Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors, thank YOu for giving the chance to read your paper. I appreciated it. Some suggestions to make it better. In "Materials and Methods" section, please define quality and type of drug intake. Gender differences should be underlined as females are more represented and lymphocyte count could differ among the two sexes. 

No further questions.

Author Response

Comment 1.1: In "Materials and Methods" section, please define quality and type of drug intake.

Response: We thank the reviewers for their valuable suggestions. We have added a "Medication Use" section to "Materials and Methods 2.4" to clearly define medication assessment. The new content is as follows:

"Medication use was assessed through structured interviews and verification of medical records at baseline. Participants were asked to report all current medications, including glucose-lowering agents (insulin, metformin, sulfonylureas, DPP-4 inhibitors, SGLT2 inhibitors, GLP-1 receptor agonists), antihypertensive medications (ACE inhibitors, angiotensin receptor blockers, calcium channel blockers, beta-blockers, diuretics), and lipid-lowering agents (statins, fibrates). Medication adherence was defined as taking prescribed medications at least 80% of the time over the past month. Information on medication type, dosage, and duration of use was systematically recorded."

 

Comment 1.2: Gender differences should be underlined as females are more represented and lymphocyte count could differ among the two sexes.

Response: This is an excellent point. We have now conducted sex-stratified analyses and added the following:

Added text in Results section 3.5 : "Sex-stratified analyses revealed no significant effect modification by sex (P for interaction = 0.421). The inverse association between PNI and renal decline was consistent in both males (β = -0.082, 95% CI: -0.145 to -0.019) and females (β = -0.091, 95% CI: -0.138 to -0.044), despite females having slightly higher baseline lymphocyte counts ."

New Table 3 : Sex-stratified baseline characteristics showing lymphocyte count differences (males: 2.1 ± 0.6 ×10⁹/L; females: 2.3 ± 0.7 ×10⁹/L, P = 0.012)

Reviewer 2 Report

Comments and Suggestions for Authors

The authors used prospective cohort data from the Shanghai Suburban Adult Cohort Biobank to examine the association between the prognostic nutritional index (PNI) and renal function decline in 1,711 community-dwelling patients with type 2 diabetes who had preserved renal function at baseline. A consistent inverse linear association was observed between PNI and the rate of decline in eGFR (nonlinearity P > 0.05). Johnson–Neyman analysis further demonstrated a statistically significant protective association of PNI in patients with HbA1c levels ranging from 7.24% to 8.71%. Additionally, the incidence of renal function decline (50.0%) was significantly higher in patients with poor glycemic control (HbA1c ≥ 8%) and low PNI (<50), whereas the inverse correlation between PNI and renal function decline was significantly weaker in patients with good glycemic control (HbA1c < 8%), with the incidence rate in the low PNI subgroup (6.7%) being lower than that in the high PNI subgroup (15.9%). These results suggest that the protective role of PNI may depend on the glycemic environment. Clinically, the authors conclude that a combined assessment of PNI and HbA1c may be a simple, practical approach to stratify risk of renal function decline and identify patient subgroups that may benefit from intervention.

The reviewers' comments are as follows:

1. Although this study used prospective cohort data, the main text does not provide information on "how many years after the start of follow-up" when renal function decline occurred, nor on "the median or maximum follow-up period." To allow readers to properly evaluate the validity of this cohort study, it is necessary to clearly state information about the follow-up period in the Methods section.

2. In Table 1, parametric and nonparametric analyses are used depending on the characteristics of the data, but the statistical methods used to compare variables are not clearly stated. It would be desirable to clearly state the statistical methods used for each measurement and survey item (e.g., t-test, Wilcoxon test, chi-square test), as well as the calculation method for P values.

3. An overview of Tables 1 and 2 suggests that the onset of renal function decline can be adequately predicted using indicators such as systolic blood pressure, fasting blood glucose (FBG), and baseline eGFR, even without focusing on PNI or HbA1c. In fact, a model combining these factors with HbA1c may have greater predictive ability. In this context, a more detailed explanation of the scientific and clinical significance of focusing on PNI is needed.

