CLINICAL PERSPECTIVE

This Article Abstract Full Text (PDF) All Versions of this Article: 113/23/2713    most recent CIRCULATIONAHA.105.577577v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Go, A. S. Articles by Shlipak, M. G. Search for Related Content PubMed PubMed Citation Articles by Go, A. S. Articles by Shlipak, M. G. Related Collections Cardio-renal physiology/pathophysiology Congestive Epidemiology Related Article

(Circulation. 2006;113:2713-2723.)
© 2006 American Heart Association, Inc.

Heart Failure

Hemoglobin Level, Chronic Kidney Disease, and the Risks of Death and Hospitalization in Adults With Chronic Heart Failure

The Anemia in Chronic Heart Failure: Outcomes and Resource Utilization (ANCHOR) Study

Alan S. Go, MD; Jingrong Yang, MA; Lynn M. Ackerson, PhD; Krista Lepper, BS; Sean Robbins, MS; Barry M. Massie, MD; Michael G. Shlipak, MD, MPH

From the Division of Research, Kaiser Permanente of Northern California, Oakland (A.S.G., J.Y., L.M.A., K.L.); the Departments of Epidemiology and Biostatistics (A.S.G., L.M.A., M.G.S.) and of Medicine (A.S.G., B.M.M., M.G.S.), University of California at San Francisco; Amgen, Thousand Oaks (S.R.); and the Cardiology (B.M.M.) and General Internal Medicine (M.G.S.) Sections, San Francisco Veterans Affairs Medical Center, San Francisco, Calif.

Correspondence to Alan S. Go, MD, Division of Research, Kaiser Permanente of Northern California, 2000 Broadway St, 3rd Floor, Oakland, CA 94612-2304. E-mail Alan.S.Go{at}kp.org

Received July 21, 2005; revision received March 7, 2006; accepted April 4, 2006.

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Background— Previous studies have associated reduced hemoglobin levels with increased adverse events in heart failure. It is unclear, however, whether this relation is explained by underlying kidney disease, treatment differences, or associated comorbidity.

Methods and Results— We examined the associations between hemoglobin level, kidney function, and risks of death and hospitalization in persons with chronic heart failure between 1996 and 2002 within a large, integrated, healthcare delivery system in northern California. Longitudinal outpatient hemoglobin and creatinine levels and clinical and treatment characteristics were obtained from health plan records. Glomerular filtration rate (GFR; mL · min^–1 · 1.73 m^–2) was estimated from the Modification of Diet in Renal Disease equation. Mortality data were obtained from state death files; heart failure admissions were identified by primary discharge diagnoses. Among 59 772 adults with heart failure, the mean age was 72 years and 46% were women. Compared with that for hemoglobin levels of 13.0 to 13.9 g/dL, the multivariable-adjusted risk of death increased with lower hemoglobin levels: an adjusted hazard ratio (HR) of 1.16 and 95% confidence interval (CI) of 1.11 to 1.21 for hemoglobin levels of 12.0 to 12.9 g/dL; HR, 1.50 and 95% CI, 1.44 to 1.57 for 11.0 to 11.9 g/dL; HR, 1.89 and 95% CI, 1.80 to 1.98 for 10.0 to 10.9; HR, 2.31 and 95% CI, 2.18 to 2.45 for 9.0 to 9.9; and HR, 3.48 and 95% CI, 3.25 to 3.73 for <9.0 g/dL. Hemoglobin levels 17.0 g/dL were associated with an increased risk of death (adjusted HR, 1.42; 95% CI, 1.24 to 1.63). Compared with those with a GFR 60 mL · min^–1 · 1.73 m^–2, persons with a GFR <45 mL · min^–1 · 1.73 m^–2 had an increased mortality risk: adjusted HR, 1.39 and 95% CI, 1.34 to 1.44 for 30 to 44; HR, 2.28 and 95% CI, 2.19 to 2.39 for 15 to 29; HR, 3.26 and 95% CI, 3.05 to 3.49 for <15; and HR, 2.44 and 95% CI, 2.28 to 2.61 for those on dialysis. Relations were similar for the risk of hospitalization. The findings did not differ among patients with preserved or reduced systolic function, and hemoglobin level was an independent predictor of outcomes at all levels of kidney function.

Conclusions— Very high (17 g/dL) or reduced (<13 g/dL) hemoglobin levels and chronic kidney disease independently predict substantially increased risks of death and hospitalization in heart failure, regardless of the level of systolic function. Randomized trials are needed to evaluate whether raising hemoglobin levels can improve outcomes in chronic heart failure.

Key Words: anemia • death • epidemiology • heart failure • hemoglobin • hospitalization • kidney

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Anemia, as evidenced by reduced blood hemoglobin (or hematocrit) levels, frequently occurs in patients with chronic heart failure (HF), reportedly ranging from 9.0%¹ to 79.1%,² depending on the population studied, the severity of HF, and the criteria used for defining anemia. Reduced hemoglobin levels may contribute to worse outcomes through higher myocardial and peripheral oxygen demand, the development and accelerated progression of left ventricular (LV) hypertrophy, worsening of edema, and promotion of inflammation.³

Clinical Perspective p 2723

Several studies suggest that a lower hemoglobin level is associated with an increased risk of death in patients with HF,^4–10 whereas others do not.^11,12 Whether anemia predicts adverse outcomes independent of other prognostic factors, such as kidney dysfunction or other confounders, is not entirely clear. The generalizability of previous studies is limited by the inclusion of participants from clinical trials or other select populations; limited ethnic diversity; use of a single measure of hemoglobin or reliance on hemoglobin at the time of hospital admission, which may not reflect stable levels; lack of control for concomitant renal insufficiency or differences in therapies; and modest sample sizes.^{8–10,12–15}

We examined a large, diverse HF population to determine the independent contributions of hemoglobin level and chronic kidney disease to the risks of death and HF-related hospitalization. We hypothesized that lower levels of hemoglobin and kidney function independently increase the risk of adverse events overall and in persons with reduced or preserved LV systolic function.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Source Population and Study Sample
All subjects were members of Kaiser Permanente of Northern California, a large, integrated, healthcare delivery system providing care for >3 million members. Its population is representative of local and statewide populations, with a slightly lower representation of the extremes of the socioeconomic spectrum and age.¹⁶ The Kaiser Foundation Research Institute’s institutional review board approved the study.

