The ECG in Healthy People

Published on 21/06/2015 by admin

Filed under Cardiovascular

Last modified 22/04/2025

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 0 (0 votes)

This article have been viewed 5978 times

1

The ECG in Healthy People

For the purposes of this chapter, we shall assume that the subject from whom the ECG was recorded is asymptomatic, and that physical examination has revealed no abnormalities. We need to consider the range of normality of the ECG, but of course we cannot escape from the fact that not all disease causes symptoms or abnormal signs, and a subject who appears healthy may not be so and may therefore have an abnormal ECG. In particular, individuals who present for screening may well have symptoms about which they have not consulted a doctor, so it cannot be assumed that an ECG obtained through a screening programme has come from a healthy subject.

The range of normality in the ECG is therefore debatable. We first have to consider the variations in the ECG that we can expect to find in completely healthy people, and then we can think about the significance of ECGs that are undoubtedly ‘abnormal’.

THE ‘NORMAL’ ECG

THE HEART RATE

There is no such thing as a normal heart rate, and the terms ‘tachycardia’ and ‘bradycardia’ should be used with care. There is no point at which a high heart rate in sinus rhythm has to be called ‘sinus tachycardia’ and there is no upper limit for ‘sinus bradycardia’. Nevertheless, unexpectedly fast or slow rates do need an explanation.

SINUS TACHYCARDIA

The ECG in Figure 1.2 was recorded from a young woman who complained of a fast heart rate. She had no other symptoms, but was anxious. There were no other abnormalities on examination, and her blood count and thyroid function tests were normal.

Box 1.1 shows possible causes of sinus rhythm with a fast heart rate.

EXTRASYSTOLES

Supraventricular extrasystoles, either atrial or junctional (AV nodal), occur commonly in normal people and are of no significance. Atrial extrasystoles ( Fig. 1.4) have an abnormal P wave; in junctional extra-systoles either there is no P wave or the P wave may follow the QRS complex.

Ventricular extrasystoles are also commonly seen in normal ECGs ( Fig. 1.5).

In healthy people, normal sinus rhythm may be replaced by what are, in effect, repeated atrial extra-systoles. This is sometimes called an ‘ectopic atrial rhythm’ and it is of no particular significance ( Fig. 1.6).

THE P WAVE

In sinus rhythm, the P wave is normally upright in all leads except VR. When the QRS complex is predominantly downward in lead VL, the P wave may also be inverted ( Fig. 1.7).

In patients with dextrocardia the P wave is inverted in lead I ( Fig. 1.8). In practice this is more often seen if the limb leads have been wrongly attached, but dextrocardia can be recognized if leads V5 and V6, which normally ‘look at’ the left ventricle, show a predominantly downward QRS complex.

If the ECG of a patient with dextrocardia is repeated with the limb leads reversed, and the chest leads are placed on the right side of the chest instead of the left, in corresponding positions, the ECG becomes like that of a normal patient ( Fig. 1.9).

A notched or bifid P wave is the hallmark of left atrial hypertrophy, and peaked P waves indicate right atrial hypertrophy – but bifid or peaked P waves can also be seen with normal hearts ( Fig. 1.10).

THE PR INTERVAL

In sinus rhythm, the PR interval is constant and its normal range is 120-200 ms (3-5 small squares of ECG paper) ( Fig. 1.11). In atrial extrasystoles, or ectopic atrial rhythms, the PR interval may be short, and a PR interval of less than 120 ms suggests pre-excitation.

A PR interval of longer than 220 ms may be due to first degree block, but the ECGs of healthy individuals, especially athletes, may have PR intervals of slightly longer than 220 ms – which can be ignored in the absence of any other indication of heart disease.

PR interval ‘abnormalities’ will be discussed further in the context of normal people in Chapter 2.

THE QRS COMPLE

THE CARDIAC AXIS

There is a fairly wide range of normality in the direction of the cardiac axis. In most people the QRS complex is tallest in lead II, but in leads I and III the QRS complex is also predominantly upright (i.e. the R wave is greater than the S wave) ( Fig. 1.12).

The cardiac axis is still perfectly normal when the R wave and S wave are equal in lead I: this is common in tall people ( Fig. 1.13).

When the S wave is greater than the R wave in lead I, right axis deviation is present. However, this is very common in perfectly normal people. The ECG in Figure 1.14 is from a professional footballer.

It is common for the S wave to be greater than the R wave in lead III, and the cardiac axis can still be considered normal when the S wave equals the R wave in lead II ( Fig. 1.15). These patterns are common in fat people and during pregnancy.

