Prenatal acetaminophen may be linked to childhood ADHD diagnosis, more review needed

The main strength of the study is objective measurement of exposure in the form of a maternal blood test during the second trimester, which was examined for acetaminophen and metabolites.

However, it comes with severe weaknesses, which compromises the clinical inferences. One weakness is this is a small and highly selected cohort within an original highly selected population. Consequently, the findings inherently come with substantial selection bias and poor external validity.

Second, the study has a low number of outcomes, resulting in poor precision of the estimates of association. Third, the authors adjusted for several covariates and confounders, but they were unable to adjust for parental ADHD, maternal fever and migraine, all of which are essential. They did use prenatal nonsteroid anti-inflammatory drugs and antibiotics use as a proxy for fever. This is unlikely to be sufficient. A pregnant woman with fever is substantially more likely to take acetaminophen for a few days than be prescribed an antibiotic.

Fourth, outcomes were assessed by parental reporting of a diagnosis of ADHD or parental reported use of ADHD-specific drugs and by several questionnaire-based tests. There does not seem to be a true physician-based diagnosis reported. ADHD is a complex and heterogeneous disorder, and a valid diagnosis requires a trained physician-based assessment according to the American Academy of Pediatrics. Finally, while measuring acetaminophen and metabolites during the second trimester is an objective exposure validation, it is also a random single time-point sample that does not allow for assessing extent or duration of exposure to substantiate the associations reported.

The issue of potential adverse childhood neurodevelopment following in utero exposure to acetaminophen, especially ADHD and autism, has been a decade-long controversy in academia. The controversy was amplified in 2021 following a paper published in Nature (Bauer AZ, et al. Nat Rev Endocrinol. 2021;doi:10.1038/s41574-021-00553-7), which stated serious concerns with respect to childhood neurodevelopment as well as potential urogenital/reproductive issues. The conclusions of this paper have been disputed, as many of the underlying studies suffer from selection bias, insufficient control of confounding factors, especially heritability of ADHD and autism, and poorly validated outcomes. A recent in-depth assessment of the most important studies concludes that clinically meaningful association between pregnancy exposure to acetaminophen and childhood ADHD/autism is unlikely and unsubstantiated by the underlying data (Damkier P, et al. Obstet Gynecol. 2025;doi:10.1097/AOG.0000000000005802).

The authors have made strong efforts to substantiate mechanistic plausibility of their findings through detailed genomic analysis of genes and variants expressed in the placenta. Such molecular and genomic approach, in general, may contribute to our understanding of human observations. However, it serves little purpose when such human observation is not reasonably well established. The proposed mechanistic links to ADHD through up- and downregulation of certain genes in the placenta are purely speculative, and the attempt by the authors to attribute proportions of a suggested causal inference to these explorative findings is neither meaningful nor scientifically justifiable. The etiology of ADHD is, but for a strong degree of heritability, not established.

In summary, the contributions from this study do not move the overall body of human evidence. It is of no consequence to clinical practice, which recommends acetaminophen as first-line treatment for pain and fever in pregnancy.

Per Damkier, MD, PhD

Professor

Department of clinical research

University of Southern Denmark

Disclosures: Damkier reports no relevant financial disclosures.