Red blood cell distribution width to albumin ratio associates with prevalence and long-term diabetes mellitus prognosis: an overview of NHANES 1999–2020 data

Background Erythrocyte dysfunction is a characteristic of diabetes mellitus (DM). However, erythrocyte-associated biomarkers do not adequately explain the high prevalence of DM. Here, we describe red blood cell distribution width to albumin ratio (RAR) as a novel inflammatory biomarker for evaluating an association with DM prevalence and prognosis of all-cause mortality. Methods Data analyzed in this study were extracted from the National Health and Nutrition Examination Survey (NHANES) 1999−2020. A total of 40,558 participants (non-DM and DM) were enrolled in the study; RAR quartiles were calibrated at Q1 [2.02,2.82] mL/g, Q2 (2.82,3.05] mL/g, Q3 (3.05,3.38] mL/g, and Q4 (3.38,12.08] mL/g. A total of 8,482 DM patients were followed (for a median of 84 months), of whom 2,411 died and 6,071 survived. The prevalence and prognosis associated with RAR and DM were analyzed; age and sex were stratified to analyze the prevalence of RAR in DM and the sensitivity of long-term prognosis. Results Among non-DM ( n =30,404) and DM ( n =10,154) volunteers, DM prevalence in RAR quartiles was 8.23%, 15.20%, 23.92%, and 36.39%. The multivariable odds ratio (OR) was significant for RAR regarding DM, at 1.68 (95% CI 1.42, 1.98). Considering Q1 as a foundation, the Q4 OR was 2.57 (95% CI 2.11, 3.13). The percentages of DM morbidity varied across RAR quartiles for dead ( n =2,411) and surviving ( n =6,071) DM patients. Specifically, RAR quartile mortality ratios were 20.31%, 24.24%, 22.65%, and 29.99% (P<0.0001). The multivariable hazard ratio (HR) for RAR was 1.80 (95% CI 1.57, 2.05). Considering Q1 as a foundation, the Q4 HR was 2.59 (95% CI 2.18, 3.09) after adjusting for confounding factors. Sensitivity analysis revealed the HR of male DM patients to be 2.27 (95% CI 1.95, 2.64), higher than females 1.56 (95% CI 1.31, 1.85). DM patients who were 60 years of age or younger had a higher HR of 2.08 (95% CI1.61, 2.70) as compared to those older than 60 years, who had an HR of 1.69 (95% CI 1.47, 1.94). The HR of RAR in DM patients was optimized by a restricted cubic spline (RCS) model; 3.22 was determined to be the inflection point of an inverse L-curve. DM patients with a RAR >3.22 mL/g suffered shorter survival and higher mortality as compared to those with RAR ≤3.22 mL/g. OR and HR RAR values were much higher than those of regular red blood cell distribution width. Conclusions The predictive value of RAR is more accurate than that of RDW for projecting DM prevalence, while RAR, a DM risk factor, has long-term prognostic power for the condition. Survival time was found to be reduced as RAR increased for those aged ≤60 years among female DM patients.