Ari Armstrong's Web Log (Main) | Archives | Terms of Use
Trump's War on Tylenol
Trump wrongly blames acetaminophen for autism and dangerously encourages pregnant women to "tough it out."
by Ari Armstrong, Copyright © 2025
Another day, another attack on American businesses by Donald Trump.
The active ingredient in Tylenol is acetaminophen a.k.a. paracetamol. Donald Trump and RFK held a bizarre media conference September 22 (transcript, release) in which they blamed pregnant women taking Tylenol for autism. Apparently Dr. Trump never got the memo that correlation does not prove causation.
James Cusack, chief executive of Autistica, a charity for autism research, told Nature, "There is no definitive evidence to suggest that paracetamol use in mothers is a cause of autism, and when you see any associations, they are very, very small." Nature reports that a study out of Sweden did find a small association between acetaminophen use and autism, but that association disappeared when adjusting for siblings. And: "A large, high-quality study from Japan of more than 200,000 children—also using sibling comparisons and published this year—found no link between acetaminophen use in pregnancy and autism."
The Swedish study finds: "Acetaminophen use during pregnancy was not associated with children's risk of autism, ADHD, or intellectual disability in sibling control analyses. This suggests that associations observed in other models may have been attributable to confounding."
Nature reports that the Swedish study "showed that around 1.42% of children exposed to acetaminophen during pregnancy were autistic, compared with 1.33% of children who were not exposed." So autism is almost entirely related to other things. Such associations can be caused by other things, or they even can just be statistical noise. One possibility is that the same conditions that cause autism also cause (prompt) some women to take more acetaminophen.
What does cause autism, then? In a word: genes (mostly). What has caused the increased diagnoses of autism? Mostly, expanded criteria and more checking. Here are bits from a 2021 study in Psychological Medicine:
More than 100 risk genes have been implicated by rare, often de novo, potentially damaging mutations in highly constrained genes. These account for substantial individual risk but a small proportion of the population risk. In contrast, most of the genetic risk is attributable to common inherited variants acting en masse, each individually with small effects. Studies have identified a handful of robustly associated common variants. Different risk genes converge on the same mechanisms, such as gene regulation and synaptic connectivity. . . .
Prevalence estimates of autism have steadily increased from less than 0.4% in the 1970s to current estimates of 1–2%. The increase is largely explained by broadening diagnostic criteria to individuals without ID [intellectual disability] and with milder impairments, and increased awareness and recognition of autistic traits. [See source for references.] . . . .
Twin studies suggest that 9–36% of the variance in autism predisposition might be explained by environmental factors. . . .
It is challenging to infer causality from observed associations, given that confounding by lifestyle, socioeconomic, or genetic factors contributes to non-causal associations between exposures and autism. Many putative exposures are associated with parental genotype (e.g. obesity, age at birth), and some are associated both with maternal and fetal genotypes (e.g. preterm birth). . . .
Twin studies were the first to demonstrate the heritability of autism. . . . Subsequently, over 30 twin studies have been published, further supporting the high heritability of autism. A meta-analysis of seven primary twin studies reported that the heritability estimates ranged from 64% to 93%.
The Trump administration bragged on X that, in 2017, Tylenol posted, "We actually don't recommend using any of our products while pregnant." But not recommending is not the same thing as dis-recommending, and it certainly is not the same thing as conceding that Tylenol use causes autism. Tylenol currently states on its web page: "If pregnant or breast-feeding, ask a health professional before use." On September 24, Tylenol posted the following message to the main page of its web site (as a pop-up):
We stand with science, and we stand with you. . . . Tylenol is one of the most studied medications in history—and is safe when used as directed by expecting mothers, infants and children. The facts remain unchanged: over a decade of rigorous research, endorsed by leading medical professionals, confirm there is no credible evidence linking acetaminophen to autism. . . . High fevers and pain are widely recognized as potential risks to a pregnancy if left untreated, especially in the first trimester. Remember to talk to your doctor. We prioritize science as the core of how we provide care, and that will never change.
The New York Times has more on this.
Trump also promoted leucovorin as a treatment for autism. Zachary Rubin writes:
Trump also used the stage to revisit old anti-vaccine talking points, repeating long-debunked claims about MMR, mercury, and hepatitis B vaccination. . . .
In August 2025, a Harvard-affiliated review suggested acetaminophen use in pregnancy might increase the risk of autism and ADHD. But the study was flawed. It relied heavily on observational data and retrospective recall, which are prone to bias, and many of the included studies failed to separate the effects of acetaminophen from the fevers, pain, or illnesses it was taken to treat. . . .
On leucovorin (folinic acid), the evidence is promising but preliminary. Small randomized controlled trials have shown modest improvements in language or adaptive behavior among subsets of autistic children, especially those with folate-receptor autoantibodies. These findings are intriguing and worthy of further study, but they do not make leucovorin a cure or a universally effective therapy.
Alison Singer told the New York Times about Trump's announcement:
It took me straight back to when moms were blamed for autism. If you can’t take the pain or deal with fever, if you can’t tough it out, then you are to blame if your child has autism. That was shocking. Simply shocking.
Helen Tager-Flusberg added:
I was expecting some of what was presented, but I have to say I was shocked and appalled to hear the extreme statements without evidence in support of what any of the presenters said. In some respects this was the most unhinged discussion of autism that I have ever listened to.
Along these lines, a headline from Stat says, "Autistic moms feel shamed and stigmatized by Trump's Tylenol warning: Blaming mothers' use of the pain reliever recalls the myth of the 'refrigerator mother.'"
William Meller raises an interesting possibility writing for Skeptic:
Renee Gardner, one of the [Swedish] study's authors, suggested that genetic tendencies that increase the risk of autism often overlap with genes influencing pain perception, meaning the genetic risk factor causes both pain (leading to paracetamol use) and autism, making the painkiller an innocent bystander. . . .
One common, though often overlooked, familial confounder involves the complex relationship between pain sensitivity and neurodevelopment. As Gardner explained, maternal conditions like hypermobility (unusually flexible joints)—which can cause joint pain requiring painkillers—are also "more likely to have autistic children." If researchers fail to account for this overlap, "painkillers can wrongly appear to be a risk factor."
Meller sensibly advises:
Untreated maternal conditions like fever, chronic pain, or inflammation are known risk factors for adverse outcomes in offspring. . . . That evidence confirms that while caution is warranted in all drug use during pregnancy, the proven risks of foregoing necessary treatment far outweigh the confounding-laden claims of a causal link between Tylenol and neurodevelopmental disorders.
Jessica Knurick also posted thoughtful commentary on the topic.
Some anti-vaccine groups are mad at RFK for linking autism to Tylenol rather than to vaccines; as Tyler Cowen says, they're "biting each others' ankles over who will be the greater fool."
Those who want to understand autism should see Is This Autism? by Donna Henderson and Sarah Wayland. Also A Kind of Spark is a lovely two-season show on BYUTV featuring autistic characters played by autistic actors.
Note: I am not a doctor, I have no medical training, and nothing in this article is intended as medical advice.