Cover Story

Over the past decade, neuroscientists have begun championing a theory of dyslexia that hinges on the idea that people with dyslexia are slower than normal to process certain types of information. This "processing-speed deficit," they posit, likely results from having less efficient connections between various portions of the brain, some of which are critical to reading. And recent findings from brain imaging studies have begun to support this theory, at least for some people with dyslexia.

Some of the earliest findings focused on the magnocellular vision pathway--a bundle of nerve cells that connect the retina to the thalamus and then the visual cortex, and which are known to facilitate fast visual processing. Several years ago, Harvard University neurologist Albert Galaburda, MD, autopsied the brains of people with dyslexia and found anatomical differences in their magnocellular pathways compared with people without reading problems.

To test whether such anatomical differences translated into functional differences, Stanford University's David Heeger, PhD, and his colleagues used functional magnetic resonance imaging (fMRI) to examine brain activity as people performed tasks that required fast visual processing--detecting differences in objects' speed of motion. The researchers found significantly less activity in the magnocellular pathways of people with dyslexia than in those of normal readers. In fact, people with lower activity in this brain region were slower readers, Heeger's team found.

Among the newest brain imaging studies (published in the February issue of the journal Neuron) is one by Stanford's John Gabrieli, PhD, Torkel Klingberg, PhD, and their colleagues. They find differences between poor readers and good readers in the brain's white matter. They used a relatively new imaging technique called "diffusion tensor imaging" that allows researchers to say something about the brain's organization. They found that the white matter underlying the junction of the brain's temporal and parietal lobes is better organized in adults without reading problems than in adults with reading problems. This is a brain area where neuroscientists believe visual and auditory information is integrated and sent to the front of the brain for further processing.

And the researchers also found that the brain images correlated well with scores on standard reading tests: the lower the reading score, the less well organized the white matter at the temporoparietal junction. The finding is particularly interesting because white matter is composed of myelin--an insulating substance that makes the movement of brain signals more efficient. The more myelin, the faster signals move. The less myelin, the more slowly and less efficiently signals move. So differences in white matter may symbolize differences in processing speed, says Gabrieli. He and his colleagues are now conducting the same study in children with reading impairments.

Yet another potential link to a processing-speed deficit in people with dyslexia arises from work by psychologists Dennis and Victoria Molfese, both PhDs, of the University of Louisville, who find 80 percent of infants who later develop dyslexia show a delayed brain response to speech sounds. In infants who do not develop dyslexia, they see two humps in the electrical signal coming from the brain--one 200 milliseconds after the sound begins and another 500 milliseconds after it begins. In babies who develop dyslexia, the first hump disappears and the second hump is exaggerated.

"It's almost as if the brain isn't orienting or responding right away," says Dennis Molfese.

--B. AZAR