The AgingElectrocardiogram JEFFREY JONES, MD, Z.M. SRODULSKI, STEVEN ROMISHER, MD
MD,
As the percentage of geriatric patients increases, objective methods of assessing the presence or absence and the severity of heart disease become more important. Frequently, elderly patients are either unable to provide accurate histories or have concomitant diseases (ie, emphysema) which exhibit symptoms overlapping those usually attributed to cardiovascular disease. Impairment of recent memory, diminished pain awareness, and a general decrease in physical activity may make typical cardiovascular symptoms more difficult to elicit than in younger patients. As a noninvasive laboratory procedure, the electrocardiogram (ECG) is a valuable aid in diagnosing and treating heart diseases in the elderly. However, the emergency physician often is confronted with a geriatric patient who has an abnormal ECG in the absence of other objective evidence of heart disease. The clinician may question whether the ECG is really a reliable index of heart disease in old people or whether aging has some influence on which findings are normal and which are abnormal. METHODOLOGlCALCONSlDEGATlONS Since 193 1 a number of conflicting studies have been published in an attempt to define the “normal” geriatric electrocardiogram. The bulk of this information has been obtained in cross-sectional studies. These studies obtain ECG recordings from different individuals in a population at a given time and may not necessarily represent true aging trends. Rather, such differences may be a result of cohort effects or selective survival.’ Only a longitudinal study of serial electrocardiograms in the same persons would be able to determine definitively which changes actually occur with age. Because most agerelated cardiovascular changes occur gradually over a number of years, relatively few such studies have been undertaken. The most important longitudinal study was that of 1,000 healthy aviation personnel (mean age 24 years) who were followed up for 24 years.* During this time serial ECGs were repeated at the mean ages of 36,42, and 48 years. In a similar study, Bachman et al’ followed 440 healthy men (aged 23 to 66 years) for 10 years as part of the Normative Aging Study. On the basis of data collected from both studies, certain cross-sectional age differences were confirmed. The analysis of electrocardiographic age trends is further From the Department of Emergency Medicine, Butterworth Hospital, Grand Rapids, Ml. Manuscript received May 24, 1989; revision accepted July 20, 1989. Address reprint requests to Dr Jones: Department of Emergency Medicine, Butterworth Hospital, 100 Michigan NE, Grand Rapids, MI 49503. Key Words: Electrocardiogram, geriatrics, arrhythmias, aging, coronary disease. 0 1990 by W.B. Saunders Company. 07358757/90/0803-0015$5.00/0 240
complicated because of interaction with other constitutional factors (eg, such as body weight, sex, race, chest contiguration, and functional variables such as blood pressure).3 The sample must be of sufficient size to compensate for these variables. Using the criterion of sample size alone, only a handful of previously reported studies may be considered to have adequate numbers.3-8 CARDIACCONDUCTIONSYSTEM A complete discussion of senile cardiomyopathy is beyond the scope of this article and has been reviewed in detail elsewhere.9~‘0 Suffice to say, with advancing age there is an increase in elastic and collagenous tissue in all parts of the conduction system. Although the resting heart rate falls little with age, there is a general decrease in the heart rate response to exercise. This decline in maximum heart rate is likely due to reduced adrenergic sensitivity in aging cells. In addition, fat accumulates around the sinoatrial (SA) node, at times producing a partial or complete separation of the node from the atrial musculature. In severe cases, this may be related to the development of sick sinus syndrome or, more commonly, sinus bradycardia. lo Most elderly people experience a decline in the number of pacemaker cells in the SA node; this change can be as dramatic as an 80% loss of nodal cells between the ages of 60 and 75 years.’ Although pronounced changes in the atrioventricular (AV) node have not been described, there are several age-related changes seen in the bundle of His. These include a loss of specialized muscle cells and an increase in fibrous and adipose tissue depositions, as well as amyloid intiltration.‘0 Also associated with aging is a variable degree of calcitication on the left side of the cardiac skeleton. Because of their proximity to skeletal structures, the AV node and proximal bundle branches may be damaged by this degenerative process, resulting in primary or idiopathic block. THE ABNORMAL ECG In view of the degenerative changes that occur in the cardiac conduction system with aging, it is not surprising that several features of the ECG are altered. These changes involve the duration of the PR and QT intervals, orientation of the electrical axis, duration and morphology of the atrial and ventricular complexes, and characteristics of the ventricular repolarization. Table 1 lists the frequency of “major” ECG abnormalities in patients 65 years of age and older.’ To accurately interpret any ECG, it is important to establish a reasonable baseline. The components of an ECG and their typical presentation in an aging patient follow. P wave P wave notching, slurring, and loss of amplitude are so common in the aged that they lose diagnostic significance
JONES, SRODULSKI, ROMISHER W THE AGING ELECTROCARDIOGRAM
TABLE1. Prevalence of ECGAbnormalliler In Subjects 66 Years or Older Entity
Percent
Abnormal ECG 1” AV block Left axis deviation (> -30”) Right axis deviation (3 + 120’) ST-T wave changes Left bundle branch block Right bundle branch block lntraventricular conduction delay Premature atrial systole Premature ventricular systole Atrial fibrillation
54 9 36 2 17 3 6 2 10 9 7
Data from references 7 and 22.
(Fig l).” In addition, the percentage of missing P waves increases. These alterations reflect loss of atrial myocardium, fibrosis, or cellular changes. In the limb leads, P waves may be so low as to be invisible. Unless the ECG is examined carefully, especially Vi, sinus rhythm may be overlooked. When P waves are not seen and the heart rhythm is regular, recording lead V, at double standardization may make the atrial depolarization apparent.” First-degree AV block The prevalence of first-degree AV block, defined as a PR interval >0.21 seconds, increases with age.i3 In a broad range of geriatric patients this prevalence is approximately lo%, but in nonagenarians the prevalence rises to 35%. A prolonged PR interval is not usually associated with clinical heart diseasei and when first-degree AV block is an isolated
FIGURE
1.
