Neurostimulation improves phonological decoding in dyslexia

1024 

Abstract Submission 

Sabrina Turker^1,2, Philipp Kuhnke^1,2, Vincent Cheung³, Gesa Hartwigsen^1,2

¹Leipzig University, Leipzig, Germany, ²Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, ³Sony Computer Science Laboratories, Tokyo

Sabrina Turker  
Leipzig University|Max Planck Institute for Human Cognitive and Brain Sciences
Leipzig, Germany|Leipzig, Germany

Philipp Kuhnke  
Leipzig University|Max Planck Institute for Human Cognitive and Brain Sciences
Leipzig, Germany|Leipzig, Germany

Vincent Cheung  
Sony Computer Science Laboratories
Tokyo

Gesa Hartwigsen  
Leipzig University|Max Planck Institute for Human Cognitive and Brain Sciences
Leipzig, Germany|Leipzig, Germany

Literacy is key to social contacts, education, and employment, and significantly influences personal well-being and mental health (Huettig, 2015). This is detrimental for the at least two children in every classroom worldwide who have dyslexia, a learning disability affecting reading and writing (ICD-11: 6A03.04; WHO, 2019/21). From a cognitive perspective, dyslexia is mainly characterized by problems in phonological decoding and accessing the respective phonological representations of words (Ramus, 2006). One of the most consistently reported neural characteristics of dyslexia is hypoactivation of the left temporo-parietal cortex (TPC), an area referred to as the 'phonological decoding' center (Richlan, 2009). Neurostimulation constitutes a promising technique to alleviate reading deficits in dyslexia (Cancer, 2018; Turker, 2021) but stimulation-induced changes in activation and connectivity remain unclear.

In the present study, we combined offline facilitatory repetitive transcranial magnetic stimulation (TMS) with fMRI in a within-subject design. We applied sham and effective TMS over the left TPC of 26 adults with dyslexia (18-39 years) before they read simple (2 syllables, 4-6 letters) and complex (3-4 syllables, 10-14 letters) words and pseudowords aloud during fMRI. Our goal was to explore TMS-induced changes in reading performance, functional activation and connectivity. To do so, we performed linear mixed models for investigating behavioural improvements and combined univariate and multivariate analyses (Multivariate Pattern Analysis; Hebart, 2015) to explore changes in functional activation. Last, we used Dynamic Causal Modeling (Friston, 2003) to explore effective connectivity during reading within the classical reading network (left inferior frontal gyrus/IFG, left TPC and left ventral occipito-temporal cortex/vOTC) and an extended network comprising hypoactive regions in the respective sample.

Behaviorally, the facilitation of the left TPC improved speech onsets for complex pseudowords and better performance was directly linked to functional coupling between the left IFG and the left vOTC. Whereas stimulation did not affect activation within the left TPC, it led to a lower recruitment of right-hemispheric areas for words and pseudowords, as well as to significantly altered neural interactions in the classical and extended reading network. In addition to a stronger recruitment of the ventral reading route (vOTC - IFG), stronger pseudoword-specific connectivity from the left supramarginal gyrus and lower inhibition from the right cerebellum on other reading areas seemed to be crucial for improving phonological decoding in dyslexia. Moreover, we found that only one subregion within the larger TPC region, the anterior SMG, showed deactivation for simple and complex pseudowords (see also Turker, 2023), suggesting that only this portion of the left TPC is causally involved in the phonological deficit in dyslexia.

We provide evidence for a crucial role of the left TPC for complex pseudoword reading and highlight the importance of network interactions within the core reading network and hypoactive brain areas in adults with dyslexia. Based on our findings on the importance of right-hemispheric regions both in terms of activation and connectivity, we conclude that current models of reading in dyslexia might need an update to capture the variety of areas involved in reading processing in dyslexia and future research should more closely investigate the contributions of the right cerebellum to reading difficulties.

TMS ²

Reading and Writing ¹

Neural Plasticity and Recovery of Function

Connectivity (eg. functional, effective, structural)

fMRI Connectivity and Network Modeling

Cerebellum

FUNCTIONAL MRI

Language

Transcranial Magnetic Stimulation (TMS)

Other - Dyslexia