Introduction This study aimed to investigate alterations in whole-brain functional connectivity after a concussion using graph-theory analysis from global and local perspectives and explore the association between changes in the functional network properties and cognitive performance. Methods Individuals with mild traumatic brain injury (mTBI, n = 29) within a month after injury, and age- and sex-matched healthy controls (n = 29) were included. Graph-theory measures on functional connectivity assessed using resting state functional magnetic resonance imaging data were acquired from each participant. These included betweenness centrality, strength, clustering coefficient, local efficiency, and global efficiency. Multi-domain cognitive functions were correlated with the graph-theory measures. Results In comparison to the controls, the mTBI group showed preserved network characteristics at a global level. However, in the local network, we observed decreased betweenness centrality, clustering coefficient, and local efficiency in several brain areas, including the fronto-parietal attention network. Network strength at the local level showed mixed-results in different areas. The betweenness centrality of the right parahippocampus showed a significant positive correlation with the cognitive scores of the verbal learning test only in the mTBI group. Conclusion The intrinsic functional connectivity after mTBI is preserved globally, but is suboptimally organized locally in several areas. This possibly reflects the neurophysiological sequelae of a concussion. The present results may imply that the network property could be used as a potential indicator for clinical outcomes after mTBI.
An exploratory study on functional connectivity after mild traumatic brain injury: Preserved global but altered local organization
Review badges
Web of Science (Free Access)
Abstract
- Contribute
- pre-publication peer review (FINAL ROUND)
Decision Letter
2022/07/2020-Jul-2022
Dear Dr. Oh:
It is a pleasure to accept your manuscript entitled "An exploratory study on functional connectivity after mild traumatic brain injury: preserved global- but altered local organization" in its current form for publication in Brain and Behavior. If there were further comments from the reviewer(s) who read your manuscript, they will be included at the foot of this letter.
FIRST LOOK: Your manuscript is accepted. HOWEVER: Please note that your manuscript will undergo an integrity check, that includes the images. Publication will only proceed on the condition that all final files comply with the journal integrity checks. In the event that any file does not comply with our integrity checks, the journal reserves the right to rescind this decision, or, alternatively, you may be contacted to resolve any concerns raised by these checks.
Please note although the manuscript is accepted the files will now be checked to ensure that everything is ready for publication, and you may be contacted if final versions of files for publication are required.
The final version of your article cannot be published until the publisher has received the appropriate signed license agreement. Once your article has been received by Wiley for production the corresponding author will receive an email from Wiley’s Author Services system which will ask them to log in and will present them with the appropriate license for completion.
Payment of your Open Access Article Publication Charge (APC):
All articles published in Brain and Behavior are fully open access: immediately and freely available to read, download and share. Brain and Behavior charges an article publication charge (APC).
Before we can publish your article, your payment must be completed. The corresponding author for this manuscript will have already received a quote email shortly after original submission with the estimated Article Publication Charge; please let us know if this has not been received. Once your accepted paper is in production, the corresponding author will receive an e-mail inviting them to register with or log in to Wiley Author Services (www.wileyauthors.com) where the publication fee can be paid by credit card, or an invoice or proforma can be requested. The option to pay via credit card and claim reimbursement from your institution may help to avoid delays with payment processing.
If your paper contains Supporting Information:
Materials submitted as Supporting Information are authorized for publication alongside the online version of the accepted paper. No further Supporting Information can be submitted after acceptance. It is the responsibility of the authors to supply any necessary permissions to the editorial office.Thank you for your fine contribution. On behalf of the Editors of Brain and Behavior, we look forward to your continued contributions to the Journal.
Sincerely,
Dr. Amanda Bischoff-Grethe
Editor in Chief, Brain and Behavior
agrethe@ucsd.eduAssociate Editor Comments to Author:
Reviewer(s)' Comments to Author:
Reviewer: 2
Comments to the Author
The manuscript has been improved by revision.P.S.- Would you be interested in publishing your proven experimental method as a detailed step-by-step protocol? Current Protocols in Neuroscience welcomes proposals from prospective authors to disseminate their experimental methodology in the rapidly evolving field of neuroscience. Please submit your proposal here: https://currentprotocols.onlinelibrary.wiley.com/hub/submitaproposal
P.S. – You can help your research get the attention it deserves! Wiley Editing Services offers professional video abstract and infographic creation to help you promote your research at www.wileyauthors.com/eeo/promotion. And, check out Wiley’s free Promotion Guide for best-practice recommendations for promoting your work at www.wileyauthors.com/eeo/guide.
This journal accepts artwork submissions for Cover Images. This is an optional service you can use to help increase article exposure and showcase your research. For more information, including artwork guidelines, pricing, and submission details, please visit the Journal Cover Image page at www.wileyauthors.com/eeo/covers.
Decision letter by
Cite this decision letter
Reviewer report
2022/07/12The manuscript has been improved by revision.
Reviewed by
Cite this review
Author Response
2022/06/27Response to reviewers’ comments
We thank the editor and the reviewers of Brain and Behavior for taking the time to review our manuscript and for the valuable comments. We have made corrections and clarifications to the manuscript in response to the reviewers’ comments as highlighted by our point-to-point response. In the main manuscript, the edited areas have been highlighted in blue for easy reference.
Associate Editor Comments to Author
Comment 1: In the introduction, the authors note that some graph theory analyses of mTBI have already been done, but they don't present the findings, which makes it hard to interpret the comment regarding the functional network still being unclear.
