Dynamic interactions between brain changes and reading development

Dynamic interactions between brain changes and reading development

First Author: Maaike Vandermosten -- KU Leuven
Additional authors/chairs: 
Jason Yeatman
Keywords: educational neuroscience, Longitudinal, early identification, Intervention, Reading development
Abstract / Summary: 

Learning to read is associated with substantial changes in brain structure and function. Studying the dynamic interplay of both behavioral and neural factors is crucial for our understanding of both typical and atypical reading development and has important clinical and educational implications. In this symposium we bring together studies that look at neural predictors of later reading ability by means of both neuroanatomical (Bouhali) and neurophysiological (González) investigations, investigate the contribution of family risk in structural connectivity pre-reading (Blockmans), and investigate training-induced neuroplasticity in both early (Economou) and advanced (Yeatman) readers. The symposium aims to further our understanding of the dynamic associations between reading development and changes at the neural level, and how these change in response to intervention.

Symposium Papers: 

Deciphering directionality in the association between cortical structure and reading skills across development

First Author/Chair:Florence Bouhali -- University of California San Francisco
Additional authors/chairs: 
Luxi Feng; Emilio Ferrer; Timothy Brown; Terry Jernigan; Richard Wagner; Fumiko Hoeft

Purpose: Brain-behavior correlations are often used to understand the mechanisms underlying literacy acquisition and the development of functional specialization for written language. Yet, these correlations are the result of complex bi-directional interactions between brain and behavior over the course of development. Here, we employ latent change score modeling of data from a longitudinal design with multiple time points to disentangle the dynamics and directionality between cortical structural properties and reading skills.
Methods: Structural T1-weighted MRI scans and reading efficiency TOWRE scores were obtained from 186 children, 130 of whom were followed longitudinally over 2 to 4 time points. Surface area and cortical thickness were derived using Freesurfer from key regions of the left-hemispheric reading network: the fusiform gyrus, the inferior frontal gryus pars opercularis and triangularis, and the supramarginal gyrus. Latent change score modeling was applied to determine the time-related associations between cortical structure and reading efficiency.
Results: Surface area of the left fusiform gyrus was found to significantly predict future gains in sight word reading efficiency, but not the reverse. No such association was observed with phonemic decoding efficiency scores, in line with the implication of the fusiform gyrus with the lexical direct reading route. Moreover, we did not find any significant interrelation between changes in cortical thickness and changes in reading scores in the studied regions.
Conclusions: Our results support a strong role of the cortical structure of high-level visual cortex as a leading indicator of reading acquisition, in particular of word reading skills.

Visual N1 sensitivity to letters in kindergarten and during primary school in children with familial risk for dyslexia

First Author/Chair:Fraga González -- University of Zurich
Additional authors/chairs: 
Georgette Pleisch; Sarah Di Pietro; Jasmin Neuenschwander; Susanne Walitza; Daniel Brandeis; Iliana Karipidis; Sylvia Brem

Specialization of visual systems for fast print recognition is key to reading acquisition. Its electrophysiological correlate is the occipito-temporal N1 response, peaking around 200 ms. N1 sensitivity to print is associated with reading ability but its development in the first school years remains unclear. An inverted U trajectory was observed in words-N1, but only few studies focused on more basic, single character processing. We aim at filling this gap by examining letter-N1 throughout primary school.
Five electroencephalography (EEG) measurements took place from kindergarten to firth grade. In total, 65 children with familial risk for dyslexia performed a visual target detection task presenting single letters and false fonts. We examined interactions between measurement and condition on N1 amplitudes and correlations with reading and familial risk for dyslexia.
Stronger N1 for letters was found in first grade, when letter knowledge gains were more pronounced. Then, bilateral letter-N1 correlated with letter knowledge, and the left letters-false fonts N1 differences correlated with family risk. No associations between N1 and reading fluency were observed. Before and after first grade, N1 did not discriminate between condition.
We found support for an inverted U trajectory in N1 letter sensitivity which reaches its maxima when letter knowledge is consolidating and fades shortly after. The results suggest N1 is a highly specialized developmental response involved in early stages of learning. Further work on a broader population in first grade could elucidate N1 relation with later reading skills and other areas of the reading network.

