Author response:
The following is the authors’ response to the original reviews.
Reviewer #1 (Public Review):
Summary:
"Neural noise", here operationalized as an imbalance between excitatory and inhibitory neural activity, has been posited as a core cause of developmental dyslexia, a prevalent learning disability that impacts reading accuracy and fluency. This study is the first to systematically evaluate the neural noise hypothesis of dyslexia. Neural noise was measured using neurophysiological (electroencephalography [EEG]) and neurochemical (magnetic resonance spectroscopy [MRS]) in adolescents and young adults with and without dyslexia. The authors did not find evidence of elevated neural noise in the dyslexia group from EEG or MRS measures, and Bayes factors generally informed against including the grouping factor in the models. Although the comparisons between groups with and without dyslexia did not support the neural noise hypothesis, a mediation model that quantified phonological processing and reading abilities continuously revealed that EEG beta power in the left superior temporal sulcus was positively associated with reading ability via phonological awareness. This finding lends support for analysis of associations between neural excitatory/inhibitory factors and reading ability along a continuum, rather than as with a case/control approach, and indicates the relevance of phonological awareness as an intermediate trait that may provide a more proximal link between neurobiology and reading ability. Further research is needed across developmental stages and over a broader set of brain regions to more comprehensively assess the neural noise hypothesis of dyslexia, and alternative neurobiological mechanisms of this disorder should be explored.
Strengths:
The inclusion of multiple methods of assessing neural noise (neurophysiological and neurochemical) is a major advantage of this paper. MRS at 7T confers an advantage of more accurately distinguishing and quantifying glutamate, which is a primary target of this study. In addition, the subject-specific functional localization of the MRS acquisition is an innovative approach. MRS acquisition and processing details are noted in the supplementary materials according to the experts' consensus-recommended checklist (https://doi.org/10.1002/nbm.4484). Commenting on the rigor, the EEG methods is beyond my expertise as a reviewer.
Participants recruited for this study included those with a clinical diagnosis of dyslexia, which strengthens confidence in the accuracy of the diagnosis. The assessment of reading and language abilities during the study further confirms the persistently poorer performance of the dyslexia group compared to the control group.
The correlational analysis and mediation analysis provide complementary information to the main case-control analyses, and the examination of associations between EEG and MRS measures of neural noise is novel and interesting.
The authors follow good practice for open science, including data and code sharing. They also apply statistical rigor, using Bayes Factors to support conclusions of null evidence rather than relying only on non-significant findings. In the discussion, they acknowledge the limitations and generalizability of the evidence and provide directions for future research on this topic.
Weaknesses:
Though the methods employed in the paper are generally strong, there are certain aspects that are not clearly described in the Materials & Methods section, such as a description of the statistical analyses used for hypothesis testing.
Thank you for pointing this out. A description of the statistical models used in the analyses of EEG biomarkers has been added to the Materials and Methods:
“First, exponent and offset values were averaged across all electrodes and analyzed using a 2x2 repeated measures ANOVA with group (dyslexic, control) as a between-subjects factor and condition (resting state, language task) as a within-subjects factor. Age was included in the analyses as a covariate due to the correlation between variables. Next, exponent and offset values were averaged across electrodes corresponding to the left (F7, FT7, FC5) and right inferior frontal gyrus (F8, FT8, FC6), and to the left (T7, TP7, TP9) and right superior temporal sulcus (T8, TP8, TP10). The electrodes were selected based on the analyses outlined by Giacometti and colleagues (2014) and Scrivener and Reader (2022). For these analyses, a 2x2x2x2 repeated measures ANOVA with age as a covariate was conducted with group (dyslexic, control) as a between-subjects factor and condition (resting state, language task), hemisphere (left, right), and region (frontal, temporal) as within-subjects factors. Results for the alpha and beta bands were calculated for the same clusters of frontal and temporal electrodes and analyzed with a similar 2x2x2x2 repeated measures ANOVA; however, for these analyses, age was not included as a covariate due to a lack of significant correlations.”
