|Home | About | Journals | Submit | Contact Us | Français|
Anemia is prevalent in heart-failure patients, and it has been associated with increased mortality rates. In a retrospective study, we evaluated the effects of anemia on long-term survival in patients who experienced purely diastolic heart failure.
Heart-failure patients with preserved systolic function (left ventricular ejection fraction, ≥0.50) were evaluated retrospectively. Of 294 patients, 162 had anemia (group 1) and 132 had no anemia (group 2) upon baseline examination. Anemia was defined as a hemoglobin level below 12 g/dL in women and below 13 g/dL in men. Multivariate Cox proportional hazards regression was conducted in order to test whether hemoglobin levels were an independent predictor of 5-year hospitalization and mortality rates in patients with diastolic heart failure. A P value less than 0.05 was considered statistically significant.
Group 1 patients had a shorter mean survival time (37.8 ± 1.8 vs 44.9 ± 1.8 mo, P = 0.01); however, there was no significant difference between the groups in hospitalization rate (7.2 ± 7.1 vs 7.5 ± 6.3, P = 0.677). In a subgroup analysis, anemia was a significant predictor of higher mortality rates in elderly patients (age, >75 yr) who had diastolic heart failure (P = 0.018).
We found that anemia is associated with increased long-term mortality rates in patients who have diastolic heart failure. In addition, anemia appears to be an independent predictor of worse outcomes in elderly heart-failure patients.
Anemia is defined by the World Health Organization as a hemoglobin (Hb) level below 12 g/dL in women and below 13 g/dL in men.1 A low Hb level is a significant risk factor in heart-failure patients: its association with increased death in this patient population has been shown, and it is an independent predictor of death.2–4 Anemia has also been associated with increased rates of hospital readmission for heart failure.
The optimal Hb level in patients with heart failure is still unclear, although a longitudinal study5 of a large population of patients with chronic heart failure showed that very high levels (≥17 g/dL) or low levels (<13 g/dL) independently predicted death and hospitalization for heart failure.
In most studies in which the effect of anemia in heart failure has been examined, the focus has been on patients who experienced systolic heart failure. In the present study, we sought to evaluate the relationship of Hb levels with short- and long-term survival in patients who experienced diastolic heart failure with preserved systolic function.
This study was approved by our institution's internal review board. Our hospital's admission records from 1999 through 2001 were reviewed for patients who had been admitted with the primary diagnosis of heart failure. Inclusion criteria were echocardiographic evidence of a left ventricular ejection fraction (LVEF) of ≥0.50 upon admission and a diagnosis of heart failure. Patients were excluded if they had severe valvular dysfunction, a history of valve replacement or heart transplantation, primary pulmonary hypertension, a hospital admission for acute myocardial infarction or unstable angina within the previous 6 months, end-stage renal disease that required hemodialysis, or an active malignancy that portended a decreased life expectancy.
Medication use was confirmed from hospital discharge summaries after the 1st admission for heart failure. The presence or absence of hypertension, diabetes mellitus, and coronary artery disease was confirmed by examining the medical records of each patient. Laboratory data, including Hb, had been recorded upon the initial admission and at 5-year follow-up, as available. Anemia was defined as a Hb concentration below 12 mg/dL in women and below 13 mg/dL in men.
The primary endpoint for this study was death after 5 years; data were obtained through the Social Security death registry. Secondary endpoints included 5-year cardiac-related and all-cause annual hospitalization rates, which were determined from hospital medical records.
All statistical analysis was performed with the use of SPSS version 13.0 software (SPSS Inc.; Chicago, Ill). Continuous data were presented as mean ± SD, categorical data were expressed as percentages, and P ≤0.05 was considered statistically significant.
Between-group comparisons of continuous numerical variables, including demographic, clinical, and echocardiographic characteristics, were performed with use of the independent-sample t test. A test of homogeneity of variances was performed for each individual variable with the Levene test and a statistically significant P value (≤0.05). For nonparametric variables, the Wilcoxon rank sum test was used. Within-group comparisons of numerical variables across times were made via the paired t test or the Wilcoxon signed rank test.
