Download Print - Circulation

Heart Failure
Body Mass Index and Prognosis in Patients With
Chronic Heart Failure
Insights From the Candesartan in Heart failure: Assessment of Reduction
in Mortality and morbidity (CHARM) Program
Satish Kenchaiah, MD, MPH; Stuart J. Pocock, PhD; Duolao Wang, PhD; Peter V. Finn, MD;
Leonardo A.M. Zornoff, MD, PhD; Hicham Skali, MD, ScM; Marc A. Pfeffer, MD, PhD;
Salim Yusuf, MD, DPhil; Karl Swedberg, MD, PhD; Eric L. Michelson, MD;
Christopher B. Granger, MD; John J.V. McMurray, MD; Scott D. Solomon, MD;
for the CHARM Investigators
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
Background—In individuals without known cardiovascular disease, elevated body mass index (BMI) (weight/height2) is
associated with an increased risk of death. However, in patients with certain specific chronic diseases, including heart
failure, low BMI has been associated with increased mortality.
Methods and Results—We examined the influence of BMI on prognosis using Cox proportional hazards models in 7599
patients (mean age, 65 years; 35% women) with symptomatic heart failure (New York Heart Association class II to IV)
and a broad spectrum of left ventricular ejection fractions (mean, 39%) in the Candesartan in Heart failure: Assessment
of Reduction in Mortality and morbidity (CHARM) program. During a median follow-up of 37.7 months, 1831 patients
died. After adjustment for potential confounders, compared with patients with BMI between 30 and 34.9, patients in
lower BMI categories had a graded increase in the risk of death. The hazard ratios (95% confidence intervals) were 1.22
(1.06 to 1.41), 1.46 (1.24 to 1.71), and 1.69 (1.43 to 2.01) among those with BMI of 25 to 29.9, 22.5 to 24.9, and ⬍22.5,
respectively. The increase in risk of death among patients with BMI ⱖ35 was not statistically significant (hazard ratio,
1.17; 95% confidence interval, 0.95 to 1.43). The association between BMI and mortality was not altered by age,
smoking status, or left ventricular ejection fraction (P for interaction ⬎0.20). However, lower BMI was associated with
a greater risk of all-cause death in patients without edema but not in patients with edema (P for interaction ⬍0.001).
Lower BMI was associated with a greater risk of cardiovascular death and noncardiovascular death. Baseline BMI did
not influence the risk of hospitalization for worsening heart failure or due to all causes.
Conclusions—In patients with symptomatic heart failure and either reduced or preserved left ventricular systolic function,
underweight or low BMI was associated with increased mortality, primarily in patients without evidence of fluid
overload (edema). (Circulation. 2007;116:&NA;-.)
Key Words: heart failure 䡲 morbidity 䡲 mortality 䡲 obesity 䡲 prognosis
E
levated body mass index (BMI) is a major risk factor for
the development of heart failure (HF).1– 4 In individuals
free of cardiovascular disease, a curvilinear association has
been described between BMI and the risk of overall mortality,
in which mortality was increased at the lowest and highest
extremes and decreased in the middle.5,6 However, in patients
with established HF, increased BMI has been noted to have
an inverse,7–13 no,14 or nonsignificant15 association with the
risk of mortality. Most of these studies were not conducted in
a contemporary time period in patients receiving modern
evidence-based treatments, and HF patients with a wide range
of left ventricular ejection fractions were not evaluated. Thus,
in patients with chronic HF, the prognostic significance of
BMI on the risk of all-cause death, cause-specific death, and
morbid outcomes remains unclear.
Editorial p ●●●
Clinical Perspective p ●●●
Accordingly, we examined the impact of BMI on prognosis
in patients with symptomatic HF and either reduced or
Received December 12, 2006; accepted May 15, 2007.
From the Division of Cardiovascular Medicine, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Mass
(S.K., P.V.F, L.A.M.Z., H.S., M.A.P, S.D.S); Medical Statistics Unit, London School of Hygiene and Tropical Medicine, London, UK (S.J.P, D.W.);
McMaster University and Hamilton Health Sciences, Hamilton, Ontario, Canada (S.Y.); Sahlgrenska Academy, Göteborg University, Göteborg, Sweden (K.S.);
AstraZeneca LP, Wilmington, Del (E.L.M); Duke University Medical Center, Durham, NC (C.B.G); and University of Glasgow, Glasgow, UK (J.J.V.M).
Correspondence to Satish Kenchaiah, MD, MPH, Brigham and Women’s Hospital, Cardiovascular Division, 15 Francis St, PBB-A-Ground, Boston,
MA 02115. E-mail [email protected]
© 2007 American Heart Association, Inc.
Circulation is available at http://circ.ahajournals.org
DOI: 10.1161/CIRCULATIONAHA.106.679779
1
2
Circulation
August 7, 2007
preserved left ventricular systolic function from the Candesartan in Heart failure: Assessment of Reduction in Mortality
and morbidity (CHARM) program. We hypothesized that in
patients with chronic HF, low BMI (and elevated BMI) would
be associated with an increased risk of death and unfavorable
cardiovascular outcomes.
Methods
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
The design and results of the CHARM program have been published
previously.16,17 Briefly, CHARM was a multinational, double-blind,
placebo-controlled program of clinical trials conducted between
March 1999 and March 2003 that evaluated the efficacy of candesartan in 7599 men and women aged ⱖ18 years who had symptomatic chronic HF (New York Heart Association class II to IV,
regardless of left ventricular ejection fraction) and were either
receiving or not receiving an angiotensin-converting enzyme inhibitor at baseline. Patients with serum creatinine ⱖ265 mmol/L; serum
potassium ⱖ5.5 mmol/L; known bilateral renal artery stenosis;
symptomatic hypotension; critical aortic or mitral stenosis; recent
myocardial infarction, stroke, or open-heart surgery (within 4
weeks); use of an angiotensin-receptor blocker in the previous 2
weeks; any noncardiac disease judged likely to limit 2-year survival;
and those unwilling to provide consent, as well as women of
childbearing potential not using adequate contraception, were excluded. Planned duration of treatment and follow-up ranged from 24
to 48 months.
Covariates and Outcomes
All information on covariates including demographics, height and
weight, cardiovascular risk factors, symptoms and signs of HF, and
concomitant medications was ascertained at the time of enrollment.
Presence or absence of pitting edema in the feet, ankle, leg, or
sacrum was documented. BMI was calculated as weight in kilograms
divided by the square of the height in meters. The individual
outcomes considered in the present investigation were all-cause
death (the primary outcome of the CHARM overall program),
cardiovascular death, noncardiovascular death, hospitalization for
worsening HF, and hospitalization for all causes. The composite
outcomes evaluated were cardiovascular death or hospitalization for
worsening HF (the primary composite outcome of each of the 3
individual CHARM trials), all-cause death or hospitalization for
worsening HF, and all-cause death or hospitalization for all causes.
