Download Cardiovascular Features of Heart Failure With Preserved Ejection

Journal of the American College of Cardiology
© 2007 by the American College of Cardiology Foundation
Published by Elsevier Inc.
Vol. 49, No. 2, 2007
ISSN 0735-1097/07/$32.00
doi:10.1016/j.jacc.2006.08.050
Heart Failure
Cardiovascular Features of Heart Failure
With Preserved Ejection Fraction Versus
Nonfailing Hypertensive Left Ventricular
Hypertrophy in the Urban Baltimore Community
The Role of Atrial Remodeling/Dysfunction
Vojtech Melenovsky, MD,*†§ Barry A. Borlaug, MD,* Boaz Rosen, MD,* Ilan Hay, MD,*
Luigi Ferruci, MD, PHD,‡ Christopher H. Morell, PHD,§ Edward G. Lakatta, MD,†
Samer S. Najjar, MD,† David A. Kass, MD*
Baltimore, Maryland
Objectives
The purpose of this study was to identify cardiovascular features of patients with heart failure with preserved
ejection fraction (HFpEF) that differ from those in individuals with hypertensive left ventricular hypertrophy
(HLVH) of similar age, gender, and racial background but without failure.
Background
Heart failure with preserved ejection fraction often develops in HLVH patients and involves multiple abnormalities. Clarification of changes most specific to HFpEF may help elucidate underlying pathophysiology.
Methods
A cross-sectional study comparing HFpEF patients (n ⫽ 37), HLVH subjects without HF (n ⫽ 40), and normotensive control subjects without LVH (n ⫽ 56). All subjects had an EF of ⬎50%, sinus rhythm, and insignificant valvular or active ischemic disease, and groups were matched for age, gender, and ethnicity. Comprehensive echoDoppler and pressure analysis was performed.
Results
The HFpEF patients were predominantly African-American women with hypertension, LVH, and obesity. They had vascular and systolic-ventricular stiffening and abnormal diastolic function compared with the control subjects. However,
most of these parameters either individually or combined were similarly abnormal in the HLVH group and poorly distinguished between these groups. The HFpEF group had quantitatively greater concentric LVH and estimated mean
pulmonary artery wedge pressure (20 mm Hg vs. 16 mm Hg) and shorter isovolumic relaxation time than the HLVH
group. They also had left atrial dilation/dysfunction unlike in HLVH and greater total epicardial volume. The product of
LV mass index and maximal left atrial (LA) volume best identified HFpEF patients (84% sensitivity, 82% specificity).
Conclusions
In an urban, principally African American, cohort, HFpEF patients share many abnormalities of systolic, diastolic,
and vascular function with nonfailing HLVH subjects but display accentuated LVH and LA dilation/failure. These
latter factors may help clarify pathophysiology and define an important HFpEF population for clinical
trials. (J Am Coll Cardiol 2007;49:198–207) © 2007 by the American College of Cardiology Foundation
Nearly one-half of patients with heart failure have an ejection fraction (EF) at or above 50% (1). Despite the high
prevalence, morbidity, and economic burden (2– 4) of heart
failure with preserved ejection fraction (HFpEF), its pathophysiology remains somewhat controversial and evidencebased guidelines for management are lacking. A majority of
HFpEF patients are elderly women with a history of systolic
hypertension, many with cardiac hypertrophy and obesity
(1,5–9). These are common features in the general population, and frequent among African-American women, who
have the highest prevalence of hypertension worldwide (10).
Most are without heart failure (HF) symptoms, and although they may have preclinical disease (11) only a
subgroup develops clinical HFpEF.
From the *Division of Cardiology, Department of Medicine, Johns Hopkins
University, Baltimore, Maryland; †Laboratory of Cardiovascular Science and ‡Clinical
Research Branch, Intramural Research Program, National Institute on Aging, National
Institutes of Health, Baltimore, Maryland; and the §Department of Mathematical
Sciences, Loyola College in Maryland, Baltimore, Maryland. Dr. Melenovsky’s current
address is the Institute of Clinical and Experimental Medicine (IKEM), Prague,
Czech Republic. Supported by a grant from the National Institute on Aging (NIA)
(AG:18324) and Intramural Research Program of the National Institutes of Health, NIA.
Manuscript received January 18, 2006; revised manuscript received August 8, 2006,
accepted August 14, 2006.
JACC Vol. 49, No. 2, 2007
January 16, 2007:198–207
Cardiovascular features that might best distinguish patients with HF symptoms from those with similar clinical
features such as hypertensive left ventricular hypertrophy
(HLVH) but without HF remain unknown, because most
studies have contrasted HFpEF patients to healthy normotensive (non-LVH) control subjects. Abnormal diastolic
function (2,12–16), vascular stiffening (17–19), and increased ventricular end-systolic stiffness (elastance [18])
have all been implicated, but it is unclear which descriptors
are most specific and independent. Accordingly, the goal of
the present study was to comprehensively assess resting
ventricular and vascular physiologic properties in an innercity cohort of HFpEF patients, and compare the findings
with those obtained in 2 control groups with similar mean
age, race, and gender with or without chronic HLVH. We
hypothesized that many ventricular and arterial abnormalities
thought to underlie HFpEF would be shared by asymptomatic
HLVH subjects, but that other changes would be observed
more selectively in HFpEF that might highlight critical maladaptations and/or novel pathophysiology.
Methods
Study subjects. The HFpEF patients were prospectively
identified from admission records at Johns Hopkins Hospitals between January 2002 and June 2005. Heart failure
was rigorously defined by Framingham criteria (see the
Appendix for supplemental Table A) (20 –22) and independently adjudicated by 2 cardiologists. All HFpEF subjects
had been hospitalized for pulmonary congestion diagnosed
by chest radiogram and clinical examination, had EF ⬎50%
within 24 to 72 h of index admission, and were required to
be in sinus rhythm. There were no other a priori inclusion
criteria. Patients with active ischemic heart disease (i.e.,
angina, positive stress test, acute coronary syndrome), segmental wall motion abnormalities, atrial flutter/fibrillation,
valvular disease (more than trace regurgitation or stenosis),
primary hypertrophic or other cardiomyopathy (e.g., amyloid),
pericardial disease, anemia (hemoglobin ⬍10 g/dl), hyperthyroidism, renal failure (creatinine ⬎3 mg/dl or requiring
dialysis), advanced pulmonary disease or pulmonary hypertension (estimated pulmonary artery systolic ⬎60 mm Hg),
or cognitive impairment precluding participation were
excluded.
Two control groups were studied. The HLVH subjects
were identified from outpatient echocardiogram records and
screened using hospital records, interview, and formal history and physical to confirm lack of diagnosis, treatment, or
symptoms for clinical HF. The other group was made up of
normotensive subjects without LVH, recruited from participants in the Baltimore Longitudinal Study of Aging (8).
The latter were asymptomatic and had no prior diagnosis,
history, or hospitalization for HF or cardiovascular disease
(except for mild and well controlled hypertension in some).