All categorical variables were compared between the 2 groups by use of the Fisher exact test. Univariate analyses were performed in order to evaluate the contribution of each relevant variable independently regarding the risk of death, and those that were found to be significant were entered into Cox regression models. Risk ratios were estimated by use of Cox proportional hazards models. Relative risks (RRs) with 95% confidence intervals (CIs) were calculated and reported when appropriate. The Kaplan-Meier method was used to estimate survival, which was stratified according to the anemia classification and 2 age subgroups; the results were displayed in graphics. The log-rank test was used to compare survival across the 2 groups.
The baseline clinical characteristics of the patients are shown in Table I. The group with anemia (group 1) comprised 162 patients, and the group with no anemia, 132 patients (group 2). In comparison with group 1 patients, group 2 patients were more likely to have lower creatinine levels, and they were less likely to have coronary artery disease. A larger proportion of group 2 patients were women (70%). Baseline echocardiographic characteristics produced no significant difference between the groups. Regarding medical therapy at baseline, there were no significant differences between the groups in their use of β-blockers, diuretic agents, calcium-channel blockers, or angiotensin-converting enzyme inhibitors or angiotensin receptor blockers.
The mean follow-up period was 3.3 ± 1.8 years for the 294 patients, during which time the observed total mortality rate was 51% (150 patients). There were no significant differences between the groups in 5-year all-cause hospitalization rates (group 1, 7.2 ± 7.1 vs group 2, 7.5 ± 6.3; P = 0.677) or cardiac-related hospitalization rates (group 1, 3 ± 3.8 vs group 2, 3.6 ± 3.5; P = 0.236). A significant difference was observed in long-term death in group 1 (93 patients; 57.4%) in comparison with group 2 (57 patients, 43.2%) (P = 0.015). Group 1 patients had a shorter mean survival (37.8 ± 1.8 vs 44.9 ± 1.8 mo; P = 0.010). Clinically and statistically significant variables were evaluated in a multivariate model in which Cox regression analysis was used to predict survival (Table II). The only variables that were statistically significant predictors of survival were age (RR, 1.06; 95% CI, 1.039–1.081; P <0.001) and anemia (RR, 1.57; 95% CI, 1.092–2.257; P = 0.015).
Men with Hb levels above 13 g/dL and women with Hb levels above 12 g/dL experienced improved survival (Fig. 1). The survival benefit of Hb levels above 11 g/dL in men and above 10 g/dL in women remained significant (P = 0.001). There was no significant survival benefit when Hb levels were ≥14 g/dL in men in comparison with men who had lower levels, or of Hb levels ≥13 g/dL in women in comparison with women who had lower levels. In a comparison of women with low Hb levels (10–12 g/dL) versus women with high Hb levels (≥13 g/dL), and of men with low Hb levels (11–13 g/dL) versus men with high Hb levels (≥14 g/dL), there was no significant survival benefit (Fig. 2). A subgroup analysis of 192 patients who were older than 75 years of age revealed a survival benefit for patients who did not have anemia (P = 0.018) (Fig. 3).
The current study shows that anemia is an independent predictor of survival in patients who have diastolic heart failure. Anemia—present in 55% of the patients at baseline examination—was a prevalent comorbidity in this study population. The presence of anemia was not associated with a statistically significant increase in either cardiac-related or all-cause hospitalization rates during long-term follow-up. Anemia was associated with higher long-term mortality rates, and this finding remained statistically significant even after factoring for age, sex, other comorbidities, and medications.
