Evidence-Based Aerobic Exercise Training in Metabolic-Associated Fatty Liver Disease: Systematic Review with Meta-Analysis

Round 1

Reviewer 1 Report

The main aim of the present meta-analysis is to examine the effect of aerobic exercise on different outcomes in adults with MADLF, and to determine the more effective protocol of aerobic exercise for the improvement of the different outcomes. 

Altohugh the idea of the article is interesting, the stregnt of the it (to compare the different aerobic protocols) was not carry out. I would have excepted subgroup analysis or at least a forest plot with the MD value of each article. The introduction it is too long, specially the description of the existing meta-analysis (and the inclusion of other lifestyles were mixed). The part of the results was not structurated and the explanation of the different aerobic intervertions was not so clear (it would be easier to understant if separating the differentes elemtns of FITT). Results of Risk of Bias were not presented. They lack many refernces in the text and there are some mistakes in the language. The numbers of tables were not in order and the figures were not included in the PDF of the article. I would have include a flow diagram following the PRISMA structure to clarigy the seerch strategy and results of it. The specific search strategy should have been specified. The terms MAFLD, NAFLD and hepatic steatosis, physical activity and exercise were sometimes used interchangeaby.

On the other hand, the use of the PICOS strategy (I would specify it in the methods) and the reference of the ACSM guidelines for exercise were well empolyed.

I would encourage the authors to develop this interesting idea, but making a marked change in the results structure and presentation and the writting of the article.

Author Response

Reviewer's comment: The main aim of the present meta-analysis is to examine the effect of aerobic exercise on different outcomes in adults with MADLF, and to determine the more effective protocol of aerobic exercise for the improvement of the different outcomes. Altohugh the idea of the article is interesting, the stregnt of the it (to compare the different aerobic protocols) was not carry out. I would have excepted subgroup analysis or at least a forest plot with the MD value of each article.

Authors' response: We thank the reviewer for their comments.  We  apologize because there appears to have been  some technical problems with the figures for the manuscript (eg. Forest plot).  These were uploaded in a separate file (not in main document) and as a result we don’t believe you had sight of them.

We have also  reorganized the results section to include subgroup analysis (3.5. Meta-regression analysis result) considering intervention time, intensity, volume and progression of the exercise protocol.

Reviewer's comment: The introduction it is too long, specially the description of the existing meta-analysis (and the inclusion of other lifestyles were mixed).

Authors' response: Thank you for that comment. We have shortened the introduction as suggested.

Reviewer's comment: The part of the results was not structurated and the explanation of the different aerobic intervertions was not so clear (it would be easier to understant if separating the differentes elemtns of FITT).

Authors' response: Thank you for that comment. We have added Figure 2 and Table 3 with further details of the included aerobic protocols detailed according to FITT principle recommended by ACSM.  We hope this makes the section clearer.

Reviewer's comment: Results of Risk of Bias were not presented.

Authors' response: We have used Cochrane Risk of Bias Tools with the  results shown in Fig. 1B. We apologise if this was not clear.  Moreover we have added in the supplementary file funnel plots and results of Begg’s test for ALT.

Reviewer's comment:  They lack many references in the text and there are some mistakes in the language.

Authors' response: We apologise for this and have  corrected the references.  The article has also been further  proofread.  We hope that this improves the language.

Reviewer's comment:  The numbers of tables were not in order and the figures were not included in the PDF of the article.

Authors' response: We apologise for this and have corrected these  mistakes.

Reviewer's comment: I would have include a flow diagram following the PRISMA structure to clarigy the seerch strategy and results of it.

Authors' response: We apologise. These figures are now visible and correctly uploaded. Fig. 1. Flowchart of studies and critical appraisal is now included. As is a Flow chart of the inclusion/exclusion process, according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.

Reviewer's comment: The specific search strategy should have been specified. The terms MAFLD, NAFLD and hepatic steatosis, physical activity and exercise were sometimes used interchangeaby. On the other hand, the use of the PICOS strategy (I would specify it in the methods) and the reference of the ACSM guidelines for exercise were well empolyed. I would encourage the authors to develop this interesting idea, but making a marked change in the results structure and presentation and the writting of the article.

Authors' response: Thank you for that valuable comment. We apologize if this ‘search strategy’ section wasn’t clear enough, we have changed the text in  this section in the hope that it is clearer.

Reviewer 2 Report

1-Page 3, line 102, please try to summarize the past literature in few lines for comparison rather to give explanations in the introduction section. This explanation can be added in the discussion section.

2- One important component is missing in this review. is the quality of evidence of outcomes by using GRADE approach and it should be added in this manuscript.

3-Forest plot for each outcome (random and fixed effect model pooled estimate) should be included in the manuscript.

