Exploring the Association of FTO rs9939609 with Type 2 Diabetes, Fasting Glucose and HbA1c in a Southeastern Mexican Region of Predominant Mayan Genetic Background

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In the manuscript entitled "Exploring the Association of FTO rs9939609 with Type 2 Diabetes, fasting glucose and HbA1c in a Southeastern Mexican Region of Predominant Mayan Genetic Background", the authors evaluate the association of this SNP with T2D in a population from Yucatan, Mexico. These are my comments to improve in the manuscript.

Introduction

Lines 35–36: The diagnosis of type 2 diabetes involves more than just hyperglycemia
Numerous papers discuss the link between FTO and type 2 diabetes in Mexican populations, including Maya youngsters from Mexico. Kindly incorporate this evidence into your explanation of the research's novelty.

Methodology

Mayan ancestry is not assured by birthplace in Merida. This study's major drawback is the absence of ancestry markers. Why not include a subanalysis that includes the data from the controls that have Maya heritage markers?

I ask that your study be approved by the Office of the Ethics Committee. 

The T2D cases in general are an heterogenous population. Please indicate the reason to adjust by age, BMI, and sex in the multivariate analysis. I suggest to include the clinical treatment, glucose control, and the period of disease progression as adjustment variables for the examination of the relationship between rs9939609 and HbA1c and fasting glucose. Please include any variable that might affect this association in the analysis. Only age, BMI, and sex adjustments are included in your study, which is limiting.   Results The title of Table 1 is absent. Please mention any traits that might affect your HbA1c and glucose levels.
Additionally, list the units, acronyms, and represented values (mean, median, standard deviation, etc.) in each table.
I propose that tables two and three be combined into a single table.
The analysis of T2D progression, management, and glucose regulation should be presented in Table 4.   Discussion In my opinion, the association of the rs9939609 with HbA1c reported by the authors is marginal. In this sense, please improve this section based on the factors that reduce or interfere with the association of the polymorphism and glucose control. 

Author Response

In the manuscript entitled "Exploring the Association of FTO rs9939609 with Type 2 Diabetes, fasting glucose and HbA1c in a Southeastern Mexican Region of Predominant Mayan Genetic Background", the authors evaluate the association of this SNP with T2D in a population from Yucatan, Mexico. These are my comments to improve in the manuscript.

R= Thank you for taking the time to review our manuscript, we really appreciate it.

Introduction

Lines 35–36: The diagnosis of type 2 diabetes involves more than just hyperglycemia

R= Thak you for pointing this out. We have rephrased lines 35-36 (now 34-37) to emphasize the complexity of type 2 diabetes. However, we point out that hyperglycemia is a key criterion of diagnosis, which is often assessed with fasting glucose, oral glucose tolerance tests or HbA1c levels.

Numerous papers discuss the link between FTO and type 2 diabetes in Mexican populations, including Maya youngsters from Mexico. Kindly incorporate this evidence into your explanation of the research's novelty.

R= Several studies have explored the association between FTO variants and metabolic traits in Mexican populations. However, to our knowledge, in southeastern populations there are no published reports showing significant associations between FTO rs9939609 and clinically diagnosed type 2 diabetes, fasting glucose (as a quantitative trait), or HbA1c. However, relevant studies have now been cited in the introduction (lines 71-77), including those reporting associations with metabolic syndrome and related traits in Northern and Western Mexican populations, as well as a study evaluating FTO in relation to HOMA-IR and HOMA-B in Mayan children, although no statistically significant results were found in the latter (PMID: 40532257). Additionally, the reference by Núñez Ortega et al. (PMCID: PMC8403373), previously cited in the discussion, which reported an association with hyperglycemia (defined as fasting glucose ≥100 mg/dL), has now been included in the introduction as well. Overall, we believe that the associations presented in our study, namely with diagnosed T2D, fasting glucose, and HbA1c in a Southeastern Mexican population, offer novel insights into the existing literature.

 

Mayan ancestry is not assured by birthplace in Merida. This study's major drawback is the absence of ancestry markers. Why not include a subanalysis that includes the data from the controls that have Maya heritage markers?

R= We acknowledge that birthplace in Mérida does not guarantee Mayan ancestry, and we have carefully phrased throughout the manuscript that the study was conducted in a region where Mayan ancestry is predominant. Regarding the suggestion to perform a subanalysis including only individuals with confirmed Mayan ancestry, we chose not to pursue this due to limited statistical power. In total, only 37 individuals with confirmed ancestry were included in the present study. Although we previously stated 45, this has now been corrected (lines 97-98 and 269). For this study we only included individuals who have all the variables presented (now mentioned in lines 108-109).

