An umbrella review found that current evidence does not clearly establish a link between maternal paracetamol use during pregnancy and autism spectrum disorder or attention-deficit/hyperactivity disorder in offspring. Published in the BMJ, the review by Jameela Sheikh, PhD, of the University of Liverpool, and colleagues, evaluated 9 systematic reviews covering 40 primary studies and identified methodological limitations that undermine confidence in previously reported associations.
The team registered a prospective protocol (PROSPERO, 26 Sept 2025) and used rapid-review methods to examine studies from Medline, Embase, PsycINFO, Cochrane, grey literature, and Epistemonikos through Sept 2025. The work was undertaken amid renewed public attention following the US president’s September 2025 comments about Tylenol use in pregnancy.
Using the A MeaSurement Tool to Assess Systematic Reviews (AMSTAR 2) criteria, the investigators found confidence in the review findings ranged from low in two reviews to critically low in seven. The corrected covered area reached 23%, indicating very high redundancy in the evidence base.
"Most reviews lacked a previously registered protocol (item 2), did not provide a comprehensive literature search strategy (item 4), and did not present a list of excluded studies with justifications (item 7)," the researchers reported. "None of the systematic reviews used the recommended tool to assess the risk of bias of non-randomised trials (either ROBINS-E or ROBINS-I)."
Meta-analyses that pooled data from conventional cohort studies reported modest associations between paracetamol exposure and attention-deficit/hyperactivity disorder (ADHD) with relative risks between 1.2 to 1.4, with smaller estimates for autism spectrum disorder (ASD). However, seven of the nine reviews explicitly cautioned against inferring causality, citing potential bias and inadequately controlled confounders in the primary studies.
The finding emerged from two studies that employed sibling-controlled designs to adjust for shared familial factors and unmeasured confounding—one from Sweden (Ahlqvist et al) and one from Norway (Gustavson et al). In both, associations observed in whole cohort analyses disappeared or markedly weakened when comparing siblings discordant for prenatal paracetamol exposure.
For ADHD, the Swedish study found only a slight difference between exposed and unexposed siblings, with the risk estimate declining from 1.07 in the general population to 0.98 after sibling adjustment. In the Norwegian study, longer exposure (defined as 29 days or more) was associated with an initial risk estimate of 2.02, which decreased to 1.06 once familial factors were controlled for.
Similar patterns emerged for ASD. In the Swedish cohort, the estimated risk for ever-use dropped from 1.05 in population analysis to 0.98 in sibling comparisons. Dose-response relationships also attenuated with higher paracetamol exposure (at the 75th percentile), declining from an estimated 1.10 to 0.88 after sibling adjustment.
Recent data from a Japanese cohort (Okubo et al, 2025) further corroborated these findings, with risk estimates for ADHD falling from 1.32 in general population analyses to 0.86 after sibling adjustment, and autism associations decreasing from 1.09 to 0.85.
The researchers noted that autism and ADHD demonstrate strong familial aggregation, making control for genetic and stable family factors important. "In the primary studies that appropriately adjusted for familial factors through a sibling controlled design, and for key time varying maternal characteristics and indications for paracetamol use, the observed association between exposure to paracetamol and risk of autism and ADHD in childhood disappeared or attenuated," they wrote.
The researchers explained that sibling fixed-effects analyses inherently account for shared genetic and environmental influences, offering advantages over conventional methods that address only measured confounders. They also noted similar attenuation patterns across other analgesic classes, suggesting the removal of family-level bias rather than negation of true effects.
"Given that alternative classes of drugs for relief of pain and fever, such as non-steroidal anti-inflammatory drugs, are known to adversely affect the fetal vascular system and can cause complications such as oligohydramnios and premature closure of the ductus arteriosus, and considering the harmful effects of pyrexia on pregnancy, women should be advised to take paracetamol when needed to treat pain and pyrexia in pregnancy," the researchers concluded.
The study received no commercial funding. Lead study author Shakila Thangaratinam, MD, of the University of Liverpool, serves as a senior investigator for the National Institute for Health and Care Research.
Source: BMJ
