SSC Journal Club: Childhood Trauma And Cognition

This month’s American Journal of Psychiatry includes Danese et al, Origins Of Cognitive Deficits In Victimized Children. Previous studies had found that abused children had lower IQ. They concluded that the severe stress of being abused must decrease brain function. Danese et al challenge that assumption in the context of the new paradigm of shared-environment-skeptical psychiatry.

They looked at some of the big developmental studies that had followed thousands of children from birth and evaluated them throughout the years. Most of these kids had gotten a bunch of different IQ tests every year, starting as young as 3. These children had also been screened for abuse, both by questions that tended to predict abuse (for example, asking parents whether they ever “punished their children harshly”) and by asking the kids, once they were grown up, whether they’d ever been abused or not.

The hope was to find children who hadn’t been abused before their first IQ test at age 3, but who started being abused before their last IQ test at age 18. Then they would see whether IQ at age 3 predicted IQ at age 18 in the same way it would in unabused children, or whether IQ was lower than predicted, presumably because of the abuse.

Before the adjustments, they found the same thing as every other study – abused children had lower IQ than unabused children (p < 0.01, B = 0.1):

Consistent with previous research, we found that adolescents and adults with a history of childhood victimization have pervasive deficits in clinically significant cognitive functions, including both general intelligence and more specific measures of executive unction, processing speed, memory, perceptual reasoning, and verbal comprehension.

After the adjustments, there was no difference between IQs in the two groups (p = 0.13, B = 0.03):

In contrast to the conventional causal interpretation of these findings, our longitudinal prospective design revealed that cognitive deficits in victimized adolescents and adults were largely explained by cognitive deficits present before the observational period for childhood victimization and by nonspecific effects of childhood socioeconomic disadvantage. On the one hand, the results are consistent with the high heritability of cognitive functions, their strong continuity across the life course, and the stable cognitive deficits previously described in children exposed to adversity. On the other hand, they are inconsistent with the causal effects of early-life stress on brain function reported in experimental animal models.

The study raises the possibility that kids from poorer families were more likely to get abused and to have lower IQs. Other possibilities include abusers (like criminals) having lower-than-average IQs and passing them on to their kids, or low-IQ kids being less able to figure out ways to escape abuse.

I find this interesting for two reasons.

First, twin studies famously find that early family experiences (the “shared environment”) are less important than previously believed. This has led to a small cottage industry of trying to deny twin studies and saying they don’t work. There’s a lot of interesting debate in this area, and by “interesting” I mean “annoying and interminable”. Maybe the best solution is to route around the whole problem by viewing twin studies as an incentive to conduct really rigorous non-twin studies to see if alternate methodologies can confirm the same surprising conclusions. Danese et al is a rare example of this being done right.

Second, a lot of discussion of the limited role of the shared environment adds the caveat “of course, we’re not talking about actual abuse here. Of course that affects outcomes.” It probably affects some of them. But here we see that it’s not necessarily true along every axis.

A recent twin study showed that child abuse doesn’t increase risk for ADHD, autism, or learning disorders. But I don’t think most people expected it to increase the risk of these disorders, which are traditionally viewed as really genetic.

A 2010 study found that it didn’t increase the risk of “conduct problems” broadly defined. These are traditionally viewed as linked with child abuse, so that’s interesting.

A 2013 study found that child abuse mostly didn’t explain personality disorders. Traditionally borderline personality disorder in particular is viewed as almost invariably associated with abuse, but if I’m reading the study right abuse really only explained about 1% of the variance in who had the condition. And although this called itself a “co-twin analysis”, it wasn’t a real twin study – it used non-abused twins as the control group for abused twins, but it didn’t check if they were identical or non-identical; this controls for family socioeconomic status but not for genetics. This leaves open the possibility that even the 1% remaining abuse-related variance might not be real.

The strongest study I could find was this one linking childhood sexual abuse to a threefold lifetime increase in psychiatric disorders. It, too, was a cotwin analysis and so not genetically controlled. And although I am not going to win any sensitivity points for saying that genetics can affect likelihood of being sexually abused, this is my impression of a lot of the cases I’ve seen – for example, children with a lot of of preexisting issues and bad relationships with their parents are more willing to secretly meet predators who have been grooming them online.

On the other hand, there are also reasons to believe these studies might understate the dangers of abuse. For example, the Danese et al study sampled 3000 kids who were supposed to be representative of the communities they came from. Suppose only 1% of children experience very severe child abuse. Then they would only have a sample of 30 – far too small to find any interesting conclusions. Maybe most of what they’re picking up is differences between not-abused-at-all kids and only-mildly-abused kids, and we shouldn’t be too surprised if they don’t find anything major.

I don’t want to underplay the role of child abuse. My impression from anecdotal evidence is that it does seem to have serious negative effects on kids later in life. For example, I’ve seen some kids from well-off, genetically-healthy families who were sexually abused by a coach or a neighbor or a priest or someone and who became nervous wrecks later in life, unable to stop thinking about it or to stop feeling guilty. Even good studies usually find that trauma is a risk factor for PTSD (thank goodness), proving that highly stressful events can be psychologically scarring and lead to future problems. Once you’ve accepted that, you’ve got most of the assumptions you need to reconstruct the case for child abuse causing psychological problems.

I think the safest assumption is that child abuse that rises to the level of trauma is associated with a PTSD-like picture (although I’ve heard some experts discourage thinking of it as literally PTSD) that seriously affects later-life outcomes. I think it’s less likely that child abuse has many effects that are wildly different from what we see in normal PTSD. And since normal PTSD doesn’t decrease IQ, I’m not surprised to hear that child abuse doesn’t do that either.

This entry was posted in Uncategorized and tagged , , , , . Bookmark the permalink.

Leave a Reply

150 Responses to SSC Journal Club: Childhood Trauma And Cognition

  1. manwhoisthursday says:

    I seriously doubt that anything in the social environment much affects IQ, even serious abuse. Maybe being locked up in a closet for your childhood, but IQ seems to be impervious to almost anything social.

    My impression is that the social environment might have a non-trivial impact on personality, though there is a large, hard-to-shake genetic influence too.

    • Murphy says:

      Social broad sense or social narrow sense? having somewhat crappy vs nurturing parents appears to have surprisingly small effects but having parents on the poor side of society, living in a house where you’re exposed to high lead levels or very poor nutrition appears to make a big difference.

    • Enkidum says:

      How about this as an example of social stuff affecting cognition?

      • fahertym says:

        From the study’s abstract:

        “We found that the same farmer shows diminished cognitive performance before harvest, when poor, as compared with after harvest, when rich. This cannot be explained by differences in time available, nutrition, or work effort. Nor can it be explained with stress: Although farmers do show more stress before harvest, that does not account for diminished cognitive performance. Instead, it appears that poverty itself reduces cognitive capacity. We suggest that this is because poverty-related concerns consume mental resources, leaving less for other tasks.”

        Isn’t “consuming mental resources” the same thing as stress?

        • wysinwygymmv says:

          I think it’s possible to perform cognitively demanding tasks without feeling stressed about it. Some people enjoy puzzle games and RTS games.

        • bobbingandweaving says:

          I think it’s more like a higher proportion of your intelligence is diverted when your mind is preoccupied with finances leaving less to dedicated to mentally exerting tasks.

          The reason they say it isn’t related to stress is that they test for biological markers of mental stress and control for that in their regressions.

          Also look at the lab study I think it illustrates the point more clearly than the field study.

          • bobbingandweaving says:

            Which is also why they use the metaphor ‘bandwidth’. Poverty apparently doesnt reduce total bandwidth it just leaves less bandwidth for the things necessary to succeed in life.

    • Evan Þ says:

      I agree. Scott’s post, and a whole lot of the literature, seems to be equivocating between “serious negative effects on kids later in life” and “serious negative effects on kids’ IQ later in life.” There can be a whole lot of other serious negative effects.

      • Ransom says:

        Also, what constitutes “abuse”? Is there a uniform definition? Does it have to be life-threatening stuff, or does spanking count? Do harsh words count?

        • sierraescape says:

          I don’t think there is a uniform definition, but I’d call abuse anything with a certain large amount less disciplinary power than hurting power. So spanking could count if it was unwarranted and/or very violent.

      • Steve Sailer says:

        A bad childhood is bad even if it doesn’t affect the rest of your life. Childhood is a big chunk of anyone’s life.

    • Yosarian2 says:

      What about the difference between children in more stimulating enviroments vs less enriched enviroments? Has there been research on how that affects IQ?