4. Focusing on PNI, the results of this study suggest that PNI is useful as a predictor of declining renal function only within a specific HbA1c range. On the other hand, the results also suggest that PNI may not be an effective predictor in patients with low HbA1c levels (normal or well-controlled). Given this, the clinical use of PNI is likely to be relatively limited. Therefore, it would be desirable to provide a more concrete and practical discussion in the Discussion section (especially in Section 4.4) regarding "in what patient backgrounds would a combined evaluation of PNI and HbA1c be useful for determining treatment strategies?"

Overall, we believe that additional explanations are needed to help readers understand the strength and uniqueness of the associations found in this study, as well as the significance of using PNI as a predictor and its clinical applicability.

Author Response

Comment 2.1: Although this study used prospective cohort data, the main text does not provide information on "how many years after the start of follow-up" when renal function decline occurred, nor on "the median or maximum follow-up period." To allow readers to properly evaluate the validity of this cohort study, it is necessary to clearly state information about the follow-up period in the Methods section.

Response: We thank the reviewer for this important observation. We have added comprehensive follow-up information to the Methods section and Abstract.

Added to Section 2.1 (Study Design and Population):

"Follow-up examinations were conducted between June 2019 and December 2020, providing a median follow-up duration of 3.2 years (interquartile range: 2.8-3.6 years; range: 2.0-4.5 years). Among the 1,711 participants with complete baseline and follow-up data, 377 (22.0%) developed the composite adverse renal outcome during follow-up."

Added to Section 2.3 (Outcome Measurement):

"The annualized eGFR decline rate was calculated as: (baseline eGFR - follow-up eGFR) / follow-up duration in years. Participants with annualized decline >3 mL/min/1.73 m² per year were classified as having rapid renal function decline, consistent with established thresholds associated with adverse long-term outcomes in diabetic populations."

Added to Table 1 footnote:

"Follow-up duration: median 3.2 years (IQR: 2.8-3.6 years)."

Comment 2.2: In Table 1, parametric and nonparametric analyses are used depending on the characteristics of the data, but the statistical methods used to compare variables are not clearly stated. It would be desirable to clearly state the statistical methods used for each measurement and survey item (e.g., t-test, Wilcoxon test, chi-square test), as well as the calculation method for P values.

Response: We agree that greater methodological transparency is needed. We have revised Table 1 footnote and Section 2.5.1 to explicitly specify the statistical tests used.

Added to Section 2.5.1 (Descriptive Statistics):

"Normality of continuous variables was assessed using the Shapiro-Wilk test. Between-group comparisons used Student's t-test for normally distributed variables, the Wilcoxon-Mann-Whitney test for non-normally distributed variables, and Pearson's chi-square test for categorical variables. All statistical tests were two-sided with α = 0.05."

Revised Table 1 Footnote:

"Note: Data are presented as mean (standard deviation) for normally distributed continuous variables, median [interquartile range] for non-normally distributed continuous variables, or n (%) for categorical variables. Between-group comparisons were performed using: Student's t-test for normally distributed continuous variables (BMI, PNI, MET, systolic and diastolic blood pressure, FBG, HbA1c, TC, TG, HDL-C, LDL-C, baseline eGFR); Wilcoxon-Mann-Whitney test for non-normally distributed continuous variables (Age); and Pearson's chi-square test for categorical variables (sex, smoking, alcohol intake). Normality was assessed using the Shapiro-Wilk test. Adverse renal outcome was defined as rapid eGFR decline (>3 mL/min/1.73 m² per year) or incident chronic kidney disease stage G3 (eGFR 30-59 mL/min/1.73 m²). Follow-up duration: median 3.2 years (IQR: 2.8-3.6 years)."

Comment 2.3: An overview of Tables 1 and 2 suggests that the onset of renal function decline can be adequately predicted using indicators such as systolic blood pressure, fasting blood glucose (FBG), and baseline eGFR, even without focusing on PNI or HbA1c. In fact, a model combining these factors with HbA1c may have greater predictive ability. In this context, a more detailed explanation of the scientific and clinical significance of focusing on PNI is needed.