The Anemia in Chronic Heart Failure: Outcomes and Resource Utilization (ANCHOR) Study analyzed data from the Kaiser Permanente chronic HF cohort, which included all adult members with diagnosed HF between January 1, 1996 and December 31, 2002. Subjects were included if they met 1 of the following criteria based on diagnoses from hospitalization, ambulatory visits, and/or emergency department databases: 1 hospitalization with a principal discharge diagnosis of HF (International Classification of Diseases, 9th revision [ICD-9] codes 398.91, 402.01, 402.11, 402.91, 428.0, 428.1, or 428.9; 21.9% of the cohort); 2 hospitalizations with a secondary discharge diagnosis of HF, for which the principal diagnosis was related to the disease (eg, coronary disease), 13.8%; 3 hospitalizations with a secondary discharge diagnosis of HF, 12.9%; 2 outpatient HF diagnoses, excluding those made in the emergency department, 45.7%; 3 emergency department diagnoses of HF, 0.001%; or 2 inpatient secondary diagnoses of HF (with any primary diagnosis) plus 1 outpatient diagnosis of HF (excluding that made in the emergency department), 5.7%. The index date was assigned at the first qualifying diagnosis date.

Hemoglobin Levels
We obtained outpatient hemoglobin measurements from health plan laboratory databases during the 12 months before a subject’s entry into the cohort until the end of follow-up or censoring. We included outpatient measurements, which are more likely to represent steady-state concentrations.

Kidney Function
We used outpatient determinations of serum creatinine to estimate the glomerular filtration rate (GFR) according to the 4-variable Modification of Diet in Renal Disease equation.^17,18 On the basis of prior work that has shown its predictive value for adverse outcomes in the general population,¹⁹ estimated GFR (in units of mL · min^–1 · 1.73 m^–2) was categorized as 60, 45 to 59, 30 to 44, 15 to 29, <15 not receiving dialysis, and maintenance dialysis.¹⁹ A GFR >60 mL · min^–1 · 1.73 m^–2 was not further delineated, given the higher degree of measurement error above this level.²⁰

Covariates
Age, sex, and self-reported race/ethnicity were identified from health plan databases. Coexisting illnesses were ascertained from relevant inpatient and outpatient diagnoses or procedures according to ICD-9 codes, laboratory results, or medication prescriptions. This information was obtained from validated hospital discharge, ambulatory visit, laboratory, and pharmacy databases, as well as validated diabetes mellitus²² and cancer²³ registries. We collected baseline and follow-up data on diagnoses of coronary disease (acute myocardial infarction, unstable angina, angina pectoris, coronary artery disease, percutaneous coronary intervention, and coronary artery bypass surgery), stroke/transient ischemic attack, peripheral arterial disease, diabetes, hypertension, cancer, thyroid disease, liver disease, lung disease, HIV infection, mitral and/or aortic valvular disease, dementia, depression, ventricular fibrillation/tachycardia, and atrial fibrillation/flutter (ICD-9 codes available on request). We ascertained information on LV systolic function status from HF management program records and hospitalization records. We defined reduced systolic function as an LV ejection fraction 40% or a qualitative description of "moderate or severely reduced" systolic function; preserved systolic function was defined by an LV ejection fraction >40% or a qualitative description of "normal or only mildly reduced" systolic function.²⁴

We controlled for use of the following medications during the 120 days before the index date and throughout follow-up, based on filled prescriptions found in pharmacy records: angiotensin-converting enzyme (ACE) inhibitors, angiotensin II receptor blockers, diuretics, ß-blockers, calcium channel blockers, statins, other lipid-lowering agents, spironolactone, direct vasodilators, and erythropoietin. Longitudinal medication use was estimated from drug refill patterns according to the calculated number of days supply for each prescription. For any 2 consecutive prescriptions, if the second prescription was filled within 30 days of the projected end date of the first, the patient was considered continually on the medication. If the second prescription was filled >30 days after the projected end date of the first, the patient was considered not taking the medication from day 31 until the start date of the next prescription. If 2 prescriptions for the same drug were filled on the same day, we used the longest estimated number of days supply to determine the end date.

Outcomes
Patients were censored at disenrollment or at the end of follow-up on December 31, 2002. Disenrollment was defined as a gap in membership of 90 days with no evidence of interim medical care. The primary outcomes were death from any cause and hospitalization for HF. Deaths were identified from health plan databases and state mortality files²⁵ through December 2002, which were the most recent data available at the time of the study. HF-related hospitalizations at health plan or nonnetwork facilities were identified from primary discharge diagnoses (ICD-9 codes 398.91, 402.01, 402.11, 402.91, and 428.x) found in hospitalization and billing claims databases. We have previously demonstrated 96% specificity for these codes²⁶ compared with Framingham clinical criteria based on medical records.²⁷

Statistical Approach
Analyses were performed with SAS statistical software version 9.0 (SAS, Inc, Cary, NC). Comparisons of baseline characteristics by the presence or absence of anemia (defined as a baseline hemoglobin value <13 g/dL in men and <12 g/dL in women²⁸) were performed with a t test for continuous variables and the ² test for categorical variables. Crude rates of death and hospitalization (dependent variables) by level of hemoglobin and estimated GFR (independent variables) were calculated from a Poisson regression, with generalized estimating equations to account for repeated measures of the independent variables within individuals. To examine the independent association between hemoglobin level, GFR, and adverse outcomes, we conducted extended Cox regression models with time-dependent covariates. Time-varying hemoglobin level and GFR categories were assigned from the last value carried forward approach. A hemoglobin value of 13.0 to 13.9 g/dL was used as the referent group after examination of crude outcome rates across hemoglobin levels. For recurrent nonfatal hospitalizations, we used a sandwich estimate of the variance-covariance matrix to obtain standard errors to accommodate the clustering of observations on subjects.²⁹ Additional variables selected for the final models included those known to be associated with outcomes and covariates that differed among subjects with or without anemia at baseline at P<0.05. We initially evaluated for the presence of interactions between hemoglobin level and either GFR or LV systolic function status with interaction terms in our multivariable analyses for both outcomes of interest. On the basis of these results, we next performed analyses stratified by baseline estimated GFR (60, 45 to 59, 30 to 44, and <30 mL · min^–1 · 1.73 m^–2) in the entire cohort and in the presence of reduced versus preserved systolic function in those with known LV systolic function status. We further examined whether the time-dependent rate of change in hemoglobin level (g/dL per 6 person-months) provided any incremental information above the absolute hemoglobin level over time.

