In an announcement on Sept. 22, President Donald Trump said that using Tylenol during pregnancy increases the risk of having a baby with autism, contradicting advice from leading obstetrics experts who say the popular drug is very safe in pregnancy.
“Don’t take Tylenol. Fight like hell not to take it,” Trump said during a White House briefing.
The president, who was flanked by U.S. Health and Human Services (HHS) Secretary Robert F. Kennedy Jr. and other cabinet officials, said pregnant women shouldn’t take acetaminophen, the active ingredient in Tylenol, unless they have an “extremely high fever.”
Several studies conducted over the past decade have probed whether acetaminophen increases the risk of autism and other neurodevelopmental disorders. The results of these studies have been mixed, with some showing a possible correlation and others showing no link. None has shown a causal connection.
The Trump Administration relied in part on the review as the basis for the announcement. But many obstetrics experts have questioned the validity of the review’s conclusions.
Dr. Nathaniel DeNicola, an ob-gyn who advises ACOG on environmental health issues, says the problem with the review is that it was based on studies that were mostly poorly designed and therefore unreliable. “The studies that have claimed an association were so deeply flawed that you can’t draw any conclusion,” he says.
Several studies, for example, relied on women’s memories to determine if and how much acetaminophen they took during their pregnancy. The women were asked to recall these details weeks, months, and in some cases years after the fact, DeNicola says. “Anyone who has been asked the question, ‘What did you have for lunch yesterday?’ knows that you tend to misremember things, even things from the recent past,” he says.
“What this tells us is that the [initial] association was most likely due to genetics,” says Brian Lee, a professor of epidemiology at Drexel University and the study’s lead investigator.
Research suggests that people with autism are more likely to suffer from pain-causing conditions such as headaches and migraines and may therefore be more likely to take pain medications. Pregnant women with a higher genetic risk of autism may therefore be more likely to take acetaminophen, Lee says.
Asked about the review published in August, which included his group’s JAMA study, Lee says its conclusions were flawed because it was based on mostly flawed studies. “Garbage in, garbage out,” he says.
Correlation is also, critically, not causation, Lee says. “The example I like to give is: Eating ice cream is statistically associated with drowning. But ice cream doesn’t cause you to drown.” It just so happens that both the rates of eating ice cream and swimming go up in hot weather.
During the press briefing, Dr. Marty Makary, chief of the U.S. Food and Drug Administration claimed that Baccarelli, who co-authored the review, had said there was a clear causal relationship between acetaminophen use during pregnancy and an increased risk of neurodevelopmental disorders. Baccarelli, however, was unequivocal in his statement to TIME that the review didn’t show a causal link and that more research was needed “to confirm the association and determine causality.”
“As the only approved medication for pain and fever reduction during pregnancy, acetaminophen remains an important tool for pregnant patients and their physicians. High fever can pose risks to both the mother and the fetus, including neural tube defects and preterm birth,” Baccarelli wrote in the statement. “After assessing the evidence, my colleagues and I recommended a balanced approach based on the precautionary principle: Patients who need fever or pain reduction during pregnancy should take the lowest effective dose of acetaminophen, for the shortest possible duration, after consultation with their physician about their individual risk-benefit calculation.”
Following the White House briefing, the FDA announced that it was issuing a notice to physicians highlighting the potential risks of acetaminophen use during pregnancy. The agency said it was also initiating a safety-label change for the drug.