APAP & ASD

Medicine can cause risks and benefits- creating this conflicted relationship we have with the medical world.

I did not think that I would be writing this blog today, but I am loving the discussion! Neurodiversity and psychiatric conditions remain murky at best so open and honest dialogue will help to ensure that we all stay up to date on the “new” information.  Multiple observational studies have reported associations between prenatal (and less consistently, early-life) acetaminophen exposure and neurodevelopmental outcomes including autism and ADHD. However, the literature is heterogeneous, and rigorous sibling-controlled and other designs that better address familial confounding produce weaker or null associations, so causality is not established. Biological mechanisms have been proposed (oxidative stress, glutathione depletion, effects on the endocannabinoid system, immune modulation) and are supported by animal and molecular work, but translation to human risk at typical therapeutic doses remains uncertain. Regulators and scientific bodies are actively reviewing the evidence; clinicians should balance the known risks of untreated fever/pain in pregnancy against uncertain associations and counsel patients accordingly. This is the summary of what we know to date, which will continue to change and develop over time.

Overview of the legal action

• In 2022, a wave of lawsuits was filed by parents alleging that manufacturers and retailers of acetaminophen failed to warn that prenatal use could increase risk of autism spectrum disorder (ASD) or ADHD in children.

• These suits typically name brand-name manufacturers (e.g. Johnson & Johnson for Tylenol) and generic/OTC acetaminophen suppliers, as well as retailers such as Walmart, Walgreens, Target, etc.

• The lawsuits include several causes of action, including: failure to warn, negligence, breach of express or implied warranty, misrepresentation, and violation of consumer protection laws.

• The plaintiffs sought to consolidate these into a multidistrict litigation (MDL): In Re: Acetaminophen - ASD/ADHD Products Liability Litigation (MDL No. 3043).

Key court rulings and procedural status

• Summary judgment and expert testimony issues: In December 2023, the MDL judge, Judge Denise Cote in the Southern District of New York, excluded the plaintiffs’ expert witnesses on general causation grounds—i.e., the experts were judged not to meet required standards for establishing that acetaminophen can cause autism or ADHD. Based largely on that exclusion, summary judgment was granted in favor of the defendants, effectively dismissing the MDL litigation at the federal level.

• Appeals: The plaintiffs have appealed the decision to grant summary judgment, seeking reversal of the exclusion of their expert testimony and a reopening of the litigation. As of mid-2025, no final decision from the appellate court had been issued. 

• State court activity: While much of the action has been in federal MDL court, plaintiffs’ attorneys are also exploring state-court claims, where evidentiary standards (for example, admissibility of expert testimony) and procedural rules may differ. That may offer alternative paths for claims that were dismissed in federal court.

• Ruling on retailer preemption: One notable ruling during the MDL was a decision rejecting the “preemption” argument made by some defendants. Preemption, in this context, is a legal doctrine under which compliance with federal law (FDA regulations) can sometimes shield a manufacturer or retailer from state-law liability. Judge Cote ruled that manufacturers and retailers could not rely on preemption to dismiss the cases (at least in certain contexts) because the plaintiffs allege failure to warn defects under state law despite federal regulation.

How the legal rulings reflect scientific and evidentiary hurdles

The legal decisions illustrate how courts currently evaluate the available science and what is required to proceed with product-liability claims. Key evidentiary points include:

• General causation requirement: To succeed in these claims, plaintiffs must show it is scientifically plausible that acetaminophen can cause the condition (autism/ADHD) under the conditions alleged (dose, timing, etc.). Courts require reliable, admissible expert testimony using accepted scientific methods. Mere statistical association in observational studies is often deemed insufficient. In this case, judges found the expert testimony offered did not meet standards of methodology, e.g. in controlling confounding, consistency, or peer support.

• Daubert standard: The legal standard for admitting expert testimony in federal courts (Daubert v. Merrell Dow Pharmaceuticals) requires that testimony be based on methods scientifically valid and reliably applied. Issues such as selection bias, cherry-picking studies, insufficiently robust methodology are grounds for exclusion. The controversies over the plaintiffs’ experts were largely about these methodological concerns.

• Causation vs. correlation: Many of the scientific studies cited in lawsuits are observational, and courts are sensitive to arguments that the higher rates of ASD/ADHD in children whose mothers report acetaminophen use may be due to underlying conditions (e.g. fever/infection, pain) rather than the drug per se. Courts have to assess whether the scientific literature addresses these confounders adequately. As we saw earlier, sibling studies and other designs reduce, but do not eliminate, evidence of association.

