Journal of the American College of Cardiology Vol. 54, No.

22, 2009
© 2009 by the American College of Cardiology Foundation ISSN 0735-1097/09/$36.00
Published by Elsevier Inc. doi:10.1016/[Link].2009.06.050

CONGENITAL HEART DISEASE

Anemia in Adults With Congenital Heart Disease

Relates to Adverse Outcome
Konstantinos Dimopoulos, MD, MSC, PHD,*† Gerhard-Paul Diller, MD, PHD,*†
Georgios Giannakoulas, MD, PHD,* Ricardo Petraco, MD,* Aikaterini Chamaidi, MD,*
Evaggelia Karaoli, MD,* Michael Mullen, MD,* Lorna Swan, MD, PHD,*
Massimo F. Piepoli, MD, PHD,† Philip A. Poole-Wilson, MD,† Darrel P. Francis, MA,‡
Michael A. Gatzoulis, MD, PHD*
London, United Kingdom

Objectives To assess the relation of anemia in noncyanotic adults with congenital heart disease (ACHD) to functional capac-
ity and mortality.

Background Anemia is common in acquired heart failure and affects prognosis. The presence of anemia and its relation to
outcome in ACHD remain unknown.

Methods Data were collected on consecutive noncyanotic ACHD patients attending our tertiary center between 2001 and
2006 in whom hemoglobin concentration was measured. Anemia was defined as hemoglobin concentration
⬍13 g/dl in males and ⬍12 g/dl in females. Cyanotic patients were excluded to avoid confounding from sec-
ondary erythrocytosis.

Results Overall, 830 noncyanotic ACHD patients (age 36.5 ⫾ 15.0 years, 49.6% male) fulfilled the inclusion criteria. The
prevalence of anemia was 13.1% and was highest in patients with congenitally corrected transposition of great
arteries and Ebstein anomaly of the tricuspid valve. Anemic patients were more likely to be receiving diuretics (p
⬍ 0.0001) and have a lower mean corpuscular volume (p ⫽ 0.0001), with a trend toward a higher New York
Heart Association functional class (p ⫽ 0.06). During a median follow-up of 47 months, 55 patients died. Ane-
mic patients had a 3-fold higher mortality risk compared with nonanemic patients, even after propensity score
adjustment for clinical variables such as systemic ventricular function, renal impairment, and diuretic therapy
(adjusted hazard ratio: 3.00; 95% confidence interval: 1.46 to 6.13).

Conclusions Anemia is not uncommon in ACHD patients attending tertiary services and is associated with a 3-fold increased
risk of death. Screening for anemia should be part of the routine assessment of ACHD patients for risk stratifica-
tion and treatment when correctable causes are identified. (J Am Coll Cardiol 2009;54:2093–100) © 2009 by
the American College of Cardiology Foundation

Anemia is common in acquired heart failure, relates to anemia in this setting is multifactorial; its prevalence in-
functional capacity, and, even when mild, is a strong creases with age, chronic kidney disease, and clinical signs of
prognostic marker for survival (1–9). The pathogenesis of congestive heart failure (10,11). Anemia has also been
reported in mildly symptomatic or even asymptomatic
patients, however, suggesting that its pathogenesis relates to
From the *Adult Congenital Heart Centre and Centre for Pulmonary Hypertension,
Royal Brompton Hospital, London, United Kingdom; †Department of Clinical
other factors than just cardiac dysfunction (12). Impaired
Cardiology, National Heart and Lung Institute, Imperial College School of Medicine, erythropoietic secretion and neurohormonal activation, for
London, United Kingdom; and the ‡International Centre for Cardiocirculatory example, seem to be important additional contributors to
Health, St. Mary’s Hospital, London, United Kingdom. Dr. Dimopoulos has been
supported by the European Society of Cardiology. Dr. Giannakoulas is supported by
the development of anemia in heart failure (13).
the Hellenic Heart Foundation, the DG Education & Culture—LLP Programme— Even though chronic renal dysfunction and neurohor-
Leonardo Da Vinci Mobility, and the Hellenic Cardiological Society. Dr. Karaoli was monal activation are commonly encountered in adults with
supported by the European Union (DaVinci programme). Dr. Francis is supported by
the British Heart Foundation. The Royal Brompton Adult Congenital Heart congenital heart disease (ACHD) (14 –18), limited data on
Programme and the Department of Clinical Cardiology have received support from the prevalence of anemia across the spectrum of ACHD are
the British Heart Foundation and the Clinical Research Committee, Royal Brompton available at present (19). We hypothesized that anemia
Hospital, London.
Manuscript received February 20, 2009; revised manuscript received May 28, 2009, would be relatively common in ACHD and relate to
accepted June 28, 2009. functional class and survival.
2094 Dimopoulos et al. JACC Vol. 54, No. 22, 2009
Anemia in Adults With Congenital Heart Disease November 24, 2009:2093–100

