Does Tylenol cause autism?

Support for Unexplainable comes from Anthropic, the team behind Claude.

You know those questions that keep you up at night?

The ones that make you fall down rabbit holes until 2 a.m.

Claude is an AI that can help you if you want to dive deeper.

Instead of rushing to surface-level answers, you can use Claude to help work through the complexity.

You can upload documents, you can search the web, and you can explore different angles on problems that genuinely interest you.

You can try Claude for free at claude.ai/slash unexplainable.

Before we start the show, I want to tell you about a new way to listen to Unexplainable by becoming a Vox member.

You'll get access to a new feed where you can listen ad-free, just a steady stream of science mysteries with no breaks.

You'll also be able to read every article on Vox.com.

You get access to a member exclusive newsletter, and you'll get a whole bunch of other perks.

More importantly, though, you'll be supporting Unexplainable.

Making this show, diving into these unanswered questions, it takes a lot of time and effort and drafts.

So many drafts.

And every new Vox member gives us a stable base of support that allows us to make even more ambitious episodes and long-term series.

We truly can't make this show without you.

And right now, Vox is offering 30% off an annual membership.

All you have to do is go to Vox.com slash members to join.

Thanks.

It's Unexplainable.

I'm Noam Hasenfeld.

And by this point, you've probably heard about Trump's Tylenol thoughts.

Don't take Tylenol.

Don't take it.

If you just can't,

I mean, it's just fight like hell not to take it.

Trump and other officials like RFK Jr.,

they seem convinced that Tylenol causes autism, that it's responsible for the increase in autism diagnoses we've seen over the past couple decades.

But when you actually look at the science, it's a lot more complicated.

And when I was reading and listening to all kinds of things on this supposed connection here, I listened to this great episode of Science versus.

And it was so good that I just wanted to share it with you.

I love the way they approach the question with an open mind.

And they do a great job of getting into the nitty-gritty of the data, not just whether Tylenol causes autism, but what specific data has led so many people to actually think it does, and why that data isn't necessarily what it seems.

Okay, here's science versus and Meryl Horn.

This is Meryl, and today we're diving into this question of whether there's a link between acetaminophen and autism.

With me is senior producer Rose Rimmler.

Hey, Rose.

Hey, Meryl.

So, yeah, you may remember that we worked on this whole autism episode a few months ago, right?

Yeah.

And we covered what causes autism, why it looks like it's gotten more common in the past few decades.

And while I I was working on that, I talked to a scientist who looked at Tylenol or acetaminophen.

So it was kind of perfect because when I saw all of this was hitting the news, I was like, oh, now I can call that guy back up.

You had this in your back pocket just waiting to get it.

Yeah.

It was all right.

Set to go.

So yeah, this is Brian Lee, professor of epidemiology at Drexel University.

Hello.

Hey, Meryl.

How are you?

It's been a very hectic last few days.

I bet.

Yeah, he's getting kind of swarmed.

And I wanted him to help me understand like where this idea came from in the first place, this idea that acetaminophen could affect a fetus.

Like if we were going to build the case against acetaminophen, according to science, where would the story start?

And apparently these concerns have been floating around for a while.

Over a decade ago, the first sort of research came out about acetaminophen during pregnancy and neurodevelopmental disorders.

And that idea isn't crazy because it's been previously linked with other sort of disorders, for example, specific birth defects,

possible sex

hormone issues.

Yes.

Huh.

So this was surprising to me, like sex hormone issues.

So Tylenol or acetaminophen, we already know it can affect a fetus.

Well, there had been research on it for sure.

Okay.

Suggesting that there might be some risks here, which was, it was surprising to me because I've been pregnant and doctors will always tell you that Tylenol and other stuff with acetaminophen in it is the safe one to take.

You're not supposed to take ibuprofen because that's thought to mess up blood flow to the fetus and has been linked to congenital birth defects.

Acetaminophen is thought to be like the safer of the two.

I didn't know that.

So like if if you're pregnant and you feel fluy or you have some pain, you're supposed to take Tylenol, not one of the other ones.

Yeah, that's that's why it's so common.

It's like the one that for so long has thought to be relatively safe.

But even though they had these suspicions from these other conditions?

Well,

there had been these kind of like trickling out of like concerns, like Brian said, that it might be messing with sex hormones, kind of acting as an endocrine disruptor.

That's so random.

Yeah, no, at first I was like, what?

But apparently back 15 years ago, some researchers pointed out that some painkillers sort of look like a type of endocrine disruptor that I'm sure you've heard of, phthalates.

Sure.

They're an element in plastics that are thought to be an endocrine disruptor.

