Heart rate

The usual resting heart rate of 65–75 beats/min reflects a substantial degree of bradycardic vagal tone, so heart rate can be increased moderately either by reducing that vagal influence or by increasing sympathetic tachycardic drive, or both. Any increase above the intrinsic pacemaker frequency of 100 beats/min, however, relies entirely on sympathetic activation. The sympathetic nerves act through activation of β-adrenoceptors and drugs that antagonize this action (the so-called β-blockers) are frequently used in patients for whom exercise is prescribed as a rehabilitative aid after heart attacks. The reduced capacity to produce normal tachycardia is an important factor in determining the absolute intensities of exercise that these individuals are able to undertake and must also be borne in mind if absolute heart rate is being used to quantify their exercise workload (see Chapter 11, p. 131).

With ageing there is a progressive reduction in the capacity of cardiac β-adrenoceptors to respond to sympathetically released catecholamines. In consequence, the absolute maximum to which heart rate can rise during exercise declines with age, being around 200 beats/min at age 20 years, but falling by around 1 beat/min per year. This imposes a progressive limit to cardiac output in older individuals, regardless of their physical fitness.

The standard equation to calculate maximum heart rate (HR_max) for an adult of a known age is:

but it is important to bear in mind that this can be applied only to adult subjects. In children, maximum heart rate appears to vary little or not at all with age and remains around a little more than 200 beats/min until approximately 18 years (see Chapter 9, p. 113). In addition, the generalization of HR_max falling by 1 beat/min per year has been derived from population studies, and absolute maxima vary by up to 10 beats/min between people of any given age. For this reason, if heart rate is to be used for quantifying workload during exercise then it is preferable to determine each individual’s HR_max directly.

In obese adults (BMI >30), the relationship between maximal heart rate and age is slightly different and the equation:

appears to be more accurate than the standard one (Miller et al 1993).

Ventricular filling time

Simple calculation would indicate that a threefold elevation of heart rate should increase cardiac output by the same amount, but it is clear from looking at the pressure–flow relationships during the cardiac cycle (Fig. 3.1) that the situation is not as straightforward as this. As heart rate increases, the interval between successive ventricular contractions decreases so that the absolute time available for refilling falls. With moderate tachycardia this is not a major problem, since most filling occurs during the first 100 ms of diastole when the atrioventricular pressure gradient is greatest. As heart rate increases further, however, filling efficiency falls dramatically. With typical ventricular action potential durations of 300–350 ms it can be calculated that a heart rate of 180 beats/min (1 beat every 330 ms) would actually allow no time for filling at all. Moreover, if there was no diastolic relaxation period then there could be virtually no coronary perfusion to the left ventricle, so myocardial metabolism could not be sustained. Since maximal exercise capacity involves heart rates in young adults of 200 beats/min and cardiac outputs in excess of 20 L/min, this scenario is clearly too simplistic.

Figure 3.1 Simulated ventricular pressure and volume curves at 80 and 110 beats/min in a heart with constant ventricular action potential duration. Note that even this moderate increase in heart rate causes a dramatic reduction of ventricular filling time and reduces stroke volume significantly.

The answer is that, in fact, ventricular action potential duration does not remain constant as heart rate increases, because the sympathetically released catecholamines that produce tachycardia also reduce the cycle time of the voltage-gated calcium channels (Fig. 3.2). In consequence, the normal HR_max for a 20-year-old of 200 beats/min is associated with an endocardial ventricular action potential plateau phase of about 200 ms rather than the 350 ms seen at rest, allowing around 100 ms for ventricular filling. In consequence, stroke volume falls only slightly even at maximal work capacity.

Figure 3.2 Simulated ventricular pressure and volume curves at 80 and 110 beats/min in a heart with ventricular action potential duration reduced as heart rate increases. Note the improved filling time and stroke volume, compared with Figure 3.1.

Atrial function

For any finite diastolic period available for ventricular filling, the efficiency of the filling process depends on the pressure gradient between atria and ventricles. This is itself determined by the efficiency of atrial filling, which reflects the pressure gradient from the peripheral veins to the heart. During exercise, two factors facilitate venous return. One is the increased negative pressure inside the thorax during inspiration that results from larger tidal volume. The other is external compression of veins in the moving limbs by muscle contraction (muscle pumping) and in the abdomen by abdominal wall muscle activity during expiration. The importance of the leg muscle pump for efficient cardiac filling is illustrated by comparing circulatory responses to arm and leg exercise. During arm exercise, heart rate rises more with given work increments, because the absence of muscle pumping limits stroke volume (see Chapter 7, p. 89).