• Statistical strength and consistency: Courts often look for replicated findings, dose–response relationships, biological plausibility, and consistency across populations. The fact that scientific reviews emphasize heterogeneity, bias, and that causality is not established undermines the plaintiffs’ ability to meet legal standards.

Recent developments & implications

• As of August 2024, a New York federal judge (Judge Cote) ruled that the remaining Tylenol/acetaminophen lawsuits claiming it causes ADHD (and by extension some ASD claims) cannot go forward because of lack of admissible expert testimony linking acetaminophen use prenatally to ADHD/ASD.

• These rulings align with statements by regulatory bodies (FDA, professional societies) that while there is some suggestive evidence, at this time the science does not support a concluded causal link. The legal rulings effectively underscore that current scientific evidence is insufficient under legal standards for causation.

• The outcome of the appellate process could influence whether future claims might succeed, set precedents regarding admissibility of epidemiologic evidence, and potentially affect product labeling or risk communications. If the appeals reverse the exclusion of experts, then cases may proceed; otherwise, many of these federal cases may be permanently dismissed. 

Intersections & tensions between legal standards and scientific uncertainty

• Courts require a high burden of proof, especially for general causation. Scientific uncertainty (e.g. confounding, inconsistent findings, dose/timing issues) often leads to expert testimony being excluded. This is lawful under Daubert, but means that even suggestive epidemiological findings may not suffice for legal liability.

• Scientific plausibility (mechanisms) is helpful but not sufficient; legal standards focus on empirical evidence that a harmful effect does occur in the exposure of concern.

• Legal rulings are not scientific verdicts; dismissal in court does not mean “no risk”—it means that under current evidence, the burden required by law has not been met.

What the epidemiologic studies show

• Numerous cohort and case–control studies (large birth cohorts, administrative datasets, and biomarker studies) have reported statistically significant associations between maternal acetaminophen use during pregnancy and increased risk of offspring neurodevelopmental disorders, including ASD and ADHD. Examples include cohort analyses and meconium biomarker studies that linked prenatal acetaminophen exposure to higher odds of ADHD and altered brain connectivity.

• However — confounding is a major concern. Mothers who take acetaminophen differ from mothers who do not in ways that relate to neurodevelopment (underlying infections, fever, pain, genetic and environmental factors). Some studies that use sibling-control designs—which control for shared familial factors—find no association between prenatal acetaminophen and autism/ADHD, suggesting at least part of the observed association in other studies may be due to confounding. A large sibling-controlled analysis published in JAMA found no increased risk when comparing siblings discordant for exposure. 

• Postnatal exposure evidence is weaker and less consistent. A smaller number of observational and animal studies raise concern about acetaminophen use in infancy/early childhood, but population-level causal evidence is limited and inconsistent.

• Systematic reviews and recent evidence syntheses reach cautious conclusions: several reviews summarize an association across many observational studies but emphasize heterogeneity, bias, and that causality has not been demonstrated. Newer 2024–2025 evaluations have re-examined the literature and debated the strength of evidence.

Plausible biological mechanisms (what could explain a causal link)

Research to date suggests several biologically plausible pathways by which acetaminophen might influence neurodevelopment (most mechanisms supported by animal or molecular studies rather than definitive human proof):

• Oxidative stress / glutathione depletion. Acetaminophen is metabolized into reactive metabolites (e.g., NAPQI) that consume glutathione; immature antioxidant systems in fetuses and infants could increase vulnerability to oxidative damage—pathways implicated in ASD. Animal and molecular studies have reported links between acetaminophen exposure, oxidative stress markers, and altered neurodevelopmental gene expression.

• Endocannabinoid system interactions. Acetaminophen (and its metabolites) modifies the endocannabinoid system, which is important for brain development (axon guidance, synapse formation). Disruption of signaling during critical windows could plausibly alter neurodevelopment. 

• Immune/inflammatory modulation. Acetaminophen affects prostaglandin and other inflammatory pathways. Maternal immune activation and altered fetal immune signaling have been implicated in ASD risk; acetaminophen could theoretically modulate those signals.

• Dose/timing-dependent effects. Experimental animal data suggest vulnerability depends on dose and developmental window; extrapolating to human therapeutic exposures requires caution. 

