JOURNAL OF INSURANCE MEDICINE
Copyright ᮊ 2004 Journal of Insurance Medicine
J Insur Med 2004;36:292–297
Diastolic Dysfunction REVIEW
Gordon A. Ewy, MD Address: Professor and Chief of
Cardiology, University of Arizona
Diastolic dysfunction of the heart is characterized by normal left College of Medicine, 1501 N. Camp-
ventricular contractility and normal ejection fraction, however ven- bell Avenue, Room 6404, Tucson,
tricular relaxation is impaired. In systolic dysfunction, ventricular AZ 85724; ph: 520-626-6332; fax:
contractility and ejection fraction are reduced, in addition to im- 520-626-6332; e-mail:
paired relaxation. The prevalence of diastolic dysfunction is in- [email protected]
creased in the elderly, especially those who have had inadequately Correspondent: Gordon A. Ewy,
treated hypertension. Both diastolic and systolic dysfunction may MD.
result in similar clinical signs and symptoms. Therefore, echocar- Keywords: Diastolic dysfunction,
diography is needed to make the distinction. Left atrial (LA) en- systolic dysfunction, echocardiogra-
largement, assessed by left atrial volume indexed to body surface phy, heart failure, hypertension, el-
area, appears to be the best measure to assess diastolic function. LA derly, presbycardia, mortality.
enlargement is likely when indexed LA volume is Ն34–40 mL/m2. Received: March 18, 2004
B-type natriuretic protein appears to be useful for the diagnosis, Accepted: April 2, 2004
assessment and prognosis of heart failure, but it does not distin-
guish between the two types of dysfunctions. Several drug treat-
ments that have effects on the mechanism of diastolic dysfunction
are under investigation.
D iastolic dysfunction is characterized by Heart failure is rarely purely systolic or di-
normal left ventricular contractility and astolic. Usually both are present with one
normal ejection fraction. In diastolic dysfunc- type of ‘‘dysfunction’’ predominating over
tion, the left ventricle contracts normally but the other. An exception is a condition called
doesn’t relax well. Diastolic dysfunction is ‘‘stunning’’ of the heart, which occurs after
more prevalent with increasing age and is blood supply is restored to the heart follow-
usually found in older people with high ing an incident such as resuscitation from car-
blood pressure. This is especially true if the diac arrest or ventricular ﬁbrillation. In stun-
hypertension has been inadequately treated. ning, there is equal systolic and diastolic dys-
function with reduced ejection fraction and
In systolic dysfunction (volume overload), increased ﬁlling pressures. Stunning, and the
the left ventricle is dilated and does not con- associated functional measures, often im-
tract well. The ventricular ﬁlling pressure can proves in 2–3 days.
be normal or elevated. The left ventricle may
not relax well leading to pulmonary conges- The presenting symptoms and signs of di-
tion because high pressure is needed to ﬁll astolic and systolic heart failure are similar.
the ventricle. Decreased systemic and renal Clinicians believe that a physical exam should
perfusion lead to sodium and ﬂuid retention, be sufﬁcient to determine which is present.
as well as other signs and symptoms of con- The Table summarizes the prevalence of
gestive heart failure. symptoms and signs in diastolic and systolic
Table. Clinical Diagnosis of Diastolic Heart Failure. Relaxation actually starts in late systole. If re-
Prevalence of Symptoms and Signs of Heart Failure in laxation is abnormal, it results in problems
Diastolic and Systolic Heart Failure1,2 with ventricular ﬁlling and a variety of echo-
cardiographic ﬁndings. These ﬁndings may
Symptom or Sign Diastolic Systolic be described as the following: prolonged de-
celeration times, slowed relaxation, pseudon-
LVEF Ͼ50% Ͻ50% ormal pattern, delayed relaxation, restrictive
Dyspnea on exertion 85% 96% pattern, and constrictive pattern. These echo-
Paroxysmal nocturnal cardiogram report terms are as confusing to
55% 50% the average physician involved in patient care
dyspnea 60% 73% as they are to an insurance medical director.
