Sleep Disorder Prevalence among Brazilian Children and Adolescents with Down Syndrome: An Observational Study

Round 1

Reviewer 1 Report

Sleep disorders in children with Down syndrome: observational study

Reviewing this article was not very easy. But the theme is interesting and must be studied. We learn by what we do not know, and in this case, it would be beneficial if we could get some results about sleep disorders in children with Down Syndrome.

A critical review of methods

The authors clearly state this is an observational study with a convenience sample. The convenience sample can be used as an exploratory tool. The disclosure of a convenience sample is essential because it sets the bounds for generalization. The paper could benefit if the authors stated what the calculations were to set the sample size in 41. However, the authors are not complying to provide that information because they do not claim the capacity of generalization and methodology is qualitative.

In this study, previously existent instruments were used. Hence, authors should present some psychometric properties so that readers can be informed about the quality of the measures. Among these measures are reliability (e.g., Cronbach’s alpha) and, when applicable, a set of other measures such as total explained variance, loadings, the goodness of fit measures, etc.

The statistical analysis description is concise but, in fact, very accurate because nothing more is done in the results section. However, it would be essential to implementing some qualitative assessments such as interviews or content analyses in a qualitative study. The authors used questionnaires and treated them with a qualitative approach. Using a quantitative approach in this paper would probably be more adequate. This use of a qualitative approach with questionnaires is “somehow” misleading because readers are expecting to see results in a much more summarized format.

A critical review of results

The authors choose an approach based on the entire data description in the results section. The authors did not base the presentation of results on summary measures. In Table 1, a complete description of the results is presented. In Table 2, again, a full report is given, and no proportions are calculated. The same happens in Table 3. Table 4 shows the scores of every participant for each instrument/ dimension. The qualitative approach that authors use has the advantage of being fully descriptive. On the other hand, readers could benefit from having summary measures for each question, dimension scores, or total scores for the questionnaires. This is what is expected in a questionnaire.

The approach followed in this paper makes the results a bit confusing. For example, in the next paragraph, there Is an extensive description of the questionnaire measure results of the participant, but not associated with other participant characteristics are performed. One of the advantages of qualitative approaches is the analysis of potential relations between variables at the “subject scale.” However, in this paper, authors extensively describe questionnaire results but make no associations. In this way, it would be more comprehensible to describe the above paragraph like this: DIMS > 21 was observed in 3 (xx %) children, SDB score > 12 was observed in 12 (xx %) children, one child had EDS > 19 points, one child scored > 7 in HDS. A total of 15 (xx %) children were above the threshold in at least one factor. Overall, 3 (xx %) children scored higher than 69 points and were considered to have sleep disorders. No child showed higher ADD and DSWT scores.

In table 4, it is possible to verify that three children (P19, P22, and P24) had higher DIMS scores (above 21 points); 12 children (P05, P07, P08, P10, P17, P19, P23, P24, P27, P30, P33, P41) had higher on the SDB score (above 6 points); no child showed higher on the ADD and the DSWT scores; the child (P03) had higher on the EDS score (above 19 points); one child (P11) had higher on the HS score (above 7 points). Although 15 children (P03, P05, P07, P08, P10, P11, P17, P19, P22, P23, P24, P27, P30, P33, P41) showed higher the score of at least one factor, in the total score, only three children (P03, P05, P41) had a higher score (above 69 points) and were considered to have sleep disorders.

There is no advantage in extensively describing results from a qualitative point of view when there is no attempt to make any associations or interpret those extensive descriptions.

In summary, the authors use instruments expected to be used in quantitative analysis (questionnaires) but then implement a qualitative approach, presenting data in categories previously existing in the questionnaires.

It would be better to change to a quantitative approach so that the paper becomes more understandable and there is a clear connection between the methods and results presentation.

The article may help discuss some Down Syndrome sleep problems and as an exploratory tool.

Author Response

Can be improved: the research design, the methods, and the conclusions.

