Emily Oster

8 minute read Emily Oster

Emily Oster

Phthalates (THAL-ates) and Risk

Do they really cause reproductive and behavior problems?

Emily Oster

8 minute read

I am a person who worries a lot. Sometimes people find this surprising, since my book titles contain phrases like “more relaxed parenting.” But in practice I participate personally in less relaxed parenting and less relaxed work and less relaxed everything else. This is how I know to lecture other people on being more relaxed. I’m worrying for you!

Worries come in many flavors. There are things you should worry about (my devoted reader mother recently suggested I could devote a column to these), like being near a pool with a kid who cannot swim. There are things you shouldn’t worry about at all, like whether it’s okay for your kids to watch 30 minutes of TV so you can take a shower (yes, it’s fine, you could even take a long bath).

And then there are things — and, honestly, this is a lot of things — which are of some concern — maybe a lot of concern — but you basically cannot control. Like asteroids. Definitely, a big asteroid hitting the earth would be really bad. It’s unlikely but not completely impossible. But the thing is you cannot do anything about it (I mean, probably not, unless that’s your job). If you walked around all day worrying about asteroids, this would just make you upset and anxious for no reason.

One of the key things I have learned from worrying a lot, and trying to worry less, is that it’s often a very good idea to think not only how concerned should I be? but, also, Is this in my control?

Which brings me to to today’s topic: phthalates.

Background and Science on phthalates

Last weekend, writing in the New York Times, Nick Kristof wrote about declining sperm volume and sperm quality, suggesting a key role for endocrine disrupters (specifically, phthalates) in driving the decline. The article has some alarming notes. For example, the person who is quoted saying: “I do not see humans becoming extinct, but…” Wait, what?

At the same time, over at CNN, phthalates were getting play for a possible role in behavior problems in children and Annie’s Mac and Cheese was pledging to remove them from their product.

For those of you who are not, say, chemistry majors: phthalates are a class of chemicals which are added to many, many plastics to make them more flexible. They are ubiquitous in the products we use and evidence of them (in the form of metabolites) appears in most tested urine. Many concerns have been raised that they have health impacts, notably on reproductive health and cancer.

This discussion is not new. Concerns about these and related chemicals have been around for decades, and it’s easy to find a lot of sources suggesting you should be concerned. But how concerned? Extinction-of-humans panic level? I- can’t- remember-which-jacket-my-keys-are-in-panic-level? Somewhere in between?

The Kristoff op-ed this weekend was largely about declining sperm counts. He discussed some other issues, although trends in those are less well measured. (For example, the the finding of increased miscarriages likely reflects early pregnancy detection, not an actual trend).

However: the change in sperm counts is real. Sperm counts have been dropping over the past decades. This basic time trend is not difficult to study (sperm is easy to get) and there are good articles (like this one on Chinese sperm donors) which demonstrate large chances in sperm count and quality over time. The extent to which sperm count impacts fertility is not completely clear. There are studies which show that while older men have a lower sperm count, they are no less likely to fertilize eggs in IVF. Of course, in the limit, you have to imagine it does matter for fertility.

Despite general agreement about the presence of this decline, the reasons are poorly understood. There are a lot of things which affect sperm count and quality. Here’s an excellent review article, which cites all of the following possible influences: alcohol, smoking, caffeine and related, illegal drugs, diet, supplements, being overweight, being underweight, heat, zinc, copper, selenium, water pollution, other pollutants, pesticides, phthalates, glycol ethers. I myself have written about sperm and tight underwear in the past.

How important are phthalates, specifically? It’s really hard to know. Research which studies this measures evidence of phthalate exposure through urine testing, and then correlates the concentration of various phthalate metabolites (by-products of exposure) with outcomes. The evidence through these approaches is a bit mixed and depend on the method of measurement and the particular metabolite considered. Moreover, the extent of exposure to phthalates is likely correlated with other lifestyle and environmental risks, making the individual association difficult to untangle.

This last point is relevant, as well, for the research on phthalate exposure and neurodevelopment in children. As that CNN article noted, there are some studies which have correlated phthalate exposure in pregnancy and childhood with neurodevelopmental problems (behavior problems, cognitive measures). But these results are inconsistent across studies.

