Left Atrial Strain as a Marker of Supraventricular Arrhythmia Risk in Type 2 Diabetes Mellitus

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors conducted a prospective cohort study showing that patients with type 2 diabetes mellitus have subclinical left atrial dysfunction and a higher incidence of supraventricular arrhythmias. Left atrial conduit strain best predicted arrhythmic risk, and its combination with left ventricular ejection fraction improved discrimination.

General comments

This is a manuscript addressing a topic “The change in left ventricular mass after mitral valve replacement surgery”. However, some concerns need to be addressed.

Specific comments

Major comments

1) Line 165: The name of the echocardiographic system and the analysis software should be specified.

2) Line 222: The left atrial filling index is not explained in the Methods section. Please provide a clear definition and describe how it was calculated.

3) Line 312: Elevated filling pressure (E/E’) was not increased in patients with DM. Additional discussion is warranted to explain this finding.

4) Line 264: “On ROC analysis in T2DM, LAScd demonstrated the best single-parameter discrimination for supraventricular arrhythmia.”; Because other parameters (e.g., LASI) have also been reported to predict arrhythmia, presenting the AUC values for all echocardiographic parameters would be informative.

Minor comments

1) Lines 69, 89, 95, 100, 162, 237, 321, 323, 337: “atrial fibrillation” should be abbreviated as AF.

2) Lines 80, 97: “left LA” should be corrected to left atrium (LA).

3) Lines 83, 94, 101, 119, 122, 168, 172, 276, 297, 332, 351, 352, 355: “left atrium” should be abbreviated as LA.

4) Figure 1: The patient’s name and date should be removed.

Author Response

Dear Reviewer,

Our team would like to sincerely thank you for the positive manuscript appreciation and for all your valuable suggestions. We have carefully addressed your comments and have revised our manuscript accordingly. Revised portions are highlighted in the text.

We hope these responses clarifies your concerns, and we remain available for any further questions.

Best regards,

All authors

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript addresses an important and clinically relevant topic: the role of left atrial (LA) strain as a marker of supraventricular arrhythmia risk in patients with type 2 diabetes mellitus (T2DM) without overt structural heart disease. The prospective design, standardized echocardiographic assessment, and focus on speckle-tracking–derived LA mechanics are strengths, and the paper is generally well written and clearly structured. However, there are several methodological, interpretative, and presentation issues to revise before the manuscript can be considered for publication.

1/ Abstract:

• Include the number of events (e.g., “supraventricular arrhythmias occurred in 20/57 (35.1%) of T2DM and 1/50 (2.0%) of controls”).

• Explicitly state that arrhythmias were assessed by 24 hour Holter at 12 months.

• In the Conclusion, mention the need for external validation.

2/ Introduction

• The Introduction cites recent work from the same group and others on LA strain and atrial myopathy in T2DM, but the incremental novelty of this study should be more clearly articulated. Currently, the “original contribution” is somewhat implicit rather than explicit.

• Consider adding 1–2 sentences at the end of the Introduction that clearly state: (i) what is already known (LA strain is altered in T2DM and linked to AF), and (ii) what this study adds (prospective, Holter-based supraventricular endpoints; focus on LAScd and simple risk models in patients without overt structural heart disease).

• Consider adding a brief, explicit definition of LAScd and why conduit function might be particularly relevant to arrhythmogenesis.

• A brief statement in the Introduction or Discussion about the broader “disease burden” and potential implications for screening and prevention at a population level might improve alignment with a disease-oriented journal audience.

3/ Study design and population

• The study is described as “prospective, single-center observational,” with 107 participants, 57 with T2DM and 50 controls. It would be helpful to clarify:

  • Whether this is a consecutive sample or a convenience sample.

  • The setting from which controls were recruited (e.g., outpatient clinics, screening programs, hospital staff) and whether they were matched beyond age (e.g., sex, hypertension).

  • Clarify whether all echocardiograms and strain analyses were performed on the same vendor and software version, and whether intra- and inter-observer variability were assessed (even in a subset).

• Exclusion criteria are appropriate, but please clarify how “significant arrhythmias” at baseline were ruled out (e.g., only ECG, or also prior Holter/clinical history) to avoid inclusion of patients with pre-existing supraventricular arrhythmias.

• The primary endpoint combines AF/AFL≥30 sec and “excessive supraventricular ectopy” (≥720 PACs/24 h and/or runs ≥20 PACs). This composite is clinically plausible but heterogeneous; the rationale for this definition should be expanded in the Methods and briefly revisited in the Discussion.

