Do techie parents have more autistic kids?

2 Oct

This is the stuff of legend. So says the Simons Foundation blog in discussing a new epidemiological study out of California. A previous study reported that fathers of ASD kids are more likely to be engineers. The idea got into the mainstream media (and deep into the public psyche) when Wired Magazine even called the phenomenon the “Geek Syndrome“.

Well, someone checked. Someone being Gayle Windham, Karen Fessel and Judith Grether. If you are an autism epidemiology geek, you will recognize at least some of those names.

Here’s the abstract:

Autism spectrum disorders in relation to parental occupation in technical fields
Gayle C. Windham, Karen Fessel, Judith K. Grether

A previous study reported that fathers of children with autism spectrum disorders (ASD) were more likely to work as engineers, requiring systemizing skills, and suggesting a distinct phenotype, but alternatively this may have been related to selection biases. We conducted a population-based study to explore whether fathers, or mothers, of children with ASD are over-represented in fields requiring highly technical skills. Subjects included 284 children with ASD and 659 gender-matched controls, born in 1994 in the San Francisco Bay Area. Parental occupation and industry were abstracted verbatim from birth certificates. Engineering, computer programming, and science were examined as highly technical occupations. To limit bias by parental socio-economic status, we selected a referent group of occupations that seemed professionally similar but of a less technical nature. Odds ratios (ORs) and 95% confidence intervals (CI) were calculated by logistic regression, adjusting for parental age, education, and child race. Mothers of cases were somewhat more likely to work in hi-tech occupations (6.7%) than mothers of controls (4.0%, P=0.07), but little difference was observed among fathers, nor for engineering separately. Compared to parents in other white collar occupations, the adjusted OR for highly technical occupations among mothers was 2.5 (95% CI: 1.2-5.3) and among fathers was 1.3 (95% CI: 0.79-2.1), with no evidence of a joint effect observed. Our results regarding maternal occupation in technical fields being associated with ASD in offspring suggest further study to distinguish parental occupation as a phenotypic marker of genetic loading vs. other social or exposure factors

To summarize it even more: mothers in “highly technical” jobs had a 2.5 times higher chance of having an autistic kid. There isn’t a clear statement for fathers. Also, there isn’t evidence of a “joint effect”. I take that to mean that the “geeks marrying geeks have more autistic kids” idea didn’t pan out.

They can’t say why mothers in highly technical jobs have more autistic kids. This makes it a study that can be quoted by everyone. “It’s genetic, see mothers in technical jobs have more autistic kids”. “It’s social, mothers in technical jobs are more likely to know about autism and get their kids diagnosed”. “It’s environmental, mothers in technical jobs are exposed to more toxins”.

In the end, this is no joke. We need to understand all the factors that can affect epidemiological studies. Epidemiological data for autism are pretty muddy, in my opinion. A lot of things have happened in the last 20 years (or more) that have changed the “rate” of autism diagnoses. Anyone who wants to find one of the many causes of autism (which is, as we all know a plural: autisms), needs to understand these factors.

26 Responses to “Do techie parents have more autistic kids?”

  1. navi October 2, 2009 at 19:25 #

    yep lossa environmental toxins while I was sitting in front of a computer doing quality assurance, and then not having a job and being unemployed. I wonder if my riding the bus caused my son’s autism. um, ya, right. ;0)

    on the environmental toxins front, though. My husband worked with epoxy in road construction prior to my pregnancy. wasn’t working with it prior to my other kids’ pregnancies. that’s about the only environmental difference between the three. I still think it’s genetic.

  2. navi October 2, 2009 at 19:26 #

    especially bc none of his other family working in road construction have autistic kids, but a distant relative on his grandmother’s side has 2 of ’em…

  3. Jeanne October 2, 2009 at 19:52 #

    My experience is not the same – it’s quite different. I am not in the least bit “techy” and neither is my son’s father. Both of us are best described as the artistic/left brained type – and so is my son. However I think it is great to exhaust all “what if’s” and “maybe’s” when researching autism – the cause, who will get it, environment, etc….it’s either something so “ou there’ that we have not figured it out yet -OR- it’s somthing so simple and right under our noses. So, even if things sound like “Nah, that can’t be it…”, let’s look at everything and one day we’ll figure it out.

