Evaluating the Mandibular Trabecular Structure in Patients with and without Bruxism: A Cross-Sectional Study

Round 1

Reviewer 1 Report

This cross-sectional study evaluated a possible association the fractal dimension of adult patients with and without bruxism using panoramic radiographs.

 

-          Periods of recruitment are missing.

-          The reasons for exclusion of patients are missing.

-          All participants were at least 50 years old. Did you apply an age restriction during recruitment?

-          “lack of the second premolar and first molar teeth:” Except for this exclusion criterion, no information about the number of teeth of the included patients is presented. However, even other missing teeth (etc. antagonist, imbalanced dentition concerning left/right site) could be regarded as a potential confounder.

Figure 1/ Methods: There is no exact definition describing the process of identifying the ROIs in a reproducible manner.  

-          Table 3: Did you perform an adequate post-hoc test after the Kruskal-Wallis-test?

-          p. 6, ll. 168-170: which statistical test was used?

-          Although stated that two raters determined FDs, it seems that neither interrater nor intrarater reliability was tested.

-          Potential confounders and sources of bias were not considered.

 

The study addresses an interesting research question, but the methods of the study as well as the reporting reveal several shortcomings.

Therefore, I suggest the “Applied Sciences” to reject this study.  

Author Response

• Periods of recruitment are missing.

They have been added now (Page 2 line 73)

• The reasons for exclusion of patients are missing.

The exclusion criteria were mostly for radiographs however the exclusion reasons for patients are added too (Page 2 line 80)

• All participants were at least 50 years old. Did you apply an age restriction during recruitment?

No, there was no age restriction.

• “lack of the second premolar and first molar teeth:” Except for this exclusion criterion, no information about the number of teeth of the included patients is presented. However, even other missing teeth (etc. antagonist, imbalanced dentition concerning left/right site) could be regarded as a potential confounder.

Since the fractal dimension from the region of interest between the second premolar and the first molar was analyzed, we believe that the lack of these teeth is the most effective as a confounder for the fractal dimensions.

• Figure 1/ Methods: There is no exact definition describing the process of identifying the ROIs in a reproducible manner.

The description of the process of identifying the ROIs is now added. (Page 3, lines 99-105)

• Table 3: Did you perform an adequate post-hoc test after the Kruskal-Wallis-test?

Yes, Mann-Whitney post hoc test was performed. (Page 7, line 187)

• 6, ll. 168-170: which statistical test was used?

Mann-Whitney post hoc test was performed. (Page 7, line 195)

• Although stated that two raters determined FDs, it seems that neither interrater nor intrarater reliability was tested.

The intraobserver and interobserver agreements have been added to the result section. (Page 5, lines 159-161)

• Potential confounders and sources of bias were not considered.

The power analysis and intrarater and interrater agreements can account for the potential biases.

Reviewer 2 Report

Brief summary: The aim of the study is to analyze the trabecular pattern differences due to bruxism as measured using fractal dimension in condyle, alveolar and genial regions. No significant difference in fractal dimension was observed in patients with bruxism compared to those without bruxism. The main strength compared to other published studies on this topic is its bigger sample size. However, the study could be improved in methods and results. Due to a lot of missing results, some of the conclusions aren’t fully supported.

 

Specific comments:

 

Introduction: Please explain how does FD values relate to clinically? Are lower FD values usually associated with lower trabecular bone pattern?

 

Methods:

 

Line 81: More information on how polysomnography test was done? What muscles were involved? Any reference technique that was used? How many researchers/clinicians were involved with bruxism diagnosis? Also, how was a diagnosis of bruxism assigned to patient if they have positive polysomnography test and no clinical symptoms vs vice versa? Steps taken to avoid any bias if multiple researches were involved

 

Line 90: How many researchers were involved with selection of these 6 regions of interest. What training did they have in selecting those 6 regions? Please add few more details on steps taken to avoid any bias if multiple researchers were involved.

 

Line 106-108: Add citations for the techniques

 

Results:

 

Table 2: If the total sample size is 365, why are the totals for each ROI only 192?

 

Table 2: What is the p-value testing here exactly? As described in text is it bruxers vs non-bruxers? If so, would recommend to modify the table to show the mean and SD by bruxism status with corresponding p-value of tests comparing bruxers to non-bruxers. Make sure the total observations are 365.

 

For Line 154-155: Create a new table showing FD values for bruxers vs non-bruxers separately for males and females separately (Sub-group analysis). Update the table with appropriate p-values

 

Line 168-169: Please specify what post-hoc test was used for pair-wise comparison following Kruskal-Wallis test

 

Discussion:

 

Line 193-199. Most of it is already presented in introduction and is just a repetition. Could be deleted.

 

Line 200- Suggest to move it to introduction.

 

Line 223- Please discuss what steps were taken in the current study to prevent any of those known errors

 

Line 251-254: No analysis/p-value was presented to test the difference of FD values between the different regions for the entire dataset.

 

Line 258-260: Was there any additional analysis done comparing different FD values across regions in bruxers and non-bruxers. No P-value or data was presented in the manuscript.

 

Line 275: No analysis/p-value was presented to test the difference of FD values between males vs females.

 

Limitations: Discuss about any possible bias in the study.

 

References:

 

Reference #1 & 12: Missing title and authors details

 

Reference #3 & 5 are the same - Kato, C.N., et al., Use of fractal analysis in dental images: a systematic review. Dentomaxillofac Radiol, 2020. 49(2): p. 20180457.

