Association of Hemoglobin Levels With Clinical Outcomes in Acute Coronary Syndromes

Marc S. Sabatine, David A. Morrow, Robert P. Giugliano, Paul B.J. Burton, Sabina A. Murphy,
Carolyn H. McCabe, C. Michael Gibson and Eugene Braunwald
Circulation. 2005;111:2042-2049; originally published online April 11, 2005;
doi: 10.1161/01.CIR.0000162477.70955.5F
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright 2005 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539

The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circ.ahajournals.org/content/111/16/2042

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located,
click Request Permissions in the middle column of the Web page under Services. Further information about
this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/

Downloaded from http://circ.ahajournals.org/ by guest on August 25, 2014

Coronary Heart Disease

Association of Hemoglobin Levels With Clinical Outcomes
in Acute Coronary Syndromes
Marc S. Sabatine, MD, MPH; David A. Morrow, MD, MPH; Robert P. Giugliano, MD, SM;
Paul B.J. Burton, MBBS, PhD; Sabina A. Murphy, MPH; Carolyn H. McCabe, BS;
C. Michael Gibson, MS, MD; Eugene Braunwald, MD
BackgroundIn the setting of an acute coronary syndrome (ACS), anemia has the potential to worsen myocardial
ischemia; however, data relating anemia to clinical outcomes in ACS remain limited.
Methods and ResultsWe examined the association between baseline hemoglobin values and major adverse cardiovascular events through 30 days in 39 922 patients enrolled in clinical trials of ACS. After adjustment for differences in
baseline characteristics and index hospitalization treatments, a reverse J-shaped relationship between baseline
hemoglobin values and major adverse cardiovascular events was observed. In patients with ST-elevation myocardial
infarction, when those with hemoglobin values between 14 and 15 g/dL were used as the reference, cardiovascular
mortality increased as hemoglobin levels fell below 14 g/dL, with an adjusted OR of 1.21 (95% CI 1.12 to 1.30,
P0.001) for each 1-g/dL decrement in hemoglobin. At the other end of the range of hemoglobin, patients with
hemoglobin values 17 g/dL also had excess mortality (OR 1.79, 95% CI 1.18 to 2.71, P0.007). In patients with
nonST-elevation ACS, with those with hemoglobin 15 to 16 g/dL used as the reference, the likelihood of cardiovascular
death, myocardial infarction, or recurrent ischemia increased as the hemoglobin fell below 11 g/dL, with an adjusted OR
of 1.45 (95% CI 1.33 to 1.58, P0.001) for each 1 g/dL decrement in hemoglobin. Patients with hemoglobin values 16
g/dL also had an increased rate of death or ischemic events (OR 1.31, 95% CI 1.03 to 1.66, P0.027).
ConclusionsAnemia is a powerful and independent predictor of major adverse cardiovascular events in patients across
the spectrum of ACS. (Circulation. 2005;111:2042-2049.)
Key Words: anemia coronary disease hemoglobin myocardial infarction risk factors
nemia has been shown to be present in 15% of patients
presenting with acute myocardial infarction (MI) and in
43% of elderly patients with acute MI.1,2 Anemia has the
potential to worsen the myocardial ischemic insult in acute
MI and other acute coronary syndromes (ACS), both by
decreasing the oxygen content of the blood supplied to the
jeopardized myocardium3 and by increasing myocardial oxygen demand through necessitating a higher cardiac output to
maintain adequate systemic oxygen delivery.4
In animal models, higher hemoglobin concentrations prevent ischemia in the setting of significant coronary artery
stenoses.5,6 In human studies, anemia has been shown to be an
independent risk factor for adverse cardiovascular outcomes
in community cohorts,7 in patients with heart failure,8,9 and in
patients undergoing percutaneous coronary intervention.10 To
this point, few studies have specifically examined anemia in
patients with ACS.2,1114 We therefore examined the association between baseline hemoglobin concentration and a range
of cardiovascular clinical outcomes in a broad cohort of
nearly 40 000 patients across the spectrum of ACS.

Methods
Patient Populations
The study cohort eligible for these analyses consisted of 41 637
patients with ACS from the following 16 Thrombolysis In Myocardial Infarction (TIMI) trials: IIIB,15 4,16 9A,17 9B,18 10A,19 10B,20
11A,21 11B,22 12,23 14,24 16 (OPUS),25 17 (InTIME II),26 18
(TACTICS),27 20 (INTEGRITI),28 23 (ENTIRE),29 and 24 (FASTER),30 the details for which have been published. Patients with active
cancer, significant liver or renal disease (typically a creatinine 2.0
mg/dL), active or recent internal bleeding, known bleeding diathesis,
and other significant comorbidities were excluded from these trials.

Baseline Data and Clinical Outcomes

Baseline characteristics were recorded by the local investigators.
Baseline hemoglobin values were available in 96% of patients
(39 922). For patients with ST-elevation MI (STEMI), the clinical
end points were cardiovascular mortality and congestive heart failure
through 30 days. For patients with nonST-elevation ACS (NSTE
ACS), the clinical end points were cardiovascular mortality, MI, and
recurrent myocardial ischemia through 30 days. Major bleeding was
defined as intracranial bleeding or any clinically overt sign of
hemorrhage that was associated with a fall in hemoglobin of 5
g/dL.31

Received August 6, 2004; revision received December 16, 2004; accepted January 11, 2005.
From the TIMI Study Group (M.S.S., D.A.M., R.P.G., S.A.M., C.H.M., C.M.G., E.B.), Cardiovascular Division, Brigham and Womens Hospital,
Boston, Mass, and Amgen, Inc (P.B.J.B.), Thousand Oaks, Calif.
Correspondence to Marc S. Sabatine, MD, MPH, Cardiovascular Division, Brigham and Womens Hospital, 75 Francis St, Boston, MA 02115. E-mail
msabatine@partners.org
2005 American Heart Association, Inc.
Circulation is available at http://www.circulationaha.org

DOI: 10.1161/01.CIR.0000162477.70955.5F

2042
Downloaded from http://circ.ahajournals.org/
by guest on August 25, 2014

Sabatine et al
TABLE 1.