Electrocardiogram
241
finding, it does not affect prognosis.” However, these individuals may be more susceptible to digitalis toxicity. QRS Comptex The duration of the QRS interval tends to be greater in the aged because of the slight slowing of the conduction system, but it is never prolonged to 0.12 seconds in the absence of a bundle branch block.‘* Epidemiologic studies have shown, however, that the R and S wave amplitudes decrease with age, while the mean QRS vector rotates to the left, in the frontal plane, and posteriorly, in the horizontal plane.14 This gradual process, which parallels the effects of obesity, is apparent from early adult life and has been attributed to’ kyphoscoliosis with increasing AP diameter of the chest, lowering of the diaphragms due to emphysema, loss of the elasticity and increased resistivity of surrounding tissues, elongation of the aorta, and partial fibrosis in the anterior fasicle of left bundle branch. Abnormal left axis deviation (ALAD) is defined as a frontal plane axis of - 30” or greater. The prevalence of ALAD in the aged is eightfold greater than in the young and is noted twice as often in the presence of clinical heart disease. l5 However, patients with isolated ALAD have no significant increases in morbidity, mortality rate, or risk factors when compared with the general population. 13*16Right axis deviation is uncommon in older people but may be found in thin, long-chested persons with a vertical heart and emphysema.” Ventricular Hypertmphy Left ventricular hypertrophy (LVH) is one of the common ECG abnormalities in the elderly. The actual alence ranges from 1% to 40% depending on the criteria in diagnosis. l3 The Framingham study” indicates that
from a healthy 73syear-old male. Note the diffuse P wave notching.
most prevused elec-
242
AMERICAN JOURNAL OF EMERGENCY MEDICINE n Volume 8, Number 3 n May 1990
trocardiographic evidence of LVH increases in frequency with age and parallels the distribution of systemic hypertension. The data also suggest that LVH is a poor prognostic sign, even in the absence of clinical evidence of heart disease. The criteria for LVH and strain in younger people apply with a few exceptions in the elderly. Tall R waves in V, or V, may be due to the proximity of the left ventricle to the chest electrode and should not be interpreted as LVH unless the total sum of RV, and SV, exceeds 35 mm and RaVL is 11 mm or greater. i2 Several series have shown that all voltage criteria lack sensitivity and specificity in the diagnosis of LVH.‘,” In view of this, Campbell et al’* suggest that the diagnosis of LVH is more specific when voltage criteria are satisfied and when discordant ST depressions or T wave changes are present. Right ventricular hypertrophy (RVH) and car pulmonale are uncommon in the aged. One investigator found only 8 cases out of 2,254 elderly subjects.‘* However, this diagnosis may be difficult to make. A right axis deviation may be secondary to emphysema and the increase of R wave voltage in V, and of S wave in V, are often missing. Furthermore, some patterns which suggest RVH are found only in extreme right precordial leads (V,R to V,R), which are not recorded. Bundle branch block The overall prevalence of bundle branch block in young military populations is considerably less than 1%.19 In collected series of aged persons, Mihalick and Fisch’ found left anterior hemiblock (LAH) in 360/o,right bundle branch block in 6%, and left bundle branch block in 3% of subjects. In almost all series of elderly persons, right bundle branch block occurs slightly more often than left bundle branch block. The prognosis of bundle branch block is dependent on its etiology. In the elderly, bundle branch block is usually caused by fibrocalcific changes in the conduction system rather than ischemic heart disease.13 Isolated left bundle branch block has been shown to have a slightly higher incidence of heart disease and subsequent mortality than agematched controls.‘,” Right bundle branch block is considered a benign condition and only when associated with ALAD is it accompanied by an increased incidence of heart disease. Arrhythmias The prevalence and complexity of most arrhythmias are increased in older populations. Routine ECG recordings performed in the nonhospitalized elderly reveal an incidence of 10% in both atrial and ventricular premature beats.’ In comparison, a younger population studied by a survey of ECG records from over 60,ooO Air Force personnel disclosed a 0.6% prevalence of extrasystoles.” Ambulatory ECG recording provides a more sensitive method to establish the prevalence of arrhythmias. Twentyfour hour ECG monitoring in carefully screened healthy elderly have shown a substantial prevalence of isolated supraventricular (88%) and ventricular (80%) ectopic beats.20,2’ Supraventricular tachyarrhythmias were found in 30% of the elderly with 26% displaying more than 100 atrial ectopic beats over the 24-hour monitoring period. Ventricu-
lar couplets or short runs of ventricular tachycardia were detected in 15%. Nearly all arrhythmias were asymptomatic, regardless of type or complexity. The mere presence of premature systoles in an older patient, therefore, does not necessarily imply either the presence of disease or an indication for treatment. Atria1 fibrillation occurs in 2% to 8% of all elderly subjects.22 A higher frequency has been reported in hospitalized geriatric patients23 and in the very old, ie, nonagenarians.13 It is universally considered an abnormal ECG finding, the vast majority of subjects with atrial fibrillation have demonstrable heart disease.’ The hemodynamic consequences of atrial fibrillation are related to the ventricular response. If the ventricular rate is 60 to 90 beats/min, which allows adequate ventricular filling, then the rhythm is usually well tolerated, even in the aged. It is signitlcant that bradycardias (sinus bradycardia, sinus arrest, sinoatrial exit block, and second- or third-degree AV block) are rare in normal elderly subjects.