I'm surprised you're not giving air pollution higher billing. Alex Tabarrock made a 5 minute video about a difference-in-difference paper showing car exhaust having all sorts of bad outcomes. And when I looked at the cited study right now it claimed there exists a "somewhat smaller literature focusing specifically on the relationship between residential proximity to busy roadways and poor pregnancy outcomes". I have no idea how tractible this stuff is or whether it necessitates e.g. moving apartments during the pregnancy if you're near a freeway, or whether there are easier interventions.
If older kids does better in school, could not a nation improve the education of their entire population by delaying the age kids start in school? So instead of people going to school for example from age 6 to 18, they go to school from age 7 to 19.
Just a random point about licorice: perhaps it's a self-feeding cycle, but if you get a candy bag in Finland and it's not specifically "fruit candy only", "chocolates only", self-selected etc. it usually contains at least 1/2-1/3 black licorice candies. That might contribute to the high ratio of formally consumed licorice here.
> Most of these studies are small and weak - and did you know that Tylenol is a deadly poison to cats? Doesn’t really bode well for inter-animal transferability of results.
Mom of a college student here. My 1st thought upon browsing this, was "I wonder how many things I did wrong". Even so many years later, this instinct is strong! My son had many food allergies and asthma, all of which I squarely blamed myself for (things like peanuts and eggs that I ate in pregnancy, despite being from a family with food allergies - just didn't know better).
The one my mother always mentioned (and yes I realize that sounds like a totally random source, but in this case my mother has a doctorate in maternal & child health from Johns Hopkins and worked in the field for many years) was the benefit of a [healthy] high-calorie diet during pregnancy. The idea is that evolutionary factors would then 'teach' [is this epigenetics?] the fetus that it lived in times of plenty, leading to a higher metabolism, which is a good thing in the modern world. Anecdotally my siblings and myself are all thin without any real effort, although my mother (albeit not my father) is also like that so it could be purely genetic. Has anyone run across this idea elsewhere, preferably with some data?
One thing that you might want to mention is the difference between the *dosage* of different types of choline -- i.e., 500 mg of choline bitartrate is not the same as 500 mg of "choline", and in particular is not the same as 500mg of choline chloride. Caudill et al uses Choline chloride, while Jacobson et al uses Choline bitartrate -- and therefore the dosages aren't directly comparable, so you probably shouldn't be directly comparing them.
It may seem like a no-brainer to do embryo selection via SNPs, if you're already going to do IVF. It's a much different story if you weren't planning on IVF in the first place. IVF is associated with higher de novo mutation counts above the rate expected based on parental age. It has an effect equivalent to roughly 6 years of paternal age or 12 years of maternal age in Wong et al. 2016.
We don't know why that is, but my pet hypothesis has to do with the uterus as a selection arena, as described by Stearns (1986). So, normally these embryos might have been weeded out, but not in vitro.
Re paternal age: Carslake et al. https://www.nature.com/articles/srep45278/tables/3 is the biggest fairly properly done study of paternal age on IQ that I know and finds significant negative effects of around half an IQ point per year, so 10 IQ points more from 20-year-old you vs. 40-year-old you. But the same study found a positive effect on having secondary education and in these Swedish studies it's extremely important how they deal with intellectually disabled people (who will not have been conscripted and hence done the IQ test). There are very clear paternal age effects on many developmental disorders that include some form intellectual disability, see the graph here https://www.nature.com/articles/nature21062
You can freeze your sperm now and some people do. I haven't found literature that investigated whether offspring from frozen sperm have more mutations. Obviously, freezing might cause problems but then again not enough for sperm banks and cattle insemination not to work.
Also re natural experiments to study stress effects on offspring, see these couple of studies of Ramadan effects. But of course the much more plausible cause is nutrition. So, maybe don't fast during pregnancy. AFAICT it's not really required, but some do so anyway.
I want to raise a methodological issue here. In the section on stress and confounds, Scott writes:
> only randomized controlled trials, or other studies that come up with clever ways around this problem, can be truly convincing
There's more than one way of reading that, depending on what you count as a 'clever way around a problem'. I want to talk about a worryingly common reading, which misrepresents how research actually works. And that's to throw away *any* study that's not causal.
That criterion certainly addresses a real problem: there's too much research there which finds a random correlation (or p-hacks one out) and leads to splashy, misleading headlines about causal effects. The problem is that it throws the baby out with the bathwater -- you'd end up rejecting, for example, continental drift.