Additionally, a previous study (PMID: 30130595) compared 45 San José Tecoh controls (of which 37 are included in our study) with 47 cases from Sisal, Yucatán. All individuals from this analysis have confirmed Mayan ancestry.  Interestingly, that study identified a trend for the rs9939609-A allele (OR = 1.32, p = 0.080), suggesting a consistent direction of effect with our findings. The statistical significance observed in our study is likely due to the increased sample size. We have now added this point to the discussion (lines 273-279).

I ask that your study be approved by the Office of the Ethics Committee. 

R= We are sorry for not including the approval number from the ethics committee.  It has now been added at lines 313-315 at the Institutional Review Board Statement section.

The T2D cases in general are an heterogenous population. Please indicate the reason to adjust by age, BMI, and sex in the multivariate analysis. I suggest to include the clinical treatment, glucose control, and the period of disease progression as adjustment variables for the examination of the relationship between rs9939609 and HbA1c and fasting glucose. Please include any variable that might affect this association in the analysis. Only age, BMI, and sex adjustments are included in your study, which is limiting.  

R= Thank you for pointing this out. We adjusted for age and sex as they are standard demographic variables known to influence metabolic traits and disease risk, including type 2 diabetes.  We now explicitly mention this in the methods (lines 128-129). BMI was included to account for adiposity, which is directly relevant to the FTO relationship under investigation and may confound the association between rs9939609 and glycemic traits. For this version, we have also included other adiposity-related measures to further support the robustness of the adjustment (see lines 129-133).

We acknowledge the heterogeneity of T2D cases and appreciate the suggestion to include additional clinical covariates. As recommended, we explored adjusting for age at diagnosis, diabetes medication categories (1. Metformin, 2. Glibenclamide + Metformin, 3. Metformin + other drugs, 4. Glibenclamide, 5. Insulin, 6. Metformin + Insulin, 7. Unknown), and disease duration in our logistic regression models, both as categorical and continuous variables.

However, including these covariates led to convergence issues, as indicated by the warning: glm.fit: algorithm did not converge; fitted probabilities numerically 0 or 1 occurred (picture attached), suggesting model overfitting and instability. These variables are structurally imbalanced between cases and controls, since controls have no diagnosis, no diabetes-related medications, and no disease duration. This imbalance exacerbates separation in the model and contributes to numerical instability. Moreover, these covariates reflect disease progression and treatment history, which occur after diagnosis and may act as downstream consequences of the disease.

To support our decision, we conducted analyses including these covariates (see attached figures in the word document), which showed unstable estimates and poor model fit. In some models, coefficient estimates were infinite or not interpretable, further indicating separation and lack of reliability.

Given our primary aim to assess the direct association between FTO rs9939609 and case-control status, we opted to exclude these covariates to maintain model robustness and interpretability.

Results The title of Table 1 is absent. Please mention any traits that might affect your HbA1c and glucose levels. Additionally, list the units, acronyms, and represented values (mean, median, standard deviation, etc.) in each table.

R= The heading of Table 1 has now been corrected.  We have also clearly added the units, acronyms and what the values represent for all the tables. As mentioned in a previous response we have included traits related to type 2 diabetes, such as waist circumference and waist-to-hip ratio.

I propose that tables two and three be combined into a single table.

R= We have combined tables 2 and 3 (now table 2). Therefore, both genotype frequencies and genotype analysis can be consulted on the same table.  

The analysis of T2D progression, management, and glucose regulation should be presented in Table 4.  

R= As stated in a previous response, these variables were not further included in the present study.

 

Discussion In my opinion, the association of the rs9939609 with HbA1c reported by the authors is marginal. In this sense, please improve this section based on the factors that reduce or interfere with the association of the polymorphism and glucose control. 

R= We appreciate this observation. Although we were unable to include the previously suggested, we now present additive, dominant, and recessive models in this version of the manuscript. Notably, the dominant model showed stronger associations with both fasting glucose and HbA1c levels, being the last actually statistical significant.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript presents a case-control genetic association study in a selected mexican population with a Mayan background, to assess the effect of the FTO rs9939609 polymorphism on the susceptibility to type 2 diabetes.