  2. Steve Sailer says:

    One methodological issue about identical twins (raised together) studies that I haven’t seen discussed much is that being raised together tends to encourage identical twins to focus upon differentiating themselves from each other, whether to assert their individuality or because there environment needs different roles. (For example, there are a lot of basketball player identical twin pairs, but usually one has to learn to play, say, center and the other power forward.) If you are a 6’11” singleton, they’ll probably teach you to play basketball with your back to the basket, but if there are two of you growing up, like the Collins twins, you or your coaches will probably decide one will be center and one power forward.

    My guess would be that these kind of circumstances would encourage identical twins raised together to be a little more different than if they were raised in identical families apart (e.g., if one identical twin was adopted by his father’s identical twin who was married to his mother’s identical twin).

    • RH28 says:

      If that was really an issue, you would see it in the twin studies. Identical twins raised in the same household would be more unlike one another than identical twins raised apart. But we don’t see that. Instead we find no effect of shared environment, just like with non-identical twins.

      • Desertopa says:

        Unless shared environment does have some effect of promoting similarity, which is countered by twins in a shared environment deliberately cultivating differentiation.

        It’s overly convenient and not very parsimonious, but considering human facility for learning and conditioning, and the presence of broad cultural differences between populations which are extremely difficult to explain in terms of genetic differences, the dearth of apparent shared environmental effects on adult character is itself a surprising result, and I don’t think we yet have the information to properly reconcile the data.

        • RH28 says:

          Ok, what about the data showing fraternal twins showing no effect of shared environment? Combine it with no shared environment effect of two adopted children being raised in the same home. So you’re getting the same result from multiple angles, making what you’re saying seem extremely unlikely.

    • Egalisator says:

      This likely happened to me and my twin brother, and we are not identical. I think we both would have turned out more alike if we wouldn’t have been twins. We often got compared to each other which had the effect that each of us chose hobbies and interests, in which the other one wasn’t that good. At least that is my perception. My parents confirmed this observation.
      I really would like to know if there are any studies about it.

      • Steve Sailer says:

        An interesting pair of twins were the playwrights Anthony (Sleuth) and Peter (Equus and Amadeus) Shaffer. They tended to claim to be fraternal twins (perhaps because Anthony was straight and Peter was gay), but theater critic Terry Teachout and twin expert Nancy Segal both told me they were identical.

        Peter worked hard to rise up the ranks of playwrights while Anthony went into other careers with more immediate money like advertising and law. But then Anthony wrote the immensely profitable “Sleuth.” I would guess that Peter’s “Amadeus” is a response: Peter perhaps saw himself as the responsible, diligent Salieri and his twin as the irresponsible, gifted Mozart.

  3. amoeba says:

    I find your third paragraph confusing. The study does NOT seem to be looking for a subgroup of children who were un-abused before 3 and abused later. Where do you see it in the paper? All the comparisons (including what you present right below your third paragraph) are between regressions with and without adjusting for various age-3 covariates such as maternal IQ etc.; but the whole sample is used with and without adjustment.

    (Also, I cannot find the exact numbers you are reporting, e.g. “p = 0.13, B = 0.03” in the paper. Is this about the Dunedin Study and about the first test they present with it (page 7, right column)? This would be Table 3A, model 1 and 5 [they say Model 4 in the text, but they must actually mean Model 5]. The numbers are B=0.11 with p<0.01 and beta=0.00 with p=0.89.)

  4. Fluffy Buffalo says:

    I’d like to add one more point to the list of possible causal relations: naturally dumb kids do more dumb stuff that gets them punished (possibly beyond the point of abuse, in the case of equally dumb parents). (Like in the joke cited in “The Nurture Assumption”: “Don’t be too hard on Johnny, he comes from a broken family” – “I’m not surprised, he could break any family.”)

    Of course, “abuse makes the child feel like shit when it’s happening, and possibly for years afterwards” should be solid enough grounds to condemn and prevent child abuse as far as possible. That it doesn’t cause lasting damage to the child’s intelligence ob top of that is a relief, not an excuse.

    • Conrad Honcho says:

      That it doesn’t cause lasting damage to the child’s intelligence ob top of that is a relief, not an excuse.

      There are also people who believe poor outcomes in life are related entirely to childhood violence (even just spanking, not abuse). Obviously child abuse should be condemned and stopped, but we should also discard fantasies that solving childhood abuse will result in a utopia for adults.

  5. reasoned argumentation says:

    For example, I’ve seen some kids from well-off, genetically-healthy families who were sexually abused by a coach or a neighbor or a priest or someone and who became nervous wrecks later in life, unable to stop thinking about it or to stop feeling guilty.

    Not quite fitting in with the rest but childhood sexual abuse is really well linked to adult homosexuality in men. That’s not IQ affecting but it sure as hell makes life much worse.

    37% of adult homosexual men report having had “encouraged or forced” sexual contact with “older or more powerful” partners before the age of 19.

    • Fluffy Buffalo says:

      “37% of adult homosexual men report having had “encouraged or forced” sexual contact with “older or more powerful” partners before the age of 19.”

      If I read the abstract correctly, the sample was from people who showed up at “sexually transmitted disease clinics”. Not exactly an unbiased sample.
      Even apart from that, the arrow of causality is not exactly clear in this case. At what age does sexual orientation manifest? Maybe the predators chose them as targets because they saw their homosexual tendencies and thought they would be more responsive than straight kids. Maybe the victims were chosen because they were particularly vulnerable – getting shunned by other kids for their strange behavior, or questioning their own identity.
      It would seriously surprise me if you could bend someone’s sexual orientation that easily…

      • cactus head says:

        >It would seriously surprise me if you could bend someone’s sexual orientation that easily…

        Bihackers on lesswrong have supposedly done it.

        • Murphy says:

          “Bihackers” oh lesswrong community. Never change. Or do, in a planned and intentional manner.

          That’s the first time I’ve ever heard the term but it completely fits.

        • Ozy Frantz says:

          The results of bihacking are pretty consistent: most people who try can get to a Kinsey 1 or Kinsey 5, but then it stalls out; a few people have epiphanies where they become bisexual, but it’s not possible to volitionally control whether or not you have them.

          • mayleaf says:

            I *feel* like I successfully bihacked, but I don’t know to what extent that best describes what actually happened. I certainly had to engage in a series of mental reframings before my platonic feelings of closeness, admiration, and affection towards certain women started to feel like “crushes” (with the associated feelings of excitement, bubbliness, and sexual attraction).

            I wrote more about this on my tumblr about a year ago:

          • Desertopa says:

            I tried to bi-hack in my mid-teens, and began and ended at a Kinsey 0. I wonder if the population of people who discuss bi-hacking at all tends to be skewed in favor of people who achieved even minimal results.

          • hlynkacg says:

            @ Desertopa
            I thought that was obvious.

      • Zephalinda says:

        It would seriously surprise me if you could bend someone’s sexual orientation that easily…

        Why? Sexual imprinting seems to be a thing in animals, and in evolutionary terms, it makes sense to have a mechanism like that to optimize sexual behavior for a particular organism’s specific social conditions. I’d expect there to be a very specific, possibly early, developmental window for it, and probably some heredity-environment interactions in terms of the strength/ flexibility/ durability of the imprint, but why, prima facie, should it be implausible?

        • Fluffy Buffalo says:

          It’s not obvious under what circumstances becoming gay constitutes “optimizing sexual behavior”, in terms of evolutionary fitness – maybe it was less of a genetic dead end in times and places where gays had to stay in the closet and live a straight lifestyle, but still, I’d expect less offspring.

          And considering that males have an evolutionary advantage if they keep other males from competing, figuring out how to turn other men gay would be a huge evolutionary advantage, or on the other hand, offering a mechanism by which other men can turn you gay would be a massive vulnerability. (I’m not saying it’s impossible, just not very plausible.)

          • Murphy says:

            Our current social system is comparatively recent. male-female pairings in some cultures would have had a lot less to do with feelings for one. People understood that the way to get children was sex and gay people are still perfectly capable of sex with the other gender when they want children so it’s not so crippling a disadvantage as you might think given copious family and social pressure.

            If you have enough power over a group of competitors to inflict whatever process will make them gay then you probably have enough power to just kill or castrate them. On the other hand being able to put on a good showing in a society where fucking the boss is the way to get ahead and isn’t gender limited might be a notable advantage.

          • Nornagest says:

            Murphy makes a good point, but Buffalo does too. In most times or places, people — especially successful people — would have been expected to get married and produce children no matter their orientation. But on the other hand it’s reasonable to expect fewer children from gay people, even in those times and places, simply because we can expect them to have been less enthusiastic about the process. 0% is unrealistic but 50% doesn’t seem unreasonable.

    • Murphy says:

      before the age of 19? so to be clear, it’s including in the numbers 18 year olds who had been “encouraged” to have sex with someone older.

      …. so basically freshers week?