Independence of Investigators
The authors were responsible for the study design, data collection, data analysis and interpretation, and manuscript preparation. The funding agency reviewed the manuscript before submission. Dr Go had complete control of all data at all times throughout the analysis and manuscript preparation process.

The authors had full access to the data and take full responsibility for its integrity. All authors have read and agree to the manuscript as written.

	Results

Top Abstract Introduction Methods Results Discussion References

 
Baseline Characteristics and Follow-Up
A total of 59 772 adults with HF were identified, with a mean age of 72 years (range, 20 to 106 years); 46% were women, and 28% were nonwhite (Table 1). At baseline, 42.6% of subjects were anemic by World Health Organization criteria. Patients with anemia at entry were more likely than those without anemia to be slightly older and male; to be a member of a minority group; to have a history of myocardial infarction, stroke, peripheral arterial disease, diabetes, hypertension, malignancy, unknown systolic function status, and reduced GFR; and to be receiving calcium channel blockers (Table 1). Anemic patients were less likely to have known lung disease and atrial fibrillation/flutter and slightly less likely to receive an ACE inhibitor or diuretic compared with nonanemic patients. Of note, patients with an elevated hemoglobin value (17 g/dL) were more likely than those without polycythemia to have diagnosed lung disease (34.8% versus 26.2%, respectively, P<0.001).

View this table: [in this window] [in a new window]   TABLE 1. Baseline Characteristics and Prior Medical Therapy Among 59 772 Adults With Diagnosed Chronic Heart Failure Between January 1, 1996, and December 31, 2002, Overall and Stratified by Presence or Absence of Anemia at Baseline

View this table: [in this window] [in a new window]   TABLE 1. Continued

Hemoglobin Level, Kidney Function, and Outcomes
Overall, 23 901 deaths occurred during 146 454 person-years of follow-up (annual incidence, 20.0%). During a median follow-up of 2.07 years (interquartile range, 0.81 to 3.87), 85.7% of subjects had 1 additional outpatient hemoglobin measurement (mean number of tests, 7.9±10.0; median, 5 [interquartile range, 2 to 10]). The unadjusted rate of death increased with lower levels of hemoglobin, increasing from 11.7 per 100 person-years for a hemoglobin value of 13.0 to 13.9 g/dL to 73.9 per 100 person-years for a hemoglobin level <9.0 g/dL (Figure 1A). The rate of death did not change significantly for those with a hemoglobin value of 13.0 to 16.9 g/dL but was significantly increased for those with a hemoglobin value 17 g/dL. The crude rate of death was also elevated for those with reduced kidney function. The rate of death increased, from 11.0 per 100 person-years for subjects with an estimated GFR of 60 mL · min^–1 · 1.73 m^–2 to a peak of 53.9 per 100 person-years for subjects with an estimated GFR <15 mL · min^–1 · 1.73 m^–2 and not receiving dialysis (Figure 1B).

View larger version (20K): [in this window] [in a new window]   Figure 1. Crude rates of death from any cause by (A) level of hemoglobin (g/dL) and (B) level of GFR (mL · min–1 · 1.73 m–2). Error bars represent 95% confidence limits.

A total of 37 192 hospitalizations for HF occurred among 23 719 cohort members (annual incidence, 25.4%). The crude rate of hospitalization was higher with lower levels of hemoglobin, increasing significantly below a hemoglobin value of 13.0 to 13.9 g/dL (19.6 per 100 person-years) to as high as 69.1 per 100 person-years for a hemoglobin value <9.0 g/dL. The rate of hospitalization was similar for those with hemoglobin levels of 13.0 to 16.9 g/dL but notably higher for those with hemoglobin values 17.0 g/dL (Figure 2A). The crude rate of HF-related hospitalization was also higher for those with poorer kidney function. The rate increased from 19.7 per 100 person-years for those with an estimated GFR 60 mL · min^–1 · 1.73 m^–2 to 69.1 per 100 person-years for those with an estimated GFR <15 mL · min^–1 · 1.73 m^–2 and not receiving dialysis (Figure 2B).

View larger version (21K): [in this window] [in a new window]   Figure 2. Crude rates of hospitalization for HF by (A) level of hemoglobin (g/dL) and (B) level of GFR (mL · min–1 · 1.73 m–2). Error bars represent 95% confidence limits.

Multivariable Analyses for Death and Hospitalization for HF
Hemoglobin Level
After adjustment for known confounders and kidney function, we found that compared with that for a hemoglobin value of 13.0 to 13.9 g/dL, the risk of death stayed relatively constant across hemoglobin levels of 14.0 to 16.9 g/dL, but the adjusted risk increased by 42% for hemoglobin values 17.0 g/dL (Table 2). Below a hemoglobin value of 13.0 to 13.9 g/dL, there was a graded, independent increased risk of death, ranging from 16% for a hemoglobin value of 12.0 to 12.9 g/dL to 248% for a hemoglobin value of <9.0 g/dL (Table 2). In separate analyses, the rate of change in hemoglobin level per 6 person-months was not significantly associated with death, after accounting for the absolute level of hemoglobin and other covariates (data not shown).

