Methodological considerations in interpreting RDW‐CV as a biomarker for renal tubular injury in type 2 diabetes

To the Editor,

We read with great interest the study by Miyataka et al. ¹ examining the association between red blood cell distribution width (RDW‐CV) and renal tubular injury in individuals with type 2 diabetes. The finding that RDW‐CV correlates specifically with urinary L‐FABP excretion, but not with albuminuria, offers a potentially cost‐effective biomarker for tubular injury in diabetic kidney disease. However, we have several methodological concerns that potentially threaten the validity of the conclusions.

Firstly, the multivariate regression models presented in tables 2, 4, and S1 appear to include a large number of predictors relative to the sample size of 490 participants. For instance, table 2 includes up to 15 covariates for each renal outcome. Although variance inflation factors (VIFs) are reported and mostly below 2, the risk of overfitting remains when many variables are included without explicit variable selection or regularization, especially in a cross‐sectional study without external validation. Overfitting can inflate effect estimates and reduce the generalizability of the reported associations. For example, the independent association between RDW‐CV and uL‐FABPCR might be less stable in a new cohort. To strengthen the analysis, the authors could consider using penalized regression techniques or bootstrapping for internal validation, and they should report model performance metrics such as adjusted R ² for each final model to better inform readers about the explanatory power of RDW‐CV beyond other factors ² .

Additionally, while the authors adjusted for several clinical and pharmacological confounders, residual confounding remains a significant concern. RDW‐CV is known to be influenced by inflammation, nutritional deficiencies (e.g., iron, B12, and folate), and subclinical comorbidities, none of which were comprehensively measured or adjusted for in this study. For instance, C‐reactive protein or interleukin‐6 levels—key markers of inflammation that affect both RDW and tubular injury—were not included. Similarly, data on iron stores, folate, or vitamin B12 were unavailable, as acknowledged in the limitations, but their absence in the multivariate models means that the observed association between RDW‐CV and tubular injury could partly reflect these unmeasured common causes. This is particularly relevant given that RDW‐CV remained significant only for tubular markers and not for albuminuria or eGFR, which might be differentially affected by such confounders. Future studies should incorporate these biomarkers into adjusted analyses or, if unavailable, explicitly discuss how their omission might bias the interpretation of RDW‐CV as a specific tubular injury marker ³ .

In summary, while the study introduces a promising and accessible biomarker for diabetic tubular injury, the robustness of the conclusions depends on addressing potential overfitting in multivariable modeling and unmeasured confounding. Ensuring model stability and accounting for key biological determinants of RDW will be essential to validate RDW‐CV as a reliable, specific indicator of renal tubular damage in type 2 diabetes. These refinements would not only strengthen the present findings but also enhance their translational relevance for clinical screening strategies.

DISCLOSURE

The authors declare no conflict of interest.

Approval of the research protocol: N/A.

Informed consent: N/A.

Approval date of registry and the registration no. of the study/trial: N/A.

Animal Studies: N/A.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.