We also expanded the description of the statistical models used in the analyses of MRS biomarkers:
“To analyze the metabolite results, separate univariate ANCOVAs were conducted for Glu, GABA+, Glu/GABA+ ratio and Glu/GABA+ imbalance measures with group (control, dyslexic) as a between-subjects factor and voxel gray matter volume (GMV) as a covariate. Additionally, for the Glu analysis, age was included as a covariate due to a correlation between variables. Both frequentist and Bayesian statistics were calculated. Glu/GABA+ imbalance measure was calculated as the square root of the absolute residual value of a linear relationship between Glu and GABA+ (McKeon et al., 2024).”
With regard to metabolite quantification, it is unclear why the authors chose to analyze and report metabolite values in terms of creatine ratios rather than quantification based on a water reference given that the MRS acquisition appears to support using a water reference.
We have decided to use the ratio of Glu and GABA to total creatine (tCr), as this is still a common practice in MRS studies at 7T (e.g., Nandi et al., 2022; Smith et al., 2021). This approach normalizes the signal, reducing the impact of intensity variations across different regions and tissue compositions. Additionally, total creatine concentration is considered relatively stable across different brain regions, which is particularly important in our study, where a functional localizer was used to establish the left STS region individually. Our decision was further influenced by previous studies on dyslexia (Del Tufo et al., 2018; Pugh et al., 2014) which have reported creatine ratios and included GM volume as a covariate in their models, thus providing comparability. It is now indicated in the Results:
“For comparability with previous studies in dyslexia (Del Tufo et al., 2018; Pugh et al., 2014) we report Glu and GABA as a ratio to total creatine (tCr).”
and in the Method sections:
“Glu and GABA+ concentrations were expressed as a ratio to total-creatine (tCr; Creatine + Phosphocreatine) following previous MRS studies in dyslexia (Del Tufo et al., 2018; Pugh et al., 2014).
We did not estimate absolute concentrations using water signals as a reference, as this would require accounting for water relaxation times, which may vary across our age range. Nevertheless, our dataset has been made publicly available for future researchers to calculate and compare absolute values.
Del Tufo, S. N., Frost, S. J., Hoeft, F., Cutting, L. E., Molfese, P. J., Mason, G. F., Rothman, D. L., Fulbright, R. K., & Pugh, K. R. (2018). Neurochemistry Predicts Convergence of Written and Spoken Language: A Proton Magnetic Resonance Spectroscopy Study of Cross-Modal Language Integration. Frontiers in Psychology, 9, 1507. https://doi.org/10.3389/fpsyg.2018.01507
Nandi, T., Puonti, O., Clarke, W. T., Nettekoven, C., Barron, H. C., Kolasinski, J., Hanayik, T., Hinson, E. L., Berrington, A., Bachtiar, V., Johnstone, A., Winkler, A. M., Thielscher, A., Johansen-Berg, H., & Stagg, C. J. (2022). tDCS induced GABA change is associated with the simulated electric field in M1, an effect mediated by grey matter volume in the MRS voxel. Brain Stimulation, 15(5), 1153–1162. https://doi.org/10.1016/j.brs.2022.07.049
Pugh, K. R., Frost, S. J., Rothman, D. L., Hoeft, F., Del Tufo, S. N., Mason, G. F., Molfese, P. J., Mencl, W. E., Grigorenko, E. L., Landi, N., Preston, J. L., Jacobsen, L., Seidenberg, M. S., & Fulbright, R. K. (2014). Glutamate and choline levels predict individual differences in reading ability in emergent readers. Journal of Neuroscience, 34(11), 4082–4089. https://doi.org/10.1523/JNEUROSCI.3907-13.2014
Smith, G. S., Oeltzschner, G., Gould, N. F., Leoutsakos, J. S., Nassery, N., Joo, J. H., Kraut, M. A., Edden, R. A. E., Barker, P. B., Wijtenburg, S. A., Rowland, L. M., & Workman, C. I. (2021). Neurotransmitters and Neurometabolites in Late-Life Depression: A Preliminary Magnetic Resonance Spectroscopy Study at 7T. Journal of Affective Disorders, 279, 417–425. https://doi.org/10.1016/j.jad.2020.10.011
GABA is typically quantified using J-editing sequences as lower field strengths (~3T), and there is some evidence that the GABA signal can be reliably measured at 7T without editing, however, the authors should discuss potential limitations, such as reliability of Glu and GABA measurements with short-TE semi-laser at 7T.