• “Data were collected between August 2024 and November 2025.” Verify that this timeframe is realistic relative to the 12 month follow-up and the number of included participants; if some had shorter follow-up, this should be explained. Specify how missing data (if any) were handled.

4/ Statistical analysis and model robustness

• Firth penalized logistic regression is appropriate given the small number of events, but the number of events (n=21 total; 20 in T2DM, 1 in controls) is very limited for multivariable modelling. The authors should:

  • Explicitly report the number of events used for each model and respect events-per-parameter considerations.

  • The authors report p-values for many comparisons; consider also reporting effect sizes and 95% CIs more consistently, including for key echo differences (e.g., LAScd, LASI) to aid clinical interpretation. The description of LAScd–LAVI correlations would benefit from adding 95% CIs for correlation coefficients and explicitly stating the sample used (full cohort vs T2DM only).

• It would strengthen the analysis to:

  • Clarify whether BP and glycaemic control (e.g., HbA1c) during follow-up were considered as potential confounders.

5/ Results/Discussion

• The Discussion appropriately frames findings within the concept of “diabetic atrial myopathy.” However, the clinical implications of using LAScd in practice need clearer boundaries:

  • Given the modest AUC and low sensitivity at the chosen cut-off, the authors should clarify whether LAScd is best seen as a risk enrichment tool to select patients for more intensive rhythm monitoring rather than as a diagnostic test.

  • The potential incremental value of LAScd beyond standard clinical variables (e.g., age, hypertension, duration of diabetes, LA size) is not fully quantified; this should be more explicitly acknowledged as a limitation.

  • Given that arrhythmias were only assessed with a single 24‑hour Holter at 12 months, the manuscript should more clearly acknowledge the possibility of under-detection of paroxysmal AF and low-burden arrhythmias, and the corresponding impact on risk estimates. This is mentioned in the limitations but could be more prominently highlighted and linked to the interpretation of the relatively modest AUC.

  • Emphasize that the LAScd+LVEF model is exploratory and at high risk of overfitting, and should be validated externally before use in practice.

  • ROC-derived cut-offs:

    • The LAScd cut-off of −8% (sensitivity 55.6%, specificity 81.8%, AUC 0.692) and the multivariable model AUC 0.772 are reasonable but not outstanding.

    • The authors should temper statements such as “LAScd emerged as the most clinically informative marker” and clearly communicate the moderate discrimination and modest sensitivity, particularly for use as a clinical screening tool.

  • Hypertension and dyslipidaemia are highly prevalent and may influence both LA remodelling and arrhythmia risk. Although baseline characteristics are “broadly similar”, dyslipidaemia at 12 months differs significantly and may reflect differential treatment or risk factor control.

  • Explore a minimal adjusted model within T2DM including at least one clinical covariate (e.g., age or hypertension) in addition to LAScd, or at least discuss why this was not feasible due to sample size constraints.

  • Table 1:

    • Consider separating clinical and echocardiographic variables more clearly with subheadings (the current table is long and dense).

    • Ensure minus signs for strain values are formatted (“−11.6 ± 4.2” vs “-11.6 ± 4.2”).

• The discussion of mechanisms (fibrosis, RAAS activation, etc.) is appropriate but could be slightly shortened to avoid redundancy with the Introduction.

• When stating that there are “no data evaluating the association between LAVI and LAScd and their prognostic value,” ensure that this remains true in light of very recent literature; if this is based on a non-systematic search, it may be safer to phrase as “To our knowledge, few/no studies have…”

• The authors might briefly relate these findings to screening strategies in T2DM and comment on how LAScd could be integrated with existing risk scores or biomarkers (NT-proBNP, etc.).

• The limitations section could be expanded to include:

• Lack of systematic assessment of antidiabetic and cardiovascular medication changes during follow-up (which may influence LA remodeling and arrhythmia burden).

• Possible selection bias in the recruitment of controls.diseases

Comments on the Quality of English Language

• The manuscript is generally written in good academic English, but a careful language edit would improve fluency and consistency. Examples:

  • “type 2 diabetes mellitus (T2DM) is responsible for 94·0%…” → consider using standard decimal punctuation (e.g., 94.0%) consistently.

  • “atrial myopathy involves atrial remodeling that occurs at the structural level…” → could be streamlined to avoid repetition.

  • Ensure consistent use of terms: “left atrial strain” vs “LA strain,” “supraventricular arrhythmia” vs “supraventricular tachyarrhythmia.”

• In the abbreviations list, “OD” should be corrected to “OR” (odds ratio).

Figure 2 and 3:

• Include the number of observations (n) in the legend or on the plots.