Exactly.

Yeah.

And so they tried testing whether painkillers, including acetaminophen, acted like phthalates when they put them on cells in a dish.

And they found that, yeah, they kind of did.

Oh, interesting.

And we think phthalates might mess up hormones like testosterone in the body.

And so researchers started looking into this and some of them did find that people who took acetaminophen during pregnancy were more likely to have a son with undescended testicles when they were born.

Huh, okay.

Yeah, so some researchers were finding these things, not in every study, but there were these hints that something might be going on here.

And that was the backdrop for this new wave of studies looking at the brain, because we know that hormones can affect brain development.

So, if something's messing with the hormones, maybe it could also change how the brain is developing in the fetus.

Okay, that makes sense.

And then studies start coming looking at whether pregnant people who took acetaminophen were more likely to have a kid with autism.

So, like, you know, some of them would get thousands of pregnant moms and ask them some questions.

They collected data on whether a mom used acetaminophen.

Did you take Tylenol?

Yes, no, that kind of thing.

And then they followed the baby after it was born to see if the baby developed autism.

Mm-hmm.

That's useful for our purposes.

Yeah, exactly.

Like they followed these kids and then checked whether they're autistic.

And these studies, they were finding stuff, but it was a little weird.

Like one of these studies found that it upped the chance for autism, but only when the autism was accompanied by hyperactivity symptoms.

Another study found that boys were more likely to be autistic with this exposure to acetaminophen in the womb, but not girls.

Huh.

Okay.

Kind of confusing.

Yeah.

Like none of these studies was like a slam dunk on its own.

But at the same time, dozens of other studies were also coming out looking at other stuff, especially ADHD, and finding that rates of ADHD were up in these kids exposed to acetaminophen in the womb.

Oh, okay.

And yeah, in some cases, it looked like the link was strongest for mothers who took acetaminophen pretty often, like in all three trimesters.

So there's like a dose response element as well.

That's what it looked like.

And in 2021, this all comes to a head.

A paper comes out from a bunch of researchers who'd been studying this called a consensus statement.

They issued a quote call for precautionary action unquote and basically suggested that people should pump the brakes on these pills, minimize the amount of acetaminophen that they take during pregnancy.

Here's Brian.

And then to top it off, they publish in a Nature journal, which is one of the most highly regarded scientific journals out there.

And so this statement really made scientists pay attention.

Up until that point, most people just kind of treated it like, oh, you know, like a new study pops up every day.

You aren't really sure what to believe or not.

But once a consensus statement came out, it gave it like a stamp of authority.

Yeah, I would feel that way about a consensus statement in nature or a nature offshoot.

Yeah.

And some parents even started suing the companies that make meds with acetaminophen in them, saying that they should have warning labels on them about these risks.

But

as all of this was happening, a lot of scientists were like, wait a second.

When you use a phrase like consensus statement, it implies a sort of every single expert in the room believes X, Y, and Z.

Like nine out of 10 dentists recommend Colgate or something of that sort.

So it implies a sort of authoritative weight to the evidence.

And implying that this was the consensus seemed to piss a lot of scientists off.

Some of them wrote responses to that consensus statement saying that it was way too soon to come out so hard against acetamenophen, saying that the evidence they cited was, quote, weak, inconsistent, and to a large extent, methodologically inadequate, unquote.

So they're saying the studies are crap.

Well, basically, yeah, that's the science-y way of saying, what?

This?

This is not enough.

And so a lot of scientists were really skeptical of this and didn't buy this story.

Scientists like Brian, does taking Tylenol actually increase, when you're pregnant, increase the chance that your kid will have autism?

The answer

right now

points towards no.

Oh, really?

Yeah.

So that's the short of it.

Wow.

Up next, the long of it.

We'll hear the other side of the story and tell you whether or not you can have your damn Tylenol after the break.

Support for the show comes from Upway.

So I didn't know how to ride a bike until I was almost in my 30s.

Somehow I missed learning when I was five, and by the time I was 10, I already felt like I was too old.

I finally learned a few years ago when I forced myself to bike to work.

It totally transformed my life.

I got around quicker.

I got more exercise.

It was just more fun to go places.

But biking can be a tough thing to get into if you aren't in the habit.

And that is where Upway comes in.

Upway makes e-bikes, so you can still get the benefits of biking without the dread of staring up at a huge hill between you and your office.

Upway makes it easy.

They offer top brands like Specialized or Cannondale and Aventon at up to 60% off retail.

Every e-bike is thoroughly inspected by a professional mechanic and comes with a one-year warranty, so you can ride with peace of mind.

You can search by brand or style.