      There seems to be some troubling run on or’s in use here. “encouraged or forced” by someone who is “older or more powerful” is like saying X% of respondents “ate bread or stabbed someone” particularly grouping in time when the individuals were adults.

      It groups together the guy who had a fling with his friend who was a year older when he was 18 with a kid who was raped at knifepoint by a stranger at age 14.

    • Ozy Frantz says:

      There’s a high rate of pederasty among gay men who go to STI clinics! Next thing you know you’ll be telling me that lesbians often use dildos and bisexuals are often poly.

    • Conrad Honcho says:

      I wonder if finding an older sexual partner is simply easier for young, interested homosexuals. A 16 year old homosexual male will have a much easier time finding a 20 year old male sexual partner than will a 16 year old heterosexual male be able to find a female sexual partner (of any age).

  6. siduri says:

    I have personally known victims of childhood sexual abuse who were absolutely scarred in their adult lives by the narrative of the “cycle of abuse.”

    I mean, the whole story of the cycle of abuse is meant to allow us to have empathy for abusers. Empathy feels good. There’s something absolving about being able to look on a child abuser and think: “OH. He’s not innately bad, he’s just a victim who’s enacting out the same pattern that was inflicted on him as child.” This pattern lets us feel something like relief. We don’t have to grapple with really hard questions about wolves in the fold. It’s very human, and very good, to want to understand the enemy.

    But I feel like not enough people thought about the burden this places *on victims*. I mean, imagine. Not only were you abused as a child. Not only are you dealing with the confusion, the fallout, the total shambles of your sexual being–asking yourself questions like “are the taboo things I get turned on by ‘normal’ taboo things that carry an erotic charge merely by their very transgressive nature? Or am I sick, deviant, have I been permanently warped by the sexual experiences that were forced on me early in life?”

    You’re constantly wrestling with these questions *already*. You have all the normal sexual guilt and self-doubt that everyone has, but IN ADDITION, you have no touchstone of ‘normal.’ You don’t have any way of knowing how much of your sexuality, and whatever kinks fuel it, is baseline and how much is a sick leftover effect of your abuse.

    AND THEN. You’re told by the scientists, the experts, the therapists, the media–that this thing you’re already wrestling to understand and control, means that YOU are, yourself, likely destined to be a predator and to harm children?

    I MEAN FUCK. What a way to take victims, and their pre-existing guilt and shame and self-doubt, and just…torque that in really the nastiest way possible. I feel really confident in saying that this dominant theraputic/media narrative has done REALLY A LOT OF HARM to abuse survivors, to the point where I personally know people who have considered suicide because they were afraid of perpetuating “the cycle of abuse.”

    And this is only one exhibit in the whole catalog of things that I’d title: “Social scientists really need to think very hard before they propose theories that maybe sound plausible but will, upon widespread uptake and naive perpetuation through the media/blogosphere/word of mouth, cause acute harm to pre-existing, incredibly vulnerable populations.”

    (I am not saying they should stop seeking truth. I am saying they should have some epistemic humility along the way. Don’t just think about how society would react if their/your ~challenging~ theory is right. Think about how it would affect people if it gets accepted, and turns out to be wrong.)

    • Murphy says:

      It’s not even just implicit. I remember an event scott linked to a while back (live by the sword?) where an individual was being witch-hunted by his fellow-feminists for saying something like “jesus would have been a feminist” or some such that riled them up. they then searched back through everything he’d ever written to snipe-quote. One item they picked up on was that he’d suffered childhood abuse, they then linked this together with where he complimented a teenage girls work for something and followed up with something like “… um just to clarify I’m not hitting on you”.

      What they then turned it into was [abuse survivor] -> [ abuse survivors become abusors]-> [confirmation that he was definitely trying to groom this girl for abuse]

      So it’s not just the abused people suffering internal worry, there’s lots of people who’ll decide that anyone who was abused is definitely a threat to children and will shout this at the victims publicly.

      • spork says:

        Wow, that is messed up, and it reinforces siduri’s point all too well. But I have to wonder if more scientifically defensible beliefs about the causes of child abuse would give more comfort to the victims. Yeah, the act of abuse itself is not going to increase their likelihood of becoming abusers – one less thing to worry about. But now they have to face up to the strong habitability of the disposition to abuse children, and… oh yeah, they are a son/daughter of a child abuser. Not only will they again feel tainted (abuser genes are in me!), but they will also have no reliable way to break the “cycle of abuse.” If it’s genetic and not behavior-perpetuated, it could pop up in the next generation even if their own behavior is angelic.

        (This is not an argument for encouraging false beliefs – just a specific instance of the general point that truth isn’t automatically comforting.)

        • Desertopa says:

          Wow, that is messed up, and it reinforces siduri’s point all too well. But I have to wonder if more scientifically defensible beliefs about the causes of child abuse would give more comfort to the victims. Yeah, the act of abuse itself is not going to increase their likelihood of becoming abusers – one less thing to worry about. But now they have to face up to the strong habitability of the disposition to abuse children, and… oh yeah, they are a son/daughter of a child abuser.

          It’s going to sound bad for a lot of abuse victims, but significantly fewer than those for whom the current paradigm already sounds bad, since a lot of child abuse victims were not victimized by their genetic parents. Particularly in terms of sexual abuse, stepparents are a lot more likely to commit it than genetic parents (not surprising in light of genetic incentives,) and a lot of abuse is perpetrated by non-parental custodial figures.

      • Loquat says:

        Also, since Paul Crowley was nice enough to link the post, you can see there the exact thing siduri was talking about:

        Target: …one of the biggest fears of a lot of sexual abuse survivors is that they’ll become their attacker. It constantly weighs on me, (name expunged), to the point where I often just wait for women to make the first move instead of risking assaulting them.

        Witch Hunters: OMG he just admitted he might sexually assault people! What an awful human being!

    • onyomi says:

      This raises the additional question of whether or not it’s accurate that childhood experiences can actually shape one’s sexual desires, preferences, fetishes, kinks, etc.? I think it’s kind of taken for granted that they can, which may result in the confusion you mention (everybody wonders a little about their sexual proclivities; add on to that experiences which everyone says are likely to warp you, and even if you only have the same kink you would have had absent abuse, you assume it’s because of the abuse).

      I don’t know if there is a scientific and/or sociological consensus on this. On the one hand, I think most people now accept the idea that being gay or straight is not only not a choice, it’s probably already destined by the time you leave the womb, if not by the time of conception. But being a straight man who likes to vacuum women’s apartments in nothing but a diaper? I am pretty skeptical one could choose or not choose this fascination any more or less, but I honestly have no idea whether it’s because you were genetically predisposed to this and/or because of some unusual experiences you had (my best guess is that the combination occurs whereby you are genetically predisposed to certain types of turn on, but the exact form that takes is somewhat cultural: if you were going to be turned on by Jesus if you’d been born in a Christian country, maybe you’ll be turned on by Krishna if born a Hindu).

      Also a bit odd about this: here, the sociology, PC consensus is absolutely no choice: gay people are born gay and that’s that. Yet these same people would seemingly strongly deny the notion that some people are just “born criminals,” presumably because they’ve accepted gayness as not being a bad thing, whereas they cannot accept the notion of being genetically predisposed, even predestined, to be bad (related, is there a consensus on pedophilia? I think there is a widespread notion that pedophiles are more likely to have been victims themselves, but I have no idea whether that’s reflected in actual studies).

      • Enkidum says:

        I keep saying this everywhere I can whenever it’s brought up, but there have been quite a few large-scale studies of influences on sexual orientation, and they have all found that somewhere between 10-40% of the variance in sexual orientation can be explained by genetic factors. Which is to say (and I keep shouting this but no one ever seems to listen) that 60-90% of the variance in sexual orientation cannot be explained by genetic factors.

        Epigenetic factors are probably hugely important as well, and it seems that the elder brother effect (so far as I’m aware, the only definitive causal factor in sexual orientation anyone’s ever identified) is likely based on interactions between the developing fetus and the amniotic environment (although the evidence is very indirect there, if I remember correctly).

        But that still leaves a lot of room for experience/learning/social stuff in general. And some of the researchers studying this stuff (I am not one of them, in case it wasn’t obvious) are explicit about that. I think the societal consensus (which I agree is what you say it is) is simply wrong here – there is very, very poor evidence that we are born with the sexual orientations we have in later life. In fact, it seems clear that either (a) no one is born with a sexual orientation, and the orientation we develop is due to a combination of various factors, including both genetic and environmental ones, or (b) some people are born with a genetically/amniotically determined sexual orientation, and they’re basically “stuck” that way, and others are not. I can’t see how you can explain discordancies between identical twins and so one, without allowing for something like this.