View this table: [in this window] [in a new window]   TABLE 2. Multivariable Associations Between Level of Hemoglobin, Level of Estimated Glomerular Filtration Rate, and Risks of Death From Any Cause and Hospitalization for Heart Failure in 59 772 Adults With Chronic Heart Failure

We observed similar patterns between hemoglobin level and hospitalization for HF, although with smaller effect sizes. Compared with that for a hemoglobin value of 13.0 to 13.9 g/dL, there was no significant difference in the adjusted risk of hospitalization for a hemoglobin value of 14.0 to 16.9 g/dL, but the adjusted risk was significantly increased by 14% for a hemoglobin value 17.0 g/dL (Table 2). Below a hemoglobin value of 13.0 to 13.9 g/dL, the adjusted risk of hospitalization increased by 12% for a hemoglobin value of 12.0 to 12.9 g/dL and doubled for a hemoglobin value <9.0 g/dL (Table 2). As with death, the rate of change in hemoglobin level per 6 person-months was not significantly associated with HF hospitalization, after accounting for the absolute level of hemoglobin and other covariates (data not shown).

In multivariable analyses, interaction terms for the hemoglobinxGFR category were significant for death (P<0.001) and hospitalization (P<0.001). In analyses stratified by baseline estimated GFR, compared with a hemoglobin value of 13.0 to 13.9 g/dL, there was no significant difference in the adjusted risk of death for a hemoglobin value 14.0 g/dL at different levels of kidney function. However, the adjusted relative risk of death with hemoglobin levels <13.0 g/dL increased in a graded fashion at all levels of GFR, with higher relative risks noted for lower hemoglobin levels at higher levels of GFR (Table 3). For HF-related hospitalizations, similar relations were observed with hemoglobin level, except that adjusted relative risks were not materially different across different levels of baseline GFR (Table 4).

View this table: [in this window] [in a new window]   TABLE 3. Multivariable Association Between Level of Hemoglobin and Death From Any Cause by Baseline Level of Estimated Glomerular Filtration Rate in 55 167 Adults With Heart Failure and Known Kidney Function

View this table: [in this window] [in a new window]   TABLE 4. Multivariable Association Between Level of Hemoglobin and Hospitalization for Heart Failure by Baseline Level of Estimated Glomerular Filtration Rate in 55 167 Adults With Heart Failure and Known Kidney Function

Kidney Function
Compared with a GFR 60 mL · min^–1 · 1.73 m^–2, there was no increased adjusted risk of death for a GFR of 45 to 59 mL · min^–1 · 1.73 m^–2 (Table 2). However, the adjusted risk of death was 39% higher for a GFR of 30 to 44 mL · min^–1 · 1.73 m^–2 and increased by 226% for a GFR <15 mL · min^–1 · 1.73 m^–2 and by 144% for patients receiving dialysis (Table 2).

After adjustment for potential confounders and hemoglobin level, compared with an estimated GFR of 60 mL · min^–1 · 1.73 m^–2, the risk of hospitalization for HF increased by 11% for a GFR of 45 to 59 mL · min^–1 · 1.73 m^–2 and nearly doubled for a GFR of either 15 to 29 mL · min^–1 · 1.73 m^–2 or <15 mL · min^–1 · 1.73 m^–2 (Table 2). The adjusted risk was only modestly higher for subjects receiving dialysis compared with a GFR 60 mL · min^–1 · 1.73 m^–2.

Analyses Stratified by Systolic Function Status
In the subset of 11 117 subjects with known reduced or preserved systolic function, we observed findings similar to those for the entire cohort. In multivariable analyses, interaction terms for hemoglobinxLV systolic function status were significant for death (P<0.001) and hospitalization (P<0.04). Compared with a hemoglobin value of 13.0 to 13.9 g/dL, there were no significant differences in the adjusted risk of death or hospitalization for HF for a hemoglobin value of 15.0 to 16.9 g/dL among subjects with reduced or preserved LV systolic function, whereas a hemoglobin value of 14.0 to 14.9 g/dL was associated with a lower risk of death in those with reduced but not preserved systolic function (Table 5). The adjusted risks of adverse outcomes associated with a hemoglobin value 17 g/dL were significantly higher only in subjects with preserved systolic function. For a hemoglobin value of <13.0 to 13.9 g/dL, there were consistent, graded, adjusted higher risks of death and hospitalization of similar magnitude in those with reduced or preserved systolic function (Table 5).

View this table: [in this window] [in a new window]   TABLE 5. Multivariable Associations Between Levels of Hemoglobin and Estimated Glomerular Filtration Rate and Risks of Death From Any Cause and Hospitalization for Heart Failure, Stratified by Left Ventricular Systolic Function Status, in 11 117 Adults With Chronic Heart Failure and Known Left Ventricular Systolic Function Status

Compared with a GFR 60 mL · min^–1 · 1.73 m^–2, there was no significantly higher adjusted risk of death or hospitalization for a GFR of 45 to 59 mL · min^–1 · 1.73 m^–2 in subjects with reduced or preserved systolic function (Table 5). However, there were progressively higher adjusted risks as GFR decreased from 30 to 44 mL · min^–1 · 1.73 m^–2 to <15 mL · min^–1 · 1.73 m^–2 in subjects with reduced or preserved systolic function (Table 5). The adjusted risk of death, but not hospitalization, was increased for maintenance dialysis regardless of systolic function status.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
We found that very high (17.0 g/dL) and reduced (<13.0 g/dL) hemoglobin levels as well as reduced kidney function (GFR <60 mL · min^–1 · 1.73 m^–2) were strong predictors of death and hospitalization among adults with HF, even after accounting for known confounders and differential medical therapy. The association with a lower hemoglobin value was a graded one, ranging from 16% and 12% increased adjusted risks of death and hospitalization, respectively, for a hemoglobin value of 12.0 to 12.9 g/dL, to 248% and 99% increased risks, respectively, for a hemoglobin value <9.0 g/dL. Similarly, the risk of adverse outcomes increased with a progressively lower GFR below an apparent threshold of 45 mL · min^–1 · 1.73 m^–2.

