Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. Methods:

• Strengths: The patient selection criteria and study design are straightforward. The use of non-parametric tests (Wilcoxon Signed Rank, Mann-Whitney U, Kruskal-Wallis) is appropriate given that the variables were not normally distributed.
• Areas for Improvement: The methodology does not state how missing data (if any) was handled during the retrospective data collection phase. Furthermore, it would be beneficial to clarify if there was a required minimum duration of stable antiretroviral therapy (ART) before the initiation of GLP-1 RAs, as initiating ART itself can lead to weight fluctuations.

2. Results, Tables, and Graphs:

• Results: The findings are reported clearly in the text, highlighting a mean HbA1c reduction of 1.0% and a mean BMI reduction of .
• Tables:
• CRITICAL: In Table 8 ("Factors associated with a percentage of weight loss > 5%"), the Odds Ratio and 95% Confidence Interval are missing for the variable "Months of GLP-1 or GLP-1/GIP RA use" (only the p-value of 0.061 is listed).
• In Table 8, there are formatting issues with the Confidence Intervals (e.g., the hyphen is missing in "1.016 (0.988 1.044)" and others).
• In Table 7, there is a missing hyphen in the Pioglitazone use CI: "0.230 (0.0301.789)". Please review all tables for typographical consistency.
• Graphs/Figures: The manuscript currently relies solely on tables. The addition of at least one graphical representation (e.g., a scatter plot showing HbA1c reduction versus months of GLP-1 RA use, or a boxplot comparing weight loss between insulin users and non-users) would significantly enhance the visual communication of your primary findings.

3. Discussion and Limitations:

• Strengths: The discussion correctly contextualizes the modest weight loss (3.55%) observed in the cohort, reasonably attributing it to the high baseline HbA1c (8.7%) and the inclusion of non-obese patients.
• Areas for Improvement: The authors mention that "adherence challenges are quite common" in the Bronx, possibly dampening the observed weight loss effects. Since medication adherence was not strictly measured, this should be explicitly listed in the formal "Limitations" paragraph alongside the lack of dosage information and the retrospective design.

4. References:

• CRITICAL: Please check your reference list for duplicates. Reference 33 and Reference 38 are completely identical (Eckard AR, Wu Q, Sattar A, Ansari-Gilani K, et al. Once-weekly semaglutide in people with HIV-associated lipohypertrophy...). You must consolidate these citations and update the numbering throughout the manuscript.

  Recommendations for the Authors:

  1. Supply the missing Odds Ratio and 95% CI data in Table 8.
  2. Fix typographical errors in the confidence intervals across Tables 7 and 8.
  3. Remove the duplicated reference (33 and 38) and adjust the in-text citations accordingly.
  4. Consider adding a graph/figure to illustrate the key findings.
  5. Briefly mention how missing data was handled in the Methods section.
  6. Add the lack of measured medication adherence to the formal limitations paragraph.

Author Response

Responses to Reviewer 1

 

Comment 1:

“ Supply the missing Odds Ratio and 95% CI data in Table 8.”

 

Response 1:

We thank the reviewer for identifying this omission. The missing odds ratio and corresponding 95% confidence interval values in Table 8 have now been included to ensure completeness and clarity of the statistical results.

 

Comment 2:

“ Fix typographical errors in the confidence intervals across Tables 7 and 8.”

 

Response 2:

We appreciate the reviewer for identifying these typographical errors. The confidence intervals in Tables 7 and 8 have been carefully reviewed and corrected for accuracy and consistency throughout the manuscript.

 

Comment 3:

“ Remove the duplicated reference (33 and 38) and adjust the in-text citations accordingly.”

 

Response 3:

We thank the reviewer for noting this duplication. The duplicate reference has been removed, and all subsequent references and corresponding in-text citations have been renumbered accordingly.

 

Comment 4:

“Consider adding a graph/figure to illustrate the key findings. “

 

Response 4:

We appreciate this helpful suggestion. To improve the presentation of our results, we have added a new figure. Please see attached

 

 

Comment 5:

“Briefly mention how missing data was handled in the Methods section.”

 

Response 5:

We thank the reviewer for this suggestion. We have revised the Methods section to clarify our approach to missing data by adding the following statement:

 “To address missing data, we employed a complete-case analysis approach. Consequently, only individuals with sufficient clinical measurements for each specific statistical evaluation were incorporated into the study.”

 

Comment 6:

“Add the lack of measured medication adherence to the formal limitations paragraph.”

