Read The Autistic Brain: Thinking Across the Spectrum Online

Authors: Temple Grandin,Richard Panek

Tags: #Non-Fiction

The Autistic Brain: Thinking Across the Spectrum (7 page)

BOOK: The Autistic Brain: Thinking Across the Spectrum
9.27Mb size Format: txt, pdf, ePub
ads

“In the case of brain trauma,” Schneider says, “we’re looking at a break in one of these cables.” Not so in autism. There, he said, “we’re looking at an anomalous growth pattern, be it genetic, be it developmental, et cetera, within that process.”

I was invited to Schneider’s lab to be scanned as part of a television program. Afterward, Schneider explained to me that he had been looking for areas in my brain that showed at least a 50 percent difference from the corresponding areas in a control subject. Two findings, he said, “really jumped out.”

One, my visual tract is huge—400 percent of a control subject’s.

Two, the “say what you see” connection in the auditory system is puny—1 percent of a control subject’s. This finding made sense. In my book
Emergence,
I discussed my childhood speech problem: “It was similar to stuttering. The words just wouldn’t come out.”

I later asked Schneider to interpret these findings for me. Because we’re still figuring out the brain, his interpretation would need to be in the nature of a hypothesis. But that’s how science works. You gather information (my brain scans), use it to formulate a hypothesis, and make a prediction you can verify.

Between birth and the age of one, Schneider explained, infants engage in two activities that developmental researchers call verbal babbling and motor babbling. Verbal babbling refers to the familiar act of babies making noises to hear what they sound like. Similarly, motor babbling refers to actions such as waving a hand just to watch it move. During this period when babies are figuring out how to engage with the world, their brains are actually building connections to make that engagement possible. During verbal babbling, fibers are growing to make the connection between the “what you’re hearing” and “what you’re saying” parts of the brain. During motor babbling, fibers are growing to make the connection between the “what you’re seeing” and “what you’re doing” parts of the brain.

Then between the ages of one and two, children reach a stage where they can say single words. What’s happening in the child’s brain at this point is that fibers are forming an interlink between those two fiber systems that were constructed during the verbal and motor babbling period. The brain is connecting “what you’re seeing” with “what you’re saying” until out pops
Mama, Dada, ball,
and so on.

In my case, Schneider hypothesized, something happened developmentally during the single-word phase so that the fibers didn’t form a connection between “what you’re seeing” and “what you’re saying.” This would be the tract that was 1 percent of the size of the control subject’s. To compensate, my brain sprouted new fibers, and they tried to go somewhere, anywhere. Where they wound up primarily was in the visual area rather than traditional language-production areas. That’s the tract that was 400 percent of the size of the control subject’s.

In such a scenario, Schneider continued, the babbling phase might be normal but language development would slow down dramatically between ages one and two.

Which would match a developmental pattern that the parents of children diagnosed with autism often report.

“Exactly,” Schneider said.

But, he emphasized, the scenario he described was still only a hypothesis. He’ll need more data, more scans that actually reflect how brains grow. “We’ve never had the technology to measure that,” he said. “The project I’m working on is to map that developmental sequence.”

He hadn’t planned to adapt the HDFT technology to map the development of the autistic brain, but a question from
60 Minutes
correspondent Lesley Stahl changed his mind. Schneider asked me for permission to show my scans to her for a segment on autism her show was preparing. (The original television program that had commissioned the scan never aired.) In order not to raise unrealistic hopes for desperate parents, Schneider wanted to mention that HDFT scanning to diagnose the autistic brain wasn’t going to be available at a local hospital in the near future—that it would be at least five to ten years before even leading hospitals had access to this technology. Stahl let him. But here’s how Schneider remembers her phrasing the question:

“So a mother with a four-year-old child who will be age fourteen before she gets a biological diagnosis of her child’s brain damage—that delay would mean a decade or more of failed treatment attempts, lost ability to communicate and educate her child, and the emotional strain that accompanies an uncertain diagnosis. What might be done to speed that process up and to make it available in five years?”

“This,” Schneider said, “is why I’m doing a project on autism.”