      • Kaj Sotala says:

        A lot of people (including me) with weird kinks report them having manifested in some form already in early childhood, even before puberty. So while they obviously can’t have been there from birth, many people at least report having experienced some form of them from a very early age.

        Of course later experience is likely to have shaped what exact form they developed into later on.

      • Controls Freak says:

        the sociology, PC consensus is absolutely no choice: gay people are born gay and that’s that.

        My queer theory prof flatly stated that she was agnostic about biological determinism, so there’s that.

    • Z says:

      Just anecdotes here, but yeah, as a childhood trauma survivor, I ruled out having children to prevent that cycle until I could be sure I wouldn’t perpetuate it. After completing trauma therapy, I no longer have that concern. A friend of mine is going through the same process.

  7. Zephalinda says:

    Not a social scientist, so perhaps I’m missing something– but shouldn’t we be concerned about the accuracy of the measuring instruments being used to identify “childhood abuse” here?

    The paper doesn’t give much detail at all (that I could find in a quick read) about how the E-Risk study assessed this condition, and the Dunedin study seems to have cobbled their measurement together from an observation for “maternal rejection,” some parental self-reports about discipline practices, and the children’s retrospective reports after reaching adulthood. On the face of it, those three measures seem like they’d correlate pretty imperfectly with actual experiences of abuse by children. Why would we expect abusers to be radically self-aware and honest about the harshness of their discipline (or to conceptualize/report physical violence as “discipline” vs. something else?) Given widespread underreporting of victimization in other areas, why should we expect people abused as children to be open, forthright and accurate when asked about those experiences by strangers a decade or more after the fact?

    In all fairness, it looks like the authors here briefly consider that possibility in the discussion, and they do mention that an earlier study found similar results using court-verified cases of abuse. But on a meta-level, I am pretty uncomfortable about the easy way that the category of “People who tick Box 5 on a survey asking “Were You Abused As A Child?” (or whatever)” gets elided uncritically into “Victims of childhood abuse”– and then, still worse, gets reframed in a post like Scott’s above as “the experience of childhood abuse, full stop, does/doesn’t cause X/Y/Z.” Nobody has observed the experience of childhood abuse here; we’re at like 3 levels of noisy remove from the actual phenomenon, the whole way through the paper.

    It’s one thing to use an imperfect instrument, find a big effect, and say “Hmm, gotta be a reason!”. But it seems like quite another to make pronouncements about the absence of phenomena when you’re effectively trying to count leaves on a tree via moving shadows on a brick wall.

    • Rachael says:

      Yeah, I thought that too. I’d have thought most abusers would either think “no, my punishments aren’t harsh, they’re pefrectly proportionate” or lie.

      • Protagoras says:

        Perhaps I’m over-optimistic, but I thought the tendency of people to lie or misrepresent on surveys was well enough understood that the claim that answering questions a certain way correlated with abuse was likely to be based on research actually finding that it correlated with abuse, rather than based on people assuming a particular answer to that question must mean that. Social science has replication problems, but most social scientists don’t seem to be quite that stupid.

    • Mary says:

      And, of course, if you have problems, citing abuse gives you some extenuation. So people with issues are more likely to describe something as abuse.

  8. Freddie deBoer says:

    “Of course that matters.” This is probably true in some situations. But here we see that it’s not necessarily true in all of them.”

    This is precisely the kind of framing I wished you’d avoid. It’s so easy to represent this sentence as the statement of a sociopath. (Or of the rationalist community.) This is the tail-wagging-the-dog aspect of quantitative indicators that turns people into cultural studies PhDs.

    • nweining says:

      Of course people who like to manufacture outrage through deliberate interpretive stinginess can pounce on this. But Scott has already given such people so much fodder that it is plausibly not worth the throat-clearing effort to avoid giving them more.

    • ADifferentAnonymous says:

      To clarify–are you worried about how this sentence can be taken out of context, or do you think there’s something problematic about it even with context?

    • Scott Alexander says:

      All right, I changed to:

      “Second, a lot of discussion of the limited role of the shared environment adds the caveat “of course, we’re not talking about actual abuse here. Of course that affects outcomes.” It probably affects some of them. But here we see that it’s not necessarily true along every axis. ”

      Is that better?

  9. RH28 says:

    Yes, an evolutionary perspective can give you insights about trauma. You think about the environment people grew up in before modern technology. Babies dropping dead all the time, constant pain from toothaches, often teetering on the edge of starvation, a high portion of men dying from modern conflict. It makes me laugh when I see college kids at American university talk about ally he pain they feel from “racism” or “microagressions” or whatever. If we were as weak as they think humans are, no way we could have survived to this point. They’re simply in an environment where they get status by complaining. Drop them off on a desert island and be amazed at how tough they get.

    The environment we evolved in must’ve selected for people who deal with and move on from trauma that is extreme by today’s standards of what individuals face.

    • fahertym says:

      Relatedly, there is an interesting historical debate as to whether PTSD is a modern phenomenon or not.

      Did 500 B.C. Athenian warriors get PTSD from seeing their friends get stabbed in the face by an enemy pike? If so, then nearly everyone in Athens and Greece would have been suffering from PTSD to one degree or another given the nature of militia armies. Considering how debilitating PTSD seems today, how could such a society even function?

      • Ozy Frantz says:

        Some aspects of PTSD, such as hypervigilance, are very adaptive in an environment where people are likely to stab you. They only become maladaptive in a stabbing-free environment.

        Anyway, Herodotus’s Histories has a couple accounts of Greek soldiers who can’t speak or function properly due to the experience of combat.

        • fahertym says:

          Do you mean that people with PTSD make better warriors?

          My impression of PTSD (based purely on second-hand knowledge and hearsay) is that it occurs when a person’s stress levels reach a mental breaking point so they cease to maintain resilience in a stressful environment. For instance, combat veterans are known to panic when hearing sudden, loud noises.

          • hlynkacg says:

            I think part of the problem is the slippery definition of PTSD. If I had to guess I’d say the number of people with severe functional impairment is small while the number of guys “sleeping with their boots on” is a good deal larger.

        • Ozy Frantz says:

          Panicking when you hear sudden loud noises is really useful if the sudden loud noises might be a bomb which is going to kill you and you have to get out of here RIGHT NOW. If the sudden loud noise is instead a car alarm or someone dropping a book, this is pretty impairing.

          • Randy M says:

            Is panic a state of mind that is ever useful, as opposed to focused and alert?

          • wildtypehuman says:

            Also, “panic” can cover fighting​rather than fleeing. Most of my mild PTSD vet friends are much more likely than baseline to, e.g, start planning how they’re going to violently respond if the seemingly innocuous guy at the bar tries to start something. I suspect they’d do fine in ancient Greece.

          • Ozy Frantz says:

            Randy: Prima facie, yes, or else it wouldn’t have evolved.

          • quanta413 says:

            Randy: Prima facie, yes, or else it wouldn’t have evolved.

            Not true. Panic could be better than some other pre-existing option and thus evolve but still be always worse than focused and alert.

            Another objection would be that evolution doesn’t optimize piece by piece. The same brain hardware that allows for emotions such as panic may be well suited for many other tasks and panic is an unfortunate side effect or overstimulation of a system that turns out to not be evolutionarily selected against strongly enough.

            That said, running away instinctively sounds like a decent strategy for at least some predators and I don’t think focused and alert then choosing to run sounds like something easy to evolve or that would do any better in that situation.

          • MNH says:

            Ozy: Identifying which features of the mind are intended and which are spandrels (in the SJ Gould sense) ranges from hard to presently-impossible. If you have evidence that lets you make that judgement, I’d like to see it, but if not, I think your claim is too hasty.

      • RH28 says:

        Good point. Four possibilities I see

        1) We’re simply biologically different from the Ancient Greeks. See Cohcran and Harpending’s book, as our environment has changed dramatically over the last 10,000 years, evolution has sped up. So as society has become less violent overall, selection for resilience has gone down as other things were selected for.

        2) PTSD hits you when you haven’t been desensitized to violence. So an Ancient Greek who saw relatives die all the time and had to slaughter animals growing up can come back from a battle just fine, while a modern American will have trouble.

        3) There’s something different about modern wars. From what I know about Iraq and Vietnam, it’s the hours and hours of uncertainty and possibly being hit from every angle that drives people crazy. Maybe there’s less long-term damage in going through a short battle where you either win or lose and then go on. Did WWII veterans get PTSD?

        4) Maybe the research on PTSD, like that on trauma, is fatally flawed some how.