When the depth of the S wave exceeds the height of the R wave in lead II, left axis deviation is present (see Figs 2.25 and 2.26).

THE SIZE OF R AND S WAVES IN THE CHEST LEADS

In lead V1 there should be a small R wave and a deep S wave, and the balance between the two should change progressively from V1 to V6. In lead V6 there should be a tall R wave and no S wave ( Fig. 1.16).

Typically the ‘transition point’, when the R and S waves are equal, is seen in lead V3 or V4 but there is quite a lot of variation. Figure 1.17 shows an ECG in which the transition point is somewhere between leads V3 and V4.

Figure 1.18 shows an ECG with a transition point between leads V4 and V5, and Figure 1.19 shows an ECG with a transition point between leads V2 and V3.

The transition point is typically seen in lead V5 or even V6 in patients with chronic lung disease (see Ch. 6), and this is called ‘clockwise rotation’. In extreme cases, the chest lead needs to be placed in the posterior axillary line, or even further round to the back (leads V7-V9) before the transition point is demonstrated. A similar ECG pattern may be seen in patients with an abnormal chest shape, particularly when depression of the sternum shifts the mediastinum to the left, although in this case the term ‘clockwise rotation’ is not used. The patient from whom the ECG in Figure 1.20 was recorded had mediastinal shift.

Occasionally the ECG of a totally normal subject will show a ‘dominant’ R wave (i.e. the height of the R wave exceeds the depth of the S wave) in lead V1. There will thus, effectively, be no transition point, and this is called ‘counterclockwise rotation’. The ECG in Figure 1.21 was recorded from a healthy footballer with a normal heart. However, a dominant R wave in lead V1 is usually due to either right ventricular hypertrophy (see Ch. 6) or a true posterior infarction (see Ch. 5).

Although the balance between the height of the R wave and the depth of the S wave is significant for the identification of cardiac axis deviation, or right ventricular hypertrophy, the absolute height of the R wave provides little useful information. Provided that the ECG is properly calibrated (1 mV causes 1 cm of vertical deflection on the ECG), the limits for the sizes of the R and S waves in normal subjects are usually said to be:

However, R waves taller than 25 mm are commonly seen in leads V5-V6 in fit and thin young people, and are perfectly normal. Thus, these ‘limits’ are not helpful. The ECGs in Figures 1.22 and 1.23 were both recorded from fit young men with normal hearts.

THE WIDTH OF THE QRS COMPLE

The QRS complex should be less than 120 ms in duration (i.e. less than 3 small squares) in all leads. If it is longer than this, then either the ventricles have been depolarized from a ventricular rather than a supraventricular focus (i.e. a ventricular rhythm is present), or there is an abnormality of conduction within the ventricles. The latter is most commonly due to bundle branch block. An RSR1 pattern, resembling that of right bundle branch block but with a narrow QRS complex, is sometimes called ‘partial right bundle branch block’ and is a normal variant ( Fig. 1.24). An RSRV pattern is also a normal variant ( Fig. 1.25), and is sometimes called a ‘splintered’ complex.

In perfectly normal hearts the normal rhythm may be replaced by an accelerated idioventricular rhythm, which looks like a run of regular ventricular extra-systoles, with wide QRS complexes ( Fig. 1.26).

THE ST SEGMENT

The ST segment (the part of the ECG between the S wave and the T wave) should be horizontal and ‘isoelectric’, which means that it should be at the same level as the baseline of the record between the end of the T wave and the next P wave. However, in the chest leads the ST segment often slopes upwards and is not easy to define ( Fig. 1.29).

An elevation of the ST segment is the hallmark of an acute myocardial infarction (see Ch. 5), and depression of the ST segment can indicate ischaemia or the effect of digoxin. However, it is perfectly normal for the ST segment to be elevated following an S wave in leads V2-V5. This is sometimes called a ‘high take-off ST segment’. The ECGs in Figures 1.30 and 1.31 were recorded from perfectly healthy young men.

The ST segment is apparently raised when there is ‘early repolarization’, which causes the ST segment to be arched, and is usually only seen in the anterior leads, not the limb leads (see Fig. 1.39).

Box 1.3 shows the possible causes of ST segment elevation, other than myocardial infarction.