 

Response 6:

We appreciate the reviewer for highlighting this important limitation. We have revised the Limitations section to explicitly acknowledge that medication adherence was not directly assessed in this retrospective study by adding the following statement:

 

 “  Furthermore, medication adherence was not assessed in this study. Variable adherence to GLP-1 RA therapy may have influenced the observed magnitude of weight loss and glycemic improvement. Therefore, this factor should be considered a potential source of residual confounding.”

Reviewer 2 Report

Comments and Suggestions for Authors

L89: Baseline measurements, including weight, glycated hemoglobin (HbA1c) and lipid panel were collected before and after the treatment. Without a comparator group, it is difficult to determine whether the observed improvements were attributable to GLP-1 therapy, concurrent diabetes management, medication adjustments, or other factors

L97: Patients eligible for inclusion were considered those who had a confirmed HIV diagnosis and who were on any dose of GLP-1 or dual GLP-1/GIP RA treatment for at least 12 weeks. This may introduce survivor bias because patients discontinuing therapy were excluded. This limitation should be adequately discussed.

L141: The ART regimens were divided into two groups. Please clarify if this is an oversimplifies the complexity of ART-associated weight gain as ARTs have distinct metabolic effects

L158: The mean weight reduction was 3.4 kg, from 95.9 kg to 92.5 kg (p<0.001). Although statistically significant, this is still modest. In this situation weight loss should be interpreted cautiously.

L279: It is also worth noting that the study was conducted in the Bronx where medication adherence challenges are quite common; therefore, the modest weight changes may partly reflect variable adherence to therapy. Please comment on how adherence was evaluated.

 

L335: Importantly, our data showed equal reduction in weight, regardless of which ART was used, suggesting that GLP-1 RA use can effectively mitigate concerns about weight gain associated with certain integrase inhibitors. Please provide evidence to support this statement or remove it.

The comparisons are generally appropriate. However, it should be discussed more thoroughly with regard to baseline HbA1c, severity of obesity and dose and duration of GLP-1 treatment

Author Response

Responses to Reviewer 2:

 

Comment 1:

"Baseline measurements, including weight, glycated hemoglobin (HbA1c) and lipid panel were collected before and after the treatment. Without a comparator group, it is difficult to determine whether the observed improvements were attributable to GLP-1 therapy, concurrent diabetes management, medication adjustments, or other factors."

 

Response 1:

We thank the reviewer for highlighting this important limitation. We agree that the absence of a comparator group limits our ability to establish causality between GLP-1 RA therapy and the observed metabolic improvements. We have revised the Discussion section accordingly with the following statement:

 

 “The absence of a comparator group prevents definitive conclusions regarding causality between GLP-1 or GLP-1/GIP RA therapy and the observed metabolic improvements. Concurrent diabetes management, modifications in other medications, lifestyle interventions, or other unmeasured factors may have influenced the observed changes.”

 

Comment 2:

“Patients eligible for inclusion were considered those who had a confirmed HIV diagnosis and who were on any dose of GLP-1 or dual GLP-1/GIP RA treatment for at least 12 weeks. This may introduce survivor bias because patients discontinuing therapy were excluded. This limitation should be adequately discussed. “

 

Response 2:

We appreciate the reviewer’s observation. We agree that requiring at least 12 weeks of treatment may introduce survivor bias by excluding patients who discontinued therapy earlier due to adverse effects, lack of efficacy, cost, access issues, or other barriers. We have added this limitation to the Discussion with the following statement:

 

 “Our inclusion criteria required at least 12 weeks of treatment with GLP-1 RAs or GLP-1/GIP dual RAs. This may introduce survivor bias, as individuals who discontinued treatment earlier because of adverse effects, access limitations, or insufficient response were not captured in our analysis. Consequently, our findings may overrepresent individuals who were able to tolerate and maintain therapy.”

 

Comment 3:

“The ART regimens were divided into two groups. Please clarify if this oversimplifies the complexity of ART-associated weight gain as ARTs have distinct metabolic effects."