Reduced hemoglobin levels and chronic kidney disease are common and interrelated factors in HF. Lower hemoglobin values in HF can be pseudoanemia due to hemodilution³⁰ or actual anemia, which can be caused or worsened by various etiologies, such as kidney dysfunction,⁶ impaired erythropoietin production or resistance, bone marrow depression from increased levels of proinflammatory cytokines,^31,32 iron or vitamin deficiencies,³³ malnutrition,³³ and possibly ACE inhibitors.^5,34 Chronic kidney disease can contribute to the development and exacerbation of HF, and progressive HF contributes to renal hypoperfusion and activation of inflammatory factors, which can lead to the development or worsening of kidney dysfunction.

Low hemoglobin levels, regardless of etiology, have been reported to be associated with lower exercise tolerance³⁵ and, in most but not all studies, an increased risk of adverse outcomes in HF.^4–10 Elevated serum creatinine levels^10,36,37 and lower creatinine clearance^38,39 have also been shown to be associated with higher risks of death in HF. Our study extends these previous observations by demonstrating a J-shaped curve for hemoglobin level and the risks of death and hospitalization for HF that is independent of underlying kidney dysfunction, major comorbidity, and longitudinal use of HF therapies. We show that the adjusted risk did not increase until a hemoglobin value 17.0 g/dL, whereas there was a steep, graded increase in risk below a hemoglobin value of 13.0 g/dL. Similarly, independent of its contribution to lower hemoglobin levels, a reduced GFR <45 mL · min^–1 · 1.73 m^–2 was also associated with higher risks in a graded fashion. A recent study reported that a change in hemoglobin level over 12 months was associated with adverse events in a clinical trial HF population.⁴⁰ We found that after adjusting for the absolute hemoglobin level during follow-up, patient characteristics, and longitudinal therapy, the rate of change in hemoglobin did not add incremental prognostic information. In addition, within a large study sample with longitudinal hemoglobin measurements, we observed that a lower hemoglobin level was associated with higher relative risks of adverse events at all levels of baseline estimated GFR, which extends previous reports.⁵ For the outcome of death, the relative risks associated with lower hemoglobin levels were notably higher at more preserved kidney function, but given the much higher absolute rates of death with a lower GFR, the impact of reduced hemoglobin levels was substantial at all levels of kidney function (Table 3 and Figure 1B). These associations were also similar in persons with reduced or preserved systolic function.

Whether reduced hemoglobin and chronic kidney disease are causally related to the excess risks of death and hospitalization in HF remains unclear. Postulated mechanisms include increased peripheral and myocardial tissue hypoxia, enhanced levels of proinflammatory cytokines, erythropoietin resistance, accelerated progression of LV hypertrophy and dilatation, abnormal ventricular remodeling, activation of the sympathetic nervous system and renin-angiotensin-aldosterone axis, and fluid overload.³

Clinical trial evidence is also limited for whether treatment of these conditions improves outcomes. Whether increasing hemoglobin can improve clinical outcomes remains controversial. For example, a randomized, open-label trial of 1223 persons with HF and/or coronary disease and receiving hemodialysis assigned to a hematocrit of 42% or 30% by the use of intravenous or subcutaneous epoetin alfa plus iron was terminated early after an interim analysis found that the risk of death was increased in the higher hematocrit group (adjusted relative risk, 1.3; 95% confidence interval, 0.9 to 1.8).⁴¹ In a 3-month randomized, placebo-controlled trial of 23 persons with severe HF (New York Heart Association [NYHA] class III/IV) and hematocrit <35%, subjects assigned to receive subcutaneous erythropoietin plus oral iron and folate were more likely to exhibit an increase in peak oxygen consumption and exercise duration and improved disease-specific quality of life scores compared with placebo-treated subjects.⁴² Silverberg and colleagues^2,43–45 have conducted 4 separate open-label interventional studies examining the use of subcutaneous erythropoietin and intravenous iron (target hemoglobin, 12.0 or 12.5 g/dL) in a total of 277 patients with medically refractory advanced HF with reduced systolic function and anemia (hemoglobin at entry, 9.5 to 11.5, 10.0 to 11.5, or <12 g/dL). Receipt of erythropoietin plus intravenous iron was associated with improved LV ejection fraction and NYHA functional class, reduced diuretic doses, and a lower number of subsequent hospitalizations compared with prestudy event rates.^2,43–45 However, these studies, though provocative, were of modest sample sizes, included only medically refractory advanced HF, and had less than optimal control groups. We are unaware of any completed clinical trial that has used clinical end points to evaluate a therapeutic strategy targeted at improving GFR in persons with HF.

The present study was strengthened by inclusion of a large, diverse, and relatively unselected population with HF diagnosed in both ambulatory and hospital settings and across a broad age spectrum. Our sample size allowed a more detailed evaluation of the level of hemoglobin and estimated GFR at which the risk of adverse outcomes increases. We had longitudinal outpatient information on hemoglobin level and kidney function, as well as updated extensive comorbidity status and use of HF therapies. We also comprehensively captured relevant hospitalizations and deaths.

Our study also has several limitations. A substantial proportion of subjects did not have available information on LV systolic function status, and NYHA severity classification and data on biomarkers such as B-type natriuretic peptide⁴⁶ were unavailable. In the subset of persons with known reduced or preserved systolic function, however, results did not differ by systolic function category. We cannot completely rule out residual confounding due to HF severity, although previous studies support the concept that anemia and chronic kidney disease are associated with a poor prognosis regardless of the severity of HF. We had limited details about the possible causes of anemia or renal insufficiency and did not have information on potential mechanisms through which these conditions enhance risk of adverse outcomes, so we cannot determine definitively if they are truly causal or mainly markers of risk. We also relied on automated clinical databases for ascertainment of the presence of comorbidities and use of medications, which may have led to some misclassification. Our study was conducted within an insured population in northern California, but we believe that our findings are generalizable to other healthcare settings and uninsured persons, given the broad spectrum of HF and the diverse nature and large size of the studied population.

In conclusion, anemia, polycythemia, and chronic kidney disease are independent predictors of mortality and excess hospitalizations in patients with HF with or without reduced LV systolic function. Given that low hemoglobin levels are common in HF and modifiable through the use of recombinant erythropoietin and/or iron therapy and because erythropoietin may have other favorable cardiovascular effects,⁴⁷ large, randomized, controlled trials are needed to determine whether a low hemoglobin level is a modifiable risk factor for adverse clinical outcomes in this high-risk population.