Author Response File: Author Response.docx

Reviewer 2 Report

Manuscript Number: applsci-2220101

Title: “Sleep disorders in children with Down syndrome: observational study”

The aim of this research was to evaluate the presence of sleep disorders in children with Down Syndrome (DS) in Brazil. To this aim, an electronic survey was administered using Google form. In particular, two questionnaires were used: Sleep Questionnaire by Reimao and Lefevre (QRL) and Sleep Disorders Scale for Children (SDSC). Overall, 41 parents participated to the survey. Qualitative evaluation showed that most DS children have some type of sleep disorder.

The topic of the manuscript is partially interesting. The manuscript is clearly written. However, the aim declared by the Authors is only partially achieved. First of all, due to the size of the sample. Second, for the lack of a control group.

Also, I wonder what knowledge this study adds to what is already known in the literature. Is there any reason why Brazilian children with Down Syndrome should have a different incidence of sleep disorders in comparison with children with Down Syndrome living in other countries?

Why is the sample size so limited despite using the internet?

Figure 1 could be deleted.

Table 4 could be deleted.

Considering the qualitative nature of the results, the discussion section is too long.

Conclusion section is no mandatory, therefore, in the present form it could be removed.

Author Response

All the sessions pointed by the reviewer were improved.

Abstract: 

- remove “most of whom live in the city of Rio de Janeiro”. It is irrelevant for the constrain of word limit of the abstract The change has been made.

- replace “person responsible for DS child” with guardian The change has been made.

- put the mean age of the 41 children Its was done

- what is the difference between “have some sleep disorder” vs “are considered to have sleep disorders”? (line 28-29) The change has been made.

- indicate with precise % what “most children” stand for. Its was done.

- delayed sleep onset and nocturnal awakenings are not type of sleep disorders. They may be considered sleep impairment, if you wish The change has been made.

Introduction: 

The introduction is very confused. This is currently the structure of the introduction: 

- first the authors introduce DS. Then they talk about sleep problems (although they talk about the general population, which does not make much sense as the study is on DS). Then they talk about cognitive impairment. Then about OSA prevalence. Then the talk about cardiovascular diseases; then they talk about RMD. They mention lower airway abnormalities and the sentence is left unsolved and unaddressed. Finally, the authors talk about memory problem with recommendation and then aging. In other words, it does not make any sense to address all of these different domains in the introduction. The introduction should lead the reader straight to the aim of the study. Why we need this research. The change has been made.

Material and methods:

2.1 call this section Ethics approval and study design

- line 103: to whom where the two questionnaires disseminated? Who were the recipients?

- line 111: not clear if there was a link sent through whatsapp. The change has been made.

- line 119: “validation via email” is not clear what the authors mean The change has been made.

- line 119: “the identify of the children” what does it mean? What was disclosed? What kind of identity? The change has been made.

- line 122: “except those descriptive responses… “ meaning is unclear. I suggest to eliminate the sentence. Were all the answers mandatory in the questionnaires? The change has been made.

 - line 127: the authors should indicate the total number of recipients of the questionnaire, because this might greatly influence the generalizability of the results and the response bias.

Inclusion and exclusion criteria should be better clarified in the Participants section. 

Also, some of the concepts are repeated among the different paragraphs. I suggest that the authors avoid repetition and collocate the concept in the appropriate subsection. The change has been made.

Results: 

Figure 2-3: it does not make any statistical sense to display in the same graph and in the same bars the same concept expressed in different unit (Number vs %). Just display under one of the two units. The change has been made.

Did the author correlated the results derived from SDSC from the ones obtained from QRL? That would be more interesting that only showing the % or a qualitative assessment. Yes, we did it.  

Discussion: 

Part of the discussion is focused on OSA. But if I understand the results of the study correctly, we don’t know what kind of sleep disorders these DS children suffer from. The change has been made.

Are the sleep disorders breathing assessed in the questionnaire only OSA? 

The discussion should compared similar studies in the literature that possibly used the same questionnaire, or that evaluated the same outcomes on SD children. The change has been made.

If possible, it should also be shortened. The change has been made.

Conclusion: 

“most DS children have some type of sleep disorder” is misleading, as it seems like disorders like OSA are very prevalent (whereas it was 29%).  The change has been made.

Limitation: No sample size calculation; no selection bias mentioned in the limitation of the study. 

The sample size was calculated.

No selection bias was also included on limitation session.