4-Many included studies have only a few patients and it also need to be included among limitations. 

5-Please include the study selection flowchart. 

Author Response

We thank the reviewer for their comments.  We  apologize because there appears to have been  some technical problems with the figures for the manuscript (eg. Forest plot, result of Cohrane Risk of Bias).  These were uploaded in a separate file (not in main document) and as a result we don’t believe you had sight of them.

Reviewer's comment: 1-Page 3, line 102, please try to summarize the past literature in few lines for comparison rather to give explanations in the introduction section. This explanation can be added in the discussion section.

Authors' response: Thank you for that comment. We have reorganized this section.

Reviewer's comment: 2- One important component is missing in this review. is the quality of evidence of outcomes by using GRADE approach and it should be added in this manuscript.

Authors' response: We apologise for this omission, we used Cochrane Risk of Bias Tools and the results are in Fig. 1B. Moreover we have added in the supplementary file funnel plots and results of Begg’s test for ALT.

Reviewer's comment: 3-Forest plot for each outcome (random and fixed effect model pooled estimate) should be included in the manuscript.

Authors' response: These figures are now visible and correctly uploaded.

Reviewer's comment: 4-Many included studies have only a few patients and it also need to be included among limitations. 

Authors' response: Thank you for that valuable comment. We added this to the study limitations in the final  paragraph of the discussion.

Reviewer's comment: 5-Please include the study selection flowchart. 

Authors' response: Changes have been made.

Round 2

Reviewer 1 Report

I would like to thank to the authors for their work and changes in the manuscript; they have carefully addressing my questions and comments and the paper has substantially improved. I would like the congratulate the authors on their work and the novel perspective of the results.

However, I have few comments before the article will be ready for its publication:

1) Methods (line 240). Please, include the reference for the PICOS strategy.

2) Lines 246. Could the authors, if possible, include the specific search strategy (specific terms and Boolean operators; not only "the combination of keywords"), at least in supplementary material. It is an important point for the reproducibility of the search of a systematic review.

3) Footnote of Figure 1. Please, include the references of PRISMA and the Cochrane risk of bias tool.

4) Footnotes of Table 2. Please, include all the abbreviations included in the Table 2.

5) Results (lines 320-323). Please, include the references of the "twelve studies evaluated". Include all the refences in all similar situations across the results section.

6) Lines 12-15 in the result section. Please, do not repeat the explanation of the abbreviation.

7) Line 30. "results showed that IHTG was significantly reduced after..." because this result it not an association.

8) Figure 3. I would include the groups (aerobic exercise) vs usual care, in the figure, not only in the footnote. Similarly, I woul include the name of the outcome in each sub-graph (ALT in A...).

9) Why the authors ordered the included articles in this way in tables and figures? Consider ordering them according to an alphabetical or year of publication criteria. On the other hand, why did the authors make the forest plots and analyses with the mean difference and not with effect size, if the ES was calculated?

10) Line 60 (3.5 section in results). "...between total intervention time and ALT difference or reduction".

11) Part 3.5 of the results? Please, rewrite sentence of lines 60-62, it is a little confusing. Did the authors analyze the differences of the different exercise modes in changes in IHTG or BMI? It could be very interesting to perform the analysis to answer the question regarding what kind of exercise (in FITT) has a higher effect in the improvement of MAFLD.

12) Discussion (line 73). Please, modify the full name by the “ACSM” abbreviation.

13) Discussion. Although the author affirmed that “First, to the best of our knowledge, this is the first 130 meta-analysis evaluating the effect of aerobic exercise to categorize training protocols according to FITT (frequency, intensity, time, type) principle recommended by ACSM, on liver function outcome”, the role of the different components of FITT on liver function has not been completely took in advantage. I would suggest performing meta-analysis by subgroups (i.e. continuous vs. interval, vigorous vs moderate intensity, more vs less than 180min/week; more vs less 10000kcal of caloric expenditure). Similarly, due to ≤ 12 weeks it is a dichotomic variable a sub-groups analysis instead a meta-regression would be more interesting (in that way, the mean change of each group would be known) and to include them at least as supplemental material.

Author Response

We would like to thank you and your reviewers for the constructive comments received on our manuscript. We have attempted to fully address all of the comments and have detailed our response to each and signposted where the corresponding changes have been made in the manuscript.

Reviewer’s comment: 1) Methods (line 240). Please, include the reference for the PICOS strategy.

Author’s response: We thank the reviewer and have added this as they recommend.

Reviewer’s comment: 2) Lines 246. Could the authors, if possible, include the specific search strategy (specific terms and Boolean operators; not only "the combination of keywords"), at least in supplementary material. It is an important point for the reproducibility of the search of a systematic review.

