Elevated Red Blood Cell Distribution Width Is Associated with Greater Risk of Morphometric Vertebral Fracture in Elderly Women Independent of Anemia, Inflammation, and Nutritional Status

Presentation Number: SAT 310
Date of Presentation: April 1st, 2017

Namki Hong*1, Da Hea Seo1, Hokyou Lee2, Chang Oh Kim3, Yoosik Youm4, Hyeon Chang Kim5 and Yumie Rhee1
1Department of Internal Medicine, Severance Hospital, Endocrine Research Institute, Yonsei University College of Medicine, Seoul, Korea, Republic of (South), 2Department of Internal Medicine, Severance Hospital, Endocrine Research Institute, Yonsei University College of Medicine, 3Yonsei University College of Medicine, Seoul, Korea, Republic of (South), 4Yonsei University, Seoul, Korea, Republic of (South), 5Yonsei University College of Medicine


Red blood cell distribution width (RDW), a conventional marker for ineffective erythropoiesis, has been found to be associated with mortality in many diseases including coronary heart disease, heart failure, and in community-dwelling elderly population (1). A recent study demonstrated that elevated RDW also predicted mortality in a hip fracture cohort (2). Furthermore, rapid hip bone loss in elderly men was associated with increased odds of anemia and low lymphocytes, suggesting the possible interdependency between hematopoiesis and bone health (3). However, whether RDW is associated with vertebral fracture (VF), the most common fracture in elderly, has not been investigated yet. Data of 576 elderly women were collected between 2012 and 2015 in a community-based cohort study (4). Primary outcome was presence of morphometric VF assessed by lateral thoracolumbar radiograph. Information regarding fracture history, comorbidities, and medication use was collected by interviewer-assisted questionnaire. Physical performance was measured by timed up-and-go test. After excluding those with estimated glomerular filtration rate (eGFR) < 30 ml/min/1.73m2, history of malignancy, or without available VF assessment or RDW measurements, a total of 559 subjects were analyzed. Mean age and body mass index of study population was 71.1 ± 4.3 years and 24.5 ± 3.1 kg/m2, respectively. Morphomectric VF was detected in 127 subjects (22.7 %). Mean RDW was higher in subjects with VF compared to those without (13.1 ± 0.5 % vs. 12.9 ± 0.5 %, P = 0.005). Prevalence of VF increased in stepwise fashion from 16.9 % in lowest RDW tertile (RDW < 12.7 %) to 27.5 % in highest tertile (RDW ≥ 13.0 %, P for trend = 0.014). RDW was positively correlated with age, serum osteocalcin, and alkaline phosphatase level, whereas hemoglobin, ferritin, albumin, and eGFR showed negative correlation with RDW. However, serum 25-hydroxyvitamin D, C-reactive protein (CRP) level, and lumbar spine areal bone mineral density (LSBMD) did not correlate with RDW. In logistic regression analysis, higher RDW value was independently associated with greater odds of morphomectric VF (odds ratio 1.49, 95% CI 1.01-2.20, P = 0.043) after adjustment for hemoglobin and conventional risk factors including age, history of fracture, smoking status, physical performance, previous steroid use and hormone replacement therapy, LSBMD, and 25-hydroxyvitamin D level. Further adjustment for serum ferritin as an indicator of iron status, osteocalcin, CRP, albumin, and eGFR did not attenuate the association of RDW with prevalent VF (odds ratio 1.55, 95% CI 1.04-2.31, P = 0.029). Our findings suggest that elevated RDW level, a readily available marker in most clinical settings, may be an independent risk factor for prevalent VF, which merits further investigation in line with crosstalk between hematopoietic system and bone metabolism.


Nothing to Disclose: NH, DHS, HL, COK, YY, HCK, YR