‘3*20 As a result, these findings in symptomatic patients may indicate the need for corrective therapy and hospitalization. ST-T Wave The most commonly observed age-related ECG changes involve ventricular repolarization: the ST segment flattens and the T wave amplitude diminishes (Fig 2). ST depression is usually found in the left precordial leads, especially in V, and V, (Fig 3).i2 The amplitude of the T wave decreases slightly with advancing years, and low or notched T waves are not uncommon. Inverted T waves in leads I, II and precordial leads V, to V, indicate myocardial damage or ischemia. Deep symmetrical T-wave inversion in the absence of voltage criteria of LVH also indicates subendocardial ischemia. Several studies indicate that the incidence of nonspecific ST-T wave changes is nearly 17% in subjects over the age of 70.3*7s,“,13 In the elderly, the multiple noncardiac causes of T wave inversions and ST segment depressions-obesity, hypoglycemia, metabolic disorders, and aging changes in the chest-result in a loss of specificity and diagnostic value. If noncardiac causes can be excluded, the ST-T wave changes in the resting ECG have the same prognostic implications as in the middle-aged population. The QT interval usually increases with age but rarely is above the upper limit for younger people. I2 Myocardial Infarction Despite the decline in cardiovascular mortality during the past decade, acute myocardial infarction still claims over 600,000 lives annually, at least half of which occur in individuals aged 65 years or older.24 Published data indicate that over 50% of elderly subjects with ECG evidence of myocardial infarction failed to manifest clinically I3 and less than 20% of elderly patients with acute infarction present with classical symptoms.” Most symptoms were secondary to impaired myocardial function resulting in low cardiac output or congestive heart failure. In the elderly, about 50% of all patients with acute myocardial infarction will have new findings of ST elevation or Q waves in two or more leads upon presentation.” Approximately another 25% will have new ECG findings of ischemia
JONES, SRODULSKI, ROMISHER W THE AGING ELECTROCARDIOGRAM
FIGURE 2.
Note the nonspecific ST-T wave changes and leftward axis in this 71-year-old patient without clinical evidence of heart disease.
Only about 25% of all or strain without ST elevation.‘*” acute myocardial infarctions will present to the clinician with ECGs that do not show any new changes of ischemia, strain, or infarction.1*‘1*13 Unfortunately, in the elderly, reliance on the electrocardiogram alone would miss about 15% to 20% of all myocardial infarctions.26
FIGURE 3. diminishes).
243
In addition, an infarction pattern on ECG may be complicated by the presence of left bundle branch block, prior infarction, cardiac amyloidosis, emphysema, or chest deformities. Septal infarction is particularly difficult to diagnose due to the usual reduction of the initial vector of ventricular activation, as a result of senile infraseptal fibrosis. This in-
81-year-old woman with age-related changes involving ventricular repolarization
(ie, ST segment flattens and T wave amplitude
244
AMERICAN JOURNAL OF EMERGENCY MEDICINE m Volume 8, Number 3 m May 1990
volves the disappearance of the R wave in V,, V, as well as of the Q wave in a VL, V,, V,. Finally, ST segment and T wave abnormalities may be the only ECG signs of nontransmural infarction. However, their presence should be interpreted with caution in elderly subjects, since they may be associated with left ventricular hypertrophy, paced rhythm, and other nonacute conditions.24 CLINICALHEART DISEASE The incidence of abnormal ECG increases with age as does the incidence of clinical heart disease (Table 2). Mihalick and Fisch’ studied the correlation between ECG variables and clinical heart disease as manifested by angina, cardiomegaly, heart failure, prior infarction, and pronounced hypertension. Those ECG abnormalities that correlated strongly with heart disease were atria1 fibrillation, left bundle branch block, and nonspecific intraventricular conduction delay. There were no significant correlations with right bundle branch block, left anterior hemiblock, prolonged PR interval, frequent premature contractions, nonspecific ST-T wave changes, or ECG evidence of myocardial infarction.’ Although these ECG abnormalities did not correlate with the presence of clinical disease, strong evidence suggests that left anterior hemiblock, right bundle branch block, and myocardial infarction, reflect anatomic heart disease, the anatomic changes not severe enough to manifest clinically.‘5 For example, prevalence of marked left axis deviation is seen twice as often in the presence of overt heart disease. Even though the statistical correlation with clinical heart disease is not significant, it is reasonable to assume that left anterior hemiblock reflects septal disease not always manifested clinically. In a study of the ECG abnormalities seen in patients 85 or TABLE 2. Characteristics of the Aging Elsdmcafdiogram 1. The incidence of abnormal ECG increases with age and heart disease. 2 Probable insignificant ECG findings in the elderly include left axis deviation without evidence of LAH, lower wave amplitudes, longer intervals, and isolated premature contractions. 3. T wave inversions and ST segment depression have too many noncardiac causes to be specific. 4. Bradycardias (sinus bradycardia, sinus arrest, exit block, and second- or third-degree AV block) are rare in normal subjects. 5. The specific ECG abnormalities that correlate strongly with heart disease are atrial fibrillation, left bundle branch block and nonspecific intraventricular conduction delay. 8. Myocardial infarction patterns, left anterior hemiblock, and right bundle branch block do not correlate with presence of clinical disease, but may reflect anatomic disease. 7. With the exception of left ventricular hypertrophy, prognosis of a specific ECG abnormality remains that of the underlying disease. Reprinted with permission from Bosker G, et al: Assessment and detection of acute coronary ischemia, In Bosker G, et al (eds): Geriatric Emergency Medicine. St Louis, MO, Mosby, in press.