The FTO rs9939609 polymorphism has been widely studied in obesity and type 2 diabetes and has been already been established as a risk allele for these disorders. Therefore, the present study is essentially a confirmatory study, with little originality or scientific interest. 

Major comment: Although the manuscript is clear and well written, there is a major methodological problem that limits any conclusions form the study. This is the small sample size. This important information is absent in the abstract and in the methods section. It is only in the results section that it becomes apparent that this is a severely underpowered study. Only 92 patients and 92 controls were studied. No power analysis is presented, but it is usual that small genetic risks can only be detected with sample sizes of hundreds of cases and controls.

Minor comments: 1) There is no discussion of the mechanism by which the genetic variant increases risk of disease. 2) Table 1 title has errors and lacks explanation of abbreviations and presented values. 3) Reference 5 is outdated.

 

 

 

Author Response

The manuscript presents a case-control genetic association study in a selected mexican population with a Mayan background, to assess the effect of the FTO rs9939609 polymorphism on the susceptibility to type 2 diabetes.

R= We appreciate the time give for reviewing this manuscript.

The FTO rs9939609 polymorphism has been widely studied in obesity and type 2 diabetes and has been already been established as a risk allele for these disorders. Therefore, the present study is essentially a confirmatory study, with little originality or scientific interest. 

R= We appreciate this observation and agree that the association between FTO rs9939609 and T2D has been previously reported. However, we believe our study contributes to the growing need for genetic and omics research in underrepresented populations, such as those in southeastern Mexico with a predominant Mayan genetic background. While our sample size is modest, it provides valuable data from a region that is rarely included in global genetic studies. This representation is crucial for understanding the transferability and population-specific effects of known risk alleles, and for ensuring that future precision medicine approaches are inclusive and equitable.

Major comment: Although the manuscript is clear and well written, there is a major methodological problem that limits any conclusions form the study. This is the small sample size. This important information is absent in the abstract and in the methods section. It is only in the results section that it becomes apparent that this is a severely underpowered study. Only 92 patients and 92 controls were studied. No power analysis is presented, but it is usual that small genetic risks can only be detected with sample sizes of hundreds of cases and controls.

R= Thank you for this important observation. We agree that the sample size is modest and have now made this more transparent by explicitly stating it in the abstract (lines 26-28). Now in the discussion we clearly mention that, since the sample size was limited, the presented associations should be interpreted with caution because it is likely that our effect sizes reported are either under- or overestimated (lines 288-291).

Nonetheless, we believe our study contributes valuable data from a population that is significantly underrepresented in genetic research. We also note that a previous study (PMID: 30130595) conducted in Yucatán, which included 45 controls and 47 cases, identified a trend toward association for the rs9939609-A allele (OR = 1.32, p = 0.080). Our study, with a slightly larger sample size, observed a statistically significant association in the same direction, suggesting that the our modest sample size increase was enough to detect the association even in smaller cohorts. We have added this comparative point to the discussion (lines 273-279).

There is no discussion of the mechanism by which the genetic variant increases risk of disease.

R= Thank you for this important comment. We have now added a paragraph in the discussion (lines 228–233) addressing the potential biological mechanisms between FTO rs9939609 polymorphism and type 2 diabetes risk. Since the variant is in the first intron of FTO, it could potentially regulate the expression of distant genes such as IRX3 and IRX5, which influence adipocyte differentiation and thermogenesis. Furthermore, FTO is expressed in the hypothalamus, where it may affect appetite regulation. These mechanisms contribute to increased obesity risk, which is a well-established risk factor for type 2 diabetes. However, it remains unclear whether the mechanism behind FTO and type 2 diabetes independently of adiposity, as our results and other literature suggest.

 

Table 1 title has errors and lacks explanation of abbreviations and presented values.

R= Sorry for not noticing this earlier, we have made Table 1 clearer and fixed the issues.

Reference 5 is outdated

R= We respectfully disagree with the assessment that reference 5 is outdated. The citation refers to the American Diabetes Association’s Standards of Care in Diabetes—2024, published in Diabetes Care (Vol. 47, Supplement 1, S20–42) in December 2023. This is one of the most recent and authoritative guidelines available on the diagnosis and classification of diabetes, and it is updated annually by the ADA Professional Practice Committee. We believe it remains appropriate and relevant to support the statement regarding established risk factors for type 2 diabetes.