The integrated end point assessed was days alive out of hospital
during 1 and 2 years of follow-up. All outcomes were determined by
an independent adjudication committee using prespecified definitions and blinded to the baseline BMI status. A death was considered
cardiovascular unless a specific noncardiovascular cause was identified. Hospitalization for worsening HF was defined as an unexpected admission to hospital for treatment of HF or when HF became
a major component of the patient’s hospital admission. A patient
admitted for this reason had to show symptoms and signs of
worsening HF and require treatment with intravenous diuretics.
Evidence of worsening HF had to include at least 1 of the following
items: increasing dyspnea on exertion, orthopnea, nocturnal dyspnea,
pulmonary edema, increasing peripheral edema, increasing fatigue or
decreasing exercise tolerance, renal hypoperfusion (ie, worsening
renal function), raised jugular venous pressure, and radiological
signs of HF.
Statistical Analyses
We evaluated BMI both as a continuous (for every 1-kg/m2 decrease)
and as a categorical variable (⬍22.5, 22.5 to 24.9, 25.0 to 29.9, 30.0
to 34.9 [referent], and ⱖ35.0 kg/m2). The BMI categories were
primarily based on cut points recommended by the World Health
Organization18 and endorsed by the National Institutes of Health.19
We calculated mean⫾SD for continuous variables and proportions
for categorical variables at baseline for each category of BMI. We
constructed linear or logistic regression models for baseline variables
with BMI as a continuous variable (per 1-kg/m2 change) and
presented probability values as an estimate of their association. To
evaluate the association between BMI and the risk of various
outcomes defined as time to first event, we calculated unadjusted,
age- and sex-adjusted, and multivariable-adjusted hazard ratios
(HRs), 95% confidence intervals (CIs), and 2-sided probability
values using Cox proportional hazards regression models. We
performed linear regression analyses to assess the relation between
BMI and the integrated end point of days alive out of hospital during
1 year and 2 years of follow-up. In multivariable models, we adjusted
for all important predictors of mortality and morbidity identified in
a previous investigation in the CHARM program.20 These variables,
in the order of strength of association with all-cause death, were age
(per 10 years over age 60 years), left ventricular ejection fraction
(per 5% decrease below 45%), diabetes mellitus (none [referent],
insulin treated, and oral therapy or diet only), male sex, New York
Heart Association class (II [referent], III, and IV), current smoking,
bundle-branch block, cardiomegaly, previous hospitalization for HF
(none [referent], ⫽ 6 months, ⬎6 months), diastolic blood pressure
(per 10-mm Hg decrease), duration of HF ⬎2 years, previous
myocardial infarction, edema, heart rate (per 10-bpm increase),
pulmonary crackles, pulmonary edema, mitral regurgitation, atrial
fibrillation, rest dyspnea, and study treatment (candesartan versus
placebo). To assess the shape of the association between BMI and
all-cause death, we introduced terms with common transformations
of BMI inclusive of linear, quadratic, linear plus quadratic, logarithmic, squared root, exponential, and categorical (groups as specified
above) in multivariable models.
In multivariable models evaluating BMI as a continuous variable
(per 1-kg/m2 change), we examined for interaction of the effect of
BMI on all-cause mortality and hospitalization for worsening HF by
age (per 1-year change and as a dichotomous variable: ⬍median and
ⱖmedian), smoking status (never/past smokers and current smokers), edema status (without edema and with edema), and left
ventricular ejection fraction (per 5% decrease below 45% and as a
dichotomous variable: ⱕ40% and ⬎40%).
We performed all analyses using SAS software version 9.1.21 We
considered a 2-sided probability value of ⬍0.05 as statistically
significant.
The authors had full access to and take full responsibility for the
integrity of the data. All authors have read and agree to the
manuscript as written.
Results
Baseline Characteristics
The mean (⫾SD) and median BMIs in the CHARM cohort
were 28.3 (⫾5.4) and 27.6, respectively, and ranged from
14.3 to 57.4 kg/m2. Approximately 40% of the patients with
HF had BMI between 25.0 and 30.0 kg/m2, and ⬇11% were
in the lowest (⬍22.5 kg/m2) and highest (ⱖ35 kg/m2) BMI
categories, respectively.
BMI was inversely related to age; patients in the heaviest
category were on average 8.6 years younger than patients in
the leanest category (Table 1). Women were more likely to be
in the leanest or the heaviest BMI category. Increasing BMI
was associated with a higher New York Heart Association
functional class, heart rate, blood pressure, and left ventricular ejection fraction. Greater proportions of individuals in
lower BMI categories were current cigarette smokers and had
previous history of myocardial infarction and bundle-branch
block. Although the duration of HF did not differ across
categories of BMI, a greater proportion of patients with lower
BMI were hospitalized for HF within 6 months before
enrollment. As anticipated, hypertension and diabetes mellitus were noted more commonly in higher BMI categories.
Whereas ischemic heart disease was the predominant cause of
HF in lower BMI categories, hypertension was the primary
Kenchaiah et al
TABLE 1.
BMI and Prognosis in Chronic HF
Baseline Characteristics According to the Category of BMI in the CHARM Study*
BMI, kg/m2
⬍22.5
(n⫽889)
22.5 to 24.9
(n⫽1277)
25 to 29.9
(n⫽3063)
30 to 34.9
(n⫽1579)
ⱖ35
(n⫽791)
P
68.8⫾11.4
67.8⫾10.4
65.8⫾10.7
63.6⫾10.6
60.2⫾11.2
⬍0.0001
41.2
29.4
26.1
33.3
42.1
Demographic characteristics
Age, y
Women, %
0.0008
Ethnic origin, %
European origin
85.5
90.8
92.7
90.4
86.6
...
Black
3.9
2.7
2.8
5.7
10.2
⬍0.0001†
Other
10.6
6.4
4.6
3.9
3.2
⬍0.0001†
II
43.8
47.3
47.1
44.6
35.0
...
III
52.6
50.5
50.4
53.3
61.7
...
Heart disease risk factors
⬍0.0001
NYHA class, %
IV
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
3.6
2.2
2.5
2.2
3.3
...
73.2⫾12.7
72.0⫾13.1
72.4⫾13.0
73.4⫾13.5
74.6⫾12.3
⬍0.0001
Systolic
125.9⫾19.3
129.1⫾19.9
130.9⫾19.0
133.3⫾18.3
134.2⫾18.8
⬍0.0001
Diastolic
73.2⫾10.5
75.0⫾10.7
76.9⫾10.4
78.2⫾10.9
79.0⫾11.0
⬍0.0001
34.9⫾14.6
37.1⫾14.1
38.5⫾14.2
40.8⫾15.3
43.4⫾16.2
⬍0.0001
ⱕ30
38.4
31.6
26.9
23.4
22.6
...
30 to ⬍40
31.0
29.4
30.7
28.6
20.6
...
40 to ⬍50
11.7
17.6
18.4
18.4
17.3
...
ⱖ50
18.9
21.3
24.0
29.6
39.4
...