Response: We appreciate the editor’s comment. We believe the editor implied the third paragraph in the original manuscript, “Despite the aforementioned studies, it is still unclear how mTBI affects the functional brain network, considering its high inter-individual variability and subtle nature”. The meaning we intended to convey was that the results of the aforementioned studies showed altered graph-theoretical measures after mTBI, most of which were reduced parameters, regarding network centrality, segregation, and efficiency. In the revised manuscript, we added sentences describing the specific findings of the previous studies in accordance with your comment [page 7 line 22 – page 8 line 8].“Previous studies assessing several graph-theory measures representing network centrality, segregation, and efficiency have shown that mTBI changes the brain in sub-optimal ways during the acute and subacute phases following mTBI (4, 12-20). Individuals with mTBI showed reduced degree and betweenness centrality in the frontal and occipital areas compared to controls (4, 18). The participation coefficient, which reflects the between-module connectivity, was also reduced after mTBI in these areas (12). Post-concussive symptoms, anxiety, depression, and tau aggregation have been found to be correlated with brain network properties (14, 17, 18, 21). The mTBI group with PCS (>6 months) had a significantly reduced mean shortest path length globally compared to controls, but mTBI without PCS did not show such changes (13). There was a positive correlation between the performance of executive function and network efficiency of the dorsolateral superior frontal and anterior cingulate regions in young adults with mTBI (19).”
Comment 2: The rationale for the study is lacking (e.g., is there too much variability in the outcomes, poor characterization of participants [this is partly noted as individual variability]).
Response: Thank you for your critical comment. We have rewritten the introduction section in the revised manuscript to clearly deliver the rationale of this study [page 7 line 2 – page 9 line 4]. The aim of this study was to explore alterations in the topological functional brain network after a concussion using graph-theory analysis and to identify the association between the topological brain changes and cognitive performance after mTBI. To date, topological changes of the functional brain network during acute and subacute phases following mTBI have been investigated mostly in individuals who have experienced blast- or sports-related mTBI, who have different characteristics of injury compared with civilians with mTBI. In addition, there were few studies investigating the association between functional network properties and cognitive performance in civilian mTBI. Therefore, we believe that further studies are needed to enrich the understanding of topological functional brain changes associated with cognitive function in civilian mTBI.Comment 3: The authors also note cognitive defects but don't describe in which domains.
Response: We appreciate the editor’s comment. In the revised manuscript, we described cognitive deficits by domain to clearly show which domains were altered and associated with functional network properties within a month after mTBI.
‘Cognitive deficits’ or ‘cognitive function’ was still used in the revised manuscript for general explanations.Comment 4: Scan parameters appear to be missing. For rsfMRI, how long was the scan, and were eyes open or closed? If open, was there a fixation?
Response: Due to the word-count limitation, we included the information on scan parameters in the Supporting Information. In detail, rsfMRI data were obtained using interleaved slice acquisition with 116 volumes consisting of 35 consecutive images (image matrix 128 mm×128 mm, voxel size 1.9 mm×1.9 mm×3.5 mm, repetition time 3500 ms, echo time 30 ms, field of view 240 mm, flip angle 90°). Participants were instructed to stay awake with closed eyes and motionless without thinking of anything.Comment 5: Results say other than delayed recall being similar, it says mTBI group showed poorer cognitive performance but this isn't broken down/its treated as a global measure. Finally, there didn't appear to be any descriptive statistics comparing the mTBI and HC groups.
Response: Thank you for your insightful comment. The individuals with mTBI and the control groups performed cognitive function tests for memory, executive function, and attention. The results of group comparison in these multiple cognitive domains were described in Table 2, which was newly generated in the revised manuscript. The results of the descriptive statistics comparing the mTBI and HC groups were also listed in Table 2.Reviewer(s)' Comments to Author:
Reviewer #1
Comments to the Author: The main goal of this study was to examine resting-state functional connectivity (RSFC) after mild traumatic brain injury (mTBI). The authors implement graph theory to examine changes in functional brain network organization after mTBI and how these changes may impact cognition in individuals with mTBI (n=29) compared to healthy controls (n=29). Although this study presents interesting findings, there are several moderate concerns that dampen enthusiasm for this submission.Comment 1: Background and Hypotheses: There is mature literature using rsfMRI and mTBI and the authors are encouraged to include in the manuscript how these findings advance the literature either through replication or extension.
Response: We appreciate the reviewer’s kind advice. In the revised manuscript, we have provided more descriptions of the previous findings, considering the similar time post injury with the present study (i.e., acute and subacute phases following mTBI). We have included conventional connectivity studies without investigating topological brain changes, showing increased functional connectivity in the mTBI group (6, 7). Extending to these findings, graph theory has been applied to investigate topological functional brain changes, and sub-optimal changes were observed in the mTBI group, mostly in the sports-related and military-based mTBI. It has been known that there were different characteristics of injury between different mTBI populations (22), which might lead to distinctive topological changes in the brain. Therefore, our study could be an extension of the previous findings to enrich the understanding of topological functional brain changes in adult mTBI by focusing on the civilian population and exploring the association with cognitive function. [page 7 lines 9-20].“Particulary, there is substantial literature using resting-state functional magnetic resonance imaging (rsfMRI) to investigate functional brain network changes in mTBI (5). The results from these rsfMRI studies have indicated that individuals with mTBI have stronger functional connectivity than controls during the acute and subacute phases of injury (6, 7), although there was also a mixture of increased and decreased connectivity findings (5). The increased connectivity was indicative of good memory, attention, and executive function, as well as clinically milder symptom severity in the mTBI group (5). In contrast, increased connectivity within a month after injury was also reported to be associated with persistent post-concussion symptom (PCS) (8).
Based on these findings, investigating topological characteristics of the brain can provide a new understanding of the complex nature of mTBI by quantitatively describing the altered characteristics of the brain in a mathematical framework (9, 10).”Comment 2: The background also requires some clarification with respect to sample and goals. The authors mention that previous papers have examined RSFC and its relationship to cognition in mTBI, but they cite several blast-related mTBI studies. This study’s sample consisted of participants with civilian mTBI so focus on blast-related TBI and PTSD is unclear. Similarly, the discussion of repetitive injury does not appear to fit this sample, where individuals presumably sustained a single injury.
