If you have a TV, phone, or  really any way to consume news, you've probably heard the claim that there's a link between acetaminophen and autism.

The current US administration has asserted that taking the drug during  pregnancy has been connected with development of the condition in offspring. And there are a lot of conflicting  takes about what that claim means, which research is “right”, and  what different studies really show.

So here’s some of the major studies  that are being talked about, their shortcomings, and what it means for all the achy or feverish  pregnant people out there. [♪INTRO] Before getting into it, a quick  disclaimer: Neurodiversity Good Actually. We have a diverse team here at SciShow, and actual autistic people worked on this video. Including me!

Hi! I got my  diagnosis a couple of months ago just in time for all of this! We support a society that meets  the needs of autistic people, not one that seeks to prevent  autism from happening.

However, it’s still important to  understand what this research is saying so that it doesn’t get  weaponized against vulnerable people. Also a note on terminology. While professionals in the field  often use person-first language, like “adult with autism”, many  from the autistic community have voiced a preference for identity-first language, such as “autistic adult”, so we’ll  use that more often in this video.

With that stated, let’s start  with why there’s so much interest in autism research right now:  the increased rate of diagnoses. The number of people diagnosed as autistic has increased over the last few decades. That’s a fact.

But that doesn’t necessarily  mean that the actual number of autistic people is increasing,  and that’s for a few reasons. First, the definition of autism has changed and expanded into a spectrum, which means more people included under  this diagnostic umbrella. There’s also been some progress in diagnosing autism in kids assigned female  at birth, Black children, and other previously underdiagnosed groups.

That would drive up the official  numbers while not changing the reality that those people  have always been autistic. Plus, kids are now regularly screened  for autism by their pediatricians, parental awareness has increased,  and adults who might have masked their symptoms all their lives are  advocating for their own evaluations. So the rate of autism could be  increasing, but it could just as easily be remaining stable while all these confounding factors explain an increase in the diagnoses.

Either way, it’s certainly not the emergency that some paint it to be. Again, autistic people existing is Fine Actually. But that’s why.

Now onto the more controversial  bit: does taking acetaminophen – aka Tylenol or paracetamol –  during pregnancy cause autism? First, it’s worth noting that not  every person with autism was birthed by someone who took acetaminophen  during pregnancy, and not every person who takes acetaminophen goes  on to have a child with autism. Autism as a behavioral profile has  existed for centuries, long before acetaminophen started being  used regularly around the 1950s.

So if you’re looking for an  “A always causes B” link, then acetaminophen is definitely not your culprit. Current research is really  looking at whether acetaminophen use increases the risk of autism in offspring, not whether it consistently causes it. But that is hard to do.

Really hard. The gold standard for  demonstrating causal relationships in medicine is a randomized  controlled trial, or RCT. Which hasn’t been done  because researchers generally strive to do what’s best for  participants in these trials.

To conduct an RCT, you would  need to split participants into groups who take acetaminophen  and others who do not. For most people that wouldn’t be a problem, because you can usually turn  to other drugs for help. Things like ibuprofen, found in  Motrin, can reduce fever or pain, or naproxen, used in Aleve,  can help with joint pain.

But it’s a different story for pregnant people, because NSAIDs like those aren’t  safe to take during pregnancy. Acetaminophen is widely considered the only safe option for fever or pain  relief in pregnant people. Because of that, telling a bunch  of achy pregnant participants that they can’t use the only  safe pain reliever is simply not something that has been done much.

Even if you had no qualms about  telling pregnant people to not take acetaminophen for 9 months,  there’s still another big problem. Acetaminophen is a fever  reducer, and high fever during pregnancy is associated with  an increased risk of autism. A 2018 study found around a 40% increase.

And a 2021 study found even  higher risk when that second trimester fever was accompanied by an infection, doubling the chances of having a baby with autism. Animal studies suggest that inflammatory molecules associated with fever might  affect parts of the brain. But in the context of our  hypothetically perfect clinical trial, it represents a huge confounding variable that would make the study worthless.

It would be borderline impossible to separate the effects of fever from the  effects of treating the fever. Oh, and untreated high fevers  during pregnancy are also dangerous for the pregnant person and the  fetus, adding another point to the ethical no-no of withholding  Tylenol from pregnant people. So instead of that, researchers have chosen to do observational studies,  which mostly means asking people whether they took  acetaminophen while pregnant and then seeing if their kiddos ended  up with an autism diagnosis.

Some studies ask this question  after a child has already been diagnosed, which means  recall bias can be an issue. Others ask throughout pregnancy or  review electronic medical records, but that can still be problematic  if patients don’t realize over-the-counter medications should be reported. Be honest with me, when’s  the last time you remembered to disclose taking regular  ol’ pain meds to your doctor?

There’s also the possibility that  stigma around taking medications discourages pregnant people from  admitting to taking the drug. Also, pregnancy fog is real. If  you’re asked at 32 weeks whether you’ve taken acetaminophen  while pregnant and how much, you might literally not remember.

Even sampling umbilical  cord blood has its problems. Acetaminophen only lasts a few  hours before breaking down, so detecting its by-products just means a person used acetaminophen in the  hours leading up to birth. And all of that doesn’t even  address the problem of figuring out dosage, frequency, and  timing of acetaminophen use.