Hemoglobin Levels and Outcomes in ACS

2043

Baseline Characteristics in STEMI Patients Stratified by Baseline Hemoglobin

Hemoglobin on Admission, g/dL
All Patients
10
10 11
1112
1213
1314
14 15
1516
16 17
17
(n25 419) (n191) (n288) (n962) (n2502) (n5077) (n6946) (n5702) (n2783) (n968)

Baseline characteristics
Age, y

60.2

65.0

68.5

66.6

65.3

62.3

60.0

57.5

55.8

55.5

0.001

Female, %

24.2

54.5

66.0

62.8

51.9

35.8

19.3

10.3

5.5

4.7

0.001

8.3

14.1

13.9

11.2

9.8

8.3

7.2

7.8

8.2

10.9

0.001

Hypertension, %

32.0

39.0

46.5

43.0

37.5

33.9

30.4

28.5

28.9

31.7

0.001

Current smoker, %

45.6

36.2

26.0

29.9

33.6

39.2

45.0

51.9

59.2

61.1

0.001

Diabetic, %

14.7

21.1

23.5

20.7

17.0

15.8

13.2

13.3

13.5

16.1

0.001

Nonwhite, %

Creatinine clearance, mL/min

82.5

69.0

61.8

65.9

70.2

77.7

83.3

88.7

92.0

91.8

0.001

Prior MI, %

15.9

18.4

23.7

20.1

17.0

16.8

16.8

13.9

12.9

15.0

0.001

Prior PCI, %

5.6

3.7

7.6

5.9

5.9

5.8

6.0

5.4

4.9

4.8

0.253

Prior CABG, %

3.0

2.8

3.7

2.9

3.8

3.6

3.2

2.6

2.1

2.7

0.007
0.001

History of CHF, %

2.7

7.8

10.3

5.8

4.5

2.6

2.3

2.0

1.5

1.9

History of CVD, %

1.1

1.7

2.9

1.4

1.6

1.3

1.0

0.9

0.8

1.0

0.005

History of PAD, %

5.3

8.3

8.5

7.7

6.9

5.4

4.4

4.8

5.3

5.0

0.001

21.5

18.7

32.2

27.1

23.4

22.9

21.7

19.6

18.7

17.6

0.001

Aspirin, %

-Blocker, %

15.3

16.8

20.1

19.7

16.6

16.6

15.8

13.5

12.7

12.9

0.001

ACE-I or ARB, %

11.9

17.3

21.9

17.6

16.2

13.8

11.0

9.8

8.6

10.6

0.001

Hypolipidemic, %

9.7

4.7

12.5

12.6

11.9

10.4

10.2

8.7

7.3

6.5

0.001

Index hospitalization, %
Anterior MI location

40.7

36.8

33.3

36.4

38.7

38.8

39.8

41.9

45.8

47.8

0.001

Aspirin

98.6

95.3

99.3

98.4

98.1

98.6

98.7

98.8

98.7

99.5

0.001

-Blocker

78.8

70.0

71.2

70.8

73.4

76.8

79.5

81.3

83.6

81.0

0.001

ACE-I or ARB

49.6

52.1

54.0

51.3

50.3

49.5

48.9

48.3

50.9

54.6

0.008

Hypolipidemic

31.3

16.8

22.8

26.3

28.7

30.0

32.8

33.4

31.5

32.4

0.001

Revascularization

34.8

23.0

25.7

25.9

30.2

32.2

34.8

38.4

39.2

40.7

0.001

PCI indicates percutaneous coronary intervention; CHF, congestive heart failure; CVD, cerebrovascular disease; PAD, peripheral arterial disease; and ARB,
angiotensin receptor blocker.

Statistical Analyses

Results

Patients were initially divided into categories based on 1-g/dL

hemoglobin increments. Univariate associations between hemoglobin categories and other baseline demographic and clinical variables
were evaluated with ANOVA and 2 tests. Unadjusted event rates for
the aforementioned cardiovascular end points were compared across
hemoglobin categories with the 2 test for trend across ordered
groups. To evaluate the independent relationship between hemoglobin and cardiovascular end points at 30 days, multivariable logistic
regression was used. Hemoglobin was coded as a multicategory
predictor in 1-g/dL increments, and the hemoglobin category that
had the lowest event rate was used as the reference group. A
comprehensive search for potential confounders was conducted.
Specifically, candidate variables for which we had data in 80% of
subjects and that either demonstrated association with baseline
hemoglobin levels (at a significance threshold of P0.25) or were of
known clinical importance were included in the final model.32 This
approach yielded the following covariates: age; gender; race; hypertension; diabetes; smoking history; creatinine clearance; prior MI;
prior congestive heart failure; prior percutaneous coronary intervention; prior CABG; cerebrovascular disease; peripheral arterial disease; prior aspirin, -blocker, ACE inhibitor, angiotensin receptor
blocker, or hypolipidemic use; index hospitalization aspirin,
-blocker, ACE inhibitor, angiotensin receptor blocker, or hypolipidemic use; index revascularization; and, for STEMI patients, anterior
location of the index MI.

There were a total of 25 419 patients with STEMI and 14 503

patients with NSTE ACS in whom baseline hemoglobin data
were available. These patients were divided into categories
based on 1-g/dL hemoglobin increments, and their baseline
characteristics are shown in Tables 1 and 2, respectively.
Across the spectrum of ACS, those with lower baseline
hemoglobin levels were more likely to be older and female,
were less likely to be a current smoker, had lower estimated
creatinine clearances, and were less likely to undergo revascularization during their index hospitalization.

Indicators of Severity of Acute Ischemic Insult

In patients with STEMI, the likelihood of hypotension,
tachycardia, and Killip class II to IV was significantly related
to baseline hemoglobin (Figures 1A through 1C, P0.001 for
each), with patients at either end of the hemoglobin spectrum
being more likely to have hemodynamic disarray. Similarly,
among patients with NSTE ACS, the presence of ST deviation was related to the baseline hemoglobin (P0.001), again
with a J-shaped pattern (Figure 2).