Abbreviations Methods variables: age, sex, functional class, previous palliative or

and Acronyms reparative surgery/sternotomy, systemic ventricular func-
Data were collected on all con- tion, serum creatinine levels, MCV, and medication (diuret-
ACHD ⴝ adults with secutive clinically stable noncya-
congenital heart disease ics, beta-blockers, digoxin, angiotensin-converting enzyme
notic ACHD patients who had inhibitors/angiotensin receptor blockers, warfarin, and as-
ccTGA ⴝ congenitally hemoglobin concentration mea-
corrected transposition of pirin) (18). Propensity scores were used to perform 5:1
great arteries
sured in our tertiary center be- nearest neighbor matching (4 nonanemic to 1 anemic
tween May 2001 and June 2006. patient) within a caliper of 0.05 SDs of the propensity score
MCV ⴝ mean corpuscular
volume Cyanotic patients (including (R version 2.8.1, package “matchit”) (22). Balance was
NYHA ⴝ New York Heart
those with Eisenmenger syn- verified by assessing standardized differences between
Association drome, where by definition, there groups for all variables in the matched cohort. A target of
is right-to-left shunting) were ⬍10% standardized difference for all variables was set and
excluded to avoid bias deriving obtained. Univariable Cox regression was then used to
from the stimulant effect of chronic hypoxia on hemopoie- compare mortality between anemic and nonanemic patients
sis. Anemia was defined as hemoglobin concentration ⬍13 in the matched cohort, thus accounting for baseline differ-
g/dl in males and ⬍12 g/dl in females, according to the ences between groups while avoiding an over-fit of the Cox
World Health Organization criteria (20). Mean corpus- model. As a sensitivity analysis, regression adjustment using
cular volume (MCV) and serum creatinine concentration propensity score quartiles was also performed. Missing
were also recorded, when available, on the same date. values were ⬍10% for all variables (mean 1%, range 0% to
Repeated measurements, other than the first, were ex- 9%). As propensity score matching requires complete data-
cluded from analysis. Measurements made after emer- sets, missing data were imputed using multiple imputation
gency admissions or within 6 months after surgery, (R version 2.8.1, package “Amelia,” which uses a
catheter intervention, and/or major bleeding were also bootstrapping-based expectation-maximization algorithm)
excluded from the analysis. Clinical data were acquired (22). Ten values for each missing cell in the data were imputed
from patient records; specific congenital heart disease to create 10 different “complete” datasets; models were then
diagnoses were previously verified by echocardiography, estimated for each dataset (23). Cox analysis was then per-
cardiovascular magnetic resonance, and/or cardiac cath- formed after propensity score matching and propensity score
eterization. Patients were classified into diagnostic regression adjustment using each of the 10 databases. Esti-
groups according to the underlying cardiac anatomy (21). mates were averaged, and the hazard ratio (HR) of anemic
Patients with more than 1 defect were classified according to versus nonanemic patients was computed. The variance of the
the prevalent lesion from a clinical and/or hemodynamic estimated log(HR) was calculated as the average of the esti-
point of view. Systemic ventricular function from transtho- mated variances from each dataset plus the sample variance in
racic echocardiograms within a year of blood sampling was point estimates across datasets (18). Moreover, Kaplan-Meier
recorded as a 3-level categorical variable: normal, mildly survival curves were plotted, and the log-rank test was used for
impaired, and moderately/severely impaired (14). comparison between groups. All p values were 2-sided, and p
Survival status and time to death were obtained from the ⬍ 0.05 was considered to indicate statistical significance.
National Health Service computer system, linked to the Analyses were performed using R version 2.8.1. (22).
national database held by the Office of National Statistics.
Approval by the local ethical committee of the Royal
Brompton Hospital was obtained. Results
Statistical analysis. Numerical values are presented as Patient population. Overall, 830 noncyanotic ACHD pa-
mean ⫾ SD, categorical variables as percentage of total. tients (mean age 36.5 ⫾ 15.0 years, 49.6% male) (Table 1)
Comparisons between groups were performed using the attending our tertiary center fulfilled inclusion criteria.
Wilcoxon rank sum and the Kruskal-Wallis tests for con- Patients from all major ACHD diagnostic groups were
tinuous variables and the Fisher exact test for categorical included; 72% had undergone previous surgical repair. The
variables. Univariate logistic regression analysis was used to majority of patients (72.2%) were asymptomatic at the time
assess the relation between anemia and clinical and demo- of the study.
graphic parameters, followed by multivariate analysis in- Hemoglobin concentration, hematocrit, and anemia.
cluding all parameters that were significant (2-sided p ⬍ Mean hemoglobin concentration in noncyanotic ACHD
0.05) on univariate analysis, to the multivariate model. Cox patients was 14.1 ⫾ 1.7 g/dl (Table 2); 13.1% of patients
regression analysis was used to assess the relation between were anemic (Table 3). As expected, hemoglobin concen-
anemia and death. To account for differences between tration was higher in men compared with women (14.8 ⫾
anemic and nonanemic patients, propensity score matching 1.6 g/dl vs. 13.4 ⫾ 1.5 g/dl; p ⬍ 0.0001) (Fig. 1). No sex
was used. Propensity scores were computed using logistic difference, however, was found in the prevalence of anemia
regression, with anemia as the dependent variable and (12.1% in men vs. 14.1% in women, p ⫽ 0.41). Mean
baseline demographic and clinical variables as independent hemoglobin concentration was within normal range for all
JACC Vol. 54, No. 22, 2009 Dimopoulos et al. 2095
November 24, 2009:2093–100 Anemia in Adults With Congenital Heart Disease

