The association between low hemoglobin-to-red blood cell distribution width ratio and macroalbuminuria in diabetic kidney disease: evidence from NHANES and an external validation cohort

Evidence suggests that hemoglobin-to-red blood cell distribution width ratio (HRR) is associated with many diseases. Nevertheless, the association between HRR and macroalbuminuria in patients with diabetic kidney disease (DKD) remains uncertain. Using NHANES data (1999-2018), DKD was clinically diagnosed as estimated glomerular filtration rate <60 mL/min/1.73 m² or urine albumin-to-creatinine ratio (UACR) ≥30 mg/g in patients with diabetes and divided into microalbuminuria and macroalbuminuria according to UACR. Propensity score matching (PSM), weighted logistics regression, restricted cubic spline, clinical subgroup and mediation analysis were performed to investigate the association between HRR and macroalbuminuria. Lasso regression and receiver operating characteristic curve were used to assess the diagnostic value. The HRR level in the macroalbuminuria group (0.96 ± 0.19) was significantly lower than that in the microalbuminuria group (1.01 ± 0.18, <i>p</i> < 0.001). Weighted binary logistic regression revealed that elevated HRR was associated with a significant decrease in UACR (95% confidence interval [CI]: 0.058-0.274, <i>p</i> < 0.001). Quartile analyses of HRR showed a significant trend across quartiles (Q4 vs. Q1, odds ratio in the fully adjusted model: 0.421, 95%CI: 0.284-0.620, <i>p</i> < 0.001). The restricted cubic spline curve demonstrated a graded decrease in the probability of macroalbuminuria with increasing HRR levels. Regression models including HRR had good detection value for macroalbuminuria. Low HRR is associated with macroalbuminuria in DKD in this cross-sectional study. HRR may be an effective inflammatory marker for identifying macroalbuminuria in DKD.