Downloaded from http://circ.ahajournals.org/ by guest on August 25, 2014

2044

Circulation

April 26, 2005

TABLE 2.

Baseline Characteristics in NSTE ACS Patients Stratified by Baseline Hemoglobin

Hemoglobin on Admission, g/dL
All Patients
8
8 9
9 10
10 11
1112
1213
1314
14 15
1516
16 17
17
(n14 503) (n137) (n306) (n342) (n343) (n976) (n2331) (n3520) (n3390) (n2130) (n812) (n216)

Baseline characteristics
Age, y

62.0

69.5

64.7

63.3

67.3

66.3

64.8

62.3

60.6

58.7

57.6

57.4

0.001

Female, %

33.0

65.7

46.4

29.5

58.0

67.4

55.3

36.4

21.0

12.0

5.8

6.0

0.001

Nonwhite, %

11.8

3.7

3.6

7.0

18.1

17.9

14.8

11.7

11.2

9.4

9.0

12.0

0.001

Hypertension, %

49.2

40.2

28.8

35.4

60.4

59.5

57.8

51.3

49.0

46.2

49.0

53.7

0.001

Current smoker, %

30.6

17.5

28.1

32.9

18.1

19.7

23.2

29.2

32.9

37.1

45.5

54.0

0.001

Diabetic, %

22.7

19.0

9.8

20.6

42.3

29.7

26.7

23.1

20.4

19.5

18.6

19.0

0.001

Creatinine clearance,
mL/min

87.8

62.8

76.8

78.1

68.6

73.1

80.5

88.4

93.6

97.4

98.0

96.9

0.001

Prior MI, %

35.2

26.3

23.6

28.1

35.0

36.0

33.4

35.4

36.2

37.6

36.3

40.7

0.001

Prior PCI, %

16.6

6.6

9.5

12.6

14.0

17.6

16.4

16.8

17.3

17.5

17.1

14.4

0.001

Prior CABG, %

15.7

14.8

9.2

11.2

15.4

17.7

14.7

15.9

16.4

16.4

16.7

13.8

0.014

History of CHF, %

6.7

3.7

3.3

5.6

15.1

13.0

8.3

6.3

5.5

4.6

5.1

4.4

0.001

History of CVD, %

5.1

8.8

3.9

6.4

9.0

7.3

5.8

5.8

3.8

3.7

3.3

5.1

0.001

8.3

12.4

10.5

10.3

10.7

12.0

8.9

7.8

7.8

6.3

8.4

6.5

0.001

Aspirin, %

62.0

88.3

84.3

80.7

64.6

62.1

60.7

58.5

61.9

61.6

59.4

63.1

0.001

History of PAD, %

-Blocker, %

34.1

46.7

36.6

31.9

34.2

36.6

35.1

34.9

34.2

32.1

29.0

26.5

0.001

ACE-I or ARB, %

19.7

28.5

13.4

16.7

26.8

26.4

23.3

20.6

17.2

16.5

16.6

18.1

0.001

Hypolipidemic, %

22.1

22.6

20.3

17.9

22.7

23.9

22.4

22.2

23.9

21.0

17.6

14.0

0.001

Aspirin

97.2

97.0

98.8

94.8

96.2

97.6

97.2

97.1

97.3

96.8

97.2

0.010

-Blocker

71.2

75.6

75.6

67.4

67.8

70.9

69.5

72.3

71.2

72.6

73.2

0.001

Index hospitalization, %
100
82.7

ACE-I or ARB

31.1

33.3

15.7

21.4

41.0

38.1

34.1

32.3

29.0

28.4

30.6

29.6

0.001

Hypolipidemic

36.8

25.9

25.5

26.5

32.3

34.7

36.1

37.6

39.7

39.1

35.4

27.4

0.001

Revascularization

35.6

25.6

24.2

27.8

25.4

33.2

33.0

36.0

37.0

40.3

39.5

37.5

0.001

Abbreviations as in Table 1.

Unadjusted Event Rates

The unadjusted rates of cardiovascular clinical events through
30 days in patients with STEMI and NSTE ACS categorized
by baseline hemoglobin are shown in Tables 3 and 4,
respectively. In general, adverse clinical event rates were
higher in those with lower hemoglobin values and tended to
decrease with increasing hemoglobin levels. For patients with
STEMI, there were highly statistically significant trends for
cardiovascular death (P0.001), congestive heart failure
(P0.001), and the composite (P0.001). For patients with
NSTE ACS, there were highly significant trends for cardiovascular death (P0.001) and recurrent ischemia (P0.001),
a trend for MI (P0.19), and a highly significant trend for
their composite (P0.001). In both STEMI and NSTE ACS
patients, the rate of adverse events started to increase in those
with very high hemoglobin values (16 to 17 g/dL). The
rates of TIMI major bleeding were low (5%) and were not
associated with hemoglobin levels in either STEMI (P0.49)
or NSTE ACS (P0.19) patients.