A review article from 2019 suggests some overall negative correlation between phthalate exposure and child outcomes, but the results vary widely across studies. In some cases the correlations are positive, not negative; sometimes they matter for boys and not girls, sometimes the other way around; some studies find effects for one metabolite but not others, and so on. Moreover, these studies aren’t generally able to adjust fully for other individual and family characteristics, leaving us as usual unsure about causality. The authors of the review article note that this is a situation ripe for publication bias (in that it may be easier to publish significant results than ones with zero effect).

It is worth layering on top of all of this the fact that there has been a concerted effort in a variety of settings to decrease phthalate usage. It’s not just Annie’s! This has been phased out of many children’s toys over time, and many retailers have pledged to eliminate these from their products. Even Taco Bell is in on this elimination.

On the flip side, phthalate exposure is ubiquitous. They’re everywhere. You can read posts like this on how to avoid (don’t microwave your meat with plastic wrap, use glass jars) but even those sites acknowledge that avoidance is basically impossible. Nearly everyone has phthalate exposure to some extent. Even the people who are really, really good about using glass jars to store stuff in.

Stepping back to the phthalate worry train

Putting this together, here is where I come out.

  • Sperm counts are declining, a topic which deserves better understanding. That review article I cited above notes at the end that getting better data on this should be possible and prioritized. Given the 10 to 12 week timeline for sperm production, one could imagine studies which evaluated at least the impact of lifestyle interventions.
  • The precise causal role of phthalates in any of this is unclear. Based on our understanding of the endocrine system and animal models, it seems likely that phthalate exposure has some impacts on the reproductive system. But the size of these effects remains really ambiguous. The effects could be there, but very small relative to lifestyle choices. Or they could be larger.
  • The neurodevelopment literature is subject to a lot of causality and probably p-hacking issues and, even if present, any causal effects are likely extremely small.
  • There is really no way to avoid phthalates completely, and we have virtually no idea how much something like switching to glass storage containers would matter.

In a sense, thinking this through put me in mind of the discussion of listeria in Expecting Better. Listeria is a pathogen which is a source of many food restriction in pregnancy (raw milk cheese, for example). Listeria infection during pregnancy is very rare, but very dangerous and should be avoided if possible. This argues for caution, but the point I make in Expecting Better is that the source of listeria is very unpredictable. It isn’t clear that (for example) avoiding ham in your sandwich is going to limit your risk very much.

Phthalates are obviously nowhere near listeria in terms of danger, so there’s no comparison there. But where I think there is a comparison is in the avoidance question. Other than advocating for policy-makers or companies to embrace reductions in phthalates, it isn’t clear what you can do about your exposure. The possible downsides are real, but not infinitely large. It simply isn’t clear what the value is to worrying about this.

At the end of Cribsheet I tell a story about bees. In brief, it is a story of the time I forced our lovely pediatrician to listen to my lengthy narrative panic attack about whether we needed to bring an epi pen on vacation just in case Penelope was stung by a bee and it turned out (despite no reason to think this) that she was allergic. Dr. Li paused and then delivered what I still think was the best parenting advice I’ve gotten: “Yeah, I’d just try not to think about that.”

It wasn’t that bee allergies weren’t a real thing, just as it is not necessarily true that phthalate exposure is completely benign. But there really wasn’t anything to do about this, and in the space of all the other concerns one might have, this was small. So, maybe, just try not to think about it.

Community Guidelines
Fitness bands and rollers for stretching on a pink background.

Feb. 24, 2023

2 minute read

What Are the Best Pre- and Post-Running Stretches?

Ask ParentData

A line graph with pink, yellow, and blue dots representing life's ups and downs.

Feb. 21, 2023

3 minute read

Wins, Woes, and Autism

Your stories for the week

A pregnant person is supported by a partner in a pool of water for a water birth.

Feb. 23, 2023

7 minute read

Labor Positions

And what I might have done differently

Close up of a postpartum belly with baby hands.

Feb. 17, 2023

2 minute read

Should I Try Stomach Binding Postpartum?

Ask ParentData