Most likely, you’ll be seeing a lot of headlines today trumpeting the findings of a new study in JAMA Pediatrics about links between acetaminophen (Tylenol, paracetamol) use during pregnancy and later risk of ADHD or hyperkinetic disorders in children. But before you withhold Tylenol from your screaming teething baby or suffer through a horrendous headache in your second trimester, it may be helpful to view the evidence with a healthy dose of caution and skepticism.
As with so many other associations that researchers are digging into, the evidence related to acetaminophen’s possible long-term effects on children are far from settled and riddled with possible methodological bias that prevents us, so far, from having much to go on. In looking at the study published today, I’ll cover the findings first, then the strengths, then what all this means and the weaknesses — don’t skip that last part!
First, what did the study involve? The researchers conducted phone interviews with 64,322 pregnant women, one in each trimester, who gave birth between 1996 and 2002. They asked the women whether they had taken any painkillers and then asked which ones if the women said yes. (Women could choose from a list of 44 or add others.) The women also reported how many weeks during each trimester they had used these painkillers. Overall, 56% of the women had taken acetaminophen at least once during their pregnancies.
The mothers’ responses were then compared against diagnoses of hyperkinetic disorders in their children, the children’s use of ADHD medications and the children’s display of ADHD-like behavior at 7 years old, based on a 25-question assessment completed by the mothers.
When the researchers crunched the numbers – taking into account a wide range of other factors that I’ll get to in a moment – they did find associations between use of acetaminophen during pregnancy and hyperkinetic disorders and/or ADHD. Children were 37 percent more likely to have been diagnosed with a hyperkinetic disorder, were 29 percent more likely to be taking ADHD medications, and were 13 percent more likely to show ADHD behaviors at age 7 if their mothers reported taking acetaminophen during pregnancy (compared to moms who didn’t).
Further, the researchers found an “exposure-response” relationship. That means the risk of these three outcomes appeared to increase along with the number of weeks that the mothers reported using the acetaminophen, particularly if it was taken during the first trimester.
The study’s findings should be taken seriously because of how many children were involved in the study – the higher the number, the more accurately trends and associations can be identified – and because it was prospective, which means the information was gathered as it happened rather than after the fact. (Asking women how often they took acetaminophen after they’ve had their babies can be subject to recall bias.)
Another major strength of the study is the number of “confounders” considered, the things the researchers took into account that might otherwise influence the results. In other words, what else could have added to the risk for ADHD or hyperkinetic disorders besides the acetaminophen? (For example, let’s say it’s possible that a certain pain condition might be associated with ADHD; if a pregnant woman with that condition took Tylenol, there would be no way to know if it was the medication or the condition that was linked to the ADHD.)
The researchers considered a range of child characteristics (birth year, birthweight, sex) and mother characteristics (age, smoking/alcohol during pregnancy, socioeconomic status, weight and mental health). Importantly, the researchers also asked mothers about health conditions – fevers, inflammation, muscle/joint diseases and infections during pregnancy – that might have influenced why they were taking acetaminophen and separately influenced the results.
Despite these strengths, though, it’s important to understand what the numbers really mean, what the study’s weaknesses are and, above all, that a study like this (observational) cannot show that taking acetaminophen during pregnancy *caused* ADHD or hyperkinetic disorders in kids.
First of all, the numbers are not big. The ones I reported above are relative risk, not absolute risk. For example, the absolute risk of showing ADHD-like behaviors at age 7 was 2.5% for mothers who didn’t take acetaminophen and 3.4% for those who did. So out of 1,000 moms who didn’t take acetaminophen, 25 of their children would show those behaviors, and out of 1,000 mothers who did take acetaminophen, 34 children would show those behaviors. That’s not a huge difference.
However, the exposure-response relationship for ADHD behaviors shows that the difference can be big for mothers taking acetaminophen more often. Consider the following chart, in which I’ve adapted some of the study’s data for easier reading:
|How long mothers took acetaminophen||Children with ADHD-like behaviors||Children lacking ADHD-like behaviors||Risk of increase|
You can see that taking more acetaminophen generally increases the risk, except an odd blip at 11-20 weeks. It turns out that blip is not statistically significant, which means it might simply be due to chance. It’s also important to notice that the numbers of children decrease as we go up in weeks of use. The fewer kids involved, the weaker the found associations are in statistical terms. That’s just one grain of salt to keep in mind in looking at this study.
Other grains of salt:
—The discussion of ADHD relies on medications and observed behaviors at age 7, not official diagnoses of the condition.
—Almost a third of the moms missed one or more phone calls during the study, so the data is incomplete and not necessarily representative, especially if those moms shared characteristics (such as having ADHD, which can have a hereditary component) that made them, as a group, less likely to participate in all three phone calls.
—Other conditions not accounted for, besides fever and the others included, might be related to the ADHD risk. Could moms who have a lot of headaches and take acetaminophen for them be more likely independently to have kids with ADHD? Or, could the type of person who more quickly pops a Tylenol to deal with a headache also be the type of person who is more genetically prone to have a kid with ADHD? We don’t know.
—Finally, the number of weeks the mothers took the acetaminophen may not translate to dosages. Consider the difference between a mom who takes one Tylenol dose each week for 20 weeks versus a mom who takes four doses of Tylenol a day for three weeks straight and then no more. Throw in the question of which trimesters they were in, and you have some complicated data to sort through that wasn’t even collected in this study.
From here, possible avenues to explore could get even more complicated: if there is a real link between acetaminophen during pregnancy and these disorders in children, it’s possible that the link only exists in women or children with a genetic predisposition for it. Such a finding has already been found with alcohol exposure in pregnancy, in which some children are more genetically susceptible to smaller amounts of alcohol during pregnancy than others, even if all other factors (including the amount of alcohol) are similar.
And, even if the association is real, does this mean pregnant women should never take acetaminophen during pregnancy? Not necessarily. Like every other medication during pregnancy, the risks AND benefits of taking the medication should be considered. Some conditions, such as high fever, may be far more dangerous to a fetus than the acetaminophen that could help reduce the fever. And, again, more than half the 64,000+ mothers took acetaminophen at least once, yet only 3.4% of their kids had ADHD-like behaviors (not even an ADHD diagnosis). That’s not something to freak out about.
So, yes, there appears to be a possible link here, but it’s the first time the link has been seen, so a lot more study is needed, and there are a lot of other factors to consider before pregnant women swear off acetaminophen during pregnancy. Perhaps the only practical advice we can pull from this study at the moment is not to pop painkillers without good cause, which is generally a good practice anyway. But if you are in serious pain or have a high fever or have some other medically legitimate reason to be taking acetaminophen, this study doesn’t give us any reason to say you shouldn’t take it while pregnant.