3.1⫾4.3
3.3⫾4.4
3.2⫾4.5
3.1⫾4.2
3.1⫾4.3
Heart rate, bpm
Blood pressure, mm Hg
Left ventricular ejection fraction, %
Medical history
Duration of HF, y
0.22
Previous hospitalization for HF, %
Within 6 mo
43.2
38.1
36.0
34.5
35.0
0.003
After 6 mo
34.4
33.4
34.3
34.9
37.6
0.19
Current cigarette smoking, %
19.8
15.6
14.2
12.7
12.9
⬍0.0001
Pharmacologically treated hypertension, %
35.3
40.8
45.2
58.5
67.1
⬍0.0001
Diabetes mellitus, %
5.1
6.1
8.1
12.7
17.1
⬍0.0001
13.0
15.0
18.4
22.7
28.6
⬍0.0001
Previous myocardial infarction, %
53.5
56.5
56.4
49.0
38.6
⬍0.0001
Current angina pectoris, %
18.7
23.4
25.3
23.4
25.2
0.064
9.4
10.0
8.6
8.2
7.3
0.034
28.9
28.8
26.9
27.2
25.9
0.072
9.2
7.4
5.6
6.8
7.3
0.57
Insulin treated
Oral therapy or diet only
Stroke, %
Atrial fibrillation, %
Cancer, %
Causes of HF, %‡
Ischemic heart disease
71.3
72.4
73.7
66.7
55.5
⬍0.0001
Hypertension
38.8
43.7
47.8
60.0
68.1
⬍0.0001
Mitral regurgitation
25.9
20.1
17.8
14.6
10.1
⬍0.0001
Dyspnea on walking or climbing
97.0
96.8
96.5
96.1
96.6
0.20
Rest dyspnea
10.7
8.8
9.3
12.9
16.7
⬍0.0001
Orthopnea
20.1
15.8
17.3
21.9
35.4
⬍0.0001
9.1
10.3
12.0
14.0
21.9
⬍0.0001
Symptoms and signs of HF, %
Paroxysmal nocturnal dyspnea
Edema
15.3
15.9
21.4
29.3
50.4
⬍0.0001
Venous congestion
28.2
26.9
31.1
38.7
56.9
⬍0.0001
Jugular venous pressure ⬎6 cm
10.6
8.1
8.9
7.9
11.8
0.46
3
4
Circulation
TABLE 1.
August 7, 2007
Continued
BMI, kg/m2
⬍22.5
(n⫽889)
22.5 to 24.9
(n⫽1277)
25 to 29.9
(n⫽3063)
30 to 34.9
(n⫽1579)
ⱖ35
(n⫽791)
P
S3 gallop
14.7
11.8
12.8
10.3
12.3
0.028
Pulmonary crackles
17.4
16.6
16.6
15.0
15.3
0.050
Pulmonary edema
3.5
2.9
2.9
2.3
2.4
0.017
2.9
1.3
1.1
0.8
0.4
⬍0.0001
25.2
22.5
21.9
21.3
18.7
0.0001
24.5
26.2
24.5
22.4
19.7
⬍0.0001
Digitalis glycoside
53.0
46.0
41.2
39.4
39.3
⬍0.0001
Diuretics
65.2
62.5
64.3
68.1
74.7
⬍0.0001
␤-Blocker
45.7
56.1
55.9
59.7
53.9
0.0008
Bilateral pleural effusions
Cardiomegaly
ECG, %
Bundle-branch block
Concomitant medications, %
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
Calcium channel blocker
13.2
17.1
18.2
25.6
31.0
⬍0.0001
Lipid-lowering drug
31.4
39.1
43.6
46.0
39.7
⬍0.0001
Angiotensin-converting enzyme inhibitor
44.4
42.1
40.1
41.0
40.1
0.20
Spironolactone
18.4
18.2
16.4
15.2
16.8
0.60
49.6
51.0
49.6
49.5
52.1
0.27
Study treatment, %
Candesartan
*P values indicate the association of each variable with BMI and were calculated from linear or logistic regression models with BMI as a continuous variable vs
each covariate.
†Compared with European origin.
‡Categories were not mutually exclusive.
factor attributable to HF in higher BMI categories. Rest
dyspnea, orthopnea, paroxysmal nocturnal dyspnea, and
edema were more likely to be noted in individuals with higher
BMI, and S3 gallop, pulmonary crackles, pulmonary edema,
and cardiomegaly were more prevalent in patients with a
lower BMI. A similar proportion of individuals were on
candesartan therapy in each category of BMI.
BMI and Risk of Death
During a median follow-up of 37.7 months, 1831 patients
(24.1%) died. For BMI categories ⬍35.0 kg/m2, crude cumulative incidence of all-cause death increased progressively
with decreasing categories of BMI (Figure 1, log-rank P of
test for equality ⬍0.0001). In unadjusted analyses, every
1-kg/m2 decrease in BMI was associated with a 5.3% increase
in the risk of all-cause death (Table 2). After adjustment for
age and sex, the increase in the risk of all-cause death was
3.3%, which did not change materially after additional
adjustment for all other important predictors of all-cause
death. In fully adjusted models, compared with patients with
BMI between 30 and 34.9 kg/m2, patients in lower BMI
categories had a graded increase in the risk of all-cause death:
22.3% increase in patients with BMI between 25 and 29.9,
45.7% increase in patients with BMI between 22.5 and 24.9,
and 69.4% increase among patients with BMI ⬍22.5 kg/m2
(Table 2). The 26.4% increase in the risk of all-cause death
among patients with BMI ⱖ35 in age- and sex-adjusted
analyses was attenuated to 16.8% and did not reach statistical
significance in fully adjusted analyses.
In multivariable analyses evaluating the shape of the
association between BMI and all-cause death, linear plus
quadratic BMI terms had the best goodness of fit (␹2
[2]⫽54.9, P⬍0.0001), suggestive of a nonlinear (U- or
J-shaped) association between BMI and all-cause death.
Logarithmic (␹2 [1]⫽41.8, P⬍0.0001) and categorical (␹2
[4]⫽48.5, P⬍0.0001) terms had similar goodness of fit,
followed by square root (␹2 [1]⫽37.5, P⬍0.0001), exponential (␹2 [1]⫽30.6, P⬍0.0001), and quadratic BMI (␹2
[1]⫽24.7, P⬍0.0001).
Patients in lower BMI groups were at a greater risk of
cardiovascular death as well as noncardiovascular death
(Table 2). A trend was evident toward increasing noncardiovascular deaths in the highest BMI category.
BMI and Risk of Hospitalization for
Worsening HF
No linear association existed between BMI and the risk of
hospitalization for worsening HF (Table 2). In age- and
sex-adjusted models, compared with the group with BMI
between 30.0 and 34.9 kg/m2, an 18.3% (borderline statistical
significance) and 26.6% increase existed in risk of hospitalization for worsening HF in the lowest and highest category,
respectively (Table 2). This association was attenuated and no
longer significant, however, in fully adjusted models.
