Hong Kong Sibling Study in JAMA Internal Medicine Finds No Prenatal Acetaminophen Link to Autism or ADHD

The study tracked prescription records across trimesters and dosages while following children up to age 23. Sibling comparisons, which control for shared genetics and household environment, eliminated the modest elevations seen in conventional cohort analyses. Pre-pregnancy and postpartum maternal use produced similar small signals, pointing to familial confounding rather than direct drug effects. Regulators including the FDA and WHO have long noted that prior observational links lacked causal proof.

Earlier reviews in journals such as JAMA Pediatrics and BMJ identified indication bias and unmeasured familial factors as likely explanations for reported associations. This design directly tests those concerns with objective pharmacy data and long-term diagnostic outcomes. The null result aligns with patterns in other sibling-controlled Scandinavian and U.S. cohorts that also found associations disappear after family matching.

Avoiding indicated pain or fever treatment carries separate risks of maternal morbidity that observational data have linked to adverse birth outcomes. Clinicians therefore retain acetaminophen as first-line when non-drug measures are insufficient. Future work requires prospective biomarker-confirmed exposure cohorts and mechanistic studies of neurodevelopment to close residual uncertainty.

Sibling-matched designs can rule out time-invariant family confounders but cannot eliminate all measurement error in prescription records or transient environmental factors; replication in independent large registries with pharmacy validation is the logical next step.