Science often advances because of new developments in technology. Think of Galileo and the telescope. He was one of the first people to point a “tube of long seeing” at the night sky, and what he found there forever changed how we conceive of the universe: mountains on the moon, moons around Jupiter, phases of Venus, and far, far more stars than met the naked eye. The same is true of neuroimaging. You can think of it as a “mindoscope” (to borrow a coinage from Hirsch), an instrument with which we have just begun to explore the universe within and to gather preliminary answers to our questions about the autistic brain: How does it look different than a normal brain? and What does it do differently than a normal brain?

We now understand the biological connections between parts of the brain and many of the behaviors that make up the current diagnosis of autism. But we don’t yet know the cause behind the biology—the answer to a third question: How did it get that way?

For that answer, we have to turn to genetics.

 

Neuroimaging isn’t perfect. In order to understand and appreciate what it can do best, let’s look at what it can and cannot do.

 

  • An fMRI can’t capture the brain’s activity during the full range of human experience. By necessity, it can observe only the brain responses that a person can have while lying still for long periods.
  • Neuroimaging also requires subjects to keep their heads still. In recent years, several studies reported that short-range connections in the brain weaken as children grow older, while long-range connections strengthen. Neuroscientists considered this news to be quite a significant advance in the understanding of the brain’s maturation process. Unfortunately, a follow-up study by the authors of the original studies showed that the supposed changes in the brain’s development disappeared once they took head movement into account. “It really, really, really sucks,” the lead investigator said.
    “My favorite result of the last five years is an artifact.”
           This finding didn’t cause scientists to rethink every brain scan out there. But it did serve as an unambiguous warning about the need to take head movement into account. This caution applies especially to studies of people with autism and other neurodevelopmental disorders. Why? Because those subjects are precisely the ones who will have the most difficulty holding still. Researchers are racing to figure out a way to factor out head motion in neuroimaging studies, but even if they’re successful, they will have to ask themselves whether the removal of data from studies of one group of subjects (like autistics) will skew comparisons with studies of neurotypical subjects.
           Even if you do manage to hold still, you can still screw up a neuroimaging result—as I know from personal experience. During one fMRI study, I was shown a flight simulation. First I was swooping over the Grand Canyon. Then I was skimming over wheat fields. Then I was skipping over mountaintops. Then I was feeling sick—which didn’t seem like a good idea when you’re inside a scanner. So I closed my eyes. Whatever else that scan was, it sure wasn’t perfect.
  • Even the best neuroimaging is only as good as current technology. Neurons fire hundreds of impulses per second, but the signal itself takes several seconds to blossom, and then it lingers for tens of seconds. Temporally precise, it’s not. And the resolution doesn’t really capture activity at the level of the neuron itself. As an article
    in Science magazine said, “Using fMRI to spy on neurons is something like using Cold War–era satellites to spy on people: Only large-scale activity is visible.”
  • And there are the researchers themselves. They have to be careful how they interpret the results. For instance, they shouldn’t assume that if a portion of the brain lights up, it’s essential for the mental process being tested. In one study, researchers found that the hippocampus was activated when subjects were performing a particular exercise, but researchers conducting another study found that lesions to the hippocampus didn’t affect the subjects’ ability to perform that same exercise. The hippocampus was indeed part of the brain’s response, but it wasn’t a necessary part of the response.
  • Researchers also can’t assume that if a patient is exhibiting abnormal behavior and the scientists find a lesion, they’ve found the source of the behavior. I remember sitting in a neurology lecture in graduate school and suspecting that linking a specific behavior with a specific lesion in the brain was wrong. I imagined myself opening the back of an old-fashioned television and starting to cut wires. If the picture went out, could I safely say I had found the “picture center”? No, because there were a lot of wires back there that I could cut that would make the TV screen go blank. I could cut the connection to the antenna, and the picture would disappear. Or I could cut the power supply, and the picture would disappear. Or I could simply pull the plug out of the wall! But would any of those parts of the television actually be the picture center? No, because the picture depends not on one specific cause but on a collection of causes, all interdependent. And this is precisely the conclusion that researchers in recent years have begun to reach about the brain—that a lot of functions depend on not just one specific source but large-scale networks.

 

So, if you ever hear that fMRI can tell us people’s political preferences, or how they respond to advertising, or whether they’re lying, don’t believe it. Science is nowhere near that level of sophistication yet—and may never be.