Adjusted ORs
Given imbalances in important baseline characteristics, multivariable logistic regression was used to evaluate the independent relationship between hemoglobin and clinical out-

comes after adjustment for a wide range of covariates,

including demographics, cardiac risk factors, prior cardiac
disease, medications and revascularization procedures, location of STEMI, and index medications and revascularization
(c-statistics for models were 0.84 for STEMI and 0.71 for
NSTE ACS). Figure 3 shows a plot of the adjusted OR and
95% CI for 30-day cardiovascular mortality in patients with
STEMI categorized by 1-g/dL hemoglobin increments. A
nonmonotonic, reverse J-shaped relationship is evident. With
patients with hemoglobin values of 14 to 15 g/dL as the
reference, cardiovascular mortality increased as hemoglobin
levels fell below 14 g/dL, with an adjusted OR of 1.21 (95%
CI 1.12 to 1.30, P0.001) for each 1-g/dL decrement in
hemoglobin. When systolic blood pressure, heart rate, and
Killip class were added to the model, the adjusted OR was
essentially unchanged (OR 1.20, 95% CI 1.11 to 1.29).
Similarly, when in-hospital medications were excluded from
the model, the adjusted OR was again essentially unchanged
(OR 1.23, 95% CI 1.11 to 1.29). For all patients with a
baseline hemoglobin 14 g/dL, the adjusted OR of cardiovascular death was 1.35 (95% CI 1.11 to 1.64, P0.003).
Reflecting the J-shaped relationship, patients with very high
hemoglobin values (17 g/dL) also had an increased risk of
dying (adjusted OR 1.79, 95% CI 1.18 to 2.71, P0.007). A

Downloaded from http://circ.ahajournals.org/ by guest on August 25, 2014

Sabatine et al

Hemoglobin Levels and Outcomes in ACS

2045

ischemia through 30 days in patients with NSTE ACS

categorized by 1-g/dL hemoglobin increments. Increased risk
of the composite end point becomes apparent as hemoglobin
values fall below 11 g/dL. With patients with hemoglobin
values of 15 to 16 g/dL as the reference, the risk of
cardiovascular death, MI, or recurrent ischemia increased as
hemoglobin levels fell below 11 g/dL, with an adjusted OR of
1.45 (95% CI 1.33 to 1.58, P0.001) for each 1-g/dL
decrement in hemoglobin. When in-hospital medications
were excluded from the model, the adjusted OR was identical
(OR 1.45, 95% CI 1.33 to 1.58). For all patients with a
baseline hemoglobin 11 g/dL, the adjusted OR of the
composite end point was 2.26 (95% CI 1.83 to 2.79,
P0.001). Directionally consistent associations were seen
between hemoglobin and the individual components of the
composite end point, including cardiovascular death (adjusted
OR 1.35, 95% CI 0.74 to 2.45), MI (adjusted OR 1.63, 95%
CI 1.07 to 2.48), and recurrent ischemia (adjusted OR 2.60,
95% CI 2.08 to 3.26). Again, reflecting the J-shaped relationship, patients with baseline hemoglobin values 16 g/dL
were also at an increased risk for the composite end point
(adjusted OR 1.31, 95% CI 1.03 to 1.66, P0.027).

Transfusion

Figure 1. Prevalence of hypotension (systolic blood pressure

100 mm Hg; A), tachycardia (heart rate 100 bpm; B), and
Killip class 2 to 4 (C) in patients with STEMI stratified by baseline hemoglobin values. SBP indicates systolic blood pressure;
HR, heart rate.

similar relationship was seen between baseline hemoglobin

and the risk of heart failure. Those with a hemoglobin 14
g/dL were at increased risk of developing heart failure
(adjusted OR 1.20, 95% CI 1.05 to 1.38, P0.009), as were
those with a hemoglobin 17 g/dL (adjusted OR 1.67, 95%
CI 1.27 to 2.20, P0.001).
Figure 4 shows a plot of the adjusted OR and 95% CI for
the composite of cardiovascular death, MI, or recurrent

Figure 2. Prevalence of ST deviation in patients with NSTE ACS

stratified by baseline hemoglobin values.

A total of 4.6% of patients with STEMI and 2.7% percent of

patients with NSTE ACS received transfusions with whole
blood or packed red blood cells. Approximately 80% of these
transfusions were in the setting of bleeding. In STEMI
patients, when transfusion, interaction terms between transfusion and hemoglobin, and bleeding were added to the
aforementioned multivariable model, transfusion was associated with a decreased risk of cardiovascular death when the
baseline hemoglobin was 12 g/dL (adjusted OR 0.42, 95%
CI 0.20 to 0.89) but not when hemoglobin was 12 g/dL
(adjusted OR 1.42, 95% CI 0.94 to 2.17). In NSTE ACS,
transfusion appeared to be associated with an increased risk
of the composite end point (adjusted OR 1.54, 95% CI 1.14 to
2.09), regardless of the hemoglobin concentration.

Discussion
In a broad cohort of patients with ACS, we found large,
highly statistically significant, and independent associations
between low hemoglobin concentrations and adverse cardiovascular outcomes. Among patients with STEMI, there was a
progressive increase in cardiovascular mortality and heart
failure as the baseline hemoglobin dropped below 14 g/dL. In
patients with NSTE ACS, an increased odds of cardiovascular
death, MI, or recurrent ischemia became apparent when the
baseline hemoglobin fell below 11 g/dL.
To this point, data linking anemia and adverse outcomes in
ACS have been limited. In one study that examined a
database of discharge abstract information in patients admitted with MI, those identified as anemic on the basis of
International Classification of Diseases, 9th Revision (ICD-9)
coding were not found to have a higher mortality.11 In
contrast, in a large database study of elderly Medicare
beneficiaries with acute MI in which actual hematocrit data
were used, a powerful, albeit unadjusted, relationship between hematocrit on admission and all-cause 30-day mortal-

Downloaded from http://circ.ahajournals.org/ by guest on August 25, 2014

2046

Circulation

April 26, 2005

TABLE 3.