Response: Thank you for your valuable comment. As the reviewer pointed out, the characteristics of civilian mTBI, blast-related TBI, and PTSD differ. Nevertheless, there has been a very limited number of studies investigating the association between graph-theory-based RSFC and its relationship to cognition in civilian mTBI. We intended to convey the necessity of the civilian mTBI study investigating alterations in the topological brain network. To clearly convey our intention, we have rewritten a paragraph in the introduction section as follows [page 8 lines 9-18].“Research on mTBI has been conducted on diverse populations, such as individuals with sports-related mTBI and military-based mTBI, as well as civilians. It has been known that there are different characteristics of injury between these populations (22), which might lead to different changes in the brain. To date, topological changes of the functional brain network during the acute and subacute phases of mTBI were investigated mostly in individuals who experienced blast- or sports-related mTBI (4, 12-16, 20). In addition, there is preliminary evidence of the association between functional network properties and cognition after a single civilian mTBI unlike that of blast- or sports-related mTBI (23). Therefore, further studies are necessary to enrich the understanding of topological functional brain changes associated with cognitive function in civilian mTBI.”
Comment 3: Hypotheses were unfortunately quite general, making it possible to support a range of findings. H1: global less significant than local, H2: “altered organization of the functional brain network may be associated with cognitive performance”. The second is particularly problematic in that it does not constrain the analyses to anything specific which casts some doubt on the reliability of the results.
Response: Thank you for your comment. Since there were few studies investigating the changes in the graph-properties of the resting-state functional connectivity after concussion in civilian individuals, an exploratory investigation should be necessary to generate a more elaborate research question for future study. Generally, individuals with mTBI still perform all major, outwardly observable functions in a normal way, but within the subdomains of cognitive, emotional, and behavioral functioning, there are subtle differences. Based on these behavioral characteristics, we hypothesized that there was altered organization of the functional network, which was related to the impaired cognitive function, but the overall network properties were maintained in the individuals with mTBI. We agree that for H2, cognitive performance was quite a large terminology, so we have rewritten the sentence as “altered organization of the functional brain network may be associated with impaired memory, executive function, and attention after mTBI.” [page 9 lines 3-4] As an exploratory investigation, we thoroughly analyzed all nodes in the brain to explore the alterations of the whole-brain functional network based on the graph-theoretical approach, without constraining the analyses to specific regions of interest.Comment 4: Sample: The authors state that the mTBI group has significantly higher BDI scores than the control group and depression/psychiatric symptoms are quite common post mTBI. The authors also state that they created a positive correlation matrix that accounted for age and sex, but they did not mention accounting for BDI scores. Depression (and other mood disorders) impact brain network properties, and the authors mention this in their introduction. Therefore, the authors should address this issue to determine if symptoms of depression influence their findings.
Response: We appreciate the reviewer’s kind advice. Since depression is one of the representative symptoms of concussion, constructing a brain network accounting for depression may also remove the relevant signals contributed by the effect of a concussion, which is of research interest. Analyzing the effect of depression on the functional brain network after mTBI using sub-group analysis should be valuable, but it was beyond the scope of this research (sub-group analysis was inappropriate in this study because of the small number of participants and the exploratory nature of this study). In addition, not only depression but also other kinds of symptoms after a concussion, such as dizziness or sleep problems, should be important factors that may influence functional network changes. In the limitations section of the revised manuscript, we have described that there is a need to unveil the effect of important symptoms after mTBI, which may influence brain network properties, and explained why we did not control the BDI scores. [page 20 lines 2-5]“Future studies are also warranted to unveil the effect of important symptoms after mTBI, such as depression or sleep problems, on brain network properties using subgroup analysis. Constructing a brain network by controlling the BDI scores was inappropriate, as it may remove the relevant brain signals contributed by the effect of a concussion.”
Comment 5: While assessment of mTBI during the first month post injury is a clear strength of the paper, n=29 remains small and the findings are tentative.
Response: Thank you for your comment. We acknowledge that the insufficient sample of this study should be a major concern and limitation of this work. Therefore, we added “an exploratory work” to the title of this manuscript and included the small sample size as one of the limitations of this study in the Limitations section [page 19 lines 13-24]. We believe that future studies could be more elaborate based on the results of this exploratory investigation.Comment 6: Methods: In this study, the preprocessed images were smoothed using a Gaussian kernel (sigma=2.55). However, spatial smoothing should be avoided when evaluating functional brain networks, especially when using the ROI approach (see Alakörkkö et al., 2017).
Response: Thank you for your insightful comment. We thoroughly reviewed the paper the reviewer mentioned (27) and understood the complex effect of spatial smoothing on the structure and properties of functional brain networks, in particular, ROI-based studies. Nevertheless, we considered that spatial smoothing is also advantageous in this study for several reasons. First, since we used an independent component analysis-based strategy for Automatic Removal of Motion Artifacts (ICA-AROMA), spatial smoothing of the data before applying ICA-AROMA can increase the ability to separate artifacts, including motion-related noise (26). Second, an increase in signal-to-noise ratio due to spatial smoothing is important to explore alterations in functional brain networks, particularly in the group with subtle characteristics and using a small number of participants. Third, because spatial smoothing is a common procedure in brain imaging data, comparing the findings with those from other studies can be helpful. Last, the size of the Gaussian kernel (sigma=2.55, FWHM=6 mm) was not large enough to introduce significant distortion of the signal. Therefore, we added the following sentences to describe the possible effect of smoothing on the results in the revised manuscript. [page 11 lines 8-11]“Spatial smoothing was applied to the data, although it may alter the structure of the functional brain networks, particularly those constructed by the regions of interest (ROIs)-based approach (27). It can increase the signal-to-noise ratio and help to identify motion-related artifacts when applying ICA-AROMA.”