What determines how strongly it's appropriate to apply the criterion? I'd argue that a key element is the presence/absence of background theory. In the case of child development, there are theories (attachment theory and life history theory) which have made remarkably good predictions to date. The other predictions of life history theory, including on how maternal stress affects children, then become highly plausible; when correlational studies fit those predictions, it's not appropriate to discard them. I make a more detailed argument on this with examples here: https://criticalscience.medium.com/against-parachute-skepticism-516414ee1815
More generally, my point is that science works with an interplay of theories and experiments, and later experiments always build on previously accepted theories. When we have findings that are divorced from theories, we should be very skeptical and apply that 'causal' criterion heavily. And vice versa. At the end of the day, the goal is always to find a parsimonious explanation of all the facts -- and deciding to ignore (replicated) definitely does not do that.
The reason I feel strongly about this is that in child development, it's just very hard to get causal data on children. Many commenters (not Scott, but certainly Emily Oster) will use 'correlation is not causation' to dismiss any evidence that doesn't fit their political preferences. But that ignores the associated theories and the wealth of evidence that supports them. [If you want to read about that evidence, see particularly The Development of the Person (Sroufe et al, 2005).]
I've been worrying over this issue for the last year, as I've been writing up topics in child development for a general audience.* In a few cases, we have some causal evidence and it's straightforward. Far more often, the state of the research depends on on a convergence of theory and multiple streams of research. For example, we might have evolutionary theories predicting that a phenomenon P will occur, biochemical evidence showing how P is actually 'implemented', psychological/psychiatric evidence that P occurs in the wild and studies that show P-ish correlations. That adds up to a very strong case; IMO often *more* convincing than an isolated causal study. But it's much harder to convey that case to a lay audience. If anyone has any thoughts on how to handle this, I'd be very grateful for them.
In the original edition of his Biodeterminist’s Guide To Parenting, Scott drew the opposite conclusion about the association between birth month and intelligence: "Study after study shows higher intelligence (2-4 points), more favorable personality traits (for example, 33% less extreme shyness) and greater height (0.6 cm) in winter/spring babies (1, 2, 3, 4)." Of these four cited studies, two links have broken since the original post was written, and the other two studies are not referenced in the current post. Why were the studies showing higher intelligence in winter/spring discarded in favor of those showing the effect in summer/fall?
Excellent post. I'm heartened to see some of the things we did after reading up for both our kids (delaying induction, reducing Tylenol, filtering water) be borne out as possibly useful. The hardest bit in pregnancy (and also in early childhood) is to have an actual evidence based conversaion with a doc - its either "don't worry here's Calpol" or "holy F let's intervene" with no gears in the middle. Also, not sure if its just the UK, but I find getting actual information from docs to be like running an interrogation chamber.
eg here the normal practice seems if you're above 37 weeks to induce if there's even the slightest question from the mother without much other analysis - we had to push back quite a bit on not doing this willy nilly and instead wait a week or two.
Thank you for the interesting review. It would be awesome to do something similar for the morning sickness too. Most of women have better or worse morning sickness during pregnancy, are there any interventions that plausibly impact its severity?
I don't know, we got rid of all the old Lenten and Advent fasting, and now the health and nutrition people are telling us we should be eating less red meat, eat more fish, cut back on sugar, etc. just like in the days of the Black Fast:
For the late 19th century/mid 20th century less intense version of this, the Black Fast was no sugar or milk in your tea, no butter on your bread, one full meal only but you can also take two collations, no meat, etc.
Small comment on the last table - you've marked the columns as "% True" and "% Relevant" but the numbers aren't percentages, right? (Otherwise you consider every single result true with less than 1% probability which seems unlikely)
I'm amused that you just give people the go-ahead to chow down on corn syrup derivatives with arbitrary (unspecified on the label) colorants and flavorants. (Unless I'm missing some sarcasm.)
I recently wrote a blog about Expecting Better pointing out statistical problems with how Oster reaches her conclusions that light drinking during pregnancy is fine:
While I think there’s a lot of uncertainty, even the studies Oster prefers seem like mild evidence in favor of a negative effect of light drinking once you account for the known biases
Perhaps I misread the tone on the receipt/cashier point, and this one is low certainty, but "it's only cashiers, so not a big deal" seems like the wrong reaction to me: there are many millions of people who are cashiers, many of whom are young women. This is a very large and relevant group of people, even at a population level!
is vitamin d deficiency really that rare? basically everyone in my family has had it and from what i heard it was more 'almost everyone who's not white will probably be deficient in the winter'.
Currently going through this with my wife and trying to imagine a doctor or other authority figure telling her she can't have any painkillers at all, lol.