Clinical Outcomes Through 30 Days in STEMI Patients Stratified by Baseline Hemoglobin

Hemoglobin on Admission, g/dL
10
(n191)

10 11
(n288)

1112
(n962)

Cardiovascular death, % (1086 events)

14.7

11.9

Congestive heart failure, % (2041 events)

12.7

12.5

Composite, % (2744 events)

21.6

19.2

End Point

1213
(n2502)

1314
(n5077)

14 15
(n6926)

1516
(n5702)

16 17
(n2783)

17
(n968)

10.1

7.1

4.8

3.3

2.8

2.8

4.2

0.001

12.5

10.8

8.8

6.8

7.1

6.8

8.9

0.001

18.4

15.1

12.2

9.0

8.8

8.7

11.2

0.001

ity was found. Similar to the present study, there was a

dose-response effect, with progressively lower survival rates
with more profound degrees of anemia.2 In a small study of
444 consecutive patients with NSTE ACS admitted to the
coronary unit of a single medical center, those with hemoglobin concentrations below 12.8 g/dL were at significantly
increased risk of death or MI.12 In a study of 936 women
undergoing evaluation for chest pain, hemoglobin was an
independent predictor of adverse cardiovascular outcomes,
with a 20% increased risk for each 1-g/dL decrement in
hemoglobin.33 Among patients undergoing percutaneous coronary intervention for ACS, anemia was associated with an
increased risk of periprocedural MI and major adverse cardiovascular events through 30 days.13,14
After careful adjustment for a broad array of baseline
characteristics, we found a striking dose-response relationship across the spectrum of ACS. Anemia has been shown to
significantly decrease oxygen delivery to myocardium downstream of coronary stenoses.3 Anemia also increases myocardial oxygen demand through necessitating a higher stroke
volume and heart rate to maintain adequate systemic oxygen
delivery.4 The combination of these processes may explain
the pathophysiology underlying the progressively worse outcomes we observed in patients with ACS with lower baseline
hemoglobin concentrations. Interestingly, the thresholds below which patients were at increased risk for major adverse
cardiovascular events differed between STEMI and NSTE
ACS. This may reflect differences in the mechanisms by
which anemia predisposes to adverse cardiovascular events in
the 2 types of ACS. In STEMI, even mildly reduced hemoglobin concentrations at the abrupt onset of the coronary
occlusion may significantly attenuate the ability of collateral
flow from nearby patent vessels to limit the extent of
myocardial necrosis and peri-infarct ischemia. We did not
TABLE 4.

have scintigraphic or core laboratory biochemical measurements of infarct size in the present cohort to test this theory
directly; however, consistent with this hypothesis is our
finding that worsening degrees of anemia were associated
with progressively higher rates of hypotension, tachycardia,
and heart failure. For NSTE ACS, coronary occlusion is
usually subtotal, and the likelihood of death and recurrent
ischemic events over the ensuing days may reflect a delicate
balance between myocardial oxygen supply and demand. In
the setting of aggressive antiischemic pharmacological therapy, a more profound degree of anemia may be necessary to
predispose a patient to recurrent ischemic events.
In both types of ACS, patients with very high baseline
hemoglobin (16 to 17 g/dL) also were at greater risk for
adverse cardiovascular events, a finding supported by observations from other studies.34 The pathophysiological basis for
these observations may be that high hemoglobin concentrations can increase blood viscosity. This, in turn, can increase
coronary vascular resistance and decrease coronary blood
flow,35 predispose to thrombosis,36 and increase myocardial
work.37
Potential limitations of this study should be considered.
The present study population was derived from clinical trials
rather than unselected community cohorts; however, the
inclusion and exclusion criteria differed between these 16
trials, thus potentially strengthening the generalizability of
our findings. Moreover, the use of clinical trials allowed us to
gather data prospectively on important baseline characteristics and clinical outcomes from dedicated case report forms
rather than, for example, unconfirmed ICD-9 coding on
discharge summaries. Furthermore, we used actual hemoglobin values, thereby minimizing misclassification and permitting a quantitative approach to the definition of anemia. The
cause of anemia in patients in the present study was unknown,

Clinical Outcomes Through 30 Days in NSTE ACS Patients Stratified by Baseline Hemoglobin
Hemoglobin on Admission, g/dL
8
(n137)

8 9
(n306)

9 10
(n342)

10 11
(n343)

1112
(n976)

1213
(n2331)

1314
(n3520)

14 15
(n3390)

1516
(n2130)

16 17
(n812)

17
(n216)

Cardiovascular death,
% (248 events)

3.7

2.3

2.3

3.8

2.4

1.8

1.7

1.5

1.3

1.5

0.9

0.001

Myocardial
(re)infarction, %
(591 events)

5.1

2.9

5.3

5.8

4.7

4.0

3.9

4.4

3.2

4.5

3.7

0.191

Recurrent ischemia, %
(1879 events)

37.2

25.8

27.1

15.6

12.8

12.8

11.7

11.7

12.0

13.6

16.7

0.001

Composite, %
(2347 events)

40.2

27.5

29.2

19.2

16.8

15.5

14.8

15.0

14.5

17.0

19.9

0.001

End Point

Downloaded from http://circ.ahajournals.org/ by guest on August 25, 2014

Sabatine et al

Hemoglobin Levels and Outcomes in ACS

2047

Figure 3. Unadjusted and adjusted ORs

and 95% CIs for association between
baseline hemoglobin concentration and
cardiovascular mortality through 30 days
in patients with STEMI. Adjusted for age;
gender; race; hypertension; diabetes;
smoking history; renal disease; prior MI;
prior heart failure; prior percutaneous
coronary intervention; prior CABG; cerebrovascular disease; peripheral arterial
disease; prior aspirin, -blocker, ACE
inhibitor or angiotensin receptor blocker,
and hypolipidemic use; anterior MI; index
hospitalization aspirin, -blocker, ACE
inhibitor or angiotensin receptor blocker,
and hypolipidemic use; and index revascularization. CV indicates cardiovascular;
Hgb, hemoglobin; and n, number of
patients.

although patients with recent bleeding, known bleeding

diathesis, or significant renal or hematologic-oncological
diseases (all important potential confounders) were excluded
from these trials. We did not measure erythropoietin levels in
these patients. In addition to stimulating erythrocyte precursors, erythropoietin has multiple cardiovascular effects that
range from myocardial protection and angiogenesis to activation of platelets and upregulation of plasminogen activator
inhibitor-1.38 Thus, some of our observations may be due to
low or high erythropoietin levels rather than anemia per se.
Lastly, comorbidities associated with hemoglobin and prognosis or differences in treatment influenced by the baseline
hemoglobin had the potential to confound our analyses;
however, we conducted a comprehensive search for potential
confounders and included a broad group of prognostic and
treatment-related variables in our multivariable analyses.
Although incompletely or unmeasured comorbidities could
cause residual confounding, given the breadth of covariates
adjusted for in the present analyses, the impact is likely to be
small.

