A large-scale study of over 700,000 mother-child pairs in Hong Kong found no causal link between prenatal acetaminophen use and the development of autism or ADHD, according to research published in JAMA Internal Medicine. The findings contradict claims suggesting that Tylenol causes autism, demonstrating that neither the dosage, timing, nor frequency of the medication influenced the risk of these neurodevelopmental conditions.
The study analyzed electronic health records spanning from 2001 to 2023. Researchers found that approximately 43 percent of the children in the cohort had been exposed to acetaminophen in utero, yet the data showed no association between the drug’s use and autism or ADHD diagnoses. The results remained consistent regardless of the mother’s age or the specific trimester during which the medication was taken.
These results emerge amid a public health debate fueled by claims from President Trump and Robert F. Kennedy Jr. Both have previously suggested that environmental factors, specifically the use of Tylenol during pregnancy, contribute to the rising rates of autism. However, Kennedy has since acknowledged that there is no proven causal link between the two.
How the Hong Kong study debunked the Tylenol-autism link
The researchers utilized a sibling-matched design to isolate the effects of acetaminophen from other variables. By comparing siblings where one was exposed to the drug in utero and the other was not, the study minimized the influence of shared genetics and home environments. When this rigorous method was applied, the link between the painkiller and autism disappeared.
Some previous studies reported a correlation between prenatal Tylenol use and autism, which critics of the drug often cite. However, the JAMA Internal Medicine study explains this through a “negative control” analysis. This process involves looking at a group that should not produce the experimental result—in this case, mothers who took acetaminophen before becoming pregnant or after giving birth.
The researchers found that these “negative control” groups also showed an association with autism in their children. Because it is biologically implausible for a medication taken before or after pregnancy to cause a fetal brain abnormality, the researchers concluded that the perceived link is not causal. Instead, the association likely reflects shared genetic or environmental factors that both trigger a mother to take pain medication and increase the likelihood of a child being born with autism.
What is the difference between correlation and causation in autism research?
In medical research, correlation means two things happen at the same time, while causation means one thing actually triggers the other. The Hong Kong study highlights a common pitfall in autism research: confusing the two. When researchers drop the sibling-matched design and simply compare exposed children to unexposed children, a correlation often appears.
This distinction is critical for pregnant women who rely on acetaminophen. The study reinforces that the medication itself does not appear to be the driver of neurodevelopmental disorders.
Why these findings matter for public health policy
The debate over acetaminophen has led to significant real-world consequences, including lawsuits against Kenvue, the manufacturer of Tylenol. By demonstrating that the link is “biologically implausible” through negative control analysis, the study provides a scientific counter-narrative to the claims being pushed by political figures.
The scale of the study—covering 700,000 pairs over two decades—provides a level of statistical power that makes the absence of a causal link highly significant.