older, Rajala et al” have reported that ECG changes had a poor association with cardiac symptoms. ST segment depression, T wave inversion, premature ventricular contractions, and atrial fibrillation were related statistically to clinical congestive heart failure whereas T wave inversion was related to clinical coronary heart disease. High left R waves, ventricular premature complexes, and atria1 fibrillation showed a significant association with cardiac enlargement and pulmonary congestion. In their study of ECGs from persons with no overt heart disease, many ECG changes were noted, including frequent T wave inversion and frequent ST segment depression. They thought that specificity of ECG changes for heart diseases was not high in very old persons. Electrocardiographic patterns indicative of heart disease rarely show a correlation with either prognosis or severity of the disease. With the possible exception of left ventricular hypertrophy, the prognosis of a specific abnormality remains that of the underlying disease. SUMMARY With advancing age, widespread histologic changes in the conduction system occur. These changes may alter several features of the aging electrocardiogram, including duration of the PR and QT intervals, orientation of the electrical axis, duration and morphology of the atrial and ventricular complexes, and characteristics of the ventricular repolarization. And although ST segment and T wave abnormalities may be the only clue to acute ischemia, they are nonspecific and associated with a multitude of noncardiac causes. With an awareness of atypical presentations and difficulties in ECG interpretation, emergency physicians may be able to improve the assessment and triage of elderly patients with acute coronary ischemia. REFERENCES 1. Bachman S, Sparrow D, Smith LK: Effect of aging on the electrocardiogram. Am J Cardiol 1981;48:513-518 2. Harlan WR, Graybiel A, Mitchell RE, et al: Serial electrocardiograms: Their reliability and prognostic validity during a 24-yr period. J Chronic Dis 1967;20:853-887 3. Simonson E: The effect of age on the electrocardiogram. Am J Cardiol 1972;29:64-73 4. Willius EA: The heart in old age: A study of 700 patients seventv-five vears of aae and older. Am J Med Sci 1931:182:1-19 ._ 5. Fisch C, Genoveie PD, Dyke RW, et al: The electrocardiogram in persons over 70. Geriatrics 1957;12:816_62CJ 8. Taran LM, Szilagyi N: Electrocardiographic changes with advancing age. Geriatrics 1958;13:352-358 7. Mihalick MJ, Fisch C: Electrocardiographic findings in the aged. Am Heart J 1974;87:117-128 8. Olbrich 0, Woodford-Williams E: The normal precordial electrocardiogram in the aged. J Gerontol 1953;8:40-55 9. Fleg JL, Gerstenblith G, Lakatta EG: Pathphysiology of the aging heart and circulation, In Messerli F (ed): Cardiovascular Disease in the Elderly. Norwell, MA, Kluwer Academic, 1984; pp 11-21 10. Gerstenblith G, Weisfeldt ML, Lakatta EG, et al: Disorders of the heart, In Andres R, Bierman EL, Hazard WR (eds): Principles and Practice of Geriatric Medicine. New York, NY, McGraw-Hill, 1985, pp 520-522 11. Rodstein M: The ECG in old age: Implications for diagnosis, therapy, and prognosis. Geriatrics 1977;32:76-79 12. Harris R: Clinical Geriatric Cardiology. Philadelphia, PA, Lippincott, 1986; pp 68-93 13. Camm AJ, Ward DE: Clinical electrocardiography, In Mar-
JONES, SRODULSKI, ROMISHER n THE AGING ELECTROCARDIOGRAM
tin A, Camm AJ (eds): Heart Disease in the Elderly. New York, NY, Wiley, 1984; pp 149-186 14. Wagner NB, Schocken DD, Lee L, et al: Evaluation of the QRS complex on the standard 1Blead electrocardiogram in normal subjects 70 to 79 years of age. Am J Cardiol1988;62:982-985 15. Fisch C: The electrocardiogram in the aged, In Noble RJ, Rothbaum DA (edsl: Geriatric Cardioloav. PA. _. Philadelohia. . Davis, 1981; pp 65-74 16. Ostrander LD: Left axis deviation: Prevalence, associated conditions, and prognosis. An epidemiologic study. Ann Intern Med 1971;75:23-28 17. Kannel WB, Gordon T, Offutt D: Left ventricular hypertrophy by electrocardiogram-prevalence, incidence, and mortality in the Framinaham studv. Ann intern Med 1969:71:89-105 18. Campb& A, Caird-FI, Jackson TFM: Prevalence of abnormalities of electrocardiogram in old people. Br Heart J 1974; 36:1005-1011 19. Johnson RL, Averill KH, Lamb LE: Electrocardiographic findings in 63,375 asymptomatic individuals. Am J Cardiol 1960;6:143-167 20. Fleg JL, Kennedy HL: Cardiac arrhythmias in a healthy
245
elderly population: Detection by 24-hour ambulatory electrocardiography. Chest 1982;81:302-397 21. Anderson S, Osterlind PO, Holmboe G, et al: Twentyfour-hour electrocardiography in a healthy elderly population. Gerontology 1988;34:139-144 22. Coodley E, Coodley G: Electrocardiographic changes associated with aging, In Coodley E (ed): Geriatric Heart Disease. Boston, MA, PSG, 1985, pp 182-187 23. Mihalick MJ, Fisch C: Should ECG criteria be modified for geriatric patients? Geriatrics 1977;32:65-72 24. Flag JL: Cardiovascular emergencies, In Wilson LB, Simson SP, Baxter CR (eds): Handbook of Geriatric Emergency Care. Baltimore, MD, University Park Press, 1984, pp 38-50 25. Pathy MS: Clinical presentation of myocardial infarction in the elderly. Br Heart J 1967;29:190-199 26. Presentation of myocardial infarction in the elderly. Lancet 1986;2:1077-1078 (editorial) 27. Rajala SA, Ulla KM, Geiger MKM: Electrocardiogram, clinical findings, and chest x-ray in persons aged 85 or older. Am J Cardiol 1985;55:1175-1178
MD,