Both control groups were selected from a larger screening
pool so that the predominant ethnicity (African American),
Melenovsky et al.
Pathophysiology of Heart Failure With Normal EF
199
gender (female), and age (⬎50
Abbreviations
and Acronyms
years) matched the mean values
for the HFpEF group. Identical
Ea ⴝ arterial elastance
exclusion criteria used for HFpEF
Ees ⴝ end-systolic
patients applied to the control
elastance
groups. The study was designed as
HF ⴝ heart failure
prospective. Owing to application
HFpEF ⴝ heart failure with
of selection criteria, studied subpreserved ejection fraction
jects were nonconsecutive. The
HLVH ⴝ hypertensive left
protocol was approved by the
ventricular hypertrophy
Johns Hopkins Medical InstituLA ⴝ left atrial
tions and National Institute on
LV ⴝ left ventricular
Aging Institutional Review Board,
LVH ⴝ left ventricular
and written informed consent was
hypertrophy
obtained.
PAWP ⴝ pulmonary artery
Study procedures. Clinical
wedge pressure
symptoms were assessed in the
HFpEF and HLVH groups by
the Minnesota Living with Heart Failure Questionnaire
(MLHFQ). All subjects were studied in a stable compensated state (⬎1 month after discharge from an HF hospitalization for HFpEF). Measurements were made after ⬎15
min rest in a supine position and involved comprehensive
echo-Doppler (Sonos 5500, Philips, Andover, Massachusetts), pulse tonometry (VP2000, Omron Healthcare Inc.,
Bannockburn, Illinois), and arm-cuff (Dinamap, Tampa,
Florida) pressure measurements. Reserve function to acute
afterload increase (sustained isometric handgrip, 30% of
maximal) was tested in the HFpEF and HLVH groups,
recording blood pressure and cardiac echo-Doppler images.
Cardiac function analysis. Echo-Doppler images were
analyzed off line by a single investigator blinded to subject
group. Stroke volume (SV) was determined as LV outflowtract area ⫻ flow velocity time integral by pulsed Doppler
(23). Ejection fraction was calculated from parasternal
long-axis views using the Teichholz formula. The LV
end-diastolic volume (EDV) was equal to SV/EF (24). The
LV mass (25) was indexed to body height (LVMI ⫽
g/m2.7), and LVH was defined using gender-specific thresholds (26). Geometry was indexed by relative wall thickness
(LV septum ⫹ posterior wall thickness)/LV internal diameter) and LVM/EDV. Total epicardial volume was calculated from 2 hemiellipsoids containing both atria or ventricles using an apical 4-chamber view.
Systolic function included EF, endocardial and midwall
fractional shortening (mFS) (27), ratio of observed (o) to
predicted (p) mFS (o/pmFS, a contractility index [12];
pmFS ⫽ 29.08 ⫺ 0.0006 ⫻ circumferential wall stress [28]),
and peak systolic chamber elastance (Ees) (29). Diastolic
function was assessed by peak early (E) and late (A) mitral
Doppler flow velocity, deceleration time, isovolumic relaxation time, pulmonary venous systolic/diastolic velocity ratio
(pvS/D), longitudinal early (E=) and late (A=) tissue velocity
(pulsed Doppler) at both mitral annular insertions, and
averaged data. Mean pulmonary artery wedge pressure
(PAWP) was estimated by E/E= (30 –32). Diastolic dys-
200
Melenovsky et al.
Pathophysiology of Heart Failure With Normal EF
JACC Vol. 49, No. 2, 2007
January 16, 2007:198–207
function was graded according to mitral, pulmonary vein
flow and myocardial tissue Doppler profiles using a modified scale (see the Appendix for supplemental Table B) (33).
Left atrial (LA) volumes were calculated by area-length
method from apical 4-chamber views (34) to measure
maximal, minimal, and diastasis LA volumes (LAVmax,
LAVmin, and LAVdiastasis) and active and passive emptying
function. In a subset of patients (HFpEF, n ⫽ 21; HLVH,
n ⫽ 25), tissue-Doppler images were obtained in color-code
mode from the apical 4-chamber view (Vivid7, GE Healthcare, Chalfont St. Giles, United Kingdom) at rest and at
peak of a 6-min sustained handgrip exercise to assess atrial
and cardiac reserve.
Arterial function. Arterial function was assessed by carotid artery applanation tonometry (35) and oscillometric
brachial pressure. Carotid systolic pressure augmentation
(augmentation index) was calculated from digitalized (1.2
kHz) waveforms by standard algorithm, and waveforms
were calibrated to match brachial mean and diastolic
blood pressure (35). Total arterial compliance was estimated by SV/carotid pulse pressure ratio (36), and
effective arterial elastance (Ea), a measure of total LV
afterload, was estimated by the LV end-systolic pressure/SV ratio (37).
Statistical analysis. Data are reported as mean ⫾ SD,
and p values are 2-sided with p ⬍ 0.05 considered to be
significant. Between-group differences were assessed by
1-way analysis of variance with a post hoc Bonferroni test
for 3 comparisons, and categorical variables were assessed
by chi-square test (version 12.0, SPSS Inc., Chicago,
Illinois). Correlations were tested by the Pearson coeffi-
cient. Between-group differences were further evaluated
by analysis of covariance to adjust for effects of specific
covariates.
To identify variables that best identified HFpEF from
HLVH groups, linear discriminant analysis with backward elimination was used to generate a linear predictor
function that provided the best between-group discrimination (version 9.1.2, SAS Institute, Cary, North Carolina). Remaining variables were entered into a classification/regression tree analysis (CART) yielding binary
splits at optimal values according to the largest improvement of goodness of the fit (S-PLUS 7.0, Insightful,
Seattle, Washington).