Diastolic heart failure, which is defined as clinical signs and symptoms of heart failure with preserved LVEF upon echocardiography, affects 40% to 60% of all patients who experience chronic heart failure.6 The prevalence of diastolic heart failure has been shown to increase with age,7 so it can increase further as the general population ages. Few studies have solely evaluated diastolic heart failure, and currently there are no accepted therapies that have been shown to improve mortality rates, despite the similar mortality rates that exist in patients with systolic heart failure.8–9
Anemia, a known comorbidity in patients with heart failure and reduced LVEF, has been shown to be associated with a higher rate of death in the long term.10–16 Studies have also suggested that anemia independently portends adverse outcomes in patients who experience heart failure with preserved left ventricular (LV) systolic function.17–19 Although heart failure with preserved LVEF is rather prevalent in the elderly patient population (age, >75 yr), little medical literature exists regarding the effect of anemia in elderly patients who have diastolic heart failure. In one of the few pertinent reports, Kerzner and colleagues20 concluded that anemia is not an independent predictor of death in elderly persons with heart failure. The investigators reached this determination on the basis of their retrospective analysis of 204 elderly heart-failure patients during a mean follow-up period of 25 months. Of note, 76 of those elderly patients were reported to have a preserved LVEF (≥0.40). In the present study, however, anemia at baseline appears to have a statistically significant association with death in the long term in elderly patients who have diastolic heart failure (Fig. 3).
Currently, there is a lack of consensus regarding the cause of anemia in patients with heart failure. Multiple interconnected theories involve “pseudoanemia” due to hemodilution, defective iron utilization, renal dysfunction, insufficient erythropoietin production, and anemia of chronic-disease pathophysiology.21,22 The mechanisms that lead to increased death in patients with heart failure and anemia are not well understood and are likely related to multiple confounding variables. Possible contributing factors are deregulation in the proinflammatory cytokine profile, poor nutritional status, and cardiac cachexia.23–27
Adverse LV remodeling may also play a role in the poor outcomes of anemic heart-failure patients. Left ventricular hypertrophy is a rather prevalent finding in patients with diastolic heart failure,28–30 and early animal studies showed LV hypertrophy in anemic rats.31 Although there have been no trials to evaluate changes in LV mass over time in association with anemia in heart-failure patients, other studies have shown that an increase in Hb is associated with a reduction in LV mass.32
Efforts to improve mortality rates are needed, because the general prevalence of heart failure is expected to escalate as the “baby boom” generation enters older age.33 Given the emerging importance of anemia in patients who have diastolic heart failure and the contradictory findings in elderly patients, further research into the causes of anemia and into the benefits of possible therapeutic options will be pertinent. Future studies are warranted in order to evaluate the potential benefit of aggressively managing treatable anemia in patients who experience diastolic heart failure. Potential therapies include iron supplementation for iron-deficiency anemia, transfusions for myelofibrosis, and recombinant erythropoietin therapy for patients who are experiencing refractory anemia and diastolic heart failure.
Limitations of this study include a relatively small sample size from a single center, where the accuracy of data and diagnoses relied chiefly upon information acquired from medical records. Data analysis focused on baseline Hb, without further investigation into possible causes of anemia. Information was not recorded in regard to the possible treatment of patients' anemia during the follow-up period. Although anemia was a statistically significant predictor of death in several multiple-variable analyses, the retrospective nature of the study incurs the possibility of additional factors that were not adjusted for, which could have contributed to long-term death.
Anemia is a prevalent comorbidity in patients who experience heart failure, and anemia is proving to be an independent predictor of increased long-term death regardless of LVEF. At present, there is sparse and contradictory information regarding the possible effects of anemia in elderly patients who have heart failure and preserved systolic function. Given the high prevalence of diastolic heart failure in elderly patients, further research is warranted in order to better evaluate the effects of anemia and possible therapeutic options in this patient population.
Address for reprints: Ernst R. Schwarz, MD, PhD, Cedars-Sinai Medical Center, 8700 Beverly Blvd., Suite 6215, Los Angeles, CA 90048. E-mail: firstname.lastname@example.org
This study was presented as a poster at the 58th annual American College of Cardiology Scientific Session in Orlando, Florida on 31 March 2009. The abstract from that presentation was published in the supplement issue of the Journal of the American College of Cardiology: Tehrani F, Phan A, Morrissey R, Chien C, Rafique A, Schwarz ER. Prognostic value of anemia in elderly patients with diastolic heart failure. J Am Coll Cardiol 2009;53(10):A196.