Comment 7: The AAL atlas is a bit outdated and has been shown to lack sensitivity to mapping connectivity to behavior. Rationale for choosing this structural atlas over more recent functional atlases (e.g., Power264, Gordon, Schaeffer atlases) should be provided.
Response: Thank you for your valuable comment. As the reviewer mentioned, the more recently developed functional atlases can provide finer nodal scales than the AAL atlas. It would be great to use such functional atlases in future studies, but in this work, we chose to use the AAL atlas due to the following reasons. First, the signal-to-noise ratio (SNR) of the time series at each node is affected by node size; a large node size increases SNR (Zaleski et al., 2010). Because we analyzed a small number of participants, maintaining the SNR as high as possible is warranted. Second, taking a finer resolution of the nodes leads to a large number of statistical tests being performed, which may reduce statistical power and increase the false-positive error rate. As an exploratory study, we aimed to investigate alterations in whole-brain functional connectivity after a concussion. This way, we were obliged to choose the rather coarse atlas instead of the fine-grained one to minimize the spurious findings. In the revised manuscript, we included the rationale for choosing the AAL atlas to give the reader a clear understanding of the choice of node made in this study.“Specifically, the way of defining the node and edge in functional connectivity may change the results of group comparison (52). In this study, we choose the AAL atlas instead of the recently developed fine-grained functional atlases (53, 54) to define the node. It is because we want to maintain the signal-to-noise ratio as high as possible, and preclude a large number of statistical tests.” [page 20 lines 8-12]
Comment 8: Also, with respect to creating the functional matrices, positive correlation matrices were retained after excluding negative values. More specifically, how did the authors handle the negative correlations in their functional connectivity matrices (absolute value, etc.)? Additionally, were the matrices Fisher r-to-z transformed? This should be addressed in the paper.
Response: According to the reviewer’s comment, we have clearly addressed how we handle the data about the positive and negative correlation matrix. A positive correlation matrix was constructed after age and sex were removed from the data using the general linear model, and negative correlation values were set to zero. This is because of general comparison across the previous findings due to the little consensus for handling or interpreting negative correlation values in the topological functional brain network (52). Fisher r-to-z transformation was not applied in this study because we did not compare the strength of the pairwise correlations between the groups but compared the network properties using graph-theory measures.“A positive correlation matrix was constructed after age and sex were removed from the data using the general linear model, after removing negative correlation values. Fisher r-to-z transformation was not applied in this study.” [page 11 lines 22-24]
Comment 9: Statistics: The results state that the betweenness centrality in the right parahippocampus of the brain was positively correlated with scores on the verbal learning test in the mTBI group, but this positive correlation was not present in the healthy controls. However, more tests are needed to reveal if the Pearson’s correlation values are significantly different between the mTBI and healthy control group when examining the relationship between cognition and local brain network organization. The authors used a non-parametric permutation test to determine the significance of the graph theory measures between the mTBI group and the healthy control group. A similar approach should be utilized to examine if local brain network organization and its correlations with cognition differ between the two groups.
Response: We appreciate the reviewer’s insightful comment. In the revised manuscript, we have compared the correlation coefficient between the groups given the prior findings (i.e., the betweenness centrality in the right parahippocampus was positively correlated with scores on the verbal learning test A1 in the mTBI group, but this positive correlation was not present in the healthy controls) using a non-parametric permutation test. In detail, we permuted betweenness centrality data across all individuals of the right parahippocampus and divided the data into the pseudo-mTBI and the pseudo-control groups. Performance of verbal learning test A1 was not randomized in the permutation test. The correlation coefficient was estimated repetitively 10,000 times using those data in each group. Thereafter, the correlation coefficient was transformed to a z-score by fisher z transformation, and a z-z comparison was conducted using the following equation.Zdiffpseudo = (Zpseudo-mTBI – Zpseudo-control) / sqrt(1/(NmTBI -3)+1/(Ncontrol-3))
In the same manner, the Zdiffreal was estimated and compared with the Zdiffpseudo. The following figure was the distribution of Zdiffpseudo of the correlation between the betweenness centrality of the right parahippocampus and performance of verbal learning test A1.
In this null distribution, 3.3426 was the value at p=0.001 (two-sided). Since the Zdiffreal was 3.5173, we can conclude that there was a significant group difference between the correlation of the betweenness centrality of the right parahippocampus and the performance of verbal learning test A1. We described these procedures in the [Materials and methods] and [Results] sections in the revised manuscript.
[Materials and methods section, page 14 lines 6-15]
“If there was a significant correlation, the non-parametric random permutation test was performed to determine that the relationship between cognitive performance and local network properties was significantly different between the groups. In detail, the local network properties of node i were randomly assigned to the pseudo mTBI group and the pseudo control group. After estimating the pseudo correlation coefficient, the value was transformed to the z value. Z comparison was performed repetitively 10,000 times to generate null distribution using the following equation.ZdiffPseudo = (Zpseudo-mTBI – Zpseudo-control) / sqrt(1/(NmTBI -3)+1/(Ncontrol-3))
The original value of z comparison was compared to the null distribution to obtain statistical significance.”
[Results section, page 16 lines 3-5].
“The relationship was significantly different between the groups using the non-parametric random permutation test (p<0.001).”Comment 10: It would also be nice to have a visual of the local brain regions that are associated with cognitive performance in the mTBI group.
Response: Thank you for your great suggestion. In the revised Figure 3, the identified brain areas, which showed a significant association with cognitive performance, are visualized.Comment 11: Minor Points/ Questions: The authors regress out signal from white matter and cerebrospinal fluid, but do not include global signal regression (GSR). This is a controversial step that is hotly debated and the reasoning for this decision should be addressed.