Results
Demographics and comorbidities. The HFpEF patients
were predominantly older African-American (76%)
women (84%). While not required for inclusion, all
HFpEF patients had hypertension and ventricular hypertrophy, and most were obese (body mass index [BMI]
⬎30 kg/m2 in 86%) (Table 1). The HFpEF patients had
a higher prevalence of noninsulin-dependent diabetes
(61%), coronary artery disease (CAD) history (42%),
anemia, and mild renal dysfunction, consistent with
earlier reports (5). The HFpEF patients were symptomatic, with a mean quality of life (QOL) score similar to
that of severe systolic failure. The HLVH subjects were
less obese (BMI ⬎30 kg/m2 in 50%), less treated with
diuretics, beta-blockers, and angiotensin-converting enzyme inhibitors (ACEI), and had a QOL score in the low
Clinical Characteristics
Table 1
Clinical Characteristics
Controls (n ⴝ 56)
Age‡, yrs
Female gender‡, %
African-American race‡, %
NYHA functional class
MLHFQ score
Height, cm
Weight, kg
HFpEF (n ⴝ 37)
65 ⫾ 11
65 ⫾ 10
70
84
61
1.2 ⫾ 0.4
76
2.6 ⫾ 0.5*†
np
50 ⫾ 34*
168 ⫾ 8
162 ⫾ 10†
78 ⫾ 18
99 ⫾ 24*†
HLVH Without HF (n ⴝ 40)
67 ⫾ 10
78
73
p Value
0.61
0.28
0.25
1.4 ⫾ 0.5
⬍0.01
10 ⫾ 12
⬍0.01
164 ⫾ 9
0.02
84 ⫾ 17
⬍0.01
Body mass index‡, kg/m2
28 ⫾ 5
37 ⫾ 8*†
31 ⫾ 6†
⬍0.01
Body surface area‡, m2
1.9 ⫾ 0.2
2.1 ⫾ 0.3*†
2.0 ⫾ 0.2
⬍0.01
Serum creatinine‡, mg/dl1
1.0 ⫾ 0.5
1.4 ⫾ 0.7*†
1.0 ⫾ 0.3
⬍0.01
13.4 ⫾ 1.4
11.6 ⫾ 1.3*†
12.8 ⫾ 1.5
⬍0.01
34
100†
Hemoglobin concentration‡, g/dl1
Comorbidity, %
93†
⬍0.01
Diabetes mellitus‡
3
61*†
35†
⬍0.01
Coronary artery disease‡
2
42*†
10
⬍0.01
Left ventricular hypertrophy
0
90†
⬍0.01
Hypertension‡
100†
Medication, %
Diuretics loop/non-loop
0/21
Calcium blocker‡/beta-blocker‡
9/11
ACEI‡/ARB/aldosterone antagonist
9/4/2
87*/20*
57/81*†
68*†/14/5
10/45
⬍0.01/0.02
35/43†
⬍0.01/⬍0.01
30†/30†/9
⬍0.01/⬍0.01/0.40
*p ⬍ 0.05 versus HLVH group; †p ⬍ 0.05 versus control group; Bonferroni post hoc test or chi-square for categorical variables; ‡variables used in discriminant analysis.
ACEI ⫽ angiotensin-converting enzyme inhibitors; ARB ⫽ angiotensin receptor blockers; HFpEF ⫽ heart failure with preserved ejection fraction; HLVH ⫽ hypertension and left ventricular hypertrophy;
MLHFQ ⫽ Minnesota Living with Heart Failure Questionnaire; np ⫽ not performed; NYHA ⫽ New York Heart Association.
Melenovsky et al.
Pathophysiology of Heart Failure With Normal EF
JACC Vol. 49, No. 2, 2007
January 16, 2007:198–207
201
Central and Peripheral Hemodynamic Parameters
Table 2
Central and Peripheral Hemodynamic Parameters
Controls (n ⴝ 56)
HFpEF (n ⴝ 37)
HLVH Without HF (n ⴝ 40)
Heart rate‡, min⫺1
65 ⫾ 8
63 ⫾ 11
64 ⫾ 11
Stroke volume, ml
81 ⫾ 21
77 ⫾ 17
77 ⫾ 11
118 ⫾ 15/65 ⫾ 10
143 ⫾ 25†/69 ⫾ 14
145 ⫾ 23†/74 ⫾ 12†
Brachial systolic‡ and diastolic‡ BP, mm Hg
Cardiac output‡, l/min⫺1
Total arterial resistance‡, dynes·s/cm5
5.0 ⫾ 0.88
1,345 ⫾ 262
5.0 ⫾ 1.1
4.9 ⫾ 1.3
1,558 ⫾ 397*†
1,707 ⫾ 532†
p Value
0.63
0.56
⬍0.01/⬍0.01
0.87
⬍0.01
Carotid mean BP, mm Hg
83 ⫾ 14
97 ⫾ 16†
103 ⫾ 19†
⬍0.01
Carotid pulse BP, mm Hg
39 ⫾ 9
62 ⫾ 18†
55 ⫾ 13†
⬍0.01
17 ⫾ 16
23 ⫾ 11
26 ⫾ 15†
0.01
Total arterial compliance, ml/mm Hg⫺1
2.08 ⫾ 0.6
1.36 ⫾ 0.5†
1.52 ⫾ 0.6†
⬍0.01
Arterial elastance‡, mm Hg/ml⫺1
1.40 ⫾ 0.2
1.67 ⫾ 0.5†
1.78 ⫾ 0.5†
⬍0.01
Carotid augmentation index‡, %
*p ⬍ 0.05 versus HLVH group; †p ⬍ 0.05 versus control group; Bonferroni post hoc test; ‡variables used in discriminant analysis.
BP ⫽ blood pressure; other abbreviations as in Table 1.
to normal range. The control subjects had a prevalence of
well controlled HTN typical for their age.
Hemodynamics and vascular properties. Heart rate and
stroke volume were nearly identical among groups (Table
2). The latter remained so after indexing to either body
height or body surface area (BSA) (p ⫽ 0.5 and 0.3).
Brachial and carotid pressures, total arterial resistance (raw
and BSA indexed), and Ea were greater in HFpEF than in
control subjects, and total arterial compliance was reduced—all
consistent with arterial stiffening and systolic hypertension.
Cardiac output was similar, although cardiac index was
slightly lower in HFpEF than in control subjects (see the
Appendix for supplemental Table C). Nearly identical
changes were observed in HLVH, suggesting they were due
to coexisting clinical features and not specific to HFpEF.
Left ventricular structure and systolic function. Ventricular hypertrophy was quantitatively greater in HFpEF
than in HLVH whether expressed as absolute mass or
indexed to body height, BSA (see the Appendix for supplemental Table C), or height2.7 (Table 3). LV end-diastolic
volume was similar among groups (Table 3) and remained
so after normalizing to body height, height2.7, or BSA (see
the Appendix for supplemental Table C). The LV mass/
volume or wall thickness/dimension concentricity indexes
were highest in HFpEF (Table 3, see the Appendix for
supplemental Table C). The EF was near-identical, whereas
the mitral-annular systolic velocity was slightly lower in
HFpEF and HLVH, consistent with LVH. Peak ventricular systolic stiffness (Ees) was higher in HFpEF than in
control subjects, as previously reported (18), but similarly
elevated in HLVH subjects. This change remained after
adjusting for medication use.
Diastolic ventricular function. The HFpEF patients had
several diastolic abnormalities compared with control subjects (Table 4); however, once again, many were similarly
altered in non-HF HLVH subjects. Only 2 parameters
differed between HFpEF and HLVH. The E/E= ratio, an
estimate of PAWP, was somewhat higher in HFpEF, with
more patients exceeding 18 mm Hg in HFpEF (57.1%)
than in HLVH (31.6%) (p ⫽ 0.03). This disparity persisted
after adjusting for LVMI or BSA (p ⫽ 0.024 and p ⫽
0.001, respectively). Isovolumic relaxation time (IVRT) was
shorter in HFpEF than in HLVH, which could relate to a
higher PAWP, although the variables were not significantly
correlated (p ⫽ 0.27).