In addition, MRS measurements of GABA are known to be influenced by macromolecules, and GABA is often denoted as GABA+ to indicate that other compounds contribute to the measured signal, especially at a short TE and in the absence of symmetric spectral editing.
A general discussion of the strengths and limitations of unedited Glu and GABA quantification at 7T is warranted given the interest of this work to researchers who may not be experts in MRS.
While we agree with the Reviewer that at 3T, it is recommended to use J-edited MRS to measure GABA (Mullins et al., 2014), the better spectral resolution at 7T allows for more reliable results for both metabolites using moderate echo-time, non-edited MRS (Finkelman et al., 2022). In this study, we used a short echo time (TE), which is optimal for Glu but not ideal for GABA, as it interferes with other signals. We are grateful to the Reviewer for suggesting the addition of a short paragraph to the Discussion, describing the practicalities of 3T and 7T MRS and changing the abbreviation to GABA+ to inform readers of possible macromolecule contamination:
“We chose ultra-high-field MRS to improve data quality (Özütemiz et al., 2023), as the increased sensitivity and spectral resolution at 7T allows for better separation of overlapping metabolites compared to lower field strengths. Additionally, 7T provides a higher signal-to-noise ratio (SNR), improving the reliability of metabolite measurements and enabling the detection of small changes in Glu and GABA concentrations. Despite these theoretical advantages, several practical obstacles should be considered, such as susceptibility artifacts and inhomogeneities at higher field strengths that can impact data quality. Interestingly, actual methodological comparisons (Pradhan et al., 2015; Terpstra et al., 2016) show only a slight practical advantage of 7T single-voxel MRS compared to optimized 3T acquisition. For example, fitting quality yielded reduced estimates of variance in concentration of Glu in 7T (CRLB) and slightly improved reproducibility levels for Glu and GABA (at both fields below 5%). Choosing the appropriate MRS sequence involves a trade-off between the accuracy of Glu and GABA measurements, as different sequences are recommended for each metabolite. J-edited MRS is recommended for measuring GABA, particularly with 3T scanners (Mullins et al., 2014). However, at 7T, more reliable results can be obtained using moderate echo-time, non-edited MRS (Finkelman et al., 2022). We have opted for a short-echo-time sequence, which is optimal for measuring Glu. However, this approach results in macromolecule contamination of the GABA signal (referred to as GABA+).”
Finkelman, T., Furman-Haran, E., Paz, R., & Tal, A. (2022). Quantifying the excitatory-inhibitory balance: A comparison of SemiLASER and MEGA-SemiLASER for simultaneously measuring GABA and glutamate at 7T. NeuroImage, 247, 118810. https://doi.org/10.1016/j.neuroimage.2021.118810
Mullins, P. G., McGonigle, D. J., O'Gorman, R. L., Puts, N. A., Vidyasagar, R., Evans, C. J., Cardiff Symposium on MRS of GABA, & Edden, R. A. (2014). Current practice in the use of MEGA-PRESS spectroscopy for the detection of GABA. NeuroImage, 86, 43–52. https://doi.org/10.1016/j.neuroimage.2012.12.004
Özütemiz, C., White, M., Elvendahl, W., Eryaman, Y., Marjańska, M., Metzger, G. J., Patriat, R., Kulesa, J., Harel, N., Watanabe, Y., Grant, A., Genovese, G., & Cayci, Z. (2023). Use of a Commercial 7-T MRI Scanner for Clinical Brain Imaging: Indications, Protocols, Challenges, and Solutions-A Single-Center Experience. AJR. American Journal of Roentgenology, 221(6), 788–804. https://doi.org/10.2214/AJR.23.29342
Pradhan, S., Bonekamp, S., Gillen, J. S., Rowland, L. M., Wijtenburg, S. A., Edden, R. A., & Barker, P. B. (2015). Comparison of single voxel brain MRS AT 3T and 7T using 32-channel head coils. Magnetic Resonance Imaging, 33(8), 1013–1018. https://doi.org/10.1016/j.mri.2015.06.003
Terpstra, M., Cheong, I., Lyu, T., Deelchand, D. K., Emir, U. E., Bednařík, P., Eberly, L. E., & Öz, G. (2016). Test-retest reproducibility of neurochemical profiles with short-echo, single-voxel MR spectroscopy at 3T and 7T. Magnetic Resonance in Medicine, 76(4), 1083–1091. https://doi.org/10.1002/mrm.26022
Further, the single MRS voxel location is a limitation of the study as neurochemistry can vary regionally within individuals, and the putative excitatory/inhibitory imbalance in dyslexia may appear in regions outside the left temporal cortex (e.g., network-wide or in frontal regions involved in top-down executive processes). While the functional localization of the MRS voxel is a novelty and a potential advantage, it is unclear whether voxel placement based on left-lateralized reading-related neural activity may bias the experiment to be more sensitive to small, activity-related fluctuations in neurotransmitters in the CON group vs. the DYS group who may have developed an altered, compensatory reading strategy.