As the percentage of geriatric patients increases, objective methods of assessing the presence or absence and the severity of heart disease become more important. Frequently, elderly patients are either unable to provide accurate histories or have concomitant diseases (ie, emphysema) which exhibit symptoms overlapping those usually attributed to cardiovascular disease. Impairment of recent memory, diminished pain awareness, and a general decrease in physical activity may make typical cardiovascular symptoms more difficult to elicit than in younger patients. As a noninvasive laboratory procedure, the electrocardiogram (ECG) is a valuable aid in diagnosing and treating heart diseases in the elderly. However, the emergency physician often is confronted with a geriatric patient who has an abnormal ECG in the absence of other objective evidence of heart disease. The clinician may question whether the ECG is really a reliable index of heart disease in old people or whether aging has some influence on which findings are normal and which are abnormal. METHODOLOGlCALCONSlDEGATlONS Since 193 1 a number of conflicting studies have been published in an attempt to define the “normal” geriatric electrocardiogram. The bulk of this information has been obtained in cross-sectional studies. These studies obtain ECG recordings from different individuals in a population at a given time and may not necessarily represent true aging trends. Rather, such differences may be a result of cohort effects or selective survival.’ Only a longitudinal study of serial electrocardiograms in the same persons would be able to determine definitively which changes actually occur with age. Because most agerelated cardiovascular changes occur gradually over a number of years, relatively few such studies have been undertaken. The most important longitudinal study was that of 1,000 healthy aviation personnel (mean age 24 years) who were followed up for 24 years.* During this time serial ECGs were repeated at the mean ages of 36,42, and 48 years. In a similar study, Bachman et al’ followed 440 healthy men (aged 23 to 66 years) for 10 years as part of the Normative Aging Study. On the basis of data collected from both studies, certain cross-sectional age differences were confirmed. The analysis of electrocardiographic age trends is further From the Department of Emergency Medicine, Butterworth Hospital, Grand Rapids, Ml. Manuscript received May 24, 1989; revision accepted July 20, 1989. Address reprint requests to Dr Jones: Department of Emergency Medicine, Butterworth Hospital, 100 Michigan NE, Grand Rapids, MI 49503. Key Words: Electrocardiogram, geriatrics, arrhythmias, aging, coronary disease. 0 1990 by W.B. Saunders Company. 07358757/90/0803-0015$5.00/0 240
complicated because of interaction with other constitutional factors (eg, such as body weight, sex, race, chest contiguration, and functional variables such as blood pressure).3 The sample must be of sufficient size to compensate for these variables. Using the criterion of sample size alone, only a handful of previously reported studies may be considered to have adequate numbers.3-8 CARDIACCONDUCTIONSYSTEM A complete discussion of senile cardiomyopathy is beyond the scope of this article and has been reviewed in detail elsewhere.9~‘0 Suffice to say, with advancing age there is an increase in elastic and collagenous tissue in all parts of the conduction system. Although the resting heart rate falls little with age, there is a general decrease in the heart rate response to exercise. This decline in maximum heart rate is likely due to reduced adrenergic sensitivity in aging cells. In addition, fat accumulates around the sinoatrial (SA) node, at times producing a partial or complete separation of the node from the atrial musculature. In severe cases, this may be related to the development of sick sinus syndrome or, more commonly, sinus bradycardia. lo Most elderly people experience a decline in the number of pacemaker cells in the SA node; this change can be as dramatic as an 80% loss of nodal cells between the ages of 60 and 75 years.’ Although pronounced changes in the atrioventricular (AV) node have not been described, there are several age-related changes seen in the bundle of His. These include a loss of specialized muscle cells and an increase in fibrous and adipose tissue depositions, as well as amyloid intiltration.‘0 Also associated with aging is a variable degree of calcitication on the left side of the cardiac skeleton. Because of their proximity to skeletal structures, the AV node and proximal bundle branches may be damaged by this degenerative process, resulting in primary or idiopathic block. THE ABNORMAL ECG In view of the degenerative changes that occur in the cardiac conduction system with aging, it is not surprising that several features of the ECG are altered. These changes involve the duration of the PR and QT intervals, orientation of the electrical axis, duration and morphology of the atrial and ventricular complexes, and characteristics of the ventricular repolarization. Table 1 lists the frequency of “major” ECG abnormalities in patients 65 years of age and older.’ To accurately interpret any ECG, it is important to establish a reasonable baseline. The components of an ECG and their typical presentation in an aging patient follow. P wave P wave notching, slurring, and loss of amplitude are so common in the aged that they lose diagnostic significance
JONES, SRODULSKI, ROMISHER W THE AGING ELECTROCARDIOGRAM
TABLE1. Prevalence of ECGAbnormalliler In Subjects 66 Years or Older Entity
Percent
Abnormal ECG 1” AV block Left axis deviation (> -30”) Right axis deviation (3 + 120’) ST-T wave changes Left bundle branch block Right bundle branch block lntraventricular conduction delay Premature atrial systole Premature ventricular systole Atrial fibrillation
54 9 36 2 17 3 6 2 10 9 7
Data from references 7 and 22.
(Fig l).” In addition, the percentage of missing P waves increases. These alterations reflect loss of atrial myocardium, fibrosis, or cellular changes. In the limb leads, P waves may be so low as to be invisible. Unless the ECG is examined carefully, especially Vi, sinus rhythm may be overlooked. When P waves are not seen and the heart rhythm is regular, recording lead V, at double standardization may make the atrial depolarization apparent.” First-degree AV block The prevalence of first-degree AV block, defined as a PR interval >0.21 seconds, increases with age.i3 In a broad range of geriatric patients this prevalence is approximately lo%, but in nonagenarians the prevalence rises to 35%. A prolonged PR interval is not usually associated with clinical heart diseasei and when first-degree AV block is an isolated
FIGURE
1.