Reviewer 3 Report

Comments to the authors

Thank you for the invitation to review the manuscript entitled “Sleep disorders in children with Down syndrome: observational study”. This study is a cross-sectional observation design aimed at quantifying the prevalence of sleep disorders in a group of 41 children affecting by Down syndrome, by means of two validated questionnaires. 

My major requests of amendment are the following: 

Abstract: 

- remove “most of whom live in the city of Rio de Janeiro”. It is irrelevant for the constrain of word limit of the abstract

- replace “person responsible for DS child” with guardian 

- put the mean age of the 41 children 

- what is the difference between “have some sleep disorder” vs “are considered to have sleep disorders”? (line 28-29)

- indicate with precise % what “most children” stand for

- delayed sleep onset and nocturnal awakenings are not type of sleep disorders. They may be considered sleep impairment, if you wish 

Introduction: 

The introduction is very confused. This is currently the structure of the introduction: 

- first the authors introduce DS. Then they talk about sleep problems (although they talk about the general population, which does not make much sense as the study is on DS). Then they talk about cognitive impairment. Then about OSA prevalence. Then the talk about cardiovascular diseases; then they talk about RMD. They mention lower airway abnormalities and the sentence is left unsolved and unaddressed. Finally, the authors talk about memory problem with recommendation and then aging. In other words, it does not make any sense to address all of these different domains in the introduction. The introduction should lead the reader straight to the aim of the study. Why we need this research. 

Material and methods:

2.1 call this section Ethics approval and study design

- line 103: to whom where the two questionnaires disseminated? Who were the recipients?

- line 111: not clear if there was a link sent through whatsapp.

- line 119: “validation via email” is not clear what the authors mean

- line 119: “the identify of the children” what does it mean? What was disclosed? What kind of identity?

- line 122: “except those descriptive responses… “ meaning is unclear. I suggest to eliminate the sentence. Were all the answers mandatory in the questionnaires?

 - line 127: the authors should indicate the total number of recipients of the quesrtionnaire, because this might greatly influence the generalizability of the results and the response bias.

Inclusion and exclusion criteria should be better clarified in the Participants section. 

Also, some of the concepts are repeated among the different paragraphs. I suggest that the authors avoid repetition and collocate the concept in the appropriate subsection. 

Results: 

Figure 2-3: it does not make any statistical sense to display in the same graph and in the same bars the same concept expressed in different unit (Number vs %). Just display under one of the two units. 

Did the author correlated the results derived from SDSC from the ones obtained from QRL? That would be more interesting that only showing the % or a qualitative assessment.  

Discussion: 

Part of the discussion is focused on OSA. But if I understand the results of the study correctly, we don’t know what kind of sleep disorders these DS children suffer from. Are the sleep disorders breathing assessed in the questionnaire only OSA? 

The discussion should compared similar studies in the literature that possibly used the same questionnaire, or that evaluated the same outcomes on SD children. 

If possible, it should also be shortened. 

Conclusion: 

“most DS children have some type of sleep disorder” is misleading, as it seems like disorders like OSA are very prevalent (whereas it was 29%).  

Limitation: No sample size calculation; no selection bias mentioned in the limitation of the study. 

Author Response

Reviewer 3-

All the sessions (the research design, methods, the results and the conclusions) pointed by the reviewer were improved.

While the data in this paper could be highly interesting, the exact aim of the paper remains unclear. This might be due to only using descriptive values and raw data without performing statistical tests. The scientific importance of this paper would highly increase if basic statistical tests were performed (comparisons between gender, tests to assess associations between different outcomes. The change has been made.

The methods regarding ethics and participants contain repetitive information, please revise and adapt accordingly. The change has been made.

The tables showing the results of the questionnaires are quite hard to follow as they are really crowded. Using percentages and different columns for each sex, could help solve this issue. I would suggest to make a less crowded table for the main paper, large tables with raw data might be a better fit in an online supplement. The change has been made.

Author Response File: Author Response.docx

Reviewer 4 Report

This manuscript describes the use of two questionnaires for detection of OSA in children with Down Syndrome.

