Professional Documents
Culture Documents
Relation between changes in red cell distribution width and clinical outcomes in
acute decompensated heart failure
Badira F. Makhoul a, d, Amal Khourieh d, Marielle Kaplan b, d, Fadel Bahouth c,
Doron Aronson c, d, e, Zaher S. Azzam a, d, e,
a
Department of Internal Medicine B, Rambam Health Care Campus, Haifa, Israel
b
The Laboratory of Clinical Biochemistry, Rambam Health Care Campus, Haifa, Israel
c
Heart Institute, Rambam Health Care Campus, Haifa, Israel
d
Ruth & Bruce Rappaport Faculty of Medicine, Technion, Israel Institute of Technology, Haifa, Israel
e
The Rappaport Family, Institute for Research in the Medical Sciences, Technion, Israel Institute of Technology, Haifa, Israel
a r t i c l e i n f o a b s t r a c t
Article history: Background: Increased red blood cell distribution (RDW) has been associated with adverse outcomes in
Received 17 September 2011 patients with heart failure. We studied the association between baseline RDW and changes in RDW during
Received in revised form 29 March 2012 hospital course with clinical outcomes in acute decompensated heart failure (ADHF) patients.
Accepted 8 April 2012 Methods and results: We prospectively studied 614 patients with ADHF. Baseline RDW and RDW change
Available online 2 May 2012
during hospital course were determined. The relationship between RDW and clinical outcomes after hospital
discharge was tested using Cox regression models, adjusting for clinical characteristics, echocardiographic
Keywords:
Anemia
ndings and brain natriuretic peptide levels. During follow up (1 year), 286 patients (46.6%) died and 84
Heart failure were readmitted for ADHF (13.7%). Median RDW was signicantly higher among patients who died com-
Red cell distribution width pared to patients who survived (15.6% interquartile range [14.5 to 17.1] vs. 14.9% mg/L interquartile range
[14.1 to 16.1], P b 0.0001). Compared with patients in the 1st RDW quartile, the adjusted hazard ratio [HR]
for death or rehospitalization was 1.9 [95% CI 1.32.6] in patients in the 4th quartile. Changes in RDW during
hospitalization were strongly associated with changes in mortality risk. Compared with patients with persis-
tent normal RDW (b 14.5%), the adjusted HR for mortality was 1.9 [95% CI 1.13.1] for patients in whom RDW
increased above 14.5% during hospital course, similar to patients with persistent elevation of RDW (HR was
1.7, 95% CI 1.22.3).
Conclusion: In patients hospitalized with ADHF, RDW is a strong independent predictor of greater morbidity
and mortality. An increase in RDW during hospitalization also portends adverse clinical outcome.
2012 Elsevier Ireland Ltd. All rights reserved.
0167-5273/$ see front matter 2012 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.ijcard.2012.04.065
B.F. Makhoul et al. / International Journal of Cardiology 167 (2013) 14121416 1413
RDW in our laboratory is 11.5 to 14.5%. The correctness of this normal range was con-
rmed by analyzing RDW data in 17293 ambulatory subjects who attended the
Rambam Center for Preventive Medicine for a medical examination and health
counseling. In this group, mean RDW was 13.1% (median 13.0%) with 95% condence
interval of RDW of 12.0 to 14.4%. Patient BNP levels were measured with the AxSYM
BNP microparticle enzyme immunoassay (Abbott Laboratories, Abbott Park, IL, USA).
All patients were followed for 12 months after hospital discharge. The primary end
point of the study was all-cause mortality [13] or readmission for the management of
HF. Readmission for HF was dened if the main reason for a new admission was new
symptoms of dyspnea with pulmonary venous congestion on X-ray with interstitial
or alveolar edema and BNP > 400 pg/mL. Following hospital discharge, clinical endpoint
information was acquired by reviewing the national death registry and by contacting
each patient individually and independently reviewing the hospital course for major
clinical events if the patient had been re-hospitalized.
Table 1
Baseline clinical characteristics according to quartiles of baseline RDW.
RDW quartile
Values are expressed as number (%) of patients, mean value SD, or Median [Interquartile Range].
ACE = angiotensin converting enzyme; ARB = Angiotensin receptor blockers.
1414 B.F. Makhoul et al. / International Journal of Cardiology 167 (2013) 14121416
Table 2 I, BNP levels and medical therapy (beta blockers, angiotensin converting-enzyme in-
Multiple linear regression analysis with Ln RDW as the dependent variable. hibitors, loop diuretics, spironolactone and digoxin). The Cox models were also adjusted
for baseline hemoglobin and for left ventricular ejection fraction (LVEF). Similar adjust-
Unadjusted Adjusted ments were used to model changes in RDW with additional adjustments for length of
Independent variable Regression P value Regression P value hospital stay in days, intravenous iron therapy, erythropoietin therapy and blood
coefcient (SE) coefcient (SE) transfusion.