The atrioventricular pressure gradient is increased directly by increased atrial filling, since this stretches the atrial walls towards their elastic limit so that intra-atrial pressure rises. In addition, activation of atrial β-adrenoceptors by sympathetically released noradrenaline (norepinephrine) and circulating adrenaline (epinephrine) speeds up action potential spread through the atrial syncytium and opens more voltage-gated calcium channels, causing larger amounts of extracellular calcium to enter the atrial cells thereby increasing sarcomere activation (increased contractility). These processes both increase active pressure development during atrial contraction.

Ventricular function

The net consequence of the atrial processes described above is that end-diastolic ventricular volume and pressure rise progressively with exercise until around 60% of maximum work capacity. There are no further changes at workloads higher than this, for three reasons: first, the ongoing reduction in diastolic time limits atrial filling; second, at diastolic volumes above a certain value the ventricular muscle begins to reach its elastic limits and so becomes relatively non-distensible; and third, at these high volumes ventricular expansion becomes restricted also by the stiff pericardial sac that surrounds the heart.

Increased ventricular filling will itself increase stroke volume, due to the Frank-Starling relationship of muscle stretch and active pressure development. In the presence of sympathetic activation, however, systolic pressure generation rises even more, due to the same effects of catecholamines on calcium channel opening as occurs in the atria. The consequences of increased contractility and more rapid contraction of the ventricular syncytium lead to an absolute reduction in end-systolic ventricular volume, so that stroke volume rises with moderate tachycardia even though diastolic filling time is reduced (Fig. 3.3).

Figure 3.3 Simulated ventricular pressure and volume curves at 80 and 110 beats/min in a heart with ventricular action potential duration reduced and myocardial contractility increased as heart rate increases. Note that despite the reduced filling time (compare with Fig. 3.2), stroke volume is now higher at the higher heart rate, due to both increased ventricular filling and increased ventricular emptying.

Ventricular afterload

Just as the atrioventricular pressure gradient influences ventricular filling, the pressure gradient between ventricles and arteries must affect the efficiency of ventricular ejection. A higher diastolic arterial pressure must delay equalization of ventricular and arterial pressures during the isovolumetric phase of ventricular contraction, so shortening the proportion of systole during which ejection occurs. In theory, therefore, cardiac output should be elevated most effectively when diastolic blood pressure is minimized. In practice, however, as we shall see in Chapter 7 (p. 85), certain types of exercise are associated with increased diastolic pressure, so this can be regarded as a further factor in setting the upper effective limit for cardiac performance.

Figure 3.4 shows how the factors discussed above contribute to the exercise response.

Figure 3.4 Expansion of Figure 1.1, with the factors discussed in this chapter denoted in red.

Turbulent flow

When fluid such as blood is pushed through tubes at low velocities, the particles move parallel to the tube wall and to each other, a state known as laminar flow. Under these conditions, all the energy applied to the fluid is used to move it. At sufficiently high flow velocities, however, the particles begin to interact radially with each other, resulting in swirling movement within the fluid that is termed turbulence. Since under these circumstances energy is expended in the inter-particle collisions, turbulent flow is less efficient than laminar flow with respect to the amount of applied force (that is, the pressure gradient) required to move a given volume of fluid through the tube. Turbulent and laminar flow also differ in that the collisions of fluid particles in turbulent flow create a noise, while laminar flow is silent.

With equal volumes of fluid movement per unit time, flow velocity is higher in narrow than wide tubes, so turbulence should be more likely in a narrow tube. Paradoxically, however, turbulence is also facilitated by large tube diameter; so absolute flow velocities that are not high enough to create turbulent flow in narrow tubes may produce turbulence in larger ones. During the cardiac cycle, the absolute velocity of blood flow is highest during the early part of ventricular ejection, when intraventricular pressure is still rising. Normally, this velocity is slightly less than that required to produce turbulence in tubes the size of the aorta and pulmonary arterial trunk. It does, however, cause turbulence in the larger diameter environment of the ventricle itself. This serves a valuable purpose.