Response: We appreciated the reviewer’s great suggestion. As the reviewer mentioned, including GSR in the brain preprocessing step is controversial and may affect the network topology significantly (28, 29). We have included a sentence as to why the GCR was excluded from the preprocessing step as follows.“Global signal regression was not applied, given concerns about bias to negative correlations (28, 29), which may lead to substantial alterations in graph-theoretical measures.” [page 11 lines 11-13]
Comment 12: Given the between-group differences in motion -- they should be presented in a Table and perhaps included in analyses to determine affects motion had on connectivity.
Response: We appreciate the reviewer’s kind advice. To control the subtle motion effect, each subject’s framewise displacement (FD) was calculated. In the Supporting Information, we described the total proportion of outlier volumes based on the FD. We generated a new table (Table 2) including between-group differences in FD.
As described in the Supporting Information, there were few numbers of volumes exceeding the threshold of FD. In addition, motion-related artifacts were identified and removed using ICA-AROMA, which automatically detects motion-related temporal and spatial components using independent component analysis.Comment 13: The authors use a range of proportional thresholds (8-26%) that allow for examining network topology but in case-control studies are problematic (see Hallquist & Hillary, 2018). While the argument is that other approaches lead to “different number of edges”, it is this possibility (differences in degree and strength) that are of critical interest after brain injury.
Response: We appreciate the reviewer for this insightful comment. We have used two different approaches to threshold the correlation matrix: primarily, the single-density threshold, and additionally, the range of sparsity threshold (8–26%). Both approaches used proportional thresholds which indicate that numbers of edges across individuals are the same, i.e., a different threshold value was applied for each participant. It allows the controlled comparison of network structure and properties among different groups (Kim et al., 2015). However, as the reviewer mentioned, the proportional threshold could be problematic when clinical groups may have changes in the number and strength of functional connections (52). Therefore, we additionally analyzed whether the individuals with mTBI and the controls differed in their mean functional connectivity. There was no significant difference in mean functional connectivity; correlation matrix (p=0.91), positive correlation matrix (p=0.24), and negative correlation matrix (p=0.28). The subtle clinical characteristics of mTBI may affect the non-significant difference in the mean functional connectivity compared to controls. Nevertheless, we agreed with the reviewer’s concern and included sentences describing the possible effect of proportional thresholding in the limitation section of the revised manuscript. [page 20 lines 12-17]“In addition, we applied single-density threshold and range-of-density threshold to threshold edge strength. This proportional thresholding could be problematic when clinical groups may have changes in the number and strength of functional connections (52). Nevertheless, there was no significant difference in mean functional connectivity between the individuals with mTBI and the controls.”
References (included in this response letter, but not in the revised manuscript)
- Zalesky A, Fornito A, Harding IH, Cocchi L, Yücel M, Pantelis C, et al. Whole-brain anatomical networks: does the choice of nodes matter? Neuroimage. 2010;50(3):970-83.
- Kim H, Yoo K, Na DL, Seo SW, Jeong J, Jeong Y. Non-monotonic reorganization of brain networks with Alzheimer's disease progression. Front Aging Neurosci. 2015;7:111.Reviewer #2
Comments to the Author: This manuscript describes functional connectivity outcomes after mild traumatic brain injury in male participants. The major findings include a change in graph theory metrics within local networks. Limitations include a small sample size that is limited to males, lack of information on education level of participants. Comments and concerns are listed below:Comment 1: Key point 3 states that fc can be used as an indicator of clinical outcomes, but this was not determined in the current study.
Response: Thank you for your comment. “Key point 3” states that “The properties of functional connectivity can be used as an indicator for clinical outcomes after mTBI.” After thoughtful consideration, we believe that the following statement would be more appropriate as the key point statement because of the explorative nature of the present study. Therefore, we replaced the statement with the following sentence.“The clinical utility of network properties should be verified in larger studies.”
Comment 2: Results: Data from the first 2 paragraphs should be included in a table.
Response: According to the reviewer’s recommendation, we have generated new tables (Table 2 and Table 3) summarizing the results of the group comparison between the clinical characteristics, cognitive performance, and global network characteristics.Table 2. Group comparison between individuals with mild traumatic brain injury (mTBI) and healthy controls (HC) with regard to demographic information, clinical characteristics, head movement during scanning, and cognitive performance.
mTBI HC P value
Age (years) 43.3±14.5 42.8±13.8 0.90
Sex (women:men) 15:14 15:14 -BDI 13.9±10.9 3.9±4.0 <0.00
K-MoCA 24.9±4.1 27.1±2.0 0.01
FAB 16.0±2.2 17.2±1.2 0.01Framewise displacement 0.09±0.06 0.08±0.03 0.56
Cognitive test,†
Card sorting test (executive function) 48.4±20.1 57.6±14.6 0.03
Digit span test
Forward 48.2±13.4 59.4±16.4 0.00
Backward 49.8±12.6 58.1±12.2 0.01
Verbal learning test (auditory working memory)
A1 49.5±9.2 56.4±9.1 0.00
A5 52.0±15.0 67.4±10.9 <0.00
Delayed recall 47.0±16.8 59.6±15.6 0.00
A1~A5 (sum) 41.4±9.8 54.1±11.3 <0.00
Visual learning test (visual working memory)
A1 57.5±11.7 60.5±7.8 0.12
A5 66.8±8.3 68.8±7.3 0.17
Delayed recall 64.6±7.4 66.9±7.9 0.14
A1~A5 (sum) 58.6±7.3 61.8±7.6 0.06
Word color test (attention)
Word (black) 48.1±11.1 57.6±11.1 0.00
Color only 40.7±10.5 48.7±12.4 0.01
Color word 41.2±10.4 48.7±10.9 0.01
Word of color word 42.4±11.0 51.7±11.5 0.00
Color of color word 39.4±13.6 47.6±13.6 0.01
Abbreviations: mTBI, mild traumatic brain injury; HC, healthy controls; BDI, Beck Depression Inventory; K-MoCA, Korean-Montreal Cognitive Assessment; FAB, Frontal Assessment Battery; A1, recalled in the first trial; A5, recalled in the fifth trial; A1~A5, recalled from the first to the fifth trials.
