UCSF-Led Team Offers New Insight Into Neurological Basis Of Dyslexia

The findings, documented through simultaneous brain imagingand behavioral tests, strongly indicate, the researcherssaid, that adult dyslexics have an enduring neurologicaldeficit in their ability to process these brief, rapidlysuccessive sounds.

They suspect that the deficiency contributes to difficultiesin early speech and language learning, and leads to aweakness in the subsequent mental leap in abstraction towords on a page that enables people to learn to read.

The study was published in the May 24 issue of Proceedingsof the National Academy of Science.

Perhaps the most provocative aspect of the finding, theresearchers said, is the clear and direct neurologicalevidence that reading deficits are generated, at least inpart, by a deficit at a very fundamental level of corticalprocessing of sound inputs.

"Our findings indicate that there is a basic problem insignal reception, as complex sound information streams intothe cerebral cortical system underlying aural speechrepresentation," said the senior author of the study,Michael Merzenich, PhD, the Francis A. Sooy Professor ofOtolaryngology and a member of the Keck Center forIntegrative Neuroscience at UC San Francisco. "The way thatthe brain processes sound in poor readers is very differentfrom its processing and representation of rapidly changingsound inputs in competent readers."

"Our research indicates that adult dyslexics arerepresenting the sound parts of words by the activation ofcortical neuron populations in a weaker and less salientform within their cortical aural speech processing system.We believe that they, therefore, are not delivering thenormal forms of representation of the separable sound partsof words to the regions of the brain involved in speechperception and reading," he said.

The authors emphasize that their findings do not discountthe additional involvement of higher levels of brainprocessing in dyslexia, where more complex combinations ofinformation lead to the recognition and interpretation ofspeech.

At the same time, they argue that the very elementary defectin the brain's processing of sound must be playing animportant role in the generation of relatively weak neuronalrepresentations of the sound parts of aural speech.

And this elementary neurological deficit, they said, couldprovide a target for remedial therapies aimed at trainingthe brain to increase the speed and accuracy with which itprocesses rapidly successive and rapidly changing sounds.

The sound-processing function occurs at a base, or entry,level of sound processing in the brain, and is believed tobe a primary step in the brain's representation of normalspeech sounds and its creation of speech andlanguage-reception abilities. The process ultimatelyculminates with a listener learning to recognize the soundparts of words, and to translate these word sounds aswritten letters.

Previous behavioral studies have suggested that theinability to parse the rapidly successive, changing soundsthat make up words, the phonemes of language, may be theprimary basis of language-learning impairments in children. Scientists have long argued that children who havedifficulty parsing word sounds are destined to havedifficulty successfully initiating reading.

Other behavioral studies have indicated that most peoplewith dyslexia, characterized by a difficulty with reading,also have impairments in the fidelity of their auditoryreception. However, because most dyslexics ultimatelydevelop facile speech reception and production capabilities,the significance of this problem for the origin of readingimpairments has been unclear.

The researchers conducted their current study in sevendyslexic adults who were of normal intelligence but severelychallenged by reading, spelling and writing. Results werecompared with those recorded in seven adults of normalintelligence who were competent readers.

The dyslexic adults performed poorly on standardized readingtests. And, as has been shown to be the case with the greatmajority of adult dyslexics, these poor readers (ages 18-42)also performed poorly on a variety of tests that measuredtheir ability to discern rapidly successive sound stimuli.

In one of these sound-discerning tests, adults were exposedto two sounds that differed in frequency and that occurred atenth or a fifth of a second apart. They were then asked toidentify the sounds and to replay the sequence in which theywere presented. Their brain activity was simultaneouslyrecorded using magnetoencephalographic brain imaging, whichmeasures magnetic field fluctuations generated by spatiallylocalizable human brain activity with millisecond precision.

In these studies, the UCSF team focused on the activitygenerated by the rapidly successive sounds evoked from the"primary" auditory cortical areas, where information aboutaural speech flows into the cerebral cortex's processingsystem for language.

Poor readers did report hearing the two very brief sounds,and often knew that in some way they weren't the same, butthey were unable to identify them, or to reliablyreconstruct the sequence in which they were represented.

"The reason," said Srikantan Nagarajan, PhD, an assistantadjunct professor of otolaryngology and a member of the KeckCenter for Integrative Neuroscience at UCSF, and the leadscientist of the study, "was demonstrated by the abnormalway that the brain of the poor-reading subjects responded tothese rapidly successive sound events."

"In normal readers, the auditory cortex generated clear,separate representations of sounds occurring within the timedimensions of a syllable," said Nagarajan. "In poorreaders, the brain separately generated only very weakrepresentations of sound events past the first sound.

"In the normal reader, successive intra-syllabic soundevents are separately represented in high fidelity withinthe processing channels of the 'primary' auditory cortex. In the impaired reader, they are not," he said.

"These findings are consistent with the increasingevidence," said Merzenich, "that language-impaired andreading-impaired children are a very broadly synonymouspopulation. Scientists have historically argued that only asmall percentage of dyslexics have a clear history of earlylanguage impairment and fundamental auditory processingdeficits. To the contrary, we have seen that most poorreaders and most language-impaired children share these samefundamental listening and brain processing abnormalities."

Moreover, he said, "The studies show that these fundamentallistening problems clearly persist across a lifetime, evenwhile the basic speech reception abilities of theseindividuals can ultimately achieve a relatively normalcompetency."

Co-authors of the study were Henry Mahncke, PhD, a researchfellow in the Keck Center at UCSF; Talya Salz, of ScientificLearning Corporation in Berkeley, CA; Paula Tallal, PhD,co-director of the Center for Molecular and BehavioralNeuroscience at Rutgers University, Newark, NJ; and TimothyRoberts, PhD, assistant professor of radiology in theBiomagnetic Imaging Laboratory, Department of Radiology, atUCSF.

The study was funded by the National Institutes of Health,the Office of Naval Research, Hearing Research Inc., and theColeman fund.