Orthopnea 35% 46%
Jugular venous distention 72% 70% Tsang et al from the Mayo Clinic devel-
Rales 72% 70% oped a simple and useful index of diastolic
Displaced LV apex 45% 65% dysfunction3 saying that, ‘‘left atrial volume
S3 45% 66% is to diastolic dysfunction, as the ejection
S4 30% 40% fraction is to systolic dysfunction.’’ Upon
Edema 90% 96% careful consideration, this statement makes a
Cardiomegaly lot of sense. With diastolic dysfunction, if the
ventricle won’t relax or if there is scarring,
heart failure. Since there is considerable over- pressure has to increase. If the pressure in the
lap in symptoms and signs derived from his- ventricle has to increase in diastole when the
tory and examination, echocardiography is mitral valve is open, the pressure in the ven-
needed to distinguish between the two types tricle and the atrium is the same. So with di-
of heart failure. astolic dysfunction, the atrium will have to
enlarge. Tsang et al studied 140 adult pa-
The major problem in any discussion of di- tients, 70 men, mean age 58 Ϯ 19 years (range
astolic heart failure is that there is no simple 18–90), referred for standard echocardio-
deﬁnition. If a patient’s ejection fraction is grams for a variety of clinical indications but
stated to be 50%, 30% or 10%, the degree of with no valvular heart disease or atrial ar-
systolic dysfunction is understood. Currently, rhythmias. Standardizing left atrial (LA) vol-
there is no simple deﬁnition of diastolic dys- ume to body surface area, these investigators
function that relies on a similar number. A found that indexed LA volume (mL/m2) was
deﬁnition that has been proposed states that strongly associated with the grade of diastolic
diastolic heart failure is a clinical syndrome dysfunction. From this, an echocardiographic
characterized by signs and symptoms of grade of diastolic dysfunction was developed.
heart failure, normal systolic function (eg, Diastolic function was graded as normal, ab-
preserved ejection fraction) and abnormal di- normal relaxation, pseudonormal or restric-
astolic function. The question remains as to tive. The mean indexed LA volume for the
how to assess the presence of abnormal dia- subgroup of patients without detectable dia-
stolic function. stolic abnormalities (n ϭ 44) was 22 Ϯ 5 mL/
m2, and there was a graded relation between
Echocardiographers have created a science indexed LA volume and severity of LV dia-
of looking at abnormal diastolic dysfunction stolic dysfunction (Figure), which could be
calling it ‘‘diastology.’’ This term refers to the expressed as a linear regression equation.
characterization of left ventricular relaxation Correspondingly, indexed LA volume Ն32
and ﬁlling dynamics and their integration mL/m2 was 100% speciﬁc (67% sensitive) for
into clinical practice. Simplistically, systole is the detection of abnormal diastolic function.
thought to be the interval starting when the In multivariate regression, a cardiovascular
ventricle begins contracting up until aortic risk score modiﬁed from the scoring method
valve closure. Diastole was taught to be the
interval between aortic valve closure, and
when the ventricle begins its next contraction.
JOURNAL OF INSURANCE MEDICINE
Figure. Relation between indexed LA volume and diastolic looks just like a normal pattern. Because there
function grade. Reprinted from Tsang et al.3 Copyright 2002 is a fairly high incidence of some ventricular
with permission from Excerpta Medica, Inc. stiffening, rapid deceleration times may be
interpreted as mild diastolic dysfunction. The
of the Framingham Heart Study, history of left atrial volume index currently reported is
vascular disease, congestive heart failure just atrial volume based on diameter. LA di-
(CHF), transient ischemic attack (TIA) or ameter must be included in the 3-plane vol-
stroke, and smoking history were indepen- ume formula and then corrected for body
dently related to the indexed LA volume. Un- surface area. Hopefully, when left atrial vol-
fortunately, there are no data as yet that relate ume corrected for body surface area is more
indexed LA volume to mortality. widely reported, we will have a better handle
on the assessment of diastolic dysfunction
Because the left atrium has more of an egg during routine echocardiographic assess-
shape than a round shape, indexed LA vol- ment.
ume is a better measure than a single LA di-
mension, because volume calculation involves BNP AND DIASTOLIC HEART FAILURE
measures in 3 different planes. The measure-
ment of LA volume is somewhat complicated, Another measure that relates to prognostic
but with the newer, more sophisticated echo signiﬁcance of patients with heart failure is a
equipment, the LA image can be outlined, protein called B-type natriuretic protein
and the volume is calculated automatically. (BNP), which is produced by the left ventri-
The normal LA volume is 22 Ϯ 6 mL/m2, cle. Its name (B-type) is due to its original
which is all one needs to remember. If in- discovery in the brain tissue of swine, but in
dexed LA volume is greater than one stan- humans it is secreted by the left ventricle in
dard deviation (Ͼ28 mL/m2), there is mild response to stretch or pressure. BNP has a
LA enlargement. Greater than 2 standard de- very short half-life, 14 Ϯ 4 hours. BNP is used
viations (Ͼ34 mL/m2) is moderate LA en- to assess patients who present in the emer-
largement, and greater than 3 standard de- gency room with dyspnea. If measured BNP
viations (Ͼ40 mL/m2) is marked LA enlarge- is Ͻ100 pg/mL, heart failure can most likely
ment. In summary, normal is less than 28 be excluded. If BNP is Ͼ100 pg/mL heart
mL/m2, and LA enlargement is greater than failure is likely. If BNP is Ͼ1000 pg/mL, se-
34–40 mL/m2. These will probably be the val- vere heart failure is usually present. If BNP
ues that are used to determine diastolic dys- remains in the 500–700 pg/mL range (even
function, at least for the next few years. with intense inpatient treatment), those indi-
viduals are likely to be readmitted soon after
Many current routine clinical echocardio- discharge and have a bad prognosis. Patients
graphic exams are inadequate to measure di- who are under age 65 years and who have
astolic dysfunction. A pseudonormal pattern persistently elevated BNP (Ͼ1000 pg/mL)
may be candidates for heart transplantation
or artiﬁcial heart devices.