	Acknowledgments

 
Sources of Funding

This study was supported by a research grant from Amgen, Inc. Dr Go is also supported by National Institute of Diabetes, Digestive, and Kidney Diseases grants UO1 DK60902 and RO1 DK58411. The funding agency reviewed the manuscript before submission. Dr Go had complete control of all data at all times throughout the analysis and manuscript preparation process.

Disclosures

Drs Go and Ackerson received research support for this study from Amgen, Inc. S. Robbins is an employee of and has ownership interest in Amgen, Inc. Dr Massie has received research support and honoraria from Amgen, Inc, and serves on its advisory board. J. Yang, K.L. Lepper, and Dr Shlipak report no conflicts of interest.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Maggioni AP, Anand I, Gottlieb SO, Latini R, Tognoni G, Cohn JN. Effects of valsartan on morbidity and mortality in patients with heart failure not receiving angiotensin-converting enzyme inhibitors. J Am Coll Cardiol. 2002; 40: 1414–1421.[Abstract/Free Full Text]
   
2. Silverberg DS, Wexler D, Blum M, Keren G, Sheps D, Leibovitch E, Brosh D, Laniado S, Schwartz D, Yachnin T, Shapira I, Gavish D, Baruch R, Koifman B, Kaplan C, Steinbruch S, Iaina A. The use of subcutaneous erythropoietin and intravenous iron for the treatment of the anemia of severe, resistant congestive heart failure improves cardiac and renal function and functional cardiac class, and markedly reduces hospitalizations. J Am Coll Cardiol. 2000; 35: 1737–1744.[Abstract/Free Full Text]
   
3. Mark DB, Felker GM. B-type natriuretic peptide: a biomarker for all seasons? N Engl J Med. 2004; 350: 718–720.[Free Full Text]
   
4. Horwich TB, Fonarow GC, Hamilton MA, MacLellan WR, Borenstein J. Anemia is associated with worse symptoms, greater impairment in functional capacity and a significant increase in mortality in patients with advanced heart failure. J Am Coll Cardiol. 2002; 39: 1780–1786.[Abstract/Free Full Text]
   
5. Al-Ahmad A, Rand WM, Manjunath G, Konstam MA, Salem DN, Levey AS, Sarnak MJ. Reduced kidney function and anemia as risk factors for mortality in patients with left ventricular dysfunction. J Am Coll Cardiol. 2001; 38: 955–962.[Abstract/Free Full Text]
   
6. Ezekowitz JA, McAlister FA, Armstrong PW. Anemia is common in heart failure and is associated with poor outcomes: insights from a cohort of 12 065 patients with new-onset heart failure. Circulation. 2003; 107: 223–225.[Abstract/Free Full Text]
   
7. Anand I, McMurray JJ, Whitmore J, Warren M, Pham A, McCamish MA, Burton PB. Anemia and its relationship to clinical outcome in heart failure. Circulation. 2004; 110: 149–154.[Abstract/Free Full Text]
   
8. Szachniewicz J, Petruk-Kowalczyk J, Majda J, Kaczmarek A, Reczuch K, Kalra PR, Piepoli MF, Anker SD, Banasiak W, Ponikowski P. Anaemia is an independent predictor of poor outcome in patients with chronic heart failure. Int J Cardiol. 2003; 90: 303–308.[CrossRef][Medline] [Order article via Infotrieve]
   
9. Kosiborod M, Smith GL, Radford MJ, Foody JM, Krumholz HM. The prognostic importance of anemia in patients with heart failure. Am J Med. 2003; 114: 112–119.[CrossRef][Medline] [Order article via Infotrieve]
   
10. McClellan WM, Flanders WD, Langston RD, Jurkovitz C, Presley R. Anemia and renal insufficiency are independent risk factors for death among patients with congestive heart failure admitted to community hospitals: a population-based study. J Am Soc Nephrol. 2002; 13: 1928–1936.[Abstract/Free Full Text]
    
11. Kalra PR, Collier T, Cowie MR, Fox KF, Wood DA, Poole-Wilson PA, Coats AJ, Sutton GC. Haemoglobin concentration and prognosis in new cases of heart failure. Lancet. 2003; 362: 211–212.[CrossRef][Medline] [Order article via Infotrieve]
    
12. Kosiborod M, Curtis JP, Wang Y, Smith GL, Masoudi FA, Foody JM, Havranek EP, Krumholz HM. Anemia and outcomes in patients with heart failure: a study from the National Heart Care Project. Arch Intern Med. 2005; 165: 2237–2244.[Abstract/Free Full Text]
    
13. Sharma R, Francis DP, Pitt B, Poole-Wilson PA, Coats AJ, Anker SD. Haemoglobin predicts survival in patients with chronic heart failure: a substudy of the ELITE II trial. Eur Heart J. 2004; 25: 1021–1028.[Abstract/Free Full Text]
    
14. McMurray J, Cohen-Solal A, Dietz R, Eichhorn E, Erhardt L, Hobbs FD, Krum H, Maggioni A, McKelvie RS, Pina IL, Soler-Soler J, Swedberg K. Practical recommendations for the use of ACE inhibitors, ß-blockers, aldosterone antagonists and angiotensin receptor blockers in heart failure: putting guidelines into practice. Eur J Heart Fail. 2005; 7: 710–721.[CrossRef][Medline] [Order article via Infotrieve]
    
15. Mozaffarian D, Nye R, Levy WC. Anemia predicts mortality in severe heart failure: the prospective randomized amlodipine survival evaluation (PRAISE). J Am Coll Cardiol. 2003; 41: 1933–1939.[Abstract/Free Full Text]
    
16. Krieger N. Overcoming the absence of socioeconomic data in medical records: validation and application of a census-based methodology. Am J Public Health. 1992; 82: 703–710.[Abstract/Free Full Text]
    
17. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med. 1999; 130: 461–470.[Abstract/Free Full Text]
    
18. Levey AS, Greene T, Kusek JW, Beck GJ. A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol. 2000; 11: 155A. Abstract.
    
19. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med. 2004; 351: 1296–1305.[Abstract/Free Full Text]
    
20. Rule AD, Larson TS, Bergstralh EJ, Slezak JM, Jacobsen SJ, Cosio FG. Using serum creatinine to estimate glomerular filtration rate: accuracy in good health and in chronic kidney disease. Ann Intern Med. 2004; 141: 929–937.[Abstract/Free Full Text]
    
21. Deleted in proof
    
22. Selby JV, Ray GT, Zhang D, Colby CJ. Excess costs of medical care for patients with diabetes in a managed care population. Diabetes Care. 1997; 20: 1396–1402.[Abstract]
    
23. Fireman BH, Fehrenbacher L, Gruskin EP, Ray GT. Cost of care for patients in cancer clinical trials. J Natl Cancer Inst. 2000; 92: 136–142.[Abstract/Free Full Text]
    
24. Lee WY, Capra AM, Jensvold NG, Gurwitz JH, Go AS. Gender and risk of adverse outcomes in heart failure. Am J Cardiol. 2004; 94: 1147–1152.[CrossRef][Medline] [Order article via Infotrieve]
    
25. Arellano MG, Petersen GR, Petitti DB, Smith RE. The California automated mortality linkage system (CAMLIS). Am J Public Health. 1984; 74: 1324–1330.[Abstract/Free Full Text]
    
26. Ruo B, Capra AM, Jensvold NG, Go AS. Racial variation in the prevalence of atrial fibrillation among patients with heart failure: the Epidemiology, Practice, Outcomes, and Costs of Heart Failure (EPOCH) study. J Am Coll Cardiol. 2004; 43: 429–435.[Abstract/Free Full Text]
    
27. McKee PA, Castelli WP, Mcnamara PM, Kannel WB. The natural history of congestive heart failure: the Framingham Study. N Engl J Med. 1971; 285: 1441–1446.[Medline] [Order article via Infotrieve]
    
28. Nutritional anaemias: report of the WHO scientific group. World Health Organ Tech Rep Ser. 1968; 405: 5–37.[Medline] [Order article via Infotrieve]
    
29. Therneau TM, Grambsch PM. Modeling Survival Data: Extending the Cox Model. New York: Springer; 2000.
    
30. Androne AS, Katz SD, Lund L, LaManca J, Hudaihed A, Hryniewicz K, Mancini DM. Hemodilution is common in patients with advanced heart failure. Circulation. 2003; 107: 226–229.[Abstract/Free Full Text]
    
31. Iversen PO, Woldbaek PR, Tonnessen T, Christensen G. Decreased hematopoiesis in bone marrow of mice with congestive heart failure. Am J Physiol Regul Integr Comp Physiol. 2002; 282: R166–R172.[Abstract/Free Full Text]
    
32. Weiss G. Pathogenesis and treatment of anaemia of chronic disease. Blood Rev. 2002; 16: 87–96.[CrossRef][Medline] [Order article via Infotrieve]
    
33. Anker SD, Negassa A, Coats AJ, Afzal R, Poole-Wilson PA, Cohn JN, Yusuf S. Prognostic importance of weight loss in chronic heart failure and the effect of treatment with angiotensin-converting-enzyme inhibitors: an observational study. Lancet. 2003; 361: 1077–1083.[CrossRef][Medline] [Order article via Infotrieve]
    
34. Herrlin B, Nyquist O, Sylven C. Induction of a reduction in haemoglobin concentration by enalapril in stable, moderate heart failure: a double blind study. Br Heart J. 1991; 66: 199–205.[Abstract/Free Full Text]
    
35. Kalra PR, Bolger AP, Francis DP, Genth-Zotz S, Sharma R, Ponikowski PP, Poole-Wilson PA, Coats AJ, Anker SD. Effect of anemia on exercise tolerance in chronic heart failure in men. Am J Cardiol. 2003; 91: 888–891.[CrossRef][Medline] [Order article via Infotrieve]
    
36. Hillege HL, Girbes AR, de Kam PJ, Boomsma F, de Zeeuw D, Charlesworth A, Hampton JR, van Veldhuisen DJ. Renal function, neurohormonal activation, and survival in patients with chronic heart failure. Circulation. 2000; 102: 203–210.[Abstract/Free Full Text]
    
37. Dries DL, Exner DV, Domanski MJ, Greenberg B, Stevenson LW. The prognostic implications of renal insufficiency in asymptomatic and symptomatic patients with left ventricular systolic dysfunction. J Am Coll Cardiol. 2000; 35: 681–689.[Abstract/Free Full Text]
    
38. Bibbins-Domingo K, Lin F, Vittinghoff E, Barrett-Connor E, Grady D, Shlipak MG. Renal insufficiency as an independent predictor of mortality among women with heart failure. J Am Coll Cardiol. 2004; 44: 1593–1600.[Abstract/Free Full Text]
    
39. McAlister FA, Ezekowitz J, Tonelli M, Armstrong PW. Renal insufficiency and heart failure: prognostic and therapeutic implications from a prospective cohort study. Circulation. 2004; 109: 1004–1009.[Abstract/Free Full Text]
    
40. Anand IS, Kuskowski MA, Rector TS, Florea VG, Glazer RD, Hester A, Chiang YT, Aknay N, Maggioni AP, Opasich C, Latini R, Cohn JN. Anemia and change in hemoglobin over time related to mortality and morbidity in patients with chronic heart failure: results from Val-HeFT. Circulation. 2005; 112: 1121–1127.[Abstract/Free Full Text]
    
41. Besarab A, Bolton WK, Browne JK, Egrie JC, Nissenson AR, Okamoto DM, Schwab SJ, Goodkin DA. The effects of normal as compared with low hematocrit values in patients with cardiac disease who are receiving hemodialysis and epoetin. N Engl J Med. 1998; 339: 584–590.[Abstract/Free Full Text]
    
42. Mancini DM, Katz SD, Lang CC, LaManca J, Hudaihed A, Androne AS. Effect of erythropoietin on exercise capacity in patients with moderate to severe chronic heart failure. Circulation. 2003; 107: 294–299.[Abstract/Free Full Text]
    
