Increased red blood cell distribution width (RDW), a marker of anisocytosis, has been associated with adverse outcomes in multiple settings. Whether RDW is predictive of mortality in patients with peripheral artery disease (PAD) is unknown. We studied 13,039 consecutive outpatients (69.5 ± 12.0 years of age, 60.9% men, 97.6% white) with PAD identified by noninvasive lower-extremity arterial testing at the Mayo Clinic from January 1997 through December 2007, with follow-up through September 2009. We defined PAD as a low (≤0.9) or high (≥1.4) ankle–brachial index (ABI). Cardiovascular risk factors and co-morbidities were ascertained using electronic medical record–based algorithms. RDW was obtained from the complete blood cell count drawn around the time of arterial evaluation. Mortality was ascertained using the Mayo electronic medical record and Accurint databases. Association of RDW with all-cause mortality was analyzed by multivariable Cox proportional hazards regression. During a median follow-up of 5.5 years, 4,039 (31.0%) deaths occurred (28.7% in low and 38.9% in high ABI subsets). After adjustment for age, gender, cardiovascular risk factors, and co-morbidities, patients in the highest quartile of RDW (>14.5%) had a 66% greater risk of mortality compared to the lowest quartile (<12.8%, p <0.0001); a 1% increment in RDW was associated with a 10% greater risk of all-cause mortality (hazard ratio 1.10, 95% confidence interval 1.08 to 1.12, p <0.0001). The adjusted hazard ratio was similar in the low (1.10, 1.08 to 1.12) and high (1.09, 1.06 to 1.12) ABI subsets. In conclusion, RDW, a routinely available measurement, is an independent prognostic marker in patients with PAD.
Peripheral arterial disease (PAD), a surrogate for systemic atherosclerotic vascular disease, affects approximately 8 million patients in the United States and is associated with increased mortality and morbidity. Predictors of mortality in patients with PAD are not well-defined and would be valuable for risk stratification and clinical decision-making, especially if routinely and inexpensively obtained. The ankle–brachial index (ABI) is an established noninvasive test for PAD, which is defined as a low (≤0.9) or high (≥1.4) ABI. Low ABI results from arterial lumen narrowing because of atherosclerosis and high ABI results from medial arterial calcification and poorly compressible arteries. Whether red blood cell distribution width (RDW) is associated with mortality in PAD is not known, and we therefore sought to investigate the prognostic value of RDW in patients with PAD.
Methods
From January 1, 1997 through December 31, 2007, 20,996 outpatients ≥18 years of age were referred to the noninvasive vascular laboratory of the Gonda Vascular Center (Mayo Clinic, Rochester, Minnesota) for lower-extremity arterial evaluation. ABI was measured using an established protocol. We excluded patients who refused participation in research (n = 760) and patients with nonatherosclerotic vascular diseases (n = 1,154) such as vasculitides using a set of appropriate International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis and procedure codes. We identified 13,039 patients with PAD defined as (1) ABI ≤0.9 at rest or 1 minutes after exercise or (2) presence of poorly compressible arteries (any ABI in 1 leg ≥1.4 or ankle systolic blood pressure ≥255 mm Hg). The index date was the date of lower-extremity arterial testing with abnormal ABI measurement. Follow-up was started at the index date and ended on September 30, 2009. The study was approved by the institutional review board of the Mayo Clinic.
ICD-9-CM codes up to 6 months after the index date and natural language processing were used to ascertain demographic information, smoking status, medications, conventional cardiovascular risk factors, and co-morbidities from the Mayo electronic medical record (EMR) as previously described. Outpatient complete blood cell count and other laboratory variables closest to the index date were obtained from the laboratory database. RDW was extracted as part of the result of the complete blood cell count obtained in the central clinical laboratory using a Sysmex XE-5000 hematology analyzer (Sysmex America, Inc., Mundelein, Illinois). Systolic and diastolic blood pressure levels at rest were obtained as structured observations from the EMR. Patients were identified as smokers if they were actively smoking or had discontinued smoking within 1 year before the index date. Hypertension was defined as systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg during 2 serial measurements within 3 months closest to the index date or a previous diagnosis of hypertension with use of antihypertensive medication. Diabetes was defined as a fasting blood glucose level ≥126 mg/dl, a random glucose measurement >200 mg/dl or hemoglobin A 1c >6.5%, or a previous diagnosis with any use of oral antidiabetic agent and/or insulin. Dyslipidemia was defined as a total cholesterol level >220 mg/dl, high-density cholesterol level <40 mg/dl in men or <45 mg/dl in women, triglyceride level >200 mg/dl, or use of a lipid-lowering medication.