Author Response

• Introduction: Please explain how does FD values relate to clinically? Are lower FD values usually associated with lower trabecular bone pattern?

Yes, lower FD value means lower complexity in the trabecular pattern. A sentence has been added to the introduction section. (Page 1, line 41).

 

Methods:

 

• Line 81: More information on how polysomnography test was done? What muscles were involved? Any reference technique that was used? How many researchers/clinicians were involved with bruxism diagnosis? Also, how was a diagnosis of bruxism assigned to patient if they have positive polysomnography test and no clinical symptoms vs vice versa? Steps taken to avoid any bias if multiple researches were involved

The polysomnography test was performed by on individuals on master muscles following the same procedures which has been mentioned in the highlighted cited study. Moreover, the combination of polysomnography, history of clinching and tooth wear used for bruxism diagnosis. (Page 2, line 88).

• Line 90: How many researchers were involved with selection of these 6 regions of interest. What training did they have in selecting those 6 regions? Please add few more details on steps taken to avoid any bias if multiple researchers were involved.

More details about ROIs selection and the researchers have been added. (Page 3 lines 99-105).

 

• Line 106-108: Add citations for the techniques

The reference has been added. (Page3).

 

• Table 2: If the total sample size is 365, why are the totals for each ROI only 192?

The tables are edited

 

• Table 2: What is the p-value testing here exactly? As described in text is it bruxers vs non-bruxers? If so, would recommend to modify the table to show the mean and SD by bruxism status with corresponding p-value of tests comparing bruxers to non-bruxers. Make sure the total observations are 365.

A clearer description is now added. (Page 5, line 175).

• For Line 154-155: Create a new table showing FD values for bruxers vs non-bruxers separately for males and females separately (Sub-group analysis). Update the table with appropriate p-values

The tables are added. (Page 6)

 

• Line 168-169: Please specify what post-hoc test was used for pair-wise comparison following Kruskal-Wallis test

A Mann-Whitney post hoc test was performed. (Page 7 line 195).

 

Discussion:

 

• Line 193-199. Most of it is already presented in introduction and is just a repetition. Could be deleted.

These sentences are deleted.

 

• Line 200- Suggest to move it to introduction.

Done. (Page 1 line 41)

 

• Line 223- Please discuss what steps were taken in the current study to prevent any of those known errors

 

The power analysis and intrarater and interrater agreements can account for the potential biases.

 

• Line 251-254: No analysis/p-value was presented to test the difference of FD values between the different regions for the entire dataset. Line 258-260: Was there any additional analysis done comparing different FD values across regions in bruxers and non-bruxers. No P-value or data was presented in the manuscript.

The p values have been added to the result section. Page 6 line 178.

 

• Line 275: No analysis/p-value was presented to test the difference of FD values between males vs females.

In both bruxers and non-bruxers groups the was no significant differences in FD values between to sexes therefore the individuals, gender did not have effect on FD values.

 

Limitations: Discuss about any possible bias in the study.

 

Population bias is discussed in limitations (Page 10, lines 336-338)

  

• Reference #1 & 12: Missing title and authors details

Corrected (reference list)

• Reference #3 & 5 are the same - Kato, C.N., et al., Use of fractal analysis in dental images: a systematic review. Dentomaxillofac Radiol, 2020. 49(2): p. 20180457.

The duplicated reference is deleted (reference list)

Reviewer 3 Report

Introduction:

1.       Introduction is well written however, the authors need to add more details on the prevalence rate, and need for this study.

2.       How these findings are benefiting a non-specialist, or lesser experienced clinician?

 

 

Methodology:

1.      Was the sample size calculated?

2.      Was age of the patient considered for inclusion?

3.      Reason for bruxism? Stress or patients with mental health issues may exhibit different characteristics

4.      Food habits of the patients may also have an effect on tooth wear

5.      In the statistical analysis section the value of statistical significance should be mention as p<0.05

 

 

Results: Well presented

Discussion: Well Written

 

Comments for author File: Comments.pdf

Author Response

Introduction:

• Introduction is well written however, the authors need to add more details on the prevalence rate, and need for this study.

The bruxism prevalence has been added to the introduction section. (Page 2, line 52)

• How these findings are benefiting a non-specialist, or lesser experienced clinician?

Due to the bruxism nature not being easy to diagnosis and because the polysomnography test, which is gold standard method for detecting bruxism, is hard to perform, diagnosing the bruxism by performing the fractal dimension on oral radiographs could be an easier approach if the bruxism leads to changes in FD values. 

Methodology:

• Was the sample size calculated?

The sample size calculation is now added to the method section. (Page 2, line 85).

• Was age of the patient considered for inclusion?

No there was no age restriction.

• Reason for bruxism? Stress or patients with mental health issues may exhibit different characteristics

This point was not considered in the sample selection as it would limit the number of included patients. Additionally, due to the multi-dimensional nature of habits such as bruxism, determination of the etiology of bruxism was beyond the scope of the present study. A short statement is now added to the limitations (Page 10, lines 333-336)

• Food habits of the patients may also have an effect on tooth wear

Polysomnography test was performed combined with the tooth wear and clinching history to ensure that the tooth wear occurred due to bruxism

• In the statistical analysis section, the value of statistical significance should be mention as p<0.05

It was mentioned in this section. (Page 4, line 153)

Round 2

Reviewer 1 Report

The authors revised the manuscript and answered all the concerns. I therefor suggest to accept this manuscript.