Is Tylenol During Pregnancy Linked to Autism?

The best available evidence does not support a causal link between Tylenol (acetaminophen) use during pregnancy and autism in children. Earlier observational studies raised alarm by reporting small statistical associations, but when researchers used study designs powerful enough to separate the drug’s effects from the conditions it was being used to treat, the associations evaporated. The story is more nuanced than either reassuring headlines or alarming ones tend to capture, and the details matter if you are pregnant and trying to decide whether to reach for a bottle of Tylenol.

Why the Concern Arose

Over the past decade, multiple observational studies found that children whose mothers used acetaminophen during pregnancy were slightly more likely to receive a diagnosis of autism or ADHD. A large Danish birth cohort study, for instance, found that prenatal acetaminophen use was associated with a roughly 50 percent increased risk of autism accompanied by hyperkinetic (hyperactive) symptoms, though not with other forms of autism. Longer use during pregnancy, beyond 20 weeks, appeared to roughly double that specific risk.¹

These findings attracted media attention and fueled ongoing lawsuits against acetaminophen manufacturers. But “associated with” does not mean “caused by.” Observational studies identify patterns; they cannot confirm that the drug itself was responsible. The women taking acetaminophen were doing so for a reason, usually pain, fever, or infection, and those underlying conditions turned out to be critical confounders that could easily produce a statistical ghost of a link where none actually exists.

What Sibling Studies Revealed

The gold standard for untangling this question, short of a randomized trial that would be unethical to run, is the sibling comparison. These studies compare children born to the same mother, where one pregnancy involved acetaminophen and another did not. Because siblings share genetics, household environment, maternal health patterns, and socioeconomic factors, this design strips away much of the confounding that plagues ordinary observational work.

A 2024 study published in JAMA examined a massive Swedish cohort. Without sibling controls, ever-use of acetaminophen was associated with a marginally increased autism risk (a hazard ratio of about 1.05). But when the analysis shifted to sibling comparisons, the association vanished entirely (hazard ratio 0.98). The authors concluded that associations seen in simpler models were likely explained by familial confounding, meaning something about the family rather than the drug was driving the pattern.²

A sibling-matched study in JAMA Internal Medicine arrived at the same conclusion: no association between prenatal acetaminophen exposure and autism (adjusted hazard ratio 1.00) or ADHD (adjusted hazard ratio 1.01). These null findings held regardless of when during pregnancy the drug was taken, how much was consumed, or whether use was sporadic or persistent.³ A critical review of the full body of evidence concluded that sibling-comparison analyses consistently show attenuation or absence of previously reported associations, suggesting acetaminophen exposure is unlikely to be causative.⁴

Fever as the Real Suspect

If acetaminophen itself is not driving the association, what is? A strong candidate is the condition it is being used to treat: fever, especially when caused by infection during pregnancy.

Data from the CHARGE study found that maternal fever during pregnancy was associated with more than double the odds of autism. But here is the twist: mothers who took fever-reducing medication, including acetaminophen, during those fevers had a substantially lower autism risk than mothers who experienced fever and did not take medication. Fever without treatment carried an odds ratio of about 2.55 for autism, while fever with antipyretic use brought the odds ratio down to about 1.30, a level that was no longer statistically significant.⁵

A separate analysis from the Study to Explore Early Development found that infection accompanied by fever during the second trimester was associated with doubled odds of autism. Over 85 percent of those infections were respiratory.⁶ A systematic review and meta-analysis also specifically examined fever during pregnancy as a risk factor for neurodevelopmental disorders, considering variables like fever severity, duration, and antipyretic use.⁷

This creates an uncomfortable paradox: the very drug being accused of causing autism may actually help reduce the risk when fever is present, by bringing down the dangerously high temperatures that appear to be the real threat to fetal brain development.

Proposed Biological Mechanisms

Even though the strongest human evidence points away from a causal link, researchers have identified several biological pathways through which acetaminophen could theoretically interfere with fetal development. These mechanisms explain why the concern has not simply gone away, even as the epidemiology grows more reassuring.

Acetaminophen crosses the placenta. In laboratory studies using human placental cells, high concentrations of the drug reduced the activity of aromatase, an enzyme critical for producing estradiol. The cells also showed a dose-dependent decrease in estradiol output. Estradiol plays a role in the development of both reproductive organs and the brain during fetal life.⁸ An ex vivo study on first-trimester human fetal ovarian tissue confirmed that acetaminophen targeted certain germ cells and disrupted steroidogenesis.⁹

A broader review in Nature Reviews Endocrinology catalogued additional ways acetaminophen disrupts hormone-dependent processes in experimental settings: inhibiting androgen production, depleting sulfated sex hormones, disturbing immune function, inducing oxidative stress, and indirectly activating the endocannabinoid system.¹⁰ The endocannabinoid pathway is itself an area of active investigation, because acetaminophen produces some of its painkilling effects through this system, and interference with endocannabinoid signaling during brain development has been hypothesized as a potential contributor to autism risk.¹¹ Oxidative stress from acetaminophen metabolism has also been proposed as a mechanism, with some researchers arguing that exposure in early childhood rather than in the womb could be more strongly linked to autism because of immature drug-metabolizing pathways in infants.¹²

The critical caveat with all of these findings is that most come from isolated cells or tissue, often at concentrations higher than what a standard dose produces in a human body. Demonstrating a mechanism in a laboratory dish is very different from demonstrating harm in a living pregnancy. The gap between “this drug can disrupt hormones in vitro” and “this drug causes autism in children” is enormous, and the human epidemiological data currently falls on the side of no causal effect.