There is also atrial natriuretic peptide
(ANP), which is secreted by the atrium. How-
ever, measurement of this peptide has not
been shown to be as helpful as BNP. More
information about BNP in the assessment of
both systolic and diastolic dysfunction will
appear as research continues. Based on stud-
ies that are available now, BNP can be used
acutely for diagnosis, as well as chronically
to monitor effectiveness of therapy and to as- the black population disproportionately.14
sess prognosis. Women are affected more because they live
longer than men and are more likely to get
PREVALENCE scarring and ﬁbrosis of the ventricle. Blacks
are more affected because they have a higher
As many as one third of individuals pre- prevalence of hypertension. The estimated
senting with clinical heart failure have nor- annual mortality due to diastolic dysfunction
mal systolic function.4–9 The prevalence varies varies widely from 9%–28% depending on
widely with the population studied. Preva- the patient series.7,16 In the Framingham
lence is high in the elderly, particularly in pa- Study, the mortality rate in patients with di-
tients who have inadequately treated hyper- astolic heart failure was 4 times that of pa-
tension. It is known that hypertension has tients without heart failure, but it was half
been inadequately treated in the US popula- that of patients with systolic heart failure.7
tion for some time. First, it was thought that
diastolic blood pressure had more signiﬁ- PROGNOSIS AND MORTALITY
cance on the development of cardiovascular
diseases, because in younger patients the di- In the V-HeFT trial, which Cohn et al con-
astolic pressure seems to increase. Then sys- ducted almost a decade and a half ago, var-
tolic blood pressure received more attention. ious forms of heart failure treatment in mid-
For many years (perhaps decades), there was dle-aged men were studied.16 They found that
the misconception that normal blood pres- a few patients in that study had heart failure
sure was 100 plus age. So for an 80 year old, but normal systolic function—therefore dia-
a blood pressure of 180 was acceptable. More stolic function. In these patients, they found
recently it has been learned that pulse pres- that the annual mortality of diastolic dys-
sure gives information on the stiffness of the function (8%) was half the annual mortality
arteries and has the greatest prognostic value. of those with systolic dysfunction (19%).
This suggests that an individual with blood
pressure of 160/60 has a much worse prog- The Olmstead County Project (Mayo Clin-
nosis than an individual with blood pressure ic) results question these differences, since
of 160/110, which is thought provoking. they showed that both systolic and diastolic
heart failure have similar prognosis, at least
Many older people have inadequately treat- for the ﬁrst 3–4 years of follow-up.6 Others
ed hypertension, so they are at high risk for have raised similar doubts.17 In a review of
diastolic dysfunction. Renal clinic or dialysis available studies from the interval 1983–2001,
patients have a high prevalence of diastolic if the study population of patients with dia-
function. On the other hand for patients in a stolic dysfunction was composed of elderly
coronary care unit or in a busy transplant patients, predominantly women (characteris-
evaluation service like the University of Ari- tic of the majority of diastolic heart failure
zona, the prevalence of diastolic dysfunction patients), then the natural history of diastolic
might be very low because these patients pre- heart failure may not be different from that
dominately are presenting for treatment of observed in patients with CHF secondary to
the consequences of systolic dysfunction. The decreased systolic function.
incidence of diastolic dysfunction is age-re-
lated. Heart failure due to diastolic dysfunc- On the other hand in younger individuals,
tion rises dramatically with age.10–12 The prev- the prognosis of diastolic failure may be bet-
alence of diastolic heart failure is 15% in ter than in secondary heart failure.17 In the
those less than age 50, 33% in those aged 50– elderly, the prognosis of heart failure second-
70, and greater than 50% in those over age ary to diastolic dysfunction may well be sim-
70.13 ilar to that of systolic failure. The difference
in the two may be related to ‘‘presbycardia,’’
Diastolic heart failure affects women and which will be discussed in more detail.