Probably just a coincidence but it's odd that she's been taking Vitamin D supplements and eating a lot of (low mercury risk) fish "because it makes me feel better," and then I read this. For my part I am just letting her figure out how to manage her pregnancy because I think stress is our biggest concern and also I love her and trust her. Still very interesting stuff.
It's worth checking whether your prenatal vitamin actually has iodine in it - the first one I took didn't. Your provider won't get into specifics, they'll just tell you to take one with folic acid.
On water filters, if you want a pitcher rather than a sink type, the Zero Water pitcher filters out fluoride (and a lot of other stuff, if the dissolved solids reader that comes with it is correct). I sent our usual tap water and some filtered water to the Maine state lab that tests well water and such, and the tap water came back with a reasonable level and the filtered water had 0 measurable fluoride.
I also got my first cavity after switching to water without fluoride during pregnancy, so I now think it's worth doing some fluoride mouthwash or something you don't swallow to make up for not having it in the water.
I think you have a typo, 1/200 for CMV should be 0.005, not 0.0005 (as you have in your table). So your estimate is ten times too low. (Edit: unless you think only 1/10 of congenital CMV cases lose IQ?)
Also there may be some confusion if you write "%" in the top of the table but then give the numbers as a fraction from 0 to 1.
Finally, regarding Tylenol vs ibuprofen, my prior is on Tylenol being generally worse due to more toxic metabolites.
But really any cyclooxygenase inhibitor will mess up prostaglandin signaling to some extent, which will interfere with some developmental processes.
When traveling in Iceland I noticed almost all of their candy contains black licorice. It would be interesting to see if they have similar results as Finland.
I think the Tylenol studies that are not sibling-studies are very likely to be unreliable, for one reason: The ability to avoid pain medication during pregnancy is a clear indicator of social privilege.
There is a clear social norm that all pills should be avoided during pregnancy, as long as it is possible. Pregnant women in more privileged situations will be able to follow that norm a lot more than pregnant women in more pressed situations.
I'm an example of this myself. When I was pregnant last year I suffered from recurring pains during the last two months. Sometimes I couldn't even speak because of the pain. The doctor said I should take Tylenol. I didn't, because I didn't really need it: When I had my pain attacks, I was free to just deal with them. I didn't have to do anything else, really. My husband always said "I'll do it, you just rest" whatever had to be done.
If I were in a situation where I needed to work or take care of children alone, I would have taken the Tylenol. I don't even believe much in the studies that say it is risky, because sibling studies tend to be much more reliable. If my environment made life the least difficult for me, I would have taken the pills to cope. But life was easy anyway, so I stayed away from them. Whatever put me in this privileged situation, I hope my children will inherit it.
On the subject of birth months and rates of schizophrenia, has there been any studies to see if the trend is reversed in, say, Australia which has the opposite flu season to the US / Europe?
Personally a bit skeptical that birth month has huge effects although the flu season is the most reasonable explanation.
Please do not promote eugenics. Also, it is disturbing how carelessly you speak of "embryos," they are also people, with rights. Whatever their potential IQ may be.
Re licorice: Does anyone know if licorice root tea also has the same problem? Because I'm pretty sure every tea that actually works to make your throat feel immediately better contains licorice root. Since I read your original article about pregnancy and licorice, I decided to wean myself off of all of them (I used to drink them because I was taking vocal lessons and sometimes you just want to soothe your throat) in case I decide to get pregnant...
Could Cytomegalovirus explain some of the birth order effects? Women are typically pretty careful with avoiding infecting when pregnant with their first, but once they have a toddler running around there is only so much they can do. So subsequent pregnancies are probably more likely to have CMV infections than first pregnancies, and the first born are less likely to be damaged than subsequent babies.
So I was first in my class in AP chemistry which qualifies me to look up some molecular masses and check these papers' dose-math.
The "2g choline" study actually made a serious error in dose-conversion when they said "Each choline supplement dose consisted of 1.25g choline bitartrate, which contained 1g of bioavailable choline cation." It's actually around half that. So they were actually testing 1g of free choline.
choline bitartrate molecules weigh 1.81 times more than choline chloride and 2.53 times more than free choline. Each of these molecules contains only one choline. Convert doses accordingly.
I think it's fine to take 2g of choline bitartrate to get ~790mg of choline. But the largest pills on amazon are only 650mg choline bitartrate per pill.
In general a dose is meaningless without specifying which form the dose-mass refers to. Someone could be off by 30x if for example one number is free lithium and the other number is lithium citrate and they forget to convert.
This is really basic stuff that everyone learns in their first chemistry class so it's disappointing that a paper got past peer review while overstating its choline dose by 2x.