Blood draining back to the right heart from different organs contains varying amounts of secreted hormones, oxygen and carbon dioxide, reflecting the different functions and metabolic rates of the tissues involved. Similarly, blood returning to the left heart from different parts of the lung varies significantly in gas content, reflecting regional variation in the efficiency of ventilation/perfusion matching. These different venous returns are usually not well mixed when they enter the heart, because laminar flow keeps them separate. In consequence, the end-diastolic left ventricular content is far from homogenous in composition and, since arterial flow is also laminar, this creates the potential for different organs to receive blood with different gas tensions and to receive variable amounts of essential hormones. Intraventricular turbulence during the initial phase of systolic ejection ensures homogeneity of the blood delivered to all tissues.

Heart sounds

The rapid pressure and flow changes in the heart during its pumping action generate noises that can be detected at the surface of the chest wall using a microphone or a stethoscope. These heart sounds constitute what is also known as the phonocardiogram and provide important information about the mechanical events that occur during the cardiac cycle. What is termed the first heart sound begins coincidently with the beginning of ventricular contraction and finishes shortly after the beginning of systolic ejection. It involves two sequential events. One corresponds to shutting of the atrioventricular valves as soon as intraventricular pressure begins to rise. The physical closure of the valves does not itself produce noise because they consist of very soft tissue, but the sudden cessation of blood movement causes vibration of the ventricular walls. This sound is followed without pause by a longer-lasting component due to the fast turbulent phase of systolic ejection. What is termed the second heart sound corresponds to the end of systole, when the rapid fall in intraventricular pressure reverses the pressure gradient between arteries and ventricles and the semi-lunar valves shut. These valves consist of relatively stiff tissue and so, unlike the situation with the atrioventricular valves and the first heart sound, the noise created directly reflects their shutting.

Because the heart sounds have a predictable relationship to mechanical events, they can be used to time the cardiac cycle (Fig. 3.5). Thus, the period between the beginning of the first and second sounds must define the period during which the ventricles are contracting. More important, since the isovolumetric phase of systole is short and varies only relatively little over a wide range of afterloads, the period between first and second sounds approximates the duration of ejection. Although the interval between R and T waves in the ECG also gives some information on this, timing of the end of ejection from the ECG is uncertain because of the long timecourse of the T wave.

Figure 3.5 Timing of the first and second heart sounds relative to pressure and flow changes during the cardiac cycle. Note that the first sound occurs during both isovolumetric contraction and the initial phase of ventricular ejection, while the second sound is much shorter and defines the cessation of ejection.

The events that dictate whether flow is laminar or turbulent, together with the basis for the heart sounds, mean that altered heart sounds can be used to detect a variety of defects of cardiac function. For instance, failure of the atrioventricular valves to close fully (valvular incompetence) would lead to retrograde leakage at high pressure during ventricular contraction. This produces turbulent noise during systolic ejection and so prolongs the first sound. Similarly, incompetence of the semi-lunar valves causes turbulent backflow at the end of systole and so prolongs the second sound. Finally, if the semi-lunar valves do not open fully (valvular stenosis) then the velocity of blood flow into the central arteries will be increased and create turbulent noise that may last through the entire ejection period.

Extrapolation from oxygen consumption

Some years ago (Astrand et al 1964) it was shown that with workloads below the lactate threshold cardiac output can be estimated from oxygen consumption according to the following equations, where both cardiac output and oxygen consumption are expressed in L/min:

Although these equations were based on careful experimentation, they may not give accurate values for cardiac output under some circumstances and more direct techniques of measurement are usually preferable. Even then, as will be seen below, the available methods all have certain limitations if they are to be applied to an exercising subject.

Application of the Fick principle to carbon-dioxide production

Carbon dioxide is produced continually by the peripheral tissues and cleared into the alveolar air. If one measures whole-body production by collecting expired air and simultaneously measures the difference in carbon-dioxide concentration across the pulmonary circulation, then these figures can be used to calculate the volume of blood that was necessary to deliver the expired carbon dioxide. Take for example the following typical values:

From the difference between pulmonary arterial and pulmonary venous contents, each 100 mL blood loses 4 mL CO₂ as it passes through the lung. If total CO₂ production is 200 mL/min then pulmonary blood flow (that is, cardiac output) over the same time must be:

that is, cardiac output is 5 L/min.