Data are the mean±standard deviation.
One participant with mTBI was excluded from the group comparison of cognitive performance tests because of poor compliance during the tasks.
†Cognitive performance was compared between the groups using a one-sided two-sample t-test.Table 3. Mean value of global network characteristics for individuals with mild traumatic brain injury (mTBI) and healthy controls (HC).
mTBI HC P value
Global efficiency 0.28±0.02 0.28±0.02 0.16
Betweenness centrality 0.01±0.00 0.01±0.00 0.12
Strength 13.78±1.19 14.21±1.97 0.23
Clustering coefficient 0.29±0.04 0.31±0.07 0.20
Local efficiency 0.38±0.04 0.39±0.07 0.21
Abbreviations: mTBI, mild traumatic brain injury; HC, healthy controls.
Data are the mean±standard deviation.
Each nodal characteristic was averaged across all nodes. Thereafter, the average and standard deviation were estimated across all individuals in each group.
Global network characteristics were compared between the groups using a one-sided two-sample t-test.Comment 3: “…but the correlation disappeared when the outlier subject was removed.” Describe the criteria for removing an outlier.
Response: Thank you for your constructive feedback. We used the median absolute deviation (MAD) to detect outliers (36). We have added an explicit statement describing the criteria for removing an outlier in the manuscript. [page 16 lines 7-8]“but the correlation disappeared when the outlier subject was removed based on the median absolute deviation (36).”
Comment 4: There is no information on education level of patients or whether patients were matched on education level.
Response: Thank you for your comment. Among the 58 participants (29 patients and 29 controls), information on education level was not acquired in 7 participants. Therefore, we did not include the information on education level in the original manuscript although it could affect the symptoms after TBI (Sumowski et al., 2013). The education level of the patients and controls is listed in the table below (not included in the revised manuscript).Education level Patients Controls
6 years (elementary school) 1 0
9 years (junior high school) 0 0
12 years (senior high school) 4 3
14~16 years (college or university) 17 18
16 years + (graduate school) 2 6
Not available 5 2
Sum 29 29
Including 2 present studentsWe included a paragraph describing the limitation of the study that cognitive reserve factors may influence brain network changes which we did not assess. [page 19 lines 10-12]
“First, although post-injury cognitive ability was assessed in individuals with mTBI, premorbid ability could not be assessed in the present study. Information on the education level of the participants could be used as an indicator of premorbid cognitive function.”
Comment 5: There are no females in the study.
Response: We did include female and male participants. The total number of participants with mTBI was 29 (14 men and 15 women). We have modified the Materials and methods to clearly describe the information of the participants. [page 9 lines 8-11]“Twenty-nine individuals with mTBI (14 men and 15 women, mean age, 43.3±14.5 years), and 29 control subjects (14 men and 15 women, mean age, 42.8±13.8 years) participated in this study.”
Comment 6: Discussion: There is little interpretation of the correlations found in mTBI patients but not controls.
Response: Thank you for your comments. The controls did not show a significant association between the network characteristics and cognitive performance. Therefore, we did not discuss the correlation in the control group. We have added a sentence that clearly describes that there was no such correlation in the control group. [page 16 lines 10-11]“There was no such correlation in the control group.”
References (included this response letter, but not in the revised manuscript)
- Sumowski JF, Chiaravalloti N, Krch D, Paxton J, DeLuca J. Education attenuates the negative impact of traumatic brain injury on cognitive status. Arch Phys Med Rehabil. 2013;94(12):2562-4.Cite this author response
- pre-publication peer review (ROUND 1)
Decision Letter
2022/05/1818-May-2022
We recognise that the impact of the COVID-19 pandemic may affect your ability to return your revised manuscript to us within the requested timeframe. If this is the case, please let us know.
Dear Dr. Oh:
Manuscript ID BRB3-2022-03-0222 entitled "An exploratory study on functional connectivity after mild traumatic brain injury; preserved global- but altered local organization" which you submitted to Brain and Behavior, has been reviewed. Some revisions to your manuscript have been recommended. Therefore, I invite you to respond to the comments appended below and revise your manuscript.
Before submitting your revisions:
Prepare a response to the reviewer comments appended below in point-by-point fashion. In order to expedite the processing of the revised manuscript, please be as specific as possible in your response and indicate the page numbers in the manuscript where you have addressed each comment.
Prepare a revised manuscript (word document), highlighting the changes you’ve made. Save this new document on your computer as you will be asked to upload it during the revision submission process. NOTE: Please be sure to keep in mind reviewer comments and incorporate your responses within the manuscript. There may well be areas where you disagree; for example, you may want to write, "A reviewer suggests that... However, I disagree because...". In any case, please try to address all of the concerns that are raised within the manuscript.
In addition to your revised manuscript with changes highlighted, please also save a “clean” copy where the changes are not marked.
To submit your revised manuscript:
- Log in by clicking on the link below
PLEASE NOTE: This is a two-step process. After clicking on the link, you will be directed to a webpage to confirm.
https://mc.manuscriptcentral.com/brainandbehavior?URL_MASK=a51dcdc9bfdc4472b04111d0ddd5f0ca
OR
Log into https://mc.manuscriptcentral.com/brainandbehavior and click on Author Center. Under author resources, use the button “Click here to submit a revision”. PLEASE DO NOT SUBMIT YOUR REVISIONS AS A NEW MANUSCRIPT.
