Reduced Structural Connectivity Between Left Auditory Thalamus and the Motion-Sensitive Planum Temporale in Developmental Dyslexia

Abstract

Developmental dyslexia is characterized by the inability to acquire typical reading and writing skills. Dyslexia has been frequently linked to cerebral cortex alterations; however, recent evidence also points toward sensory thalamus dysfunctions: dyslexics showed reduced responses in the left auditory thalamus (medial geniculate body, MGB) during speech processing in contrast to neurotypical readers. In addition, in the visual modality, dyslexics have reduced structural connectivity between the left visual thalamus (lateral geniculate nucleus, LGN) and V5/MT, a cerebral cortex region involved in visual movement processing. Higher LGN-V5/MT connectivity in dyslexics was associated with the faster rapid naming of letters and numbers (RANln), a measure that is highly correlated with reading proficiency. Here, we tested two hypotheses that were directly derived from these previous findings. First, we tested the hypothesis that dyslexics have reduced structural connectivity between the left MGB and the auditory-motion-sensitive part of the left planum temporale (mPT). Second, we hypothesized that the amount of left mPT-MGB connectivity correlates with dyslexics RANln scores. Using diffusion tensor imaging-based probabilistic tracking, we show that male adults with developmental dyslexia have reduced structural connectivity between the left MGB and the left mPT, confirming the first hypothesis. Stronger left mPT-MGB connectivity was not associated with faster RANln scores in dyslexics, but was in neurotypical readers. Our findings provide the first evidence that reduced cortico-thalamic connectivity in the auditory modality is a feature of developmental dyslexia and it may also affect reading-related cognitive abilities in neurotypical readers.SIGNIFICANCE STATEMENT Developmental dyslexia is one of the most widespread learning disabilities. Although previous neuroimaging research mainly focused on pathomechanisms of dyslexia at the cerebral cortex level, several lines of evidence suggest an atypical functioning of subcortical sensory structures. By means of diffusion tensor imaging, we here show that dyslexic male adults have reduced white matter connectivity in a cortico-thalamic auditory pathway between the left auditory motion-sensitive planum temporale and the left medial geniculate body. Connectivity strength of this pathway was associated with measures of reading fluency in neurotypical readers. This is novel evidence on the neurocognitive correlates of reading proficiency, highlighting the importance of cortico-subcortical interactions between regions involved in the processing of spectrotemporally complex sound.

Keywords: DWI; RAN; developmental dyslexia; medial geniculate body (MGB); planum temporale (PT); primary auditory cortex (A1).

Copyright © 2019 the authors 0270-6474/19/391720-13$15.00/0.

Figures

Figure 1.

The left side of the figure displays the statistical parametrical map of the localizer contrast “Sentences - Silence”. The color bar represents t-values. The crosshair over the zoomed-in medial geniculate body responses indicates the statistic peak location from the localizer contrast “Sentences – Silence” used for definition of ROIs. The right side shows the ROIs centered at the statistic peak location of the localizer contrast “Sentences – Silence”. The maps and ROIs are superimposed on the same section of the MNI152 structural T1 volume.

Figure 2.

Localization of the MGB masks on the individual subject T1 brain presented for six randomly chosen dyslexic and neurotypical participants.

Figure 3.

ROIs in MNI standard space superimposed on sections of the MNI152 structural T1 volume.

Figure 4.

A, Averaged probabilistic white matter connectivity for neurotypicals and dyslexics between the left mPT and the left MGB (green). The log-normalized and averaged tracks are presented in MNI standard space and thresholded to the same minimum value of 0.08. B, Mean connectivity strength of the left mPT–MGB connection for neurotypicals and dyslexics. Error bars indicate ± 1 SEM.

Figure 5.

Mean connectivity strength for neurotypicals and dyslexics between the mPT and the MGB. Error bars indicate ± 1 SEM.

Figure 6.

A, Averaged probabilistic white matter connectivity for neurotypicals and dyslexics between the mPT and the MGB (dark green) and between A1 and the MGB (blue). The log-normalized and averaged tracks are presented in MNI standard space and thresholded to the same minimum value of 0.08. B, Mean strength of the mPT–MGB and the A1–MGB connectivity for neurotypicals and dyslexics. Error bars indicate ± 1 SEM.

Figure 7.

Averaged probabilistic white matter connectivity for neurotypicals and dyslexics between the MGB and the IC. A, Log-normalized and averaged tracks are presented in MNI standard space and thresholded to the same minimum value of 0.50. B, Mean connectivity strength of tracks between IC and MGB for neurotypicals and dyslexics. Error bars indicate ± 1 SEM.

Figure 8.

Correlation between the RANln and the connectivity index of the white matter pathway between the left mPT and the left MGB. A significantly negative correlation emerged for neurotypical participants, suggesting that stronger mPT–MGB connectivity was associated with faster RANln.