Figure 1 shows diastolic dysfunction grades, revealing
substantial overlap in HFpEF and HLVH subjects. Although 59.4% of HFpEF patients had mild to moderate
Left Ventricular Structure and Systolic Function
Table 3
Left Ventricular Structure and Systolic Function
LV mass‡, g·m2.7
Controls (n ⴝ 56)
HFpEF (n ⴝ 37)
HLVH Without HF (n ⴝ 40)
p Value
36 ⫾ 7
81 ⫾ 23*†
58 ⫾ 12†
⬍0.01
LV relative wall thickness‡
0.48 ⫾ 0.09
0.68 ⫾ 0.14*†
0.60 ⫾ 0.10†
⬍0.01
LV ED wall thickness, mm
10 ⫾ 1.5/10 ⫾ 1.2
16 ⫾ 2.6*†/15 ⫾ 2.6*†
14 ⫾ 1.7†/13 ⫾ 1.7†
⬍0.01
42 ⫾ 5
45 ⫾ 5†
43 ⫾ 4
111 ⫾ 24
115 ⫾ 33
112 ⫾ 32
0.54
LV ejection fraction‡, %
72 ⫾ 10
72 ⫾ 11
72 ⫾ 10
0.98
Endocardial fractional shortening, %
41 ⫾ 8
42 ⫾ 10
41 ⫾ 8
0.80
120 ⫾ 34
106 ⫾ 40
120 ⫾ 35
0.15
Midwall fractional shortening, %
22 ⫾ 6
19 ⫾ 6
19 ⫾ 6
0.053
Observed/predicted midwall FS‡, %
98 ⫾ 28
85 ⫾ 24
85 ⫾ 28
2.73 ⫾ 0.7
3.51 ⫾ 1.8†
3.72 ⫾ 1.4†
⬍0.01
8.3 ⫾ 1.7
6.4 ⫾ 1.4†
7.2 ⫾ 1.4†
⬍0.01
LV ED diameter, mm
LV ED volume‡, ml
Circumferential ES stress, kdyne/cm2
LV ES elastance‡, mm Hg/ml⫺1
Systolic annular tissue velocity (Sm), cm/s‡
⬍0.01
0.025
*p ⬍ 0.05 versus HLVH group; †p ⬍ 0.05 versus control group; Bonferroni post hoc test; ‡variables used in discriminant analysis. Sm was averaged from septal and lateral part of mitral annulus.
ED ⫽ end-diastolic; ES ⫽ end-systolic; FS ⫽ fractional shortening; LV ⫽ left ventricular; other abbreviations as in Table 1.
202
Melenovsky et al.
Pathophysiology of Heart Failure With Normal EF
JACC Vol. 49, No. 2, 2007
January 16, 2007:198–207
Left Ventricular Diastolic Function and Left Atrial Morphology and Function
Table 4
Left Ventricular Diastolic Function and Left Atrial Morphology and Function
Controls (n ⴝ 56)
HFpEF (n ⴝ 37)
80 ⫾ 16/78 ⫾ 20
HLVH Without HF (n ⴝ 40)
p Value
0.05/⬍0.01
Mitral flow
86 ⫾ 27/92 ⫾ 25†
79 ⫾ 18/93 ⫾ 21†
1.1 ⫾ 0.3
1.0 ⫾ 0.4
0.9 ⫾ 0.3†
219 ⫾ 42
257 ⫾ 112†
Isovolumic relaxation time‡, ms
78 ⫾ 11
85 ⫾ 22*
96 ⫾ 17†
Pulmonary vein S/D velocity ratio‡
1.5 ⫾ 0.9
1.2 ⫾ 0.4
1.5 ⫾ 0.3
E=‡, cm/s
9.8 ⫾ 2.3
6.6 ⫾ 2.4†
7.6 ⫾ 2.7†
A=‡, cm/s
9.3 ⫾ 2.1
8.1 ⫾ 3.2
9.4 ⫾ 1.9
8.4 ⫾ 2.2
15 ⫾ 5.3*†
11 ⫾ 4.5†
⬍0.01
E‡ and A velocity, cm/s
E/A ratio‡
E–deceleration time‡, ms
247 ⫾ 64
0.02
0.03
⬍0.01
0.04
Diastolic annular tissue velocities
E/E= ratio‡
⬍0.01
0.07
PAWP estimate (from E/E=), mm Hg
12 ⫾ 2.7
20 ⫾ 6.5*†
1.6 ⫾ 5.6†
⬍0.01
Diastolic dysfunction grade‡, 0–3
0.6 ⫾ 0.9
1.4 ⫾ 0.9†
1.1 ⫾ 0.9†
⬍0.01
Maximal‡, ml
50 ⫾ 13
84 ⫾ 26*†
60 ⫾ 16
⬍0.01
Diastasis, ml
32 ⫾ 11
61 ⫾ 23*†
39 ⫾ 14
⬍0.01
Minimal, ml
19 ⫾ 8
42 ⫾ 18*†
24 ⫾ 10
⬍0.01
Total‡, %
63 ⫾ 12
51 ⫾ 11*†
61 ⫾ 10
⬍0.01
Passive, %
36 ⫾ 11
28 ⫾ 12*†
34 ⫾ 11
⬍0.01
Active, %
42 ⫾ 13
32 ⫾ 11*†
40 ⫾ 12
⬍0.01
6,988 ⫾ 2,974*†
3,516 ⫾ 1,367†
⬍0.01
LA volume
LA emptying fraction
Maximal LA volume ⫻ LV mass/height2.7, ml·g/m2.7
1,850 ⫾ 862
*p ⬍ 0.05 versus HLVH group; †p ⬍ 0.05 versus control group; Bonferroni post hoc test; ‡variables used in discriminant analysis. Annular tissue velocities were averaged from the septal and lateral parts
of mitral annulus.
A= ⫽ late diastolic tissue velocity; E= ⫽ early diastolic tissue velocity; LA ⫽ left atrial; PAWP ⫽ pulmonary artery wedge pressure; other abbreviations as in Table 1.
dysfunction (grades I and II), this also applied to 67.5% of
HLVH subjects (p ⬎ 0.4). Advanced dysfunction (grade
III) was rare in both groups (p ⫽ 0.26), and no dysfunction
occurred in 16.2% of HFpEF versus 30% of HLVH (p ⫽
0.15). Even 35.7% of the control subjects displayed some
dysfunction (p ⬍ 0.001). Remaining subjects were not
classifiable by this scheme.