Follow the on-screen instructions. First you will be asked to provide your “Response to Decision Letter”—this is the response to reviewer comments that you prepared earlier.
Click through the next few screens to verify that all previously provided information is correct.
File Upload: Delete any files that you will be replacing (this includes your old manuscript). Upload your new revised manuscript file with changes highlighted, a “clean” copy of your revised manuscript file, any replacement figures/tables, or any new files. Once this is complete, the list of files in the “My Files” section should ONLY contain the final versions of everything. REMEMBER: figures/tables should be in jpeg, tiff, or eps format. We hope that you will designate one of the figures in your paper to be considered for our online covers and potential publication on our blog.
Review and submit: please be sure to double-check everything carefully so that your manuscript can be processed as quickly as possible.
Deadlines:
Because we are trying to facilitate timely publication of manuscripts submitted to Brain and Behavior,your revised manuscript should be uploaded as soon as possible. If it is not possible for you to submit your revision in 2 months, we may have to consider your paper as a new submission. If you feel that you will be unable to submit your revision within the time allowed please contact me to discuss the possibility of extending the revision time.Wiley Editing Services Available to All Authors
Should you be interested, Wiley Editing Services offers expert help with manuscript, language, and format editing, along with other article preparation services. You can learn more about this service option at www.wileyauthors.com/eeo/preparation. You can also check out Wiley’s collection of free article preparation resources for general guidance about writing and preparing your manuscript at www.wileyauthors.com/eeo/prepresources.Once again, thank you for submitting your manuscript to Brain and Behavior and I look forward to receiving your revision.
Sincerely,
Dr. Amanda Bischoff-Grethe
Editor in Chief, Brain and Behavior
agrethe@ucsd.eduAssociate Editor Comments to Author:
In the introduction, the authors note that some graph theory analyses of mTBI have already been done, but they don't present the findings, which makes it hard to interpret the comment regarding the functional network still being unclear. The rationale for the study is lacking (e.g., is there too much variability in the outcomes, poor characterization of participants [this is partly noted as individual variability]). The authors also note cognitive defects but don't describe in which domains.
Scan parameters appear to be missing. For rsfMRI, how long was the scan, and were eyes open or closed? If open, was there a fixation? Results say other than delayed recall being similar, it says mTBI group showed poorer cognitive performance but this isn't broken down/its treated as a global measure. Finally, there didn't appear to be any descriptive statistics comparing the mTBI and HC groups.Reviewer(s)' Comments to Author:
Reviewer: 1
Comments to the Author
The main goal of this study was to examine resting-state functional connectivity (RSFC) after mild traumatic brain injury (mTBI). The authors implement graph theory to examine changes in functional brain network organization after mTBI and how these changes may impact cognition in individuals with mTBI (n=29) compared to healthy controls (n=29). Although this study presents interesting findings, there are several moderate concerns that dampen enthusiasm for this submission.Background and Hypotheses:
1. There is mature literature using rsfMRI and mTBI and the authors are encouraged to include in the manuscript how these findings advance the literature either through replication or extension.
2. The background also requires some clarification with respect to sample and goals. The authors mention that previous papers have examined RSFC and its relationship to cognition in mTBI, but they cite several blast-related mTBI studies. This study’s sample consisted of participants with civilian mTBI so focus on blast-related TBI and PTSD is unclear. Similarly, the discussion of repetitive injury does not appear to fit this sample, where individuals presumably sustained a single injury.
3. Hypotheses were unfortunately quite general, making it possible to support a range of findings. H1: global less significant than local, H2: “altered organization of the functional brain network may be associated with cognitive performance”. The second is particularly problematic in that it does not constrain the analyses to anything specific which casts some doubt on the reliability of the results.
Sample
4. The authors state that the mTBI group has significantly higher BDI scores than the control group and depression/psychiatric symptoms are quite common post mTBI. The authors also state that they created a positive correlation matrix that accounted for age and sex, but they did not mention accounting for BDI scores. Depression (and other mood disorders) impact brain network properties, and the authors mention this in their introduction. Therefore, the authors should address this issue to determine if symptoms of depression influence their findings.
5. While assessment of mTBI during the first month post injury is a clear strength of the paper, n=29 remains small and the findings are tentative.
Methods:
6. In this study, the preprocessed images were smoothed using a Gaussian kernel (sigma=2.55). However, spatial smoothing should be avoided when evaluating functional brain networks, especially when using the ROI approach (see Alakörkkö et al., 2017).The AAL atlas is a bit outdated and has been shown to lack sensitivity to mapping connectivity to behavior. Rationale for choosing this structural atlas over more recent functional atlases (e.g., Power264, Gordon, Schaeffer atlases) should be provided.
Also, with respect to creating the functional matrices, positive correlation matrices were retained after excluding negative values. More specifically, how did the authors handle the negative correlations in their functional connectivity matrices (absolute value, etc.)? Additionally, were the matrices Fisher r-to-z transformed? This should be addressed in the paper.
Statistics:The results state that the betweenness centrality in the right parahippocampus of the brain was positively correlated with scores on the verbal learning test in the mTBI group, but this positive correlation was not present in the healthy controls. However, more tests are needed to reveal if the Pearson’s correlation values are significantly different between the mTBI and healthy control group when examining the relationship between cognition and local brain network organization. The authors used a non-parametric permutation test to determine the significance of the graph theory measures between the mTBI group and the healthy control group. A similar approach should be utilized to examine if local brain network organization and its correlations with cognition differ between the two groups.
• It would also be nice to have a visual of the local brain regions that are associated with cognitive performance in the mTBI group.