We agree that including only one region of interest for the MRS measurements is a potential limitation of our study, and we have now added this information to the Discussion:
“Moreover, since the MRS data was collected only from the left STS, it is plausible that other areas might be associated with differences in Glu or GABA concentrations in dyslexia.”
However, differences in Glu and GABA concentrations in this region were directly predicted by the neural noise hypothesis of dyslexia. We acknowledge that this information was missing in the previous version of the manuscript. It is now included in the Results:
“Moreover, the neural noise hypothesis of dyslexia identifies perisylvian areas as being affected by increased glutamatergic signaling, and directly predicts associations between Glu and GABA levels in the superior temporal regions and phonological skills (Hancock et al., 2017).”
as well as in the Discussion:
“Nevertheless, the neural noise hypothesis predicted increased glutamatergic signaling in perisylvian regions, specifically in the left superior temporal cortex (Hancock et al., 2017).”
Figure 1 contains a lot of information, and it may be helpful to split it into 2 figures (EEG vs. MRS) so that the plots could be made larger and the reader could more easily digest the information.
(a) I would also recommend displaying separate metabolite fit plots for each group, since the current presentation in panel F makes it appear that the MRS data is examined by testing differences between groups across the full spectrum (where the lines diverge), which really isn't the case.
(b) The GABA peak is not visible in the spectrum, and Glutamate and GABA both have multiple peaks that should be shown on the spectrum. This may be best achieved by displaying the individual metabolite sub-spectra below the full spectrum
Thank you for these suggestions. We have split the information into two Figures following the Reviewer’s recommendations.
It is not clear why the 3T structural images were used for segmentation and calculation of tissue fraction if 7T structural images were also acquired (which would presumably have higher resolution).
Generally, T1-weighted images from the 7T scanner exhibit more artifacts than those from the 3T scanner due to higher magnetic field inhomogeneity. These artifacts are especially pronounced in regions near air-tissue interfaces, such as the temporal lobes. Therefore, we chose the 3T structural images for segmentation and tissue fraction calculations and clarified this in the Method section:
“Voxel segmentation was performed on structural images from a 3T scanner, coregistered to 7T structural images in SPM12, as the latter exhibited excessive artifacts and intensity bias in the temporal regions”.
The basis set includes a large number of metabolites (27), including many low-concentration metabolites/compounds (e.g., bHG, bHB, Citrate, Threonine, ethanol) that are typically only included in studies targeting specific metabolites in disease/pathology. Please justify the inclusion of this maximal set of metabolites in the basis set, given that the inclusion of overlapping low-concentration metabolites may influence metabolite measurements of interest (https://doi.org/10.1002/mrm.10246).
There is still no consensus in the MR community on which metabolites should be included in the model of human cerebral 1H-MR spectra. Typically, only major contributors such as NAA, Cr, Cho, Lac, mI, and possibly Glx are evaluated. Some studies also include additional metabolites like Ace, Ala, Asp, GABA, Glc, Gly, sI, NAAG, and Tau. In this study, as in a few others, further metabolites such as PCh, GPC, PCr, GSH, PE, and Thr were introduced and this approach seems suitable for high-field spectra (Hofmann et al., 2002).
Hofmann, L., Slotboom, J., Jung, B., Maloca, P., Boesch, C., & Kreis, R. (2002). Quantitative 1H-magnetic resonance spectroscopy of human brain: Influence of composition and parameterization of the basis set in linear combination model-fitting. Magnetic Resonance in Medicine, 48(3), 440–453. https://doi.org/10.1002/mrm.10246
Please provide a figure indicating the localization of the MRS voxel for a sample subject.