        • chamomilegeodebackup says:

          yeah, wwii vets got ptsd. it was called battle fatigue then, just like it was called shell shock in wwi. and all those things were definitely considered the same disease, just with name updates–unlike say the old diagnosis of hysteria, which a lot of people now think was probably used to name ptsd showing up in women, esp as a result of rape.

          of course, wwii and wwi were more similar to vietnam in the ways you mentioned than any of those wars were to ancient conflicts

          • HeelBearCub says:

            I feel like I am supposed to link the George Carlin bit now, but I also don’t actually agree with Carlin’s take, although I admire how brilliantly he delivers it.

          • hlynkacg says:

            I’m curious, what is the precise nature of your disagreement? My personal take is that he’s on to something but the theory is incomplete.

          • HeelBearCub says:

            What he is offering there is an anti-science take, whether he knows it or not. He basically just wants to spit on people who want to talk about things more accurately.

            Shell shock is a descriptive of PTSD, but essentially colloquial in origin and applies to the specifics of WWI. Battle fatigue is linking together a set of symptoms that seem to be able to come out of combat in general. It’s a more accurate description brought about by study, not embraced by a random bureaucrat because it sounds softer. Post traumatic stress disorder names the same symptoms, brought about by the same mechanisms, whether or not they come out of combat.

            He starts off by talking about hating words and phrases that hide meaning, but actually is talking about hating people who think things are complex.

        • Nancy Lebovitz says:

          There was an essay by David Drake (probably in one of the Men Hunting Things anthologies) about war becoming more stressful. It’s become more continuous– no breaks for night, and no going home for planting and harvest. It’s louder– a stressor in itself and disrupts sleep. Death and injury are more random.

        • Aapje says:


          Modern war is a constant grind, where people just get worn down over time (hence: rotating out units to prevent disintegration). Ancient war was often a 1 day affair. There was a single battle and one side won.

          • lycotic says:

            We were talking about the Athenians, right? The Peloponnesian War went on for 27 years, and contained plenty of disturbing stuff (cf. Melos). I’m unconvinced.

          • Aapje says:

            War back then was a whole lot of travel and waiting, but very little fighting. That war could take 27 years, because it was low intensity.

            WW 1 and 2 combat often consisted of continuous contact with the enemy, with a lot of shelling and where any mistake could be your end. It’s constant stress vs a day of stress here and there in the past.

        • hlynkacg says:

          @ RH28

          #1 Feels obviously false. To be honest the writings of Pyrrho and Aeschylus feel much more immediate and “human” to me than those of most modern rationalists. I find it hard to credit the idea that they are/were fundamentally different from I.

          #2 and 3# certainly seem plausible. #3 more so than #2.

          #4 is almost certain

        • FacelessCraven says:

          RH28 – “3) There’s something different about modern wars. ”

          Way more ambush threat, and threat from things similar to ambush.

          WWI introduces artillery firing supersonic shells from beyond line of sight, along with sniping as a chronic threat. Up till this point, large-scale wars had a much cleaner demarcation between combat and “on the march”, with a great deal more warning and time to prepare when moving into danger, a shorter dwell-time in danger, a heavy emphasis on rote, mechanistic drills while in danger, and lots of time to decompress coming out of danger. All of these factors seem like they would help minimize PTSD quite a lot, and WWI is when they all go away.

      • publiusvarinius says:

        After seeing literally everyone on the Ellis Island Immigration Photos having the thousand-yard stare, I entertained the following hypothesis: PTSD is just normal human behavior in nature, and what we call normal behavior today would be considered pathological in most settings.

        Do wild animals naturally display what we call symptoms of PTSD: irritable, agressive, reckless, self-destructive behavior; hypervigilance; strong startle response; restless sleep?

        • keranih says:

          Do wild animals naturally display what we call symptoms of PTSD: irritable, agressive, reckless, self-destructive behavior; hypervigilance; strong startle response; restless sleep?

          If by “wild” you mean “free living, non-confined animals that have never been domesticated” then we have no idea. We can’t (yet) observe those animals well enough to be able to tell.

          If by “wild” you mean just “not domesticated” then, yes – and domesticated animals will do this, too, if not managed appropriately. Unpredictable environment, frequent startling, and isolation/overcrowding (which is very species specific) is the easiest way to get to that state.

          The US army’s done some work on bomb dogs who show these signs, but it’s more anecdotal than confirmed.

      • teageegeepea says:

        Greg Cochran seems to think PTSD (or “combat fatigue”) is iatrogenic, since the Germans & Soviets didn’t seem to have much of it during WW2. He also thinks obligate male homosexuality is caused by a pathogen affecting the brain (same deal with narcolepsy, and likely with schizophrenia).

        • John Schilling says:

          On the first, there is a strong case that the acute symptoms of SS/BF/OE/PTSD are greatly diminished when the policy is a quick bullet to the back of the head of anyone disabled by those symptoms. Unfortunately the armies which implement that treatment don’t generally do rigorous follow-up studies on the long-term effects.

          Pathogenic homosexuality, interesting and not wholly implausible theory, but I think anyone proposing to study it today will find themselves wishing they’d picked a relatively uncontroversial topic like HBD instead.

        • publiusvarinius says:

          since the Germans & Soviets didn’t seem to have much of it during WW2

          I’m not sure what Greg Cochran is referring to here. Shell-shock was a very serious problem for the Nazis on the Eastern Front. Beevor’s “The Second World War” discusses several effects of this, including the serious troubles military policeman had with keeping order at Pitomnik Airfield due to the seriously shell-shocked staff.

          On a more personal note, family members told me wartime stories about the German shell-shock patients being treated in Szolnok, Hungary. A group of naked patients would run out of one of the barracks every time there was a loud noise.

          The world of Soviet psychiatry was not much different. Even POWs were screened and treated for shell-shock. Unfortunately, people who were transferred to mental institutions for treatment were removed from the POW camp records, which subsequently made finding and returning them very difficult. In an extreme case, a POW returned home after spending 53 years in a Siberian mental hospital.

        • hlynkacg says:

          Cochran is full of shit. the Soviets had shit-tons of issues with PTSD and there’s no reason to believe the Germans wouldn’t have, if more of them had lived to talk about it.

          This probably makes me a hypocrite as under most circumstances I’d be the first one to tell you that humans, as a collective, are nothing but a bunch of sex-crazed murder-monkeys with delusions of significance. However, I find it somewhat infuriating that people insist on injecting sex into everything when it’s clear that their understanding of the same is so far outside the median. Phillia != Agape != Eros. Do I really have to be the one to explain to everyone that it is possible to have trust and intimacy without sex?

      • Sanchez says:

        There is evidence that at least part of what gets called PTSD in soldiers today is the result of explosives specifically and not trauma in general.

        From Dan Carlin’s piece on WWI, I gathered that many people back then thought of the particular blend of symptoms they called “shell shock” as something unprecedented and specifically related to shelling (hence the name, of course).

        • John Schilling says:

          Interesting piece, but it barely mentions Vietnam. Since the enemy in that war made relatively little use of IEDs and artillery, I’d expect relatively less PTSD among Vietnam veterans than WWII or Iraq. Also, I’d expect you would be able to sort the population of Vietnam vets by exposure to zero, one, or many blast incidents and see if there is a difference in the PTSD rate.

        • keranih says:

          I gathered that many people back then thought of the particular blend of symptoms they called “shell shock” as something unprecedented and specifically related to shelling (hence the name, of course).

          In the American Civil War, it was called “soldier’s heart”.

          One could argue that it occurs more commonly in sustained massive campaigns combined with physical distress, explosions and noise, but I’ll bet any number of marshmellows that it was around during the 30 years war.

          I think it’s worth noting that a majority of soldiers don’t develop PTSD. Nor are all rape victims compromised for life. I think it’s certainly not unreasonable for the data to show that not all child abuse victim are distinguishable from their null state contemporaries as adults.

    • JohnBuridan says:

      Tying this thread together with @reasoned argumentation’s thread above about sexuality and bi-hacking… because I’m just going just take a classical angle on both.

      On the one hand, the Ancient Greeks obviously knew PTSD by its symptoms. There are numerous reports of soldiers defecating themselves during battle, panicking, weeping, etc (Victor Hansen) or doing the opposite, breaking ranks and, bringing this back to Achilles in Vietnam, going berserk. The hoplite formations got deeper as time went on, the deeper the formation the more moral support soldiers had (Adrian Goldworthy). 40 ranks deep is pretty dang far! And going no thinner than 8

      As a side note: this is why Hannibal’s victory at Cannae is so astonishing, he let the center collapse, and he made his troops assume a very thin ranks so that he could surround the enemy. Thinning ranks requires morale in spades, battle-hardened discipline, and absolute trust.