The nullification / "reversal" in the age advantage amongst elite athletes may be because if you are a really skilled athlete born in the end of the year and are the youngest on your team but still good enough, at some point, age stops mattering and your natural ability level shines through. Thus, we'd expect the elite not to see much of a difference because those who are good enough can still play; you're weeding out the merely really good but not quite superelite players.
I would be skeptical of a reversal, though one possibility would be that constantly playing at a disadvantage gives them a slight training advantage, so they end up a bit better than they would otherwise if they actually stick with it.
Wittgenstein said that by looking at language, he let the philosophical fly out of the bottle and thus ended the maddening work, and this post strikes me as doing the same for pregnancy optimization.
Whether or not that is your intention, it's a thought-provoking post (and makes me glad I'm well past the age to worry about any of it).
On the birth month thing and sports, one of the things that defined my socialization was being born on the cusp of our rec. league baseball program's cutoff date. I alternated year to year being a bench-sitting, error-prone right fielder, to being a star pitcher. I didn't play past high school, but I learned many life lessons about failure and success.
I would estimate that the birth month disparity would have some impact on IQ. Schizophrenia generally has far more of a detrimental effect on IQ than autism spectrum disorders. And theoretically, a child born up to ~11.9 months earlier than their peers would perform slightly better academically and moderately better athletically, thus receiving more encouragement from teachers. But this would obviously be very difficult -if not impossible- to quantify with currently available data.
The idea that fish oil is defective isn't a stretch at all. We know what we want from fish oil: omega-3 fatty acids, so we measure it and they usually don't have it. And that shouldn't be so surprising because this is specifically the kind of fat that oxidizes easily and so that's just what it does.
Aluminum itself is a neurotoxin although it has low solubility in water and low oral bioavailability. Should I be concerned about the quantity of aluminum that an activated alumina filter would put in the water?
The major biodeterminist parenting intervention is to have kids young. If that's not going to happen, freeze your eggs or sperm. Feels weird recommending something I didn't do, but I would do it if I were young now - especially for men, probably for women.
To clarify "avoid CMV", I think a solid piece of advice is "don't start a job in healthcare or at a school/daycare" right when you're trying to conceive, because newcomers tend to get sick quite often when they start working in these environments (with CMV & everything else--nearly all bedside nurses have had CMV)
Another confounder for birth month is whether or not your planning is effective. In practice, planning the birth month of your child is really pretty tricky, as a lot of people don't get pregnant on the first try (or even second, third, etc). One would assume that a better ability to plan this intentionally might be due to higher fertility in general, which could be correlated to a lot of things, including a healthier embryo/fetus and subsequently higher IQ.
I respect the dedication for exploring all these, to end up with 2 IQ points equivalent. What a journey. (Also, that Elon Musk decompression baby theory-not-theory is a gem!)
Thanks so much for doing this (both research and communication) - wife & I are implementing some of these which we probably wouldn't have without this article.
wait, arent the Finns doing really really well overall?
"Glycyrrhizin" sounds like a terrible Scrabble hand.
Epistemic status: saw a video once.
I'm surprised you're not giving air pollution higher billing. Alex Tabarrock made a 5 minute video about a difference-in-difference paper showing car exhaust having all sorts of bad outcomes. And when I looked at the cited study right now it claimed there exists a "somewhat smaller literature focusing specifically on the relationship between residential proximity to busy roadways and poor pregnancy outcomes". I have no idea how tractible this stuff is or whether it necessitates e.g. moving apartments during the pregnancy if you're near a freeway, or whether there are easier interventions.
Here's the video: https://marginalrevolution.com/marginalrevolution/2021/11/pollution-kills-and-more.html
Here's the paper: https://www.nber.org/system/files/working_papers/w15413/w15413.pdf
If older kids does better in school, could not a nation improve the education of their entire population by delaying the age kids start in school? So instead of people going to school for example from age 6 to 18, they go to school from age 7 to 19.
Just a random point about licorice: perhaps it's a self-feeding cycle, but if you get a candy bag in Finland and it's not specifically "fruit candy only", "chocolates only", self-selected etc. it usually contains at least 1/2-1/3 black licorice candies. That might contribute to the high ratio of formally consumed licorice here.
> Most of these studies are small and weak - and did you know that Tylenol is a deadly poison to cats? Doesn’t really bode well for inter-animal transferability of results.
Given that Tylenol is also a notoriously deadly poison to humans (see: https://en.wikipedia.org/wiki/Paracetamol_poisoning ), I don't think this has the significance you think it does.