What Animal Studies Show

Mouse and rat experiments have tried to bridge the gap between laboratory mechanisms and real-world outcomes. A mouse study found that developmental acetaminophen exposure led to changes in pup vocalizations and reduced activity in adult male offspring, but not females. Gene expression changes in the prefrontal cortex were sex-specific and involved glutathione metabolism, DNA damage pathways, and endocrine and immune system genes.¹³ A rat study found that gestational exposure impaired nest-seeking behavior and altered responses to certain stimuli, leading the authors to characterize acetaminophen as a potential developmental neurotoxicant.¹⁴

These findings are genuine, but they carry major translation problems. Rodent studies often use doses scaled differently than human therapeutic doses, and rodent brain development follows a different timeline. Animal work can generate hypotheses and flag areas for further study, but it does not tell you what happens in human pregnancies at normal doses. The human sibling studies, with their far more direct relevance, remain the strongest evidence base for answering the question.

Duration and Dose Patterns

One consistent thread in the observational literature is that associations tend to look stronger with longer or heavier acetaminophen use. An umbrella review of systematic reviews published in the BMJ found stronger associations with longer duration or higher frequency of use and with exposure during the third trimester.¹⁵ A meta-regression analysis similarly found that the association between prenatal exposure and ADHD increased with the mean duration of exposure.¹⁶

Whether this reflects a genuine dose-response relationship or simply more opportunity for confounding is unclear. Women who take more acetaminophen tend to have more severe or chronic conditions. The sibling studies that found no association were designed to address exactly this kind of confounding, and they found null results even when examining cumulative dose and usage patterns.¹⁷ This suggests the dose-response pattern in standard analyses may be another artifact of confounding rather than evidence of a biological threshold.

How Common Acetaminophen Use Is During Pregnancy

Acetaminophen is currently the only pain and fever reducer generally recommended for use during pregnancy. Other common analgesics like ibuprofen carry risks, particularly after the first trimester. Professional groups including the American College of Obstetricians and Gynecologists have reassured patients that acetaminophen is safe. More than 60 percent of pregnant women use acetaminophen at some point, typically for headaches, other pain, or fever, and about 20 percent use it for more than 20 days over the course of their pregnancy.¹⁸

Those numbers frame the question in an important way. If acetaminophen truly caused autism at any meaningful rate, you would expect a very strong correlation between a drug used by the majority of pregnant women and autism rates. The roughly 2 percent prevalence of autism coexisting with more than 60 percent acetaminophen exposure during pregnancy is difficult to square with a meaningful causal role for the drug. This does not rule out a tiny effect in a genetically susceptible subgroup, but it argues against the kind of broad causation that public fears have imagined.

The Political Dimension

The question of whether Tylenol causes autism has moved beyond scientific journals and into political discourse. In 2025, Donald Trump publicly advised pregnant women not to take paracetamol, claiming that its use in pregnancy is linked to autism in children.¹⁹ This statement was made without the caveats researchers attach to the observational data and did not reflect the findings of the sibling-controlled studies.

Thousands of lawsuits have also been filed against acetaminophen manufacturers, alleging that the drug caused neurodevelopmental disorders. These lawsuits have amplified public anxiety, though legal standards for evidence differ from scientific ones. The existence of lawsuits does not mean the scientific case has been established. A review published in the International Journal of Gynecology and Obstetrics directly addressed this concern, concluding that the established risks of untreated pain and fever during pregnancy significantly outweigh theoretical concerns based on methodologically limited studies.²⁰ Biomarker studies in selected cohorts have reported higher odds of autism across exposure levels, but generalizability and residual confounding remain concerns.²¹

What This Means If You Are Pregnant

For anyone reading this while pregnant and wondering what to do, the practical picture is more reassuring than the headlines suggest. The strongest evidence available says that acetaminophen at standard doses for limited periods does not cause autism. Untreated fever, on the other hand, does appear to carry real risks for fetal brain development.²²

The general guidance from medical professionals has been to use the lowest effective dose for the shortest necessary duration and not to take it preemptively. If you have a fever or significant pain, treating it is medically reasonable. If you have mild discomfort you can manage without medication, skipping the pill is also reasonable. What you should not do is refuse to treat a high fever during pregnancy out of fear of autism. The data specifically suggest that fever without treatment carries a higher autism risk than fever with treatment.

Why This Question Keeps Resurfacing

Even as sibling-controlled studies and critical reviews accumulate, this topic refuses to settle down. Part of the reason is structural: autism diagnoses have risen substantially over recent decades, and acetaminophen use has been widespread during the same period. When two trends run in parallel, people naturally look for a connection. The biological plausibility from lab and animal studies keeps the hypothesis alive in scientific circles, even as the human epidemiological evidence grows more robust against a causal interpretation.

The other part is emotional. Autism is a condition that affects families profoundly, and the desire to find a preventable cause is powerful. Acetaminophen, as a cheap and ubiquitous over-the-counter drug taken during a vulnerable period, makes an intuitive target. Researchers will continue refining their methods, and future studies may identify a narrow subpopulation with heightened genetic susceptibility to acetaminophen’s endocrine effects. But for the broad population of pregnant women, the current weight of evidence lands firmly on one side: acetaminophen use during pregnancy does not appear to cause autism, and avoiding it when you genuinely need it may create more risk than it removes.