No statistically significant association was present between
BMI and the risk of composite of cardiovascular death or
hospitalization for worsening HF (data not shown). The
association of BMI with the risk of the composite of all-cause
Kenchaiah et al
BMI and Prognosis in Chronic HF
5
BMI Categories
Cumulative Incidence of All-Cause Death (%)
40
<22.5
Log-Rank p-value <0.0001
22.5 to 24.9
30
25 to 29.9
≥35
30 to 34.9
20
10
0
0
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
No. at risk
BMI categories
<22. 5
22.5 to 24.9
25 to 29.9
30 to 34.9
≥35
0.5
1.0
1.5
2.0
2.5
3.0
3.5
597
923
2384
1311
659
448
700
1802
974
448
157
263
647
329
109
4.0
Time (Years)
889
1277
3063
1579
791
828
1210
2934
1554
771
777
1162
2828
1513
747
722
1100
2722
1470
728
672
1037
2605
1424
703
Figure 1. Cumulative incidence curves for all-cause death according to category of BMI at the baseline examination. Vertical bars indicate SEs of the incidence estimates at 3 years of follow-up. Data shown are truncated at 3.5 years.
death or hospitalization for worsening HF was similar to that
with all-cause death alone (Table 2).
BMI and Risk of Hospitalization for All Causes
During follow-up, 4797 (63.1%; 40.8 per 100 person-years of
follow-up) were hospitalized for all causes. Baseline BMI did
not influence the risk of hospitalization for all causes. In fully
adjusted models, compared with patients with BMI of 30 to
34.9, the HRs (95% CIs) were 1.07 (0.96 to 1.19; P⫽0.20),
0.98 (0.89 to 1.08; P⫽0.73), 1.03 (0.89 to 1.05; P⫽0.39), and
1.09 (0.94 to 1.17; P⫽0.36) for patients with BMI ⬍22.5,
22.5 to 24.9, 25 to 29.9, and ⱖ35, respectively. The results
for composite of all-cause death or hospitalization for all
causes were similar to those for hospitalization for all causes
(data not shown).
BMI and Days Alive Out of Hospital
In fully adjusted models, every 1-kg/m2 decrease in BMI was
associated with 0.6 (95% CI, 0.3 to 0.9; P⬍0.0001) and 1.9
(95% CI, 1.1 to 2.6; P⬍0.0001) fewer days alive out of
hospital during 1 and 2 years of follow-up, respectively. In
models evaluating BMI as a categorical variable, however,
evidence existed of a possible inverted J-shaped association
between BMI and days alive out of hospital characterized by
a statistically significant stepwise decrease below and a
decrease (albeit statistically nonsignificant) above BMI of 30
to 34.9 (Figure 2).
Effect Modification
The association between BMI and the risk of all-cause death
and hospitalization for worsening HF did not depend on age,
smoking status, or left ventricular ejection fraction (all P for
interaction ⬎0.20). Lower BMI was associated with higher
risk of mortality in patients with impaired as well as preserved left ventricular ejection fraction (Table 3). However,
lower BMI was associated with a greater risk of all-cause
death in patients without edema but not in patients with
edema (Table 3, P for interaction ⬍0.001). The HRs for each
1-kg/m2 decrease in BMI were 1.05 (95% CI, 1.03 to 1.06;
P⬍0.0001) in patients without edema and 1.00 (95% CI, 0.99
to 1.02; P⫽0.71) in patients with edema. Similar results were
noted for both cardiovascular and noncardiovascular death
(data not shown). Edema status or left ventricular ejection
fraction at baseline did not influence the association between
BMI and the risk of hospitalization for worsening HF (Table
3; P for interaction ⬎0.30).
Discussion
In CHARM patients, lowest mortality was noted in patients
with BMI between 30 and 34.9, a progressive increase in
mortality was evident below this range, and a plateau or
increase was observed above this range (U- or J-shaped
association). By the same token, highest number of days alive
out of hospital during 1 or 2 years of follow-up was observed
in patients with BMI between 30 and 34.9, with a stepwise
decrease in days alive out of hospital below and a modest
decrease above this range (inverted J-shaped association).
Lower BMI was associated with a greater risk of death,
cardiovascular death, and noncardiovascular death (including
cancer deaths). The relation between BMI and the risk of
all-cause death was not altered by age, smoking status, or left
6
Circulation
TABLE 2.
August 7, 2007
Results of Multivariable Cox Proportional Hazards Models Examining the Relation of BMI to the Risk of Various Outcomes
Unadjusted
Age- and Sex-Adjusted
Fully Adjusted*
No. of Events
(%)
Rate/100
Person-years
HR (95% CI)
P
HR (95% CI)
P
HR (95% CI)
P
1831 (25.0)
8.5
1.05 (1.04 to 1.06)
⬍00001
1.03 (1.02 to 1.04)
⬍0.0001
1.03 (1.02 to 1.04)
⬍0.0001
⬍22.5 kg/m2
328 (36.9)
14.0
2.36 (2.01 to 2.77)
⬍0.0001
1.93 (1.64 to 2.27)
⬍0.0001
1.69 (1.43 to 2.01)
⬍0.0001
22.5 to 24.9 kg/m2
365 (28.6)
10.3
1.74 (1.48 to 2.03)
⬍0.0001
1.47 (1.26 to 1.72)
⬍0.0001
1.46 (1.24 to 1.71)
⬍0.0001
25 to 29.9 kg/m2
710 (23.2)
8.1
1.36 (1.18 to 1.56)
⬍0.0001
1.22 (1.06 to 1.40)
0.006
1.22 (1.06 to 1.41)
0.005
30 to 34.9 kg/m2
280 (17.7)
6.0
1.00 (referent)
...
1.00 (referent)
...
1.00 (referent)
...
ⱖ35 kg/m
148 (18.7)
6.5
1.09 (0.89 to 1.32)
0.42
1.26 (1.03 to 1.54)
0.022
1.17 (0.95 to 1.43)
0.13
1460 (19.2)
6.8
1.05 (1.04 to 1.06)
⬍0.0001
1.04 (1.02 to 1.05)
⬍0.0001
1.03 (1.02 to 1.04)
⬍0.0001
⬍22.5 kg/m2
247 (27.8)
10.6
2.21 (1.85 to 2.65)
⬍0.0001
1.87 (1.56 to 2.25)
⬍0.0001
1.61 (1.33 to 1.95)
⬍0.0001
22.5 to 24.9 kg/m2
297 (23.3)
8.4
1.76 (1.48 to 2.09)
⬍0.0001
1.52 (1.28 to 1.82)
⬍0.0001
1.50 (1.25 to 1.79)
⬍0.0001
2
25 to 29.9 kg/m
581 (19.0)
6.6
1.39 (1.19 to 1.62)
⬍0.0001
1.26 (1.08 to 1.47)
0.004
1.26 (1.08 to 1.47)
0.004
30 to 34.9 kg/m2
224 (14.2)
4.8
1.00 (referent)
...
1.00 (referent)
ⱖ35 kg/m2
111 (14.0)
4.8
1.02 (0.81 to 1.28)
0.89
1.16 (0.93 to 1.46)
371 (4.9)
1.7
1.06 (1.04 to 1.08)
⬍0.0001
⬍22.5 kg/m2
81 (9.1)
3.5
2.96 (2.10 to 4.16)
22.5 to 24.9 kg/m2
68 (5.3)
1.9
1.63 (1.15 to 2.32)
All-cause death
BMI as continuous variable
per 1-kg/m2 decrease
BMI as categorical variable
2
Cardiovascular death
BMI as continuous variable
per 1-kg/m2 decrease
BMI as categorical variable
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
1.00 (referent)
...