While the data in this paper could be highly interesting, the exact aim of the paper remains unclear. This might be due to only using descriptive values and raw data without performing statistical tests. The scientific importance of this paper would highly increase if basic statistical tests were performed (comparisons between gender, tests to assess associations between different outcomes,…)

The methods regarding ethics and participants contain repetitive information, please revise and adapt accordingly.

The tables showing the results of the questionnaires are quite hard to follow as they are really crowded. Using percentages and different columns for each sex, could help solve this issue. I would suggest to make a less crowded table for the main paper, large tables with raw data might be a better fit in an online supplement.

Author Response

All the sessions (the research design, methods, the results and the conclusions) pointed by the reviewer were improved. 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

All the recommendations made were followed, and others were made to give coherence to work. It is clear that there was work to improve the article, and this has merit. In short, a good job has been done following up on the recommendations and reorganizing the work.

Author Response

We appreciate your relevant contributions and recognize that with the suggestions of the 4 reviewers, this current version is effectively more qualified for a possible publication.

Author Response File: Author Response.pdf

Reviewer 2 Report

The revised version of the manuscript is improved.

However Authors did not perform all ruquire changes.  

Author Response

We appreciate the identification of the adjustments we have made in order to improve the manuscript. Nevertheless, we consider all possible changes. Those that did not change, it was due to a consensus among the comments of the 4 reviewers, and not for lack of attempt to result in a better version of this manuscript.

Reviewer 3 Report

Comment to the authors

 The authors did an extensive job in trying to improve the soundness of their study. While I recognize the hard work and efforts that clearly increased the scientific merit of the paper, some improvement still needs to be done. Here some suggestions. 

Abstract: 

- line 21: replace “aimed to try to quantify” with “aimed at investigating the prevalence”

- lines 29-32: reformulate the conclusion, adding the most important findings. In the current version of the conclusion, the authors here repropose the same sentence 3 times. 

Introduction: 

line 72: same comment as above for the aim of the study (see abstract)

- line 72: replace “so” with “therefore”

Methods: 

- line 92: errors in English grammar. 

- line 93: what are these groups about? What is the estimate sample size of these groups? Which individuals are part of these groups? Do they only need to be residents of Brazil to be part of these groups? Why were these groups chosen? 

- line 110: English syntax not correct 

- line 143: something is wrong with this sample size calculation. Normally you would have an average, SD that produce a power of … at an alpha = 0.05. Please, reconsider this sample size calculation. 

- statistical analysis: what was the level of statistical significance selected? Which statistical package was selected to carry out the analysis? 

Table 1:

- how were daytime sleepiness and OSA diagnosis detected? Through the answer to ad-hoc questionnaire? It is not clear from the methods which are the questionnaire that provide the answer of such questions. 

Results:

- where are the data presented in the tables consistent with the paragraph of lines 188-193? Also, as there is a comparison between genders, can the authors run a simple statistical analysis of chi-square for proportion between boys and girls to see if there is any significant difference? And actually, all the results can also be investigated through chi-square test for proportion

Limitation:

- line 386: what does it mean that “no selection bias was done”?  

Author Response

Abstract: 

- line 21: replace “aimed to try to quantify” with “aimed at investigating the prevalence” This change has been made.

- lines 29-32: reformulate the conclusion, adding the most important findings. In the current version of the conclusion, the authors here repropose the same sentence 3 times. This change has been made.

Introduction: 

line 72: same comment as above for the aim of the study (see abstract) This change has been made.

- line 72: replace “so” with “therefore” This change has been made.

Methods: 

- line 92: errors in English grammar. This change has been made.

- line 93: what are these groups about? What is the estimate sample size of these groups? Which individuals are part of these groups? Do they only need to be residents of Brazil to be part of these groups? Why were these groups chosen? It was removed.

- line 110: English syntax not correct. This change has been made.

- line 143: something is wrong with this sample size calculation. Normally you would have an average, SD that produce a power of … at an alpha = 0.05. Please, reconsider this sample size calculation. This change has been made.

- statistical analysis: what was the level of statistical significance selected? Which statistical package was selected to carry out the analysis? The information has been added.

Table 1:

- how were daytime sleepiness and OSA diagnosis detected? Through the answer to ad-hoc questionnaire? It is not clear from the methods which are the questionnaire that provide the answer of such questions. The information has been added.