Response: This is an excellent point that deserves thorough discussion. We have added a new subsection to the Discussion to address the incremental value of PNI beyond traditional risk factors.

Added New Subsection 4.6: Incremental Value of PNI Beyond Traditional Risk Factors

We propose several complementary reasons why PNI merits clinical attention despite the availability of established predictors:

First, PNI captures distinct pathophysiological processes. While systolic blood pressure reflects hemodynamic stress and FBG/HbA1c capture glycemic exposure, PNI integrates nutritional reserves (albumin) and immune-inflammatory competence (lymphocyte count). These represent upstream vulnerability factors that may precede overt metabolic derangement. Thus, PNI may identify patients with subclinical vulnerability who would be missed by traditional risk stratification alone.

Second, PNI provides actionable, modifiable targets. While baseline eGFR is a powerful predictor, it is a marker of existing damage rather than a therapeutic target. In contrast, low PNI identifies patients who may specifically benefit from nutritional intervention—a therapeutic avenue often underutilized in diabetes management. Our finding that the PNI-risk association is strongest at intermediate HbA1c levels (7.24-8.71%) suggests a "window of opportunity" where nutritional optimization could provide maximal benefit.

Third, the synergistic interaction between PNI and HbA1c reveals context-dependent risk. Our stratified analysis (Figure 4) demonstrates that the combination of poor glycemic control (HbA1c ≥8%) and low PNI (<50) confers disproportionate risk (50% incidence), far exceeding what would be predicted by either factor alone. Identifying such synergistic risk profiles has important implications for resource allocation.

Fourth, PNI offers practical advantages for risk stratification in resource-limited settings. PNI is calculated from routine laboratory tests (serum albumin and complete blood count) that are already performed in standard diabetes care, requiring no additional cost or specialized equipment.

We acknowledge that a comprehensive risk prediction model incorporating blood pressure, glucose, baseline eGFR, and PNI would likely achieve superior discrimination compared to any single biomarker. However, our goal was not to replace traditional risk factors but to demonstrate that PNI provides complementary information—identifying a distinct dimension of vulnerability (nutritional-inflammatory status) and highlighting patient subgroups who may benefit from tailored intervention strategies."

Comment 2.4: Focusing on PNI, the results of this study suggest that PNI is useful as a predictor of declining renal function only within a specific HbA1c range. On the other hand, the results also suggest that PNI may not be an effective predictor in patients with low HbA1c levels (normal or well-controlled). Given this, the clinical use of PNI is likely to be relatively limited. Therefore, it would be desirable to provide a more concrete and practical discussion in the Discussion section (especially in Section 4.4) regarding "in what patient backgrounds would a combined evaluation of PNI and HbA1c be useful for determining treatment strategies?"

Response: We appreciate this request for greater clinical specificity. We have substantially expanded Section 4.4 (Clinical Implications) to provide detailed, actionable guidance on patient selection, risk stratification, and personalized management pathways.

Revised and Expanded Section 4.4 (Clinical Implications) - Key additions include:

"1. Risk Stratification Tool: We propose routine calculation of PNI [serum albumin (g/L) + 5 × lymphocyte count (×10⁹/L)] in all patients with T2D during annual assessment, particularly for those with HbA1c ≥7.5%. This simple calculation uses readily available laboratory data without additional cost.

2. Identification of Ultra-High-Risk Patients: Cross-classification of PNI (<50 vs. ≥50) and HbA1c (<8% vs. ≥8%) enables rapid identification of patients requiring intensive intervention. The ultra-high-risk subgroup (HbA1c ≥8% and PNI <50) with 50% event rate should trigger immediate multidisciplinary referral.

3. Personalized Management Pathways:

• Ultra-high-risk patients (HbA1c ≥8% and PNI <50): Urgent nephrology and dietitian referral, intensive glycemic optimization with SGLT2 inhibitors and/or GLP-1 receptor agonists, structured nutritional intervention focusing on high-biological-value protein, Mediterranean dietary pattern, and potential omega-3 supplementation
• PNI-sensitive patients (HbA1c 7.24-8.71%): Focus on improving nutritional status alongside moderate glycemic control through lifestyle modification, dietary counseling, and anti-inflammatory nutrition strategies
• Glycemia-dominant patients (high HbA1c, high PNI): Prioritize aggressive glucose-lowering therapy while maintaining adequate nutritional support.