Author’s response: We thank the reviewer and have added this as they recommend.

Reviewer’s comment: 3) Footnote of Figure 1. Please, include the references of PRISMA and the Cochrane risk of bias tool.

Authors’ response: We thank the reviewer and have added as suggested.

Reviewer’s comment: 4) Footnotes of Table 2. Please, include all the abbreviations included in the Table 2.

Author’s response: We apologize and have defined the abbreviations and ensured that the missing parameters are included in the footnotes of Table 2.

Reviewer’s comment: 5) Results (lines 320-323). Please, include the references of the "twelve studies evaluated". Include all the refences in all similar situations across the results section.

Authors’ response: We thank the reviewer and have added as suggested.

Reviewer’s comment: 6) Lines 12-15 in the result section. Please, do not repeat the explanation of the abbreviation.

Authors’ response: All mistakes has been corrected.

Reviewer’s comment: 7) Line 30. "results showed that IHTG was significantly reduced after..." because this result it not an association.

Authors’ response: We thank the reviewer and have changed as suggested.

Reviewer’s comment: 8) Figure 3. I would include the groups (aerobic exercise) vs usual care, in the figure, not only in the footnote. Similarly, I would include the name of the outcome in each sub-graph (ALT in A...).

Authors’ response: We thank the reviewer and have changed as suggested.

Reviewer’s comment: 9) Why the authors ordered the included articles in this way in tables and figures? Consider ordering them according to an alphabetical or year of publication criteria. On the other hand, why did the authors make the forest plots and analyses with the mean difference and not with effect size, if the ES was calculated?

Authors’ response: Thank you for that comment, we we have reorganize figures and tables (now it’s alphabetical).

Effect Size can be described  in a broad or  a narrow sense. In a broad sense, this term can be applied to mean differences (MD). In our article we used MD in forest plots.  In a narrow sense the term it can  refer to some normalized (unitless)  measure such as standardised mean diferences (SMD). We decided to use  MD instead of SMD. This approach can be found in many other meta-analyses.  To support our choice, we have cited the paper  entitled  „A Mean Difference is an Effect Size” by Pierre Dragicevic [Project-Team Aviz Research Report n° 9354, July 2020]: The normal implication of the term effect size is that it indicates the size of the difference between the means of the conditions or groups on the dependent variable.

Reviewer’s comment: 10) Line 60 (3.5 section in results). "...between total intervention time and ALT difference or reduction".

Authors’ response: Changes have been made.

Reviewer’s comment: 11) Part 3.5 of the results? Please, rewrite sentence of lines 60-62, it is a little confusing. Did the authors analyze the differences of the different exercise modes in changes in IHTG or BMI? It could be very interesting to perform the analysis to answer the question regarding what kind of exercise (in FITT) has a higher effect in the improvement of MAFLD.

Authors’ response: Thank you for the reviewers comment, we reorganized result and discussion sections. We tried to analyze the differences in the different exercise modes ( e.g. continuous vs. interval, more vs less than 180min/week) in changes in ALT, AST, GGT, IHTG and BMI using meta-regression. Meta-regression analysis demonstrated a significant correlation only between total intervention time and ALT. Other factors that we assessed: type, intensity, volume of the exercise protocol did not significantly impact the magnitude of the AST, GGT, IHTG or BMI. Unfortunately there were only a small number of studies to explore the relationship  between the dose of physical activity and the effectiveness of the intervention (e.g. When we analyzed continuous training vs interval there was only one study from interval training with GGT or HOMA-IR or IHTG measurements).  

Reviewer’s comment: 12) Discussion (line 73). Please, modify the full name by the “ACSM” abbreviation.

Authors’ response: Changes have been made as requested.

Reviewer’s comment: 13) Discussion. Although the author affirmed that “First, to the best of our knowledge, this is the first 130 meta-analysis evaluating the effect of aerobic exercise to categorize training protocols according to FITT (frequency, intensity, time, type) principle recommended by ACSM, on liver function outcome”, the role of the different components of FITT on liver function has not been completely took in advantage. I would suggest performing meta-analysis by subgroups (i.e. continuous vs. interval, vigorous vs moderate intensity, more vs less than 180min/week; more vs less 10000kcal of caloric expenditure). Similarly, due to ≤ 12 weeks it is a dichotomic variable a sub-groups analysis instead a meta-regression would be more interesting (in that way, the mean change of each group would be known) and to include them at least as supplemental material.

Authors’ response: We thank the reviewer, they are correct, because of the small number of studies the role of all components of FITT principle on liver function has not been analyzed. We have therefore changed the discussion and limitation section. We have added whole result of meta-regression in the supplementary file (2).

Reviewer 2 Report

The present form with recommended changes is acceptable for publication.

Author Response

Thank you.