The painkiller acetaminophen or paracetamol is one of the most widely taken drugs during pregnancy, used by roughly half of all pregnant people worldwide.

But today the US Food and Drug Administration (FDA) announced that it would slap a new warning label on the drug, citing a “possible association” between autism in children and the use of acetaminophen, also called Tyenol in the United States, during pregnancy.

US President Donald Trump, however, was not as nuanced in his language: “Don’t take Tylenol,” he said repeatedly in a widely anticipated announcement about autism. “Fight like hell not to take it,” he said.

Autistic people show differences in social communication and interaction, and reported prevalence of autism has risen in some countries in recent decades. But many researchers who study autism caution that there is insufficient data to link autism and acetaminophen and that focusing on such a link is no more than a distraction.

Autism is on the rise: what’s really behind the increase?

“There is no definitive evidence to suggest that paracetamol use in mothers is a cause of autism, and when you see any associations, they are very, very small,” says James Cusack, chief executive of Autistica, a UK autism research and campaigning charity in London, who is autistic. “At the heart of this is people trying to look for simple answers to complex problems.”

At today’s announcement, US health secretary Robert F. Kennedy Jr, acknowledged that autism has many possible causes. FDA commissioner Martin Makary noted that, in some cases, insufficient transport of the nutrient folate into the brain has been linked with traits common in autism. The FDA today announced that it would initiate approval of a form of folate called leucovorin for people with low levels of folate in the fluid surrounding the brain and spinal cord.

But Kennedy, a long-time anti-vaccine advocate, and Trump focused their attention today on two other factors that they said could be linked to autism: acetaminophen and vaccines. Decades of research have repeatedly found no reproducible link between vaccines and autism.

“The evidence does not support a causal link between acetaminophen or vaccines and autism,” says Sura Alwan, a clinical teratologist at the University of British Columbia in Vancouver, Canada, and executive director of PEAR-Net Society, a nonprofit advocating for maternal fetal health and research. “Suggesting otherwise may fuel misinformation and undermine confidence in safe treatments and immunizations.”

Here Nature examines the evidence for a link between acetaminophen, which is used to treat fever as well as pain, and autism.

How strong are the data connecting autism and acetaminophen?

Scientists say that the most robust research does not link autism and acetaminophen use in pregnancy. “The better controlled studies are less likely to find even a small risk,” says Helen Tager-Flusberg, a psychologist who studies autism at Boston University, Massachusetts, “And even then, what we’re talking about is a minor association. … We do not think that taking acetaminophen is in any way contributing to actually causing autism.”

Working out whether there is a link between the drug and autism is difficult, says Viktor Ahlqvist, an epidemiologist at the Karolinska Institute in Stockholm and co-author of what might be the biggest study¹ on the link. The medication is available over the counter, so a lot of usage is not recorded in medical databases. This means researchers rely on self-reports, which can be unreliable.

Confounding factors are an even bigger problem. Women who take paracetamol in pregnancy are usually in poorer health than those who don’t, perhaps because they had an infection or an underlying condition. Any apparent link between acetaminophen and autism might therefore be explained by these other health factors rather than the drug itself. Although scientists try to adjust for such confounders in their studies, such adjustments are “rarely sufficient,” Ahlqvist says. This is one reason why studies looking for a link have produced conflicting results.

The study led by Ahlqvist harnessed data on nearly 2.5 million children born in Sweden between 1995 and 2019 and — from the country’s extensive health records — data on acetaminophen prescriptions during pregnancy and on self-reported use collected by midwives, as well as whether children later received autism diagnoses.

The study showed that around 1.42% of children exposed to acetaminophen during pregnancy were autistic, compared to 1.33% of children who were not exposed ─ a “very small” difference, says Ahlqvist.

The team also compared pairs of siblings (born to the same mother), one of whom had been exposed to acetaminophen and one who had not. Siblings share half of their genome, and share a similar upbringing and mother’s background health, so any detected difference in autism between siblings is more likely to be due to the drug. The researchers found no association between acetaminophen and autism using this method — which supports the idea that links found in other studies were really explained by confounding factors.

Another large, high-quality study² from Japan including over 200,000 children — also using sibling comparisons and published this year — found no link between acetaminophen use in pregnancy and autism.

Are there any studies that do link the medication and autism?

A review³ of studies published in August in the journal Environmental Health concluded that there is an association. But researchers interviewed by Nature point instead to the contrasting results of a review⁴ of high-quality studies on the topic. That review, published in February, concluded that “in utero exposure to acetaminophen is unlikely to confer a clinically important increased risk” of autism spectrum disorder.

“There is no robust evidence or convincing studies to suggest there is any causal relationship and any conclusions being drawn to the contrary are often motivated, under-evidenced, and unsupported by the most robust methods,” Monique Botha, associate professor in social and developmental psychology at Durham University, UK, said in comments to the Science Media Centre, a UK press office.

“We believe independent, sound science clearly shows that taking acetaminophen does not cause autism,” says a spokesperson for Tylenol’s manufacturer, Kenvue Inc. in Summit, New Jersey.