Researcher on Tylenol-Autism Connection: Not the Best Science

You will hear a lot in the next few days about Tylenol (acetaminophen) during pregnancy and the risk of autism and attention deficit hyperactivity disorder (ADHD). President Donald Trump and Health & Human Services Secretary Robert F. Kennedy Jr. have teased about releasing a report with new evidence about acetaminophen and neurodevelopmental disorders.

Just this afternoon, HHS said the FDA will issue a physician notice and begin the process to initiate a safety label change for acetaminophen

But physicians, researchers and autism experts — many of whom are posting on LinkedIn to try to clear up the confusion — continue to say that acetaminophen can be used safely during pregnancy to treat pain and fever and that there is no one cause of autism.

Brian K. Lee, Ph.D.

“The science hasn’t changed, regardless of what comes out from the report. The science is the science, and it points towards a lack of good evidence for a causal effect of acetaminophen on autism and ADHD, Brian K. Lee, Ph.D., Drexel University Dornsife School of Public Health, told reporters today. “The administration has done some things that raise eyebrows, at least as far as scientists are concerned, and they aren’t grounded in the best scientific evidence.”

Lee, who took part in a comprehensive study of 2.5 million people examining acetaminophen and neurodevelopmental disorders, said that current research points in one direction—away from a causal connection.

What makes pregnancy safety difficult to assess is that these studies are observational and not randomized, controlled trials. “If you aren’t achieving this gold standard of a randomized, controlled trial, then that always leaves open the door for some possibility that there is a causal effect of anything,” Lee said.

Trying to assess causation in studies that are not randomized can have many challenges that impact the validity of the findings. They require strong assumptions, careful statistical analysis, and considerations for alternative explanations.

Lee points out that observational trials looking at safety in pregnancy can have low sample sizes, selection bias, and the inability to control for genetics. Some studies that have found a link between acetaminophen and neurodevelopmental disorders may not have gone to the highest lengths to account for confounders, which are variables that could impact the findings but are not controlled for in the study. A confounding variable may distort or mask effects of another variable.

In the case of autism and ADHD, genetic confounding in some studies may account for their conclusions, Lee said. “We know autism, ADHD and other neurodevelopmental disorders are highly heritable,” he said.

If researchers don’t control for genetics, there’s a possibility that conclusions can be misleading. Lee argues that many studies fail to account for the possibility that the same genetic factors that predispose children to these conditions might also make mothers more likely to experience pain and therefore use acetaminophen during pregnancy.

Autism is likely the result of genetics and the environment, including parents’ age at conception, prematurity or very low birth weight, and exposures that affect brain development, such as fever or illness during pregnancy.

In a statement, Alycia Halladay, Ph.D., chief science officer at the Autism Science Foundation, said, “It is disingenuous and misleading to boil autism’s causes down to one simple thing. We know that autism is incredibly complicated, and we need to move away from studies that simplify it down to one exposure without any other considerations.”

Lee and his colleagues in their study controlled for this variable with a sibling control. Published in April 2024 in JAMA, they collected data from Medical Birth Register, the National Patient Register, and the longitudinal integrated database for health insurance and labor market studies of 2,480,797 children born between 1995 and 2019 in Sweden. They found that acetaminophen use during pregnancy was not associated with children’s risk of autism, ADHD, or intellectual disability.

So how did we get here?

The debate has serious real-world implications for pregnant women’s healthcare. Dr. Lee warned that limiting acetaminophen use could be dangerous, noting that alternative pain relievers like NSAIDs are contraindicated during pregnancy, and untreated fever poses known risks to both mothers and unborn children. “Fever is a known harmful quantity for mothers and their unborn children, and if you don't treat a fever, the consequences can be severe,” he said.

What makes this all so confusing is that some recent studies have looked at research of prenatal acetaminophen exposure and autism and have come to a different conclusion, although Lee said there may be challenges related to bias, controlling for genetics, lack of dosing information and reliance on self-reported data.

Complicating this analysis is that researchers don’t entirely know how acetaminophen works to relieve pain. “I'm not an animal model or a physiological expert, but there are some open questions about the physiological pathways,” he said.

Several studies of acetaminophen have found a link to acetaminophen and autism. One study published in August 2025 in BMC Environmental Health analyzed results from 46 previous studies and represents data from over 100,000 participants. This study from researchers at Mount Sinai and Harvard T.H. Chan School of Public Health found an association between acetaminophen and increased incidence of neurodevelopmental disorders (NDDs), including autism and ADHD.

Lee, however, pointed out that this study was a meta-analysis — a study of studies — rather than new original research, and he suggests that this approach may weigh lower-quality studies higher than other studies.

Another study, published earlier this month in Paediatric and Perinatal Epidemiology, assessed more than 200,000 people in Japan. This study used sibling control analysis and found a small increase in risk of neurodevelopmental disorder diagnoses in children whose mothers used acetaminophen during pregnancy. But the researchers in this study said that sensitivity analyses suggested that unmeasured confounding, misclassification, and other biases may partially explain the finding.

Lee also discussed research from Norway, the Norwegian Mother, Father, and Child Cohort Study (MoBa), that initially found acetaminophen was associated with symptoms at three years of age for ADHD and autism. But a later study using the same data and that used a sibling control found a lowered risk for ADHD and autism. Lee noted significant limitations in interpreting these results, including the instability of childhood symptoms and differences between early symptom manifestation and later clinical diagnosis.

Additionally, Lee discussed the ongoing litigation against the makers of Tylenol and store-branded acetaminophen in the Southern District of New York. In this lawsuit, the plaintiffs relied heavily on biomarker studies that suggested that acetaminophen metabolites were found in umbilical cord blood and that showed a strong association with neurodevelopmental disorders.

But Lee said the half-life of acetaminophen is just a few days. “You’re really measuring acetaminophen use just before or at the time of delivery, so it doesn’t really say much about the acetaminophen use over the pregnancy during a neurodevelopmentally relevant period.

“Every single study out there has flaws, and it’s going to be a subjective opinion as to what are flaws and what are not,” he said.