Acetaminophen use during pregnancy is not likely to raise the risk of having a child with autism, ADHD, or intellectual disability, according to a new study published Friday.
Researchers across Europe reexamined evidence from multiple studies investigating the link between these conditions and the use of paracetamol — called acetaminophen in the United States — and found that these purported associations fell apart after controlling for confounding factors.
Acetaminophen, the active ingredient in Tylenol, has long been a frontline medication for relieving pain or fever in people who are pregnant. The scientific literature shows that this practice shouldn’t change, said co-author Asma Khalil, obstetrics and maternal fetal medicine professor at St. George’s Hospital, University of London.
“The message really is clear,” said Khalil. “Paracetamol remains a safe option during pregnancy when taken as guided — for the duration that’s needed, with a correct dose.”
Independent health experts praised the study’s methods and conclusions, which diverge sharply from the recommendations offered by President Trump and top health officials in September, when they said that acetaminophen should be used only if absolutely needed during pregnancy and launched a nationwide public information campaign to inform doctors and families of the purported risk. Their announcement came as part of their bid to find the cause of autism, which decades of research has already linked to primarily genetic factors.
The findings of this new study, published in The Lancet, are unlikely to change these recommendations. Health and Human Services officials said the study did not refute claims from other researchers.
“Many experts have expressed concern of the use of acetaminophen during pregnancy, including Dr. Andrea Baccarelli, the Dean of the Harvard School of Public Health,” said HHS spokesperson Andrew Nixon.
The U.S. government is an outlier in discouraging acetaminophen use — the American College of Obstetricians and Gynecologists, its British counterpart the Royal College of Obstetricians and Gynaecologists, as well as the European Medicines Agency continue to recommend the drug as a first-line treatment for pain and fever in pregnant people.
The U.S. health officials’ stance that exposure to acetaminophen during pregnancy is associated with autism spectrum disorder spurred the European researchers into a new review of the available evidence. Acetaminophen is a critical tool for physicians, the go-to drug to use during pregnancy because other commonly available painkillers, such as Advil (ibuprofen), and opioid medications can put a fetus at risk of kidney damage, preterm birth, or even stillbirth. Acetaminophen is not just a pain reliever for people who are pregnant, it also mitigates fevers that can increase the likelihood of birth defects.
Scientists have had ample time to investigate the potential adverse outcomes of acetaminophen, a drug first used clinically in 1893so much so that “there might be more reviews of acetaminophen studies than actual acetaminophen studies with primary data,” quipped Brian Lee, a professor of epidemiology at Drexel University. Generally, they have found that the benefit of acetaminophen outweighs the risk associated with using it. Why, then, did Khalil and her team add another meta-analysis to the large pile of studies?
Lee and other experts see the Lancet paper as a direct rebuke of a meta-analysis published in 2025 and co-authored by Baccarelli. The two papers analyze many of the same studies, but the 2025 review arrived at a drastically different conclusion, finding “significant links” between acetaminophen use during pregnancy and increased risks of conditions like autism or ADHD. Baccarelli’s paper was cited by top health officials in September, saying that the dean has established a “causal relationship” between the drug and these adverse outcomes.
Baccarelli has so far declined to comment on health officials citing his work to urge caution in the use of acetaminophen, or about the paid testimony he has given to strongly assert a link between acetaminophen use during pregnancy and conditions like autism. A Harvard T.H. Chan School of Public Health spokesperson said Baccarelli was unavailable to discuss the new Lancet study.
Authors of the new Lancet study winnowed down thousands of published papers to 43 studies based on risk for “bias,” or how thoroughly papers’ study designs and protocols accounted for confounding factors, such as genetics. The authors prioritized studies in which a mother took acetaminophen during one pregnancy, but not another. If autism or other neurodevelopmental outcomes were associated with an environmental factor such as taking Tylenol, then researchers would find that differential in the siblings’ neurodevelopmental outcomes.
But that’s not what the researchers found, especially when evaluating the data from two notable studies out of Sweden and Japanwhich used this method across massive datasets. When the researchers conducted a meta-analysis of these and 15 other papers, they found no association between maternal acetaminophen use and a heightened risk of autism, ADHD, or intellectual disability.
David Mandell, a professor of psychiatry at the University of Pennsylvania’s Perelman School of Medicine, applauded the Lancet study’s robust design and commitment to controlling for confounding variables, calling it a “pretty solid rebuke” of Baccarelli’s paper, which devalued sibling control studies in its analysis.
“As much as there exists a scientific best practice to conducting reviews and meta-analyses, they are still very much completely resulting from subjective interpretation,” said Lee. “Beauty is in the eye of the beholder, but compared with the horrifically inaccurate Prada et al. review, this is a significant improvement.” (Diddier Prada, an assistant professor of population health science and policy at Mount Sinai’s Icahn School of Medicine, was a co-author on the Baccarelli 2025 paper.)
The conflicting conclusions of Baccarelli’s paper and this paper from European researchers highlight the critical choices that scientists make when crafting a study. There’s no perfect design that eliminates all confounders, especially when sifting through the endless variables that influence pregnancy and the historic omission of pregnant people from clinical trials.
Sibling control studies like those out of Sweden and Japan have become a common in epidemiological circles on the backs of a tantalizing promise: having the same mother means an association found between, say, Tylenol and autism will be free from the confounding factors constant between pregnancies, such as genetics, socioeconomic status, and more. If you have a large enough dataset, you can tackle critical health questions.
Zeroing in on this specific type of differential exposure within a family has limits, such as excluding families with a single child. It can also bias results toward the null, meaning toward finding no association between acetaminophen and certain disabilities. National Institutes of Health Director Jay Bhattacharya raised these points when excoriating siblings studies on social media in September.
But the limitations cut both ways, said Jonathan Sebat, professor of psychiatry and cellular biology at the University of California, San Diego. Observational studies like the ones favored in Baccarelli’s meta-analysis can overestimate the existence of an association. He said it is critical to use a study design that controls for genetic factors when autism is mainly inherited, adding that it is hard to summarily dismiss these sibling cohort studies when their accounting for genetic and familial confounding did not just weaken the association. It disappeared.
“That picture is not very supportive of the hypothesis,” said Sebat. “I need a lot more evidence now to really believe that there’s an acetaminophen association.”
The myriad factors that influence child development may complicate a definitive pronouncement such as Tylenol use does not cause autism or other neurodevelopmental disorders. But health experts say the latest review of evidence shows that physicians should continue to use acetaminophen to treat pain and fevers during pregnancy.
“This [study] confirmed that there is no relationship between taking paracetamol in pregnancy and a higher likelihood of autism, ADHD, or intellectual disabilities in the offspring,” said Grainne McAlonan, a translational neuroscience professor at King’s College London. “While the impact of last year’s announcement has been extensive, I hope the findings of this study bring the matter to a close.”