Electrocardiogram
241
finding, it does not affect prognosis.” However, these individuals may be more susceptible to digitalis toxicity. QRS Comptex The duration of the QRS interval tends to be greater in the aged because of the slight slowing of the conduction system, but it is never prolonged to 0.12 seconds in the absence of a bundle branch block.‘* Epidemiologic studies have shown, however, that the R and S wave amplitudes decrease with age, while the mean QRS vector rotates to the left, in the frontal plane, and posteriorly, in the horizontal plane.14 This gradual process, which parallels the effects of obesity, is apparent from early adult life and has been attributed to’ kyphoscoliosis with increasing AP diameter of the chest, lowering of the diaphragms due to emphysema, loss of the elasticity and increased resistivity of surrounding tissues, elongation of the aorta, and partial fibrosis in the anterior fasicle of left bundle branch. Abnormal left axis deviation (ALAD) is defined as a frontal plane axis of - 30” or greater. The prevalence of ALAD in the aged is eightfold greater than in the young and is noted twice as often in the presence of clinical heart disease. l5 However, patients with isolated ALAD have no significant increases in morbidity, mortality rate, or risk factors when compared with the general population. 13*16Right axis deviation is uncommon in older people but may be found in thin, long-chested persons with a vertical heart and emphysema.” Ventricular Hypertmphy Left ventricular hypertrophy (LVH) is one of the common ECG abnormalities in the elderly. The actual alence ranges from 1% to 40% depending on the criteria in diagnosis. l3 The Framingham study” indicates that
from a healthy 73syear-old male. Note the diffuse P wave notching.
most prevused elec-
242
AMERICAN JOURNAL OF EMERGENCY MEDICINE n Volume 8, Number 3 n May 1990
trocardiographic evidence of LVH increases in frequency with age and parallels the distribution of systemic hypertension. The data also suggest that LVH is a poor prognostic sign, even in the absence of clinical evidence of heart disease. The criteria for LVH and strain in younger people apply with a few exceptions in the elderly. Tall R waves in V, or V, may be due to the proximity of the left ventricle to the chest electrode and should not be interpreted as LVH unless the total sum of RV, and SV, exceeds 35 mm and RaVL is 11 mm or greater. i2 Several series have shown that all voltage criteria lack sensitivity and specificity in the diagnosis of LVH.‘,” In view of this, Campbell et al’* suggest that the diagnosis of LVH is more specific when voltage criteria are satisfied and when discordant ST depressions or T wave changes are present. Right ventricular hypertrophy (RVH) and car pulmonale are uncommon in the aged. One investigator found only 8 cases out of 2,254 elderly subjects.‘* However, this diagnosis may be difficult to make. A right axis deviation may be secondary to emphysema and the increase of R wave voltage in V, and of S wave in V, are often missing. Furthermore, some patterns which suggest RVH are found only in extreme right precordial leads (V,R to V,R), which are not recorded. Bundle branch block The overall prevalence of bundle branch block in young military populations is considerably less than 1%.19 In collected series of aged persons, Mihalick and Fisch’ found left anterior hemiblock (LAH) in 360/o,right bundle branch block in 6%, and left bundle branch block in 3% of subjects. In almost all series of elderly persons, right bundle branch block occurs slightly more often than left bundle branch block. The prognosis of bundle branch block is dependent on its etiology. In the elderly, bundle branch block is usually caused by fibrocalcific changes in the conduction system rather than ischemic heart disease.13 Isolated left bundle branch block has been shown to have a slightly higher incidence of heart disease and subsequent mortality than agematched controls.‘,” Right bundle branch block is considered a benign condition and only when associated with ALAD is it accompanied by an increased incidence of heart disease. Arrhythmias The prevalence and complexity of most arrhythmias are increased in older populations. Routine ECG recordings performed in the nonhospitalized elderly reveal an incidence of 10% in both atrial and ventricular premature beats.’ In comparison, a younger population studied by a survey of ECG records from over 60,ooO Air Force personnel disclosed a 0.6% prevalence of extrasystoles.” Ambulatory ECG recording provides a more sensitive method to establish the prevalence of arrhythmias. Twentyfour hour ECG monitoring in carefully screened healthy elderly have shown a substantial prevalence of isolated supraventricular (88%) and ventricular (80%) ectopic beats.20,2’ Supraventricular tachyarrhythmias were found in 30% of the elderly with 26% displaying more than 100 atrial ectopic beats over the 24-hour monitoring period. Ventricu-
lar couplets or short runs of ventricular tachycardia were detected in 15%. Nearly all arrhythmias were asymptomatic, regardless of type or complexity. The mere presence of premature systoles in an older patient, therefore, does not necessarily imply either the presence of disease or an indication for treatment. Atria1 fibrillation occurs in 2% to 8% of all elderly subjects.22 A higher frequency has been reported in hospitalized geriatric patients23 and in the very old, ie, nonagenarians.13 It is universally considered an abnormal ECG finding, the vast majority of subjects with atrial fibrillation have demonstrable heart disease.’ The hemodynamic consequences of atrial fibrillation are related to the ventricular response. If the ventricular rate is 60 to 90 beats/min, which allows adequate ventricular filling, then the rhythm is usually well tolerated, even in the aged. It is signitlcant that bradycardias (sinus bradycardia, sinus arrest, sinoatrial exit block, and second- or third-degree AV block) are rare in normal elderly subjects.