A figure indicating the localization of the MRS voxel for a sample subject was added to the MRS checklist.
It would be helpful to include Table S1 in the main article.
Table S1 from the Supplementary Material has now been added to the main manuscript as Table 1 in the Results section.
Please report descriptive statistics for EEG and MRS measures in Table S1.
We have added a new Table S1 in the Supplementary Material, providing descriptive statistics for EEG and MRS E/I balance measures, presented separately for the dyslexic and control groups.
I recommend avoiding using the terms "direct" and "indirect" to contrast MRS and EEG measures of E/I balance. Both of these measures are imperfect and it is misleading to say that MRS is a "direct" measure of neurotransmitters. There is also ambiguity in what is meant by "direct": in contrast to EEG, MRS does not measure neural activity and does not provide high-resolution temporal information, so in a sense, it is less direct.
Thank you for this suggestion. We have replaced the terms 'direct' and 'indirect' biomarkers with 'MRS' and 'EEG' biomarkers throughout the text.
There are many cases throughout the results in which Bayes and frequentist stats seem to contradict each other in terms of significance and what should be included in the models, especially with regard to the interaction effects (the Bayes factors appear to favor non-significant interactions). I think this is worth considering and describing to offer more clarity for the readers.
We agree that a discussion of the divergent results between Bayesian and frequentist models was missing in the previous version of the manuscript. To provide greater clarity for the readers, we have conducted follow-up Bayesian t-tests in every case where the results indicated the inclusion of non-significant interactions with the effect of group in the model. These additional analyses have been performed for the exponent, offset, as well as for beta bandwidth in the Supplementary Material. We have also added a paragraph addressing these discrepancies in the Discussion:
“Remarkably, in some models, results from Bayesian and frequentist statistics yielded divergent conclusions regarding the inclusion of non-significant effects. This was observed in more complex ANOVA models, whereas no such discrepancies appeared in t-tests or correlations. Given reports of high variability in Bayesian ANOVA estimates across repeated runs of the same analysis (Pfister, 2021), these results should be interpreted with caution. Therefore, following the recommendation to simplify complex models into Bayesian t-tests for more reliable estimates (Pfister, 2021), we conducted follow-up Bayesian t-tests in every case that favored the inclusion of non-significant interactions with the group factor. These analyses provided further evidence for the lack of differences between the dyslexic and control groups. Another source of discrepancy between the two methods may stem from the inclusion of interactions between covariates and within-subject effects in frequentist ANOVA, which were not included in Bayesian ANOVA to adhere to the recommendation for simpler Bayesian models (Pfister, 2021).”
Pfister, R. (2021). Variability of Bayes factor estimates in Bayesian analysis of variance. The Quantitative Methods for Psychology, 17(1), 40-45. doi:10.20982/tqmp.17.1.p040
It would be helpful to indicate whether participants in the DYS group had a history of reading intervention/remediation. In addition to showing that the DYS group performed lower than the CON group on reading assessments as a whole and given their age, was the performance on the reading assessments at an individual level considered for inclusion in the study? (i.e., were participants' persistent poor reading abilities confirmed with the research assessments?)
We were unable to assess individual reading skills due to the lack of standardized diagnostic norms for adult dyslexia in Poland. Therefore, participants in the dyslexic group were recruited based on a previous clinical diagnosis of dyslexia, and reading and reading-related tasks were used for group-level comparisons only. This information has been added to the Methods section:
“Since there are no standardized diagnostic norms for dyslexia in adults in Poland, individuals were assigned to the dyslexic group based on a past diagnosis of dyslexia.”
Unfortunately, we did not collect information about participants' history of reading intervention or remediation. In this context, we acknowledge that including a sample of adult participants is a potential limitation of our study, however, this was already mentioned in the Discussion.
Regarding the fMRI task, please indicate whether the participants whose threshold and/or contrast was changed for localization were from the DYS or CON group.
This information is now added to the Method section:
“For 6 participants (DYS n = 2, CON n = 4), the threshold was lowered to p < .05 uncorrected, while for another 6 participants (DYS n = 3, CON n = 3) the contrast from the auditory run was changed to auditory words versus fixation cross due to a lack of activation for other contrasts.”
