Autism as a Disorder of High Intelligence | Hacker News Reader

Autism as a Disorder of High Intelligence

journal.frontiersin.org

90 pointsjonbaer34 comments

One popular anecdote or perhaps paraphrased statistical tidbit in the last few years was that incidences of autism were unusually high in Silicon Valley and particularly children of developers. Not that all developers have "high" intelligence--but higher than average on the whole? Just wondering if there might be any strengthening arguments to be made from connected studies. Of course, could be a red herring that masks the critical factors, but if we had to bet on the likelihood of a connection...maybe something there.

you might as well check a sample of developers for this disease and check against a sample of the general population. its not just the children, you know.

This article reports on a study finding a link between "systems thinking" oriented parents and autism.

http://arstechnica.com/science/2011/06/kids-autism-may-be-re...

One possible mechanism: One of the proposed reasons for the recent rise in levels of autism is that more autistic children are sent to the psychiatrist.

You need to be relatively "rich" to be able to afford sending your child to the psychiatrist. Most programmers in Silicon Valley are relatively rich. Therefore, more children are sent to the psychiatrist in Silicon Valley than in the "rest" of the population, so more children with autism are diagnosed.

This is confusing genetic markers for intelligence with intelligence. Building half a bridge is not useful, and individual markers don't stand alone.

He mentions that a: autists are dumber (i.e. do not express some or most of the markers for high intelligence) but posess the genetic markers, and b: the expression of intelligence markers of siblings/parents/higher-order relatives of autists (specifically) is understudied and recommends it for further study. [As far as I read it anyway]

I do believe that the "imbalance"-hypothesis he presents is a bit vague and proposes several imbalances according to different (and possibly orthogonal) models for intelligence, making it seem a bit unfalsifiable to me. Who knows. Behavioral neuroscience based on MRI(or similar modalities) is a giant sandcastle anyway, best not stack those results too high.

I agree that the details in some areas were vague, but this is a broad overview and could lead to something testable if the research is followed through.

The short version, to me, was that general intelligence is based on several cognitive axes working together. I believe the paper lists 3-4 most of which were separable, like verbal scores, mental rotation tests, etc. And with Autism, you have 1-2 of those axes working above average but the rest not. To falsify, you could break down IQ tests for individuals with autism and compare some areas with the general population and also with each other. If they match the expected pattern, they may be onto something, otherwise the theory is incorrect.

From what I read, that is exactly the point the author is trying to make. Namely that having some of the markers for high IQ, and missing others causes their overall performance to drop even if there are a few areas where they are above average.

Some background on this particular journal: http://www.nature.com/news/backlash-after-frontiers-journals...

I am not at all an expert on autism, but my understanding is that one of its features is a strong deficiency in the ability to read human nonverbal communication.

All of their markers for intelligence are suspect.

> including full-scale IQ and a PCA-based measure of g (Clarke et al., 2015), childhood IQ, college attendance, and years of education (Bulik-Sullivan et al., 2015), cognitive function in childhood and educational attainment (Hill et al., 2015), and verbal-numerical reasoning and educational level reached (Hagenaars et al., 2016)

I bet all those markers are more closely correlated with income levels in childhood than with occurrence of autism (or anything else).

Knowing, in isolation, that IQ is more closely correlated with income levels in childhood than IQ is with thing X, does not imply that IQ is not a causal contributor of X.

It could just be the case that IQ is partly genetic (in fact, it is) and that IQ may be a causal factor in income levels to a stronger degree than it is for autism or other things.

Right, and in fact this is known to be the case: http://biorxiv.org/content/early/2016/04/16/048991

My summary: https://brainslab.wordpress.com/2016/05/07/shared-genetic-co...

Hasn't IQ been proven to be not just partly, but mostly genetic? Barring exposure to outright toxic environmental conditions.

IQ has the highest genetic correlation out of any trait we've observed in Psychology - r ~ 0.7 if I recall correctly from my college days.

This is actually interesting in terms of GWAS - there used to be a large difference between "heritability" (~the measurable variance offspring shares with their parents, like height in cm, which can be influenced by a ton of parameters if you don't know about them/don't control them [like nutrition]). In the age of genome-wide association studies the linked SNPs often correlate much weaker than the heritability predicts (example: you could explain 10% of variance with SNPs, but heritability says it should be 70%) - this was called the problem of "missing heritability".

Some more complex models recover some of that missing heritability, these slides are a nice summary: http://jvanderw.une.edu.au/Mod9Lecture_SNP_Her.pdf

There is a paper from 2011 looking at human intelligence (http://www.nature.com/mp/journal/v16/n10/full/mp201185a.html) and using a little bit of modeling they got the correlation up high:

>We estimate that 40% of the variation in crystallized-type intelligence and 51% of the variation in fluid-type intelligence between individuals is accounted for by linkage disequilibrium between genotyped common SNP markers and unknown causal variants.