Limitations, biases, and why studies disagree

• Indication (confounding by indication): Fever, viral infection, and pain (the reasons for acetaminophen use) themselves may increase neurodevelopmental risk; separating medication effects from underlying illness is difficult. Consider this- pregnant mothers should not “tough it out” either.

• Measurement error: Many studies rely on maternal recall, prescriptions, or over-the-counter use reporting; such measures can misclassify exposure dose and timing. Biomarker studies (meconium) reduce misclassification but are fewer.

• Residual familial confounding: Genetics and family environment are powerful determinants of ASD; sibling designs and negative-control analyses that account for these often attenuate associations. 

• Heterogeneity across studies: Differences in exposure windows (trimester-specific vs. any prenatal use), outcomes measured (ASD diagnosis vs. traits), and covariates adjusted produce inconsistent results. 

Recent regulatory and public-health developments

• Regulatory review (2025): U.S. authorities announced an active review and have initiated label changes and communications reflecting possible associations between prenatal acetaminophen use and neurodevelopmental outcomes; public agencies emphasize ongoing review and the need for cautious interpretation. This reflects the accumulation of observational data and the need to weigh risks of untreated fever/pain against uncertain medication risks.

• Scientific community stance: Professional societies (e.g., obstetrics, pediatrics) generally continue to recommend acetaminophen as first-line antipyretic/analgesic in pregnancy because fever itself poses risks and because causal evidence for acetaminophen is not established—while also calling for more research and careful counseling. (See coverage from academic centers summarizing the debate.)

Clinical implications and practical guidance

• Do not alarm; counsel on risks and benefits. Given uncertainty, clinicians should avoid definitive statements that acetaminophen causes autism. Instead, discuss that some observational studies report associations but that causality is unproven and studies differ. Emphasize that high fevers and uncontrolled pain in pregnancy can be harmful, and acetaminophen remains an option when indicated. J

• Minimize exposure when reasonable but treat indications appropriately. For pregnant patients: treat high fever promptly (fever can harm the fetus), use the lowest effective dose of acetaminophen when needed, and consider nonpharmacologic measures when appropriate. Shared decision-making is key. For infants/children, follow current pediatric guidance for fever/pain management and dosing.

• Document indications and timing. If using acetaminophen in pregnancy, document indication (fever vs. pain), dose, and duration—this transparency helps clinical decision-making and future research.

Research gaps — what we still need

1. Well-designed causal studies (prospective cohorts with biomarkers, randomized trials where ethical, or strong quasi-experimental designs) that better separate medication effects from maternal illness.

2. Dose–response and timing analyses to determine whether specific trimesters or cumulative dose thresholds are relevant.

3. Mechanistic human studies linking exposure at therapeutic doses to measurable oxidative, immune, or endocannabinoid changes in the fetus or neonate.

4. Translation of animal findings: reconcile species/dose differences and study clinically relevant exposure ranges.

Key references (more available on request)

• Ahlqvist V.H. et al., JAMA 2024 — sibling-control analysis: no association between prenatal acetaminophen and ASD/ADHD when controlling for familial confounding.

• Baker B.H. et al., JAMA Pediatrics 2020 — meconium biomarker study associating prenatal acetaminophen exposure with ADHD and altered brain connectivity.

• Khan F.Y. et al., systematic review 2022 — summarizes observational evidence for prenatal acetaminophen and ASD/ADHD and discusses limitations.

• Parker W. et al., reviews on oxidative stress and acetaminophen hypothesized mechanisms (2017–2023) — mechanistic plausibility via oxidative stress and immune pathways.

• U.S. FDA / HHS communications (2025) — agencies responding to accumulating evidence and initiating label/review actions.

• Prada D. et al., Environmental Health 2025 — recent evaluation of the evidence on acetaminophen use and neurodevelopment.

Conclusion

There is a growing body of observational evidence linking prenatal acetaminophen use to neurodevelopmental outcomes including autism and ADHD, and biological mechanisms that could plausibly mediate harm have been proposed and shown in animal and molecular studies. However, causality remains unproven because of confounding, measurement limitations, and inconsistent results across more rigorous designs (e.g., sibling controls). Regulatory bodies are reviewing the data and updating communications; clinicians should use shared decision-making, treating fever when necessary and using the lowest effective acetaminophen dose when indicated, while conveying the current uncertainty to patients. Continued high-quality research (biomarker-based cohorts, mechanistic human studies, and clearer dose/timing data) is needed to resolve the question.