Kang Muslim
For decades, millions of expectant mothers worldwide have relied on acetaminophen—universally known as paracetamol and frequently purchased under brand names like Tylenol—as the standard first-line defense against pain and fevers. Because alternative analgesics such as ibuprofen or aspirin carry documented risks of birth defects during gestation, acetaminophen became the medical community’s default recommendation. However, this foundational pillar of prenatal care was severely shaken over the past ten years by a cascade of observational studies, conflicting regulatory policies, and sensationalized headlines suggesting a link between the common pain reliever and neurodevelopmental conditions such as autism spectrum disorder (ASD) and attention-deficit/hyperactivity disorder (ADHD).
A comprehensive new umbrella review published in the journal Developmental Medicine & Child Neurology aims to cut through the noise. Led by researcher Hoda Tayebi-Hillali of the University of Santiago de Compostela in Spain, the research team conducted a rigorous re-evaluation of existing literature. Their findings offer substantial reassurance to expectant parents: when methodologically weaker studies and high-risk biases are systematically set aside, there is little evidence that prenatal acetaminophen use causes autism or ADHD in children. While small statistical associations can still be found in lower-quality datasets, the researchers conclude that these numbers largely reflect confounding variables, genetic backgrounds, and methodological flaws rather than a direct pharmacological hazard.
The Chronology of a Medical and Public Controversy
To understand the weight of Tayebi-Hillali’s new umbrella review, it is necessary to trace how the scientific and public debate evolved over the course of a decade. The modern phase of the controversy began taking shape in the late 2010s, driven largely by observational studies that relied on retrospective maternal questionnaires.
In 2019, a prominent meta-analysis pooled early observational research and concluded that prenatal exposure to acetaminophen was associated with higher rates of ADHD in children. Although the researchers behind these early papers urged caution and noted limitations, the findings quickly entered the public sphere. Regulatory bodies and medical professionals found themselves divided, debating whether the documented risk was a genuine physiological threat or the artifact of flawed research methodologies.
The scientific pushback began gaining momentum shortly thereafter. In 2021, a sibling-comparison study altered the conversation by comparing brothers and sisters within the same households. Because siblings share a vast majority of their genetic background and home environment, sibling-matched designs help control for unmeasured familial factors. This study suggested that shared family background and genetics—rather than the maternal consumption of acetaminophen—best explained the apparent statistical link to ADHD.
The sibling-control hypothesis received massive empirical backing in 2024 through a sweeping, population-wide study of nearly 2.5 million Swedish children. When researchers evaluated sibling cohorts, taking acetaminophen during pregnancy demonstrated no link to an increased risk of autism or ADHD.
Despite this emerging scientific consensus favoring safety, the debate abruptly entered the geopolitical arena in September 2025. U.S. President Donald Trump publicly advised against acetaminophen use during pregnancy, citing potential autism risks. Prompted by the political climate and ongoing public scrutiny, the U.S. Food and Drug Administration (FDA) initiated steps to update the drug’s product labeling to note a potential association.
The regulatory reaction was swift and starkly contradictory internationally. Just one day after the FDA’s announcement, the European Medicines Agency (EMA) publicly stated that it had found no credible scientific evidence indicating that acetaminophen causes autism—a position promptly echoed by health authorities in the United Kingdom and Australia. This jarring international divergence left expectant mothers and healthcare providers caught in a crossfire of confusing public health messaging.
Deconstructing the Evidence: The Methodology of the New Umbrella Review
Faced with mounting confusion, Tayebi-Hillali and her colleagues set out to conduct the most rigorous synthesis of the literature to date. Recognizing that previous systematic reviews often suffered from overlapping datasets—frequently counting the exact same primary studies multiple times under different iterations—the team designed an umbrella review.
An umbrella review is a high-level statistical analysis that gathers, appraises, and synthesizes data from multiple previously published meta-analyses. Searching six major scientific databases up to November 2025, the research team identified seven meta-analyses encompassing more than three million participants. Three of these reviews examined autism, and all seven examined ADHD.
Initially, the researchers performed an unrestricted pooling of all available data, regardless of the quality of the individual meta-analyses. In this broad sweep, they identified small positive associations: the pooled risk ratio was 1.17 for autism and 1.29 for ADHD. These numbers suggested that children exposed to the drug prenatally had an estimated 17 percent and 29 percent higher relative risk of diagnosis, respectively, compared to unexposed peers.
However, when the team rated the methodological quality of the seven meta-analyses using a standardized appraisal checklist, a critical pattern emerged. Most of the existing reviews were categorized as low quality. Common shortcomings included failing to quantify the precise dose of acetaminophen ingested, omitting lists of excluded studies, and ignoring the overlap of primary source data.
When the researchers dropped the weakest reviews, the results barely budged (1.17 for autism and 1.28 for ADHD). Only a single meta-analysis, published in 2025, earned a high-confidence rating. That specific high-quality review reported much smaller risk ratios: 1.10 for autism—a result that lacked statistical significance—and 1.16 for ADHD.
Going Beyond the Meta-Analyses: Assessing Primary Study Bias
Because the existing meta-analyses relied on heavily overlapping pools of primary research, Tayebi-Hillali’s team took their investigation a step further. They bypassed the intermediary reviews entirely and conducted a direct analysis of the underlying primary studies, evaluating each for risk of bias using standardized tools that accounted for confounding factors, exposure measurement accuracy, and participant selection criteria.
When the analysis was strictly restricted to primary studies exhibiting a low risk of bias—such as large-scale population studies from Sweden and Japan—the risk ratios dropped dramatically. The risk ratio for autism fell to 1.05 based on two major studies, while the ratio for ADHD fell to 1.17 based on three studies. Both estimates hovered marginally above 1.0, indicating that a very minor statistical association persisted, though the underlying ADHD studies varied considerably in their individual conclusions.
“One of the most striking findings was how much the estimates changed once we took the quality of the evidence into account,” Tayebi-Hillali explained in an interview. “Associations that initially appeared consistent were substantially attenuated when we excluded lower-quality reviews and primary studies at higher risk of bias. It was also particularly telling that some sibling-comparison studies, which help control for the genetic and environmental factors shared within a family, did not find the associations originally reported.”
Understanding Absolute Risk vs. Relative Risk
To evaluate what these remaining small risk ratios mean for everyday clinical practice, public health experts emphasize the distinction between relative risk and absolute risk.
Given the established global baseline prevalence of autism—affecting roughly 1 in 117 children—and ADHD—affecting roughly 1 in 60 children—any absolute increase in risk resulting from the drug would be exceptionally modest, even if the remaining association were proven to be genuine.
“Even taking the figures at face value, the impact would be very small,” Tayebi-Hillali noted. “Moreover, in observational studies, a risk ratio this small can readily be explained by bias, without the drug playing any role at all. The size of the effect, therefore, does not justify alarm.”
The problem of residual statistical noise in observational research is frequently driven by a phenomenon known in epidemiology as “confounding by indication.” Pregnant women almost exclusively take acetaminophen to treat underlying health issues, such as high fevers, infections, severe pain, or systemic inflammation. Extensive medical research demonstrates that severe maternal fevers and untreated infections during gestation can independently disrupt fetal brain development. This creates an analytical hurdle for researchers: isolating the biochemical impact of the medication from the physiological impact of the illness it was administered to treat is exceedingly difficult.
Furthermore, many primary studies relied on maternal self-reported questionnaires administered months or years after pregnancy. Human memory is imperfect, and retrospective self-reporting frequently leads to misclassification errors, with participants struggling to accurately recall the exact dosage, frequency, timing, or duration of over-the-counter drug use.
Contrasting Perspectives within the Scientific Community
While the umbrella review aligns closely with recent sibling-matched investigations, it directly challenges other notable papers in the field.
For instance, the findings stand in tension with a 2025 review that concluded associations were strongest in highly rigorous studies—particularly those utilizing biological samples to measure drug metabolites. Tayebi-Hillali’s team, however, regarded biomarker measurements taken from umbilical cord blood, meconium, or maternal blood as carrying uncertain clinical validity regarding long-term neurodevelopment.
Similarly, the new review diverges from older 2016 and 2022 studies that linked prenatal acetaminophen to early behavioral and attention problems. Those studies relied primarily on parent-reported behavioral symptom checklists rather than formal clinical diagnoses. While those papers were included in the primary literature pool, none satisfied the strict risk-of-bias thresholds implemented in the new umbrella review.
Clinical Guidance and Recommendations for Expectant Mothers
As the scientific community continues to refine its understanding of prenatal pharmacology, researchers emphasize the importance of balanced communication. Misinterpretations can cut both ways: it is scientifically incorrect to claim the umbrella review proves paracetamol is entirely free of risk under every possible dosage scenario, but it is equally inaccurate to present observational associations as proven causal pathways.
“The main misinterpretation to avoid is the claim that our study proves paracetamol causes autism or ADHD; equally, it would be wrong to conclude that we have shown any exposure, regardless of dose or duration, to be entirely free of risk,” Tayebi-Hillali cautioned. “Our work is a synthesis of observational studies and therefore cannot establish causality on its own.”
Medical authorities universally agree that managing acute pain and reducing high fevers during pregnancy is vital for preventing maternal and fetal complications. Untreated hyperthermia (high fever) during the first trimester is a well-established risk factor for neural tube defects and other congenital abnormalities. Consequently, leading obstetric organizations maintain that acetaminophen remains the most appropriate first-line treatment when medically necessary, provided it is used at the lowest effective dose for the shortest possible duration.
“Headlines that present an association as if it were proven causation can convey an alarm that the evidence does not warrant,” Tayebi-Hillali concluded. “At the same time, it is essential to acknowledge uncertainty and to keep investigating. We should also remember that untreated fever and pain during pregnancy can have consequences, so avoiding a clinically necessary medication without consulting a healthcare professional is not a risk-free decision either.”
Looking forward, the authors emphasize that future research must move beyond broad observational metrics. Progress will depend on prospective studies that capture precise data regarding dosage, frequency, duration, gestational timing, and the specific clinical indications prompting medication use, all while deploying advanced sibling-comparison methodologies to eliminate familial confounding. Until then, comprehensive umbrella evidence indicates that mothers can continue to use acetaminophen safely when clinical needs dictate, guided by the advice of their healthcare providers rather than alarming media narratives.