ST segment depression is measured relative to the baseline (between the T and P waves), 60-80 ms after the ‘J’ point, which is the point of inflection at the junction of the S wave and the ST segment. Minor depression of the ST segment is not uncommon in normal people, and is then called ‘nonspecific’; theadvantage of using this word is that it leaves the way open for a later change of diagnosis. ST segment depression in lead III but not VF is likely to be nonspecific ( Fig. 1.32). Nonspecific ST segment depression should not be more than 2 mm ( Fig. 1.33), and the segment often slopes upwards. Horizontal ST segment depression of more than 2 mm indicates ischaemia (see Ch. 5).

THE T WAVE

In a normal ECG the T wave is always inverted in lead VR, and often in lead V1, but is usually upright in all the other leads ( Fig. 1.34).

The T wave is also often inverted in lead III but not VF. However, its inversion in lead III may be reversed on deep inspiration ( Figs 1.35 and 1.36).

T wave inversion in lead VL as well as in VR can be normal, particularly if the P wave in lead VL is inverted. The ECG in Figure 1.37 was recorded from a completely healthy young woman.

T wave inversion in leads V2-V3 as well as in V1 occurs in pulmonary embolism and in right ventricular hypertrophy (see Chs 5 and 6) but it can be a normal variant. This is particularly true in black people. The ECG in Figure 1.38 was recorded from a healthy young white man, and that shown in Figure 1.39 from a young black professional footballer. The ECG in Figure 1.40 was recorded from a middle-aged black woman with rather nonspecific chest pain, whose coronary arteries and left ventricle were shown to be entirely normal on catheterization.

Box 1.4 summarizes the situations in which T wave inversion is seen.

Generalized flattening of the T waves with a normal QT interval is best described as ‘nonspecific’. In a patient without symptoms and whose heart is clinically normal, the finding has little prognostic significance. This was the case with the patient whose ECG is shown in Figure 1.41. In patients with symptoms suggestive of cardiovascular disease, however, such an ECG would require further investigation.

Peaked T waves are one of the features of hyper-kalaemia, but they can also be very prominent in healthy people ( Fig. 1.42). Tall and peaked T waves are sometimes seen in the early stages of a myocardial infarction, when they are described as ‘hyperacute’. They are, however, an extremely unreliable sign of infarction.

The T wave is the most variable part of the ECG. It may become inverted in some leads simply by hyperventilation associated with anxiety.

An extra hump on the end of the T wave, a ‘U’ wave, is characteristic of hypokalemia. However, U waves are commonly seen in the anterior chest leads of normal ECGs ( Fig. 1.43), where they can be remarkably prominent ( Fig. 1.44). It is thought that they represent repolarization of the papillary muscles. A U wave is probably only important if it follows a flat T wave.

THE ECG IN ATHLETES

Any of the normal variations discussed above can be found in athletes. There can be changes in rhythm and/or ECG pattern, and the ECGs of athletes mayalso show some features that might be considered abnormal in non-athletic subjects, but are normal in athletes (see Box 1.5). Figure 1.45 shows the short and varying PR interval of an ‘accelerated idionodal rhythm’ (also known as a ‘wandering atrial pacemaker’). Here the sinus node rate has slowed, and the heart rate is controlled by the AV node, which is discharging faster than the SA node.

The ECGs in Figures 1.45, 1.46 and 1.47 were all recorded during the screening examinations of healthy young footballers.

THE ECG IN CHILDREN

The normal heart rate in the first year of life is 140-160/min, falling slowly to about 80/min by puberty. Sinus arrhythmia is usually quite marked in children.

At birth, the muscle of the right ventricle is as thick as that of the left ventricle. The ECG of a normal child in the first year of life has a pattern that would indicate right ventricular hypertrophy in an adult. The ECG in Figure 1.48 was recorded from a normal 1-month-old child.

The changes suggestive of right ventricular hypertrophy disappear during the first few years of life. All the features other than the inverted T waves in leads V1 and V2 should have disappeared by the age of 2 years, and the adult ECG pattern should have developed by the age of 10 years. In general, if the infant ECG pattern persists beyond the age of 2 years, then right ventricular hypertrophy is indeed present. If the normal adult pattern is present in the first year of life, then left ventricular hypertrophy is present.

The ECG changes associated with childhood are summarized in Box 1.7.

FREQUENCY OF ECG ABNORMALITIES IN HEALTHY PEOPLE

The ECG findings we have discussed so far can all be considered to be within the normal range. Certain findings are undoubtedly abnormal as far as the ECG is concerned, yet do occur in apparently healthy people.

