Purpose

Previous reports have shown that resting concentrations of gamma-amino butyric acid (GABA) will predict resting state Functional Connectivity (rsFC) measures¹ as well as amplitude of task activation^2,3. Furthermore, the correlations quantified from rsFC have been shown to covary with behavioral performance in the context of the brain’s reading circuit⁴. The goal of this study is to model how the neurochemical profile interplays with the connectivity underlying the decoding and fluency components of the reading circuit.

Methods

Subjects: Nine typical and six adult struggling readers were recruited from the Center for the Study of Adult Literacy, classified based on reading assessments using the Woodcock-Johnson Test of Achievement-III. MRI/MRS acquisition: High-resolution T1-weighted MPRAGE and rsfMRI (TR=2sec, TE=30ms, voxel=3.4x3.4x4mm3, 32 slices) images were acquired on a Siemens 3T Tim Trio with a 12 ch-head coil. Magnetic Resonance Spectroscopy (MRS) data was collected using MEGA-PRESS, which applies a J-editing pulse to identify the GABA metabolite (TR=2000ms, TE=68ms, edit pulse frequency=1.9ppm, control pulse frequency=7.5ppm, VAPOR water suppression). The 3x3x3cm³ voxel was placed in the Inferior Frontal Gyrus (IFG). An unsuppressed water spectrum from the same area was also acquired for eddy current compensation. Processing of rsfCMRI: The rsFC images were corrected for slice timing, global head motion, EPI distortions, physiological noise, and spatially normalized to MNI, followed by masking of the ventricles, low-pass filtering between 0.001 and 0.1Hz, and spatially smoothed (FWHM=6mm). Seed-based CC analysis was applied in a whole-brain manner by seeding in Left pars Opercularis (L-pOP), and transformed using a Fisher Z-transform. Processing of MRS: The MRS data was pre-processed with in-house Matlab scripts to perform frequency and phase correction, align editing and control spectra on the Creatine (Cr) peak, subtract editing and control spectra to generate the difference spectrum, and 2Hz line broadening. The difference and control spectra were separately fitted to a simulated basis set using LCModel.

Results

The fit of the GABA+ signal in the difference spectrum was excellent (CRLB 3-8%; Fig1B). The rsFC measures were also of high quality, such that they were sensitized predominantly to grey matter neuronal activity. Seeding in L-pOP, we found significant connections with Left Superior Temporal Gyrus (L-STG) in both typical and struggling readers (corrected p<0.001, cluster size=100), but only detected significant connections to Right pars Triangularis (R-pTR) in struggling readers (Fig1C). The rsFC strength of L-pOP to L-STG is significantly and positively related to the GABA+/Cr concentration in IFG (Fig1E). Furthermore, the amount of GABA+/Cr in IFG is positively related with a subject’s ability to decode written language (Fig1F). The ability to decode written language is directly and significantly related to how fluently the subject is able to read (Fig1G), but is negatively impacted by right hemisphere (RH) frontal connections (Fig1H). These results culminate in the neurochemical-neurovascular-cognitive model shown in Fig1D, where the left hemisphere (LH) connections in tandem with the neurochemical profile promote reading fluency, whereas RH connections decrease fluency.

Discussion and Conclusion

The LpOP to LSTG connection is known as the dorsal path in the reading network, and has previously been linked to decoding ability⁵. In task-fMRI studies, it has been suggested that some struggling readers have a right-lateralized reading circuit, which is less efficient than the typical left-lateralized reading network in typical subjects⁶. Our model is in support of these observations, but goes beyond by interrogating the underlying networks that support the behavior, and combining that information with the neurochemistry that characterize the trait. These preliminary results of combining MRS and rsFC with neuropsychological measures are promising, and will help identify the underlying dysfunction in struggling adult readers’ brain circuitry. Future work will explore the neurochemical profile in other related brain regions of struggling readers, which will be combined with information from whole brain rsFC.

Acknowledgements

No acknowledgement found.

References

[1]Kwan SH, et al. Neonatology 2014. [2]Hu Y, et al. J Neurosci 2013. [3]Muthukumaraswamy SD, et al. Proc Natl Acad Sci USA 2009. [4] Hampson M, et al. Neuroimage 2006. [5] Pugh KR, et al. Ment Retard Dev Disabil Res Rev 2000.[6] Waldie KE, et al. Brain Sci 2013.