‘3*20 As a result, these findings in symptomatic patients may indicate the need for corrective therapy and hospitalization. ST-T Wave The most commonly observed age-related ECG changes involve ventricular repolarization: the ST segment flattens and the T wave amplitude diminishes (Fig 2). ST depression is usually found in the left precordial leads, especially in V, and V, (Fig 3).i2 The amplitude of the T wave decreases slightly with advancing years, and low or notched T waves are not uncommon. Inverted T waves in leads I, II and precordial leads V, to V, indicate myocardial damage or ischemia. Deep symmetrical T-wave inversion in the absence of voltage criteria of LVH also indicates subendocardial ischemia. Several studies indicate that the incidence of nonspecific ST-T wave changes is nearly 17% in subjects over the age of 70.3*7s,“,13 In the elderly, the multiple noncardiac causes of T wave inversions and ST segment depressions-obesity, hypoglycemia, metabolic disorders, and aging changes in the chest-result in a loss of specificity and diagnostic value. If noncardiac causes can be excluded, the ST-T wave changes in the resting ECG have the same prognostic implications as in the middle-aged population. The QT interval usually increases with age but rarely is above the upper limit for younger people. I2 Myocardial Infarction Despite the decline in cardiovascular mortality during the past decade, acute myocardial infarction still claims over 600,000 lives annually, at least half of which occur in individuals aged 65 years or older.24 Published data indicate that over 50% of elderly subjects with ECG evidence of myocardial infarction failed to manifest clinically I3 and less than 20% of elderly patients with acute infarction present with classical symptoms.” Most symptoms were secondary to impaired myocardial function resulting in low cardiac output or congestive heart failure. In the elderly, about 50% of all patients with acute myocardial infarction will have new findings of ST elevation or Q waves in two or more leads upon presentation.” Approximately another 25% will have new ECG findings of ischemia
JONES, SRODULSKI, ROMISHER W THE AGING ELECTROCARDIOGRAM
FIGURE 2.
Note the nonspecific ST-T wave changes and leftward axis in this 71-year-old patient without clinical evidence of heart disease.
Only about 25% of all or strain without ST elevation.‘*” acute myocardial infarctions will present to the clinician with ECGs that do not show any new changes of ischemia, strain, or infarction.1*‘1*13 Unfortunately, in the elderly, reliance on the electrocardiogram alone would miss about 15% to 20% of all myocardial infarctions.26
FIGURE 3. diminishes).
243
In addition, an infarction pattern on ECG may be complicated by the presence of left bundle branch block, prior infarction, cardiac amyloidosis, emphysema, or chest deformities. Septal infarction is particularly difficult to diagnose due to the usual reduction of the initial vector of ventricular activation, as a result of senile infraseptal fibrosis. This in-
81-year-old woman with age-related changes involving ventricular repolarization
(ie, ST segment flattens and T wave amplitude
244
AMERICAN JOURNAL OF EMERGENCY MEDICINE m Volume 8, Number 3 m May 1990
volves the disappearance of the R wave in V,, V, as well as of the Q wave in a VL, V,, V,. Finally, ST segment and T wave abnormalities may be the only ECG signs of nontransmural infarction. However, their presence should be interpreted with caution in elderly subjects, since they may be associated with left ventricular hypertrophy, paced rhythm, and other nonacute conditions.24 CLINICALHEART DISEASE The incidence of abnormal ECG increases with age as does the incidence of clinical heart disease (Table 2). Mihalick and Fisch’ studied the correlation between ECG variables and clinical heart disease as manifested by angina, cardiomegaly, heart failure, prior infarction, and pronounced hypertension. Those ECG abnormalities that correlated strongly with heart disease were atria1 fibrillation, left bundle branch block, and nonspecific intraventricular conduction delay. There were no significant correlations with right bundle branch block, left anterior hemiblock, prolonged PR interval, frequent premature contractions, nonspecific ST-T wave changes, or ECG evidence of myocardial infarction.’ Although these ECG abnormalities did not correlate with the presence of clinical disease, strong evidence suggests that left anterior hemiblock, right bundle branch block, and myocardial infarction, reflect anatomic heart disease, the anatomic changes not severe enough to manifest clinically.‘5 For example, prevalence of marked left axis deviation is seen twice as often in the presence of overt heart disease. Even though the statistical correlation with clinical heart disease is not significant, it is reasonable to assume that left anterior hemiblock reflects septal disease not always manifested clinically. In a study of the ECG abnormalities seen in patients 85 or TABLE 2. Characteristics of the Aging Elsdmcafdiogram 1. The incidence of abnormal ECG increases with age and heart disease. 2 Probable insignificant ECG findings in the elderly include left axis deviation without evidence of LAH, lower wave amplitudes, longer intervals, and isolated premature contractions. 3. T wave inversions and ST segment depression have too many noncardiac causes to be specific. 4. Bradycardias (sinus bradycardia, sinus arrest, exit block, and second- or third-degree AV block) are rare in normal subjects. 5. The specific ECG abnormalities that correlate strongly with heart disease are atrial fibrillation, left bundle branch block and nonspecific intraventricular conduction delay. 8. Myocardial infarction patterns, left anterior hemiblock, and right bundle branch block do not correlate with presence of clinical disease, but may reflect anatomic disease. 7. With the exception of left ventricular hypertrophy, prognosis of a specific ECG abnormality remains that of the underlying disease. Reprinted with permission from Bosker G, et al: Assessment and detection of acute coronary ischemia, In Bosker G, et al (eds): Geriatric Emergency Medicine. St Louis, MO, Mosby, in press.