However, look at that "unknown", because the same abstract says:

>Finally, using just SNP data we predicted ~1% of the variance of crystallized and fluid cognitive phenotypes in an independent sample.

Using their SNPs alone it's only 1%. I assume the "true" genetic variability (~~~heritability) is somewhere in between those two values, since the 40/50% number seems to assume that these unknown variants will be discovered (they haven't yet AFAIK, maybe they don't exist, maybe they do).

Many of the studies showing this have some serious flaws, because of the types of people they tended to use. The twins used in the studies where disproportionately from higher income families. You don't find many low SES families who adopt, so for this and other reasons it's very difficult to find twins raised apart in low SES families.

The summary is that it's pretty clear that among higher SES families IQ is more than 50% genetic (but environment still plays a very large role). However, many studies indicate that among lower SES families, environment is the dominant factor.

There's still conflicting evidence here though, because there are some studies that indicate that SES has little impact on IQ heritability. The argument is that SES has a much greater on the heritability of IQ in children than on that of adults.

One idea that I ran across is that as SES rises, genetic contributions to success start to become the greatest factors.

Low SES environments are likely to cause malnourishment and other problems that cause people to fail to live up to their potential.

Thus, genetic explanations for inequality are more plausible among populations of high SES.

Right, but I wasn't trying to make the whole argument in one little post. The problem is that all our measures of intelligence have never been shown to be good measures. We've just sort of assumed that intelligence is a thing. When it fact it may be that any child who's provided with the resources and support to meet high expectations will meet them.

So what they could do here is show that Autism occurrence is more closely related to their measures than to income levels in childhood, but I don't think they did that (I confess I didn't read the whole thing).

It seems odd that you left out the start of that sentence:

"All four of these studies, which used diverse, independent populations and tests (or correlates) of cognitive abilities, have reported significant, substantial genetically-based positive associations of autism risk with intelligence, notably including full-scale IQ..."

I think the part of the sentence where they used "diverse, independent populations" and obviously tried to control for effects like income levels in childhood undermines your criticism. More to the point, the studies mentioned were about measuring the IQ of individuals with known alleles for autism. You can attempt to argue that childhood income may have a stronger affect on IQ than autism does, but the studies themselves showed there was some positive effect of Autism on IQ.

Well, I know I'm definitely too weird to get a date. It's nice to think that means I'm intelligent.

Eric Raymond's take :

"Neurotypicals spend most of their cognitive bandwidth on mutual grooming and status-maintainance activity."

http://esr.ibiblio.org/?p=7060#more-7060

It just doesn't stand up to scrutiny, especially given how many people are autistic.

An estimated 1% of the population currently (including those estimated as not diagnosed). Subject to rise as the DM-4/5 diagnostic criteria adapts. Why does this fact of 1% of the population, mean the article is not viable?

I'd think the amount dropped significantly once Asperger's was removed from the spectrum in DSM-5 (although this article seems to ignore that fact).

Asperger's was folded into the spectrum, and the actual condition was removed as a "standalone".

Correct me if I'm wrong, but my understanding was that references to "Aspergers" were removed, but the diagnosis was basically just rolled into the greater "Autism spectrum" diagnosis.

That's not my understanding of what happened in DSM-5. My understanding is that Asperger's was removed as a separate diagnoses, as it now falls under Autistic Spectrum Disorder. The Wikipedia article seems to match my prior understanding (https://en.wikipedia.org/wiki/Autism_spectrum), "The DSM-5 redefined the autism spectrum to encompass the previous (DSM-IV-TR) diagnoses of autism, Asperger syndrome, pervasive developmental disorder not otherwise specified (PDD-NOS), and childhood disintegrative disorder."

>The strength of each of the editions of DSM has been “reliability” – each edition has ensured that clinicians use the same terms in the same ways. The weakness is its lack of validity.

http://www.nimh.nih.gov/about/director/2013/transforming-dia...

Time for a new book.

I like Steven Novella's nuanced take on this issue: https://www.sciencebasedmedicine.org/dsm-5-and-the-fight-for...

It was not removed - it was pushed into the "Umbrella" of "Autistic spectrum disorder" - why? Because there was no agreement as to what "High functioning autism" vs "Aspergers" was so different people diagnosing would put different people under different things.

Oh, so it was the other way around, basically? Well, thanks for correcting me.

Because that 1% demonstrates a staggeringly wide range of intelligence, and there are enough to study without too many confounding factors.

Its hard to discuss since the article goes into considerable detail in opposition to your position, and you provide no detail beyond something paraphrased to it would be too difficult to do what was actually done successfully in the article.

I read the article; or at least a deep skim. Its examples, data, and arguments are pretty convincing. Lots of interesting data and citations provided in support of the theory.

If its a wrong theory, at least its well documented and interesting to think about and leads down interesting paths. Insert fairly obvious analogy to string theory.

Reply on news.ycombinator.com