 

3

Sequencing the Autistic Brain

O
N SEPTEMBER
6, 2012, I was doing what I usually do when I need to kill time in an airport—lingering at a newsstand, flipping through magazines, browsing the front pages of newspapers—when a page 1 headline in the
New York Times
caught my eye: “Study Discovers Road Map of DNA.” I grabbed the paper and read on
: “The human genome is packed with at least four million gene switches that reside in bits of DNA that once were dismissed as ‘junk’ but that turn out to play critical roles in controlling how cells, organs and other tissues behave.”

Well, it’s about time,
I thought. The idea of junk DNA had never made sense to me. I remember in graduate school hearing about junk DNA. I heard references to it in the classroom. I saw peer-reviewed research articles about it in
Science
and
Nature.
Junk DNA
is not a nickname, even though it may sound like one; it is an actual scientific term. It’s called junk DNA because, unlike the sequences of DNA that code for proteins, these sequences didn’t seem to have any purpose.

That idea was ridiculous to me. The double helix had always reminded me of a computer program, and you would never write code that had a lot of unnecessary stuff. The “junk”
had to
serve a purpose. It had to be something like the gene’s operating system. If you go into your computer and find a lot of weird files, you might wonder what they’re for, but you wouldn’t conclude that they served no purpose. And you sure wouldn’t want to reverse a couple of zeros and ones just to see what happened. Same thing with junk DNA. If you messed around with it, the gene’s “computer program” would not work.

I was hardly alone in harboring this deep suspicion. For years, scientists had been taking the idea of junk DNA less and less seriously. In fact, geneticists had started preferring the terms
noncoding DNA
and
dark matter,
both of which suggested that this kind of DNA was simply a mystery, not garbage. As I stood reading the article in the airport, I felt vindicated after so many years, which is always nice, but that’s not what jumped out at me.

The article—amid many others that day and in the weeks to come that emphasized the non–junk DNA angle—was based on the results of a massive federal research effort called the Encyclopedia of DNA Elements, or Encode. The project involved 440 scientists from thirty-two laboratories around the world, and the group’s first thirty papers had appeared a day earlier in
Nature,
Genome Research,
and
Genome Biology.
In one common analogy, the earlier sequencing of the human genome by the Human Genome Project and by Craig Venter’s Celera Genomics in 2001 “was like getting a picture of Earth from space,” as one scientist told the
Times
, while Encode was like Google Maps: It told us “where the roads are,” “what traffic is like at what time of the day,” “where the good restaurants are, or the hospitals or the cities or the rivers.” The Human Genome Project told us what the genome was. Encode has begun to tell us what it does.

But what really interested me was the article’s explanation of how the genome does what it does. In order to appreciate its significance, you first have to understand what DNA looks like. We’ve all seen the popular image of the double helix: that corkscrew of seemingly endless combinations of A (adenine), C (cytosine), G (guanine), and T (thymine) bases. But that Tinker Toy model represents a strand of DNA that’s stretched out. A strand of DNA completely unfurled would be about ten feet long. But it’s not unfurled. Instead, DNA is so tightly coiled that it fits inside the microscopic cell nucleus. By looking at DNA in its natural state, Encode researchers found, as the
Times
reported, “that small segments of dark-matter DNA are often quite close to genes they control.”

Now that,
I thought,
is a mindblower.

Until then, scientists had been thinking about DNA in its stretched-out form. But if you envision DNA as a tightly wound coil—and while I was standing in the airport, holding the
Times
in my hands, that’s exactly what my picture brain was doing—then a noncoding piece of DNA could be flipping switches on coding DNA that’s hundreds of thousands of base pairs away. In the stretched-out helix, they’re nowhere near each other; in the coiled-up helix, they’re adjacent to each other.

BOOK: The Autistic Brain: Thinking Across the Spectrum
9.27Mb size Format: txt, pdf, ePub
ads

Other books

Breaking Shaun by Abel, E.M.
Words With Fiends by Ali Brandon
Stowaway by Becky Black
The Seal Wife by Kathryn Harrison
Brothers In Law by Henry Cecil
Carnal Harvest by Robin L. Rotham
Nothing but the Truth by Jarkko Sipila
A Tradition of Victory by Alexander Kent
The Drawing of the Three by Stephen King