the “association is strongest when acetaminophen is taken for four weeks or longer.” The study itself also notes that “acetaminophen remains the preferred analgesic due to its relatively favorable safety profile compared to other medications[.]”

Baccarelli’s comments and the study are markedly different from Trump’s public statement. Astonishingly, Trump told pregnant women, “Don’t take Tylenol. Don’t take it. Fight like hell not to take it.” He encouraged women to “tough it out” rather than take acetaminophen. He also claimed that “[t]here’s no downside” to not taking the over-the-counter fever-reducing drug.³

Certainly, my friends who are defending the Trump administration over this claim would push back if I said that there’s a correlation between states with liberalized gun laws and higher homicide rates. Now, I’m defining “liberalized gun laws” as requiring the issuance of a concealed carry permit (“shall issue”) and constitutional carry. The states with the highest homicide rates are Louisiana (10.8 murders per 100,000 inhabitants), New Mexico (10.5), Alabama (8.7), Tennessee (7.9), Missouri (7.8), South Carolina (7.5), North Carolina (7.5), Mississippi (7.4), Arkansas (7.3), and Maryland (7.2).⁴ Of these states, only North Carolina, New Mexico, and Maryland have not enacted constitutional carry.⁵

Correlation, of course, does not equal causation. There are a number of potential other factors that could also play a role in the homicide rates in these states, and it’s likely a combination of them.

The administration is also using a March 2017 tweet from Tylenol—see posts from the White House and the Department of Health and Human Services here and here—as part of their campaign. The tweet from Tylenol stated, “We actually don’t recommend using any of our products while pregnant. Thank you for taking the time to voice your concerns today.” The maker of Tylenol says that the tweet has been taken out of context. Indeed, the post was a reply to a user who no longer has an account on the platform.

The Tylenol account notes in a response to a user, “[W]e haven’t tested Tylenol to be used during pregnancy[.]” In response to a user, who has since deleted their account, Tylenol says, “If you are pregnant/nursing, seek the advice of your healthcare professional before using Tylenol or any other medication.” In another response to a user who gave Tylenol a “shout out for being pregnancy safe,” Tylenol replied, “Thanks for the shout out Carrie! Just make sure to talk to your doctor before taking Tylenol while you’re pregnant.”

Tylenol isn’t simply taking the beating. The brand is pushing back. This pop-up now appears on its website.

I do find it interesting, though, that those pushing the narrative aren’t addressing one major glaring issue. Baccarelli has been a paid expert witness in lawsuits on the supposed connection between NDDs and acetaminophen. Now, Harvard points out, “Baccarelli noted in the ‘competing interests’ section of the research paper that he has served as an expert witness for plaintiffs in a case involving potential links between acetaminophen use during pregnancy and neurodevelopmental disorders.”

The study does indeed note that “competing interest.” On page 39, the study explains, “Dr. Baccarelli served as an expert witness for the plaintiff’s legal team on matters of general causation involving acetaminophen use during pregnancy and its potential links to neurodevelopmental disorders. This involvement may be perceived as a conflict of interest regarding the information presented in this paper on acetaminophen and neurodevelopmental outcomes. Dr. Baccarelli has made every effort to ensure that this current work—like his past work as an expert witness on this matter—was conducted with the highest standards of scientific integrity and objectivity.”

Baccarelli was paid “about $150,000” to serve as an expert witness. During that case, District Court Judge Densie Cote agreed with the defense’s contention that Baccarelli “cherry-picked and misrepresented study results and refused to acknowledge the role of genetics in the etiology of either ASD or ADHD.”⁶ Cote granted the defense’s motion to exclude Baccarelli’s opinions.⁷ To be clear, I am not implying or claiming that the meta-analysis Baccrelli recently co-authored doesn’t meet scientific standards. However, there is a perception of a conflict of interest that’s hard to avoid.

We’ve got to recognize that there are multiple factors that contribute to NDDs, including genetic factors. There’s also a greater awareness of NDDs than in the past. That’s a good thing because, in children, it allows schools to catch learning needs and behavioral issues that could be associated with NDDs. Pregnant women should talk to their doctors about acetaminophen or any other drug they use during the course of their pregnancy and while nursing their baby. That is, and always has been, the right approach.

1. At least a couple of friends have children with severe cases of autism.

2. I’ve gone through the initial evaluation and will undergo a test and sleep study soon.

3. Uh, yes, there can be downsides if the fever is high enough.

4. The District of Columbia has the highest homicide rate, at 25.5 murders per 100,000 inhabitants. I would argue that the District, despite having relatively strict gun laws, in a unique position because of its geographical proximity to states with relaxed gun laws.

5. In New York State Rifle and Pistol Association, Inc. v. Bruen (2022), the Supreme Court held that a New York law restricting the issuance of a permit to carry a firearm unless there was a demonstrable or justifiable need was unconstitutional. Although many states still require a permit to carry a firearm, I believe only Connecticut and Rhode Island are “may issue” states.

6. See p. 62. The full quote from Judge Cote’s opinion: “Defendants argue that Dr. Baccarelli’s opinions are unreliable for several reasons. They contend that he improperly applied a ‘transdiagnostic’ approach to neurodevelopmental disorders that elides meaningful differences between ADHD and ASD. They contend as well that he did not conduct a reliable Bradford Hill or Navigation Guide analysis for several reasons, including that he cherry-picked and misrepresented study results and refused to acknowledge the role of genetics in the etiology of either ASD or ADHD. They are correct.”

7. There are two specific reasons for the exclusion. Please see p. 70 and p. 86 of Cote’s opinion.