  4. passionlessDrone October 2, 2009 at 20:40 #

    Hello friends –

    I’m conflicted by these findings. On one hand, it’s great fun to blame my wife (engineer), her [idiot] father (engineer), or my father (engineer) for Luke’s autism. On the other hand, as a software developer, I now can point to this as evidence that it wasn’t me. Ho ho.

    On the serious side, nice write up of a study I wouldn’t have run into otherwise. Thank you, Sullivan.

    – pD

  5. Emily October 2, 2009 at 20:51 #

    Hmmmm…spouse is software developer, I’m a (female) biologist with a humanities brain. Dunno.

  6. Laurentius Rex October 2, 2009 at 22:28 #

    Trouble is if you know your history you will realise this is an old throwback to Kanner, it is actually an artefact of the particular advantage of the educated middle classes in getting access to diagnosis and claiming the commanding heights of autism.

    FFS get real there is a whole substratum there, it has to be demographically and sociologically.

    And you all say your science is pure, well what would you know? what would you all know?

    • Sullivan October 2, 2009 at 22:35 #

      “…it is actually an artefact of the particular advantage of the educated middle classes in getting access to diagnosis and claiming the commanding heights of autism.”

      Isn’t that what they tried to account for by comparing to other “white collar” jobs?

      This isn’t the be-all and end-all of autism epidemiology. It could be wrong, or only apply to families in the San Fransisco bay area in the 1990’s.

      I’m just interested that someone actually went out and checked some real data.

      “what would you all know?’

      It’s safe to say that what we all know is a lot less than what we don’t know. Pretty safe to say on just about any subject. I missed the part where I said that my science is pure, though.

  7. Sullivan October 2, 2009 at 22:40 #

    My experience is not the same – it’s quite different. I am not in the least bit “techy” and neither is my son’s father. Both of us are best described as the artistic/left brained type – and so is my son.

    The “legends” have you covered as well–I’ve heard that musicians are more often found with autistic traits.

    We live in Lake Wobegone, where all children have an above average chance to be diagnosed autistic.

  8. passionlessDrone October 2, 2009 at 22:43 #

    Hello Laurentius Rex –

    Trouble is if you know your history you will realise this is an old throwback to Kanner, it is actually an artefact of the particular advantage of the educated middle classes in getting access to diagnosis and claiming the commanding heights of autism.

    I felt that the theory made sense based on what I’d been told, and seemed to have read, that in many cases people with autism have spectacular spatial skills and/or the ability to focus very tightly on a problem; skills and abilities that in fact, are quite useful for a profession in engineering.

    Your experiences may be different.

    – pD

  9. Joseph October 2, 2009 at 23:13 #

    The equivalent Baron-Cohen work was a survey of NAS membership. Maybe that’s where the selection bias is. I recall they did address the possibility of a general professional over-representation in a follow-up.

    They also had a survey of university students, where they looked at relatives of engineering/math students vs. humanities students. Of course, this would tend to be about familial autism, and maybe it matters if autism is familial. (Clearly, if the hypothesis is correct at all, it would be because of familiality.)

    Perhaps they need to do a follow-up involving sibling pairs who are autistic.

    It’s interesting, though, that the Baron-Cohen work was completely about fathers and grandfathers, and the new work says it’s the mothers. Reminds me of the father’s age vs. mother’s age debate that is currently being ironed out.

  10. Irene Burton October 3, 2009 at 09:52 #

    This is fascinating. 20 or so years ago It was all to do with ‘cold’ mothers, or if that didn’t fit it was the result of ‘too intelligent’ parents. In my experience at that time also, diagnosis was delayed.