Demographic
Table 1 and Clinical and
Demographic Characteristics According toAccording
Clinical Characteristics UnderlyingtoDiagnosis
Underlying Diagnosis

Systemic Ventricular
NYHA Functional Class Previous Surgery Dysfunction

n (%) Age (yrs) Male I II III IV Palliation Repair Mild Moderate-Severe

Atrial septal defect 144 (17.3) 45.0 ⫾ 17.5 36.8 72.7 19.7 7.6 0.0 0.0 41.7 2.1 1.4
Ventricular septal defect 44 (5.3) 32.0 ⫾ 11.6 45.5 79.5 13.6 6.8 0.0 6.8 70.5 17.1 2.4
Aortic coarctation 107 (12.9) 33.5 ⫾ 13.9 56.1 85.8 8.5 5.7 0.0 0.0 71.0 2.8 5.7
Valve/outflow tract disease 136 (16.4) 36.6 ⫾ 14.8 56.6 71.9 21.5 6.6 0.0 0.0 83.8 12.5 4.7
Atrioventricular septal defects 45 (5.4) 36.2 ⫾ 14.1 42.2 79.5 17.9 2.6 0.0 6.7 74.4 7.1 0.0
Repaired tetralogy of Fallot 120 (14.5) 36.4 ⫾ 12.8 56.7 65.1 22.9 11.0 0.9 43.3 100.0 12.8 2.6
Ebstein anomaly of the tricuspid 20 (2.4) 43.9 ⫾ 14.2 40.0 64.3 21.4 14.3 0.0 0.0 60.0 5.3 0.0
ccTGA 22 (2.7) 39.8 ⫾ 14.5 54.5 30.0 50.0 10.0 10.0 13.6 22.7 14.3 38.1
Mustard 37 (4.5) 29.9 ⫾ 7.1 51.4 73.3 23.3 3.3 0.0 27.0 100.0 36.1 22.2
Fontan 35 (4.2) 25.4 ⫾ 8.1 45.7 60.0 26.7 6.7 6.7 61.8 100.0 37.9 6.9
Complex 32 (3.9) 29.2 ⫾ 10.7 59.4 61.5 26.9 11.5 0.0 37.5 65.6 3.6 10.7
Other congenital defects 88 (10.6) 36.3 ⫾ 15.5 46.6 75.0 17.9 7.1 0.0 9.1 54.5 8.5 9.8
Total 830 (100) 36.5 ⫾ 15.0 49.6 72.2 19.7 7.4 0.7 13.5 71.9 10.5 5.9

Values are percentages unless otherwise indicated.

ccTGA ⫽ congenitally corrected transposition of great arteries; NYHA ⫽ New York Heart Association.