The frequency with which abnormalities are detected depends on the population studied: most abnormalities are found least often in healthy young people recruited to the armed services, and become progressively more common in populations of increasing age. An exception to this rule is that frequent ventricular extrasystoles are very common in pregnancy. The frequency of right and left bundle branch block has been found to be 0.3% and 0.1% respectively in populations of young recruits to the services, but in older working populations these abnormalities have been detected in 2% and 0.7% respectively of apparently healthy people.

WHAT TO DOimage

When an apparently healthy subject has an ECG record that appears abnormal, the most important thing is not to cause unnecessary alarm. There are four questions to ask:

THE RANGE OF NORMALITY

Normal variations in the P waves, QRS complexes and T waves have been described in detail. T wave changes usually give the most trouble in terms of ECG interpretation, because changes in repolarization occur in many different circumstances, and in any individual, and variations in T wave morphology can occur from day to day.

Box 1.8 lists some of the ECG patterns that can be accepted as normal in healthy patients, and some that must be regarded with suspicion.

THE PROGNOSIS OF PATIENTS WITH AN ABNORMAL ECG

In general, the prognosis is related to the patient’s clinical history and to the findings on physical examination, rather than to the ECG. An abnormal ECG is much more significant in a patient with symptoms and signs of heart disease than it is in a truly healthy subject. In the absence of any other evidence of heart disease, the prognosis of an individual with one of the more common ECG abnormalities is as follows.

CONDUCTION DEFECTS

First degree block (especially when the PR interval is only slightly prolonged) has little effect on prognosis. Second and third degree block indicate heart disease and the prognosis is worse, though the congenital form of complete block is less serious than the acquired form in adults.

Left anterior hemiblock has a good prognosis, as does right bundle branch block (RBBB). The presence of left bundle branch block (LBBB) in the absence of other manifestations of cardiac disease is associated with about a 30% increase in the risk of death compared with that of individuals with a normal ECG. The risk of death doubles if a subject known to have a normal ECG suddenly develops LBBB, even if there are no symptoms – the ECG change presumably indicates progressive cardiac disease, probably most often ischaemia. Bifascicular block seldom progresses to complete block, but is always an indication of underlying heart disease – the prognosis is therefore relatively poor compared to that of patients with LBBB alone.

ARRHYTHMIAS

Supraventricular extrasystoles are of no importance whatsoever. Ventricular extrasystoles are almost universal, but when frequent or multiform they indicate populations with a statistically increased risk of death, presumably because in some people they indicate subclinical heart disease. The increased risk to an individual is, however, insignificant and there is no evidence that treating ventricular extrasystoles prolongs survival.

Atrial fibrillation is frequently the result of rheumatic or ischaemic heart disease or cardiomyopathy, and the prognosis is then relatively poor. In about one third of individuals with atrial fibrillation no cardiac disease can be demonstrated. However, even in these people the risk of death is increased by three or four times, and the risk of stroke is increased perhaps tenfold, compared with people of the same age whose hearts are in sinus rhythm.

FURTHER INVESTIGATIONS

Complex and expensive investigations are seldom justified in asymptomatic patients whose hearts are clinically normal, but who have been found to have an abnormal ECG.

An echocardiogram should be recorded in all patients with bundle branch block, to assess the size and function of the individual heart chambers. Patients with LBBB may have a dilated cardiomyopathy, and the echocardiogram will then show a dilated left ventricle which contracts poorly. Alternatively they may have ischaemia, and the echocardiogram will show some segments of the left ventricle failing to contract or contracting poorly. Patients with LBBB may also have unsuspected aortic stenosis. Patients with RBBB may have an atrial septal defect or pulmonary hypertension, but quite frequently the echocardiogram shows no abnormality.

Echocardiography may be helpful in establishing the cause of T wave inversion, which might be due to ischaemia, ventricular hypertrophy or cardiomyopathy.

Patients with frequent ventricular extrasystoles seldom need detailed investigation, but if there is any question of underlying heart disease an echocardiogram may help to exclude the possibility of a cardiomyopathy. It is also worth checking their blood haemoglobin level.

In patients with atrial fibrillation, an echocardiogram is useful for defining or excluding structural abnormalities, and for studying left ventricular function. An echocardiogram is indicated if there is anything that might suggest rheumatic heart disease. Since atrial fibrillation can be the only manifestation of thyrotoxicosis, thyroid function must be checked. Atrial fibrillation may also be the result of alcoholism, and this may be denied by the patient, so it may be fair to check liver function.

Table 1.1 shows investigations that should be considered in the case of various cardiac rhythms and indicates which underlying diseases may be present.

Share this: