Independent New Hampshire Journalism · Northfield, NHCorrection and update, Sept. 19, 2026, 3:10 p.m. EDT: An earlier version of this column linked the phrase "completely went away" to an NIH news release, presented a European Medicines Agency quote as the shared conclusion of several agencies, said siblings in the Swedish study differed only in acetaminophen exposure, and called acetaminophen the only safe over-the-counter fever reducer in pregnancy. In fact the phrase is from study coauthor Brian Lee (PBS News), the quote is the EMA's (NPR), siblings share much but not all of their genetics and environment, and the FDA calls acetaminophen the "safest" option (FDA notice). We corrected those passages, dated the announcement, and added an update on later findings and the label's status.
The pattern repeats itself with disturbing precision.
First came COVID misinformation. Now it's prenatal care. The Trump administration's September 22, 2025, announcement linking acetaminophen to autism follows the same playbook: make bold health claims without scientific backing, contradict medical consensus, leave vulnerable people confused and scared.
I've watched this unfold before. The difference this time? Pregnant women are caught in the crossfire.
The Evidence Problem
When Trump declared "Taking Tylenol is not good" and claimed communities without the medicine have "no autism," he ignored decades of rigorous research. The scientific community doesn't just disagree with this position. They've actively disproven it.
The most comprehensive study analyzed data from more than 2 million children in Sweden. When researchers compared siblings born to the same mother, the association between acetaminophen use and autism disappeared. "When we did a sibling analysis that compared siblings (born to the same mother), the association completely went away," Drexel University epidemiologist Brian Lee, a coauthor, said, as reported by PBS News.
This is the critical distinction the announcement ignores. Some observational studies have found statistical associations. But association isn't causation.
Observational studies can't control for all variables. Maybe mothers who take more acetaminophen during pregnancy have more infections, more stress, or other health conditions. These factors, not the medication, could explain any statistical relationship.
The Swedish study addressed this problem by comparing siblings. Brothers and sisters born to the same mother share much of their genetics and home environment, so the comparison strips out many family factors that can make a medicine look risky. Result? No link to autism.
The Medical Reality
Here's what makes this announcement particularly dangerous: fever during pregnancy poses real risks.
Untreated fever can cause miscarriage, birth defects, and premature birth. The Society for Maternal-Fetal Medicine warns these risks are especially severe early in pregnancy, when many women don't even know they're pregnant yet.
The FDA's own notice to physicians calls acetaminophen "the safest over-the-counter alternative in pregnancy among all analgesics and antipyretics," and pregnant women are told to avoid aspirin and ibuprofen. When the administration tells pregnant women to avoid it unless "medically necessary," they're creating a false choice between two manufactured risks.
The American College of Obstetricians and Gynecologists called the announcement "highly unsettling" precisely because it affects "the health and well-being of millions of people without the backing of reliable data."
The Trust Erosion
This pattern of health misinformation creates a deeper problem: it erodes trust in institutions that pregnant women need to rely on.
When political announcements contradict medical consensus, people stop knowing who to believe. Some will avoid acetaminophen and risk dangerous fevers. Others will dismiss all health warnings as political theater.
Both responses are dangerous.
The announcement even cited research from Harvard's Andrea Baccarelli, whose work on acetaminophen and autism was previously rejected by a federal judge as "unreliable" expert testimony. Using disputed research to support predetermined policy positions isn't science. It's politics dressed up as medicine.
The Real Standard
I understand the appeal of simple answers to complex problems. Autism rates have increased, and parents desperately want explanations. But good science doesn't work by finding convenient scapegoats.
Real evidence requires controlled studies, peer review, and replication. It means accepting uncertainty when the data isn't clear. It means distinguishing between statistical associations and actual causation.
Regulators abroad rejected the claim. "Available evidence has found no link between the use of paracetamol during pregnancy and autism," the European Medicines Agency said, NPR reported. Australia's chief medical officer and its drug regulator said they "join with other global medicines regulators, leading clinicians and scientists worldwide in rejecting claims regarding the use of paracetamol in pregnancy, and the subsequent risk of development of autism in children."
This isn't about politics. It's about protecting pregnant women from having to choose between manufactured fears and real medical risks.
Moving Forward
Pregnant women deserve better than having their healthcare decisions influenced by political announcements that contradict scientific evidence.
They deserve clear, honest information about real risks and benefits. They deserve medical guidance based on the best available evidence, not the loudest political voices.
Most importantly, they deserve health policies that prioritize their wellbeing over political messaging.
The pattern of health misinformation doesn't have to continue. But stopping it requires choosing evidence over politics, even when the evidence is more complex than we'd like it to be.
Update, as of September 19, 2026: The FDA's September 2025 notice itself said "a causal relationship has not been established." On January 20, 2026, the European Medicines Agency pointed to a systematic review and meta-analysis in The Lancet Obstetrics, Gynaecology, & Women's Health that found "no evidence that using paracetamol at recommended doses during pregnancy increases the risk of autism spectrum disorder, ADHD or intellectual disability in children." As of September 9, 2026, the acetaminophen label had not changed, Medical Daily reported.
Corrections: Granite State Report corrects verified errors promptly and appends a dated note identifying what changed and when. The corrections policy and the corrections register are public, and correction notes are never removed.