ACHD diagnostic groups and was lowest in patients with 10-fl decrease; 95% CI: 1.43 to 2.75; p ⬍ 0.0001) and
congenitally corrected transposition of great arteries treatment with diuretics (adjusted OR: 2.86; 95% CI: 1.85
(ccTGA) (13.3 ⫾ 2.1 g/dl). Patients with ccTGA had the to 4.39; p ⬍ 0.0001) were the only multivariate predictors of
highest prevalence of anemia (27.3%), followed by patients anemia. Despite the clear association between MCV levels
with Ebstein anomaly of the tricuspid valve (20.0%) and anemia, only 25.7% of anemic patients were microcytic
(Fig. 2). (MCV ⬍84 fl).
To further investigate the relation between a low MCV,
PREDICTORS OF ANEMIA. Univariate predictors of anemia
were a lower MCV (odds ratio [OR]: 1.89 per 10-fl an indirect marker of bleeding, and anemia, we performed a
decrease; 95% confidence interval [CI]: 1.37 to 2.61; p ⫽ subanalysis in patients on warfarin therapy. Microcytosis
0.0001), treatment with diuretics (OR: 2.66; 95% CI: 1.74 was found in 29.0% of anemic compared with 9.1% of
to 4.03; p ⬍ 0.0001), and treatment with warfarin (OR: nonanemic patients undergoing warfarin therapy. However,
1.59; 95% CI: 1.01 to 2.50; p ⫽ 0.046). In addition, a strong 22 of the 31 (71%) anemic patients on warfarin had no
trend for anemia among patients in New York Heart microcytosis.
Association (NYHA) functional class III or IV (OR: 1.86; Hemoglobin concentration and anemia as a predictor of
95% CI: 0.94 to 3.48; p ⫽ 0.06) and in those with higher outcome. During a median follow-up of 47.3 months, 55
creatinine concentration (OR: 1.50; 95% CI: 0.98 to 2.31; patients died. Overall mortality was 1.7% per year (95% CI:
p ⫽ 0.06) was also observed. MCV (adjusted OR: 1.98 per 1.3% to 2.2%). The highest mortality was observed in the
ccTGA group (10.0% per year; 95% CI: 4.0% to 20.7%),
Diagnosis Values
Laboratory According
Laboratory ValuestoAccording
Specific ACHD
to Specific ACHD followed by the Fontan population (5.3% per year; 95% CI:
Table 2
Diagnosis 2.1% to 11.0%). Median time to death was 48.6 months,
and median age at death was 36.3 years.
Hb MCV Creatinine
(g/dl) (fl) (mg/dl) Anemic patients were at significantly higher risk of death
Atrial septal defects 13.9 ⫾ 1.5 89.6 ⫾ 7.4 0.92 ⫾ 0.33 compared with patients without anemia (5-year mortality of
Ventricular septal defect 14.1 ⫾ 1.5 88.5 ⫾ 5.4 0.93 ⫾ 0.22 17.7% vs. 6.0% in nonanemic patients; log-rank p ⬍ 0.0001)
Aortic coarctation 13.9 ⫾ 1.4 88.1 ⫾ 4.4 0.91 ⫾ 0.19 (Fig. 3). On univariate Cox analysis, anemia was a strong
Valve/outflow tract disease 13.9 ⫾ 1.8 88.6 ⫾ 6.0 0.91 ⫾ 0.22 predictor of mortality (HR: 3.13; 95% CI: 1.77 to 5.55;
14.0 ⫾ 2.0 91.0 ⫾ 5.3 0.90 ⫾ 0.21
Atrioventricular septal defects
p ⬍ 0.0001) (Fig. 4). Other univariate predictors of death
Repaired tetralogy of Fallot 14.5 ⫾ 1.5 88.9 ⫾ 5.7 1.03 ⫾ 0.79
were a higher NYHA functional class; treatment with
Ebstein anomaly of the tricuspid 13.9 ⫾ 1.7 91.3 ⫾ 4.5 0.98 ⫾ 0.28
ccTGA 13.3 ⫾ 2.1 91.9 ⫾ 6.9 1.01 ⫾ 0.47
diuretics, angiotensin-converting enzyme inhibitors or an-
Mustard 14.8 ⫾ 1.7 89.3 ⫾ 4.5 0.93 ⫾ 0.19
giotensin receptor blockers, beta-blockers, digoxin, or war-
Fontan 14.6 ⫾ 2.3 90.2 ⫾ 6.0 1.02 ⫾ 0.53 farin; a higher creatinine concentration; and moderate-
Complex 14.6 ⫾ 2.0 88.1 ⫾ 5.2 0.90 ⫾ 0.19 severe systemic ventricular dysfunction. On multivariable
Other congenital defects 13.8 ⫾ 1.6 88.8 ⫾ 6.0 0.95 ⫾ 0.32 analysis including all univariable predictors, anemia re-
Total 14.1 ⫾ 1.7 89.1 ⫾ 5.9 0.94 ⫾ 0.40 mained a predictor of outcome (adjusted HR: 2.26; 95% CI:
ACHD ⫽ adults with congenital heart disease; ccTGA ⫽ congenitally corrected transposition of
1.12 to 4.52). After propensity-score analysis, anemia re-
great arteries; Hb ⫽ hemoglobin; MCV ⫽ mean corpuscular volume. mained a strong predictor of outcome (HR: 3.00; 95% CI:
2096 Dimopoulos et al. JACC Vol. 54, No. 22, 2009
Anemia in Adults With Congenital Heart Disease November 24, 2009:2093–100

Comparison
Table 3 Between Anemic
Comparison and Anemic
Between Nonanemic
and Patients
Nonanemic Patients