...
0.20
1.09 (0.87 to 1.38)
0.45
1.03 (1.00 to 1.05)
0.019
1.03 (1.01 to 1.05)
0.0001
⬍0.0001
2.12 (1.50 to 3.00)
⬍0.0001
2.03 (1.41 to 2.91)
0.0001
0.007
1.28 (0.89 to 1.82)
0.18
1.32 (0.91 to 1.89)
0.14
Noncardiovascular death
BMI as continuous variable
per 1-kg/m2 decrease
BMI as categorical variable
25 to 29.9 kg/m2
129 (4.2)
1.5
1.24 (0.91 to 1.70)
0.18
1.06 (0.77 to 1.46)
0.71
1.09 (0.79 to 1.5)
0.60
30 to 34.9 kg/m2
56 (3.5)
1.2
1.00 (referent)
...
1.00 (referent)
...
1.00 (referent)
...
ⱖ35 kg/m
37 (4.7)
1.6
1.36 (0.90 to 2.06)
0.15
1.71 (1.13 to 2.59)
0.012
1.52 (0.99 to 2.32)
0.053
1675 (22.7)
8.6
1.00 (1.00 to 1.02)
0.076
1.00 (0.99 to 1.00)
0.34
1.00 (0.99 to 1.01)
0.94
⬍22.5 kg/m2
229 (25.8)
11.0
1.36 (1.15 to 1.61)
0.0003
1.18 (1.00 to 1.40)
0.054
1.10 (0.92 to 1.31)
0.30
22.5 to 24.9 kg/m2
281 (22.0)
8.8
1.10 (0.94 to 1.29)
0.23
0.99 (0.84 to 1.16)
0.88
1.03 (0.87 to 1.21)
0.72
2
25 to 29.9 kg/m
646 (21.1)
8.2
1.03 (0.91 to 1.18)
0.63
0.97 (0.85 to 1.11)
0.64
1.02 (0.89 to 1.16)
0.80
30 to 34.9 kg/m2
334 (21.2)
7.9
1.00 (referent)
...
1.00 (referent)
...
1.00 (referent)
...
ⱖ35 kg/m2
185 (23.4)
9.1
1.14 (0.95 to 1.36)
0.15
1.27 (1.06 to 1.52)
0.011
1.10 (0.92 to 1.33)
0.29
2
Hospitalization for worsening HF
BMI as continuous variable
per 1-kg/m2 decrease
BMI as categorical variable
*See text for details.
ventricular ejection fraction. The increased risk of death
among patients with low BMI was evident primarily in the
absence of edema, an indicator of fluid overload status.
Few studies have addressed the issue of shape of the
association between BMI and all-cause death in chronic HF
patients. In a report from the Digitalis Investigation Group
(DIG) trial, unadjusted analyses using polynomial logistic
regression revealed a nonlinear association between BMI and
death.13 In the present investigation, age- and sex-adjusted as
well as multivariable time-to-event analyses that accounted for
major indicators of poor prognosis also suggest a nonlinear (Uor J-shaped) relation between BMI and mortality. Commensurate with this finding, an inverted J-shaped association was noted
between BMI and days alive out of hospital, a type of quality of
life-adjusted survival that accounted for multiple hospitalizations
during the period of evaluation.
Edema increases BMI on the basis of fluid excess rather
than solid tissue mass. Results of our investigation suggest
that in patients with chronic HF, baseline BMI can be a useful
indicator of mortality primarily in the absence of edema.
In the DIG trial13 as well as in our present investigation,
in which relatively younger (mean age, 63 to 65 years) and
stable HF patients were enrolled, left ventricular systolic
dysfunction, as assessed by estimates of left ventricular
ejection fraction, did not alter the effects of BMI on the
risk of all-cause death. In the Danish Investigations of
Kenchaiah et al
During 1 year of follow-up
During 2 years of follow-up
Body mass index categories
Body mass index categories
0.0
-5 . 1
-6.6
- 7.5
-9.5
-10.0
-10.4
-11.3
-11.8
-14.1
p=0.061
*
-15.0
*
*
-20.0
Body mass index
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
No of patients
Days alive out of
hospital (Mean±SD)
0.0
Days alive out of hospital (estimated mean difference)
Days alive out of hospital (estimated mean difference)
-3.7
-4.9
- 5.0
0.0
0.2
0.0
-17.1
BMI and Prognosis in Chronic HF
7
1.5
-9.5
-10.0
-12.0
-16.9
-19.0
-20.0
-25.4
-26.0
-28.6
-28.7
-30.0
p=0.081
*
-39.3
-40.0
-40.4
*
-50.0
-52.6
-60.0
*
<22.5
22.5-24.9
25-29.9
30-34.9
≥35
<22.5
22.5-24.9
25-29.9
30-34.9
889
1277
3063
1579
791
889
1277
3063
1579
791
334.1
±76.7
339.8
±70.7
343.5
±65.3
352.8
±43.7
348.9
±53.1
626.4
±196.2
646.7
±179.4
661.5
±164.4
686.4
±123.6
677.8
±140.6
≥35
Figure 2. Results of fully adjusted linear regression analyses evaluating the relation between BMI and days alive out of hospital during
1 and 2 years of follow-up. BMI between 30 and 34.9 was the reference category. Vertical bars represent 95% CIs. *P⬍0.0001.
Arrhythmia and Mortality (DIAMOND) congestive heart
failure study, in which older (mean age, 72 years) HF
patients admitted for new-onset or worsening HF were
enrolled, in patients with left ventricular systolic dysfunction, as assessed by wall motion index, higher BMI was
associated with a greater risk of all-cause death, whereas in
patients with normal left ventricular systolic function,
lower BMI posed a greater risk of all-cause death.12 The
differences in study population and mode of ascertainment
of left ventricular systolic function may partially explain
the disparity in interaction of left ventricular function and
BMI and the risk of mortality.
In our analyses as well as in the study by DIG investigators,13 BMI did not influence the risk of hospitalization for
HF in multivariable analyses. The lack of association with
this sensitive clinical measure of HF severity, along with the
similarity of effect of BMI on cardiovascular and noncardiovascular deaths, suggests that low BMI is a general marker
(not a causal factor) of ill health.
A number of hypotheses have been put forth to explain the
association between lower BMI and increased risk of mortality in HF patients.22–25 Lower BMI in patients with “prevalent” HF, as in the present and other previously published
studies,7–9,11–14 may have been due to preceding involuntary
weight loss or cardiac cachexia, a heightened metabolic or
increased catabolic state26 associated with worse prognosis.8,27,28 By the same token, higher BMI, except when
produced by edema, might indicate a greater metabolic
reserve, increased tolerance to metabolic stress,29 and/or a
less severe form of HF, and consequently may have a better
prognosis.30 Elevated BMI may be associated with reduced
cardiac sympathetic activity,31 attenuated neurohormonal response to stress,32 lower proinflammatory cytokine levels,
and lesser catabolic-anabolic imbalance,33 all of which bode
a better prognosis. Increased production of leptin34 and
soluble receptors of tumor necrosis factor35 in adipose tissues
may reduce the detrimental effects of tumor necrosis factor, a
proinflammatory cytokine. Individuals with higher BMI,
presumably due to lower natriuretic peptide levels,36,37 may
manifest symptoms and signs of HF at a younger age with
consequent longer survival after diagnosis (lead-time bias).