4. Monitoring Strategy: Serial PNI measurements (every 6-12 months) could track response to interventions and identify emerging risk in stable patients. A declining PNI trend should prompt reassessment of nutritional status and inflammatory burden.

5. Resource Allocation: This stratification approach helps healthcare systems allocate limited resources (specialist referrals, intensive programs, dietitian consultations) to those most likely to benefit, improving cost-effectiveness of preventive interventions."

We have also added specific guidance on when PNI assessment is most valuable (patients with HbA1c 7.0-9.0%, multiple risk factors, resource-limited settings) and when it may be lower priority (excellent glycemic control with HbA1c <7.0%, very advanced hyperglycemia >9.0%).

Reviewer 3 Report

Comments and Suggestions for Authors

• The aim in abstract is not clear
• Separate inclusion /and exclusion criteria in a subsection
• For the following make a separate subsection in method and material and place it above statistics The protocol was approved by the Ethical Review Com mittee of the School of Public Health, Fudan University (IRB No. 2016-04-0586) and con-112 ducted in accordance with the Declaration of Helsinki.
• Specify the type of diabetes (type 1, type 2, or both)
• Preserved renal function is defined solely as eGFR > 60 mL/min/1.73 m². where there any other criteria?
• The composite outcome does not include albuminuria progression, which is a key component of CKD staging and prognosis. Please shorty provide a reason
• This may limit the ability to capture early or non–eGFR-based renal damage.
• Where there any bias in your study?
• How will the study be useful in practice?
• What are the future perspectives?

Author Response

Comment 3.1: The aim in abstract is not clear.

Response: We have revised the abstract to more clearly state the study aims. The revised objectives now read:

This study aimed to: (1) characterize the dose-response relationship between PNI and early renal function decline in type 2 diabetes using restricted cubic splines; (2) identify whether glycemic control (HbA1c) modifies the PNI-renal decline association; and (3) evaluate the clinical utility of combining PNI and HbA1c for risk stratification.

Comment 3.2: Separate inclusion and exclusion criteria in a subsection.

Response: We have restructured section 2.1 with clear subsections:

2.1.1 Inclusion Criteria

• Age 20-74 years
• Confirmed diagnosis of type 2 diabetes mellitus
• Preserved renal function (eGFR > 60 mL/min/1.73 m²) at baseline
• Complete data on exposure, outcomes, and key covariates

2.1.2 Exclusion Criteria

• Pre-existing chronic kidney disease (CKD stages 3-5)
• History of end-stage renal disease or dialysis
• Pregnancy
• Severe comorbidities (cancer, cirrhosis, cardiopulmonary failure) at baseline
• Missing data on PNI components, eGFR, or essential covariates

Comment 3.3: For the ethical approval, make a separate subsection in method and material and place it above statistics.

Response: We have created a new subsection 2.6 Ethical Considerations placed immediately after the Covariates section and before Statistical Analysis:

2.6. Ethical Considerations

All participants provided written informed consent prior to enrollment. The study protocol was approved by the Ethical Review Committee of the School of Public Health, Fudan University (IRB No. 2016-04-0586) and conducted in accordance with the Declaration of Helsinki. Data confidentiality and participant privacy were maintained throughout the study.

Comment 3.4: Specify the type of diabetes (type 1, type 2, or both).

Response: We have clarified this throughout the manuscript:

• Title: Now specifies "Type 2 Diabetes"
• Abstract line 3: "...among patients with type 2 diabetes (T2D)..."
• Section 2.1.1: " Participants were eligible if they met all of the following criteria:
• Age 20-74 years at baseline
• Confirmed diagnosis of type 2 diabetes mellitus, defined as fasting plasma glucose ≥7.0 mmol/L, 2-hour post-load glucose ≥11.1 mmol/L, HbA1c ≥6.5%, or physi-cian-diagnosed T2D with current use of glucose-lowering medications, according to American Diabetes Association criteria
• Preserved renal function (eGFR >60 mL/min/1.73 m²) at baseline, corresponding to CKD stages G1-G2
• Complete data on exposure variables (PNI components), outcomes, and key co-variates"

Comment 3.5: Preserved renal function is defined solely as eGFR > 60 mL/min/1.73 m². Were there any other criteria?