We used the following ICD-9-CM codes to identify co-morbidities: 410.xx to 414.xx for coronary heart disease including ischemic heart disease or 36.10 to 36.14 for a history of percutaneous coronary intervention or coronary artery bypass surgery; 428 for heart failure; 430.xx to 438.xx for cerebrovascular disease or 00.61, 00.63, and 38.10 for a history of carotid stenting or endarterectomy; 490 to 496 and 500 to 505, 506.4 for chronic obstructive pulmonary disease; 582 to 583.7, 585, 586, and 588 for chronic kidney disease; and 140 to 208 for malignancy. Estimated glomerular filtration rate was calculated based on the Modification of Diet in Renal Disease study formula. All-cause mortality was ascertained from the Mayo EMR and the Accurint databases.
Continuous variables are reported as mean ± SD or median. Categorical variables are reported as frequency and percentage. RDW was examined as a continuous variable and categorized into quartiles using cutoffs <12.8%, 12.8% to 13.5%, 13.5% to 14.5%, and >14.5%. We compared distribution of demographic characteristics, conventional cardiovascular risk factors, co-morbidities, and laboratory variables across RDW quartiles by analysis of variance for continuous variables and chi-square test for categorical variables. Statistically significant differences for variables were assessed after adjustment for age, gender, and body mass index using multivariable logistic regression. Cox proportional hazards regression was performed to evaluate the association of RDW with all-cause mortality adjusting for age and gender (model 1) and adjusting for body mass index, smoking status, hypertension, dyslipidemia, diabetes, cerebrovascular disease, coronary heart disease, heart failure, chronic obstructive pulmonary disease, chronic kidney disease, malignancy, aspirin and statin use, and hemoglobin level (model 2) in the overall sample and the 2 ABI subsets. Kaplan–Meier method was used to compare survival in different quartiles of RDW groups using log-rank test. Hypothesis testing was 2-tailed with a p value <0.05 considered statistically significant. Statistical analyses were performed with JMP and SAS 8.2 (SAS Institute, Cary, North Carolina).
Results
Of 13,039 patients with PAD, most were white (97.6%), 60.9% were men, and mean age was 69.5 ± 12.0 years. Prevalence of low ABI was 77.4% and that of high ABI was 22.6%. Mean RDWs were 13.9 ± 1.9% in patients with PAD, 13.8 ± 1.8 in the low ABI group, and 14.3 ± 2.1 in the high ABI group (p <0.001 after adjustment for age and gender). A larger proportion of patients in the high ABI group (18.9% in high ABI vs 10.3% in low, p <0.001 after adjustment for age and gender) had RDWs >15.6% (upper limit of Mayo Clinic normal values). Clinical characteristics of patients with PAD in quartiles of RDW are presented in Table 1 . Patients in the highest RDW quartile were older and with higher prevalence of cardiovascular risk factors and co-morbidities except for cerebrovascular disease and malignancy. There was no statistically significant difference in aspirin use across quartiles (p = 0.25). The higher quartile was associated with decreased mean corpuscular volume, hemoglobin, and estimated glomerular filtration rate. After adjustment for age and gender, the high ABI subset had more patients in the highest quartile (35.7% in high vs 23.1% in low ABI, p <0.001) and fewer patients in the lowest quartile of RDW (17.8% in high vs 24.6% in low ABI).