Elevated BMI may increase the risk of death and unfavorable cardiovascular outcomes by promoting risk factors for
cardiovascular disease19,38 and by influencing abnormal left
ventricular remodeling.39 Extreme BMI may continue to pose
greater hemodynamic stress on the heart that could be
deleterious. The improvement in functional class and pump
function noted in smaller studies in which extremely obese
HF patients have undergone marked weight reductions40,41
lends credence to this notion.
The strengths of our investigation include a large sample size,
multicenter setting, inclusion of a broad spectrum of left ventricular ejection fractions in patients with symptomatic HF on
evidence-based treatments in a contemporary time period, assessment of edema status, availability of information on most
covariates of interest (including duration of HF) for multivariable analyses, nearly complete follow-up of patients during the
2- to 4-year study period, and objective ascertainment of all
study end points by an independent committee blinded to BMI
status at enrollment. Several limitations of these analyses should
also be noted. We did not have data on body weights at the time
of first diagnosis of HF to ascertain the influence of weight
change subsequent to the time of first diagnosis on the risk of
death or cardiovascular outcomes. Selection bias is likely in all
prognostic studies conducted on patients with prevalent HF (in
which BMI at the time of onset of HF is unavailable, and
changes in body weight preceding their inclusion in various
categories of BMI are uncertain).24 Except for edema status, we
did not have data on body composition or percent body fat
to evaluate the impact of each component (muscle, bone,
8
Circulation
August 7, 2007
TABLE 3. Results of Multivariable Cox Proportional Hazards Models Examining the Relation of BMI to the Risk of All-Cause Death
and Hospitalization for Worsening HF According to Left Ventricular Ejection Fraction and Edema Status at Baseline Examination
BMI Categories, kg/m2
⬍22.5
22.5 to 24.9
25 to 29.9
30 to 34.9
ⱖ35
No. of events/at
risk (%), rate/100 py
258/641 (40.2), 15.5
274/818 (33.5), 12.4
527/1876 (28.1), 10.0
201/878 (22.9), 7.8
90/363 (24.8), 8.6
Multivariable HR (95%
CI), P*
1.69 (1.39 to 2.06), ⬍0.0001
1.44 (1.19 to 1.74), 0.0001
1.20 (1.02 to 1.41), 0.032
1.00 (referent)
1.19 (0.92 to 1.53), 0.18
No. of events/at
risk (%), rate/100 py
70/248 (28.2), 10.4
91/459 (19.8), 6.9
183/1187 (15.4), 5.3
79/701 (11.3), 3.8
58/428 (13.6), 4.7
Multivariable HR (95%
CI), P*
1.99 (1.42 to 2.80), ⬍0.0001
1.67 (1.23 to 2.29), 0.001
1.33 (1.01 to 1.74), 0.039
1.00 (referent)
1.25 (0.88 to 1.77), 0.21
No. of events/at
risk (%), rate/100 py
272/753 (36.1), 13.6
285/1074 (26.5), 9.5
493/2409 (20.5), 7.0
174/1117 (15.6), 5.2
56/392 (14.3), 4.7
Multivariable HR (95%
CI), P*
1.92 (1.57 to 2.34), ⬍0.0001
1.56 (1.29 to 1.89), ⬍0.0001
1.22 (1.03 to 1.46), 0.024
1.00 (referent)
1.09 (0.80 to 1.47), 0.60
No. of events/at
risk (%), rate/100 py
56/136 (41.2), 16.3
80/203 (39.4), 15.2
217/654 (33.2), 12.5
106/462 (22.9), 8.1
92/399 (23.1), 8.3
Multivariable HR (95%
CI), P*
1.10 (0.78 to 1.56), 0.57
1.21 (0.89 to 1.64), 0.22
1.25 (0.99 to 1.59), 0.060
1.00 (referent)
1.24 (0.93 to 1.64), 0.14
No. of events/at
risk (%), rate/100 py
182/641 (28.4), 12.4
210/818 (25.7), 10.8
454/1876 (24.2), 9.7
221/878 (25.2), 9.7
91/363 (25.1), 9.7
Multivariable HR (95%
CI), P*
1.18 (0.96 to 1.46), 0.11
1.08 (0.89 to 1.31), 0.45
1.00 (0.85 to 1.18), 0.99
1.00 (referent)
1.00 (0.78 to 1.28), 1.00
No. of events/at
risk (%), rate/100 py
47/248 (19.0), 7.6
71/459 (15.5), 5.7
192/1187 (16.2), 6.0
113/701 (16.1), 5.8
94/428 (22.0), 8.5
Multivariable HR (95%
CI), P*
1.02 (0.71 to 1.45), 0.93
0.97 (0.71 to 1.32), 0.83
1.09 (0.86 to 1.39), 0.47
1.00 (referent)
1.28 (0.97 to 1.69), 0.087
No. of events/at
risk (%), rate/100 py
182/753 (24.2), 10.1
217/1074 (20.2), 7.9
452/2409 (18.8), 7.1
198/1117 (17.7), 6.4
74/392 (18.9), 6.8
Multivariable HR (95%
CI), P*
1.18 (0.96 to 1.46), 0.12
1.11 (0.91 to 1.35), 0.30
1.07 (0.9 to 1.27), 0.43
1.00 (referent)
1.14 (0.87 to 1.49), 0.35
No. of events/at
risk (%), rate/100 py
47/136 (34.6), 16.3
64/203 (31.5), 14.2
194/654 (29.7), 13.0
136/462 (29.4), 11.9
111/399 (27.8), 11.7
Multivariable HR (95%
CI), P*
0.95 (0.67 to 1.34), 0.75
0.93 (0.68 to 1.26), 0.63
0.98 (0.78 to 1.23), 0.88
1.00 (referent)
1.06 (0.82 to 1.36), 0.67
All-cause death
Left ventricular ejection
fraction ⱕ40
Left ventricular ejection
fraction ⬎40
Without edema
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
With edema
Hospitalization for worsening
HF
Left ventricular ejection
fraction ⱕ40
Left ventricular ejection
fraction ⬎40
Without edema
With edema
py indicates person-years.
*See text for details.
and fat tissue) on the risk of clinical outcomes.42 Results of
analyses stratified according to baseline edema status
should be interpreted with caution given the small sample
size of patients with edema in lower-BMI groups. We did
not have other baseline data such as serum albumin levels
for all participants, which may have allowed further
elucidation of underweight status versus cardiac cachexia (or protein-calorie malnutrition) in patients with the
lowest BMI values. Higher-BMI patients with HF were
younger at the time of enrollment than lower-BMI patients,
suggesting lead-time bias as a possible explanation for the
association between higher BMI and better survival. However, the similar duration of HF in all BMI groups and
absence of interaction between age and BMI and the risk of
death in multivariable analyses do not support this explanation. Because a majority of patients studied were of
European origin, the generalizability of our findings to
other races and ethnic groups is another limitation.