Mom of a college student here. My 1st thought upon browsing this, was "I wonder how many things I did wrong". Even so many years later, this instinct is strong! My son had many food allergies and asthma, all of which I squarely blamed myself for (things like peanuts and eggs that I ate in pregnancy, despite being from a family with food allergies - just didn't know better).
Thyroid!
Both hypothyroid conditions and hyperthyroid conditions during pregnancy can be bad.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6209822/
The one my mother always mentioned (and yes I realize that sounds like a totally random source, but in this case my mother has a doctorate in maternal & child health from Johns Hopkins and worked in the field for many years) was the benefit of a [healthy] high-calorie diet during pregnancy. The idea is that evolutionary factors would then 'teach' [is this epigenetics?] the fetus that it lived in times of plenty, leading to a higher metabolism, which is a good thing in the modern world. Anecdotally my siblings and myself are all thin without any real effort, although my mother (albeit not my father) is also like that so it could be purely genetic. Has anyone run across this idea elsewhere, preferably with some data?
Scott, did you see my Choline review?
https://old.reddit.com/r/slatestarcodex/comments/96vs5d/choline_supplementation_during_pregnancy/
One thing that you might want to mention is the difference between the *dosage* of different types of choline -- i.e., 500 mg of choline bitartrate is not the same as 500 mg of "choline", and in particular is not the same as 500mg of choline chloride. Caudill et al uses Choline chloride, while Jacobson et al uses Choline bitartrate -- and therefore the dosages aren't directly comparable, so you probably shouldn't be directly comparing them.
Re IVF:
It may seem like a no-brainer to do embryo selection via SNPs, if you're already going to do IVF. It's a much different story if you weren't planning on IVF in the first place. IVF is associated with higher de novo mutation counts above the rate expected based on parental age. It has an effect equivalent to roughly 6 years of paternal age or 12 years of maternal age in Wong et al. 2016.
We don't know why that is, but my pet hypothesis has to do with the uterus as a selection arena, as described by Stearns (1986). So, normally these embryos might have been weeded out, but not in vitro.
Re paternal age: Carslake et al. https://www.nature.com/articles/srep45278/tables/3 is the biggest fairly properly done study of paternal age on IQ that I know and finds significant negative effects of around half an IQ point per year, so 10 IQ points more from 20-year-old you vs. 40-year-old you. But the same study found a positive effect on having secondary education and in these Swedish studies it's extremely important how they deal with intellectually disabled people (who will not have been conscripted and hence done the IQ test). There are very clear paternal age effects on many developmental disorders that include some form intellectual disability, see the graph here https://www.nature.com/articles/nature21062
You can freeze your sperm now and some people do. I haven't found literature that investigated whether offspring from frozen sperm have more mutations. Obviously, freezing might cause problems but then again not enough for sperm banks and cattle insemination not to work.
Also re natural experiments to study stress effects on offspring, see these couple of studies of Ramadan effects. But of course the much more plausible cause is nutrition. So, maybe don't fast during pregnancy. AFAICT it's not really required, but some do so anyway.
https://gh.bmj.com/content/4/3/e001185.abstract
https://www.sciencedirect.com/science/article/pii/S1570677X18300911?casa_token=elsnjf_iNcQAAAAA:42zOuGYKU4X7NHqogMbv1XqO6XzOXaeZXZeDbl4Dix7v7AsPUG_-NzV3Sr5bZUPmgOPq4nRfKw
https://www.sciencedirect.com/science/article/pii/S030438781500067X?casa_token=ShzCnoJBtVUAAAAA:4lcMmOiNViEc795RQHNynyXHEcAP_HIJeAp5qe1Qe3Log5x5GXJOitw1BYKGX-ykqZr6TtAJlw
I want to raise a methodological issue here. In the section on stress and confounds, Scott writes:
> only randomized controlled trials, or other studies that come up with clever ways around this problem, can be truly convincing
There's more than one way of reading that, depending on what you count as a 'clever way around a problem'. I want to talk about a worryingly common reading, which misrepresents how research actually works. And that's to throw away *any* study that's not causal.
That criterion certainly addresses a real problem: there's too much research there which finds a random correlation (or p-hacks one out) and leads to splashy, misleading headlines about causal effects. The problem is that it throws the baby out with the bathwater -- you'd end up rejecting, for example, continental drift.
