Amid heated public debate, the most robust evidence to date shows that using paracetamol as directed during pregnancy is not linked to adverse neurodevelopmental outcomes, reinforcing current clinical guidance.
Study: Prenatal paracetamol exposure and child neurodevelopment: a systematic review and meta-analysis. Image credit: SNeG17/Shutterstock.com
Paracetamol is the leading pain-relieving and fever-lowering drug used in pregnancy. With emerging claims that it is dangerous in pregnancy due to its adverse impact on fetal neurodevelopment, researchers considered it essential to synthesize existing evidence to examine this hypothesis. The results were recently published in the journal The Lancet Obstetrics, Gynaecology & Women’s Health.
Paracetamol safety in pregnancy sparks renewed controversy
Paracetamol is recommended and prescribed to treat pain and fever in pregnancy worldwide. Its safety profile is much better than that of non-steroidal anti-inflammatory drugs (NSAIDs) like aspirin or ibuprofen, and of opioids.
Recent announcements and public statements by US political leaders in September 2025 aim to discourage paracetamol use in pregnancy, based on its apparent link to autism risk. This was sourced from a review that suffers from heterogeneous data and uses atypical definitions for both exposure and outcomes.
If valid, this could flag a major public health concern since millions of pregnant women use the drug. Conversely, unfounded fears about paracetamol use in pregnancy could leave pregnant women and their babies open to the risks associated with untreated pain and fever, which include miscarriage and preterm birth.
The American College of Obstetricians and Gynecologists (USA) and the Royal College of Obstetricians and Gynaecologists (UK), along with other professional bodies such as FIGO and SMFM, and regulators including the EMA, MHRA, and Health Canada, are among the organizations that continue to recommend the use of paracetamol in keeping with current prescribing standards.
Past reviews of the literature in this area have produced conflicting findings due to exposure misclassification, confounding factors, and biases in the studies used. For instance, with prognostic-factor bias, differences in baseline attributes between the comparison groups affect the outcomes irrespective of the exposure of interest.
Several rigorous, large studies, such as a Swedish study covering 2.48 million births, have found null associations after sibling-comparison analyses that account for shared familial and genetic confounding. The current study aimed to provide a stringent evaluation of multiple neurodevelopmental outcomes following prenatal exposure to paracetamol.
Filtered evidence minimizes confounding
The paper reviewed 43 studies that explored associations between paracetamol use during pregnancy and neurodevelopmental disorders like autism spectrum disorder (ASD), attention deficit hyperactivity disorder (ADHD), and intellectual disability, with risk estimates. Only cohort studies with clearly defined outcomes and adjusted risk estimates were included, while studies reporting only unadjusted results were excluded.
Sibling comparison studies were used to obtain primary evidence, as they compensated for shared familial and genetic factors. However, time-lapse errors and residual confounding remain potential sources of error.
While all 43 studies were included in the systematic review, only 17 were eligible for the meta-analysis. The latter attempted to stratify risk by trimester of paracetamol exposure and duration of use, sex of the baby, and the duration of follow-up, but this proved impossible within sibling-comparison and low-risk-of-bias subsets, because too few studies reported stratified data with sufficient statistical power to support stable subgroup estimates.
Meta-analyses were based on a generic inverse-variance random-effects model to account for differences in study populations, intervention types, measured outcomes, and study designs. This allowed for the assessment of the pooled estimate from the distribution of true effects across the various studies. The odds ratios were calculated for each outcome.
Robust sibling analyses show no neurodevelopmental risk
The sibling comparison studies revealed no risk of ASD, ADHD, or intellectual disability with paracetamol exposure in pregnancy. This persisted as null when only low-bias studies were analyzed. When only studies with more than 5 years of longitudinal data or only studies with adjusted risk estimates were analyzed, they continued to show no associations.
By contrast, some conventional observational analyses showed small elevations in risk estimates, which were heterogeneous and attenuated in more robust analyses, suggesting residual confounding. However, rare or delayed events may still have been missed.
The authors point out that this study is unique as the first systematic review and meta-analysis to focus on sibling-comparison studies and to ensure that each included study was assessed for prognostic-factor bias using the Quality in Prognosis Studies (QUIPS) tool. However, sibling studies are less generalizable and may be affected by time-lapse effects on the mother’s health across pregnancies.
The study presents the results of a comprehensive search of major databases and examines multiple outcomes, thereby strengthening the conclusions. Its findings clearly indicate that “prenatal paracetamol exposure was not associated with increased risks of autism spectrum disorder, ADHD, or intellectual disability.” This remains true regardless of longer follow-up and with low-bias studies.
Reassuringly, these conclusions are supported by earlier large registry-based sibling-comparison studies that fail to show an association. These studies have the greatest homogeneity and the lowest chances of bias. The current study also aligns with the recommendations of well-regarded professional organizations such as the ACOG and the RCOG.
Apparent associations reported in prior observational studies are likely due to unaccounted confounding factors, such as maternal illness, fever, genetic predisposition, and environmental factors. Any of these increases the risk of analgesic use and of neurodevelopmental abnormality.
In particular, observational studies based on subjective measures are subject to recall bias. For instance, mothers with neurodivergent children may be more likely later to incorrectly remember or overreport their use of medication, which may not always have been paracetamol.
The limitations of this systematic review include the predominance of non-sibling studies, which increases the risk of confounding, variability in exposure and outcome definitions, differences in how confounders were adjusted for, and variable assessment of neurodevelopmental status, which is inevitable in this rapidly changing area of practice and research.
The authors also note that some studies measured autistic traits using screening questionnaires rather than formal diagnostic criteria, contributing to heterogeneity. Data were also insufficient to draw reliable conclusions about sex-specific effects. Notably, only 17 of 43 studies were suitable for meta-analysis, limiting the accuracy of its estimates.
The long-term use of paracetamol in pregnancy is rare. In such a scenario, the neurodevelopmental risk may be traceable to the underlying health condition rather than to the drug, requiring further dose-response studies.
There is no human evidence establishing a causal link between paracetamol exposure and adverse neurodevelopmental outcomes when the drug is used as recommended. However, preclinical studies indicate endocrine disturbances and oxidative stress caused by paracetamol metabolites, alterations in prostaglandin pathways, and epigenetic changes. It is possible that the immature fetal liver puts the fetus at greater risk of paracetamol toxicity via these pathways, especially as paracetamol freely crosses the placenta.
Avoiding paracetamol may pose greater maternal, fetal risk
Current evidence does not indicate a clinically important increase in the likelihood of autism spectrum disorder, ADHD, or intellectual disability in children of pregnant individuals who use paracetamol as directed, supporting existing recommendations on its safety. Avoidance of paracetamol based on inconclusive evidence could expose pregnant women and their babies to known risks associated with untreated fever or severe pain.
Further work should enhance the accuracy and credibility of these conclusions by using more accurate, for instance, biomarker-based, exposure measures, standardized outcome definitions, and modifying the study design to incorporate the underlying mechanisms and baseline familial risk. Involving the ASD community is also key to future research, ensuring that it reflects their goals.
Download your PDF copy now!
Research shows racial-ethnic differences in the risk of pregnancy-related high blood pressure