    My daughter was a year old when I started to question her development, at 2 years they took my worries seriously, at 3 years I was given a reading list, but no diagnosis. The pediatrician would not use the word autism. It was almost another 3 years before we gained access to a specialist who actually said and more importantly put into writing in our daughters medical record, severe autism.

    My husband and myself are both orientated towards the humanities and the arts, our other daughter is writer, technology scares us to death, its nothing short of a miracle you will be reading this, I struggle with the internet etc.

    These studies are important, they all give us a bit more information, we all need that. I do not pretend to understand the ins and outs of how these studies are put together, but what I do notice, is that diagnosis and the correct recording of it, is probably the key. Worrying about what the parents do for living seems to me to be just harking back to those old stale ‘are they intelligent or not?’ questions that have got us nowhere in the last 20 years.

    We’re going round and round in circles folks, and you know what we’re all going to end up disappearing into if we don’t give it a rest!

  11. Anna Hayward October 3, 2009 at 11:26 #

    As an Asperger mother of autistic kids, who used to work in IT, the reason is as plain as the nose on your face: People with mild autism, and autistic-type personalities (too mild for a diagnosis) are often good at techie, systemising jobs. If you’re bad with people and good with computers, you’re going to be attracted to that kind of job. Since autism has a huge genetic component, these techie, semi-autistic parents get together (birds of a feather and all that…) and produce autistic kids.

    Maybe the problem is that people only ever see the bad side of autistic traits and are unaware that intense focus, extreme logic and mathematical brains can be a distinct advantage?

    To me, it is almost funny that people can’t see this and frett about the “Silicone Valley Effect”, particularly when the very parents doing the fretting are (to me) blatantly autistic themselves (albeit high functioning). “The apple doesn’t fall far from the tree” as they say.

  12. Laurentius Rex October 3, 2009 at 13:46 #

    Well Joseph, it is indeterminacy at work, a number of conflicting but equally surface plausible data and hypotheses.

    Sommats wrong somewhere and it won’t be determinate so long as people stick rigidly to whichever school of thought they are (somewhat predetermined by circumstance if not neurology) to follow.

    Besides it raise the age old nature or nurture theme, that possibly social learning is at work, as I say again Who knows?

    BTW isn’t “silicone valley” where Baywatch is set 🙂

    As for all autism science, I am determined (predetermined maybe) to remain sceptical until they widen the diagnostic criteria to take account of sensory and perceptual difference, simulataneously acknowledging the bleed over into the complex Venn diagrams of what really constitutes neurodiversity, given that a lot of the so called spatial superior traits are also said to be common amongst dyslexics, architects standing in for engineers with that particular meme.

  13. Laurentius Rex October 3, 2009 at 14:43 #

    Anna I am somewhat frustrated when people on our side of the cure debate use phrases such as “the reason is as plain as the nose on your face:” to support individual observations because that is precisley what those who insist that vaccines damage there children use as an argument too, that is to say the triumph of personal observation and anecdotal evidence over sustained and rigouros scientific study.

    I think there are so many flaws to going with the SBC team on this. I am concerned that all that is happening is that a self defining set is being created, whereas the validity of that needs to be determined from outside even if it means the walls of the notional set become rather porous in the process.

    The systematising extreme male hypothesis is essentially prederming the argument by saying that autism is and only is X Y and Z thus excluding any other possibility or argument. “If it is not X Y and Z then it is something else we won’t allow in.”

    It makes little sense at all with longitudinal studies because what is now X Y Z has heretore been X and P or P Q and Z to stretch the analogy.

  14. Laurentius Rex October 3, 2009 at 14:53 #

    And now taking on Passionless Drone and Anna together, (a formidable task)

    I anser, yes my experiences are different because I tend to fall in with Artistic Aspies, it is perhaps as Anna says “birds of a feather and all that.