• Consider adding 95% CI bands to the regression line in Figure 2 if available.

Author Response

Dear Reviewer,

Our team would like to sincerely thank you for the positive manuscript appreciation and for all your valuable suggestions. We have carefully addressed your comments and have revised our manuscript accordingly. Revised portions are highlighted in the text.

We hope these responses clarifies your concerns, and we remain available for any further questions.

Best regards,

All authors

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript by Benchea and col. presents the results of a prospective, single-center observational study evaluating speckle-tracking echocardiography Left atrial strain parameters as predictors of supraventricular arrhythmias in patients with type 2 diabetes mellitus (T2DM) without overt structural heart disease.

The authors demonstrate that LA strain is impaired in T2DM and Left atrium conduit strain provides the best discrimination for supraventricular arrhythmic risk at 12-month follow-up.

The topic is clinically relevant and current, the methodology is appropriate, and the results are generally well presented.

The manuscript adds to the knowledge in the field of diabetic atrial myopathy and arrhythmic risk stratification.

I do not have major comments regarding the paper.

As a minor comment, I suggest mentioning the vendor/software for echocardiography and Holter.

Author Response

Dear Reviewer,

Our team would like to sincerely thank you for the positive manuscript appreciation and for all your valuable suggestions. We have carefully addressed your comments and have revised our manuscript accordingly. Revised portions are highlighted in the text.

We hope these responses clarifies your concerns, and we remain available for any further questions.

Best regards,

All authors

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have corrected the manuscript according to the reviewer's comments.

Author Response

Dear Reviewer,

Our team would like to sincerely thank you for the positive manuscript appreciation.

Best regards,

All authors

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have responded constructively to the previous review and have implemented the majority of requested changes in a satisfactory manner. The manuscript is now scientifically sound, clearly presented, and appropriately.

Minor suggestion:

1. Consider adding one short phrase on the clinical roleof LAScd (e.g., “as a risk-enrichment marker rather than a stand-alone diagnostic test”) to further temper expectations already nuanced later in the text.
2. The text notes the limited events and restrictions on model complexity, but does not always explicitly state the number of events used in each model within the Results paragraph describing the logistic regressions
3. Adding 95% CIs for LAScd and LASI differences (and possibly LASr) in the Results would align better with modern reporting standards and the earlier request.
4. If feasible, including a 95% CI band around the regression line in Figure 2 would visually reinforce the correlation strength, though this is optional.
5. Some mechanistic paragraphs (myopathy, RAAS, fibrosis, etc.) are still partly redundant with the Discussion; a small reduction in overlapping content would further improve flow and concision

Author Response

Dear Reviewer,

Our team would like to sincerely thank you for the positive manuscript appreciation and for all your valuable suggestions. We have carefully addressed your comments and have revised our manuscript accordingly. Revised portions are highlighted in the text.

Q1: Consider adding one short phrase on the clinical roleof LAScd (e.g., “as a risk-enrichment marker rather than a stand-alone diagnostic test”) to further temper expectations already nuanced later in the text.

R1: We agree and have clarified the clinical role of LAScd to avoid overinterpretation (lines 97-100).

Q2: The text notes the limited events and restrictions on model complexity, but does not always explicitly state the number of events used in each model within the Results paragraph describing the logistic regressions.

R2: We explicitly stated the number of supraventricular arrhythmia events used to support the regression modelling (lines 295-296).

Q3: Adding 95% CIs for LAScd and LASI differences (and possibly LASr) in the Results would align better with modern reporting standards and the earlier request.

R3: We revised the Results section to report effect sizes with 95% confidence intervals for the key left atrial strain measures (LAScd and LASI, and also LASr), in addition to p-values, to improve interpretability and align with current reporting standards (lines 249-262).

Q4: If feasible, including a 95% CI band around the regression line in Figure 2 would visually reinforce the correlation strength, though this is optional.

R4: We agree this could improve visual interpretation. Because the figure is currently based on aggregated reporting, we addressed this request by adding 95% CIs for the correlation coefficient (r) directly in the Results text, which quantifies uncertainty around the association. This provides similar interpretive value even without a shaded band.

Q5: Some mechanistic paragraphs (myopathy, RAAS, fibrosis, etc.) are still partly redundant with the Discussion; a small reduction in overlapping content would further improve flow and concision.

R5: We agree and have streamlined redundant mechanistic content to improve clarity and concision while preserving essential pathophysiologic context (lines 333-395).

We hope these responses clarifies your concerns, and we remain available for any further questions.

Author Response File: Author Response.docx