Overall Anemic Nonanemic p Value*

n (%) 830 (100) 109 (13.1) 721 (86.9)
Age (yrs) 36.5 ⫾ 15.0 38.1 ⫾ 15.0 36.3 ⫾ 14.9 0.19
Sex (male, %) 49.6 45.9 50.2 0.41
NYHA functional class (%)
I 72.2 66.3 73.1 0.73
II 19.7 20.8 19.6 0.79
III 7.4 10.9 6.9 0.16
IV 0.7 2.0 0.5 0.27
Diuretics (%) 24.3 42.2 21.6 ⬍0.0001
Warfarin (%) 21.1 28.4 20.0 0.06
ACE-I/ARB (%) 26.9 28.4 26.6 0.73
Beta-blockers (%) 22.8 25.7 22.4 0.46
Aspirin (%) 16.1 14.7 16.3 0.78
Digoxin (%) 11.0 12.8 10.7 0.51
Previous palliation (%) 13.5 7.3 14.4 0.06
Previous repair (%) 71.9 70.6 72.1 0.84
Systemic ventricular dysfunction (%)
Mild 10.5 13.7 10.0 0.33
Moderate-severe 5.9 8.8 5.5 0.27
Mean cellular volume (fl) 89.1 ⫾ 6.0 87.0 ⫾ 9.1 89.4 ⫾ 5.2 0.005
Plasma creatinine (mg/dl) 0.94 ⫾ 0.39 1.02 ⫾ 0.85 0.92 ⫾ 0.27 0.16

*p value between patients with and without anemia.

ACE-I ⫽ angiotensin-converting enzyme inhibitor; ARB ⫽ angiotensin receptor blocker; NYHA ⫽ New York Heart Association.

Figure 1 Hemoglobin Concentration Distribution

Distribution of hemoglobin concentration in adults with congenital heart disease according to sex.
JACC Vol. 54, No. 22, 2009 Dimopoulos et al. 2097
November 24, 2009:2093–100 Anemia in Adults With Congenital Heart Disease

The pathogenesis of anemia in congenital heart disease. The

etiology of anemia in ACHD is likely to be multifactorial
(12). Patients in the ccTGA diagnostic subgroup, who had
the highest prevalence of anemia (27.3%), also had the
highest prevalence of symptoms (70%) (Table 1) and the
highest prevalence of systemic ventricular dysfunction, sug-
gesting a possible role of ventricular dysfunction in patho-
genesis. However, a high prevalence of anemia was also
observed in patients with Ebstein anomaly of the tricuspid
valve, a primarily right-sided lesion with resultant low
systemic cardiac output. MCV and the use of diuretics were
independent predictors of anemia in the noncyanotic pop-
ulation, suggesting that renal impairment, abnormal iron
metabolism, and circulatory congestion contribute indepen-
dently toward the occurrence of anemia in ACHD. Fur-
thermore, anemia in ACHD patients can occur as the result
of acute or chronic blood loss due to abnormal hemostasis,
vascular bleeding (arteriovenous malformations or collateral
vessels), use of anticoagulants and antiplatelets, hemolysis
(prosthetic valves, intracardiac patches, or conduits), inter-
vention, or surgery. The strong relation between anemia and
MCV found in this study is in favor of iron deficiency being
a predisposing factor for anemia. Although we have ex-
cluded patients who had recent catheter interventions or
surgery to account for post-operative anemia (common), we
cannot exclude that anemia is due to inadequate iron intake
(dietary or due to malabsorption), although this, we specu-
Figure 2 Prevalence of Anemia late, is unlikely in the noncyanotic ACHD patients (24).
Despite the strong correlation between MCV, treatment
The prevalence of anemia in various adults with congenital heart disease diag-
nostic groups. ccTGA ⫽ congenitally corrected transposition of great arteries.
with warfarin, and anemia, only a minority of patients from
the subgroup of patients undergoing warfarin therapy were
microcytic, suggesting a multifactorial etiology of anemia in
1.46 to 6.13 by matching and HR: 2.83; 95% CI: 1.57 to this population.
5.10 by regression adjustment). Reduced hemopoiesis is another potential mechanism of
anemia in ACHD. Reduced erythropoietin production is
associated with renal dysfunction, which we have recently
Discussion shown to be common in ACHD (18). However, no significant

Anemia was relatively common in this young population of

noncyanotic ACHD patients. Low MCV and treatment
with diuretics were significant predictors of anemia, sug-
gesting a possible role of iron deficiency and of the heart
failure syndrome in its pathogenesis. Anemic ACHD pa-
tients were at a 3-fold risk of death, even after adjustment
for functional class, systemic ventricular function, and other
established risk factors.
The prevalence of anemia in ACHD. The prevalence of
anemia in this ACHD population appears to be lower than
that reported for acquired heart failure, even though re-
ported estimates of the prevalence of anemia in acquired
heart disease vary widely (ranging between 9% and 79%,
according to the characteristics of the population studied)
(1– 4,6 – 8). ACHD patients in our study were much
Figure 3 Unadjusted Mortality Curves
younger than the average cohort of patients with ischemic
heart disease or heart failure and were, in their vast majority Unadjusted Kaplan-Meier mortality curves according to the presence of anemia.
(72%), asymptomatic.
2098 Dimopoulos et al. JACC Vol. 54, No. 22, 2009
Anemia in Adults With Congenital Heart Disease November 24, 2009:2093–100