If the association between hemoglobin levels and adverse

cardiovascular events is, in fact, causal, the present findings
would support the current practice guidelines from the American College of Cardiology/American Heart Association that
suggest screening for and correcting anemia in ACS.39,40 The
guidelines, however, do not specify what hemoglobin level to
target. Studies of transfusion thresholds have provided some
data relevant to patients with coronary disease. In a randomized clinical trial in critical illness, there was no apparent
benefit to a more liberal transfusion strategy (hemoglobin
target 10 to 12 g/dL) versus a more restrictive strategy (target
7 to 9 g/dL)41; however, in an analysis specifically restricted
to those with ischemic heart disease, there was a trend toward
higher mortality in those who were randomized to a restrictive strategy.42 Two recent nonrandomized studies have
yielded conflicting results. In a study of patients with ACS,
transfusion was associated with an increased risk of 30-day
mortality, although this effect was no longer apparent if the
nadir hematocrit was below 25%.43 In contrast, in elderly
patients with acute MI, transfusion appeared to be beneficial

Figure 4. Unadjusted and adjusted ORs

and 95% CIs for association between
baseline hemoglobin concentration and
cardiovascular death, MI, or recurrent
ischemia through 30 days in patients
with NSTE ACS. Adjusted for age; gender; race; hypertension; diabetes; smoking history; renal disease; prior MI; prior
heart failure; prior percutaneous coronary
intervention; prior CABG; cerebrovascular disease; peripheral arterial disease;
prior aspirin, -blocker, ACE inhibitor or
angiotensin receptor blocker, and hypolipidemic use; index hospitalization aspirin, -blocker, ACE inhibitor or angiotensin receptor blocker, and hypolipidemic
use; and index revascularization. CVD
indicates cardiovascular death; RI, recurrent ischemia; Hgb, hemoglobin; and n,
number of patients.

Downloaded from http://circ.ahajournals.org/ by guest on August 25, 2014

2048

Circulation

April 26, 2005

if the hematocrit was 33%.2 The present data show a

reduction in cardiovascular mortality with transfusion in
STEMI patients with a hemoglobin 12 g/dL (approximately
equivalent to a hematocrit 36%) but an increased risk of
cardiovascular death, MI, or recurrent ischemia in patients
with NSTE ACS who were transfused; however, we would
underscore that all nonrandomized comparisons must be
viewed with caution because, despite multivariable analyses,
there can be residual confounding by indication, with physicians more likely to transfuse sicker patients. Only randomized trials can definitely resolve the benefit of transfusions in
ACS.
In conclusion, we have found that in 39 922 patients with
ACS enrolled in clinical trials, anemia was a powerful
predictor of cardiovascular mortality and ischemic events.
The graded relationship between hemoglobin levels and
clinical outcomes persisted after adjustment for a wide array
of baseline prognostic factors and in-hospital treatments.
Given these data, a prospective, randomized clinical trial may
be warranted to determine whether precise targeting of
hemoglobin levels improves outcomes in patients with ACS.

Acknowledgments
Dr Sabatine is supported in part by National Heart, Lung, and Blood
Institute grants R01 HL072879 and R01 HL072872. Statistical
analyses supported by Amgen, Inc.

References
1. Chesebro JH, Knatterud G, Roberts R, Borer J, Cohen LS, Dalen J, Dodge
HT, Francis CK, Hillis D, Ludbrook P. Thrombolysis in Myocardial
Infarction (TIMI) Trial, phase I: a comparison between intravenous tissue
plasminogen activator and intravenous streptokinase: clinical findings
through hospital discharge. Circulation. 1987;76:142154.
2. Wu WC, Rathore SS, Wang Y, Radford MJ, Krumholz HM. Blood
transfusion in elderly patients with acute myocardial infarction. N Engl
J Med. 2001;345:1230 1236.
3. Most AS, Ruocco NA Jr, Gewirtz H. Effect of a reduction in blood
viscosity on maximal myocardial oxygen delivery distal to a moderate
coronary stenosis. Circulation. 1986;74:10851092.
4. Levy PS, Quigley RL, Gould SA. Acute dilutional anemia and critical left
anterior descending coronary artery stenosis impairs end organ oxygen
delivery. J Trauma. 1996;41:416 423.
5. Yoshikawa H, Powell WJ Jr, Bland JH, Lowenstein E. Effect of acute
anemia on experimental myocardial ischemia. Am J Cardiol. 1973;32:
670 678.
6. Levy PS, Kim SJ, Eckel PK, Chavez R, Ismail EF, Gould SA, Ramez
Salem M, Crystal GJ. Limit to cardiac compensation during acute isovolemic hemodilution: influence of coronary stenosis. Am J Physiol. 1993;
265:H340 H349.
7. Sarnak MJ, Tighiouart H, Manjunath G, MacLeod B, Griffith J, Salem D,
Levey AS. Anemia as a risk factor for cardiovascular disease in the
Atherosclerosis Risk in Communities (ARIC) study. J Am Coll Cardiol.
2002;40:2733.
8. 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;38:955962.
9. 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:
223225.
10. McKechnie RS, Smith D, Montoye C, Kline-Rogers E, ODonnell MJ,
DeFranco AC, Meengs WL, McNamara R, McGinnity JG, Patel K, Share
D, Riba A, Khanal S, Moscucci M. Prognostic implication of anemia on
in-hospital outcomes after percutaneous coronary intervention. Circulation. 2004;110:271277.

