Symposium Chair: Brett Miller, Ph.D., Program Director, Reading, Writing, & Related Learning Disabilities Research Program, Child Development & Behavior Branch, Eunice Kennedy Shriver National Institute of Child Health and Human Development
Cognitive neuroscience has provided significant insight into the development of reading skills and the underlying causes and consequences of reading difficulties and dyslexia. This full-day symposium includes a range of approaches to understanding the brain-basis of reading disability. These include: big-data approaches to inform our understanding of dyslexia; foundational research examining the relationship between general cognitive skills and reading specific skills; research leveraging twin designs to inform our understanding of the genetic etiology of the neural structure and function of reading comprehension and provide insight into networks that may be influenced by environmental variation; utilization of neurobiological indices to predict learners’ responses to intervention; and an examination of the nature of the neural response to a school-based reading intervention, specifically focusing on reading comprehension and its interactions with executive functions. These studies demonstrate some of the range of ongoing, scientific research to improve our understanding of dyslexia, with the goal of improving long-term outcomes for our learners.
Shared Big Data Can Advance Our Understanding of Dyslexia
Mark A. Eckert, Ph.D., Associate Professor of Otolaryngology, Head and Neck Surgery
The availability of open access data and novel data mining analyses provide exciting opportunities to advance our understanding of dyslexia. Distinct advantages of research using shared big data include increased statistical power from large sample sizes, the integration of genetic, neurobiologic, and behavioral data, as well as new collaborative opportunities. Pitfalls can arise when integrating data from different research sites because of missing data and varied sampling approaches, for example, that may confound results or limit interpretations. These topics will be discussed in the context of a multi-site study involving retrospective structural imaging and behavioral data that includes dyslexia group and reading ability dimensional datasets from 23 international research groups. The behavioral profiles and gross brain morphology of people with dyslexia in this large and growing dataset is quite varied. We have observed modest but consistent gray matter effects across dyslexic children with varied behavioral profiles compared to controls and after dealing with missing data within and between research sites. In particular, children with dyslexia are more likely to exhibit lower gray matter volume in a left superior temporal gyrus/sulcus relative to controls. Consistent effects across research sites and implications for understanding dyslexia highlight the value of using shared data.
Neurocognitive Components of Reading, Task Difficulty, and Some Notes on Feedback
William W. Graves, Ph.D., Assistant Professor, Department of Psychology, Rutgers University, Newark
Much work from my lab and others’ has attempted to better understand the neural basis of reading by breaking it down into its cognitive components in order to determine where and how those component processes are carried out in the brain. Recently our ongoing investigations uncovered a curious finding. Comparing neural activation for words relative to pseudowords (pronounceable nonwords) typically activates a consistent set of areas for words that have come to be thought of as the semantic system, while pseudowords largely activate the dorsal attention system. We have found conditions under which essentially the opposite pattern obtains–words can activate the dorsal attention system, and pseudowords can activate the putative semantic system. This seems to be related to task difficulty effects. In studies manipulating the difficulty with which words are distinguished from pseudowords, we have demonstrated changes in these networks based on how easily discriminable words are from pseudowords, and enhanced top-down influence for less familiar words from brain regions involved in cognitive control. One hypothesis this leads to is that reading training might benefit most from a combination of reading-specific and domain-general feedback. Preliminary behavioral and neuroimaging data will be presented that support this hypothesis.
Neurobiological Underpinnings of Passage Comprehension: A Twin Study
Stephen Petrill, Ph.D., Professor of Psychology, Ohio State University
Integrating behavioral genetics and neuroimaging approaches could yield important and foundational understanding not only of the genetic etiology of neural structure and neural function of reading comprehension, but also insight into networks that may be influenced by environmental variation. Data is drawn from N = 300 pairs of 14 to16-year-old MZ and DZ twins, who are participating in ongoing longitudinal twin projects at Ohio State and the University of Colorado. Twins are concurrently assessed on a battery of behavioral measures of reading, including word-level skills, reading fluency, and reading comprehension, as well as functional measures of reading comprehension, numeric estimation, and working memory. Structural and resting state measures are also being collected. Results suggest that expository text comprehension was associated with regions related to mental model updating and integration (e.g. posterior cingulate cortex and precuneus). Moreover, over the course of the task, activation in this semantic control region increased, whereas regions associated with attention decreased. We are currently examining whether these regions account for a portion of the genetic and environmental variance in reading comprehension, and how these relate to behavioral and brain-based measures of working memory and mathematical cognition.
Neurobiological Indices as Predictors of Intervention Response: Implications for Reading Outcomes
Laurie Cutting, Ph.D., Patricia and Rodes Hart Professor of Special Education, Psychology, Radiology & Pediatrics, Vanderbilt University
Studies have shown that implementing evidenced based interventions results in improved reading abilities, along with changes in brain function; more recently, research has begun to ask whether neurobiological measures (neuroimaging) can predict response to intervention. This presentation will review neurobiological indices (structural, functional, and connectivity), along with behavioral measures, that predicted outcome to a short-term (15 hour) reading intervention focusing on building word-level skills. Additionally, the role of executive function in responsiveness to intervention will be discussed.
A fMRI Study of Sentence Reading and Response Inhibition in Pre- and Post-Intervention Struggling Readers
Jessica Church, Ph.D., Assistant Professor, Dept of Psychology, University of Texas at Austin
A multi-modal neuroimaging study investigated brain activity in 4th grade children recruited from a multi-city, in-school reading intervention project. We were particularly focused on reading comprehension and its interactions with executive functions, and thus we collected neuroimaging data during a sentence reading task as well as during a stop-signal task that has been commonly used to measure response inhibition, an important executive function component. Inhibition may play an important role in reading; for example, path analysis studies suggest that inhibitory control predicts reading comprehension performance in 4th grade students (Kieffer, Vukovic, & Berry, 2013).
Our fMRI analysis of the sentence reading task found significantly greater activity in ventral occipito-temporal, motor, and frontal cortex in struggling readers relative to non-struggling readers, suggestive of greater item-by-item (less fluent) processing in struggling readers. Our investigation of the stop-signal task found significantly different activation in putative task control regions during correct “go” trials and failed “stop” trials in the post-intervention readers relative to the pre-intervention readers. These data suggest that reading intervention-related brain change may occur in control-related brain regions on a different timeframe than reading-related regions. Our results are interpreted in the context of age and the outcomes of the reading intervention.