We conclude that in patients with symptomatic and chronic
HF, underweight status or low BMI is associated with a
Kenchaiah et al
greater risk of all-cause death, and mild to moderate overweight status is associated with the lowest risk. The implication of moderate to extreme BMI on prognosis is uncertain.
Lower BMI is an indicator of poor survival, primarily in the
absence of fluid overload. Our findings warrant additional
research, especially in community-based settings among patients with new-onset HF, to assess the impact of BMI,
ascertained in close proximity to the first diagnosis of HF, on
morbidity and mortality.
Source of Funding
The CHARM program was funded by AstraZeneca, which was
responsible for data collection.
Disclosures
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
The data analysis for this article was performed independently by
Drs Kenchaiah, Pocock, and Wang. The Executive Committee (Drs
Pfeffer, Swedberg, Granger, McMurray, and Yusuf) supervised the
management of the study and were primarily responsible for the
interpretation of the data and review and approval of the manuscript.
Dr Michelson is an employee of AstraZeneca. Drs Pocock, Pfeffer,
Yusuf, Swedberg, Granger, McMurray, and Solomon have received
research grants, honoraria for lectures, and/or consulting fees from
AstraZeneca. The remaining authors report no conflicts.
References
1. Chen YT, Vaccarino V, Williams CS, Butler J, Berkman LF, Krumholz
HM. Risk factors for heart failure in the elderly: a prospective
community-based study. Am J Med. 1999;106:605– 612.
2. He J, Ogden LG, Bazzano LA, Vupputuri S, Loria C, Whelton PK. Risk
factors for congestive heart failure in US men and women: NHANES I
epidemiologic follow-up study. Arch Intern Med. 2001;161:996 –1002.
3. Wilhelmsen L, Rosengren A, Eriksson H, Lappas G. Heart failure in the
general population of men: morbidity, risk factors and prognosis. J Intern
Med. 2001;249:253–261.
4. Kenchaiah S, Evans JC, Levy D, Wilson PW, Benjamin EJ, Larson MG,
Kannel WB, Vasan RS. Obesity and the risk of heart failure. N Engl
J Med. 2002;347:305–313.
5. Manson JE, Willett WC, Stampfer MJ, Colditz GA, Hunter DJ,
Hankinson SE, Hennekens CH, Speizer FE. Body weight and mortality
among women. N Engl J Med. 1995;333:677– 685.
6. Calle EE, Thun MJ, Petrelli JM, Rodriguez C, Heath CW Jr. Body-mass
index and mortality in a prospective cohort of U.S. adults. N Engl J Med.
1999;341:1097–1105.
7. Lissner L, Odell PM, D’Agostino RB, Stokes J III, Kreger BE, Belanger
AJ, Brownell KD. Variability of body weight and health outcomes in the
Framingham population. N Engl J Med. 1991;324:1839 –1844.
8. Anker SD, Ponikowski P, Varney S, Chua TP, Clark AL, Webb-Peploe
KM, Harrington D, Kox WJ, Poole-Wilson PA, Coats AJ. Wasting as
independent risk factor for mortality in chronic heart failure. Lancet.
1997;349:1050 –1053.
9. Mosterd A, Cost B, Hoes AW, de Bruijne MC, Deckers JW, Hofman A,
Grobbee DE. The prognosis of heart failure in the general population: the
Rotterdam Study. Eur Heart J. 2001;22:1318 –1327.
10. Lissin LW, Gauri AJ, Froelicher VF, Ghayoumi A, Myers J, Giacommini
J. The prognostic value of body mass index and standard exercise testing
in male veterans with congestive heart failure. J Card Fail. 2002;8:
206 –215.
11. Davos CH, Doehner W, Rauchhaus M, Cicoira M, Francis DP, Coats AJ,
Clark AL, Anker SD. Body mass and survival in patients with chronic
heart failure without cachexia: the importance of obesity. J Card Fail.
2003;9:29 –35.
12. Gustafsson F, Kragelund CB, Torp-Pedersen C, Seibaek M, Burchardt H,
Akkan D, Thune JJ, Kober L. Effect of obesity and being overweight on
long-term mortality in congestive heart failure: influence of left ventricular systolic function. Eur Heart J. 2005;26:58 – 64.
13. Curtis JP, Selter JG, Wang Y, Rathore SS, Jovin IS, Jadbabaie F,
Kosiborod M, Portnay EL, Sokol SI, Bader F, Krumholz HM. The obesity
paradox: body mass index and outcomes in patients with heart failure.
Arch Intern Med. 2005;165:55– 61.
BMI and Prognosis in Chronic HF
9
14. Alla F, Briancon S, Juilliere Y, Mertes PM, Villemot JP, Zannad F.
Differential clinical prognostic classifications in dilated and ischemic
advanced heart failure: the EPICAL study. Am Heart J. 2000;139:
895–904.
15. Ho KK, Anderson KM, Kannel WB, Grossman W, Levy D. Survival after
the onset of congestive heart failure in Framingham Heart Study subjects.
Circulation. 1993;88:107–115.
16. Swedberg K, Pfeffer M, Granger C, Held P, McMurray J, Ohlin G,
Olofsson B, Ostergren J, Yusuf S; CHARM-Programme Investigators.
Candesartan in Heart failure: Assessment of Reduction in Mortality and
morbidity (CHARM): rationale and design. J Card Fail. 1999;5:
276 –282.
17. Pfeffer MA, Swedberg K, Granger CB, Held P, McMurray JJ, Michelson
EL, Olofsson B, Ostergren J, Yusuf S, Pocock S. Effects of candesartan
on mortality and morbidity in patients with chronic heart failure: the
CHARM-Overall programme. Lancet. 2003;362:759 –766.
18. Obesity: Preventing and Managing the Global Epidemic: Report of a
WHO Consultation on Obesity, Geneva, June 3–5, 1997. Geneva, Switzerland: World Health Organization; 1998.
19. Clinical Guidelines on the Identification, Evaluation, and Treatment of
Overweight and Obesity in Adults. Bethesda, Md: National Institutes of
Health; 1998. Publication NIH 98-4083.
20. Pocock SJ, Wang D, Pfeffer MA, Yusuf S, McMurray JJ, Swedberg KB,
Ostergren J, Michelson EL, Pieper KS, Granger CB. Predictors of mortality and morbidity in patients with chronic heart failure. Eur Heart J.
2006;27:65–75.
21. SAS Institute Inc. SAS/STAT User’s Guide, Version 9.1. Cary, NC: SAS
Institute Inc; 2004.
22. Lavie CJ, Milani R, Mehra MR, Ventura HO, Messerli FH. Obesity,
weight reduction and survival in heart failure. J Am Coll Cardiol. 2002;
39:1563–1564.