older, Rajala et al” have reported that ECG changes had a poor association with cardiac symptoms. ST segment depression, T wave inversion, premature ventricular contractions, and atrial fibrillation were related statistically to clinical congestive heart failure whereas T wave inversion was related to clinical coronary heart disease. High left R waves, ventricular premature complexes, and atria1 fibrillation showed a significant association with cardiac enlargement and pulmonary congestion. In their study of ECGs from persons with no overt heart disease, many ECG changes were noted, including frequent T wave inversion and frequent ST segment depression. They thought that specificity of ECG changes for heart diseases was not high in very old persons. Electrocardiographic patterns indicative of heart disease rarely show a correlation with either prognosis or severity of the disease. With the possible exception of left ventricular hypertrophy, the prognosis of a specific abnormality remains that of the underlying disease. SUMMARY With advancing age, widespread histologic changes in the conduction system occur. These changes may alter several features of the aging electrocardiogram, including duration of the PR and QT intervals, orientation of the electrical axis, duration and morphology of the atrial and ventricular complexes, and characteristics of the ventricular repolarization. And although ST segment and T wave abnormalities may be the only clue to acute ischemia, they are nonspecific and associated with a multitude of noncardiac causes. With an awareness of atypical presentations and difficulties in ECG interpretation, emergency physicians may be able to improve the assessment and triage of elderly patients with acute coronary ischemia. REFERENCES 1. Bachman S, Sparrow D, Smith LK: Effect of aging on the electrocardiogram. Am J Cardiol 1981;48:513-518 2. Harlan WR, Graybiel A, Mitchell RE, et al: Serial electrocardiograms: Their reliability and prognostic validity during a 24-yr period. J Chronic Dis 1967;20:853-887 3. Simonson E: The effect of age on the electrocardiogram. Am J Cardiol 1972;29:64-73 4. Willius EA: The heart in old age: A study of 700 patients seventv-five vears of aae and older. Am J Med Sci 1931:182:1-19 ._ 5. Fisch C, Genoveie PD, Dyke RW, et al: The electrocardiogram in persons over 70. Geriatrics 1957;12:816_62CJ 8. Taran LM, Szilagyi N: Electrocardiographic changes with advancing age. Geriatrics 1958;13:352-358 7. Mihalick MJ, Fisch C: Electrocardiographic findings in the aged. Am Heart J 1974;87:117-128 8. Olbrich 0, Woodford-Williams E: The normal precordial electrocardiogram in the aged. J Gerontol 1953;8:40-55 9. Fleg JL, Gerstenblith G, Lakatta EG: Pathphysiology of the aging heart and circulation, In Messerli F (ed): Cardiovascular Disease in the Elderly. Norwell, MA, Kluwer Academic, 1984; pp 11-21 10. Gerstenblith G, Weisfeldt ML, Lakatta EG, et al: Disorders of the heart, In Andres R, Bierman EL, Hazard WR (eds): Principles and Practice of Geriatric Medicine. New York, NY, McGraw-Hill, 1985, pp 520-522 11. Rodstein M: The ECG in old age: Implications for diagnosis, therapy, and prognosis. Geriatrics 1977;32:76-79 12. Harris R: Clinical Geriatric Cardiology. Philadelphia, PA, Lippincott, 1986; pp 68-93 13. Camm AJ, Ward DE: Clinical electrocardiography, In Mar-
JONES, SRODULSKI, ROMISHER n THE AGING ELECTROCARDIOGRAM
tin A, Camm AJ (eds): Heart Disease in the Elderly. New York, NY, Wiley, 1984; pp 149-186 14. Wagner NB, Schocken DD, Lee L, et al: Evaluation of the QRS complex on the standard 1Blead electrocardiogram in normal subjects 70 to 79 years of age. Am J Cardiol1988;62:982-985 15. Fisch C: The electrocardiogram in the aged, In Noble RJ, Rothbaum DA (edsl: Geriatric Cardioloav. PA. _. Philadelohia. . Davis, 1981; pp 65-74 16. Ostrander LD: Left axis deviation: Prevalence, associated conditions, and prognosis. An epidemiologic study. Ann Intern Med 1971;75:23-28 17. Kannel WB, Gordon T, Offutt D: Left ventricular hypertrophy by electrocardiogram-prevalence, incidence, and mortality in the Framinaham studv. Ann intern Med 1969:71:89-105 18. Campb& A, Caird-FI, Jackson TFM: Prevalence of abnormalities of electrocardiogram in old people. Br Heart J 1974; 36:1005-1011 19. Johnson RL, Averill KH, Lamb LE: Electrocardiographic findings in 63,375 asymptomatic individuals. Am J Cardiol 1960;6:143-167 20. Fleg JL, Kennedy HL: Cardiac arrhythmias in a healthy
245
elderly population: Detection by 24-hour ambulatory electrocardiography. Chest 1982;81:302-397 21. Anderson S, Osterlind PO, Holmboe G, et al: Twentyfour-hour electrocardiography in a healthy elderly population. Gerontology 1988;34:139-144 22. Coodley E, Coodley G: Electrocardiographic changes associated with aging, In Coodley E (ed): Geriatric Heart Disease. Boston, MA, PSG, 1985, pp 182-187 23. Mihalick MJ, Fisch C: Should ECG criteria be modified for geriatric patients? Geriatrics 1977;32:65-72 24. Flag JL: Cardiovascular emergencies, In Wilson LB, Simson SP, Baxter CR (eds): Handbook of Geriatric Emergency Care. Baltimore, MD, University Park Press, 1984, pp 38-50 25. Pathy MS: Clinical presentation of myocardial infarction in the elderly. Br Heart J 1967;29:190-199 26. Presentation of myocardial infarction in the elderly. Lancet 1986;2:1077-1078 (editorial) 27. Rajala SA, Ulla KM, Geiger MKM: Electrocardiogram, clinical findings, and chest x-ray in persons aged 85 or older. Am J Cardiol 1985;55:1175-1178
