The Intense World Theory of Autism is an appealing one to me, because it describes my experience, and resonates with much that I have heard described by other autistic people.
It was pleasing, therefore, to learn of new research that seems to offer some empirical support to this idea, and perhaps a partial explanation about what is different about the autistic brain that could explain the intense experience.
Still, I find myself being more than a little skeptical about the claims and explanations being made. Some of the statements made about this research appear to me to be self-contradictory, although that may be coming from the reporting of the studies rather than in the research itself.
And, like too much research these days, it leaves me with the impression of a sales pitch for the next grant, with sweeping claims being made based on small samples.
All that said, I want to believe the claims and explanations given, so it’s worth commenting on them, even though it is clear to me that we have only just begun a long journey on the way to understanding what it going on inside the autistic brain.
In a post on the Medical Express website about a year ago, headlined as “Detecting autism from brain activity,” it was announced that “Neuroscientists from Case Western Reserve University School of Medicine and the University of Toronto have developed an efficient and reliable method of analyzing brain activity to detect autism in children.” Red Flag #1. Why children? Or, why just children? I don’t know if there has ever been a literature study to determine what percentage of autism research has been devoted to children. My guess is that it is a very high number. Yet, autistic children grow up to be autistic adults, and what is true of children may or may not be true of adults. Some work I’ve seen by the Nancy Kanwisher Lab suggests to me that many earlier studies of children may have missed the boat in (at least) one important aspect.
In many, if not most, of the studies of children I’ve seen, the autistic kids being studied are compared with a “control group” of neurotypical (NT) children, which is carefully matched for such things as IQ, gender, and age. Problem: one of Kanwisher’s studies that measured a certain skill discovered, as expected, that autistic children did not perform as well as their NT counterparts. Yet, when the data were plotted on an age graph, it was clear that both groups improved over time. Even more striking, the scores and rate of improvement were almost identical if one shifted the age scale by 12 to 18 months for the autistic group. In other words, the autistic kids were every bit as talented, but it took them a year or two longer than their NT counterparts to develop the skills.
By the time they are adults, my guess is that there is no discernible difference. This reminds me of the debate that was raging a few short years ago as to whether there was a meaningful difference between what was called “high-functioning autism” and Asperger Syndrome. I asked a psychiatrist friend of mine if he had an opinion, and he responded that technically, the difference was the age of language acquisition, but by the time people reached adolescence there was no detectable difference, so it really wasn’t a very meaningful distinction in pragmatic terms.
Back to the brain activity test: Red Flag #2. The study claimed to be able to detect autism with “with 94 percent accuracy.” What does this claim really mean? I suspect it is tied to Red Flag #3, which is that they studied only 19 children. If they were “wrong” about only one of these children, that would mean they correctly identified about 95% of the group. Of course, that assumes the kids were labeled correctly in the first place, which is a potential source of error that was not mentioned in the short write-up.
I’m not a neuroscientist, nor am I a statistician by trade, although I did have a long and storied career as a financial economist, and statistical analysis was an essential tool in my bag of tricks. In order to make meaningful statistical inferences, one needs a confidence interval and some measure of potential error (called a confidence level). A single number is meaningless. In fact, “It is crucial to know the confidence level associated with a confidence interval: The interval by itself is meaningless” according to a text on statistics from Berkeley. Ceteris paribus, the smaller the sample size, the higher the error term. A sample of 19 is pretty tiny.
Sorry for the statistical digression, but I wanted to be clear why I am waving the warning flag of skepticism. It may very well be that subsequent studies will confirm these results, but I would want to see larger samples and also independent verification from other laboratories.
More to the point of relevance to the Intense World Theory, the write-up contains this statement:
Their approach also allows them to measure background noise, or the spontaneous input driving the brain’s activity while at rest. A spatial map of these inputs demonstrated there was more complexity and structure in the control group than the [autism] group, which had less variety and intricacy.
Let’s contrast that to what the same group of researchers said in a more recent study:
…this study is a follow-up to the authors’ prior finding that brain connections are different in autistic children. [The earlier] paper determined that the differences account for the increased complexity within their brains.
Now I’m confused. It may be, as I said, that these summary write-ups mischaracterized what the original papers said, so I will have to obtain some clarification on that. These statements seem to be directly contradictory.
According to the write-up of the second study,
They showed that autistic children’s brains at rest generate more information than non-autistic children. This may explain their lack of interest in external stimuli, including interactions with other people.
This may be a bit of speculative over-interpretation. I’m not sure that their observation (assuming it can be verified with larger samples, and, importantly, with adults) actually “explains” anything, especially a “lack of interest.” This interpretation strikes me as an all-too-common lack of theory of mind by the researchers. They are attributing their own state of mind to an autistic child because of externally observed behaviors. What is going on inside the autistic mind may not be predictable by such a simple analogy.
This certainly is an interesting line of research that deserves to be pursued. It seems to me, though, that it is way too early to be making some of the dramatic claims that I read in these descriptions.
I also wonder about possible comparisons with similar research being done at the University of Pittsburgh (and perhaps elsewhere) using somewhat different technology. My friends Temple Grandin and John Elder Robison have participated in that study, and John urged me to sign up (which I did, and am on their waiting list). Temple was in Boston a few months ago and showed some slides of her brain as part of her talk. The pictures suggested to me that it may be more accurate to describe the autistic brain not as “more” or “less” active or connected, but as “differently” wired.
Brain imaging and other techniques are producing information perhaps faster than we can interpret it, but I suppose that’s a good problem to have. I look forward to learning more in the coming months.