By Curt Suplee
Washington Post Staff Writer
Tuesday, March 3, 1998; Page A01

For the first time, scientists have been able to identify specific brain malfunctions involved in dyslexia -- a discovery that could substantially improve understanding of the chronic reading problem that afflicts some 10 million Americans.

Equally important, said lead researcher Sally E. Shaywitz of Yale University School of Medicine, the work provides scientific confirmation "for what has previously been a hidden disability."

"If you break your arm, you can hold up an X-ray and see it," she said. "But I get calls from distraught parents and even teachers who say, 'My school system denies that there's such a thing as dyslexia.' Well, now they can say, 'Here's the evidence.' "

Most members of the general public regard dyslexia as a condition that has something to do with "reading letters backwards." But recently a broad expert consensus has determined that it has surprisingly little to do with recognizing the visual form of words. Rather, dyslexia, which affects 80 percent of all those labeled "learning disabled," entails an inability to break the letters of written words into the 44 distinct sounds (called phonemes) in American English -- a capability called "phonologic awareness."

Using high-tech imaging equipment, Shaywitz's team found that the brains of dyslexic subjects show very little activity in areas known to be important in linking the written form of words with phonic components.

The new work will be "very useful" in designing better therapies for dyslexic readers, said Paula A. Tallal, co-director of the Center for Molecular and Behavioral Neuroscience at Rutgers University. "To do that, you really have to understand the neural mechanism," said Tallal, who was not involved in the research.

Shaywitz and colleagues used a brain-mapping procedure called functional magnetic resonance imaging (fMRI) to monitor mental activity in 61 subjects. Like the MRI scans familiar to many patients, fMRI provides an image of internal body structures in very fine resolution. But it also shows which parts of the brain are most active at a given time.

The subjects were given five kinds of exercises. Each required progressively more effort in processing the sound aspects of written language, as distinct from merely recognizing shapes, letters or words. One task, for example, asked whether several nonsense words (such as "leat" and "jete") would rhyme -- a difficult problem for dyslexics.

By comparing patterns of activity in dyslexic and normal subjects, the team detected what Shaywitz suspects is "the neural signature of dyslexia" embodied in two very different kinds of responses.

The 32 readers who were not impaired made intensive use of various rear-brain areas including the angular gyrus and Wernicke's area. The 29 dyslexic readers, by contrast, showed very little activity in that critical region. Instead, they seem to have compensated by overusing a front-brain section, called Broca's area, traditionally associated with other aspects of language processing and speech. Why this occurs in dyslexics remains unknown.

The research, reported in today's Proceedings of the National Academy of Sciences, was funded by the National Institute of Child Health and Human Development.

Knowing the telltale brain pattern of dyslexic readers, Shaywitz said, means "we could diagnose people early." At present, most children are not identified as dyslexic until they are in the third grade. Shaywitz and others would like to see diagnosis occur years earlier, long before a student starts to fail. Although dyslexia persists throughout life, skills improve consistently with practice.

Neuroscientist Albert M. Galaburda of Harvard Medical School said that he found the Shaywitz study interesting, though he regards it as far from a paradigm shift. "It illustrates, but does not resolve" a major source of debate in the field, he said: whether the phonological difficulty "is a primary problem, or whether it is secondary to some other perceptual processing problem."

Tallal's research indicates that there may indeed be other processes at work. "What we've found, and which is consistent with [the Shaywitz] paper, is that in addition to phonological problems, the [dyslexic] brain shows a limited capacity to organize brief and rapidly changing input," she said.

Because it appears that dyslexics are "characterized by how fast they can organize information, especially acoustic information," Tallal and others have created a software program intended to aid children with reading disabilities by slowing down or stretching out the sounds, thus giving the child's brain more time to process each unit of information.

So far, such interventions are not widespread. Nor, said Shaywitz, is much provision made for dyslexic persons who are highly intelligent and motivated, but require extra reading time and other assistance.

"We as scientists can give people the information," she said, "but then society has to grab it and do something with it."

Anatomy of the Problem

New research shows that during some verbal tasks, people with learning disabilities have a different pattern of brain activity than other subjects. The darker the shading below, the greater the activity; white areas were not studied.

Normal:

Minimal front-brain activity.

High activity in visual cortex and rear-brain areas such as the angular gyrus and Wernicke's areas, which link visual and phonic aspects of words.

Dyslexic:

High activity in front-brain regions such as Broca's area, which is involved with vocalizing speech.

Little activity in rear-brain areas.

SOURCE: Proceedings of the National Academy of Sciences