 

Response 3:

We thank the reviewer for this important point. We agree that our current grouping of ART medications needs to be revised. For that reason we have changed the ART groups as follows:

Group 1: tenofovir alafenamide (TAF)-based regimens,

Group 2: tenofovir disoproxil fumarate (TDF) - based regimens

Group 3: NNRTI  (Dolutegravir/rilpivirine)

Group 4: Non- Tenofovir NRTI (Dolutegravir/lamivudine)

Group 5: Others

 

These changes are outlined in the revised manuscript in table 1 and with the following statement:

“The ART regimens were divided into five groups: Group I [tenofovir alafenamide (TAF)-based regimens], which included emtricitabine/rilpivirine/tenofovir alafenamide, bictegravir/emtricitabine/tenofovir alafenamide, darunavir/cobicistat/emtricitabine/tenofovir alafenamide and elvitegravir/cobicistat/emtricitabine/tenofovir alafenamide, accounting for 72.3% of patients; Group II [Tenofovir disoproxil fumarate (TDF)-based regimens], including emtricitabine/rilpivirine/tenofovir disoproxil fumarate, elvitegravir/cobicistat/emtricitabine/tenofovir disoproxil fumarate, efavirenz/emtricitabine/tenofovir disoproxil fumarate, doravirine/lamivudine/tenofovir disoproxil fumarate, making up 3% of the cohort; 4.5% of patients were on Group III (NNRTI regimen), including dolutegravir/rilpivirine; 11.9% of patients were on group IV (Non-tenofovir NRTI), including dolutegravir/lamivudine; lastly 8.4 % were on group V (other medications).”

 

We believe that this current categorization does not simplify but rather stratifies in further sub groups, based on meaningful pharmacological distinction. Having said that, some of those groups were already small and further subcategorization would further fragment the groups, so unfortunately we haven’t originally performed further subgroup analysis based on ART.

 

Comment 4:

“ The mean weight reduction was 3.4 kg, from 95.9 kg to 92.5 kg (p<0.001). Although statistically significant, this is still modest. In this situation weight loss should be interpreted cautiously."

 

Response 4:

We agree with the reviewer that the magnitude of weight reduction observed was modest compared with clinical trials evaluating GLP-1 therapies in populations without HIV. We have revised the Discussion sections to provide a more cautious interpretation with the following statement:

 

“Although, the overall degree of weight loss over 24 months is modest and should be interpreted cautiously, it is important to consider that the mean HbA1c in our cohort was 8.7%, indicating poor diabetes control.”

 

Comment 5:

"It is also worth noting that the study was conducted in the Bronx where medication adherence challenges are quite common; therefore, the modest weight changes may partly reflect variable adherence to therapy. Please comment on how adherence was evaluated."

 

Response 5:

We appreciate this important comment. Medication adherence was not directly measured in our retrospective analysis, and we agree that this should have been explicitly stated. We have added the following statement:

 

“Furthermore, medication adherence was not assessed in this study. Variable adherence to GLP-1 RA therapy may have influenced the observed magnitude of weight loss and glycemic improvement. Therefore, this factor should be considered a potential source of residual confounding. “

 

Comment 6:

"Importantly, our data showed equal reduction in weight, regardless of which ART was used, suggesting that GLP-1 RA use can effectively mitigate concerns about weight gain associated with certain integrase inhibitors. Please provide evidence to support this statement or remove it."

 

Response 6:

We thank the reviewer for identifying this overinterpretation. We have changed this statement to the following:

 

“Importantly, our data showed that no significant differences in weight loss were observed when the population was stratified by ART groups.”

 

Comment 7:

"The comparisons are generally appropriate. However, it should be discussed more thoroughly with regard to baseline HbA1c, severity of obesity and dose and duration of GLP-1 treatment."

 

Response 7:

We appreciate the reviewer’s suggestion. We have expanded our discussion regarding the influence of baseline HbA1c, obesity severity, and treatment duration on therapeutic response, as described in the following statements:

 

“ Firstly, GLP-1 RA therapy led to a mean HbA1c reduction of 1%, with the duration of treatment emerging as the only independent predictor of more significant reductions. Treatment duration was relatively prolonged in our cohort, with a mean duration of 24.2 months.”

 

“Accordingly, only 62% of the cohort was obese, which could have diluted the observed weight loss effect, as GLP-1 RA-related weight reduction tends to be more pronounced in individuals with obesity.”

 

“ Although, the overall degree of weight loss over 24 months is modest and should be interpreted cautiously, it is important to consider that the mean HbA1c in our cohort was 8.7%, indicating poor diabetes control. Given that our sample also included people without diabetes, the mean HbA1c among those with T2DM was likely even higher. Studies and clinical experience support that patients with poorly controlled T2DM may experience less weight loss on GLP-1 RAs, compared to those with better glycemic control.”