What determines how strongly it's appropriate to apply the criterion? I'd argue that a key element is the presence/absence of background theory. In the case of child development, there are theories (attachment theory and life history theory) which have made remarkably good predictions to date. The other predictions of life history theory, including on how maternal stress affects children, then become highly plausible; when correlational studies fit those predictions, it's not appropriate to discard them. I make a more detailed argument on this with examples here: https://criticalscience.medium.com/against-parachute-skepticism-516414ee1815
More generally, my point is that science works with an interplay of theories and experiments, and later experiments always build on previously accepted theories. When we have findings that are divorced from theories, we should be very skeptical and apply that 'causal' criterion heavily. And vice versa. At the end of the day, the goal is always to find a parsimonious explanation of all the facts -- and deciding to ignore (replicated) definitely does not do that.
The reason I feel strongly about this is that in child development, it's just very hard to get causal data on children. Many commenters (not Scott, but certainly Emily Oster) will use 'correlation is not causation' to dismiss any evidence that doesn't fit their political preferences. But that ignores the associated theories and the wealth of evidence that supports them. [If you want to read about that evidence, see particularly The Development of the Person (Sroufe et al, 2005).]
I've been worrying over this issue for the last year, as I've been writing up topics in child development for a general audience.* In a few cases, we have some causal evidence and it's straightforward. Far more often, the state of the research depends on on a convergence of theory and multiple streams of research. For example, we might have evolutionary theories predicting that a phenomenon P will occur, biochemical evidence showing how P is actually 'implemented', psychological/psychiatric evidence that P occurs in the wild and studies that show P-ish correlations. That adds up to a very strong case; IMO often *more* convincing than an isolated causal study. But it's much harder to convey that case to a lay audience. If anyone has any thoughts on how to handle this, I'd be very grateful for them.
* E.g. https://criticalscience.medium.com/on-the-science-of-daycare-4d1ab4c2efb4
-- in that case we happen to have good causal evidence, but the case was actually strong before that evidence came out.
In the original edition of his Biodeterminist’s Guide To Parenting, Scott drew the opposite conclusion about the association between birth month and intelligence: "Study after study shows higher intelligence (2-4 points), more favorable personality traits (for example, 33% less extreme shyness) and greater height (0.6 cm) in winter/spring babies (1, 2, 3, 4)." Of these four cited studies, two links have broken since the original post was written, and the other two studies are not referenced in the current post. Why were the studies showing higher intelligence in winter/spring discarded in favor of those showing the effect in summer/fall?
Excellent post. I'm heartened to see some of the things we did after reading up for both our kids (delaying induction, reducing Tylenol, filtering water) be borne out as possibly useful. The hardest bit in pregnancy (and also in early childhood) is to have an actual evidence based conversaion with a doc - its either "don't worry here's Calpol" or "holy F let's intervene" with no gears in the middle. Also, not sure if its just the UK, but I find getting actual information from docs to be like running an interrogation chamber.
eg here the normal practice seems if you're above 37 weeks to induce if there's even the slightest question from the mother without much other analysis - we had to push back quite a bit on not doing this willy nilly and instead wait a week or two.
Thank you for the interesting review. It would be awesome to do something similar for the morning sickness too. Most of women have better or worse morning sickness during pregnancy, are there any interventions that plausibly impact its severity?
Thanks, Scott. I'm going to start taking choline and finally plug in my air filters today.
So, back to fish on Fridays, make sure it's mackerel, cod and salmon. Trout if you can get it! 😀
https://www.mayoclinic.org/diseases-conditions/heart-disease/in-depth/omega-3/art-20045614
I don't know, we got rid of all the old Lenten and Advent fasting, and now the health and nutrition people are telling us we should be eating less red meat, eat more fish, cut back on sugar, etc. just like in the days of the Black Fast:
https://en.wikipedia.org/wiki/Black_Fast
For the late 19th century/mid 20th century less intense version of this, the Black Fast was no sugar or milk in your tea, no butter on your bread, one full meal only but you can also take two collations, no meat, etc.
Small comment on the last table - you've marked the columns as "% True" and "% Relevant" but the numbers aren't percentages, right? (Otherwise you consider every single result true with less than 1% probability which seems unlikely)
I'm amused that you just give people the go-ahead to chow down on corn syrup derivatives with arbitrary (unspecified on the label) colorants and flavorants. (Unless I'm missing some sarcasm.)
I recently wrote a blog about Expecting Better pointing out statistical problems with how Oster reaches her conclusions that light drinking during pregnancy is fine:
https://www.filedrawer.blog/post/oster_pregnancy_alcohol/
While I think there’s a lot of uncertainty, even the studies Oster prefers seem like mild evidence in favor of a negative effect of light drinking once you account for the known biases
Perhaps I misread the tone on the receipt/cashier point, and this one is low certainty, but "it's only cashiers, so not a big deal" seems like the wrong reaction to me: there are many millions of people who are cashiers, many of whom are young women. This is a very large and relevant group of people, even at a population level!
is vitamin d deficiency really that rare? basically everyone in my family has had it and from what i heard it was more 'almost everyone who's not white will probably be deficient in the winter'.
