All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
Thanks for clarifying the remaining points.
[# PeerJ Staff Note - this decision was reviewed and approved by Vladimir Uversky, a PeerJ Section Editor covering this Section #]
The original reviewer of your manuscript has examined the changes. S/He is of the opinion that while the changes are appropriate, there are still a few areas which require elaboration. These are detailed in the review.
The reviewer and I feel that could be addressed by relatively minor revisions of the text.
Please respond to these suggestions and outline changes in a cover letter.
The authors have had the manuscript proof read.
Statistical Analysis (Lines 150 – 161) needs some additional work:
Line 152: Continuous variables were assessed for normality using the nonparametric test.
There is no single non-parametric test, there are non-parametric tests for all sort of hypotheses. The query here was with regard to which test was actually used – a name or a reference is all that is required.
Line 153: Variables with a non-normal distribution were expressed as the median (interquartile range).
Better would be:
Variables with a non-normal distribution are summarised by the median and interquartile range.
Line 153 – 155:
Then, the difference of variables with a non-normal distribution between two groups was assessed by the Mann-Whitney test, the categorical data were assessed by the chi-squared test, and the relationship between two continuous variables was assessed by the Spearman correlation coefficient.
The “Then” at the start of the sentence is not needed. Non-normal distribution also applies to categorical variables. Better to re-state with something like:
Differences between two groups of continuous variables with non-normal distributions were assessed by the Mann-Whitney test. Differences between categorical variables were assessed by the chi-squared test. The relationship between two continuous variables were assessed by the Spearman correlation coefficient.
Line 157: Logarithmic transformation was performed to generate data with a normal distribution.
Was the transformed data assessed for normality?
While I realise that duration is just one aspect of a hyperglycemia episode it is used in the paper's title and to convey the results and conclusions.
The authors state in their rebuttal:
"However, our conclusion has not changed."
The conclusions (line 58 - 62) is:
The serum I-FABP level was positively associated with the duration of
hyperglycemia and glycemic variability but negatively associated with islet beta-cell function in type 2 diabetes patients; moreover, the serum I-FABP level was higher in patients with retinopathy than in those without retinopathy, suggesting that the IB dysfunction got worse with
the progression of diabetes.
Let me explain the issue I have with this statement.
1) for the inpatients there is a (statistical) relationship between duration and I-FAB,
2) for the outpatients here is no evidence of a relationship.
As the sub-groups have been pre-defined, this is a planned finding for the experiment.
When the data is combined there is statistical evidence of a relationship, but we know that for the outpatients no such relationship has been established.
This may occur in the combined group for a number of reasons - group sizes, effect sizes etc. So, while there is statistical evidence (i.e. small p-value) this does not change the outpatient result from the outpatient subgroup analysis and therefore, despite the statistical evidence, we cannot say that the relationship exists for all patients. This needs to be reflected in the conclusions and results.
Are there are systematic differences between the groups that are not captured in the analysis of all patients together? It appears that the analysis for all patients is confounded in some manner.
In my opinion, it is not enough that this is discussed at lines 365 – 367 / 321 – 322, it needs to be acknowledged in the Conclusions and Results (and possibly also the title).
For example: You could state that when all patients are combined there is some evidence of a relationship but on further investigation it turns out that there appears to be different behaviours in each subgroup with some further discussion.
Given that there are 3 groupings, inpatients, outpatient and all patients, why are only the results for inpatient and all patients reported, the result for outpatient is of interest too. You need to mention the results for all groupings and any discussion of “all” patients with regard to duration needs to acknowledge this.
I would recommend that the following need to be changed to reflect this difference between inpatient/outpatient/all:
Conclusions (line 58 - 62)
Results (46 - 56) Only inpatient and all patient results are reported - outpatient results are need too.
Discussion (209-210)
I would like to thank the authors for the revisions made in respect to the comments on the first revision of the article. Many of the issues raised in the comments have been resolved.
However, I feel there are still a couple of remaining issues that need to be looked at, detailed in the Validity of findings. These are related to how the results are reported.
As you will see, all reviewers agree that the work is interesting and should be published; however, it is also clear that a number of important issues need to be addressed.