You can book a call with a real human for help.

And you can even stop by their showrooms in New York or LA.

You can go to upway.co and use code unexplainable to get $150 off your first e-bike purchase of $1,000 or more.

That's upway.co code unexplainable.

Go ahead.

You won't regret it.

Support for Unexplainable comes from Anthropic, the team behind Claude.

Claude is an AI that can help you explore different angles on all kinds of questions.

Like the other day, I was watching the Red Sox and they played this sound after a strikeout.

And I was like, huh, I don't remember when that became a thing.

But it also wasn't something that I could just search for easily.

So I asked Claude.

He gave me some sources.

I tracked him down.

And before I knew it, I had the definitive history of the...

I also use Claude when I'm curious about genuinely complicated things.

I'll ask it something big.

We'll go back and forth to hone exactly what I'm looking for.

And yeah, it does get things wrong sometimes.

But when I double check or ask for a citation, it'll give me a better source.

Now, it's important to know how to use Claude well.

Like, ask it follow-up questions.

Actually, check out the citations it gives you.

But if you do, Claude can be a powerful tool.

It helps me find articles and primary documents that I never would have come across.

And it helps me dig wider and deeper, all while staying in control of the digging.

You can try Claude for free at claude.ai/slash unexplainable and see why the world's best problem solvers choose Claude as their thinking partner.

To remind you that 60% of sales on Amazon come from independent sellers, here's Scott from String Joy.

Hey, y'all, we make guitar strings right here in Nashville, Tennessee.

Scott grows his business through Amazon.

They pick up, store, and deliver his products all across the country.

I love how musicians everywhere can rock out with our guitar strings.

One, two, three, four.

Rock on, Scott.

Shop small business like mine on Amazon.

Welcome back.

This is Meryl.

So far, we've gone through a lot of the evidence against acetaminophen, the ingredient in medicines like Tylenol.

Lots of studies seem to find that taking it while you're pregnant does increase the chance that your kid will be autistic or have ADHD.

Are you ready for the rest of the science, Rose Rimmler?

Uh

yes.

Are you ready?

Yes, I'm ready.

All right, so here's where Brian Lee, our epidemiologist hero, comes in gallantly with a shiny new approach to looking at this question.

His paper came out just last year.

We had a study funded by NIH before everything went to hell.

And we were like, hey, we could actually look at this.

And so Brian teams up with some Swedish researchers because he's actually really into IKEA.

I don't think that's why.

Yeah, you got me.

I think it's because the Scandinavians keep really good records, public health records.

Yeah, yep.

It's because they have an amazing database.

Sweden having its lovely universal healthcare system means that you can track everyone.

They don't just disappear in the system.

They're there.

They're tracked.

Everyone has access to this, right?

So, yeah, they got data from a ton of kids, almost two and a half million of them.

Basically, every kid who was born there over a 24-year period.

And they also had data from the moms, details about their medical care, like the medicines that they took during pregnancy.

And so, now Brian's team looked to see: were the kids exposed to acetaminophen in utero more likely to be autistic?

And right off the bat, they found that, yes, they were.

Oh, so they found that the kids who were exposed to acetaminophen had a 20 to 30 percent higher chance of being autistic.

Oh,

that's not what I was expecting.

You were about to tell me.

Right.

But actually, Brian's team was sort of relieved when they found this.

This shows that we're not in crazy land, for example, that somehow our data are not flawed.

We're not flawed.

Everything is working as anticipated.

And that's because the effects that Brian was seeing in his data was in the same ballpark as what other groups had found.

But here is where Brian and his team took a different path.

They wanted to see if other factors might be muddling up the data here, making it look like it's the acidaminophen to blame when really it's something else entirely.

Yeah, okay.

So

Brian is trying to think of Are there other factors that might come into play here and trying to like do some science to sort of suss those out basically yeah and so to try to do that sussing out he starts checking like okay what if some of the other factors that we know are linked to autism could be contributing to that effect that he saw that 20 to 30 percent increased chance

because like we talked about in our earlier episode on autism we know that the age of the mom can affect the chance of whether her kid will be autistic.

Like older moms are more likely to have autistic kids.

Or if the mom has some other health condition or got sick during pregnancy, that might increase the chance for autism too.

Hmm.

And so those factors have been known for a while and are generally pretty well accepted.

Yeah, yeah, exactly.

And so Brian's team took another look at the data.

And when they started accounting for some of that stuff, like how old the moms were, the effects of cenamenophen got smaller.

It went from increasing the chance of having an autistic kid by 20 or 30 percent down to 5 percent.

