Developmental Dyslexia: An Educational Neuroscience Approach

Around 7% of children have specific problems in acquiring reading and writing skills, with a ratio of 4:1 boys:girls. When there is no obvious sensory, neural or educational basis, this learning difficulty is called dyslexia. The causes of dyslexia are poorly understood, but some features are common across different languages. One is difficulties in phonological tasks like counting the number of syllables in words and deciding whether words rhyme. Another is difficulty in distinguishing how rapidly sounds begin - dyslexics find it difficult to distinguish the abruptness of onset of sounds, for example a note played by a trumpet (sharp onset or ?rise time?) vs a violin (more extended onset or ?rise time?). We think these difficulties with phonology and rise time are linked, because the brain uses the onsets of sounds to define where the modulation (the important frequency and amplitude information) is in speech. The brain also uses sound onsets to reset the oscillation patterns of groups of brain cells that track the incoming acoustic signal by locking their oscillatory frequency to this signal. If the perception of sound onsets is impaired, then this tracking mechanism would not work as efficiently in a dyslexic brain. Some of these cell networks are driven by syllable-level fluctuations in the incoming signal, which means they are driven by lower frequency modulations. These low frequency modulations are crucial for speech intelligibility and are also important for hearing rhythm in music and syllable stress in speech. Our current rise time data suggest that children with dyslexia find both hearing metrical structure in music and syllable stress in speech difficult. In our work, we will test whether the brains of children with dyslexia have a specific difficulty in perceiving these low frequency modulations, using a variety of behavioural and brain imaging approaches. If we find that there is this proposed sensory/neural difficulty in dyslexia, then interventions based on music and metrical language (like poetry and nursery rhymes) may be very beneficial early in development for enriching this kind of auditory perception. We hope to improve our understanding of the causes of dyslexia through this research, and to find biomarkers (specific neural signatures of processing difficulty) that could help identify children at risk early in life. If the factors we propose turn out to be important, they will also help us in designing better educational programmes for children with dyslexia.

Developmental dyslexia (DD) affects 7% of schoolchildren and presents a major obstacle to educational success. Children with DD have specific problems with reading and spelling that are not due to low intelligence, poor educational opportunities or overt sensory or neurological damage. Subtle sensory/neural problems are however likely. Here we adopt an educational neuroscience approach better to understand the potential sensory and neural causes of DD with the aim of identifying biomarkers for the early identification of children at risk, which would enable earlier remediation. We focus on understanding the mechanisms of atypical phonological development (the core feature of DD across different languages) at the neural, cognitive and sensory levels. To this end we propose a ?temporal sampling? framework based on prior MRC-funded work, intended to capture both the temporal properties of speech and the temporal properties of the oscillatory neural networks that are thought to sample the auditory input. We test this framework using behavioural, physiological and neuroimaging approaches. This innovative perspective integrates adult neuroimaging and contemporary speech science data with infancy and child phonological data, bringing together literatures that have to date not overlapped. Our proposal is that auditory neural and sensory processes work differently in dyslexic brains, with impaired envelope perception and possibly enhanced perception of temporal fine structure. Behaviourally this may lead to impaired syllable level mechanisms and (possibly) over-specified segment-level mechanisms. This results developmentally in speech-based representations that work for language but not for literacy. Complementary studies of cognitive and linguistic features provide information for theory-driven educational interventions, which are also informed by current work with typically-developing 4- and 5-year-olds. This multi-level research strategy should identify a range of neural and cognitive biomarkers of risk for reading/ language disorders, as well as identify factors relevant to designing educational interventions.

The proposal tests the temporal sampling framework with 112 children with and without dyslexia who have been followed since age 7. Now aged 13-15 years on average, the participants will be tested with EEG and fNIRS imaging, and MEG once they are aged 16 years. The imaging work examines the integrity of phase locking to lower modulation frequencies, which our psychoacoustic data on amplitude envelope onset (rise time) suggest is impaired across languages. The physiology explores peripheral auditory neural mechanisms. The behavioural work tests the predicted associated impairments in syllable stress, prosodic perception and musical rhythm, seeking factors that can form the basis of remediation.