11. Al Falluji N, Lawrence-Nelson J, Kostis JB, Lacy CR, Ranjan R, Wilson

AC. Effect of anemia on 1-year mortality in patients with acute myocardial infarction. Am Heart J. 2002;144:636 641.
12. Goncalves AG, Ferreira J, Aguiar C, Trabulo M, Silva JA, Seabra-Gomes
R. Prognostic value of baseline hemoglobin in acute coronary syndromes.
Circulation. 2002;106(suppl II):II-402.
13. Lee PC, Kini AS, Ahsan C, Fisher E, Sharma SK. Anemia is an independent predictor of mortality after percutaneous coronary intervention.
J Am Coll Cardiol. 2004;44:541546.
14. Nikolsky E, Aymong ED, Halkin A, Grines CL, Cox DA, Garcia E,
Mehran R, Tcheng JE, Griffin JJ, Guagliumi G, Stuckey T, Turco M,
Cohen DA, Negoita M, Lansky AJ, Stone GW. Impact of anemia in
patients with acute myocardial infarction undergoing primary percutaneous coronary intervention: analysis from the Controlled Abciximab and
Device Investigation to Lower Late Angioplasty Complications
(CADILLAC) Trial. J Am Coll Cardiol. 2004;44:547553.
15. The TIMI IIIB Investigators. Effects of tissue plasminogen activator and
a comparison of early invasive and conservative strategies in unstable
angina and non-Q-wave myocardial infarction: results of the TIMI IIIB
Trial. Circulation. 1994;89:15451556.
16. Cannon CP, McCabe CH, Diver DJ, Herson S, Greene RM, Shah PK,
Sequeira RF, Leya F, Kirshenbaum JM, Magorien RD, Palmeri ST, Davis
V, Gibson CM, Poole WK, Braunwald E, the TIMI 4 Investigators.
Comparison of front-loaded recombinant tissue-type plasminogen activator, anistreplase and combination thrombolytic therapy for acute myocardial infarction: results of the Thrombolysis in Myocardial Infarction
(TIMI) 4 Trial. J Am Coll Cardiol. 1994;24:16021610.
17. Antman EM, for the TIMI 9A Investigators. Hirudin in acute myocardial
infarction: safety report from the Thrombolysis and Thrombin Inhibition
in Myocardial Infarction (TIMI) 9A trial. Circulation. 1994;90:
1624 1630.
18. Antman EM, for the TIMI 9B Investigators. Hirudin in acute myocardial
infarction: thrombolysis and thrombin inhibitors in myocardial infarction
(TIMI) 9B trial. Circulation. 1996;94:911922.
19. Cannon CP, McCabe CH, Gibson CM, Ghali M, Sequeira RF, McKendall
GR, Breed J, Modi NB, Fox NL, Tracy RP, Love TW, Braunwald E.
TNK-tissue plasminogen activator in acute myocardial infarction: results
of the Thrombolysis in Myocardial Infarction (TIMI) 10A dose-ranging
trial. Circulation. 1997;95:351356.
20. Cannon CP, Gibson CM, McCabe CH, Adgey AA, Schweiger MJ,
Sequeira RF, Grollier G, Giugliano RP, Frey M, Mueller HS, Steingart
RM, Weaver WD, Van de Werf F, Braunwald E, for the Thrombolysis in
Myocardial Infarction (TIMI) 10B Investigators. TNK-tissue plasminogen activator compared with front-loaded alteplase in acute myocardial
infarction: results of the TIMI 10B trial. Circulation. 1998;98:28052814.
21. The Thrombolysis in Myocardial Infarction (TIMI) 11A Trial Investigators. Dose-ranging trial of enoxaparin for unstable angina: results of
TIMI 11A. J Am Coll Cardiol. 1997;29:1474 1482.
22. Antman EM, McCabe CH, Gurfinkel EP, Turpie AGG, Bernink PJLM,
Salein D, Bayes de Luna A, Fox K, Lablanche J-M, Radley D,
Premmereur J, Braunwald E, for the TIMI 11B Investigators. Enoxaparin
prevents death and cardiac ischemic events in unstable angina/nonQ-wave myocardial infarction: results of the Thrombolysis In Myocardial
Infarction (TIMI) 11B trial. Circulation. 1999;100:15931601.
23. Cannon CP, McCabe CH, Borzak S, Henry TD, Tischler MD, Mueller
HS, Feldman R, Palmeri ST, Ault K, Hamilton SA, Rothman JM,
Novotny WF, Braunwald E, for the TIMI 12 Investigators. Randomized
trial of an oral platelet glycoprotein IIb/IIIa antagonist, sibrafiban, in
patients after an acute coronary syndrome: results of the TIMI 12 trial.
Circulation. 1998;97:340 349.
24. Antman EM, Giugliano RP, Gibson CM, McCabe CH, Coussement P,
Kleiman NS, Vahanian A, Adgey AAJ, Menown I, Rupprecht H-J, Van
der Wieken R, Ducas J, Scherer J, Anderson K, Van der Werf F,
Braunwald E, for the TIMI 14 Investigators. Abciximab facilitates the
rate and extent of thrombolysis: results of the Thrombolysis in Myocardial Infarction (TIMI) 14 trial. Circulation. 1999;99:2720 2732.
25. Cannon CP, McCabe CH, Wilcox RG, Langer A, Caspi A, Berink P,
Lopez-Sendon J, Toman J, Charlesworth A, Anders RJ, Alexander JC,
Skene A, Braunwald E. Oral glycoprotein IIb/IIIa inhibition with
orbofiban in patients with unstable coronary syndromes (OPUS-TIMI 16)
trial. Circulation. 2000;102:149 156.
26. The InTIME-II Investigators. Intravenous NPA for the Treatment of
Infarcting Myocardium Early: InTIME-II, a double-blind comparison
of single-bolus lanoteplase vs accelerated alteplase for the treatment of

Downloaded from http://circ.ahajournals.org/ by guest on August 25, 2014

Sabatine et al

27.

28.

29.

30.

31.

32.
33.

