When Public Health Becomes Politics:
HHS Bases Autism Claim on Inconclusive Tylenol Studies
Dr. José M. Zuniga, President/CEO, Fast-Track Health
As a society, we entrust our public health institutions with an enormous responsibility: to ground policy and clinical guidance in the best and most rigorous science available. Anything less imperils trust, protection, and lives. It is therefore deeply troubling for the President of the United States, flanked by his Secretary of Health and Human Services, FDA Commissioner, NIH Director, and CMS Administrator, assert a link between acetaminophen (Tylenol) use in pregnancy and an increased risk of autism, even as many leading experts caution that the evidence remains inconclusive.
The problem is not asking questions. It is critical to study environmental, genetic, and other prenatal factors that might interact to influence neurodevelopment. But putting forth sweeping public health guidance or policy shifts based on observational studies – or early reviews with mixed or contradictory findings – risks misinforming the public, giving false hope, or, worse, causing harm. Observational associations can help generate hypotheses, but they very often fail to establish causation. Pregnancy is one of the most sensitive periods in development. Women, caregivers, and clinicians deserve clarity not sensationalism.
We must be wary of overreaching conclusions. According to prominent medical associations, acetaminophen is considered safe for pregnant women when used appropriately, especially when needed to treat pain or fever. The risk of leaving fever or pain untreated during pregnancy can itself be serious, potentially leading to other outcomes of concern. Guidance that leans toward prohibition or avoidance of commonly used medications – before conclusive trials or overwhelming consensus – can lead to panic, confusion, and unintended negative health consequences.
Public health policy needs transparency, nuance, and humility. If new research suggests possible risks, so be it – those findings should be shared, debated, and subjected to peer review. But issuing definitive-style warnings or policy shifts in advance of solid, reproducible evidence undermines the credibility of the institutions themselves. People look to the CDC, the NIH, the FDA, CMS, and HHS for impartial science, not politics cloaked in scientific claims. When policy follows ideology rather than accumulated, vetted data, trust erodes.
HHS says it has reviewed prior research, including an August 2025 review by Mount Sinai and Harvard researchers,” that suggests a possible link between acetaminophen use during early pregnancy and increased risk of autism in offspring. (1) Also under consideration are other cohort studies, including a large Swedish study, (2) as well as evidence concerning maternal folate levels during pregnancy.
With respect to the latter, the Swedish sibling-control cohort study (2024, ~2.5 million children) using standard models without sibling controls they found marginally increased risks for autism and ADHD, but the increases were small. Crucially, when controlling for shared familial factors (using full sibling comparisons), no association was found between acetaminophen use and autism, ADHD, or intellectual disability.
The 2025 review “Evaluation of the Evidence on Acetaminophen Use and Neurodevelopmental Disorders (NDDs)” using the Navigation Guide methodology included 63 relevant papers. Many of the studies reported positive associations (higher risk) between prenatal acetaminophen exposure and NDDs including ADHD, autism, etc. But it also found many null findings, plus significant concerns about bias, confounding, and study quality.
Bottomline, no causal relationship has been definitively established. Most of the evidence is observational. That means we see associations (i.e., pregnant person used acetaminophen → child later diagnosed with autism/NDD), but that does not prove cause. Confounding variables (fever, infections, other medications, genetic/familial risk) could also explain or partly explain associations. The sibling-control design in the Swedish study is one attempt to correct for unmeasured familial/genetic confounders; its finding of no association in that design suggests that at least some of the observed associations in naïve models may be due to those confounders.
The studies reviewed in the Navigation Guide review (2025) differ in how exposure is measured (self-report, records, prescription vs over-the-counter use), what outcomes are measured (autism, ADHD, “neurodevelopmental disorder” more broadly), follow-up time, and control of confounders. etc. Some studies have sibling-controls; many do not. That variation matters a lot for reliability.
The review also tried to rate the studies by strength of evidence and bias. While many of the studies found associations, higher bias risk was common, and the evidence strength is not uniform. Moreover, some studies report exposure-response relationships, but the exact timing in pregnancy (which trimester), the dosage, and other factors vary. That makes it difficult to translate into concrete guidance (“avoid Tylenol during pregnancy”) because what dose, when, and under what conditions matters.
Another study cited during the White House press conference today is the Johns Hopkins/Boston Birth Cohort analysis, which measured acetaminophen biomarkers in umbilical cord blood and found that newborns with the highest exposure had roughly three times the risk of later ADHD or autism compared to those with lowest exposure. However, this study relies on biomarker exposure measured at birth and cannot fully account for all potential confounding variables (such as maternal illness, dosage timing, or environmental exposures), nor does it establish causation. (4)
Of note, The Navigation Guide review (2025) concludes with a caveat: “While this association warrants caution, untreated maternal fever and pain pose risks such as neural tube defects and preterm birth, necessitating a balanced approach. We recommend judicious acetaminophen use – lowest effective dose, shortest duration – under medical guidance, tailored to individual risk–benefit assessments, rather than a broad limitation.”
Equally concerning is the simultaneous promotion of leucovorin (a form of folinic acid) as a potential “treatment” for autism, despite limited and inconclusive evidence. Small, preliminary studies have examined leucovorin for subgroups of children with folate pathway abnormalities, but systematic reviews conclude that the data are insufficient to recommend its use broadly for autism spectrum disorder (ASD). (3) Endorsing or elevating therapies without rigorous validation risks misleading families, wasting resources, and undermining confidence in legitimate avenues of autism care and support.
At its best, public health is a balance of urgency and caution, innovation and conservatism. We must not allow our institutions to sacrifice scientific rigor at the altar of political expediency. To respect both truth and trust, we need policies and guidance as carefully constructed as the stakes demand.
Citations:
- Ahlqvist VH, Sjöqvist H, Dalman C, Karlsson H, et al. acetaminophen use during pregnancy and children’s risk of autism, ADHD, and intellectual disability. JAMA. 2024;331(14):1205-1214. doi: 10.1001/jama.2024.3172
- Prada D, Ritz B, Bauer AZ, and Baccarelli AA. Evaluation of the evidence on acetaminophen use and neurodevelopmental disorders using the Navigation Guide methodology. Environ Health. 2025 Aug 14;24:56. doi: 1186/s12940-025-01208-0
- Sathe N, Andrews JC, McPheeters ML, Warren ZE. Nutritional and dietary interventions for autism spectrum disorder: a systematic review. Pediatrics. 2017;139(6):e20170346. doi: 10.1542/peds.2017-0346
- Ji Y, Wang X, et al. Acetaminophen metabolites in cord blood and risk of attention-deficit/hyperactivity disorder and autism spectrum disorder in childhood. JAMA Psychiatry. 2020;77(2):180–189. doi: 10.1001/jamapsychiatry.2019.3259