Pre-reading differences along white matter tract profiles in children at risk for dyslexia

First Author/Chair:Lauren Blockmans -- Research group Experimental ORL, Department of Neurosciences, KU Leuven, Belgium
Additional authors/chairs: 
Fumiko Hoeft; Jan Wouters; Pol Ghesquière; Maaike Vandermosten

Purpose. Reading is a complex ability relying on a widespread neural network along the left hemisphere. Diffusion imaging allows for investigation of white matter pathways implicated in reading and for dyslexia-related differences. Previous studies have consistently shown different white matter organization in the left arcuate fasciculus (AF) in both adults and children with dyslexia, even prior to reading onset. However, longitudinal studies in pre-readers are biased towards persons with dyslexia with a family risk (FR). Given that recent studies suggested that brain differences in auditory and phonological regions of the temporal lobe are driven by FR rather than reading outcome, we aim to determine the influence of FR on the neural predictors of dyslexia.
Method. We selected pre-reading children with and without FR and/or cognitive risk (CR) for dyslexia. We administered behavioral tasks tapping into reading-related abilities and acquired diffusion-weighted MRI scans in 134 pre-readers: 35 FR⁺CR⁺, 22 FR⁺CR⁻, 28 FR⁻CR⁻ and 49 FR⁻CR⁺. Using Automated Fiber Quantification, we determined fractional anisotropy (FA) along the course of the left AF.
Results. Analyses in a subgroup (n = 52) already revealed that FR does not drive the FA differences in the direct segment of the left AF between pre-readers who develop typical reading skills and pre-readers who develop dyslexia. Extending these analyses to our entire study sample will allow for a more detailed look into the contribution of FR at the pre-reading stage.
Conclusion. The findings of this study can clarify how FR relates to the connectivity differences in at-risk pre-readers.

Training-induced white matter plasticity in pre-readers at risk for dyslexia

First Author/Chair:Maria Economou -- KU Leuven, Belgium
Additional authors/chairs: 
Jolijn Vanderauwera; Toivo Glatz; Femke Vanden Bempt; Shauni Van Herck; Pol Ghesquière; Jan Wouters; Maaike Vandermosten

Purpose: Recent findings suggest that reading remediation in dyslexia is most effective when it takes place during the early stages of reading development, yet it usually only starts after years of severe and persistent reading problems. Despite the importance of preventive intervention in overcoming this paradox, its neural impact remains relatively unexplored. We will address this by investigating the neural effects of an early literacy intervention in pre-readers at risk for dyslexia. Specifically, we aim to characterize white matter changes in response to intervention and explore the relation with reading-related behavioral outcomes.
Method: Sixty kindergarteners with an increased cognitive risk for developing dyslexia received either tablet-based literacy training (GraphoGame Flemish) or active control training for 12 weeks. In addition, a control group of 30 typically developing kindergarteners was included. Prior to and after intervention, an extensive battery of cognitive measures, as well as diffusion-weighted and T2-relaxation images, were acquired in the 90 children.
Results: A preliminary analysis of the left arcuate fasciculus in a subsample (n=45) revealed significant changes in white matter properties for the literacy training group but not for the control groups, suggesting literacy training-specific neuroplasticity. We are currently optimizing the processing pipeline to best deal with the motion artifacts inherent to pediatric MRI data and extending the analyses to include the full sample. Analyses are ongoing but detailed results will be presented.
Conclusions: The present study can contribute to our understanding of early training-induced plasticity in children at risk for developing dyslexia, in relation to typical development.

Plasticity and learning: How interventions shape white matter development

First Author/Chair:Jason Yeatman -- Stanford University
Additional authors/chairs: 
Elizabeth Huber

Purpose: The process of learning to read has a profound impact on brain development, affecting the way sensory systems encode information, changing the structure of neural connections between visual and language regions, and leading to a new form of communication through written language. Over the past decade a wealth of data has demonstrated links between individual differences in white matter structure and reading skills but what has remained unclear is the extent to which these differences remain consistent throughout development or change as a function of a child’s educational environment. Here we use an intensive intervention program to examine the stability/plasticity of individual differences in white matter structure.
Method: 36 children with reading difficulties were recruited for an intensive summer intervention program (160 hours of one-on-one instruction in Seeing Stars). Diffusion MRI data and behavioral measures of reading skills were collected longitudinally (4 measurements per subjects).
Results: Intervention induced large-scale changes in white matter diffusion properties, that occurred over a rapid timescale, and in synchrony with reading improvements. Intervention subjects showed significant changes in diffusion properties throughout an extensive network of white matter tracts and measures were stable in control subjects.
Conclusions: White matter constitutes a highly plastic system that flexibly adapts to environmental demands and changes as a function of a reading education.