    Now to avoid the pitfalls of retrodiagnosis, I would not want to recruit JS Bach into the fold (unless prof Fitzgerald has been at him already)

    But what was he? Accomplished performer, creative genius or sound engineer?

    Well in truth part of the tools in trade of an organist in those days was recommmending the specification of organs to be fitted in churches, but that is no different from say recommending a Gibson guitar over a Fender.

    No the interesting part about it’s music is it’s mathematical precision, although of course Moondog disagreed with that, finding nitpicking faults in the counterpoint, but then I do claim Moondog as a retro dx, he was way too eccentric to be NT and definately my kind of aspie 🙂

  15. Bill October 3, 2009 at 18:38 #

    I am an engineer, and I am endowed with Asperger’s Syndrome (professionally diagnosed).
    I have been studying Asperger’s/Autism for over ten years because of my curiosity about the syndrome, and I reasonably think I have a pretty good ability to spot the symptoms in other people I interact with frequently. At the last large consulting engineering firm I worked at, I estimate 25% of the engineers, designers and drafters have some form of autism.
    The recent British study of autism in adults suggested that autism spectrum adults were less likely to finish college, so if they are less likely to finish college, yet 25% of college educated engineers have Asperger’s, it would appear I have noticed the tip of a rather large iceberg. Either that or men endowed with Asperger’s in the US are more likely to get admitted to and finish college educations than they are in the UK.
    There certainly seems to be a strong genetic component involved; I suspect my father and two of his brothers were engineers with AS. Their father was an engineer. I have a brother diagnosed with AS (a systems programmer) and two engineer brothers I suspect have AS. Another brother has AS so bad he barely functions in society. Four sisters are obviously not normal, and all four have worked highly technical jobs in the computer/electronics field.
    Of my three sons, One barely functions in society, one made it through college but has attention deficits, one is working on a highly technical degree and pretty clearly has AS. I have several cousins in technical fields who I suspect have Asperger’s.
    It could all be coincidence. It could be my family shares a unique and unusual autism/Asperger’s gene. But please forgive me for suspecting a relationship between engineers and Asperger’s.

  16. Joseph October 3, 2009 at 22:43 #

    @Bill: The number of engineering jobs in the US is about 2 million. That would be about 0.8% of the adult population. Even if 25% of them are autistic, prevalence of autistic engineers would be 0.2%, or about 1/5th of all autistics. Then you have to consider that most of those engineers might not meet threshold criteria in something like an ADOS evaluation, even if they do have autistic traits. So yeah, engineers are a small minority of all autistics.

  17. ebohlman October 4, 2009 at 02:56 #

    To summarize it even more: mothers in “highly technical” jobs had a 2.5 times higher chance of having an autistic kid.

    Nope. You’re misinterpreting an odds ratio as a relative risk. You can’t even properly estimate relative risk from a retrospective design like this (one where the number of subjects with and without the outcome condition was determined by the researchers) and even in a prospective study the odds ratio is only close to the relative risk if the probability of the outcome condition (autism in this case) is very low.

    For readers not familiar with the terminology, relative risk is the ratio of the probability of a “positive” outcome under one condition (e.g. mother in technical field) to the probability of a positive outcome under another condition. An odds ratio, on the other hand, is the ratio of the odds of a positive outcome under one condition to the odds of a positive outcome under the other condition. The odds of a positive outcome is (“odds” is a singular noun) the ratio of the probability of the outcome occurring to the probability of its not occurring. If p is the probability of the outcome, the odds of the outcome is p/(1-p).

    The odds ratio is meaningful for any 2X2 association, regardless of how the data were obtained. The relative risk is only meaningful for certain kinds of study designs; that’s why the odds ratio is what’s usually reported.

  18. Jake Crosby October 4, 2009 at 12:41 #

    @Joseph: ADOS evaluations are practically worthless, and the reason why autism is associated with some engineers is because they are toxic like you.