Figure 4 Univariate Predictors of Mortality

Univariate predictors of death: hazard ratio and 95% confidence interval. ACEi/ARB ⫽ angiotensin-converting
enzyme inhibitor/angiotensin receptor blockers; NYHA ⫽ New York Heart Association.

relation was found between creatinine concentration and ane- there was only a trend toward higher prevalence of anemia in
mia in the present study. “Anemia of chronic disease” is more impaired patients. This may reflect the poor ability of
another plausible cause of anemia in ACHD patients. Acute or subjective assessment of functional status in predicting true
chronic immune activation is the basis of anemia of chronic functional capacity, rather than a lack of relation between
disease, as cytokines and the reticuloendothelial system affect anemia and exercise capacity. It has, in fact, been shown that
iron homeostasis, erythropoietin production, and the life du- subjective measures of functional status such as the NYHA
ration of erythrocytes (10,18,25–28). Immune activation in- functional classification tend to underestimate the degree of
creases hepcidin concentration, which interacts with intestinal impairment in ACHD patients (21,34 –36).
iron transport proteins such as ferroportin, and inhibits iron The prognostic power of anemia could also derive from its
absorption (29,30). Elevated cytokine levels such as tumor established relation to renal dysfunction. Decreased renal
necrosis factor-alpha have been reported in ACHD pa- function and anemia are risk factors for all-cause mortality in
tients, especially in the presence of cyanosis and peripheral patients with left ventricular dysfunction, especially when both
edema (31–33). The strong relation between diuretic use are present, and renal dysfunction is a strong predictor of
and anemia underlines the importance of the heart failure outcome in ACHD (1,2,6,8,11,18,37). However, anemia was
syndrome in the pathophysiology of ACHD (34). a strong predictor of death even after adjustment for systemic
The relation between anemia and outcome of ACHD ventricular function, functional class, and renal impairment.
patients. The prognostic power of anemia may derive from The established relation between neurohormonal and cytokine
its relation to disease severity and exercise capacity (4,6,11). activation and anemia in acquired heart failure may also explain
The oxygen-carrying capacity of the blood is, in fact, an the prognostic power of anemia in ACHD.
essential component of the cardiorespiratory chain responsible Clinical implications. In this population, anemia was a
for providing peripheral tissues with oxygen adequate for their strong and independent prognostic marker. Screening for
metabolic needs. Anemia results in reduced oxygen-carrying anemia should become part of the routine assessment of
capacity and a premature shift to anaerobic metabolism during ACHD patients as a measure of risk stratification. Correctable
exertion. Anemia may also precipitate heart failure and cause causes of anemia, such as bleeding, should always be sought
deterioration in exercise capacity by increasing venous return and treated. Trials of administration of recombinant erythro-
and reducing oxygen delivery to the myocardium. However, poietin in anemic patients with acquired heart failure have been
JACC Vol. 54, No. 22, 2009 Dimopoulos et al. 2099
November 24, 2009:2093–100 Anemia in Adults With Congenital Heart Disease

promising (38 – 40). Erythropoietin and iron administration and Resource Utilization (ANCHOR) Study. Circulation 2006;113:
2713–23.
have resulted in reduction of symptoms, improvement of
3. Sharma R, Francis DP, Pitt B, Poole-Wilson PA, Coats AJS, Anker
exercise tolerance, and reduced need for diuretics. Whether or SD. Haemoglobin predicts survival in patients with chronic heart
not chronic treatment of anemia with iron supplementation failure: a substudy of the ELITE II trial. Eur Heart J 2004;25:1021– 8.
and erythropoietin leads to an improvement in outcome in 4. Anand I, McMurray JJV, Whitmore J, et al. Anemia and its relation-
ship to clinical outcome in heart failure. Circulation 2004;110:149 –54.
ACHD patients needs to be established. 5. Anand IS. Pathogenesis of anemia in cardiorenal disease. Rev Car-
Study limitations. Cyanotic ACHD patients (including diovasc Med 2005;6 Suppl 3:S13–21.
those with Eisenmenger syndrome) were excluded from this 6. Kosiborod M, Smith GL, Radford MJ, Foody JM, Krumholz HM.
The prognostic importance of anemia in patients with heart failure.
study to avoid confounding from secondary erythrocytosis
Am J Med 2003;114:112–9.
and the inherent difficulty of defining anemia in this 7. McClellan WM, Flanders WD, Langston RD, Jurkovitz C, Presley R.
population. Increased hemoglobin levels in cyanotic patients Anemia and renal insufficiency are independent risk factors for death
represent an appropriate physiological adaptation to chron- among patients with congestive heart failure admitted to community
hospitals: a population-based study. J Am Soc Nephrol 2002;13:
ically low oxygen saturations. Iron deficiency “relative ane- 1928 –36.
mia” is frequent in cyanotic patients, but occurs at hemo- 8. Szachniewicz J, Petruk-Kowalczyk J, Majda J, et al. Anaemia is an
globin levels much higher than those of noncyanotic independent predictor of poor outcome in patients with chronic heart
failure. Int J Cardiol 2003;90:303– 8.
individuals, and its diagnosis requires information in addi- 9. Valeur N, Nielsen OW, McMurray JJ, Torp-Pedersen C, Kober L.
tion to mere hemoglobin concentration (e.g., transferrin Anaemia is an independent predictor of mortality in patients with left
saturation and ferritin levels) (24,41– 44). Thus, a uniform ventricular systolic dysfunction following acute myocardial infarction.
definition of anemia across the entire ACHD spectrum is Eur J Heart Fail 2006;8:577– 84.
10. Tang Y-D, Katz SD. Anemia in chronic heart failure: prevalence,
not feasible. etiology, clinical correlates, and treatment options. Circulation 2006;
Although not every single patient seen in our center 113:2454 – 61.
during the study period underwent blood testing, this was 11. Kosiborod M, Curtis JP, Wang Y, et al. Anemia and outcomes in
patients with heart failure: a study from the National Heart Care
more likely due to patient preference rather than physician Project. Arch Intern Med 2005;165:2237– 44.
choice. Our tertiary ACHD center manages patients from 12. O’Meara E, Clayton T, McEntegart MB, et al. Clinical correlates and
England and Wales, and many of them opt to have their consequences of anemia in a broad spectrum of patients with heart
blood tests locally. Thus, a selection bias cannot be ex- failure: results of the Candesartan in Heart Failure: Assessment of
Reduction in Mortality and Morbidity (CHARM) Program. Circula-
cluded. Furthermore, the case mix of ACHD diagnostic tion 2006;113:986 –94.
subgroups in our study, excluding the cyanotic patients, is 13. Kazory A, Ross EA. Anemia: The point of convergence or divergence
very representative of ACHD tertiary work. Future, larger for kidney disease and heart failure? J Am Coll Cardiol 2009;53:
639 – 47.
prospective studies with longer follow-up and including 14. Bolger AP, Sharma R, Li W, et al. Neurohormonal activation and the
patients who are not under tertiary care may provide chronic heart failure syndrome in adults with congenital heart disease.
additional information on anemia mechanisms as well as its Circulation 2002;106:92–9.
prognostic value in ACHD patients in general and specific 15. Ohuchi H, Takasugi H, Ohashi H, et al. Abnormalities of neurohor-
monal and cardiac autonomic nervous activities relate poorly to
ACHD diagnostic subgroups in particular. functional status in Fontan patients. Circulation 2004;110:2601– 8.
16. Davos CH, Davlouros PA, Wensel R, et al. Global impairment of
cardiac autonomic nervous activity late after repair of tetralogy of
Conclusions Fallot. Circulation 2002;106:I69 –75.
17. Davos CH, Francis DP, Leenarts MF, et al. Global impairment of
cardiac autonomic nervous activity late after the Fontan operation.
Anemia is not uncommon in ACHD patients attending
Circulation 2003;108 Suppl 1:II180 –5.
tertiary centers. Anemic ACHD patients have a 3-fold 18. Dimopoulos K, Diller GP, Koltsida E, et al. Prevalence, predictors,
increased mortality risk in the medium term. Screening for and prognostic value of renal dysfunction in adults with congenital
anemia should be part of the routine assessment of ACHD heart disease. Circulation 2008;117:2320 – 8.
19. Collins N, Piran S, Harrison J, Azevedo E, Oechslin E, Silversides
patients for risk stratification, and for treatment, when CK. Prevalence and determinants of anemia in adults with complex
correctable causes are identified. congenital heart disease and ventricular dysfunction (subaortic right
ventricle and single ventricle physiology). Am J Cardiol 2008;102:
625– 8.
Reprint requests and correspondence: Dr. Konstantinos Dimo- 20. Nutritional anaemias. Report of a WHO scientific group. World
poulos, Adult Congenital Heart Centre and Centre for Pulmonary Health Organ Tech Rep Ser 1968;405:5–37.
Hypertension, Royal Brompton Hospital, Sydney Street, London 21. Dimopoulos K, Okonko DO, Diller GP, et al. Abnormal ventilatory
SW3 6NP, United Kingdom. E-mail: k.dimopoulos02@[Link]. response to exercise in adults with congenital heart disease relates to
cyanosis and predicts survival. Circulation 2006;113:2796 – 802.
22. R Development Core Team. R: A language and environment for
REFERENCES statistical computing. Vienna, Austria: R Foundation for Statistical
Computing, 2008.
1. Al-Ahmad A, Rand WM, Manjunath G, et al. Reduced kidney 23. Rubin DB. Multiple Imputation for Nonresponse in Surveys. New
function and anemia as risk factors for mortality in patients with left York, NY: Wiley, 1987.
ventricular dysfunction. J Am Coll Cardiol 2001;38:955– 62. 24. Broberg CS, Bax BE, Okonko DO, et al. Blood viscosity and its
2. Go AS, Yang J, Ackerson LM, et al. Hemoglobin level, chronic kidney relationship to iron deficiency, symptoms, and exercise capacity in
disease, and the risks of death and hospitalization in adults with adults with cyanotic congenital heart disease. J Am Coll Cardiol
chronic heart failure: the Anemia in Chronic Heart Failure: Outcomes 2006;48:356 – 65.
2100 Dimopoulos et al. JACC Vol. 54, No. 22, 2009
Anemia in Adults With Congenital Heart Disease November 24, 2009:2093–100