When it's down to like a 5% increase in risk, you start to actually wonder whether or not it's real or not.

Could it be like moms who are older are more likely to take Tylenol because they're like, have more aches and pains or something like that?

And we know that's separately associated with autism, that kind of a thing.

Yeah, exactly.

Or a mom being sicker, right?

Like you don't just take Tylenol for funsies.

Yeah.

So the other studies weren't adjusting for stuff like age of the parent,

you know, other health conditions, stuff like that.

Yeah, basically they weren't all doing the same type of adjustments.

Okay.

So that's, that's one thing.

And then there's another massive factor that contributes to whether someone's autistic, their genes.

Oh, yeah.

Yeah.

They think that this is like the number one thing that determines whether someone's autistic.

It's like 70 to 90 percent of it.

So Brian's team wanted to know, could that be part of the story too?

And that's sort of tricky to sort out.

Like ideally, you'd look at identical twins, right?

And expose one to acetaminophen in the womb and then see if there's like a difference in autism rates.

I guess.

I don't know how you would do that.

Yeah, exactly.

You can't do that.

But there's another way to get at this.

They could look at siblings in their data.

What if we did this within the same family unit?

So basically kids born to the same mom, but during one pregnancy, the mom used acinamanfin and the other pregnancy, the mom didn't, right?

So then you control at least partially for genetics.

Yeah, that's great.

So Brian's team looked in their giant Swedish database and gathered thousands of cases like this.

They crunched the numbers and here's what they found.

Everything flatlined in terms of risk.

You know, what was already a very minimal increase in risk just flatlined to absolutely no difference in risk for a child who is exposed.

No increased chance of autism.

No increased risk for autism, for ADHD, for intellectual disability.

It was just a pure flatline.

Yeah, so what this means is that within a family, whether or not the mom took acetamenophen, the chance that her kid was autistic was the same.

So can you put this all together for me?

Like, what,

what's going on here exactly?

Okay, yes.

So, the more I learned about this, the more I realized that there could be a completely different explanation for why acetaminophen and autism might seem like they're linked.

Because, like, if you imagine, like, say you're a mom of an autistic kid, that means that you're also more likely to be on the autism spectrum yourself, right?

You probably have some of these autism genes too.

And being on the spectrum, having some autistic traits, that's actually been linked to having more pain when you're pregnant.

It's also been linked to having migraines.

Yeah.

And so, like,

we have these reasons why people on the spectrum might be more likely to take acetaminophen.

So maybe that's why it's looking like it's the medicine causing the autism.

But we have a way of explaining all of this without blaming the acetaminophen at all.

So, all right.

So, this is like a classic sort of red herring science story.

Yeah.

So like it looked like it was this one thing, but it's actually that one thing is associated with this other thing that it was sort of covering up.

Yeah, that could really be what's going on here.

I mean, it would be like if you looked at your chance of playing outdoor pickleball,

that might go up along with how likely you are to have had ice cream recently.

But really, both of these things are just linked because it's warmer outside uh-huh

yeah exact exactly that's what i mean yeah it's like they're not act directly influencing the other it's just like they're both responding to the same third variable yes there's a third variable um exactly class that's what it seems like to me um

But there is still like a back and forth here, a bit of a nerd fight.

Like

other researchers saying Brian's study isn't perfect either.

And there are still reviews coming out that say there is a link between acetaminophen and stuff like autism.

So I asked Brian about these studies.

I guess like normally I would trust a meta-analysis looking at a bunch of studies over

like a couple studies here and there saying something else.

Like, why should I trust your study more than this meta-analysis?

So one of the fun things with the meta-analysis is that it's a garbage in, garbage out kind of a thing.

You can feed it garbage, but your meta-analysis is not going to turn this garbage into something beautiful.

You can repeat the wrong thing many times in completely different samples and get the same wrong results in every single sample.

And yeah, it's not like he's saying that these are all terrible scientists,

but more that these reviews are mostly looking at studies that didn't have that genetic stuff.

Like they couldn't look at siblings.

They just didn't have that data.

And so they couldn't do what Brian's team did.

But Brian's study, it's not the only one that did do that kind of like sibling analysis.

And so, like, there's one that just came out a couple weeks ago from Japan that did the same thing and it backed up Brian's study.

Okay, so given all that,

what do you make of all this?

Like,

if you were pregnant right now, would you take Tylenol?

Well, for the autism question, I do find studies like Brian's to be the most convincing.

Like, I don't think there's a clear link here.

And even to go back to those other concerns around Tylenol, like that idea that it might be a hormone disruptor.

Oh, the undescended testicles.

Exactly.

What about the undescended testicles?

