July 31, 2004

Autism as lack of neurological coordination

In the August issue of Brain, there's an article by Marcel Just, Vladimir Cherkassky, Timothy Keller and Nancy Minshew, presenting evidence from functional MRI brain imaging for a new hypothesis about autism. They suggest that autism is not mindblindness due to a faulty theory-of-mind module, nor is it runaway maleness overwhelming empathy with analysis. Instead, it's underconnectivity: "a deficiency in the coordination among brain areas".

According to the CMU press release:

In explaining the theory, Marcel Just, one of the study's lead authors and director of Carnegie Mellon's Center for Cognitive Brain Imaging, compared the brain of a normal person to a sports team in which the members cooperate and coordinate their efforts. In an autistic person, though some "players" may be highly skilled, they do not work effectively as a team, thus impairing an autistic's ability to complete broad intellectual tasks. Because this type of coordination is critical to complex thinking and social interaction, a wide range of behaviors are affected in autism.

Here's the full abstract:

The brain activation of a group of high-functioning autistic participants was measured using functional MRI during sentence comprehension and the results compared with those of a Verbal IQ-matched control group. The groups differed in the distribution of activation in two of the key language areas. The autism group produced reliably more activation than the control group in Wernicke's (left laterosuperior temporal) area and reliably less activation than the control group in Broca's (left inferior frontal gyrus) area. Furthermore, the functional connectivity, i.e. the degree of synchronization or correlation of the time series of the activation, between the various participating cortical areas was consistently lower for the autistic than the control participants. These findings suggest that the neural basis of disordered language in autism entails a lower degree of information integration and synchronization across the large-scale cortical network for language processing. The article presents a theoretical account of the findings, related to neurobiological foundations of underconnectivity in autism.

Although the findings deal only with language, and in fact only with specific aspects of sentence comprehension, the paper's discussion extends the hypothesis to much broader ideas about autism as characterized by, or even caused by, lack of adequate integration among different brain areas.

This is a plausible and interesting idea, for which the authors cite a range of other evidence. But the contribution of this particular experiment should be interpreted a bit more cautiously, it seems to me. The task studied is a very specific and limited one: visually presented text with binary choices between interpretations:

The cook thanked the father.
Who was thanked? cook -- father

The editor was saved by the secretary.
Who was saving? editor -- secretary.

It's certainly interesting that there's a significant difference between the autistic and the control groups in the distribution of brain activity in performing this simple task. The autistic group showed more activity in Wernicke's area, and less activity in Broca's area. Putting it more generally, the autistics showed more activity in posterior language-related parts of the brain, and less activity in anterior language-related parts of the brain. This suggests greater focus on the words and their meanings, and less focus on how the words were put together. It seems to me to raise a host of interesting questions -- were the autistics just trying harder to figure out who did what to whom on the basis of meaning rather than form, as the authors suggest? or were they composing elaborate taxonomic theories about the entity classes involved? "Let's see, cook::father, secretary::editor, teacher::??" or might they have more efficient Broca's areas, which therefore were working at a lower duty cycle, needed less blood flow, and therefore looked less active to fMRI?

But I'm more concerned about the argument for lack of coordination, which depends on finding lower correlations among activity levels in different brain "regions of interest" (RoIs). fMRI measurements of activity levels are very noisy. A lower correlation in activity levels between two regions might reflect the fact that they are genuinely less coordinated, but it might also reflect the fact that the measurement of one of them has a lower signal-to-noise level. Which it would, given that its task-related activity level is lower.

So in evaluating this particular argument, I'd like to see the full dataset. There are some convincing-looking pictures

Fig. 2 Examples of functional connectivity between LDLPFC and LIFG (Broca's area) in individual participants, shown as the activation time series in the two brain regions, with vertical bars indicating boundaries between seven epochs of sentences of the same type. (A) Autistic participant with low functional connectivity, r = 0.31, where the two time series do not closely track each other. (B) Control participant with high functional connectivity, r = 0.79, where the activation time series in the two regions is highly similar.

and some significant (though not overwhelming) overall statistics ("When the functional connectivities of the two groups were compared in each ROI pair separately, every single one of the 10 reliable (P < 0.05) differences (out of 186 comparisons) showed a lower functional connectivity in the autistic group ... Although about nine differences might be expected to be reliable by chance, the uniform direction of difference is not expected by chance.") But I'd like to be able to play with the original data, to convince myself that they're not seeing the results of SNR differences rather than coordination differences.

Let me emphasize that I find the coordination hypothesis interesting and attractive. On one level, it seems directly opposed to the mindblindness or "theory of mind module" hypothesis, due originally to Frith et al.:

"...in a normally developing child, the computational capacity to represent mental states has an innate neurological basis. In the autistic child damage to the circumscribed system of the brain has occurred, and this prevents the normal operation of the critical cognitive mechanism"

If you subtract the idea that "theory of mind" is a highly localized function, the two ideas are less opposed -- "theory of mind" is a late-developing ability that plausibly depends on difficult coordination of several different brain regions, and (to the extent that a single brain region is singled out) shows heavy involvement of anterior regions near those that were less active in the autistic group in this experiment (which of course required no "theory of mind" reasoning at all).

The coordination idea seems less easy to square with Simon Baron Cohen's notion of autism as " extreme maleness", runaway male analytic thought (with concomitant deficits in female-associated empathizing, natch). Though you could spin out a theory about female brains being more integrated, etc. -- and of course some people have done that...

So maybe Marcel Just and Simon Baron Cohen are working from different ends of the autistic elephant. Or maybe one of them is struggling in the dark with a palm tree that happened to be nearby. Stay tuned.

[The Just et al. paper was brought to my attention by Fernando Pereira]


Posted by Mark Liberman at July 31, 2004 08:07 AM