Sibling-matched design

The researchers then performed a sibling-matched analysis, comparing autism and ADHD cases among siblings, some of whom were exposed to acetaminophen in utero and some who weren’t. This study design helps account for unmeasured family factors that influence the likelihood of the conditions, particularly genetics and shared environmental conditions. The autism analysis included over 124,000 sibling-matched children, and an analysis of attention-deficit/hyperactivity disorder (ADHD) included a cohort of over 97,000 sibling-matched children.

The researchers saw no link between prenatal acetaminophen use and either condition. It didn’t matter what dosage of acetaminophen was taken, when it was taken during the pregnancy (which trimester), how often it was taken, or how old the mother was at the time. There was simply no link between acetaminophen and autism or ADHD.

Interestingly, there was a link when the researchers dropped the sibling-matched design and instead compared acetaminophen-exposed with unexposed children, which is a finding that has come up in other studies. But when the researchers performed a “negative control” analysis and compared children whose mothers had taken acetaminophen before ever getting pregnant or after they had given birth compared to mothers who didn’t use the painkiller, they also saw an association—one that is “biologically implausible.”

“Collectively, these findings suggest that the positive signal observed in both conventional and negative control analyses reflect residual familial confounding, rather than a true pharmacologic effect of prenatal paracetamol exposure,” the researchers concluded.

The finding of no association between acetaminophen use in pregnancy and neurodevelopmental conditions in children was also found in large sibling-matched studies in Sweden in 2024 and Japan in 2025.