Reviewer 3 Report

Comments and Suggestions for Authors

This paper tackles a really important question: does the FTO rs9939609 variant play a role in type 2 diabetes for people with Mayan ancestry? Since Mayan ancestry is pretty rare in genetic studies, this research is super timely and worth checking out. But, the study’s small size and some indirect ancestry info mean we should take the results with a grain of salt. The fact that there are an equal number of cases and controls is good, but with only 184 participants, it’s tough to spot tiny genetic effects. It would be great if they could say what the study’s likely effect size would be. Using birthplace to guess Mayan ancestry is fine, but it’s not the same as using genetic markers. They should make that clear in the abstract and discussion.

The diabetes diagnosis criteria seem good, but it’s not clear if glucose and HbA1c were measured the same way for everyone, or if some people were already diagnosed and treated, which could mess with the results. Adding BMI as a factor is a good idea, but FTO is super linked to body fat, so waist circumference or waist-to-hip ratio would give a clearer picture. If those numbers aren’t available, they should say that’s a limitation.

From a statistical point of view, the additive genetic model is standard, but since not many people have two copies of the risk allele, the model might not be super stable. Checking out dominant or recessive models could give us more confidence. If the results have wide confidence intervals and are borderline significant, they should be described as early signs that need more testing. The mediation analysis is a bit much for such a small group; while the methods are solid, the estimates are all over the place and should be seen as exploratory.

The results tables are pretty clear, but let’s polish them up a bit! Could we add units for BMI, glucose, and HbA1c? Also, it would be great to show allele frequencies with percentages and confidence intervals. For the regression tables, please include the sample size for each model, as exclusions might change it. Since the mediation analysis is exploratory, let’s move those results to supplementary material so we can focus the main text on the most important stuff.

The idea that FTO acts “beyond adiposity” is super interesting, but maybe we should tone it down a bit since we haven’t made many adjustments. The association still held up after we accounted for BMI, but the biological pathways are still a bit of a mystery. Let’s put the results in the context of what’s already known in the global research world, like previous meta-analyses and GWAS findings. And just a heads-up: we really need larger studies, ideally in Mayan or other Indigenous groups, before we can make any strong conclusions.

The abstract could be a bit more focused: let’s highlight the main finding—that there’s an association with type 2 diabetes that’s independent of BMI—and mention the main limitations: the small sample size and the fact that we don’t have direct ancestry markers. The discussion should stick to the data and avoid jumping to conclusions. Also, let’s use consistent abbreviations, like using BMI throughout.

Overall, this is a really important piece of work that adds to the evidence in a population that’s often overlooked in genetic studies. The preliminary findings are interesting, but let’s be more measured in our interpretation, and make the limitations crystal clear. With these tweaks, the paper could really make a valuable contribution.

Author Response

This paper tackles a really important question: does the FTO rs9939609 variant play a role in type 2 diabetes for people with Mayan ancestry? Since Mayan ancestry is pretty rare in genetic studies, this research is super timely and worth checking out.

R= Thank you for the encouraging comment and for your time given to review this article.

But, the study’s small size and some indirect ancestry info mean we should take the results with a grain of salt. The fact that there are an equal number of cases and controls is good, but with only 184 participants, it’s tough to spot tiny genetic effects. It would be great if they could say what the study’s likely effect size would be.

R= Thank you for this thoughtful comment. Based on a post hoc power analysis using our sample size (92 cases and 92 controls), a minor allele frequency of 15.2% (as observed in our population), and a significance level of 0.05, our study has 80% power to detect odds ratios of approximately 2.3 or higher under additive or dominant models, and 4.5 or more for the recessive model. We acknowledge that this limits our ability to detect small to moderate genetic effects, which are more typical for common variants associated with complex traits such as T2D.

However, previous research has suggested that the genetic effects of FTO may be detectable even in relatively small samples. For instance, a prior study (PMID: 30130595) including 45 Mayan controls and 47 Mayan cases reported a trend toward association for the rs9939609-A allele (OR = 1.32, p = 0.080), which we were able to detect as statistically significant in our study, likely due to our slightly larger sample size. We have added this comparison to the Discussion section to better contextualize our findings considering the aforementioned result (lines 273-279).

Given that our study is underpowered to detect small effect sizes typical of common variants, we have opted not to include the full power calculation in the manuscript. Nonetheless, we acknowledge this limitation explicitly in the Discussion (lines 287-291), noting that the observed effect sizes presented may be overestimated or underestimated due to the limited sample size.