Please pay particular attention to the statistical analysis comments from reviewer-2; these need to be individually addressed or clearly rebutted in your re-submission.
Reviewer-1 and to a lesser extent reviewer-3 have some scientific questions which should be fairly straight-forward for you to deal with. However, it is important that each of the points they raised is addressed clearly in the manuscript, either via new data or through commentary in the discussion.
I look forward to seeing your re-submission. Given the nature of these reviews, the re-submission will probably need to go back to the reviewers for inspection.
[# PeerJ Staff Note: Please ensure that all review comments are addressed in a rebuttal letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate. It is a common mistake to address reviewer questions in the rebuttal letter but not in the revised manuscript. If a reviewer raised a question then your readers will probably have the same question so you should ensure that the manuscript can stand alone without the rebuttal letter. Directions on how to prepare a rebuttal letter can be found at: https://peerj.com/benefits/academic-rebuttal-letters/ #]
Although the manuscript has been proof red an edited by Medjaden bioscience limited, there are still several sections of the paper that require careful editing for English language. For example in the abstract the statement "however this association disappeared in followed multiple linear regression analysis" needs to be corrected and in the discussion line 293 the word "wildly" needs to be replaced by "widely"
The introduction needs to discuss the sensitivity of I-FABP as a marker of intestinal per me ability or barrier dysfunction rather than as a marker of intestinal epithelial cell damage. We does of the article will wish to know the relative sensitivity of the blood test compared to other biomarkers of permeability such as saccharide absorption tests
The article structure, figures, tables and raw data are appropriate
The conclusion regarding the progression of diabetes is not in place it to the association with duration of hyperglycemia
This research study is within the Aims and scope of the Journal
The research question it is well defined and relevant, however the title needs to specify that the study was conducted exclusively in type 2 diabetes and that type 1 diabetes as well as patient below the age of 18 years were excluded
The investigation is performed in at technically adequate and ethical manner
The methods are sufficiently detailed
The discussion includes appropriate for text regarding the limitations of the study specifically the concomitant pathological conditions that may have been impacting the results in the cohort studied as in patients, and therefore the associated comorbidity may have impacted the result in serum I-FABP. A 2nd major limitation that is acknowledged is the presence of confounders, such as the use of anti diabetes drugs as well as the exclusion of severe nephropathy, which may impact the conclusions of the study.
The underlying data have been provided in the supplementary files specifically the EXCEL spread sheet with all the data available
The conclusions are somewhat suspect that since the only positive finding that remains significant after multi very it analysis is the duration of hyperglycemia, and the conclusion statement of the abstract which suggests that duration of hyperglycemia implies progression of diabetes is not documented in the cohort. For example it is not probe in that duration of type 2 diabetes in this cohort actually caused a more diabetic tropathy or vascular complications.
1. The title needs to specify that the study was conducted exclusively in type 2 diabetes
2. The conclusions are somewhat suspect that since the only positive finding that remains significant after multi very it analysis is the duration of hyperglycemia, and the conclusion statement of the abstract which suggests that duration of hyperglycemia implies progression of diabetes is not documented in the cohort. For example it is not proven that duration of type 2 diabetes in this cohort actually caused a more diabetic tropathy or vascular complications, and in addition, the study shows no association of serum I-FABP with insulin levels which would be an appropriate marker of progression of diabetes..
3. The discussion includes appropriate text regarding the limitations of the study specifically the concomitant pathological conditions that may have been impacting the results in the cohort studied as in patients, and therefore the associated comorbidity may have impacted the result in serum I-FABP. A second major limitation that is acknowledged is the presence of confounders, such as the use of anti diabetes drugs as well as the exclusion of severe nephropathy, which may impact the conclusions of the study.