      The Greeks also theorized that homosexual activity was good for soldiers increasing the bonds of brotherhood (see Achilles and Patroclus and Achilles’ reaction to Pat’s loss) and good for politician’s giving them a disposition to look after the affairs of State (see Plato’s Symposium). Relationship’s between older men and young men in Athens had both an intellectual and physical component to them. However, it was expected that the older man would also help the young man find a wife… this is quite a different society from our own, and it’s hard (meaning impossible) to tell how often this played out across the Hellenic world. But we do know it was called scandalous or at least very weird for two men to be exclusive lifetime lovers in Classical Athens. Does this mean the sexual make-up of Classical Greece than today? That’s hard to tell. But it does seem to me that all male environments have more homosexual activity (boarding school).

      If you imagine sexuality as perhaps Plato imagined education, something which exists inherently in you already, but gets drawn out, developed, and cultivated over time, then it makes sense that both environment and personal will could allow for some flexibility in developing one’s sexual persona.

      • John Schilling says:

        On the one hand, the Ancient Greeks obviously knew PTSD by its symptoms. There are numerous reports of soldiers defecating themselves during battle, panicking, weeping, etc (Victor Hansen) or doing the opposite, breaking ranks and, bringing this back to Achilles in Vietnam, going berserk.

        That’s fear, not PTSD. Note in particular that what you and the Ancients describe is decidedly mid-traumatic, not post.

        • JohnBuridan says:

          Fair enough. I was definitely conflating there. But on the other hand, Ajax kills himself after the battle, Odysseus slays people in his home, Herodotus recounts the soldier who turned into a mute.

          We don’t have a clear picture of PTSD in the Ancient World, but we do get glimpses… Is that fair?

          • Loquat says:

            I wouldn’t say Odysseus counts as PTSD. He slays his wife’s suitors after calmly investigating the situation and planning out his actions, not in a sudden fit of rage or panic. Furthermore, since some of the suitors had been plotting to murder his son for being an obstacle to them, it’s likely at least some of them would have tried to kill Odysseus himself if he’d just showed up and tried to throw them out.

          • John Schilling says:

            Yes, Odysseus’s case is properly orderly, and argues for “PTSD” as a positive adaptation in an environment where people aren’t going to stop trying to kill you.

            Odysseus is also imaginary, whereas Hereodotus was reporting on real events, so we do get glimpses. From Xenophon as well, IIRC.

  10. nweining says:

    This feels like an inappropriate thing to point out in the midst of discussing such an awful problem, but: “executive unction” is a remarkably thought provoking typo which you should maybe try to work into Unsong somehow.

  11. dcarroll001 says:

    There has been significant work done in the last 10-15 years on the physiological impact of early childhood trauma on the development of the brain (including trauma incurred during pregnancy), as well as interventions that can mitigate the impact. I have printed literature on the research in my file cabinet, as it was distributed as part of my training as a foster parent. In my part of the country, Karen Purvis (recently deceased) is the leader in the field ( Most of the technical details are beyond my pay grade, but the gist of the research is that early childhood trauma alters the brain chemistry and the developmental path of the child, especially in years -1 to 3. I’m not sure this has much impact on “intelligence”, except these children have difficulty concentrating and paying attention in school. However, that trauma can be reversed, because the brain is still growing – the earlier the intervention the better. I never fully understood her TBRI stuff, because I am a math guy and it was presented to me by social workers who seem to speak a different language than I. But it is mostly techniques to build trust and form bonds with traumatized children (who don’t know how to trust or form bonds with adults), de-escalate conflict and panic (they escalate very easily), and teach them alternative coping skills.

    While I haven’t looked closely at the your study on this post, from what I can tell many of these kinds of statistical studies group together various forms of “child abuse” from mild spanking to raping and then suffer from an averaging effect – statistically averaging out the variations, overstating the effects of mild abuse and understating the effects of severe abuse. Also, I am always suspicious of studies that rely on self-reported and survey data, especially regarding behaviors deemed socially unacceptable and experiences that cause shame.

  12. alwhite says:

    Every time I see the phrase “et al.” I replace it with the phrase “and friends”. I think this practice should be adopted by the entire community for the betterment of all.

  13. Aceso Under Glass says:

    It seems weird to talk about the effects of abuse without reference to genes like COMT, where one variant makes people more sensitive to their environment (good environments produce even better outcomes, bad environments produce worse outcomes). I can’t find a succinct reference to the overall pattern and I don’t want to post a million links and get caught in the spam filter, but I’ll see if I can find something more specific.

  14. xXxanonxXx says:

    I’m mentally updating my summary of your Biodeterminist’s Guide to Parenting from “don’t beat or starve your children” to “don’t starve your children.” I used to be incredibly scared of the idea of child-rearing, but now it’s looking like something I could handle.

    • fahertym says:

      “Scott Alexander encourages parents to not not beat their children.”

      – for use in future debates.

    • Ozy Frantz says:

      I feel like “beating people probably doesn’t cause long-term negative outcomes” is not a good argument in favor of beating people?

      Like, it probably wouldn’t cause long-term negative outcomes if I punched my housemate in the face, but that doesn’t mean it’s a good way of resolving disputes over the dishes.

      It is uncontroversial that you have a lot of control over whether a person is happy for two decades of their life. That should be sobering whether or not their unhappiness leads them to become unsuccessful in the future.

      • xXxanonxXx says:

        Part of what I love about this community is what I see as attempts to construct systems of empathy so rigorous they are sociopath proof, but my comment was an attempt at humor. If someone truthfully looked at the the “don’t beat or starve your children” prescription, decided that was too high a bar to hurdle, but then came around and started considering procreation after hearing they only needed to do that last bit?

        I would do everything in my power to persuade that person not to have children.

      • Lila Rieber says:

        “It is uncontroversial that you have a lot of control over whether a person is happy for two decades of their life. That should be sobering whether or not their unhappiness leads them to become unsuccessful in the future.”

        Definitely. Also probably affects their happiness for every year afterwards… Even if someone’s low test scores are genetic, it’s possible for them to live a happy fulfilled life. Abusing them makes that a lot less likely.

      • teageegeepea says:

        Judith Harris said the reason not to mistreat children is the same as the reason not to mistreat adults.

  15. Marshayne Lonehand says:

    This thoughtful and well-expressed SSC essay might be further improved (as it seems to me at least) by mentioning the evident postulate, that in a hypothetical world in which genetic traits were entirely responsible for all personality disorders, it might none-the-less plausibly be the case — very plausibly the case (as it seems to me) — that the construction of coherent abuse-narratives improved therapy outcomes.

    Partly this postulate is spiritual (and medical) common-sense. Doesn’t everyone long for better answers to tough questions like “Why do bad (psychological) things happen to me, and to people I care about?” Hence the resident-training maxim “Never deprecate a patient’s/client’s illusions, unless you can offer them better illusions.”

    In contrast, an ultra-rational response — along the lines that “the best available evidence suggests that a mix of bad genes, bad experiences, and plain bad luck too, is chiefly responsible for psychological suffering in general, and personality disorders in particular” — is not self-evidently the most therapeutically effective answer, or the most socially effective answer either, is it?

    More broadly, it’s a humility-inducing exercise to query the NIH’s Clinical Trials database for trials concerning “personality” whose enrollment is presently “open”. The resulting 229 open clinical studies (as of this writing) compose a peer-reviewed survey of the most urgent questions in regard to personality disorders whose answers we do not know.

    An evident sobering fact is, at present we don’t understand all that much, scientifically speaking, about the origins and treatments of personality disorders … and learning more will require a dismayingly long time and dauntingly much work.

    Conversely, an evident hopeful fact is, these 229 ongoing clinical trials encompass pretty much every cutting-edge tool of science, technology, engineering, art, and mathematics. To appreciate this optimism-inducing reality, just restrict the clinical trial-search according to your favorite STEAM-tool (variously “drug”, “talk”, “narrative”, “artwork”, “connectome”, “epigenetic”, or “tensor”, to suggest only a few).

    So good on `yah STEAM-workers! Keep up the (literally) good work! 🙂

    • hlynkacg says:

      John, you’re drunk. Go to bed.

    • Marshayne Lonehand says:

      As a concrete followup, a substantial subset of the questions that are posed in Scott Alexander’s OP are being studied in the just-begun “Preschooler Emotion Regulation in the Context of Maternal Borderline Personality Disorder

      Offspring of mothers with Borderline Personality Disorder (BPD) are at serious risk for developing mental illness at every stage of their life, and yet little is known about how this risk is transmitted. … Difficulty managing emotions is a hallmark feature of BPD, and yet the ability to do so is necessary for responding effectively to childrens’ emotions. This process is called maternal emotion socialization, which has a major impact on how children develop their own emotion regulation (ER) skills. … This proposal will leverage Dialectical Behavior Therapy (DBT) Skills training, a robust and effective method for improving ER as a tool to change maternal ER in mothers with BPD. … Findings from this proposal will identify a modifiable pathway by which offspring of mothers with BPD are at risk, determine the extent to which child ER can be restored by treating mothers, and will be the first DBT Skills trial to measure outcomes in offspring.