Minor Points/ Questions:The authors regress out signal from white matter and cerebrospinal fluid, but do not include global signal regression (GSR). This is a controversial step that is hotly debated and the reasoning for this decision should be addressed.
Given the between-group differences in motion -- they should be presented in a Table and perhaps included in analyses to determine affects motion had on connectivity.
The authors use a range of proportional thresholds (8-26%) that allow for examining network topology but in case-control studies are problematic (see Hallquist & Hillary, 2018). While the argument is that other approaches lead to “different number of edges”, it is this possibility (differences in degree and strength) that are of critical interest after brain injury.
Reviewer: 2
Comments to the Author
This manuscript describes functional connectivity outcomes after mild traumatic brain injury in male participants. The major findings include a change in graph theory metrics within local networks. Limitations include a small sample size that is limited to males, lack of information on education level of participants. Comments and concerns are listed below:- Key point 3 states that fc can be used as an indicator of clinical outcomes, but this was not determined in the current study
Results:
- Data from the first 2 paragraphs should be included in a table.
- “…but the correlation disappeared when the outlier subject was removed.” Describe the criteria for removing an outlier.
- There is no information on education level of patients or whether patients were matched on education level.
- There are no females in the study.Discussion:
- There is little interpretation of the correlations found in mTBI patients but not controlsDecision letter by
Cite this decision letter
Reviewer report
2022/05/16This manuscript describes functional connectivity outcomes after mild traumatic brain injury in male participants. The major findings include a change in graph theory metrics within local networks. Limitations include a small sample size that is limited to males, lack of information on education level of participants. Comments and concerns are listed below:
- Key point 3 states that fc can be used as an indicator of clinical outcomes, but this was not determined in the current study
Results:
- Data from the first 2 paragraphs should be included in a table.
- “…but the correlation disappeared when the outlier subject was removed.” Describe the criteria for removing an outlier.
- There is no information on education level of patients or whether patients were matched on education level.
- There are no females in the study.Discussion:
- There is little interpretation of the correlations found in mTBI patients but not controlsReviewed by
Cite this review
Reviewer report
2022/04/26The main goal of this study was to examine resting-state functional connectivity (RSFC) after mild traumatic brain injury (mTBI). The authors implement graph theory to examine changes in functional brain network organization after mTBI and how these changes may impact cognition in individuals with mTBI (n=29) compared to healthy controls (n=29). Although this study presents interesting findings, there are several moderate concerns that dampen enthusiasm for this submission.
Background and Hypotheses:
1. There is mature literature using rsfMRI and mTBI and the authors are encouraged to include in the manuscript how these findings advance the literature either through replication or extension.
2. The background also requires some clarification with respect to sample and goals. The authors mention that previous papers have examined RSFC and its relationship to cognition in mTBI, but they cite several blast-related mTBI studies. This study’s sample consisted of participants with civilian mTBI so focus on blast-related TBI and PTSD is unclear. Similarly, the discussion of repetitive injury does not appear to fit this sample, where individuals presumably sustained a single injury.
3. Hypotheses were unfortunately quite general, making it possible to support a range of findings. H1: global less significant than local, H2: “altered organization of the functional brain network may be associated with cognitive performance”. The second is particularly problematic in that it does not constrain the analyses to anything specific which casts some doubt on the reliability of the results.
Sample
4. The authors state that the mTBI group has significantly higher BDI scores than the control group and depression/psychiatric symptoms are quite common post mTBI. The authors also state that they created a positive correlation matrix that accounted for age and sex, but they did not mention accounting for BDI scores. Depression (and other mood disorders) impact brain network properties, and the authors mention this in their introduction. Therefore, the authors should address this issue to determine if symptoms of depression influence their findings.
5. While assessment of mTBI during the first month post injury is a clear strength of the paper, n=29 remains small and the findings are tentative.
Methods:
6. In this study, the preprocessed images were smoothed using a Gaussian kernel (sigma=2.55). However, spatial smoothing should be avoided when evaluating functional brain networks, especially when using the ROI approach (see Alakörkkö et al., 2017).The AAL atlas is a bit outdated and has been shown to lack sensitivity to mapping connectivity to behavior. Rationale for choosing this structural atlas over more recent functional atlases (e.g., Power264, Gordon, Schaeffer atlases) should be provided.
Also, with respect to creating the functional matrices, positive correlation matrices were retained after excluding negative values. More specifically, how did the authors handle the negative correlations in their functional connectivity matrices (absolute value, etc.)? Additionally, were the matrices Fisher r-to-z transformed? This should be addressed in the paper.
Statistics:The results state that the betweenness centrality in the right parahippocampus of the brain was positively correlated with scores on the verbal learning test in the mTBI group, but this positive correlation was not present in the healthy controls. However, more tests are needed to reveal if the Pearson’s correlation values are significantly different between the mTBI and healthy control group when examining the relationship between cognition and local brain network organization. The authors used a non-parametric permutation test to determine the significance of the graph theory measures between the mTBI group and the healthy control group. A similar approach should be utilized to examine if local brain network organization and its correlations with cognition differ between the two groups.
• It would also be nice to have a visual of the local brain regions that are associated with cognitive performance in the mTBI group.
Minor Points/ Questions:The authors regress out signal from white matter and cerebrospinal fluid, but do not include global signal regression (GSR). This is a controversial step that is hotly debated and the reasoning for this decision should be addressed.
Given the between-group differences in motion -- they should be presented in a Table and perhaps included in analyses to determine affects motion had on connectivity.
The authors use a range of proportional thresholds (8-26%) that allow for examining network topology but in case-control studies are problematic (see Hallquist & Hillary, 2018). While the argument is that other approaches lead to “different number of edges”, it is this possibility (differences in degree and strength) that are of critical interest after brain injury.
Reviewed by
Cite this review