23. Kalantar-Zadeh K, Block G, Horwich T, Fonarow GC. Reverse epidemiology of conventional cardiovascular risk factors in patients with chronic
heart failure. J Am Coll Cardiol. 2004;43:1439 –1444.
24. Kenchaiah S, Gaziano JM, Vasan RS. Impact of obesity on the risk of
heart failure and survival after the onset of heart failure. Med Clin North
Am. 2004;88:1273–1294.
25. Lavie CJ, Mehra MR, Milani RV. Obesity and heart failure prognosis:
paradox or reverse epidemiology? Eur Heart J. 2005;26:5–7.
26. Berry C, Clark AL. Catabolism in chronic heart failure. Eur Heart J.
2000;21:521–532.
27. Anker SD, Rauchhaus M. Insights into the pathogenesis of chronic heart
failure: immune activation and cachexia. Curr Opin Cardiol. 1999;14:
211–216.
28. Anker SD, Negassa A, Coats AJ, Afzal R, Poole-Wilson PA, Cohn JN,
Yusuf S. Prognostic importance of weight loss in chronic heart failure and
the effect of treatment with angiotensin-converting-enzyme inhibitors: an
observational study. Lancet. 2003;361:1077–1083.
29. Coats A, Anker SD, Roecker EB, Schultz MK, Staiger C, Amann-Zalan
I, Shusterman N, Packer M. Prevention and reversal of cardiac cachexia
in patients with severe heart failure by carvedilol: results of the
COPERNICUS study. Circulation. 2001;104 II-437.
30. Horwich TB, Fonarow GC, Hamilton MA, MacLellan WR, Woo MA,
Tillisch JH. The relationship between obesity and mortality in patients
with heart failure. J Am Coll Cardiol. 2001;38:789 –795.
31. Vaz M, Jennings G, Turner A, Cox H, Lambert G, Esler M. Regional
sympathetic nervous activity and oxygen consumption in obese normotensive human subjects. Circulation. 1997;96:3423–3429.
32. Weber MA, Neutel JM, Smith DH. Contrasting clinical properties and
exercise responses in obese and lean hypertensive patients. J Am Coll
Cardiol. 2001;37:169 –174.
33. Anker SD, Clark AL, Kemp M, Salsbury C, Teixeira MM, Hellewell PG,
Coats AJ. Tumor necrosis factor and steroid metabolism in chronic heart
failure: possible relation to muscle wasting. J Am Coll Cardiol. 1997;30:
997–1001.
34. Takahashi N, Waelput W, Guisez Y. Leptin is an endogenous protective
protein against the toxicity exerted by tumor necrosis factor. J Exp Med.
1999;189:207–212.
35. Mohamed-Ali V, Goodrick S, Bulmer K, Holly JM, Yudkin JS, Coppack
SW. Production of soluble tumor necrosis factor receptors by human
subcutaneous adipose tissue in vivo. Am J Physiol. 1999;277:E971–E975.
36. Mehra MR, Uber PA, Park MH, Scott RL, Ventura HO, Harris BC,
Frohlich ED. Obesity and suppressed B-type natriuretic peptide levels in
heart failure. J Am Coll Cardiol. 2004;43:1590 –1595.
37. Wang TJ, Larson MG, Levy D, Benjamin EJ, Leip EP, Wilson PW,
10
Circulation
August 7, 2007
Vasan RS. Impact of obesity on plasma natriuretic peptide levels. Circulation. 2004;109:594 – 600.
38. Eckel RH, Krauss RM; AHA Nutrition Committee. American Heart
Association call to action: obesity as a major risk factor for coronary heart
disease. Circulation. 1998;97:2099 –2100.
39. Lauer MS, Anderson KM, Kannel WB, Levy D. The impact of obesity on
left ventricular mass and geometry: the Framingham Heart Study. JAMA.
1991;266:231–236.
40. Alpert MA, Lambert CR, Terry BE, Cohen MV, Mulekar M, Massey CV,
Hashimi MW, Panayiotou H, Mukerji V. Effect of weight loss on left
ventricular diastolic filling in morbid obesity. Am J Cardiol. 1995;76:
1198 –1201.
41. Alpert MA, Terry BE, Mulekar M, Cohen MV, Massey CV, Fan TM,
Panayiotou H, Mukerji V. Cardiac morphology and left ventricular
function in normotensive morbidly obese patients with and without congestive heart failure, and effect of weight loss. Am J Cardiol. 1997;80:
736 –740.
42. Lavie CJ, Osman AF, Milani RV, Mehra MR. Body composition and
prognosis in chronic systolic heart failure: the obesity paradox.
Am J Cardiol. 2003;91:891– 894.
CLINICAL PERSPECTIVE
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
In 7599 symptomatic heart failure (New York Heart Association class II to IV) patients (mean age, 65 years; 35% women)
with a broad spectrum of left ventricular ejection fractions (mean, 39%) in the Candesartan in Heart failure: Assessment
of Reduction in Mortality and morbidity (CHARM) clinical trials, we examined the influence of body mass index (BMI)
(defined as weight in kilograms divided by the square of height in meters) on prognosis. During a median follow-up of 38
months, 1831 patients died. Lowest mortality was noted in patients with BMI between 30 and ⬍35, and progressive
increase in mortality was evident below this range (22%, 46%, and 69% increase in patients with BMI 25 to ⬍30, 22.5 to
⬍25, and ⬍22.5, respectively), and a plateau or increase was observed above this range (17% increase, albeit statistically
nonsignificant, among patients with BMI ⱖ35). Similarly, highest number of days alive out of hospital was observed in
patients with BMI between 30 and 34.9 (39, 29, 19, and 12 fewer days alive out of hospital during 2 years of follow-up
in patients with BMI ⬍22.5, 22.5 to ⬍25, 25 to ⬍30, and ⱖ35, respectively). The association between BMI and mortality
was similar in patients with impaired as well as preserved left ventricular ejection fraction. However, lower BMI was
associated with a greater risk of death primarily in patients without edema, an indicator of fluid overload status. Lower BMI
(likely a consequence of involuntary weight loss or cachexia) was linked with a greater risk of cardiovascular death and
noncardiovascular death (including cancer deaths), suggesting that among patients with heart failure, low BMI is a general
marker of ill health.
Downloaded from http://circ.ahajournals.org/ by guest on June 15, 2017
Body Mass Index and Prognosis in Patients With Chronic Heart Failure. Insights From
the Candesartan in Heart failure: Assessment of Reduction in Mortality and morbidity
(CHARM) Program
Satish Kenchaiah, Stuart J. Pocock, Duolao Wang, Peter V. Finn, Leonardo A.M. Zornoff,
Hicham Skali, Marc A. Pfeffer, Salim Yusuf, Karl Swedberg, Eric L. Michelson, Christopher B.
Granger, John J.V. McMurray and Scott D. Solomon
for the CHARM Investigators
Circulation. published online July 16, 2007;
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2007 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/early/2007/07/16/CIRCULATIONAHA.106.679779.citation
Data Supplement (unedited) at:
http://circ.ahajournals.org/content/suppl/2007/08/09/CIRCULATIONAHA.106.679779.DC1
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/