 

Using birthplace to guess Mayan ancestry is fine, but it’s not the same as using genetic markers. They should make that clear in the abstract and discussion.

R= Thank you for your comment. We have amended the abstract per your recommendations, and the potential implications of ancestry and the presented results was expanded in lines 265-279 of the discussion.

The diabetes diagnosis criteria seem good, but it’s not clear if glucose and HbA1c were measured the same way for everyone, or if some people were already diagnosed and treated, which could mess with the results.

R=All cases had previously confirmed type 2 diabetes, however we did measure all the biochemical measurements with the same equipment. We have added a sentence in the methods (lines 108-109) to make this clear.

Adding BMI as a factor is a good idea, but FTO is super linked to body fat, so waist circumference or waist-to-hip ratio would give a clearer picture. If those numbers aren’t available, they should say that’s a limitation.

R= As suggested, we have included waist circumference and waist-hip ratio as variables in our study. However, as now detailed in the Methods section (lines 129-133), we opted to construct separate models including BMI, waist circumference, or waist-hip ratio individually, rather than including all three in the same model. This approach was chosen to avoid redundancy, as these variables all serve as proxies for adiposity. Overall, the associations observed with the FTO variant appear to be independent of adiposity, regardless of the specific measurement used.

From a statistical point of view, the additive genetic model is standard, but since not many people have two copies of the risk allele, the model might not be super stable. Checking out dominant or recessive models could give us more confidence. If the results have wide confidence intervals and are borderline significant, they should be described as early signs that need more testing.

R= Thank you for this valuable observation. As suggested, we have now included both dominant and recessive models for all associations presented. Interestingly, the additive and dominant models yielded consistent results, with stronger associations observed in the dominant mode; for example, the association with HbA1c reached statistical significance only under this model. In contrast, the recessive model consistently showed non-significant results, which is expected given the low frequency of the TT genotype in our sample. These additional models help reinforce the robustness of our findings, although we agree that all results should be interpreted with caution due to the limited sample size.

The mediation analysis is a bit much for such a small group; while the methods are solid, the estimates are all over the place and should be seen as exploratory.

R= Thank you for this important observation. We agree that, given the limited sample size, the mediation analysis should be interpreted with caution. As now clarified in the revised Discussion section (lines 259-263), our intention was not to infer causality but to explore potential relationships and generate hypotheses for future research. Based on the consistency and strength of the associations observed, we chose to focus on the dominant model for the mediation analysis. This model yielded more stable coefficients and stronger effect sizes across outcomes, which supports its use in this exploratory context. We have emphasized that these findings are preliminary and require validation in larger, independent cohorts (lines 258-264). Of note, the mediation analysis was performed using the percentile method to calculate confidence intervals, rather than the bias-corrected and accelerated (BCa) method, as the percentile approach provided more interpretable intervals, particularly for the proportion mediated part (see Methods, lines 146-148). Furthermore, we now do not adjust for other covariates, as only T2D status seems to be relevant for this model

The results tables are pretty clear, but let’s polish them up a bit! Could we add units for BMI, glucose, and HbA1c? Also, it would be great to show allele frequencies with percentages and confidence intervals. For the regression tables, please include the sample size for each model, as exclusions might change it.

We have added the requested units, particularly in Table 1 where they were previously missing. As suggested by another reviewer, we have merged the allele frequency table with the association results (now Tables 2 and 3), which we hope improves readability and helps contextualize the findings. Allele frequencies were already presented as n (%) at the beginning of the table. Regarding the regression models, we did not include sample sizes in the tables because all individuals had complete data for the covariates used. This has now been clarified in the Methods section (lines 108-109).

Since the mediation analysis is exploratory, let’s move those results to supplementary material so we can focus the main text on the most important stuff.

R= We respectfully disagree with the suggestion to move the mediation analysis to the supplementary material. Given that the manuscript is relatively concise, we believe there is sufficient space to present these results in the main text. As noted in the revised Discussion lines 258-261, we clearly state that the mediation analysis is exploratory and intended for hypothesis generation rather than causal inference. We feel that keeping these results in the main manuscript helps illustrate relationships worth investigating in future studies.