JOURNAL OF INSURANCE MEDICINE
Looking at diastolic heart failure alone, an diastolic dysfunction. Microscopically, these
age effect is seen with both mortality and ﬁbrous ‘‘ropes’’ can be seen as complex coils
morbidity. The 5-year mortality rate for pa- during contraction, becoming straight and
tients under age 50 is 15%. For patients aged parallel during relaxation limiting this phase.
50–70, it is 33%, and for patients over age 70,
5-year mortality is 50%. For morbidity, as as- Fibrosis, or scarring, is just a matter of in-
sessed by 1-year incidence of hospital admis- creased collagen in the heart. The normal col-
sion for CHF, the rate is 25%, 50%, and 50% lagen content is about 3%–5%, and more than
for ages Ͻ50, 50–70, and Ͼ70, respectively.13 that result in scarring. Special histologic
stains can demonstrate excess ﬁbrous tissue.
RISK FACTORS Fibrous tissue accumulation in the myocar-
dium is the major determinate of abnormal
The chief risk factors for diastolic heart fail- ventricular stiffness.18–20
ure are advanced age, hypertension, diabetes,
left ventricular hypertrophy, and coronary ar- Presbycardia is the ‘‘old heart.’’ Adult car-
tery disease. The mechanical factors which diac myocytes are terminally differentiated.21
lead to cardiac muscle thickening or scarring These cells begin as multipotent cells, but
includes hypertrophy, ﬁbrosis (collagen), once associated with the heart in embryologic
presbycardia, pericardial disease or restraint, development, they differentiate into heart
coronary turgor, and inﬁltration of the myo- cells and remain so. Therefore, they are re-
cardium. Functional factors that contribute to garded as ‘‘terminally differentiated.’’ One
the development of diastolic dysfunction are mechanism of aging is that cell loss occurs
tachycardia and arrhythmias, ischemia, de- via apoptosis, ‘‘programmed’’ cell death and
layed relaxation, altered relaxation patterns, by necrosis. When heart cells are lost, ‘‘new’’
and calcium overload (stone heart). heart cells are not regenerated. Compensato-
ry hypertrophy develops, therefore old hearts
When considering hypertrophy as a mech- are hypertrophied hearts. Presbycardia is the
anism, if a normal left ventricular (LV) wall loss of myocardial cells. If a heart attack oc-
is 1 cm thick, then in an individual with se- curs in a 40 year old, almost all of those pa-
vere left ventricular hypertrophy with a 2 cm tients will survive. If a heart attack occurs in
thick LV wall, the ventricle contracts well but an 80 year old, they are less likely to survive
has difﬁculty relaxing. because half of the heart cells have already
been lost, and those that remain are hyper-
One of the major mechanisms contributing trophied.
to diastolic dysfunction, particularly with ag-
ing, is ﬁbrosis. In the ventricular myocardi- Therefore, presbycardia is a hypertrophied
um, myocytes occupy two thirds of the vol- heart and a stiffer heart. With aging alone,
ume, and only account for one third of the hypertrophy and ﬁbrosis develop even in the
cells. On the other hand, the myocardial in- absence of hypertension. Half of the US pop-
terstitium, which serves to hold actin, myo- ulation over the age of 65 has hypertension.
sin, and other contractile elements together, The combination of age and hypertension on
occupies one third of the volume, but ac- the development of diastolic heart failure is
counts for two thirds of the cells. Myocytes dramatic. Prognosis in diastolic dysfunction
are made up a series of sarcomeres, bound does appear to be related to etiology. Unfor-
together into myoﬁbrils by a collagen weave tunately, detailed data that would establish
similar to tendons. If there is a lot of scarring, conclusive associations is not yet available.
these can be seen as little ‘‘ropes’’ that won’t
let the ventricle relax well. Since there is a lot TREATMENT
of muscle, there is no problem with contrac-
tion, so systolic function is adequate, but im- A number of drugs that can decrease car-
paired relaxation in diastole leads to isolated diac scar tissue and collagen content are be-
ing developed. It has been thought that a scar
was unmodiﬁable, but there are active build- of heart failure and left ventricular dysfunction in
ing and destructive processes within a scar. the general population; The Rotterdam Study. Eur
The rate and relative balance of these pro- Heart J. 1999;20:447–455.