Currently going through this with my wife and trying to imagine a doctor or other authority figure telling her she can't have any painkillers at all, lol.
Probably just a coincidence but it's odd that she's been taking Vitamin D supplements and eating a lot of (low mercury risk) fish "because it makes me feel better," and then I read this. For my part I am just letting her figure out how to manage her pregnancy because I think stress is our biggest concern and also I love her and trust her. Still very interesting stuff.
I'm so excited for the return of this post!
It's worth checking whether your prenatal vitamin actually has iodine in it - the first one I took didn't. Your provider won't get into specifics, they'll just tell you to take one with folic acid.
On water filters, if you want a pitcher rather than a sink type, the Zero Water pitcher filters out fluoride (and a lot of other stuff, if the dissolved solids reader that comes with it is correct). I sent our usual tap water and some filtered water to the Maine state lab that tests well water and such, and the tap water came back with a reasonable level and the filtered water had 0 measurable fluoride.
I also got my first cavity after switching to water without fluoride during pregnancy, so I now think it's worth doing some fluoride mouthwash or something you don't swallow to make up for not having it in the water.
Breastmilk contains very little fluoride from mothers who drink fluoridated tap water, so it's not worth continuing to avoid after pregnancy. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2612944/
On the subject of pregnancy, does anyone here have strong opinions (medical or otherwise) about the site Evidence Based Birth?
https://evidencebasedbirth.com/
I think you have a typo, 1/200 for CMV should be 0.005, not 0.0005 (as you have in your table). So your estimate is ten times too low. (Edit: unless you think only 1/10 of congenital CMV cases lose IQ?)
Also there may be some confusion if you write "%" in the top of the table but then give the numbers as a fraction from 0 to 1.
Finally, regarding Tylenol vs ibuprofen, my prior is on Tylenol being generally worse due to more toxic metabolites.
But really any cyclooxygenase inhibitor will mess up prostaglandin signaling to some extent, which will interfere with some developmental processes.
Does this mean you and new wife are pregnant now or just trying?
If you're trying stop drinking totally and start taking pre-natals even though you ain't pregnant yet! :) that's my goofy advice.
Good and important content. Thank you!
When traveling in Iceland I noticed almost all of their candy contains black licorice. It would be interesting to see if they have similar results as Finland.
I wonder if there'll be an astralbaby.com in 30 years for people whose parents tried all of the above.
I think the Tylenol studies that are not sibling-studies are very likely to be unreliable, for one reason: The ability to avoid pain medication during pregnancy is a clear indicator of social privilege.
There is a clear social norm that all pills should be avoided during pregnancy, as long as it is possible. Pregnant women in more privileged situations will be able to follow that norm a lot more than pregnant women in more pressed situations.
I'm an example of this myself. When I was pregnant last year I suffered from recurring pains during the last two months. Sometimes I couldn't even speak because of the pain. The doctor said I should take Tylenol. I didn't, because I didn't really need it: When I had my pain attacks, I was free to just deal with them. I didn't have to do anything else, really. My husband always said "I'll do it, you just rest" whatever had to be done.
If I were in a situation where I needed to work or take care of children alone, I would have taken the Tylenol. I don't even believe much in the studies that say it is risky, because sibling studies tend to be much more reliable. If my environment made life the least difficult for me, I would have taken the pills to cope. But life was easy anyway, so I stayed away from them. Whatever put me in this privileged situation, I hope my children will inherit it.
On the subject of birth months and rates of schizophrenia, has there been any studies to see if the trend is reversed in, say, Australia which has the opposite flu season to the US / Europe?
Personally a bit skeptical that birth month has huge effects although the flu season is the most reasonable explanation.
So where does listening to Mozart fall? Tier 5?
Wow great timing, my wife and I are starting trying for a baby this week.
test
Just as I'm starting to wrap my head around conceiving :) thanks for the intro, I now feel so prepared!
Please do not promote eugenics. Also, it is disturbing how carelessly you speak of "embryos," they are also people, with rights. Whatever their potential IQ may be.
Re licorice: Does anyone know if licorice root tea also has the same problem? Because I'm pretty sure every tea that actually works to make your throat feel immediately better contains licorice root. Since I read your original article about pregnancy and licorice, I decided to wean myself off of all of them (I used to drink them because I was taking vocal lessons and sometimes you just want to soothe your throat) in case I decide to get pregnant...