Reviewer #2 (Public Review):
Summary:
This study utilized two complementary techniques (EEG and 7T MRI/MRS) to directly test a theory of dyslexia: the neural noise hypothesis. The authors report finding no evidence to support an excitatory/inhibitory balance, as quantified by beta in EEG and Glutamate/GABA ratio in MRS. This is important work and speaks to one potential mechanism by which increased neural noise may occur in dyslexia.
Strengths:
This is a well-conceived study with in-depth analyses and publicly available data for independent review. The authors provide transparency with their statistics and display the raw data points along with the averages in figures for review and interpretation. The data suggest that an E/I balance issue may not underlie deficits in dyslexia and is a meaningful and needed test of a possible mechanism for increased neural noise.
Weaknesses:
The researchers did not include a visual print task in the EEG task, which limits analysis of reading-specific regions such as the visual word form area, which is a commonly hypoactivated region in dyslexia. This region is a common one of interest in dyslexia, yet the researchers measured the I/E balance in only one region of interest, specific to the language network.
We agree with the Reviewer that including different tasks for the EEG biomarkers assessment would be valuable. However, this limitation was already addressed in the Discussion:
“Importantly, our study focused on adolescents and young adults, and the EEG recordings were conducted during rest and a spoken language task. These factors may limit the generalizability of our results. Future research should include younger populations and incorporate a broader array of tasks, such as reading and phonological processing, to provide a more comprehensive evaluation of the E/I balance hypothesis.”
Further, this work does not consider prior studies reporting neural inconsistency; a potential consequence of increased neural noise, which has been reported in several studies and linked with candidate-dyslexia gene variants (e.g., Centanni et al., 2018, 2022; Hornickel & Kraus, 2013; Neef et al., 2017). While E/I imbalance may not be a cause of increased neural noise, other potential mechanisms remain and should be discussed.
Thank you for referring us to other works reporting neural variability in dyslexia. We agree that a broader context regarding sources of reduced neural synchronization, beyond E/I imbalance, was missing in the previous version of the manuscript. We have now included these references in the Discussion:
“Furthermore, although our results do not support the idea of E/I balance alterations as a source of neural noise in dyslexia, they do not preclude other mechanisms leading to less synchronous neural firing posited by the hypothesis. In this context, there is evidence showing increased trial-to-trial inconsistency of neural responses in individuals with dyslexia (Centanni et al., 2022) or poor readers (Hornickel and Kraus, 2013) and its associations with specific dyslexia risk genes (Centanni et al., 2018; Neef et al., 2017). At the same time, the observed trial-to-trial inconsistency was either present only in a subset of participants (Centanni et al., 2018), limited to some experimental conditions (Centanni et al., 2022), or specific brain regions – e.g., brainstem in Hornickel and Kraus (2013), left auditory cortex in Centanni et al. (2018), or left supramarginal gyrus in Centanni et al. (2022).”
A better description of the exponent and offset components is needed at the beginning of the results, given that the methods are presented in detail at the end. I also do not see a clear description of these components in the methods.
A description of the aperiodic components is now included in the Results:
“In the initial step of the analysis, we analyzed the aperiodic (exponent and offset) components of the EEG spectrum. The exponent reflects the steepness of the EEG power spectrum, with a higher exponent indicating a steeper signal; while the offset represents a uniform shift in power across frequencies, with a higher offset indicating greater power across the entire EEG spectrum (Donoghue et al., 2020).”
as well as in the Materials and Methods:
“Two broadband aperiodic parameters were extracted: the exponent, which quantifies the steepness of the EEG power spectrum, and the offset, which indicates signal’s power across the entire frequency spectrum.”
Reviewer #3 (Public Review):
Summary:
This study by Glica and colleagues utilized EEG (i.e., Beta power, Gamma power, and aperiodic activity) and 7T MRS (i.e., MRS IE ratio, IE balance) to reevaluate the neural noise hypothesis in Dyslexia. Supported by Bayesian statistics, their results show solid 'no evidence' of EI balance differences between groups, challenging the neural noise hypothesis. The work will be of broad interest to neuroscientists, and educational and clinical psychologists.
Strengths:
Combining EEG and 7T MRS, this study utilized both the indirect (i.e., Beta power, Gamma power, and aperiodic activity) and direct (i.e., MRS IE ratio, IE balance) measures to reevaluate the neural noise hypothesis in Dyslexia.