So knowing the facts around  acetaminophen use during pregnancy is pretty hard for any study to pinpoint, which means all their results need  to be taken with a grain of salt. But let’s hypothetically say that  people were recruited right when they found out they were pregnant and there was perfect reporting of acetaminophen use. There’s still other problems  those studies would run into.

Chronic pain is common in  autistic people, and some research suggests they may have a higher  sensitivity to pain in general. Research also shows that up to 80% of autism risk is related to genetics alone. That means pregnant people who  are autistic themselves or have some of the right genetic makeup  may be in more pain and need to take acetaminophen more during  pregnancy, and their child getting diagnosed with autism might’ve  happened anyway due to genetics.

The pregnant person could  also be undiagnosed autistic, since diagnosing girls with  autism a few decades ago was pretty uncommon, so studies trying to take any familial diagnoses into  account might even miss that. The studies also don’t always  inquire why pregnant people took acetaminophen, which is a huge issue when we already know there’s a link  between fever and autism. And that’s just scratching the surface of potential confounding variables.

Some can be factored in fairly  easily, like autoimmune disorders, gestational diabetes, obesity, and  parental age, but others are more difficult to quantify, like infection, stress, and exposure to environmental pollution. Just to name a few. No singular thing causes autism.

There are a lot of associations and  things that might increase risk, but trying to parse out  correlation and causation is nearly impossible when there are so  many variables to consider. Which brings us to the 2025  study at the crux of this claim. This 2025 study, published in  the journal Environmental Health, was a review article, meaning the  researchers looked at a bunch of existing studies that investigated  the link between acetaminophen use during pregnancy and  neurodevelopmental disorders.

Then, they rated the studies on things like funding bias or how they recorded drug exposure. Only eight of the studies  they looked at focused on the association between specifically autism and acetaminophen use during pregnancy. And based on the results of those  studies and the researchers’ ratings, the authors concluded that  there was strong evidence of an association between autism and acetaminophen.

But remember it was just a review article, meaning it can only assess what’s  already out there in the literature. Reviews can’t establish causation on their own. Plus, as great as it would  be for the researchers to be all-knowing when it comes to  rating other people’s studies, there’s still a level of subjectivity there.

And all the studies they looked  at still suffer from the same shortcomings of trying to  track acetaminophen usage, so any conclusions this review  article reached could also be flawed. Still, the researchers here made  it clear that they do believe there’s adequate evidence  to link acetaminophen use during pregnancy to autism, and even leave the possibility of a causal  relationship on the table. But not all studies reach that same conclusion.

A study in The BMJ used a  reliable assessment tool called AMSTAR 2 to rate the quality of  the review articles it looked at. It showed they were all of  low or critically low quality, including the controversial 2025 paper. The researchers concluded that  the best studies were the ones that included sibling controls,  and that based on those, any association between acetaminophen and autism shrinks or disappears when  properly controlled for.

Another vote against a link comes from a very large 2024 Swedish study. With a sample size of over 2 million  and over two decades of data, they found a slight increase in risk when patients reported using acetaminophen  at all during pregnancy. But that association once  again disappeared when they looked at sibling controls,  and stayed absent even when looking at different doses  taken throughout the pregnancy.

That paper concluded that the initial association must have been caused by some other variables, such as genetics or other environmental factors. Critics question the results of  that study because the rate of acetaminophen use was much  lower than reported in others, suggesting they suffered from  some data collection problems. I sound like a broken record, but that’s definitely a theme  in all of this research.

But a study on over 200,000 children  in Japan had similar findings, with a slight association disappearing  when the sibling comparison was done, and with such a  large number of participants, neither of these studies  should be totally dismissed. To sum it up, the data is mixed  and riddled with design flaws. Maybe there’s an association, or maybe the mountain of confounding variables are to blame.

There’s really just not enough evidence to say there is or isn’t a correlation, never mind trying to prove  acetaminophen itself causes autism. Even if there is an association,  some researchers suggest that chronic or frequent use  is where that link emerges. And doctors recommend that anyone,  but especially pregnant folks, use the lowest effective  dose when taking medications.

That means pregnant people are probably already “fighting like hell” to minimize  acetaminophen use during pregnancy. People aren’t gulping acetaminophen  down just for fun of it. But if they were, the WHO still  states that there is no conclusive evidence linking acetaminophen  use during pregnancy to autism.

And in response to uproar from the  scientific and medical community, even people in the US administration have walked back some of the  claims they initially made. Unlike the unproven risk of acetaminophen use, high fevers have well documented dangers and should always be treated during pregnancy. The research is very clear on that.

For the less clear research, keep  in mind that no study is perfect, and strong claims in any direction could always benefit from a healthy dose of scrutiny. To harken back to your intro to  psychology class, correlation does not mean causation, and some  skepticism can be a good thing. If you’re pregnant and struggling  to figure out what’s safe to take, always talk to your doctor.

If you’re the parent of an autistic  person and took Tylenol during your pregnancy, you shouldn’t  feel any guilt or blame for your child’s neurotype – because you  followed the best guidance available, and because, say it with me, autistic  people existing is Fine Actually. And if you’re autistic, know that  no matter where the research lands on this, your existence can never be boiled down to a medication your parent did  or didn’t take while pregnant. [♪OUTRO]