Could Cytomegalovirus explain some of the birth order effects? Women are typically pretty careful with avoiding infecting when pregnant with their first, but once they have a toddler running around there is only so much they can do. So subsequent pregnancies are probably more likely to have CMV infections than first pregnancies, and the first born are less likely to be damaged than subsequent babies.
So I was first in my class in AP chemistry which qualifies me to look up some molecular masses and check these papers' dose-math.
The "2g choline" study actually made a serious error in dose-conversion when they said "Each choline supplement dose consisted of 1.25g choline bitartrate, which contained 1g of bioavailable choline cation." It's actually around half that. So they were actually testing 1g of free choline.
choline bitartrate molecules weigh 1.81 times more than choline chloride and 2.53 times more than free choline. Each of these molecules contains only one choline. Convert doses accordingly.
I think it's fine to take 2g of choline bitartrate to get ~790mg of choline. But the largest pills on amazon are only 650mg choline bitartrate per pill.
In general a dose is meaningless without specifying which form the dose-mass refers to. Someone could be off by 30x if for example one number is free lithium and the other number is lithium citrate and they forget to convert.
This is really basic stuff that everyone learns in their first chemistry class so it's disappointing that a paper got past peer review while overstating its choline dose by 2x.
The nullification / "reversal" in the age advantage amongst elite athletes may be because if you are a really skilled athlete born in the end of the year and are the youngest on your team but still good enough, at some point, age stops mattering and your natural ability level shines through. Thus, we'd expect the elite not to see much of a difference because those who are good enough can still play; you're weeding out the merely really good but not quite superelite players.
I would be skeptical of a reversal, though one possibility would be that constantly playing at a disadvantage gives them a slight training advantage, so they end up a bit better than they would otherwise if they actually stick with it.
However, I suspect it is probably the former.
Wittgenstein said that by looking at language, he let the philosophical fly out of the bottle and thus ended the maddening work, and this post strikes me as doing the same for pregnancy optimization.
Whether or not that is your intention, it's a thought-provoking post (and makes me glad I'm well past the age to worry about any of it).
On the birth month thing and sports, one of the things that defined my socialization was being born on the cusp of our rec. league baseball program's cutoff date. I alternated year to year being a bench-sitting, error-prone right fielder, to being a star pitcher. I didn't play past high school, but I learned many life lessons about failure and success.
Scott - the timing of this redux makes me think that you are considering having kids. Congrats!
I would estimate that the birth month disparity would have some impact on IQ. Schizophrenia generally has far more of a detrimental effect on IQ than autism spectrum disorders. And theoretically, a child born up to ~11.9 months earlier than their peers would perform slightly better academically and moderately better athletically, thus receiving more encouragement from teachers. But this would obviously be very difficult -if not impossible- to quantify with currently available data.
So this market should resolve now, right? https://manifold.markets/ScottAlexander/which-of-these-interventions-will-i
The idea that fish oil is defective isn't a stretch at all. We know what we want from fish oil: omega-3 fatty acids, so we measure it and they usually don't have it. And that shouldn't be so surprising because this is specifically the kind of fat that oxidizes easily and so that's just what it does.
This is a pretty good piece! Thanks for doing all this work.
Aluminum itself is a neurotoxin although it has low solubility in water and low oral bioavailability. Should I be concerned about the quantity of aluminum that an activated alumina filter would put in the water?
The major biodeterminist parenting intervention is to have kids young. If that's not going to happen, freeze your eggs or sperm. Feels weird recommending something I didn't do, but I would do it if I were young now - especially for men, probably for women.
To clarify "avoid CMV", I think a solid piece of advice is "don't start a job in healthcare or at a school/daycare" right when you're trying to conceive, because newcomers tend to get sick quite often when they start working in these environments (with CMV & everything else--nearly all bedside nurses have had CMV)
Another confounder for birth month is whether or not your planning is effective. In practice, planning the birth month of your child is really pretty tricky, as a lot of people don't get pregnant on the first try (or even second, third, etc). One would assume that a better ability to plan this intentionally might be due to higher fertility in general, which could be correlated to a lot of things, including a healthier embryo/fetus and subsequently higher IQ.
I respect the dedication for exploring all these, to end up with 2 IQ points equivalent. What a journey. (Also, that Elon Musk decompression baby theory-not-theory is a gem!)
Thanks so much for doing this (both research and communication) - wife & I are implementing some of these which we probably wouldn't have without this article.