Thursday, May 15, 2008

Autism subtypes and analytical skills

Francesca Happé (of Kings College, London) opened the 2008 IMFAR (International Conference for Autism Research) conference with a keynote address that argues for something I’ve suspected for years: that autism has many distinct subtypes, and that the autistic spectrum is multi-dimensional.

Happé’s proposed dimensions:

1. The Theory of Mind deficit, or subnormal awareness of other people’s minds.
2. The Executive Function deficit, or impairments in planning and cognitive flexibility.
3. Weak Central Coherence, or a focus on details.

Happé’s scheme recasts my son from “moderately autistic” plain and simple to moderately autistic along dimensions 1 and 3. He’s weak in empathy and strong in details, but unimpaired in the kind of planning and cognitive flexibility considered deficient in many with autism: he aces puzzles like Towers of Hanoi.

But if Happé’s three dimensions capture autism, where do the unusually strong analytic (or “left-brain”) skills of my son--and others of his subtype--come from?

A later talk by A. Harrison (et al, University College, London) suggests a partial answer. Harrison discussed the Framing Effect, or the influence of context on decisions.

Here’s my best recollection of his example:

Frame 1:
An epidemic strikes 600 people. Which containment strategy do you prefer?
1. One in which exactly 200 people will survive.
2. One in which there’s a 2 in 3 chance that everyone will die, and a 1 in 3 chance that everyone will survive.

Frame 2:
An epidemic strikes 600 people. Which containment strategy do you prefer?
3. One in which exactly 400 people will die.
4. One in which there’s a 2 in 3 chance that everyone will die, and a 1 in 3 chance that everyone will survive.

Even though Frames 1 and 2 describe identical choices, non-autistics are about 4 times more likely to pick the first choice in the positive frame (1), and the second choice in the negative one (2). Autistics are far less influenced by framing context.

Harrison proposes that their Weak Central Coherence, or detail focus, explains this hyper-analytic or rational decision process, which contrasts with the more “emotional” heuristics used by non-autistics.

Whether detail focus fully explains my son's hyper-analytic skills--his facility, e.g., with systems of switches, gears, linear equations, and computational subroutines—begs further research.


concerned heart said...

Some autism is familial with certain genes affected and some autism is non-familial and caused de novo by new mutations in older fathers (33+). Some autism is due it seems to mitochondrial disorders. In all cases different genes are mutated and different personalities and deficits result.

concerned heart said...

I left out the fact that it is the sperm's DNA that has the mutations due to copy errors that increase with age and toxic exposure.

lefty said...

Thanks for the link.

There's been some talk about genetic factors at the conference, with a couple of dozen proposed mutations, not yet explored.

The only specific mapping between genes and autistic profiles that has come up so far is a case where having 3 copies of a particular gene results in one developmental outcome, and having 0 copies at that gene results in another. The details escape me.

Liz Ditz said...

Thanks for these posts, Lefty.

Some kids with autism have significant expressive language delays (and perhaps receptive as well).

Where does Happé put the language-delay feature?

I am also very curious about the intersection of autism and culture, as RR Grinker explored in Unstrange Minds.

lefty said...

Hi, Liz,

I finally noticed your message!

Happé and others have proposed that the language deficits stem from the social disconnect. Kids on the spectrum aren't tuning in and soaking up language as others do, and so aren't getting complete immersion. They're also not attending as much to gaze and other indicators of what people are referring to, and thus often make the wrong inferences about the meanings of new words.

What makes this direction of causality particularly compelling is the strong correlation between language skills and Theory of Mind skills, with current language skills more strongly predicting later TOM skills than current TOM skills predict later language skills.