25. Deswal A, Petersen NJ, Feldman AM, Young JB, White BG, Mann 36. Raphael C, Briscoe C, Davies J, et al. Limitations of the New York
DL. Cytokines and cytokine receptors in advanced heart failure: an Heart Association functional classification system and self-reported
analysis of the cytokine database from the Vesnarinone trial (VEST). walking distances in chronic heart failure. Heart 2007;93:476 – 82.
Circulation 2001;103:2055–9. 37. Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic
26. Macdougall IC, Cooper AC. Erythropoietin resistance: the role of kidney disease and the risks of death, cardiovascular events, and
inflammation and pro-inflammatory cytokines. Nephrol Dial Trans- hospitalization. N Engl J Med 2004;351:1296 –305.
plant 2002;17 Suppl 11:39 – 43. 38. Silverberg DS, Wexler D, Blum M, et al. Effects of treatment with
27. Weiss G, Goodnough LT. Anemia of chronic disease. N Engl J Med epoetin Beta on outcomes in patients with anaemia and chronic heart
2005;352:1011–23. failure. Kidney Blood Press Res 2005;28:41–7.
28. Macdougall IC, Cooper AC. Hyporesponsiveness to erythropoietic 39. Silverberg DS, Wexler D, Blum M, et al. The use of subcutaneous
therapy due to chronic inflammation. Eur J Clin Invest 2005;35 Suppl erythropoietin and intravenous iron for the treatment of the anemia of
3:32–5. severe, resistant congestive heart failure improves cardiac and renal
29. Ganz T. Hepcidin, a key regulator of iron metabolism and mediator of function and functional cardiac class, and markedly reduces hospital-
anemia of inflammation. Blood 2003;102:783– 8. izations. J Am Coll Cardiol 2000;35:1737– 44.
30. Nemeth E, Tuttle MS, Powelson J, et al. Hepcidin regulates cellular 40. Silverberg DS, Wexler D, Sheps D, et al. The effect of correction of
iron efflux by binding to ferroportin and inducing its internalization. mild anemia in severe, resistant congestive heart failure using subcu-
Science 2004;306:2090 –3. taneous erythropoietin and intravenous iron: a randomized controlled
31. Hasper D, Hummel M, Kleber FX, Reindl I, Volk HD. Systemic study. J Am Coll Cardiol 2001;37:1775– 80.
inflammation in patients with heart failure. Eur Heart J 1998;19: 41. Perloff JK, Rosove MH, Child JS, Wright GB. Adults with cyanotic
761–5. congenital heart disease: hematologic management. Ann Intern Med
32. Sharma R, Bolger AP, Li W, et al. Elevated circulating levels of 1988;109:406 –13.
inflammatory cytokines and bacterial endotoxin in adults with con- 42. Rosove MH, Perloff JK, Hocking WG, Child JS, Canobbio MM,
genital heart disease. Am J Cardiol 2003;92:188 –93. Skorton DJ. Chronic hypoxaemia and decompensated erythrocytosis
33. Niebauer J, Volk HD, Kemp M, et al. Endotoxin and immune in cyanotic congenital heart disease. Lancet 1986;2:313–5.
activation in chronic heart failure: a prospective cohort study. Lancet 43. Oechslin E. Hematological management of the cyanotic adult with
1999;353:1838 – 42. congenital heart disease. Int J Cardiol 2004;97 Suppl 1:109 –15.
34. Dimopoulos K, Diller GP, Piepoli MF, Gatzoulis MA. Exercise 44. Spence MS, Balaratnam MS, Gatzoulis MA. Clinical update: cyanotic
intolerance in adults with congenital heart disease. Cardiol Clin adult congenital heart disease. Lancet 2007;370:1530 –2.
2006;24:641– 60.
35. Diller GP, Dimopoulos K, Okonko D, et al. Exercise intolerance in
adult congenital heart disease: comparative severity, correlates, and
prognostic implication. Circulation 2005;112:828 –35. Key Words: congenital heart defects y anemia y prognosis.