Emergency departments across the United States saw a significant drop in acetaminophen orders for pregnant women following President Trump’s September warning linking the common painkiller to autism, according to a new study published Thursday in The Lancet.

The research found that orders of acetaminophen, the active ingredient in Tylenol, initially dropped 16% before stabilizing at a 10% decrease in the twelve weeks after the Trump administration cautioned against the drug’s use during pregnancy. Meanwhile, prescriptions for leucovorin, a cancer treatment drug promoted by health officials as a treatment for autism, rose 71% during the same period.

“This is an embodiment of how much power our federal health officials have,” said Michael Barnett, a study author and professor at the Brown University School of Public Health. “Even when nothing in the evidence base has changed, even for something as ubiquitous and familiar as Tylenol, it can still shift things to a measurable degree within days.”

The president’s message was clear during the September press conference: “Don’t take Tylenol. Don’t take it,” according to the study. Trump claimed that taking the painkiller increased the risk of having a child with autism, despite scientific consensus finding no such link and evidence that untreated fevers carry their own risks to pregnant people and fetal neurodevelopment.

Researchers analyzed electronic health records from hospitals and clinics nationwide using Epic’s Cosmos database. They examined data from 88,857 emergency department visits among pregnant individuals and 853,216 visits among non-pregnant individuals. The 10% drop translated to 22 fewer Tylenol orders per 1,000 visits, with each “order” indicating a medical professional accessed a hospital’s medication dispensing system or sent a prescription to a pharmacy.

Notably, the study found no statistically significant drop in acetaminophen use among non-pregnant women, suggesting the federal messaging specifically influenced treatment decisions for pregnant patients.

Study co-author Jeremy Faust, an emergency physician at Brigham and Women’s Hospital in Boston and assistant professor at Harvard Medical School, worked with Barnett to mine the national health records data.

The research raises concerns about potential unintended consequences. Lisa Croen, director of the Kaiser Permanente Autism Research Program who was not affiliated with the study, warned about possible medication substitution.

“Tylenol is the medication that is safe, the one medication that is safe to control pain and fever during pregnancy,” Croen said. “We don’t know if people were substituting paracetamol for other pain meds that are known to be hazardous [during pregnancy], like ibuprofen.”

The data cannot determine whether clinicians or patients drove the decrease, nor does it track over-the-counter acetaminophen use outside emergency departments.

While acetaminophen orders leveled off by the study’s end, leucovorin prescriptions continued accelerating. The 71% increase reflects the drug’s smaller market—the actual increase equated to only 17 prescriptions per 100,000 visits—but occurred after the Trump administration promoted the treatment on national television.

Leucovorin has been a niche treatment with modest effects for a subset of people with autism. According to Barnett, the administration’s promotion resembled “a direct-to-consumer pharmaceutical ad.”

The Trump administration’s broader autism agenda has drawn scrutiny from advocates. Health Secretary Robert F. Kennedy Jr. has called rising autism diagnosis rates an “epidemic,” while officials have removed federal webpages about controversial autism treatments and proposed a national autism registry that sparked community concerns.

The study demonstrates how federal health messaging can rapidly influence medical practice patterns, even when contradicting established scientific evidence about medication safety during pregnancy.