Response: Thank you for this clarification request. We have expanded the definition in section 2.1.1:

Preserved renal function was defined as baseline eGFR >60 mL/min/1.73 m² (corresponding to CKD stages G1-G2). Although urinary albumin-to-creatinine ratio (ACR) was not systematically measured in all participants at baseline and therefore could not be used as an inclusion criterion, participants with known advanced albuminuria (ACR >300 mg/g) documented in medical records were excluded. This approach ensured that the study population had relatively preserved renal function at baseline.

Comment 3.6: The composite outcome does not include albuminuria progression, which is a key component of CKD staging and prognosis. Please shortly provide a reason.

Response: This is an important limitation. We have added the following explanation in section 2.3 and expanded the limitations section:

Urinary albumin-to-creatinine ratio (ACR) was not systematically measured in all participants at baseline or follow-up in the SSACB cohort, as the primary focus of the parent study was broader cardiovascular and metabolic outcomes rather than kidney-specific markers. Consequently, we could not incorporate albuminuria or its progression into our composite outcome definition.

We acknowledge this represents a significant limitation, as the Kidney Disease: Improving Global Outcomes (KDIGO) classification of CKD in diabetes integrates both GFR categories (G1-G5) and albuminuria categories (A1: <30 mg/g; A2: 30-300 mg/g; A3: >300 mg/g). Albuminuria is an independent predictor of CKD progression and cardiovascular events in diabetes, and eGFR decline and albuminuria progression may represent partially distinct pathophysiological processes.

Our study therefore captures primarily GFR-based kidney function decline and may miss early glomerular damage manifesting as isolated albuminuria. Where available from medical records (n=412, 24%), baseline ACR was <30 mg/g (A1) in 68%, 30-300 mg/g (A2) in 27%, and >300 mg/g (A3) in 5% of participants.

According to the complete KDIGO classification, participants developing incident CKD stage G3 in our study would be classified as G3aA1, G3aA2, G3aA3, G3bA1, G3bA2, or G3bA3, but albuminuria category could not be determined. This GFR-based definition, while incomplete, captures clinically significant kidney function decline and has been validated as a meaningful endpoint in diabetic populations. Future investiga-tions incorporating comprehensive assessment of both eGFR and ACR trajectories are warranted.

Comment 3.7: This may limit the ability to capture early or non-eGFR-based renal damage.

Response: We fully agree and have added this to the limitations section (Section 4.5):

Several important limitations warrant consideration. First, the observational design precludes causal inference, and despite adjustment for a broad set of confounders including medications, residual confounding from unmeasured factors (e.g., detailed dietary intake, specific medication dosages, subclinical infections) may remain. Second, reliance on single baseline measurements of PNI and covariates does not capture temporal variability, potentially introducing non-differential misclassification and attenuating associations through regression dilution bias; longitudinal PNI trajectories war-rant further investigation. Third, generalizability may be limited given the Chinese suburban cohort, necessitating external validation in diverse ethnicities and healthcare settings. Fourth, the absence of systematic albuminuria assessment represents a significant limitation, as the reliance on eGFR-based outcomes may have missed early glomerular damage manifesting primarily as proteinuria, potentially misclassifying participants with early albuminuric kidney disease and underestimating the true bur-den of diabetic kidney disease. Fifth, several potential biases merit acknowledgment: selection bias from excluding participants with severe comorbidities or those lost to follow-up may have resulted in a healthier cohort; survival bias from excluding indi-viduals who died before follow-up; measurement bias from single-time-point assessments; and information bias from self-reported lifestyle factors. These limitations underscore the need for future studies incorporating comprehensive kidney disease phenotyping with both eGFR and albuminuria trajectories, repeated biomarker measure-ments, and diverse populations to validate and extend our findings.