      How cool would it be — scientifically speaking — if this clinical outcome study were adjoined to genomic studies (cf.Methylation Status of BDNF Gene After Dialectical Behavior Therapy in BPD“)? Or to anatomic studies (cf.Overlapping Neural Circuits in Pediatric OCD“)? Or to drug studies (cf.A Novel Drug for Borderline Personality Disorder“)?

      It was the worst of times, it was the best of times” … the worst of times in respect to the depths of our present medical ignorance … the best of times in respect to the superabundance of viable research paths toward a better medical future.

      Is it any wonder, that researchers, patients, and families alike feel “drunk” — drunk with newfound hopes? 🙂

  16. Ilya Shpitser says:

    They are not using the right method for determining causality in longitudinal studies (because of time-varying confounders).

    In particular, victimization is the repeated treatment here, and any intermediate outcome “interleaved” within the period of abuse is a time dependent confounder, and needs to be handled carefully.

  17. mingyuan says:

    I was hoping you would write about this paper! I’m currently in a class on child poverty and they’ve been pushing the ‘differences in IQ are entirely explicable by postnatal environmental factors’ angle really hard, and I’ve been hesitant to accept this because I can’t imagine being assigned a paper that came to the opposite conclusion. But at the same time they act like they have robust evidence! Here’s some of it.

    For one thing there’s the paper “Excessive Stress Disrupts the Architecture of the Developing Brain” from Harvard’s Center on the Developing Child, whose title pretty much speaks for itself. Paul Tough, in How Children Succeed, says that early stress affects the prefrontal cortex, which makes children worse at emotional and cognitive self-regulation, i.e. it leads to behavior problems. He also says that executive function (as measured using a game that tested working memory capacity) is affected by allostatic load as measured by ACEs, and that constant stress in childhood overloads the HPA axis, which messes up your stress response. Unfortunately Tough doesn’t cite any of the studies by name, so I’m not sure if these are the same studies that Danese et al reviewed and found to be faulty, but I don’t think so.

    Also from the little I know about attachment theory I would at least expect more emotional problems in children raised by abusive parents, and I would expect that to lead to more behavior problems. And isn’t there some evidence that behavior problems can affect IQ scores? There was a study which gave children an IQ test, then gave them another IQ test where they were given an M&M for each correct answer, and the scores of the children who had performed worst on the first test jumped an average of 20 points. So for the lowest-performing children, the problem is not so much inability to solve problems as inability to concentrate.

    Yeah so I don’t really know anything about social science or psychology and I’m just regurgitating what I’ve read, but I’m still confused about the state of the evidence. If anyone has a way of reconciling Danese et al’s results with this ‘biology of stress’ research please tell me; I really want to understand.

    Agh! Social science is so impossible to figure out! This is why I studied physics!

    • reasoned argumentation says:

      And isn’t there some evidence that behavior problems can affect IQ scores? There was a study which gave children an IQ test, then gave them another IQ test where they were given an M&M for each correct answer, and the scores of the children who had performed worst on the first test jumped an average of 20 points.

      Published in “the Journal of Fails to Replicate” right between studies about stereotype threat and priming.

      Agh! Social science is so impossible to figure out!

      Well if you start out with a premise that’s false then dedicate yourself to finding evidence that justifies that premise or conducing studies that rely on that premise being true to produce results (and are willing to fake results) then yeah, it’s impossible to figure out.

    • Lila Rieber says:

      Completely agree with this. There’s no reason to expect that children with severe behavioral problems would be able to be compliant for anything, much less an activity as unstimulating as a standardized test.

  18. Schibes says:

    Scott says:

    The hope was to find children who hadn’t been abused before their first IQ test at age 3, but who started being abused before their last IQ test at age 18.

    My siblings and I would have made great datapoints for this study. I was 3 years old with a sky high IQ when my mother came down with schizophrenia, while my sister was 5 at that time and my brother was 3 or so months old. My sister was above average IQ at the time and my brother couldn’t be tested for obvious reasons.

    My mother was most abusive to my sister during those years, I won’t go into detail, but I would call it “moderate” with one or two “severe” episodes. I got away easy with just a couple episodes of light abuse, and my brother was more or less ignored completely by my mother for the entire duration of her illness in our home (4 years, after which she was institutionalized for the following 10 years).

    Fast forward a few decades into the future and my sister and I are both doing fine with our bachelors degrees and white collar jobs, while my brother, the ignored one, struggled with learning disabilities all through elementary and high school, stopped taking college courses after a single semester, and now bounces aimlessly from one menial blue-collar no-benefits job to the next with long periods of unemployment in between. I feel like I can tell from his interests and hobbies (watching PBS, writing/recording/performing music) that he could have developed into an intellectual too but his development was squelched by neglect.

  19. Lila Rieber says:

    I don’t buy that there’s a significant number of people who were abused after the age of 3 but received adequate parenting prior to that. Infants and toddlers are if anything more challenging than young children. I also don’t buy that there are families in which one identical twin is abused and the other is completely spared.

    Somewhat unrelated, I’m not saying it’s good to write articles falsifying evidence for the anti-child-abuse/neglect position, but I don’t understand the motivation of researchers who seek out justification that child abuse/neglect is fine. It sends a noxious message, and I’m not sure what it accomplishes.

    • Gazeboist says:

      I don’t think “justification that child abuse/neglect is fine” is the goal for most researchers, at least when you phrase it that way (though “abuse victims are not condemned to be permanently warped by their trauma” is a plausible one, based on conversations above). We should always be checking our own evidence, especially when we’re making “consequentialism as communication” arguments. *If* we are wrong, and someone who was convinced by a false argument finds out, they are likely to discard the conclusion as well, despite the fact that we might have made a better one. Take sodomy laws, for example. It’d be pretty terrible if (hypothetically) Texas and Utah reinstated their sodomy laws just because someone found out that bihacking was plausible. Same with “conversion therapy” – the problem with conversion therapy isn’t that it won’t work, it’s that a parent has no business interfering with their child’s sexuality.

      The first rule of arguing from facts is that your facts must be correct. Working to ensure that this is the case is admirable.

      • Ozy Frantz says:

        “Conversion therapy doesn’t work” is a really important argument in favor of banning conversion therapy for adults. Ethically, it’s really dubious to refuse to let people get therapy to change something about themselves that’s worsening their quality of life unless the therapist is fraudulent.

        (Ex-gay therapy is mostly fraudulent.)

        • Controls Freak says:

          For the record, even the APA (which is a politicized organization in this space) believes that the current research for conversion therapy is shit. That is not, “Conversion therapy doesn’t work.” That is, “We have no bloody clue, because the research is shit.” At least, it is if we’re being honest.

          • Jameson Quinn says:

            Well, some people have a clue. The people for whom conversion therapy has not worked, and (if any exist) those for whom it has worked. We know that at least some of the former exist. This may not be the strongest clue, but it is a bloody clue.

          • John Schilling says:

            Do lottery winners generally understand the odds of winning the lottery? Do the losers?

          • Jameson Quinn says:

            At what odds of working would conversion therapy be worth it?

            Say you have two people. Person A says “I want conversion therapy, even if the odds are just 10%. If it works, I can save my relationships with my family and friends.” Person B says “10 years ago, I would have said the same as person A. Boy, was I stupid. I was ignoring the problems with my existing relationships and pretending CT would be a magic bullet. Now, I wouldn’t retroactively take it even if the chance it worked were 100%”.

            Obviously, you can’t let person B choose for person A. But I’m not sure you can trust person A to clearheadedly evaluate their breakeven odds either.

          • Controls Freak says:

            I don’t see how constructing different utility functions that drive differing behavior amongst individuals who decide to play the lottery or not play the lottery has anything to do with whether we can compute an objective, mathematical description of lottery success.

          • Jameson Quinn says:

            My point is that the exact odds of lottery success are irrelevant.

            You have group A saying “we built this lottery, because we believe the prize is highly valuable. We’re not very interested in testing any of our assumptions scientifically.

            Group Z is saying “that prize has negative value, and the lottery never pays out anyway, and we’re not very interested in testing either of those things scientifically either.”

            Group M says “We can say for sure that it doesn’t always pay out, and we can’t say any more, because group A are bad at pretending to be scientists.”

            There’s also groups C and X saying that they played the lottery and it did/didn’t pay out for them, but between those two there are reasons to doubt C’s veracity, and enough X members to suspect that at least some of them are truthful.