The relation between changes in RDW during hospital course and the primary end-
Age (per 10 years) 0.009 (0.001) 0.004 0.009 (0.003) 0.001 point was also analyzed by categorizing the patients according to RDW at baseline and
Atrial brillation 0.02 (0.01) 0.007 0.03 (0.01) 0.002 follow-up values. Differences were considered statistically signicant at the 2-sided
Beta blockers 0.02 (0.01) 0.02 P b 0.05 level. Statistical analyses were performed using the SPSS statistical software
Loop diuretic at 0.04 (0.01) b 0.0001 0.02 (0.001) 0.046 version 16.0 (Chicago, IL) and STATA version 11.0 (College station, TX).
presentation
Estimated GFR 0.004 (0.002) 0.04
(per 10 ml/min 3. Results
increase)
Blood urea nitrogen 0.01 (0.002) b 0.0001 0.01 (0.002) b0.0001
(per 10 mg/ml
During the study period, 614 patients who met the inclusion cri-
increase) teria were recruited. The median time between the baseline and in-
Ln BNP 0.02 (0.007) 0.0007 0.01 (0.006) 0.02 hospital RDW measurements was 5 days (interquartile range 2 to
Hemoglobin (g/dL) 0.02 (0.002) b 0.0001 0.01 (0.002) b0.0001 11 days). The distribution for RDW in normal subjects and the study
MCV () 0.005 (0.001) b 0.0001 0.004 (0.001) b0.0001
HF patients is shown in Fig. 1. The RDW distribution curve is shifted
to the right in HF patients with marked increase in RDW values.
The clinical characteristics of patients according to quartiles of
baseline RDW are shown in Table 1. Patients with higher RDW values
had reduced renal function, lower baseline hemoglobin and MCV, and
presented with higher BNP levels. They were more likely to be on
loop diuretics. Intravenous iron, erythropoietin and blood transfusion
were more likely to be used in the highest RDW quartile.
Table 2 depicts clinical and laboratory parameters that were inde-
pendently associated with RDW. Of the measures of renal function,
RDW was more strongly associated with BUN than with eGFR. There
was a poor correlation between RDW and BNP in both patients with
(r = 0.11, P = 0.14) and without anemia (r = 0.12, P = 0.003). Of
note, there was no association between RDW and left ventricular
ejection fraction. The nal model explained only 22% of the variability
in RDW.
Table 3
Unadjusted and adjusted Cox's proportional hazards model for all-cause mortality according to quartile of RDW.
Baseline RDW
1st Quartile 1.0 b0.0001 1.0 b 0.0001
2nd Quartile 0.9 (0.61.3) 0.45 0.9 (0.61.3) 0.63
3rd Quartile 1.3 (0.91.8) 0.15 1.2 (0.81.7) 0.28
4th Quartile 1.8 (1.32.5) 0.0002 1.9 (1.32.6) b0.0001
BNP
1st Quartile 1.0 b0.0001 1.0 0.08 0.01
2nd Quartile 1.4 (0.05) 0.05 1.4 (1.02.9) 0.13
3rd Quartile 1.5 (1.12.2) 0.02 1.3 (0.91.9) 0.01
4th Quartile 2.1 (1.52.9) b0.0001 1.6 (1.12.4)
Age (per 10 years 1.3 (1.11.4) b0.0001 1.2 (1.11.4) b0.0001
increase)
Bun (per 10 mg/dL 1.2 (1.101.3) b0.0001 1.1 (1.01.2) 0.001
increase)
Elevated cTnI 1.6 (1.22.2) 0.002 1.5 (1.12.1) 0.02
LVEF b 50% 1.8 (1.42.4) b0.0001 1.6 (1.22.1) 0.004
Beta blocker 0.8 (0.61.0) 0.03 0.7 (0.60.9) 0.02
* The Cox models were adjusted for age, gender, history of diabetes, hypertension, smoking status, estimated glomerular ltration rate (eGFR), blood urea nitrogen (BUN), serum
sodium, atrial brillation, elevated cardiac troponin I, BNP levels and medical therapy (beta blockers, angiotensin converting-enzyme inhibitors, loop diuretics, spironolactone and
digoxin), baseline hemoglobin and for left ventricular ejection fraction.
B.F. Makhoul et al. / International Journal of Cardiology 167 (2013) 14121416 1415
rates were 44.9%, 43.0%, 54.7% and 67.8% in the rst, second, third and
fourth RDW quartile respectively.
The results of a multivariable Cox model are shown in Table 3. Both
RDW and BNP levels remained independent predictors of 1-year mor-
tality or rehospitalization for HF together with age, reduced LVEF,
elevated BUN, elevated cTnI and use of beta-blockers. The most pro-
nounced increase in mortality occurred from the third to the forth
RDW quartile. Using RDW as a continuous rather than categorical vari-
able in the same model, the adjusted HR for 1% increase in RDW during
hospital course was 1.15 (95% CI 1.081.21, P b 0.0001).
Acknowledgement