Results:

- where are the data presented in the tables consistent with the paragraph of lines 188-193? Also, as there is a comparison between genders, can the authors run a simple statistical analysis of chi-square for proportion between boys and girls to see if there is any significant difference? And actually, all the results can also be investigated through chi-square test for proportion. This change has been made.

Limitation:

- line 386: what does it mean that “no selection bias was done”? We removed it.

Author Response File: Author Response.pdf

Reviewer 4 Report

I would like to thank the reviewers for taking into account the suggestions, the modifications have increased the paper’s quality. Some remarks remain:

General comment: please critically read through the paper to remove typos and grammatical errors. Especially the discussion would benefit from a critical evaluation of the text.

Abstract:

·       I would suggest to remove ‘to try to’ and just use ‘This current study aimed to quantify the prevalence of sleep disorders in a group of 41 DS children and adolescents’ as primary aim.

·         Please add percentages for boys and other parameters mentioned

Introduction:

·         The study of Lee showed a high prevalence of OSA regardless of measurement technique. Due to the differences in calculation of AHI and sleep duration, it is expected to have lower AHI under polygraphy compared to polysomnography, regardless of the prevalence of Down Syndrome. This might not be correctly reflected by the current wording.

Results:

·         Table 2 is quite hard to follow. Is it correct that the first column contains the questions of the questionnaire while the other columns represent the possible answers? It is quite confusion at first that not all columns are filled in, but I understand it is hard to visualize. Maybe a stacked bar chart might be a good way to visualize? Or an additional heading can be added to the table?

·         If you write this paragraph: “It is possible to verify that most children (68,28%) wake up at least once during the night. The answers indicate that boys presented a more worrying behavior compared to girls with DS. Most boys sleep during the day and take longer to sleep at night and spend more time to go back to sleep. The habit of waking up at night is reported in a balanced way between boys and girls. And for both, sleep does not interfere with your daily routine”, comparing boys to girls, these statements should be backed up with statistical tests

·         Did you find any correlation between the results of both tests?

·         In Figure 4B, a lot of obesity values are equal to 0. How is this possible? Are these missing values? Please verify and adapt as needed.

Author Response

Reviewer 4

I would like to thank the reviewers for taking into account the suggestions, the modifications have increased the paper’s quality. Some remarks remain:

General comment: please critically read through the paper to remove typos and grammatical errors. Especially the discussion would benefit from a critical evaluation of the text.

Abstract:

•      I would suggest to remove ‘to try to’ and just use ‘This current study aimed to quantify the prevalence of sleep disorders in a group of 41 DS children and adolescents’ as primary aim. This change has been made.
• Please add percentages for boys and other parameters mentioned. We chose don’t compared genders in this current study.

Introduction:

• The study of Lee showed a high prevalence of OSA regardless of measurement technique. Due to the differences in calculation of AHI and sleep duration, it is expected to have lower AHI under polygraphy compared to polysomnography, regardless of the prevalence of Down Syndrome. This might not be correctly reflected by the current wording. This change has been made.

Results:

• Table 2 is quite hard to follow. Is it correct that the first column contains the questions of the questionnaire while the other columns represent the possible answers? It is quite confusion at first that not all columns are filled in, but I understand it is hard to visualize. Maybe a stacked bar chart might be a good way to visualize? Or an additional heading can be added to the table? This change has been made. We improved it.
• If you write this paragraph: “It is possible to verify that most children (68,28%) wake up at least once during the night. The answers indicate that boys presented a more worrying behavior compared to girls with DS. Most boys sleep during the day and take longer to sleep at night and spend more time to go back to sleep. The habit of waking up at night is reported in a balanced way between boys and girls. And for both, sleep does not interfere with your daily routine”, comparing boys to girls, these statements should be backed up with statistical tests. We chose don’t compared genders in this current study.
• Did you find any correlation between the results of both tests? We chose don’t compared genders in this current study.
• In Figure 4B, a lot of obesity values are equal to 0. How is this possible? Are these missing values? Please verify and adapt as needed. The information has been added.

Author Response File: Author Response.pdf