Atrial volumes and function. Left atrial size and function
were similar in the HLVH and control subjects but abnormal in the HFpEF subjects (Table 4). The LAVmax was
68% higher in HFpEF than in control subjects and 40%
Figure 1
The Prevalence of DD Grades in Each Subject Group
Patients unclassifiable by the scheme are noted as indeterminate. There was
substantial overlap in the prevalence of diastolic dysfunction (DD) in symptomatic heart failure with preserved ejection fraction (HFpEF) and asymptomatic
hypertensive left ventricular hypertrophy (HLVH) subjects.
greater than in HLVH (both p ⬍ 0.0001). These disparities
were preserved after indexing to BSA or by multivariate
adjustment to BSA, LV mass, and estimated PAWP (p ⬍
0.007) (Fig. 2). Therefore, differences in LA volume were
not related to resting high filling pressures, altered ventricular geometry, or differences in anthropometry, although
they could still have related to differential elevation of LA
pressures during exercise. The LA volume before and after
atrial contraction showed similar disparities. Atrial total, passive, and active emptying fraction was reduced in HFpEF
versus control or HLVH subjects. The LAVmax was best
predicted by the combination of LVMI, BSA, E/E=, and
age (z-standardized ␤ coefficients 0.34, 0.26, 0.24, and 0.15,
respectively, by multiple regression with backward elimination; final r2 ⫽ 0.41).
The combination of biatrial and right ventricular enlargement, ventricular hypertrophy, and normal LV cavity size
led to greater total heart (epicardial) volumes (Fig. 3, see the
Appendix for supplemental Table C). This is potentially
important, because this volume relates to pericardial constraint and right-left heart interaction and could contribute
to higher LV diastolic pressures.
Atrial reserve was assessed during sustained isometric
handgrip in HFpEF and HLVH groups. The increases
(percentage change from baseline) of heart rate (8% vs.
11% for HFpEF and HLVH, respectively; p ⫽ 0.2),
cardiac output (12% vs. 15%; p ⫽ 0.8), and total arterial
resistance (12% vs. 15%, p ⫽ 0.9) were similar. However,
arterial pressure response was diminished in HFpEF
(systolic: 18% vs. 25%, p ⫽ 0.02; diastolic: 22% vs. 33%,
Melenovsky et al.
Pathophysiology of Heart Failure With Normal EF
JACC Vol. 49, No. 2, 2007
January 16, 2007:198–207
Figure 2
203
Comparison of LA Volumes and Emptying Fractions Adjusted for LV Mass Body
Surface Area and Estimated Pulmonary Artery Wedge Pressure for Each Subject Group
Plots show the adjusted mean and 95% confidence interval. HFpEF subjects had significantly greater atrial size and reduced function. *p ⬍ 0.05 HFpEF versus LVH.
LA ⫽ left atrial; LV ⫽ left ventricular; other abbreviations as in Figure 1.
p ⫽ 0.02). Handgrip exercise did not alter LV EF,
end-diastolic volume, Ees, or systolic and early-diastolic
tissue velocities. However, late-diastolic annular tissue
velocity (A=) rose in HLVH subjects but was unchanged
in HFpEF (⌬ 34.9% vs. ⌬ 5.1%; p ⫽ 0.004) (Fig. 4),
supporting reduced active atrial contractile reserve in the
HF group.
What parameters best differentiate HFpEF from HLVH
subjects? Discriminate function analysis was performed to
determine which parameter set best separated HFpEF from
HLVH subjects. Thirty-nine variables (indicated by ‡ in
tables, all reflecting parameters with significant differences
between groups but avoiding colinearity) were entered, and
8 remained (misclassification error rate 0.145). These were
(in order of influence): LVMI, history of CAD, LAVmax,
relative wall thickness, diastolic blood pressure, history of
CAD, total LA emptying fraction , Ees, and ACEI therapy.
Interestingly, no resting vascular or conventional LV diastolic or systolic parameter (including diastolic dysfunction
grade and tissue Doppler-derived indices) remained. Analysis of the remaining variables by CART-derived binary
classification yielded LVMI (⬎71 g/m2.7) as the major
predictor for HFpEF. A smaller mass but history of CAD
and LAVmax ⬎83 ml also predicted HFpEF, with a
misclassification rate of 0.148.
The discriminatory capacity of individual parameters was
further examined by histograms and receiver-operating
curves (ROC). Distributions for E/E= ratio and diastolic
dysfunction grade (Fig. 5A) revealed substantial overlap
Figure 4
Figure 3
Total Epicardial Versus Left Ventricular
End-Diastolic Chamber Volumes
Increase in total (epicardial) heart volume with similar end-diastolic volume (EDV)
in HFpEF versus HLVH, and both versus control subjects. *p ⬍ 0.01 versus HLVH;
†p ⬍ 0.001 versus control subjects. Abbreviations as in Figure 1.
Reduced Atrial Systolic Reserve in HFpEF Patients
The effect of isometric handgrip exercise on late diastolic (A=) annular tissue
velocity in patients with HFpEF (n ⫽ 21) and asymptomatic HLVH subjects (n ⫽
25). Unlike HLVH subjects, HFpEF patients had minimal increases in A= with
handgrip-exercise. Data shown are mean ⫾ SEM. *Between group ANOVA.
Abbreviations as in Figure 1.
204
Figure 5
Melenovsky et al.
Pathophysiology of Heart Failure With Normal EF
JACC Vol. 49, No. 2, 2007
January 16, 2007:198–207
LV Mass and Atrial Volumes Better Distinguish HFpEF From HLVH Subjects Than Indexes of Diastolic Dysfunction
Histograms for group distributions of E/E= ratio and diastolic dysfunction (DD) grade (A) versus LVMI and LAVmax (B). The latter show better separation between HFpEF
and HLVH subjects. (C) LA enlargement was weakly correlated to LVMI, particularly in the HFpEF subjects. (D) Product of LVMI and LAVmax provides the best separation
among study groups. (E) Receiver-operating curves for discriminating between HFpEF and HLVH. The most optimal characteristic was found for LVMI ⫻ LAVmax (AUC ⫽
0.85), followed by LVMI (AUC ⫽ 0.79) and LAVmax (AUC ⫽ 0.77). E/E= ratio (AUC ⫽ 0.69) and DD grade (AUC ⫽ 0.68) were poorer discriminators. The optimal cut-off
value (4,418 ml ⫻ g/m2.7) of LVMI ⫻ LAVmax (arrow) separated the groups with a sensitivity 0.838 and specificity 0.825. AUC ⫽ area under the curve; E ⫽ early diastolic tissue velocity; E= ⫽ longitudinal early tissue velocity; LAV ⫽ left atrial volume; LVMI ⫽ left ventricular mass index; other abbreviations as in Figures 1 and 2.
(particularly for HFpEF and HLVH), whereas those for
LVMI and LAVmax (Fig. 5B) had greater separation. The
LVMI and LAVmax were only weakly correlated (Fig 5C);
therefore, their product (LVMI ⫻ LA-Volmax)—a marker
of combined LV and LA morphologic change—provided
the best discrimination (Figs. 5D and 5E) with an area
under the ROC of 0.85. An optimal cut-off value of
4,418 ml·g/m2.7 separated both groups, with a sensitivity
of 83.8% and a specificity of 82.5%.