cesses determine whether scar extent and size 9. Dauterman K, Massie B, Gheorghiade M. Heart
remains stable, increases or decreases. There failure associated with preserved systolic function:
are some data to suggest that the angiotensin a common and costly clinical entity. Am Heart J.
converting enzyme (ACE) inhibitors, the an- 1998;135:S310-S319.
giotensin receptor blockers (ARB), and aldo- 10. Tresch D, McGough M. Heart failure with normal
sterone blockers (eg, Aldactone) will affect systolic function: a common disorder in older peo-
the amount of scar tissue present. Further ple. J Am Geriatr Soc. 1995;43:1035–1042.
study of these possible actions is anticipated. 11. Benjamin E, Levy D, Anderson K, et al. Determi-
nants of Doppler indexes of left ventricular dia-
Based on a presentation to the American Academy of In- stolic function in normal subjects (the Framing-
surance Medicine in Scottsdale, Ariz, on October 15, 2003. ham Heart Study). Am J Cardiol. 1992;70:508–515.
12. Sagie A, Benjamin E, Galderisi M, et al. Reference
REFERENCES values for Doppler indexes of left ventricular dia-
stolic ﬁlling in the elderly. J Am Soc Echocardiogr.
1. McDermott MM, Feinglass J, Sy J, Gheorghiade M. 1993;6:570–576.
Hospitalized congestive heart failure patients with 13. Zile M, Brutsaert D. New Concepts in Diastolic
preserved versus abnormal left ventricular systolic Dysfunction and Diastolic Heart Failure: Part I: Di-
function: clinical characteristics and drug therapy. agnosis, Prognosis, and Measurements of Diastolic
Am J Med. 1995;99:629–635. Function. Circulation. 2002;105:1387–1393.
14. Topol E, Traill T, Fortuin N. Hypertensive hyper-
2. Echeverria H, Bilsker M, Myerburg R, Kessler K. trophic cardiomyopathy of the elderly. N Engl J
Congestive heart failure: echocardiographic in- Med. 1985;312:277–283.
sights. Am J Med. 1983;75:750–755. 15. Pernenkil R, Vinson J, Shah A, Beckham V, Witten-
berg C, Rich M. Course and prognosis in patients
3. Tsang T, Barnes M, Gersh B, Bailey K, Seward J. Ͼ or ϭ 70 years of age with congestive heart fail-
Left atrial volume as a morphophysiologic expres- ure and normal versus abnormal left ventricular
sion of left ventricular diastolic dysfunction and ejection fraction. Am J Cardiol. 1997;79:216–219.
relation to cardiovascular risk burden. Am J Car- 16. Cohn J. Treatment of patients with heart failure.
diol. 2002;90:1284–1289. Circulation. 1990;81:1440a–1441a.
17. Senni M, Redﬁeld MM. Heart failure with pre-
4. Dougherty A, Naccarelli G, Gray E, Hicks C, Gold- served systolic function: A different natural his-
stein R. Congestive heart failure with normal sys- tory? J Am Coll Cardiol. 2001;38:1277–1282.
tolic function. Am J Cardiol. 1984;54:778–782. 18. Weber K, Brilla C, Janicki J. Myocardial ﬁbrosis:
functional signiﬁcance and regulatory factors. Car-
5. Dodek A, Kassebaum D, Bristow J. Pulmonary diovasc Res. 1993;27:341–348.
edema in coronary-artery disease without cardio- 19. Sugihara N, Genda A, Shimizu M, et al. Diastolic
megaly. Paradox of the stiff heart. N Engl J Med. dysfunction and its relation to myocardial ﬁbrosis
1972;286:1347–1350. in essential hypertension. J Cardiol. 1988;18:353–
6. Senni M, Tribouilloy C, Rodeheffer R, et al. Con- 20. Ohsato K, Shimizu M, Sugihara N, Konishi K, Tak-
gestive Heart Failure in the Community: A Study eda R. Histopathological factors related to diastolic
of All Incident Cases in Olmsted County, Minne- function in myocardial hypertrophy. Jpn Circ J.
sota, in 1991. Circulation. 1998;98:2282–2289. 1992;56:325–333.
21. Katz A. The Cardiomyopathy of Overload. An Un-
7. Vasan R, Larson M, Benjamin E, Evans J, Reiss C, natural Growth Response in the Hypertrophied
Levy D. Congestive heart failure in subjects with Heart. Ann Intern Med. 1994;121:363–371.
normal versus reduced left ventricular ejection
fraction: Prevalence and mortality in a population-
based cohort. J Am Coll Cardiol. 1999;33:1948–1955.
8. Mosterd A, Hoes A, de Bruyne M, et al. Prevalence