4. In addition the discussion needs to place in perspective the relative contribution of hyperglycemia to the abnormal levels of I-FABP; thus, if the correlation coefficients with duration of hyperglycemia are approximately 0.33, the variance in serum I-FABP attributable to duration of hyperglycemia is estimated at only approximately 10%, which reduces the importance of hyperglycemia as a risk factor
5. The introduction needs to discuss the sensitivity of I-FABP as a marker of intestinal per me ability or barrier dysfunction rather than as a marker of intestinal epithelial cell damage. We does of the article will wish to know the relative sensitivity of the blood test compared to other biomarkers of permeability such as saccharide absorption tests
6. Although the manuscript has been proof red an edited by Medjaden bioscience limited, there are still several sections of the paper that require careful editing for English language. For example in the abstract the statement "however this association disappeared in followed multiple linear regression analysis" needs to be corrected and in the discussion line 293 the word "wildly" needs to be replaced by "widely"
7. In line 104, use of the word "had" to describe the age of the patients is incorrect, and S1H abbreviation does not appear to be explained anywhere
While the article had a good structure and flow, I would suggest that the manuscript should be proof read again to ensure clarity of expression, correct tenses etc. I have detailed a few examples of what I mean by this:
Line 49-50: “Serum I-FABP is positively associated with the duration of hyperglycemia, which suggested that the IB dysfunction got worse with the progression of diabetes.”
Reviewer comment:
=================
Mixture of tenses here – possibly better as:
“Serum I-FABP is positively associated with the duration of hyperglycemia, which suggests that the IB dysfunction gets worse with the progression of diabetes.”
Lines 63 – 64: “since hyperglycemia has existed long before other diseases happen in diabetic patients,”
Reviewer comment:
=================
This may be better as “since hyperglycemia exists long before other diseases occur in diabetic patients,”
Lines 68 – 69: “Nowadays intestinal barrier (IB) dysfunction induced by hyperglycemia was considered to be”
Reviewer comment:
=================
Here the tenses do not appear to be correct: was/is:
Instead I would suggest:
“Nowadays intestinal barrier (IB) dysfunction induced by hyperglycemia is considered to be”
Lines 69 – 70 “was considered to be the underlying mechanism of systemic infection and inflammatory in diabetes.”
Reviewer comment:
=================
Ii looks to me like there is a word missing after “inflammatory”? Possibly “response”?
Reviewer comment:
=================
Lines 100 – 101: “who had been tested hyperglycemia”
Possibly missing word: “for”
“who had been tested for hyperglycemia”
While it is possible to understand what is being said, clarity would be increased for the reader with some further changes like these examples.
This is an observational study on both inpatients and outpatients with a relatively straightforward statistical analysis. At the centre of the discussion is the relationship between IB dysfunction (as expressed through I-FABP) and duration of hyperglycemia (duration).
The main conclusion is:
Line 49-50: Serum I-FABP is positively associated with the duration of hyperglycemia, which suggested that the IB dysfunction got worse with the progression of diabetes.
There are a couple of points to be made here about the relationship.
One line 74: “Thaiss’ study, hyperglycemia was confirmed to be the independent risk factor of IB dysfunction”
On line 68 – 70 we have: “Nowadays intestinal barrier (IB) dysfunction induced by hyperglycemia was considered be the underlying mechanism of systemic infection and inflammatory in diabetes.”
This seems to indicate that some that there is some sort of causal relationship between hyperglycemia and IB.
The study in this paper states (lines 280 – 281):
“The association between hyperglycemia and IB dysfunction was clarified in our study, but the causal relationship between them is yet to be discussed.”
(Reviewer comment: As it is duration of hyperglycemia that is being investigated, I would expect that this should be ““The association between the duration of hyperglycemia and IB dysfunction…”
So no causal assumptions for the analysis. Just an observational investigation.
Reviewer comment:
==================
In the multiple linear regressions (lines 138, 152) I-FABP is always the response or independent variable and duration is the explanatory variable, which could be considered to at least indicate that the causal direction is as discussed lines 74, 68-70. In particular, in Table 1.3 both duration and age are explanatory variables, with age always affecting I-FABP.
As the causal relationship is not discussed in this paper, when the model is introduced it should be stated more clearly that this is just way of determining possible associations between I-FABP and the other variables – with, for example, some further discussion of how age may effect both I-FABP, duration.
Study participants (97 - 107)
=======================
Were all patients who agreed to the study and met the criteria included in the analysis or did any drop out (as this is a cross-sectional study drop may not be an issue). While the inclusion and exclusion criteria are clearly given I think a small statement about possible bias in the study population would be helpful.