    • Sullivan October 4, 2009 at 15:18 #

      @Joseph: ADOS evaluations are practically worthless, and the reason why autism is associated with some engineers is because they are toxic like you.

      Jake Crosby,

      volunteer to go into IEP meetings and tell the school administrators this. See how many of your readers take you up on the offer.

      My guess is you would get booed off the stage, so to speak. ADOS is the standard. It is the number one test that is being translated for use in other countries.

      As a second exercise–only consider people who were diagnosed via instruments other than ADOS. Since most kids are, and far fewer adults were (sit didn’t exist when they were young) you will have an even harder time convincing people of an autism epidemic,

  19. Joseph October 4, 2009 at 13:58 #

    ADOS evaluations are practically worthless

    @Jake: Did you think that before or after the NHS study? I believe we all know which one it is.

    In reality, the ADOS is one of the best diagnostic instruments there is, if not the best.

  20. Joseph October 4, 2009 at 14:24 #

    In fact, it boggles the mind a little that people have no problem with the methods of the CDC to count autistic children, for example, which consist of two questions by phone to the parents:

    1. Has a health professional ever told you that child has ASD?

    2. Does child have ASD currently?

    So this is OK. An ADOS evaluation, not so much. Yeah, in Bizarro World. Wait, I think I just described AoA.

  21. Dedj October 4, 2009 at 16:40 #

    The ADOS is one of the best diagnostic tools available.

    It’s not flawless. It has some limitations but so does COTNAB, MOHOST, AMPS and numerous other widely known and used assessments.

    People with autism are noticably ‘atypical’ in the areas that are tested by the ADOS. Someone with even pre-registration clinical and academic experience of autism would have likely made these observations themselves.

    “I don’t like it” is not an arguement.

  22. Laurentius Rex October 4, 2009 at 17:52 #

    I do have issues with ADOS, it would be rather surprising if I did not coming from the direction I do on the social construction of autism and disability and in deconstructing the social practice of medical research and diagnosis.

    The reason it has become a “gold standard” has much to do with the stranglehold it has got over the research in terms of the need for some kind of standardisation to intervalidate various studies.

    That does not mean it is the best, or even more the most equitable as it actually is an exclusive academic piece of sharp practice that leaves out those researchers who cannot afford the costs of ADOS evaluation for there research cohort.

    That is one of the reasons why the Cambridge Battery of tests which are available free to any researchers without the various restrictions on ADOS and other proprietory observation schedules and instruments.

    There was an interesting discussion on the merits and demerits of these various tests at the most recent NAS international conference.

    There is a lot more to it, and behind the scenes than most people who merely read the research realise.

    I happen to think that the entirety of Autism science is flawed, and it is remarkable that there is as much consistency as there is.

    I see the role of the researcher is rescuing what can be from the embers of this burnt out edifice, rather like archeology in a way, because ADOS certainly does not give the whole picture.

    What I do see in these discussions though is a microcosm of the social practice of research, in that people cleave to the studies they prefer and are willing to have lesser standards for what they are predisposed to agree with than that they wish to contest.

    It is a subtle bias that can be seen in almost any reviewer.

  23. Audrey October 11, 2009 at 02:09 #

    Another hypothesis the findings are consistent with (and one that explains why mothers’ nerdiness increases the OR more than fathers’) is that the contributory factor is the hormonal milieu in which the fetus develops. In other words: Nerdy women may be more likely to have higher levels of circulating androgens which the fetus is then exposed to in utero. I know Baron-Cohen’s team in Britain is investigating a similar hypothesis. Not sure I’m sold on it yet, but it sounds plausible.


  1. Tweets that mention Autism Blog - Do techie parents have more autistic kids? « Left Brain/Right Brain -- - October 2, 2009

    […] This post was mentioned on Twitter by Dominique DeShields. Dominique DeShields said: Reading: "Autism Blog – Do techie parents have more autistic kids? My experience is the opposite! ( ) […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: