The Role of Reticulocyte-Derived Parameters in the Detection of Iron-Restricted Erythropoiesis in the Elderly

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I examined your article titled The role of Reticulocyte derived parameters in the detection of iron restricted erythropoiesis in the elderly in detail. The authors compared some values ​​provided by new generation blood count devices to standard laboratory tests in order to detect iron deficiency in patients over 65 years of age. The study is primarily a clinical trial. An alternative measurement method is proposed for situations where traditional tests are inadequate or misleading. Some points in the article require review and further elaboration.

1. The introduction notes that there is no consensus on the definition of iron deficiency in the elderly and that biomarker threshold values ​​are controversial. However, the introduction does not provide detailed information on how this study will fill a gap in the literature.
2. A theoretical explanation of why MRV biologically changes in iron deficiency would be helpful.
3. Only outpatients were included in the study. Exclusion of inpatients or those with more severe conditions makes it difficult to generalize the findings to the elderly population. This requires a more detailed discussion.
4. Presenting the method in a flowchart will help readers understand the study more clearly.
5. It would be helpful to present the results in the form of comparative graphs.

In conclusion, the study needs to be revised in line with the suggestions made. As it stands, its contribution to the literature will be limited.

Author Response

Reviewer 1

Thank you for the time and effort to review the manuscript , and for the meaningful suggestions

I examined your article titled The role of Reticulocyte derived parameters in the detection of iron restricted erythropoiesis in the elderly in detail. The authors compared some values ​​provided by new generation blood count devices to standard laboratory tests in order to detect iron deficiency in patients over 65 years of age. The study is primarily a clinical trial. An alternative measurement method is proposed for situations where traditional tests are inadequate or misleading. Some points in the article require review and further elaboration.

1. The introduction notes that there is no consensus on the definition of iron deficiency in the elderly and that biomarker threshold values ​​are controversial. However, the introduction does not provide detailed information on how this study will fill a gap in the literature.

We agree and have clarified how our work addresses the current gap. We explicitly state that, in elderly patients, there is no consensus on iron deficiency definitions and that data on Mindray‑derived reticulocyte parameters in this population are scarce.  We added a final paragraph specifying that our study aims to establish MRV and RHe cut‑off values for detecting iron‑restricted erythropoiesis in older outpatients, using sTfR as reference, and to compare their performance with conventional iron markers in this age group.

 

1. A theoretical explanation of why MRV biologically changes in iron deficiency would be helpful.

We acknowledge this point and have added a brief biological explanation. In the revised manuscript we explain that, under iron‑restricted erythropoiesis, hemoglobin synthesis is impaired in late erythroblasts, leading to the release of smaller and hypochromic reticulocytes; consequently, mean reticulocyte volume (MRV) decreases in parallel with reticulocyte hemoglobin content (RHe). We emphasise that MRV thus reflects the interplay between iron availability and erythroid maturation in the marrow.​

 

1. Only outpatients were included in the study. Exclusion of inpatients or those with more severe conditions makes it difficult to generalize the findings to the elderly population. This requires a more detailed discussion.

 

We agree that the outpatient setting limits generalizability. We have expanded the Limitations subsection to clarify that our cohort consisted exclusively of elderly outpatients attending routine controls, with relatively stable clinical status, and that results may not be directly extrapolated to acutely ill inpatients or institutionalised frail elders. We now explicitly state that additional studies in hospitalized and higher‑dependency geriatric populations are needed to confirm the applicability of the proposed cut‑offs.​

1. Presenting the method in a flowchart will help readers understand the study more clearly.

We appreciate this suggestion and have added a flowchart summarizing patient selection, exclusion criteria, and subsequent diagnostic classification into IDA, ACD, and mixed IDA/ACD based on ferritin, TSAT, CRP, and sTfR. This visual scheme complements the text in the Materials and Methods section and facilitates understanding of the study workflow.​

1. It would be helpful to present the results in the form of comparative graphs.

We agree that graphical presentation improves readability. In the revised version we have added comparative figures (Box &whiskers plots) illustrating the distribution of reticulocyte derived parameters (MRV and RHe)  in the groups. We now present ROC curves for MRV, RHe and ferritin. These figures complement the Tables  highlight the superior diagnostic performance of MRV and RHe versus s‑ferritin.​

 

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for permit me to review your manuscript addressing the diagnostic value of reticulocyte-derived parameters (RHe and MRV) for detecting iron-restricted erythropoiesis in elderly patients. The study addresses a clinically important issue in geriatric hematology. The manuscript addresses an important diagnostic challenge and proposes potentially useful automated parameters. However, clarification of the diagnostic reference standard, reduction of potential bias, and more robust statistical reporting are necessary before the conclusions can be considered reliable.

Major comments: 

1.- Concept of diagnosis: the definition stablish that soluble transferrin receptor (sTfR >52 nmol/L) is used as the “gold standard” for iron deficiency. However the groups (IDA, ACD, mixed anemia) are defined using ferritin, TSAT, CRP, and sTfR but there is no table, even as supplementary material with distribution (median/IQR if not normal) of CRP and other markers by group; and clarify how many meet CRP>5. CRP, serum iron, or transferrin values.

2.- ROC analyses are presented for “iron deficient erythropoiesis,” but the exact definition of this endpoint is not consistently described. Only includes ferritin sensitivity/specificity for some points; 95% CI is missing and the list appears incomplete. It is unclear whether the study evaluates: Absolute iron deficiency, Functional iron deficiency, Iron-restricted erythropoiesis, or a composite condition (IDA + mixed + FID). This must be explicitly clarified.

3.-  Inflammation and Chronic Kidney Disease, as this is a large ederly cohort (mean age 80.5 years) with a high burden of comorbidities, including chronic renal disease and inflammatory conditions. This situations may influence sTfR and erythropoietic indices, CRP affects ferritin interpretation, and then the absence of stratified or multivariable analysis limits interpretability.

I am suggesting to stratify analyses according to CRP levels (e.g., <5 vs ≥5 mg/L), Stratification by renal function, or to plan logistic regression models adjusting for CRP and renal function to assess incremental diagnostic value.

Minor comments:

The study is described as both “prospective” and “cross-sectional.” Please clarify whether this refers to prospective enrollment with cross-sectional analysis.

Normality testing is reported, yet results are presented as mean ± SD throughout. If variables were not normally distributed, medians and IQR should be considered.

The terms “iron deficient erythropoiesis” and “iron restricted erythropoiesis” are used interchangeably. Please standardize terminology.

The ferritin cutoff of 100 µg/L (Youden index) differs from the <30 µg/L IDA definition. This discrepancy should be clearly contextualized if it is due to an adjustment for underlying inflamation.

 

Author Response

Reviewer 2

Thank you for the time and effort to review the manuscript , and for the meaningful suggestions

Thank you for permit me to review your manuscript addressing the diagnostic value of reticulocyte-derived parameters (RHe and MRV) for detecting iron-restricted erythropoiesis in elderly patients. The study addresses a clinically important issue in geriatric hematology. The manuscript addresses an important diagnostic challenge and proposes potentially useful automated parameters. However, clarification of the diagnostic reference standard, reduction of potential bias, and more robust statistical reporting are necessary before the conclusions can be considered reliable.

Major comments: 

1.- Concept of diagnosis: the definition stablish that soluble transferrin receptor (sTfR >52 nmol/L) is used as the “gold standard” for iron deficiency. However the groups (IDA, ACD, mixed anemia) are defined using ferritin, TSAT, CRP, and sTfR but there is no table, even as supplementary material with distribution (median/IQR if not normal) of CRP and other markers by group; and clarify how many meet CRP>5. CRP, serum iron, or transferrin values.

We thank the reviewer for this important remark and have clarified the diagnostic reference standard. In the revised Methods we explicitly state that soluble transferrin receptor (sTfR ≥ 52 nmol/L) was used as the reference standard for iron‑restricted erythropoiesis, while ferritin, TSAT and CRP were used to classify patients into IDA, ACD and mixed ACD/IDA. We have added , s‑iron, transferrin and CRP values to Table 2 

 

2.- ROC analyses are presented for “ron restricted erythropoiesis,” but the exact definition of this endpoint is not consistently described. Only includes ferritin sensitivity/specificity for some points; 95% CI is missing and the list appears incomplete. It is unclear whether the study evaluates: Absolute iron deficiency, Functional iron deficiency, Iron-restricted erythropoiesis, or a composite condition (IDA + mixed + FID). This must be explicitly clarified.

We acknowledge the lack of an explicit definition of the ROC endpoint and have now clarified this. In the revised Methods and Results we specify that “iron‑restricted erythropoiesis” for ROC analyses was defined by sTfR ≥ 52 nmol/L, irrespective of the underlying clinical category (IDA, ACD or mixed IDA/ACD). Thus, patients with absolute or functional iron deficiency (IDA + mixed IDA/ACD) were grouped as having iron‑restricted erythropoiesis, while those with sTfR < 52 nmol/L were considered to have sufficient iron supply.

We have also standardised terminology throughout the manuscript to “iron‑restricted erythropoiesis” and now explicitly distinguish absolute iron deficiency (IDA), functional iron deficiency within inflammatory states, and iron‑restricted erythropoiesis as the functional endpoint captured by sTfR, RHe and MRV.

3.-  Inflammation and Chronic Kidney Disease, as this is a large ederly cohort (mean age 80.5 years) with a high burden of comorbidities, including chronic renal disease and inflammatory conditions. This situations may influence sTfR and erythropoietic indices, CRP affects ferritin interpretation, and then the absence of stratified or multivariable analysis limits interpretability.

I am suggesting to stratify analyses according to CRP levels (e.g., <5 vs ≥5 mg/L), Stratification by renal function, or to plan logistic regression models adjusting for CRP and renal function to assess incremental diagnostic value.

We agree that inflammation and renal function are key modifiers in this elderly cohort,but only 18.7% of our patients suffered CKD, so the results only can be considered as preliminary. We have added a description of eGFR in the Results section and a short paragraph in the Discussion explicitly addressing their potential impact on reticulocyte indices.

We performed an exploratory stratified ROC analysis according to CRP (≤ 5 vs > 5 mg/L) and renal function (eGFR ≥ 60 vs < 60 mL/min/1.73 m²). The AUCs for MRV and RHe remained high and of similar magnitude across strata, whereas ferritin performance deteriorated in the high‑CRP and reduced eGFR subgroups; these findings are now briefly reported in the Results, given the sample size those are preliminary data.

​We considered multivariable logistic regression but judged that the number of outcome events relative to covariates would not allow a robust model without overfitting. We therefore restricted the additional analysis to stratification and emphasise this limitation in the Discussion.​

Minor comments:

The study is described as both “prospective” and “cross-sectional.” Please clarify whether this refers to prospective enrollment with cross-sectional analysis.

We appreciate the need for clearer wording. We have revised the description of the design to “prospectively enrolled cohort with cross‑sectional analysis” and removed ambiguous phrasing. This is now stated consistently in the Methods.

Normality testing is reported, yet results are presented as mean ± SD throughout. If variables were not normally distributed, medians and IQR should be considered.

We thank the reviewer for this observation. We used the Kolmogorov–Smirnov test to assess distribution; variables were normally distributed so are  expressed as mean ± SD.​

The terms “iron restricted erythropoiesis” and “iron restricted erythropoiesis” are used interchangeably. Please standardize terminology.

We agree that consistent terminology is important. The manuscript has been revised to preferentially use the term “iron‑restricted erythropoiesis” for the functional endpoint, while “iron deficiency anemia” and “functional iron deficiency” are reserved for etiologic categories, in line with current literature.​

The ferritin cutoff of 100 µg/L (Youden index) differs from the <30 µg/L IDA definition. This discrepancy should be clearly contextualized if it is due to an adjustment for underlying inflamation.

We thank the reviewer for highlighting this potential confusion. The s- ferritin cut‑off of 100 µg/L derives from ROC analysis (Youden index) in this elderly, multimorbid cohort, and is distinct from the conventional 30 µg/L threshold used to define absolute iron deficiency (IDA) in the absence of inflammation.

 We explicitly state in the discussion that, in older patients with frequent chronic inflammation, higher ferritin thresholds (e.g., 50–100 µg/L) are often recommended to improve sensitivity for iron deficiency, which is consistent with our finding that 100 µg/L provided the best diagnostic balance in this population.

 

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Upon reviewing the revised article in detail, I was pleased to see that the authors had thoroughly addressed the suggestions. In its current form, the article has reached a level that will contribute significantly to the literature. The authors conducted a thorough and successful revision process, making necessary text additions and adding diagrams and graphs. These changes clearly enhanced the scientific validity and readability of the article.

Author Response

Thank you for your time and effort to review the manuscript, pleased to see the reviewer considers V1 a good contribution to the topic 

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript entitled:"The role of Reticulocyte derived parameters in the detection of iron restricted erythropoiesis in the elderly"  has improved substantially compared with the previous version, particularly in to clarify the study aim, expand the pathophysiological rationale and improve the discussion regarding ferritin thresholds in elderly populations with exploratory analyses regarding renal function and inflammation

However, several methodological and reporting issues still limit the interpretability and generalizability of the results:

1.- It is neccesary to clarify the study design, because appears to be a cross-sectional diagnostic accuracy study using prospectively collected samples, rather than a prospective cohort. Possibly is better to describe as:"This study represents a cross-sectional diagnostic accuracy analysis performed on prospectively collected samples from elderly outpatients undergoing routine laboratory testing."

2.- The reference standard sTfR ≥ 52 nmol/L as the gold standard for iron-restricted erythropoiesis, is not universally accepted as a definitive reference standard because the inflammation and erythropoietic activity can influence sTfR levels and the classical gold standard remains bone marrow iron staining. In this way a short paragraph discussing this limitation would strengthen the methodological transparency.

3.- Concerning statistical analysis: the dataset appears sufficiently large to allow a simple multivariable logistic regression model, including: MRV, RHe, CRP, ferritin, eGFR. In addition a formal comparison between the ROC curves is needed to ensure that MRV and RHe significantly outperform ferritin.

4.- The discussion is somewhat brief in terms of novelty, since the MRV cutoffs have not previously been reported for elderly populations. Authors should emphasize the novelty of the work, highlighting the biological explanation of MRV decrease during iron-restricted erythropoiesis and the potential analytical differences between platforms, supporting the advantages offered by MRV.

There are some typographical errors for example: "22.3 % had abnormal renal function (eGFR ≥ 60 mL/min/1.73 m²)" should correspond to eGFR < 60 mL/min/1.73 m²

Author Response

The manuscript entitled:"The role of Reticulocyte derived parameters in the detection of iron restricted erythropoiesis in the elderly"  has improved substantially compared with the previous version, particularly in to clarify the study aim, expand the pathophysiological rationale and improve the discussion regarding ferritin thresholds in elderly populations with exploratory analyses regarding renal function and inflammation.

Thank you for the time and effort inn the revision of the maniuscript.

However, several methodological and reporting issues still limit the interpretability and generalizability of the results:

1.- It is neccesary to clarify the study design, because appears to be a cross-sectional diagnostic accuracy study using prospectively collected samples, rather than a prospective cohort. Possibly is better to describe as:"This study represents a cross-sectional diagnostic accuracy analysis performed on prospectively collected samples from elderly outpatients undergoing routine laboratory testing."

We thank the reviewer for this helpful clarification.
We have revised the description of the study design in the Methods section to state that this is a cross‑sectional diagnostic accuracy analysis based on prospectively collected samples from elderly outpatients undergoing routine laboratory testing, as suggested. We adjusted the Study Design to ensure consistent terminology throughout the manuscript.

2.- The reference standard sTfR ≥ 52 nmol/L as the gold standard for iron-restricted erythropoiesis, is not universally accepted as a definitive reference standard because the inflammation and erythropoietic activity can influence sTfR levels and the classical gold standard remains bone marrow iron staining. In this way a short paragraph discussing this limitation would strengthen the methodological transparency.

We agree that sTfR ≥ 52 nmol/L is not a universally accepted definitive reference standard and that inflammation and erythropoietic activity may influence sTfR concentrations.​
We have added a paragraph in the Methods/Limitations section explicitly acknowledging that bone marrow iron staining remains the classical gold standard and discussing how these factors may affect sTfR, thus clarifying the limitations of our chosen reference standard. In addition, we briefly justify our choice of sTfR in this elderly outpatient population, emphasizing its practicality and reduced invasiveness compared with bone marrow examination.

3.- Concerning statistical analysis: the dataset appears sufficiently large to allow a simple multivariable logistic regression model, including: MRV, RHe, CRP, ferritin, eGFR. In addition a formal comparison between the ROC curves is needed to ensure that MRV and RHe significantly outperform ferritin.

We have now fitted a multivariable logistic regression model with iron‑restricted erythropoiesis stratified  according to inflammation and renal function . The results of this model, including adjusted odds ratios and confidence intervals, have been added to the Results section . We have added the comparisons of the ROC curves to assess whether MRV and RHe significantly outperform ferritin .

4.- The discussion is somewhat brief in terms of novelty, since the MRV cutoffs have not previously been reported for elderly populations. Authors should emphasize the novelty of the work, highlighting the biological explanation of MRV decrease during iron-restricted erythropoiesis and the potential analytical differences between platforms, supporting the advantages offered by MRV.

We thank the reviewer for pointing out the need to better highlight the novelty of our findings. We have expanded the Discussion to emphasize that, to our knowledge, MRV cut‑off values have not previously been reported specifically for elderly populations.

 

There are some typographical errors for example: "22.3 % had abnormal renal function (eGFR ≥ 60 mL/min/1.73 m²)" should correspond to eGFR < 60 mL/min/1.73 m²

The typographical error in the description of renal function has been corrected.