| Variable | Quartiles | |||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| (n = 3,008) | (n = 3,378) | (n = 3,273) | (n = 3,380) | |
| Age (years) | 67.8 (12.9%) | 69.4 (11.9%) | 70.6 (11.3%) | 70.4 (11.6%) |
| Men | 1,758 (58.4%) | 2,134 (63.2%) | 2,014 (61.5%) | 2,039 (60.3%) |
| Body mass index (kg/cm 2 ) | 28.5 (5.1%) | 28.9 (5.4%) | 29.4 (5.8%) | 29.2 (6.6%) |
| Low ankle–brachial index | 2,482 (82.5%) | 2,729 (80.8%) | 2,550 (77.9%) | 2,326 (68.8%) |
| High ankle–brachial index | 526 (17.5%) | 649 (19.2%) | 723 (22.1%) | 1,054 (31.2%) |
| Risk factors | ||||
| Dyslipidemia ⁎ | 2,336 (77.7%) | 2,646 (78.3%) | 2,577 (78.7%) | 2,573 (76.1%) |
| Hypertension | 2,132 (70.9%) | 2,506 (74.2%) | 2,527 (77.2%) | 2,543 (75.2%) |
| Diabetes | 960 (31.9%) | 1,070 (31.7%) | 1,178 (36.0%) | 1,416 (41.9%) |
| Smoking | 2,079 (79.3%) | 2,494 (82.8%) | 2,388 (80.5%) | 2,392 (79.6%) |
| Co-morbidities | ||||
| Cerebrovascular disease ⁎ | 910 (30.3%) | 1,078 (31.9%) | 1,053 (32.2%) | 1,035 (30.6%) |
| Coronary heart disease † | 1,480 (49.2%) | 1,826 (54.1%) | 1,926 (58.9%) | 2,051 (60.7%) |
| Heart failure | 267 (8.9%) | 386 (11.4%) | 539 (16.5%) | 969 (28.7%) |
| Chronic obstructive pulmonary disease | 408 (13.6%) | 530 (15.6%) | 580 (17.7%) | 647 (19.2%) |
| Chronic kidney disease | 141 (4.7%) | 202 (6.0%) | 317 (9.7%) | 644 (19.1%) |
| Malignancy ⁎ | 536 (17.8%) | 669 (19.8%) | 703 (21.5%) | 729 (21.6%) |
| Medications | ||||
| Aspirin use ⁎ | 1,345 (44.7%) | 1,549 (45.9%) | 1,546 (47.2%) | 1,556 (46.0%) |
| Statin use | 1,021 (33.9%) | 1,283 (38.0%) | 1,306 (39.9%) | 1,300 (38.5%) |
| Laboratory variables | ||||
| Red blood cell distribution width (%) | 12.3 ± 0.3 | 13.1 ± 0.2 | 13.9 ± 0.3 | 16.2 ± 2.1 |
| Erythrocyte count (10 12 /L) | 4.3 ± 0.5 | 4.3 ± 0.6 | 4.2 ± 0.6 | 4.1 ± 0.7 |
| Mean corpuscular volume (fl) | 92.0 ± 4.1 | 91.6 ± 4.5 | 90.9 ± 5.0 | 89.3 ± 8.1 |
| Hemoglobin (g/dl) | 13.5 ± 1.6 | 13.5 ± 1.7 | 13.1 ± 1.8 | 12.1 ± 1.9 |
| Creatinine (mg/dl) | 1.2 ± 0.5 | 1.3 ± 0.6 | 1.7 ± 0.7 | 1.7 ± 1.5 |
| Estimated glomerular filtration rate (ml/min/1.73 cm 2 ) | 62.1 ± 18.0 | 60.9 ± 17.8 | 58.1 ± 19.4 | 53.6 ± 24.3 |
⁎ Variables not significant at p <0.01.
† History of myocardial infarction, documented angina, abnormal stress test result, or coronary revascularization.
The association of RDW with mortality in the entire sample and in the 2 ABI subsets during the median follow-up of 5.5 years is presented in Table 2 . Patients who died were older (73.5 ± 10.3 vs 67.8 ± 12.3 years for patients alive), more often were men (63.6%), and had a higher prevalence of all cardiovascular risk factors and co-morbidities. After adjustment for age and gender, decedents had greater RDWs (14.3 ± 2.1 vs 13.7 ± 1.6 for patients alive, p <0.001), lower statin use (25.1% vs 34.5% for patients alive, p <0.001), and lower aspirin use (29.9% vs 31.9% for patients alive, p <0.001). A 1% increment in RDW was independently associated with a 15% greater risk of all-cause mortality and similar risk in the 2 ABI subsets. After additional adjustment for body mass index, cardiovascular risk factors, co-morbidities, malignancy, medications, and hemoglobin level, RDW remained an independent predictor of all-cause mortality in these groups ( Table 2 ). When RDW was analyzed according to quartiles, patients in the higher quartile (RDW >13.5%) had higher mortality compared to those in the referent quartile (RDW <12.8%; Table 3 ). We also found a graded increased risk of mortality in these patients, including patients in the fourth quartile with 45% higher and those in the third quartile with 14% higher risk of death compared to the next lower quartile. As shown by Kaplan–Meier curves ( Figure 1 ), patients in the highest quartile of RDW (>14.5%) had the poorest survival during a median follow up of 4.5 years (log-rank p <0.0001).
| Variable | Deaths/Patients (%) | Model 1 ⁎ , HR (95% CI) | Model 2 † , HR (95% CI) |
|---|---|---|---|
| Peripheral artery disease | 4,039/13,039 (31.0%) | 1.15 (1.13–1.16) | 1.10 (1.08–1.12) |
| Low ankle–brachial index | 2,891/10,087 (28.7%) | 1.15 (1.13–1.17) | 1.10 (1.08–1.12) |
| High ankle–brachial index | 1,148/2,952 (38.9%) | 1.13 (1.11–1.15) | 1.09 (1.06–1.12) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