43. Silverberg DS, Wexler D, Sheps D, Blum M, Keren G, Baruch R, Schwartz D, Yachnin T, Steinbruch S, Shapira I, Laniado S, Iaina A. The effect of correction of mild anemia in severe, resistant congestive heart failure using subcutaneous erythropoietin and intravenous iron: a randomized controlled study. J Am Coll Cardiol. 2001; 37: 1775–1780.[Abstract/Free Full Text]
    
44. Silverberg DS, Wexler D, Blum M, Tchebiner JZ, Sheps D, Keren G, Schwartz D, Baruch R, Yachnin T, Shaked M, Schwartz I, Steinbruch S, Iaina A. The effect of correction of anaemia in diabetics and non-diabetics with severe resistant congestive heart failure and chronic renal failure by subcutaneous erythropoietin and intravenous iron. Nephrol Dial Transplant. 2003; 18: 141–146.[Abstract/Free Full Text]
    
45. Silverberg DS, Wexler D, Blum M, Schwartz D, Keren G, Sheps D, Iaina A. Effect of correction of anemia with erythropoietin and intravenous iron in resistant heart failure in octogenarians. Isr Med Assoc J. 2003; 5: 337–339.[Medline] [Order article via Infotrieve]
    
46. Maisel AS, Krishnaswamy P, Nowak RM, McCord J, Hollander JE, Duc P, Omland T, Storrow AB, Abraham WT, Wu AH, Clopton P, Steg PG, Westheim A, Knudsen CW, Perez A, Kazanegra R, Herrmann HC, McCullough PA. Rapid measurement of B-type natriuretic peptide in the emergency diagnosis of heart failure. N Engl J Med. 2002; 347: 161–167.[Abstract/Free Full Text]
    
47. Smith KJ, Bleyer AJ, Little WC, Sane DC. The cardiovascular effects of erythropoietin. Cardiovasc Res. 2003; 59: 538–548.[Abstract/Free Full Text]
    

---

 

CLINICAL PERSPECTIVE

With the expanding population of patients with heart failure (HF), identifying the subset of patients with HF who are at the highest risk for poor outcomes is important, so that clinicians can provide patients more accurate information about their prognosis and treat modifiable risk factors. The roles that anemia (ie, low hemoglobin level) and chronic kidney disease (ie, reduced glomerular filtration rate) together play in affecting clinical outcomes in HF are not well understood. This study examined a large population of >59 000 ambulatory and hospitalized patients with chronic HF and found that the levels of hemoglobin and underlying kidney function independently predicted the risks for death and hospitalization for HF, even after accounting for differences in other patient characteristics and the types of treatment that the patients received. Both high (17 g/dL) and lower (<13.0 g/dL) hemoglobin levels and an estimated glomerular filtration rate <45 mL · min^–1 · 1.73 m^–2 were strong, graded, independent risk factors for adverse outcomes in the setting of chronic HF. These findings can assist clinicians to systematically identify patients who are at higher risk for death and hospitalization. In addition, patients with low hemoglobin levels may benefit from evaluation and treatment of reversible causes of anemia, and those identified with reduced kidney function should avoid nephrotoxic agents, if possible. This study lends further support for randomized trials to determine whether raising hemoglobin levels in patients with HF-complicated anemia improves clinical outcomes.

Related Article:

Issue Highlights
Circulation 2006 113: 2675. [Full Text]

This article has been cited by other articles:

				D. M. Henkel, M. M. Redfield, S. A. Weston, Y. Gerber, and V. L. Roger Death in Heart Failure: A Community Perspective Circ Heart Fail, July 1, 2008; 1(2): 91 - 97. [Abstract] [Full Text] [PDF]

				W.H. W. Tang, W. Tong, A. Jain, G. S. Francis, C. M. Harris, and J. B. Young Evaluation and long-term prognosis of new-onset, transient, and persistent anemia in ambulatory patients with chronic heart failure. J. Am. Coll. Cardiol., February 5, 2008; 51(5): 569 - 576. [Abstract] [Full Text] [PDF]

				W. Husemann, M. Fobker, M. Pohlen, C. Bruch, M. Hausberg, G. Breithardt, and H. Reinecke Impact of haemoglobin concentration and chronic kidney disease in patients with coronary heart disease undergoing percutaneous coronary interventions Nephrol. Dial. Transplant., September 1, 2007; 22(9): 2563 - 2570. [Abstract] [Full Text] [PDF]

				T. C. Dowling Prevalence, etiology, and consequences of anemia and clinical and economic benefits of anemia correction in patients with chronic kidney disease: An overview Am. J. Health Syst. Pharm., July 1, 2007; 64(13_Supplement_8): S3 - S7. [Abstract] [Full Text] [PDF]

				F. Andreotti, G. Coluzzi, and F. Crea Letter by Andreotti et al Regarding Article, "Hemoglobin Level, Chronic Kidney Disease, and the Risks of Death and Hospitalization in Adults With Chronic Heart Failure: The Anemia in Chronic Heart Failure: Outcomes and Resource Utilization (ANCHOR) Study" Circulation, March 6, 2007; 115(9): e313 - e313. [Full Text] [PDF]

				A. S. Go, J. Yang, L. M. Ackerson, K. Lepper, M. G. Shlipak, S. Robbins, and B. M. Massie Response to Letter Regarding Article, "Hemoglobin Level, Chronic Kidney Disease, and the Risks of Death and Hospitalization in Adults With Chronic Heart Failure: The Anemia in Chronic Heart Failure: Outcomes and Resource Utilization (ANCHOR) Study" Circulation, March 6, 2007; 115(9): e314 - e314. [Full Text] [PDF]

				A. S. Go, W. Y. Lee, J. Yang, J. C. Lo, and J. H. Gurwitz Statin therapy and risks for death and hospitalization in chronic heart failure. JAMA, November 1, 2006; 296(17): 2105 - 2111. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 113/23/2713    most recent CIRCULATIONAHA.105.577577v1 Alert me when this article is cited Alert me if a correction is posted Citation Map