34.

patients with acute myocardial infarction. Eur Heart J. 2000;21:

20052013.
Cannon CP, Weintraub WS, Demopoulos LA, Vicari R, Frey MJ, Lakkis
N, Neumann FJ, Robertson DH, DeLucca PT, DiBattiste PM, Gibson
CM, Braunwald E, for the Tactics-Thrombolysis in Myocardial Infarction
18 Investigators. Comparison of early invasive and conservative
strategies in patients with unstable coronary syndromes treated with the
glycoprotein IIb/IIIa inhibitor tirofiban. N Engl J Med. 2001;344:
1879 1887.
Giugliano RP, Roe MT, Harrington RA, Gibson CM, Zeymer U, Van de
Werf F, Baran KW, Hobbach HP, Woodlief LH, Hannan KL, Greenberg
S, Miller J, Kitt MM, Strony J, McCabe CH, Braunwald E, Califf RM.
Combination reperfusion therapy with eptifibatide and reduced-dose
tenecteplase for ST-elevation myocardial infarction: results of the Integrilin and Tenecteplase in Acute Myocardial Infarction (INTEGRITI)
phase II angiographic trial. J Am Coll Cardiol. 2003;41:12511260.
Antman EM, Louwerenburg HW, Baars HF, Wesdorp JC, Hamer B,
Bassand JP, Bigonzi F, Pisapia G, Gibson CM, Heidbuchel H, Braunwald
E, Van de Werf F. Enoxaparin as adjunctive antithrombin therapy for
ST-elevation myocardial infarction: results of the ENTIRE-Thrombolysis
in Myocardial Infarction (TIMI) 23 Trial. Circulation. 2002;105:
16421649.
Ohman EM. FASTER-TIMI 24. Presented at the 18th International Symposium on Thrombolysis and Interventional Therapy in Acute Myocardial
Infarction, November 16, 2002, Chicago, Ill.
Bovill EG, Terrin ML, Stump DC, Berke AD, Frederick M, Collen D,
Feit F, Gore JM, Hillis LD, Lambrew CT, et al. Hemorrhagic events
during therapy with recombinant tissue-type plasminogen activator,
heparin, and aspirin for acute myocardial infarction: results of the Thrombolysis in Myocardial Infarction (TIMI), Phase II Trial. Ann Intern Med.
1991;115:256 265.
Hosmer DW, Lemeshow S. Applied Logistic Regression. 2nd ed. New
York, NY: John Wiley & Sons; 2000.
Arant CB, Wessel TR, Olson MB, Bairey Merz CN, Sopko G, Rogers
WJ, Sharaf BL, Reis SE, Smith KM, Johnson BD, Handberg E, Mankad
S, Pepine CJ. Hemoglobin level is an independent predictor for adverse
cardiovascular outcomes in women undergoing evaluation for chest pain:
results from the National Heart, Lung, and Blood Institute Womens
Ischemia Syndrome Evaluation Study. J Am Coll Cardiol. 2004;43:
2009 2014.
Gagnon DR, Zhang TJ, Brand FN, Kannel WB. Hematocrit and the risk
of cardiovascular disease: the Framingham study: a 34-year follow-up.
Am Heart J. 1994;127:674 682.

Hemoglobin Levels and Outcomes in ACS

2049

35. Kershenovich S, Modiano M, Ewy GA. Markedly decreased coronary

blood flow in secondary polycythemia. Am Heart J. 1992;123:521523.
36. Lowe GD, Forbes CD. Blood rheology and thrombosis. Clin Haematol.
1981;10:343367.
37. Dormandy JA, Hoare E, Colley J, Arrowsmith DE, Dormandy TL.
Clinical, haemodynamic, rheological, and biochemical findings in 126
patients with intermittent claudication. BMJ. 1973;4:576 581.
38. Smith KJ, Bleyer AJ, Little WC, Sane DC. The cardiovascular effects of
erythropoietin. Cardiovasc Res. 2003;59:538 548.
39. Ryan TJ, Antman EM, Brooks NH, Califf RM, Hillis LD, Hiratzka LF,
Rapaport E, Riegel B, Russell RO, Smith EE III, Weaver WD, Gibbons
RJ, Alpert JS, Eagle KA, Gardner TJ, Garson A Jr, Gregoratos G, Smith
SC Jr. 1999 update: ACC/AHA guidelines for the management of patients
with acute myocardial infarction: executive summary and recommendations: a report of the American College of Cardiology/American Heart
Association Task Force on Practice Guidelines (Committee on Management of Acute Myocardial Infarction). Circulation. 1999;100:
1016 1030.
40. Braunwald E, Antman EM, Beasley JW, Califf RM, Cheitlin MD,
Hochman JS, Jones RH, Kereiakes D, Kupersmith J, Levin TN, Pepine
CJ, Schaeffer JW, Smith EE III, Steward DE, Theroux P. ACC/AHA
guideline update for the management of patients with unstable angina and
non-ST-segment elevation myocardial infarction: a report of the
American College of Cardiology/American Heart Association Task Force
on Practice Guidelines (Committee on the Management of Patients With
Unstable Angina). 2002. American College of Cardiology Web site.
Available at: http://www.acc.org/clinical/guidelines/unstable/unstable.pdf. Accessed August 6, 2004.
41. Hebert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G,
Tweeddale M, Schweitzer I, Yetsir E, the Transfusion Requirements in
Critical Care Investigators for the Canadian Critical Care Trials Group. A
multicenter, randomized, controlled clinical trial of transfusion
requirements in critical care. N Engl J Med. 1999;340:409 417.
42. Hebert PC, Yetisir E, Martin C, Blajchman MA, Wells G, Marshall J,
Tweeddale M, Pagliarello G, Schweitzer I. Is a low transfusion threshold
safe in critically ill patients with cardiovascular diseases? Crit Care Med.
2001;29:227234.
43. Rao SV, Jollis JG, Harrington RA, Granger CB, Newby LK, Armstrong
PW, Moliterno DJ, Lindblad L, Pieper K, Topol EJ, Stamler JS, Califf
RM. Relationship of blood transfusion and clinical outcomes in patients
with acute coronary syndromes. JAMA. 2004;292:15551562.

Downloaded from http://circ.ahajournals.org/ by guest on August 25, 2014