Red blood cell distribution width is an indicator of physiological dysregulation and reveals the relationship between biomarkers and health status in hemodialysis patients

Introduction: Red blood cell distribution width (RDW) and intra-individual variability (IIV) of biomarkers are different types of variation. Although their high levels are commonly associated with high mortality risks and poor clinical outcomes, the reason for this remains unclear. Methods: We explored the significance of this similarity using time-series data (from 143,475 samples) collected over a 9-year period from 1,216 patients on long-term hemodialysis. The IIV of each of the 22 blood-based biomarkers was assessed using the moving coefficient of variation (mCV). Non-linear longitudinal/temporal changes of the variables and their mutual relationships were analyzed using generalized additive (mixed) models. Results: RDW and all of the IIVs showed similar trends throughout the entire period of dialysis treatment: a rapid decrease after hemodialysis initiation followed by an increase that accelerated prior to death. The slopes of these downward and upward trends during each one-year period were statistically different from zero for RDW and the majority of the IIVs. Given the reported increase in the RDW and IIVs of several biomarkers in the pre-dialysis stages of chronic kidney disease, these shared changes might reflect alterations in homeostatic capacity. The RDW exhibited positive correlations with the mCVs of all 22 biomarkers (all p-values < 0.0009), suggesting that it represents dysregulation of the overall physiological system. Furthermore, the levels of the biomarkers showed a specific (nearly linear or U-shaped) relationship with the RDW, analogous to their reported relationship to all-cause mortality in the hemodialysis population. Conclusion: Increased RDW could serve as a marker of physiological dysregulation, and measurement of the RDW can be an effective tool for assessing the overall health status and identifying optimal ranges for various health indicators.