Weaknesses:
The authors may need to provide more data to assess the quality of the MRS data.
We have addressed the following specific recommendations of the Reviewer providing more data about the quality of the MRS data.
The authors may need to explain how the number of subjects is determined in the MRS section.
We have clarified the MRS sample description in the Results section:
“Due to financial and logistical constraints, 59 out of the 120 recruited subjects, selected progressively as the study unfolded, were examined with MRS. Subjects were matched by age and sex between the dyslexic and control groups. Due to technical issues and to prevent delays and discomfort for the participants, we collected 54 complete sessions. Additionally, four datasets were excluded based on our quality control criteria, and three GABA+ estimates exceeded the selected CRLB threshold. Ultimately, we report 50 estimates for Glu (21 participants with dyslexia) and 47 for GABA+ and Glu/GABA+ ratios (20 participants with dyslexia).”
Is there a reason why theta and gamma peaks were not observed in the majority of participants? What are the possible reasons that likely caused the discrepancy between this study and previously reported relevant studies?
We have now added a discussion about the absence of oscillatory peaks in the theta and gamma bands to the Discussion section:
“We could not perform analyses for the gamma oscillations since in the majority of participants the gamma peak was not detected above the aperiodic component. Due to the 1/f properties of the EEG spectrum, both aperiodic and periodic components should be disentangled to analyze ‘true’ gamma oscillations; however, this approach is not typically recognized in electrophysiology research (Hudson and Jones, 2022). Indeed, previous studies that analyzed gamma activity in dyslexia (Babiloni et al., 2012; Lasnick et al., 2023; Rufener and Zaehle, 2021) did not separate the background aperiodic activity. For the same reason, we could not analyze results for the theta band, which often does not meet the criteria for an oscillatory component manifested as a peak in the power spectrum (Klimesch, 1999). Moreover, results from a study investigating developmental changes in both periodic and aperiodic components suggest that theta oscillations in older participants are mostly observed in frontal midline electrodes (Cellier et al., 2021), which were not analyzed in the current study.”
Hudson, M. R., & Jones, N. C. (2022). Deciphering the code: Identifying true gamma neural oscillations. Experimental Neurology, 357, 114205. https://doi.org/10.1016/j.expneurol.2022.114205
Klimesch, W. (1999). EEG alpha and theta oscillations reflect cognitive and memory performance: A review and analysis. Brain Research Reviews, 29(2-3), 169-195. https://doi.org/10.1016/S0165-0173(98)00056-3
Based on Figure 1F, the quality of the MRS data may be contaminated by the lipid signal, especially for the DYS group. To better evaluate the MRS data, especially the GABA measurements, the authors need to show:
(a) the placement of the MRS voxel on the anatomical images;
Averaged MRS voxel placement was already presented in Figure 1 (now Figure 2) in the manuscript. Now, we have also added exemplary single-subject images to the MRS checklist in the Supplement.
(b) Glu and GABA model functions
We have now provided more meaningful Glu and GABA indications in Figure 2.
(c) CRLB for GABA
We have added respective estimates to the Supplement:
%CRLB of Glu: mean 2.96, SD = 0.79
%CRLB of GABA: mean 10.59, SD = 2.76
%CRLB of NAA: 1.76 SD = 0.46
Further, the authors added voxel's gray matter volume as a covariate when performing separate ANCOVAs. The authors may need to use alpha correction or 1-fCSF correction to corroborate these results.
We chose to use the ratio of Glu and GABA to total creatine (tCr), as this remains a common practice in MRS studies at 7T (e.g., Nandi et al., 2022; Smith et al., 2021). This decision was also influenced by previous dyslexia studies (Del Tufo et al., 2018; Pugh et al., 2014) and is now clarified in the Results and Methods sections.
Regarding alpha correction, a recent paper (García-Pérez et al., 2023) recommends: 'In general, avoid corrections for multiple testing if statistical claims are to be made for each individual test, in the absence of an omnibus null hypothesis.' Since we report null findings, further alpha correction would not significantly impact the results.
García-Pérez, M. A. (2023). Use and misuse of corrections for multiple testing. Methods in Psychology, 8, 100120. https://doi.org/10.1016/j.metip.2023.100120
