Higher Red Cell Distribution Width and Poorer Hospitalization-related Outcomes in Elderly Patients

Kyoung Min Kim MD, PHD; Ridhima Nerlekar MS; Gregory J. Tranah PHD; Warren S. Browner MD, PHD; Steven R. Cummings MD, PHD


J Am Geriatr Soc. 2022;70(8):2354-2362. 

In This Article

Abstract and Introduction


Background: Red cell distribution width (RDW), an index for variation of red blood cell (RBC) size, has been proposed as a potential marker for poorer outcomes in several aging-related diseases and conditions. We tested whether greater variability of RBC size, presented as a higher RDW value, predicts poor prognoses among hospitalized patients over 60 years old.

Methods: We retrospectively collected data from older hospitalized patients aged ≥60 years between January 2013 to December 2017 at Sutter Health, a large integrated health system in Northern California. The RDW was measured during hospital admission and categorized with 1% intervals (≤13.9, 14.0–14.9, 15.0–15.9, 16.0–16.9, 17.0–17.9 and ≥18.0%). The primary outcome was the rate of in-hospital mortality and secondary outcomes included 30-day re-admission rate and length of hospital stay (in days).

Results: A total of 167,292 admissions from 94,617 patients were included. The overall in-hospital mortality rate was 6.3%. As the RDW value increased, the rate of in-hospital mortality gradually increased from 2.7% for the lowest RDW category to 12.2% in the highest category (p-trend <0.001). The overall 30-day re-admission rate after discharge was 12.5% and the rate of 30-day re-admission also increased with increasing RDW categories (7.4% in the lowest group vs. 15.8% in the highest group, p-trend <0.001). Patients with the highest RDW values at admission stayed 1.5–2.0 times longer in the hospital than patients with lower RDW values who were admitted for the same causes.

Conclusions: Greater variability of RBC size is significantly associated with worse prognosis in hospitalized elderly patients, indicating higher mortality, greater risk of early re-admission, and longer hospital stay days. Risk stratification strategies for hospitalized elderly should include RDW value.


Red blood cells (RBCs), one of the most common peripheral blood cell populations, maintain relatively uniform size and red cell distribution width (RDW), reported as a component of complete blood cell (CBC), is an index for the variability of RBCs' size.[1] RDW normally ranges from 11% to 15% and lower RDW values indicate lesser variability of RBC size, whereas a higher value means greater variability in size.[1] Beyond the traditional use of RDW in hematologic fields that distinguish causes of anemia, several studies have reported that greater variability of RBC size, presented as higher RDW values, are associated with the occurrence and poorer outcomes in non-hematologic diseases, including cardiovascular diseases, osteoporotic fractures, dementia, stroke.[2–4] Furthermore, more recent clinical evidence suggests that associations between greater variability of RBC size and adverse health outcomes are also observed in community-based studies, not limited to the subjects with specific diagnoses.[5–7]

The biological mechanisms linking greater variability of RBC size and adverse health-related outcomes are not known, however emerging evidence supports the use of RDW as a marker for biologic aging, disease vulnerability, oxidative stress, and chronic inflammation.[8–10] Meanwhile, these associations, higher RDW values, and worse outcomes were reported to be independent of nutrient status or inflammation, which affects both RDW values and poor health conditions.[11,12]

Predicting the risk of mortality and re-admission are critical for in-patient care.[8] An accurate risk classification strategy for patients can also improve outcomes in clinical practice.[13,14] However, there are no strong predictors of mortality or early re-admission that can be applied across multiple aging-related diseases. In the present study, we tried to confirm the potential of the RDW as a biomarker for poorer outcomes using data from a large integrated health system comprised of diverse races/ethnicities and different diagnoses. To this end, we tested whether RDW was significantly associated with higher mortality in hospitalized elderly patients in general, and in separate diseases classes as well. We also assessed the association of RDW with immediate medical progression, including 30-day readmission rates and length of stay.