Statistical Analysis (132 - 140)
========================
Lines 133 – 134 “Non-parametric tests were performed to determine the normality of continuous variables, whereas the non-normal distribution variables were expressed as median (interquartile range).”
Reviewer comment:
================
This sentence contains two things: assessing normality and summary statistics for non-normal variables. I would suggest this be changed into 2 sentences, for example: Continuous variables were assessed for normality using the <test-name> test. Non-normal variables are…
Results (Line 142 – 153)
Table 1.1
Reviewer comment:
================
The Male/Female p-value reported is from a Fisher exact test not chi-square test.
Lines 150 – 152 and (Table 1.2)
Reviewer comments:
=================
Some of these correlations are weak, for example FPG is -0.19. I think a comment on the strengths of the relationships is need.
Lines 152 – 153: “Further, multivariate linear regression analysis showed that serum I-FABP was positively associated with age and duration of hyperglycemia in all patients (Table 1.3).”
Reviewer comments:
=================
I-FABP is right-skewed and looks to be non-normal and therefore model assumptions are violated. Any estimates and inferences are likely to be biased.
I suggest you try a transformation of the data to reduce skewness e.g. log transform, and re-run the analysis.
A general question: what model was fit here? No intercept is reported in the Table.
Were all variable including intercept fitted or just duration and age? This is what I would expect. If so then they should also be reported in the Table.
Please state clearly the model.
This model really doesn’t really tell much more than Table 1.2 where duration and age were the two largest positive correlations with I-FABP.
Why might it be the case than some of the variable with correlations to I-FABP are not significant in the model?
The headings for the Table need to be explained. Is B the estimate of the parameter from the model? (This looks to be taken directly from the SPSS output).
Serum I-FABP in outpatient subgroup
===============================
Reviewer comments:
=================
Lines 169 – 171 / Table 3.2 “However, the duration of hyperglycemia was the only metric that statistically associated with serum I-FABP in the followed multiple linear regression 171 analysis (P<.001) (Table 3.2).”
Same comments apply as for the previous regression.
Serum I-FABP in inpatient subgroup
================================
It is not clear how Table 3.3 is formed from the supplementary material.
This needs to be explained further please.
I have concerns here about the overall conclusion statement.
Taken together the data overall does show an association or correlation between I-FAPB and hyperglycemia duration.
However, the two subgroups show different behaviours.
For outpatients all the durations are 0 (or rounded to 0). There is no variation in the data, the correlation can’t be calculated and no (statistical) relationship is established between the two variables.
The case is different for the inpatients where durations have a median of 10 years.
My concern here is that the relationship between I-FAPB and hyperglycemia duration reported for the overall population is driven by the sub-group sizes. One could imagine a situation where, if a larger number of outpatients were recruited, it would eventually mask the signal in the pooled data.
Looking at the outpatient duration, even large values of I-FABP (max 5221.39) have 0 durations, whereas for inpatient the durations tend to be non-zero, with positive association. (Is the rounding of duration for outpatients hiding systematic changes in duration?)
The conclusion needs to be adjusted for the different behaviours in the sub-groups, with discussion of why this is so. Given the 0 durations for the outpatients for high levels of I-FAPB, what is driving the relationship for the inpatients data?
I enjoyed reading the article and found it to be an interesting study. My comments detailed above are mainly in the area of the statistical analysis, where some clarification is needed, with some additional comments on the grammar, but more importantly in the area of the different behaviours of the subgroups and how these should be reported in the findings.
Authors reported that serum I-FABP was significantly associated with duration of hyperglycemia. The manuscript is well written and easy to read. Additionally, the results are interesting but some issues should be solved. I think the data are sound and this manuscript can be published in "PeerJ" after major revision. In the section of introduction, The sentence " Generally, the nature......... to evaluate IB dysfunction in our study" in the line 87-89, please check the citations of references 8 and 22. Are these two references belong to authors' studies ? . Please correct study to studies.
Please add the country name of ELISA kit in the line 129.
Please describe "subtitles "more detail.
1. Please cite the references in the section of discussion in the lines 220-230.
2. Please discuss the role of I-FABP in oxidative stress, inflammatory response, and infection.
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.