Comment 3.8: Were there any biases in your study?

Response: We have added potential biases to section 4.5 of the discussion:

Fifth, several potential biases merit acknowledgment: selection bias from excluding participants with severe comorbidities or those lost to follow-up may have resulted in a healthier cohort; survival bias from excluding individuals who died before follow-up; measurement bias from single-time-point assess-ments; and information bias from self-reported lifestyle factors.

Comment 3.9: How will the study be useful in practice?

Response: We have enhanced the Clinical Implications section (4.4) with more specific practical recommendations:

"These findings offer immediately implementable clinical applications:

1. Risk Stratification Tool: We propose routine calculation of PNI [serum albumin (g/L) + 5 × lymphocyte count (×10⁹/L)] in all patients with T2D during annual assessment, particularly for those with HbA1c ≥7.5%. This simple calculation uses readily available laboratory data without additional cost.
2. Identification of Ultra-High-Risk Patients: Cross-classification of PNI (<50 vs. ≥50) and HbA1c (<8% vs. ≥8%) enables rapid identification of patients requiring intensive intervention. The ultra-high-risk subgroup (HbA1c ≥8% and PNI <50) with 50% event rate should trigger immediate multidisciplinary referral.
3. Personalized Management Pathways:

• Ultra-high-risk patients: Urgent nephrology and dietitian referral, intensive glycemic optimization with SGLT2i/GLP-1RA, structured nutritional intervention
• PNI-sensitive patients (HbA1c 7.24-8.71%): Focus on improving nutritional status alongside moderate glycemic control
• Glycemia-dominant patients (high HbA1c, high PNI): Prioritize aggressive glucose-lowering therapy

4. Monitoring Strategy: Serial PNI measurements (every 6-12 months) could track response to interventions and identify emerging risk in stable patients.
5. Resource Allocation: This stratification approach helps healthcare systems allocate limited resources (specialist referrals, intensive programs) to those most likely to benefit."

Comment 3.10: What are the future perspectives?

Response: We have added a new paragraph at the end of the Conclusions section:

Future research should pursue several directions. First, prospective validation studies in diverse ethnic populations and healthcare settings are needed to confirm the generalizability of our PNI-HbA1c stratification framework. Second, interventional trials should test whether targeted improvement of PNI through nutritional counseling, protein supplementation, or anti-inflammatory dietary patterns (e.g., Mediterranean diet, omega-3 fatty acids) can slow renal decline, particularly in the identified high-risk subgroup. Such trials should incorporate serial PNI measurements to assess dose-response relationships and optimal therapeutic targets. Third, mechanistic studies using proteomics and metabolomics could elucidate the molecular pathways linking nutritional-inflammatory status to glomerular and tubular injury under different glycemic conditions. Fourth, future cohorts should systematically collect both eGFR and albuminuria data to comprehensively phenotype diabetic kidney disease trajectories. Fifth, machine learning approaches integrating PNI, HbA1c, and other biomarkers may enable development of more precise predictive models. Finally, implementation science research is needed to evaluate the feasibility, cost-effectiveness, and clinical impact of incorporating PNI-based risk stratification into routine diabetes care pathways.

Reviewer 4 Report

Comments and Suggestions for Authors

DM is a pathology often associated with renal impairment, therefore an early diagnosis or prediction of renal involvement is of high importance in clinical practice. This study can be beneficial for both practitioners and patients and can improve the management of diabetic patients in order to highlight the risk of DKD onset even from the early stages. The methodology and results were well described and the conclusions were in accordance with the assessed findings. One minor recommendation - better define CKD stage G3: eGFR between 30-59 mL/min/1.73mp. Considering that CKD includes also stages G1 and G2, it should be added if the patients presented any congenital renal malformation / diseases (probably not, but it should be highlighted this aspect, as well). In addition, no data regarding albumin-creatinine ratio was included - was this ratio evaluated? (knowing the clear association between this test and renal impairment in diabetic patients, in my opinion, it should be mentioned, as well. Additionally, the correct classification of CKD in diabetic patients includes the values of eGFR and ACR, therefore when CKD definition is presented, ACR values should added as well - i.e., A1, A2 or A3 - probably the stage was G3A1, but it should be clearly defined).