The idea that FTO acts “beyond adiposity” is super interesting, but maybe we should tone it down a bit since we haven’t made many adjustments. The association still held up after we accounted for BMI, but the biological pathways are still a bit of a mystery. Let’s put the results in the context of what’s already known in the global research world, like previous meta-analyses and GWAS findings. And just a heads-up: we really need larger studies, ideally in Mayan or other Indigenous groups, before we can make any strong conclusions.

R= We appreciate the reviewer’s comment and agree that the idea of FTO acting beyond adiposity is still not fully understood. As we now highlight in the revised Discussion (lined 228-238), most existing research links FTO to T2D risk primarily through its effects on adiposity. To date, there is no clear mechanistic explanation for how FTO may influence T2D independently of adiposity. We have retained the statement in the manuscript because the associations persist regardless of the adiposity-related measurement used (BMI, waist circumference, or waist-hip ratio). However, we have now added a note emphasizing the need for larger studies, particularly in Mayan or other populations (lines 238-240), to confirm these findings.

The abstract could be a bit more focused: let’s highlight the main finding—that there’s an association with type 2 diabetes that’s independent of BMI—and mention the main limitations: the small sample size and the fact that we don’t have direct ancestry markers. The discussion should stick to the data and avoid jumping to conclusions. Also, let’s use consistent abbreviations, like using BMI throughout.

R= Thank you for your comment, as mentioned previously we revised the abstract and the discussion. We now try to be consistent with the abbreviations as well.

Overall, this is a really important piece of work that adds to the evidence in a population that’s often overlooked in genetic studies. The preliminary findings are interesting, but let’s be more measured in our interpretation, and make the limitations crystal clear. With these tweaks, the paper could really make a valuable contribution.

R= Once again thank you for your valuable input that have helped us to improve our manuscript! We hope that the current version presents better the obtained finings.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The main problem of the study remains, which is the very small sample size and lack of power analysis that would provide important information about the necessary sample size to detect a significant association. This limits any useful conclusion from the study.

Minor comment: ref 5 is outdated because (as the authors acknowledge in their response) the ADA Standards of Care in Diabetes are published every year. This means that there is a more recent version (jan2025) that could have been cited.

 

Author Response

Thank you for your time given to review our manuscript, your feedback has been very helpful to improve it. Below, I answered the points raised on the last revision of the paper.

The main problem of the study remains, which is the very small sample size and lack of power analysis that would provide important information about the necessary sample size to detect a significant association. This limits any useful conclusion from the study.

Thank you for your comment. We agree that the sample size is modest and have acknowledged this limitation in the previous version of the manuscript. While we did not perform an a priori power analysis, we conducted a post hoc assessment using the Genetic Association Study (GAS) Power Calculator tool (R package), applying our current sample size (n = 184), the observed minor allele frequency in all the population (15.2%), and the reported prevalence of type 2 diabetes in Mexico (18.2%) as parameters. This analysis indicates that our study has approximately 78% power to detect an association under the dominant model and 77% under the additive model, using the unadjusted odds ratios (2.09 and 1.88, respectively). Furthermore, we have 83% power to detect the strongest odds ratio reported for the dominant model (2.18) and 81% for the additive model (1.93), based on the analysis adjusted for waist-hip ratio. Although we chose not to include this post hoc analysis in the manuscript to avoid retrospective justification, we appreciate the opportunity to clarify that the study appears to be sufficiently powered (around 80%) to detect the observed associations. The review history will be available, so the data discussed here will be accessible to readers regardless.

Minor comment: ref 5 is outdated because (as the authors acknowledge in their response) the ADA Standards of Care in Diabetes are published every year. This means that there is a more recent version (jan2025) that could have been cited.

Thank you for this comment, and we are sorry to have overlooked this. We have updated the reference accordingly. 

Reviewer 3 Report

Comments and Suggestions for Authors

Dear authors, your responses demonstrate statistical correctness, conceptual rigor and transparency. Congratulations! All my concerns have been adequately and thoughtfully addressed. Now the manuscript conveys its findings with appropriate caution, clear acknowledgment of limitations and improved methodological coherence. I have no further comments.

Author Response

Dear authors, your responses demonstrate statistical correctness, conceptual rigor and transparency. Congratulations! All my concerns have been adequately and thoughtfully addressed. Now the manuscript conveys its findings with appropriate caution, clear acknowledgment of limitations and improved methodological coherence. I have no further comments.

We truly appreciate your encouraging comments, and we are sincerely grateful for your valuable feedback, which greatly helped us improve our manuscript. It was genuinely a pleasure to have you involved in the review process!