Discussion
The present study provides the first comprehensive comparison between cardiovascular function in HFpEF subjects
and individuals with hypertension and LVH without symptoms, diagnosis, or treatment history of HF. We further
matched age, gender, and ethnicity and included a third
group without HLVH. Multiple ventricular systolic, diastolic, and vascular abnormalities were observed in HFpEF
JACC Vol. 49, No. 2, 2007
January 16, 2007:198–207
patients compared with normotensive non-LVH control
subjects, as previously reported (14,15,17,18). However,
most of these changes were similarly observed in HLVH
subjects, suggesting they were primarily related to coexisting
clinical features. Between the 2 groups, LVH magnitude
and atrial dilation/failure differed most, providing the best
separation based on regression modeling and ROC analysis.
The results suggest that atrial function may be a key
compensatory mechanism countering evolution of HFpEF,
and they highlight the diagnostic utility of atrial failure as a
marker of the disease.
The term “heart failure with a preserved ejection fraction”
is descriptive of a syndrome involving multiple cardiovascular abnormalities that may each contribute to symptoms.
Vascular and ventricular stiffening exacerbate blood pressure
fluctuations, promote ventricular hypertrophic response,
increase cardiac systolic load particularly during stress, and
compromise systolic reserve (18,38). Diastolic abnormalities
lead to higher filling pressures to achieve adequate filling,
contributing to the redistribution of blood to the central
thorax. Increased heart volume and exaggerated epicardial
constraint may also contribute (18). Finally, atrial dilation
and dysfunction likely represent loss of an important compensation which becomes increasingly important given the
other abnormalities. By comparing HFpEF with HLVH,
we were better able to resolve what may be the “last straw”
that finally separates preclinical hypertensive heart disease
from HFpEF. Our data do not refute results of earlier work
designed to elucidate the pathophysiology of HFpEF (i.e.,
contribution of individual factors). Rather, they show that
when comparing preclinical HLVH with HFpEF, the
amount of concentric LV hypertrophy, LA remodeling, and
epicardial volume enlargement are particularly abnormal.
The present results are in concert with earlier studies that
compared HFpEF subjects with normotensive and/or nonhypertrophied control subjects (1,9,13–18,20,39). However,
the fact that many similar abnormalities were observed in
subjects with HLVH without HF makes it more difficult to
assign primary causality to any one change. Advanced age,
race, hypertension, gender, LVH, and obesity each significantly influences vascular and ventricular function
(8,38,40,41), yet their presence does not guarantee HF
symptoms. The present HFpEF patients were predominantly African American, reflecting our inner-city referral
population, but this is an important group that often has
been overlooked. African Americans have a high prevalence
of hypertension and a higher propensity to develop LVH
(7), display an earlier presentation of HFpEF (5), and may
have a worse prognosis with this disease (42).
Ventricular function in HFpEF. Rest contractility appeared normal in HFpEF, consistent with recent data (12),
although some systolic parameters were altered. Systolic
tissue velocity was lower, as previously reported (43), but
was similarly reduced in HLVH, likely reflecting concentric
LVH. Also as previously reported, HFpEF subjects had
greater end-systolic stiffening (Ees) than non-LVH control
Melenovsky et al.
Pathophysiology of Heart Failure With Normal EF
205
subjects (18), yet this too was similarly observed in HLVH.
It remains unclear if contractility reserve is more limited in
HFpEF, as some studies have suggested (44).
Both E/E= and IVRT differed between HFpEF and
HLVH. As with the higher E/E=, a shorter IVRT could
reflect elevated diastolic pressures in HFpEF, although the
lack of correlation between variables may suggest another
mechanism. Higher diastolic pressures can also be due to
increased LV filling (39), although this was not observed in
the present study, or to reduced chamber compliance (14),
although E-wave deceleration time, which estimates diastolic compliance (45), was similar between HFpEF and
HLVH groups. Another cause is exaggerated ventricular
interdependence from pericardial constraint, which can
elevate diastolic pressures even without diastolic ventricular
stiffening. This would be expected if total epicardial heart
volume increased (owing to atrial dilation and increased LV
mass), which we observed. In this setting, pericardial contributions to intracavitary LV pressure become more apparent, a finding supported by invasive analysis of the enddiastolic pressure-volume relations in HFpEF (18).
Approximately 35% of resting LV diastolic pressure in
humans is attributable to external forces from right heart
and pericardial loads (46). This does not imply pericardial
constriction but rather a greater pericardial contribution to
measured LV diastolic pressures.
Concentric LVH, atrial enlargement, and dysfunction.
Increased LA volumes are increasingly viewed as a marker
for diastolic dysfunction (33) and are an independent
predictor of coronary heart failure and worsened mortality
in previously asymptomatic elderly subjects with preserved
EF (47). Atrial dilation occurs with age and can contribute
to alterations in LA pressure and therefore reduced early
diastolic filling (48). In the present study, age was similar
across groups, yet further LA enlargement was observed in
HFpEF. Although this could be explained by the greater
LVH or higher diastolic pressures in this cohort, disparate
LA size and function persisted after adjusting for these
parameters (Fig. 2). There may be disproportionate elevation of diastolic pressures during exercise in HFpEF compared with HLVH subjects, which could underly chronic
atrial dilation/dysfunction and/or the blunted A= response
during handgrip. However, this remains speculative. Furthermore, ventricles exposed to chronic pressure overload do
not uniformly dilate or develop depressed function, and one
cannot presume that all atria do either. The HFpEF
subjects also had a greater prevalence of CAD, diabetes
(49), impaired renal function, and obesity (50), and these
factors can favor volume retention and potentially contribute to muscle damage in an overloaded atria.
The unique feature of atrial size (and function) was that
it was significantly altered only in HFpEF and, along with
LVMI, provided the most sensitive and specific parameters
to differentiate HFpEF from HLVH subjects. The product
of the 2 does not convey a single physical property but serves
as a simple way to combine both findings in a diagnostic
206
Melenovsky et al.
Pathophysiology of Heart Failure With Normal EF
index. This could also be done by a regression model. The
lack of specific easily employed diagnostic entry criteria has
been a barrier for clinical trials in HFpEF, and this index
may be useful in this regard and further help identify
preclinical patients at higher risk to develop incident HF.
Study limitations. Although we observed highly significant differences for many parameters in HLVH and
HFpEF versus control subjects, lack of differences between
the former 2 groups could relate to sample size (type II
error). However, the fact that we observed marked differences in LV mass, atrial size, and function underscores the
robustness of the comparison. Furthermore, the sample size
was similar or greater than in most other detailed cardiovascular studies (12,14,15,18).
We used noninvasive measures of cardiovascular function,
and although these have limitations and can yield values that
depend on the specific methods used (e.g., pulsed versus
color-mode Doppler), differences between HFpEF and
control subjects were compatible with earlier invasive data
(12,14,18). We estimated PAWP by E/E=, which may be
less accurate in subjects without LV dysfunction (32) but
has been found to be reliable in HFpEF (31). The HFpEF
group was 76% African American, reflecting our urban
inner-city referral population. This may limit its applicability to the broader population of HFpEF, because the degree
of ventricular remodeling is affected by race (7). Nevertheless, many of the HFpEF results are similar to those
reported in a cohort that was 30% African American (5).
Coronary artery disease was not directly assessed in most
subjects, and its role in HFpEF potentially was overestimated, because HLVH subjects were less often tested for
CAD, and the incidence of oligosymptomatic CAD was
probably higher than their records or history conveyed (51).
Finally, given the modest sample size, there is potential
instability in the pooled statistical analysis, and some other
features might prove more discriminatory in a larger data set.
Conclusions. Although multiple indexes of LV systolic,
diastolic, and vascular function are abnormal in HFpEF,
HLVH subjects of similar age, gender, ethnicity without
HF share many of these abnormalities. Quantitative differences in diastolic function exist between the groups, but
with substantial overlap they appear difficult to use as a
robust distinguishing feature. The presence of more marked
concentric hypertrophic remodeling of LV and left atrial
dysfunction/remodeling may provide more specific positive
diagnostic identifiers for HFpEF subjects and help assess
prognosis. The fact that these differences were present when
comparing HFpEF and HLVH suggests that excessive
LVH, atrial dysfunction, and epicardial volume enlargement may represent important decompensations in the
pathogenesis of HFpEF. It would be intriguing if reversal of
these changes converted symptomatic HFpEF patients back
to more compensated HLVH subjects, but this remains to
be determined.
JACC Vol. 49, No. 2, 2007
January 16, 2007:198–207
Acknowledgments
The authors greatly thank Tricia Marhin, RN, Patricia
Fizgerald, RN, Laura Shively, RN, Beverly Gorman, Ela
Chamera, and Kristy Kessler, RN, for their excellent work
on the project.
Reprint requests and correspondence: Dr. David A. Kass,
Ross 835, Division of Medicine, Johns Hopkins Medical Institutions, 720 Rutland Avenue, Baltimore, Maryland 21205. E-mail:
[email protected].
REFERENCES
1. Owan TE, Hodge DO, Herges RM, Jacobsen SJ, Rogers VL, Redfield
MM. Trends in prevalence and outcome of heart failure with preserved
ejection fraction. N Engl J Med 2006;355:251–9.
2. Kitzman DW. Diastolic heart failure in the elderly. Heart Fail Rev
2002;7:17–27.
3. Bursi F, Weston SA, Redfield MM, et al. Systolic and diastolic heart
failure in the community. JAMA 2006;296:2209 –16.
4. Franklin KM, Aurigemma GP. Prognosis in diastolic heart failure.
Prog Cardiovasc Dis 2005;47:333–9.
5. Klapholz M, Maurer M, Lowe AM, et al. Hospitalization for heart
failure in the presence of a normal left ventricular ejection fraction:
results of the New York Heart Failure Registry. J Am Coll Cardiol
2004;43:1432– 8.
6. Masoudi FA, Havranek EP, Smith G, et al. Gender, age, and heart
failure with preserved left ventricular systolic function. J Am Coll
Cardiol 2003;41:217–23.
7. Kizer JR, Arnett DK, Bella JN, et al. Differences in left ventricular
structure between black and white hypertensive adults: the Hypertension Genetic Epidemiology Network study. Hypertension 2004;43:
1182– 8.
8. Lakatta EG, Levy D. Arterial and cardiac aging: major shareholders in
cardiovascular disease enterprises: part II: the aging heart in health:
links to heart disease. Circulation 2003;107:346 –54.
9. Chen HH, Lainchbury JG, Senni M, Bailey KR, Redfield MM.
Diastolic heart failure in the community: clinical profile, natural
history, therapy, and impact of proposed diagnostic criteria. J Card Fail
2002;8:279 – 87.
10. Thom T, Haase N, Rosamond W, et al. Heart disease and stroke
statistics—2006 update: a report from the American Heart Association
Statistics Committee and Stroke Statistics Subcommittee. Circulation
2006;113:e85–151.
11. Hunt SA, Abraham WT, Chin MH, et al. ACC/AHA 2005 guideline
update for the diagnosis and management of chronic heart failure in
the adult: a report of the American College of Cardiology/American
Heart Association Task Force on Practice Guidelines (Writing Committee to Update the 2001 Guidelines for the Evaluation and Management of Heart Failure): developed in collaboration with the
American College of Chest Physicians and the International Society
for Heart and Lung Transplantation: endorsed by the Heart Rhythm
Society. Circulation 2005;112:e154 –235.
12. Baicu CF, Zile MR, Aurigemma GP, Gaasch WH. Left ventricular
systolic performance, function, and contractility in patients with
diastolic heart failure. Circulation 2005;111:2306 –12.
13. Zile MR, Brutsaert DL. New concepts in diastolic dysfunction and
diastolic heart failure: part I: diagnosis, prognosis, and measurements
of diastolic function. Circulation 2002;105:1387–93.
14. Zile MR, Baicu CF, Gaasch WH. Diastolic heart failure—
abnormalities in active relaxation and passive stiffness of the left
ventricle. N Engl J Med 2004;350:1953–9.
15. Zile MR, Gaasch WH, Carroll JD, et al. Heart failure with a normal
ejection fraction: is measurement of diastolic function necessary to
make the diagnosis of diastolic heart failure? Circulation 2001;104:
779 – 82.
16. Kitzman DW, Little WC, Brubaker PH, et al. Pathophysiological
characterization of isolated diastolic heart failure in comparison to
systolic heart failure. JAMA 2002;288:2144 –50.
JACC Vol. 49, No. 2, 2007
January 16, 2007:198–207
17. Hundley WG, Kitzman DW, Morgan TM, et al. Cardiac cycledependent changes in aortic area and distensibility are reduced in older
patients with isolated diastolic heart failure and correlate with exercise
intolerance. J Am Coll Cardiol 2001;38:796 – 802.
18. Kawaguchi M, Hay I, Fetics B, Kass DA. Combined ventricular
systolic and arterial stiffening in patients with heart failure and
preserved ejection fraction: implications for systolic and diastolic
reserve limitations. Circulation 2003;107:714 –20.
19. Kass DA, Bronzwaer JG, Paulus WJ. What mechanisms underlie
diastolic dysfunction in heart failure? Circ Res 2004;94:1533– 42.
20. Redfield MM, Jacobsen SJ, Burnett JC Jr., Mahoney DW, Bailey KR,
Rodeheffer RJ. Burden of systolic and diastolic ventricular dysfunction
in the community: appreciating the scope of the heart failure epidemic.
JAMA 2003;289:194 –202.
21. Senni M, Tribouilloy CM, Rodeheffer RJ, et al. Congestive heart
failure in the community: a study of all incident cases in Olmsted
County, Minnesota, in 1991. Circulation 1998;98:2282–9.
22. Devereux RB, Roman MJ, Liu JE, et al. Congestive heart failure
despite normal left ventricular systolic function in a population-based
sample: the Strong Heart Study. Am J Cardiol 2000;86:1090 – 6.
23. Loeppky JA, Hoekenga DE, Greene ER, Luft UC. Comparison of
noninvasive pulsed Doppler and Fick measurements of stroke volume
in cardiac patients. Am Heart J 1984;107:339 – 46.
24. Hart JP, Cabreriza SE, Gallup CG, Hsu D, Spotnitzt HM. Validation
of left ventricular end-diastolic volume from stroke volume and
ejection fraction. ASAIO J 2002;48:654 –7.
25. Devereux RB, Alonso DR, Lutas EM, et al. Echocardiographic
assessment of left ventricular hypertrophy: comparison to necropsy
findings. Am J Cardiol 1986;57:450 – 8.
26. Palmieri V, de Simone G, Arnett DK, et al. Relation of various
degrees of body mass index in patients with systemic hypertension to
left ventricular mass, cardiac output, and peripheral resistance (the
Hypertension Genetic Epidemiology Network study). Am J Cardiol
2001;88:1163– 8.
27. Shimizu G, Hirota Y, Kita Y, Kawamura K, Saito T, Gaasch WH.
Left ventricular midwall mechanics in systemic arterial hypertension.
Myocardial function is depressed in pressure-overload hypertrophy.
Circulation 1991;83:1676 – 84.
28. Devereux RB, de Simone G, Pickering TG, Schwartz JE, Roman MJ.
Relation of left ventricular midwall function to cardiovascular risk factors
and arterial structure and function. Hypertension 1998;31:929 –36.
29. Chen CH, Fetics B, Nevo E, et al. Noninvasive single-beat determination of left ventricular end-systolic elastance in humans. J Am Coll
Cardiol 2001;38:2028 –34.
30. Nagueh SF, Middleton KJ, Kopelen HA, Zoghbi WA, Quinones MA.
Doppler tissue imaging: a noninvasive technique for evaluation of left
ventricular relaxation and estimation of filling pressures. J Am Coll
Cardiol 1997;30:1527–33.
31. Bruch C, Grude M, Muller J, Breithardt G, Wichter T. Usefulness of
tissue Doppler imaging for estimation of left ventricular filling pressures in patients with systolic and diastolic heart failure. Am J Cardiol
2005;95:892–5.
32. Rivas-Gotz C, Manolios M, Thohan V, Nagueh SF. Impact of left
ventricular ejection fraction on estimation of left ventricular filling
pressures using tissue Doppler and flow propagation velocity. Am J
Cardiol 2003;91:780 – 4.
33. Pritchett AM, Mahoney DW, Jacobsen SJ, Rodeheffer RJ, Karon BL,
Redfield MM. Diastolic dysfunction and left atrial volume: a
population-based study. J Am Coll Cardiol 2005;45:87–92.
34. Gottdiener JS, Bednarz J, Devereux R, et al. American Society of
Echocardiography recommendations for use of echocardiography in
clinical trials. J Am Soc Echocardiogr 2004;17:1086 –119.
Melenovsky et al.
Pathophysiology of Heart Failure With Normal EF
207
35. Chen CH, Ting CT, Nussbacher A, et al. Validation of carotid artery
tonometry as a means of estimating augmentation index of ascending
aortic pressure. Hypertension 1996;27:168 –75.
36. Chemla D, Hebert JL, Coirault C, et al. Total arterial compliance
estimated by stroke volume-to-aortic pulse pressure ratio in humans.
Am J Physiol 1998;274:H500 –5.
37. Kelly RP, Ting CT, Yang TM, et al. Effective arterial elastance as
index of arterial vascular load in humans. Circulation 1992;86:513–21.
38. Chen CH, Nakayama M, Nevo E, Fetics BJ, Maughan WL, Kass DA.
Coupled systolic-ventricular and vascular stiffening with age: implications for pressure regulation and cardiac reserve in the elderly. J Am
Coll Cardiol 1998;32:1221–7.
39. Maurer MS, King DL, El Khoury RL, Packer M, Burkhoff D. Left
heart failure with a normal ejection fraction: identification of different
pathophysiologic mechanisms. J Card Fail 2005;11:177– 87.
40. de Simone G, Kitzman DW, Palmieri V, et al. Association of
inappropriate left ventricular mass with systolic and diastolic dysfunction: the HyperGEN study. Am J Hypertens 2004;17:828 –33.
41. Redfield MM, Jacobsen SJ, Borlaug BA, Rodeheffer RJ, Kass DA.
Age- and gender-related ventricular-vascular stiffening: a community
based study. Circulation 2005;112:2254 – 62.
42. East MA, Peterson ED, Shaw LK, Gattis WA, O’Connor CM. Racial
differences in the outcomes of patients with diastolic heart failure. Am
Heart J 2004;148:151– 6.
43. Yu CM, Lin H, Yang H, Kong SL, Zhang Q, Lee SW. Progression
of systolic abnormalities in patients with “isolated” diastolic heart
failure and diastolic dysfunction. Circulation 2002;105:1195–201.
44. Liu CP, Ting CT, Lawrence W, Maughan WL, Chang MS, Kass DA.
Diminished contractile response to increased heart rate in intact
human left ventricular hypertrophy. Systolic versus diastolic determinants. Circulation 1993;88:1893–906.
45. Little WC, Ohno M, Kitzman DW, Thomas JD, Cheng CP.
Determination of left ventricular chamber stiffness from the time for
deceleration of early left ventricular filling. Circulation 1995;92:
1933–9.
46. Dauterman K, Pak PH, Maughan WL, et al. Contribution of external
forces to left ventricular diastolic pressure. Implications for the clinical
use of the Starling law. Ann Intern Med 1995;122:737– 42.
47. Takemoto Y, Barnes ME, Seward JB, et al. Usefulness of left atrial
volume in predicting first congestive heart failure in patients ⱖ65 years
of age with well-preserved left ventricular systolic function. Am J
Cardiol 2005;96:832– 6.
48. Hees PS, Fleg JL, Dong SJ, Shapiro EP. MRI and echocardiographic
assessment of the diastolic dysfunction of normal aging: altered LV
pressure decline or load? Am J Physiol Heart Circ Physiol 2004;286:
H782– 8.
49. Piccini JP, Klein L, Gheorghiade M, Bonow RO. New insights into
diastolic heart failure: role of diabetes mellitus. Am J Med 2004;116
Suppl 5A:64S–75S.
50. Wang TJ, Parise H, Levy D, et al. Obesity and the risk of new-onset
atrial fibrillation. JAMA 2004;292:2471–7.
51. Liao Y, Cooper RS, Ghali JK, Szocka A. Survival rates with coronary
artery disease for black women compared with black men. JAMA
1992;268:1867–71.
APPENDIX
For supplemental Tables A, B, and C, please see the online
version of this article.