And to go back to those, those studies on the undescended testicles are super mixed.

So like some studies will find a link there other studies don't and actually the most recent review that i could find on this said that most studies don't find more undescended testicles in babies exposed to acetaminophen and utero so even the the thing that started this whole suspicion of acetaminophen during pregnancy off

that has actually upon further investigation not really borne out like there there might not actually be a link there either no yeah exactly And a bunch of big medical groups agree here.

They say that we don't have evidence that taking acetaminophen during pregnancy causes autism or issues with the development of the fetus, like the FDA, regulators in Europe, a big group of OBGYNs.

But like, yeah, just in case, if I did take acetaminophen, I'd try not to take it for like weeks at a time.

Because generally, some of the studies that do find risks only find them if you take acetaminophen for like longer periods of time.

It also depends I guess on what you're taking it for.

I mean I'm not a I'm not a doctor.

This is not medical advice, but it does make me wonder like

if you have a high fever,

that might be worse for the fetus than the like infantismal risk of some Tylenol on their brain development, right?

Yes, actually, that's true too.

So having a fever when you're pregnant, we know that ups the chance for for lots of serious stuff like birth defects for the kid, heart malformations,

premature births.

And fevers also happen to be linked to a kid being autistic later.

Oh.

Okay, that's very interesting and fairly satisfying, actually.

Yeah.

All right.

And by the way, we did reach out to the company that owns Tylenol, and they sent us a big statement pointing to some of the science that we've talked about.

And they said, quote, we continue to believe that taking acetaminophen does not cause neurodevelopmental disorders, including autism, unquote.

And for Brian's part, he really doesn't think the science supports the idea that taking a little Tylenol when you're pregnant makes the kids later become autistic or have ADHD.

Like taking off my epidemiologist hat and just, you know, as a person hat, like blame

is damaging, it's harmful, like, and it doesn't exactly help.

People might be blaming themselves for like that one random Tylenol they took when, you you know, they were pregnant 10 years ago.

And, you know, if we can help, you know, take away some of that guilt that's there, I think that's, that's nice because people shouldn't be feeling guilty about these things in the past.

Yeah, feel guilty about something else.

Go back to feeling guilty about eating lunch meat.

Yeah.

All right.

Well, thanks, Rose.

Thanks, Meryl.

That's science versus.

All right.

Rose, do you want to ask me how many citations I have in this episode?

Oh, yeah, Meryl.

How many citations do you have in this episode?

We have 59 citations.

Woo!

So, if you want to see all the science that went into this episode, check out the transcripts.

You can find links to that, which has all these beautiful citations in the show notes.

And we did reach out to the U.S.

Department of Health and Human Services about this, and a spokesperson said,

until we release the final report, any claims about its contents are nothing more than speculation, unquote.

And if you want to learn more about autism, check out our other episode on this called Autism, The Real Reason It's Going Up.

And if you want to tell us what you think of this episode or have episode ideas, reach out to us.

Our Instagram is science underscore VS.

This episode was produced by me, Meryl Horn, with help from Rose Rimmler, Michelle Deng, and Aketi Foster Keys.

We're edited by Blythe Sorrell.

Mix and sound design by Bumi Hidaka.

Our executive producer is Wendy Zuckerman.

Fact-checking and consulting by Erica Akiko Howard.

Music written by Peter Leonard, Bobby Lorde, So Wiley, Emma Munger, and Bumi Hidaka.

Thanks to all the researchers that we spoke to for this episode and our other episode on autism.

Science Versus is a Spotify Studios original.

Listen for free on Spotify or wherever you get your podcasts.

And if you do listen on Spotify, follow us and tap the bell for episode notifications.

Adobe Acrobat Studio, so brand new.

Show me all the things PDFs can do.

Do your work with ease and speed.

PDF spaces is all you need.

Do hours of research in an instant.

With key insights from an AI assistant.

Pick a template with a click.

Now your preso looks super slick.

Close that deal, yeah, you won.

Do that, doing that, did that, done.

Now you can do that, do that with Acrobat.

Now you can do that, do that with the all-new Acrobat.

It's time to do your best work with the all-new Adobe Acrobat Studio.

Oh, the car from Carvana's here.

Well, will you look at that?

It's exactly what I ordered.

Like, precisely.

It would be crazy if there were any catches.

But there aren't, right?

Right.

Because that's how car buying should be.

With Carvana, you get the car you want.

Choose delivery or pickup and a week to love it or return it.

Buy your car today with Carvana.

Delivery your pickup fees may apply.

Limitations and exclusions may apply.

See our seven-day return policy at carvana.com.