Author Response

Comment 4.1: Better define CKD stage G3: eGFR between 30-59 mL/min/1.73m².

Response: We have clarified this throughout the manuscript:

• Section 2.3 now reads: "...or (2) incident CKD stage G3, defined as eGFR 30-59 mL/min/1.73 m² at follow-up (corresponding to stages G3a [45-59] or G3b [30-44]) among participants with baseline eGFR ≥60 mL/min/1.73 m²."
• Table 1 footnote: Updated to specify "...or incident chronic kidney disease stage G3 (eGFR 30-59 mL/min/1.73 m²)."

Comment 4.2: Considering that CKD includes also stages G1 and G2, it should be added if the patients presented any congenital renal malformation/diseases.

Response: Excellent point. We have added this to the exclusion criteria and methods:

Section 2.1.2 (Exclusion Criteria) now includes: "- Known congenital renal malformations or hereditary kidney diseases (polycystic kidney disease, Alport syndrome, etc.) - History of renal surgery, nephrectomy, or renal transplantation"

Section 2.1 narrative now states: "Participants were excluded if they had...congenital renal anomalies or hereditary kidney diseases documented in medical records or identified through baseline imaging/clinical assessment. This ensured that observed renal function changes reflected diabetic kidney disease progression rather than pre-existing structural abnormalities."

Comment 4.3: No data regarding albumin-creatinine ratio was included - was this ratio evaluated?

Response: We appreciate this important observation. We have added comprehensive clarification:

The limitations of outcome assessment are added in section 2.3:

Urinary albumin-to-creatinine ratio (ACR) was not systematically measured in all participants at baseline or follow-up in the SSACB cohort, as the primary focus of the parent study was broader cardiovascular and metabolic outcomes rather than kidney-specific markers. Consequently, we could not incorporate albuminuria or its progression into our composite outcome definition. We acknowledge this represents a significant limitation, as the KDIGO classification of CKD in diabetes integizes both GFR categories (G1-G5) and albuminuria categories (A1: <30 mg/g; A2: 30-300 mg/g; A3: >300 mg/g). Albuminuria is an independent predictor of CKD progression and cardiovascular events in diabetes, and eGFR decline and albuminuria progression may represent partially distinct pathophysiological processes. Our study therefore captures primarily GFR-based kidney function decline and may miss early glomerular damage manifesting as isolated albuminuria. Where available from medical records (n=412, 24%), baseline ACR was <30 mg/g (A1) in 68%, 30-300 mg/g (A2) in 27%, and >300 mg/g (A3) in 5% of participants. Future investigations incorporating comprehensive assessment of both eGFR and ACR trajectories are warranted.

Section 4.5 (Limitations) has been expanded: Fifth, several potential biases merit acknowledgment: selection bias from excluding participants with severe comorbidities or those lost to follow-up may have resulted in a healthier cohort; survival bias from excluding individuals who died before follow-up; measurement bias from single-time-point assessments; and information bias from self-reported lifestyle factors.

Comment 4.4: The correct classification of CKD in diabetic patients includes the values of eGFR and ACR, therefore when CKD definition is presented, ACR values should be added as well.

Response: We have added clarification:

Section 2.3: According to the Kidney Disease: Improving Global Outcomes (KDIGO) classification, CKD in diabetes is optimally staged using both GFR categories (G1-G5) and albuminuria categories (A1-A3). However, due to the absence of systematic ACR assessment in our cohort, we defined incident CKD based solely on GFR criteria (stage G3: eGFR 30-59 mL/min/1.73 m²). Participants developing incident CKD stage G3 would be classified as G3aA1, G3aA2, G3aA3, G3bA1, G3bA2, or G3bA3 in the complete KDIGO system, but albuminuria category could not be determined in our study. This GFR-based definition, while incomplete, captures clinically significant kidney function decline and has been validated as a meaningful endpoint in diabetic populations.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

no more comments