Is it true that Tylenol use during pregnancy is linked to Autism in the newborn?

Venkat Rao

The recent news on the possible association between the use of Tylenol, the brand name for Acetaminophen and other comparable brandname products by pregnant women, with increased risk of Autism and Attention Deficient Hyperactivity Disorder (ADHD) in children, was met with an avalanche of rebuttals and knee-jerk factchecking by media outlets. The conversation turned—what should have been a serious national discourse on public health risks posed by widely marketed and consumed drugs and food products—into a political theater by the usual cast of television and media reporters almost totally and completely refuting such associative evidence, even if one such existed. This was undertaken without even considering the basic scientific data and weight-of-evidence based approach that was used the Health and Human Services in making such a determination.

Even a highly respected journal such as Nature sensationalized the FDA policy with a title, “Trump links autism and Tylenol” Ironically, the same Nature article says reported association between paracetamol use by pregnant women and autism is exceedingly small clearly implying such a link could not be totally ruled out. Would a pregnant women accept even a “small risk,” however small it might be, when informed of the potential, low-level risk of autism in her expectant child, as an acceptable risk?

The study did not unequivocally conclude there is no association whatsoever between use of Tylenol during pregnancy and autism. The Nature report is accurate that Paracetamol, the active analgesic ingredient in Tylenol, has a long history of use by pregnant women with an established safety record. Also, very few alternatives to paracetamol exists that are safe and effective as painkillers for use by pregnant women. Nevertheless, is it safe from a public health policy standpoint to accept the small potential risk of Tylenol use-associated autism? Is such a public health policy considered sound if it overrides a low-level risk, and not include a label claim that such a risk may exist?

The Harvard School of Public Health published a report in the summer of 2025 on the plausible link between fetus exposed to acetaminophen—also known by the brand name Tylenol or as paracetamol—by the expectant mother during pregnancy, may more likely result in neurodevelopmental disorders (NDDs) including autism and attention-deficit/hyperactivity disorder (ADHD), according to a new study published by the dean of the faculty at the Harvard School of Public Health.

The reported study is a summary analysis of 46 previous research investigations worldwide on the potential link between acetaminophen use by pregnant women and pediatric neurodevelopmental disabilities (NDD) in children. A rigorous and comprehensive meta-analysis of published research literature using the Navigation Guide Systematic Review—the most widely accepted analytical framework for analysis of multiple research publications, indicated association between acetaminophen (Tylenol) exposure during pregnancy and increase incidence of NDD.

An earlier prospective cohort study of 1,011 women who reported taking Tylenol during pregnancy, children who were exposed to the painkiller in the uterus scored significantly higher on 3 of the 7 Child Behavior Checklist (CBCL) syndrome scale, reporting withdrawal, sleep problems, and attention deficiency. All scores reported were significantly associated with prenatal stress, with two scores on sleep and attention problem significantly higher among children with maternal exposure to Tylenol during pregnancy.

There are other studies as well on prenatal stress and behavioral disorders when pregnant women use of acetaminophen during pregnancy resulting in behavioral problems among children, with researchers interpreting the mechanism of action as the outcome of immune system dysfunction and increased action in the hypothalamic-pituitary-adrenal axis. Going by this mechanism of action pregnant women experiencing stress such as pain results in the over activation of the axis, which in prolonged or severe cases, could lead to increased cortisol level and exposure to the fetus. Extreme stress during pregnancy is not widespread, but not uncommon either, necessitating use of painkillers or other similar medical countermeasures.

The U.S. Food and Drug Administration  initiated the process for a label change for acetaminophen (Tylenol and comparable products) to reflect evidence suggesting that the use of acetaminophen by pregnant women may be associated with an increased risk of neurological conditions such as autism and ADHD in children.

An activist section of the science policy influencers ruled out any whatsoever associative relationship, rejecting FDA notice to modify the label on acetaminophen containers to include a safety note on the label of Tylenol would be highly dangerous.

FDA did not ban the use of acetaminophen during pregnancy or imposed any restrictions to its use. The new measure requires a notation on the label of the container that such a risk may exist, and leaves the decision to take the medicine in the hands of the consumer.  Since acetaminophen is a over-the-counter medication needing no prescription, the informed decision to use the painkiller is left to the consumer as it should be. 

The counterfactual argument that even a precautionary note on the label may increase the “fear in pregnant women at a time when they truly don’t need to,” and it is absolutely not supported by published scientific evidence is how this interpretation has gone.

I would like to draw a similar parallel policy standpoint from the Environmental Protection Agency under the Superfund program, aimed at reducing or eliminating the risk of cancer and other diseases posed by toxic chemicals from hazardous dump sites. Using a cancer risk extrapolation model, an excess potential cancer risk of one in a million people living close to the dump site and exposed to the cancer-causing chemical as the cut-off level. Any projected risk in excess of one in a million-cancer incident would require remediation of the site and its restoration to acceptable risk levels.

Now compare this to another policy setting approach used by US Environmenal Protection Agency to assess threat of expsoure to hazardous chemicals from Superfund landfill sites. Uing a hypothetical exposure and statistical extrapolation of cancer incidence, the science-based policy sets a safe and acceptable level of risk at in excess of one in million cancers. This is no matter whether cancer incidence has increased or not in the community living close to a dump site. EPA policy makers consider excess of one case in a million, as sufficiently protective, without any causal or associative evidence to back up. Hazardous chemicals from a dump site pose serious risks, including risk of getting cancer. The comparison is merely based on the science-based policy setting rationale and approach.

Let us compare this to autism. According to the US Health and Human Services, autism prevalence has increased in the US to 1 in 31 children and 1 in 45 adults with Autism Spectrum Disorder (ASD) and approximately 1 in 44 children in the U.S. is diagnosed with an autism spectrum disorder annually. Out of these 44 children diagnosed, approximately 1 in 27 boys are identified with autism, and 1 in 116 girls identified with autism (Autism Speaks, 2022). Over the next decade, an estimated 707,000 to 1,116,000 teens (70,700 to 111,600 each year) will enter adulthood and age out of school-based autism services.

Let us pause for a moment. A hypothetical cancer risk based on hypothetical exposure to hazardous chemicals from a waste site resulting in an estimated excess cancer incidence of one in a million is the governing policy of EPA. Anything above this estimate requires site remediation and restoration. The one in a million excess lifetime cancer risk primarily based on hypothetical exposure and risk scenarios for cancer from toxic chemicals from a Superfund site as the acceptable level and anything above requires site remediation. Whereas a small, but notable association of  paracetamol (Tylenol) linked autism—now with an increasing prevalence of the number of cases in the US population is questionable public health policy under the same premise of science-based policy setting.

Given that there is a notable body of evidence linking use of acetaminophen during pregnancy and autism in the newborn children; given that we are witnessing a notable surge in the prevalence of autism in recent years with approximately 1 in 31 children exhibiting some form of autism and related developmental disabilities, a public health policy protective of children with a label change is a prudent measure leaving the ultimate decision to take the medicine in the hands of the patient-consumers.

National discourse on topics relating to public health is greatly aggravated since the COVID-19 pandemic, when other-than-scientific biases tilted/distorted key facts on the pandemic and the needed public health response. Scientific analysis and reporting should be rigorously guarded from other influences, for it will compromise the very basis of science-based policy making framework.