            The lottery should not exist if any of the following are true:
            1. The prize has negative value.
            1a. The cases where the prize has positive value have alternate solutions with higher benefits and/or lower costs.
            2. The lottery never pays out.
            3. The odds are low enough that the costs are higher than the benefits.

            I believe we know that 1a is true almost always; that the evidence suggests strongly that 3 would be true, even if you assume 1 isn’t; and that simple heuristics suggest that 2 may be true. I think that’s all the evidence we need; so the exact odds are not the point.

          • Controls Freak says:

            I never said we need exact odds, so your entire comment is tilting at a windmill. I’m not even going to reiterate my point; I’ll just let you take a mulligan on the whole reading thing. But this time, please try to avoid arguing against things I haven’t said.

      • keranih says:

        Same with “conversion therapy” – the problem with conversion therapy isn’t that it won’t work, it’s that a parent has no business interfering with their child’s sexuality.

        Given the variety of expressions of sexuality in the human range, and given the variety of definitions of “interferring”, I think this is a postion very far out of the normal accepted range.

        I would not hold it at all as abusive for a parent to attempt to introduce a child to potiental dates who were not the childern of drug dealer or thieves. I don’t even think it’s at all harmful to suggest that a partner from the same religion would be better for a long term relationship than one of a different religious background.

        I absolutely think it’s appropriate for a parent to steer their child away from sexuality focused on inflicting pain on their partner, involvement with close relations, and with beasts. And I think this is both correct and widely accepted.

        We might quibble about where to draw the line, but the line is there.

        • Gazeboist says:

          I suppose I’d say the line sits where the parent has an interest as a member of a society – there’s no special interest as a parent, just a position of special power (and thus increased responsibility to avoid doing harm).

  20. Tyrrell McAllister says:

    > more specific measures of executive unction

    Do we really want more executive unction?

  21. Charlie__ says:

    I have two complaints: One is the usual sort of complaint I make about studies like this, and another is something I consider bleeding obvious.

    I think the most likely causal path from abuse to lower IQ is through malnutrition, and the second is neglect as an infant (unless you think that being raised by wolves does not make you do worse on IQ tests), but with n=3000, among parents who are conscientious enough to participate in this study, how big of a signal are you really going to get?

    In my middle school class of 400, I would estimate there were 2 kids whose poor nutrition was linked to abuse, and about 20 more who just had bad eating habits, unstable homes, poverty, etc. Would the parents of those 2 kids have signed up for a study? And even if they did, would the study have been able to tease out the signal of 2 kids being deliberately mistreated by their parents from 20 kids just not eating right, just from their IQ scores? Heck no!

    The only effect this study was qualified to find was whether hitting your child as a form a punishment made them score worse on IQ tests, because there they could actually get a decent sample and have some hope of fighting selection bias. Even if one argues that types of abuse correlate with each other, the effect size is still too small and the noise too big. This study is interesting, but I think that it’s searching under the lamppost, rather than looking at the types of abuse we’d expect to have the strongest effect on life outcomes.

    And now the bleeding obvious: Why no adoption studies? If you want to avoid genetics, just track adopted kids. Of course, if I think it’s obvious, that probably means there are already published papers on it out there. Right?

    • Charlie__ says:

      Huh. It is surprisingly hard to find a study about abuse and adoption that’s about the effect of abuse after adoption, rather than before.

      In the course of a short literature search, though, I’ve learned that there are two semi-contradictory genres of studies of adoption and IQ. The first genre (ex) finds that if you take kids from one demographic and have them raised by another demographic, you can predict their IQ by just about averaging their birth demographic and adoptive demographic. This genre has been done with children of rich and poor parents, of white and black skin, and of different national origins.

      The second genre (ex) finds that if you look at adopted children’s IQs over time, by the time they hit puberty their IQ is only weakly correlated with their adoptive parents’ IQs, but is correlated with their biological parents’ IQs similar to how non-adopted children’s IQ is correlated with their biological parents’. This data mostly comes from detailed adoption studies that are localized (e.g. in Colorado), but take measures to try to be representative within that location. Papers in the second genre will often mysteriously forget that the first genre exists, and vice versa.

      If you’re clever, you may notice that the second genre of study seems like it should contain enough data to also replicate the apparently-contradictory first genre of study. I think this has been done with the colorado adoption data from my second example above, but the results of the first genre were not replicated, and the contradiction remained! This may be the sort of thing where asking a roomful of subject experts may be the only way to figure out what’s going on.

    • Marshayne Lonehand says:

      Along similar lines …

      Question  Why is there no outcome-measure overlap between the OP-referenced study “Origins of Cognitive Deficits in Victimized Children” and the above-referenced study “Preschooler Emotion Regulation in the Context of Maternal Borderline Personality Disorder”?

      Motivation  The study “Origins of Cognitive Deficits …” focused chiefly upon (what might be called) “Cluster R” cognitive measures, that is, measures associated to ratiocinative cognition (cf. “intelligence coefficient”, IQ). In striking contrast, the study “Preschooler Emotion Regulation …” focused chiefly upon (what might be called) “Cluster E” cognitive measures, that is, measures associated to empathic cognition (cf. “emotion regulation”, ER).

      It’s thought-provoking — and even discomfiting (isn’t it?) — that the former study provides no rationale for not studying ER, while the latter study provides no rationale for not studying IQ. Perhaps these two research communities should talk to each other more than they do?

      The contrast between these two studies provides motivation to ponder, in respect to clinical practice, which cognitive cluster is more characteristically associated to acute psychological distress and dysfunction: “Cluster R” (ratiocinative) deficits or “Cluster E” (empathic) deficits?

      More simply, which is the more prevalent clinical complaint: “My high IQ does not protect me from being emotionally miserable and socially isolated”, versus “My richly fulfilling emotional life and loving circle of family and friends does not translate into outstanding GRE/MCAT/LSAT/Mensa scores”?

      According to Hollywood stereotype, clients with exceptional “Cluster R” skills show up in therapy all the time, whereas exceptional “Cluster E” skills are the domain of the therapist (or even the Creator). Why is the reverse stereotype less commonly seen — save to shavian effect … such as the ever-popular genre of romantic comedies in which the risably “R”-strong / “E”-deficient therapist receives healing and/or learns how to love … the ability to whole-heartfully give and joyfully receive love being, for Hollywood at least, the most-prized and most-universal Cluster E skill (by far)?

      Summary question  Are there appreciable reasons to provisionally regard IQ-centric/”Cluster R” abuse-studies as generically insensitive to the “Cluster E” cognitive injuries and empathic deficits that, both clinically and culturally, are (apparently) the most harmful psychological sequelae to severe childhood abuse?

  22. Jameson Quinn says:

    When you adjust for something, and the p value goes up, that could be because your adjustment removed the effect. Or it could be because you used up too many degrees of freedom in making your subgroups and applying your adjustment. Furthermore, even if that’s not the problem, going from p<.01 to p=.13 when re-analyzing adjusted data is in itself a change that you'd expect to happen with p=.08 or so for real effects of the appropriate size… that is, something that you wouldn't call significant on its own.

    In other words: this is interesting enough to merit further study, but Scott (like many others in this thread) is reading way too much into it.

    …I just skimmed the study. I was biased against it before I read it. It did not surprise me — it was about as bad as I expected. That is, there were not any glaring statistical errors obvious without going over the numbers with a calculator, but it certainly did talk as if failure to reject the null was synonymous with evidence against the hypothesis. Without checking it more closely, I’d say that this paper is probably about as good evidence for its conclusions as the average psychology paper, which is to say, better than nothing, but with a significant chance that its findings wouldn’t replicate.

  23. JayMan says:

    It’s worth pointing out that susceptibility to PTSD is heritable, like everything else. It is possible that childhood trauma can lead to adult PTSD in those genetically susceptible to such – a true gene-environment interaction. But the track record on GxE has been pretty bad so I’m not placing my chips there, either.

    Twin control studies can only really prove negatives (i.e., no effect) than show positives because MZ twins aren’t really identical. You have no idea if what you’re seeing is an effect of whatever environmental exposure in question or just a manifestation of pre-existing twin differences (due to developmental stochasticity, for example).

    So I’d say the trauma exception doesn’t have a lot going for it.

  24. deciusbrutus says:

    ” For example, I’ve seen some kids from well-off, genetically-healthy families who were sexually abused by a coach or a neighbor or a priest or someone and who became nervous wrecks later in life, unable to stop thinking about it or to stop feeling guilty. ”

    Have you seen kids from well-off, genetically-healthy families who were not abused by anyone and became nervous wrecks later in life?

    Do you have numbers backing up the causal claim, because that quoted sentence is both being used to justify a conclusion and made of pure confirmation bias.

Leave a Reply