In practice, the rapid equilibration of CO₂ between blood and alveolar air means that we do not need to measure blood concentrations of CO₂ directly in order to make reasonably accurate estimates of cardiac output, provided that pulmonary function is normal and the subject is at a steady state for CO₂ production. The end-expiratory CO₂ concentration can be regarded as in equilibrium with that leaving the lung in the pulmonary venous blood (PaCO₂) and this can be calculated as:

and then converted to arterial CO₂ content by reference to a CO₂ dissociation table. The assumption that complete equilibrium is achieved between air and plasma has to be recognized as a potential source of error, because the steep slope of the CO₂ dissociation curve means that significant changes in CO₂ content could occur with only minor shifts in alveolar PCO₂.

Pulmonary arterial CO₂ concentration (P_VCO₂) is approximately 6 mmHg higher than PaCO₂, under almost all circumstances in normal individuals, regardless of whether they are at rest or exercising. Accurate measurement of PvCO_2, however, requires a rebreathing technique. The subject breathes into a bag that contains oxygen plus a concentration of around 10% CO₂, which is significantly higher than PvCO₂. This reversal of the normal concentration gradient causes CO₂ to diffuse into the bloodstream until, after a few breaths, alveolar and plasma concentrations equilibrate and the PCO₂ remaining in the bag is identical to P_VCO₂.

With continuous gas analysis using a metabolic cart and approximation of P_VCO₂ this technique allows cardiac output determination on a breath-to-breath basis. The rebreathing procedure, however, requires around 5 or 6 breaths, so accurate measurements can be made only around once per minute. Clearly it is not technically feasible to measure true beat-to-beat output (stroke volume).

Nitrous oxide rebreathing

This is another method that employs the Fick principle, but which measures pulmonary exchange of a foreign gas as a marker rather than that of endogenous CO₂. The subject breathes from a bag of oxygen containing known concentrations of a soluble inert gas (usually nitrous oxide) and of an insoluble gas such as sulphur hexafluoride. The rate of decrease in nitrous oxide concentration in the bag must be proportional to the rate at which it is taken up by the pulmonary bloodstream, and computer algorithms allow the volume flow involved to be calculated. In healthy lungs, the only other factor that could affect gas uptake is a change in pulmonary exchange area caused by altered depth of ventilation. This is adjusted for by monitoring any change in the concentration of the insoluble marker. The inert gas technique avoids the approximations that are inherent with methods that use CO₂ and the literature suggests that this method provides closely similar values of cardiac output to those obtained using invasive techniques in the same subjects. For the purposes of the exercise physiologist, its main limitation is that all the absorbed nitrous oxide has to be expired again before a second measurement can be made. In practice this means that successive measurements can be obtained only once every several minutes, which can limit the accuracy of tracking cardiac output during graded exercise.

Thoracic conductance and impedance

If the resting electrical resistance of the thorax is measured along the rostrocaudal axis, a decreased resistance can be detected during the rostral movement of blood into the aortic arch that is associated with cardiac ejection. The magnitude of this change is proportional to the volume of blood involved and is, therefore, an index of stroke volume. Resistance measurement is by means of adhesive skin electrodes positioned above and below the rostrocaudal limits of the heart and stroke volume is calculated using an algorithm that incorporates the resistivity of blood and the period of ejection as measured from the ECG or a phonocardiogram.

These techniques have been calibrated against invasive methods for assessment of cardiac output. They appear acceptably accurate and are able to provide beat-to-beat measurement of cardiac output at rest and during light to moderate whole-body stationery exercise. With intense exercise, however, movement artefacts preclude readable records and, with prolonged exercise or exposure to heat, sweat-induced loss of electrode adhesion becomes a problem.

Imaging

Like the impedance and conductance techniques, Doppler echocardiography provides beat-to-beat information on stroke volume. It involves simultaneous determination of aortic arch diameter and the velocity of blood flow through this vessel, by means of a detector placed manually on the thoracic wall. The technique is reliable for assessing cardiac output at rest, but it relies on very accurate and stable positioning of the imaging detector on the skin. Therefore, although it could be employed during exercise that involves only lower body movements it is not useful in any circumstances where the subject is ambulatory.