This chapter is closely related to Chapter 9, which described chemical control of breathing. This division of the subject of control is a semantic one, designed to make learning easier. All chemical control involves neural sensory mechanisms, and it is neural mechanisms that determine and bring about breathing, which in turn plays such an important part in the homeostatic control of the chemical composition of the body. The difference between chemical and neural control is really a matter of time: chemical control, which we have dealt with, takes place over seconds or minutes; neural control reacts in fractions of a second to influence breathing breath by breath.

The coordinated act of ‘breathing’ is ancient in origin. Sea anemones open their mouths and contract their bodies to expel and replace their coelenteric fluid every half-hour or so to ensure O₂ supply to their cells. This opening of the mouth and body contraction obviously involves a coordinating system, which is most probably neural, but next to nothing is known of this. Even more wonderful, the humble sea-cucumber achieves the majority of its respiratory exchange by ‘breathing’ water into and out of its cloaca, and, in a way that inevitably invites comparison with the gasp and breath-hold we make when startled, the sea-cucumber ‘holds its breath’ when poked.

We are not normally conscious of the automatic systems of our bodies. The kidneys, gut, cardiovascular and respiratory systems, for example, carry on their tasks of homeostasis largely beyond our control and without affecting our consciousness. That is, with the exception of the respiratory system which, when required, can be subjugated to assist in conscious tasks such as speaking, and even to take part in non-respiratory acts, such as when we fix our ribcage to act as a framework against which the arms can work when lifting a heavy weight. When not being used in this conscious way the respiratory system carries on automatically, producing a minute ventilation which is appropriate for metabolism at that time and controlling levels of O₂, CO₂ and [H⁺] in the blood. It is the neural control of breathing that determines the pattern of that minute ventilation.

Minute ventilation is described by the equation:

where V_T and f are the tidal volume of each individual breath and the frequency of breathing, respectively. The equation tells us that a particular minute ventilation can be made up of an infinite variety of volumes and frequencies, from high frequencies and small volumes to low frequencies and large volumes. How and why we unconsciously ‘choose’ a particular pattern is the province of neural control of breathing.

Economy of energy is an evolutionary advantage, and the pattern of breathing chosen by our bodies is aimed at minimizing the amount of work we have to do to produce a particular minute ventilation. This work is directly related to the force exerted by the respiratory muscles, and it may be that we aim to minimize the tension in our respiratory muscles and/or minimize work. It seems that we get our respiratory systems to ‘resonate’. The resonant frequency for any particular minute ventilation depends on the value of that minute ventilation and the mechanical properties of the lungs (which have been dealt with in Chapters 3 and 4).

The process of matching pattern of breathing to the physical properties of the lungs can be likened to pushing someone on a swing. If you get the timing right it requires very little effort to keep the swing going, and the timing of the push depends on the physical properties of the swing (the length of the ropes).

Breathing originates in the brainstem (Fig. 10.1) which is made up of the medulla (which means marrow, as in bone marrow) and the pons (which means bridge), which connects the medulla to the rest of the brain. The neural basis of breathing within the brainstem is the central pattern generator, whose output to the respiratory muscles is modulated by numerous afferent inputs. Nerve impulses leave the central nervous system via the phrenic and intercostal nerves to bring about breathing by contraction of the respiratory muscles, mainly the diaphragm and intercostals. Other nerves to accessory muscles (e.g. of the larynx) synchronize their contractions with the phases of breathing.

Breathing is a rhythmic process, and this rhythm starts in a generator in the central nervous system appropriately called the central pattern generator. This produces a ‘rough and ready’ pattern of breathing. This pattern is modified and improved on in terms of efficiency by other regions of the brain, and by afferent inputs from receptors in the lungs and chest in particular, to produce a pattern which is efficient and can respond to changed conditions.

The basic pattern of breathing originates in that part of the brainstem which joins the spinal cord to the midbrain and cerebellum, shown in Figure 10.1 and in more detail in Figure 10.5. This region consists of the medulla and pons. If the brain above the medulla is removed (as by Transection II in Figure 10.5) the breathing pattern is remarkably normal. Breathing only ceases when connections between the medulla and spinal cord are cut. (There is some evidence that there are rhythm generators capable of producing breathing movements in the spinal cord itself, but this is a very minor effect occurring under very limited conditions.) The major generator of basic respiratory rhythm is situated in the medulla and is influenced by higher regions of the brain and by activity from receptors in other parts of the body.

The importance and site of the central pattern generator is clearly seen when people are executed by the process of hanging. The cause of death is not asphyxiation by the noose, as many people think, but rather a snapping of the spinal cord. This arrests breathing by cutting off the output from the medulla to the phrenic nerve and diaphragm.

The neural mechanisms that bring about the rhythm of breathing, consisting of an oscillation of inspiration followed by expiration, followed by inspiration and so on, are still not completely clear.

The idea that oscillating activity in the phrenic nerve could originate in a group of neurons that simply increase and decrease their activity is untenable because there is no reason why such a system should not simply ‘stick’ in the on or off position, which would result in the subject being ‘stuck’ in inspiration or expiration. This argument also applies to the old but persistent idea of two groups of neurons that produce either inspiration or expiration and reciprocally inhibit each other (Fig. 10.2).

All plausible models of the central pattern generator start with inspiratory neurons, because to generate a pattern of quiet breathing that is all that is required, expiration in quiet breathing being passive. Most of these models involve some sort of self-limiting negative feedback in the medulla, which operates an ‘off switch’ that limits inspiration (Fig. 10.3).

The durations of the two phases of breathing (inspiratory duration, t_I, and expiratory duration, t_E) are under independent control and so can change independent of each other, or both can change at the same time. Both are influenced by the volume of the lungs. Thus when breathing is accelerated t_E is the first to be shortened as V_T increases, with t_I remaining fairly constant until a threshold is reached, after which it begins to shorten significantly (Fig. 10.4).

These relationships are partly the result of the influence of peripheral mechanoreceptor activity on the rhythm generator (p. 131). These influences are still not completely understood, but some simple statements about the generation of the basic pattern of breathing can be made:

The relationship between V_T, t_I and t_E shown in Figure 10.4 is disrupted in disease. In chronic obstructive pulmonary disease, for example, overall minute ventilation (), which can be thought of as the physical expression of the neural drive to breathe, increases as the disease progresses. This offsets to some extent the decrease in efficiency caused by mismatching and changes in lung mechanics.

Within this increase in the pattern of breathing also changes. There is initially an increase in V_T, but as airways resistance increases as the disease progresses V_T decreases below normal. Frequency of breathing increases throughout the progress of the disease. The relationship of inspiratory duration to tidal volume (t_I to V_T) reflects the drive to breathe. The relationship between inspiratory duration and the total breath duration (t_I to t_Tot) reflects the way a single breath is divided up into the time to fill the lungs (t_I) and the time to empty back to the start position (t_E). These two phases are of course governed by the mechanical properties of the lungs and airways. Central drive to breathe must increase as airflow limitation progresses, reaching a maximum with respiratory failure. This increased drive effectively increases V_T until the increased work of breathing resulting from airflow limitation overpowers it and actually causes a fall in V_T. The only way the patient can now increase his minute ventilation is to increase his frequency of breathing. The problem with this strategy is that the expiratory airflow limitation produced by the disease demands that a greater proportion of each breath be devoted to expiration, and the fraction of each breath devoted to inspiration has to be reduced. Air trapping and a subjective sensation of relief when breathing at high volume (which holds the airways open in what has been termed ‘auto-PEEP’) causes the patient with severe COPD to breathe with a rapid shallow pattern at increased lung volumes. This is an inefficient pattern because, in addition to the airflow obstruction, it places the respiratory muscles at a mechanical disadvantage to such an extent that the increased work of breathing can exhaust the patient.

The different effects of damage at different levels of the brainstem, and the profound effects of cutting off afferent information in the vagus nerves by cutting or blocking activity in them, led early investigators to the erroneous idea that there were a variety of anatomical ‘centres’ in the pons. These centres, together with afferent activity from the peripheral nervous system (mainly in the vagus nerve), modified the activity of the medullary rhythm generator into an efficient pattern of breathing. The term ‘respiratory centre’ is incorrect if it is taken as implying that there are discrete anatomical bodies or regions of the brain that can be identified macroscopically or microscopically. Respiratory ‘centres’ are more correctly thought of as diffuse networks of neurons which are active together to bring about the same respiratory effect. Higher densities of neurons with a common purpose are, however, found in specific regions of the brain, and these regions can, if you wish, be considered as centres for ease of description.

Disconnecting the upper pons from the brainstem (Transection I in Fig. 10.5) removes the effect of the pontine respiratory group (PRG). The neurons that make up this centre are found in and around the nucleus parabranchialis medialis (NPBM).

When Transection II (Fig. 10.5) is made, with the vagi cut, breathing becomes slower and deeper. This rate-controlling, volume-limiting effect of the PRG is probably a result of the inspiratory neuron group of the medulla stimulating the PRG during inspiration. When stimulated in this way the PRG, after a short delay, sends inhibitory impulses back to the inspiratory neurons, cutting short their activity (and hence phrenic nerve discharge to the diaphragm) in a classic negative-feedback arrangement. Shortening one breath in this way means that the next breath can start earlier, that is, the frequency of breathing is increased and the volume of breaths reduced. It has been suggested that the PRG evolved as a means of mediating rapid shallow breathing (panting), initiated for thermal or emotional reasons by higher parts of the brain.

The effects of removing the PRG are only seen when the vagi are cut, because either the PRG or the vagi can produce the same effect of cutting short inspiration and limiting inspiratory volume.

Removing the PRG and cutting the vagi results in apneusis – long powerful inspiratory efforts interspersed with brief expirations (Greek, apneusis, without breath). This led to the suggestion that there was an apneustic centre in the lower pons. This idea is no longer popular, and the effects produced by this transection are thought to be the result of general damage rather than the disconnecting of a specific centre.

The pons connects the medulla to the higher regions of the brain. From direct electrical recording and observations of the clock-like unmodulated rhythms (slow deep patterns of breathing and apneusis) seen when these regions are damaged two functionally distinct groups of neurons in the medulla have been identified (Fig. 10.6).

The control of breathing described so far is automatic and independent of the brain above the pons. In unanaesthetized humans, however, the higher parts of the brain affect breathing both involuntarily, during emotion, hyperthermia, and probably during exercise, but also during conditions that require voluntary control of our respiratory muscles. These conditions may be primarily respiratory, speaking, blowing, whistling; or be for other purposes, such as when we lock the ribcage to provide a framework against which the arms can act, or when we increase intra-abdominal pressure during defecation. In all cases voluntary control is always bilateral: we cannot contract one half of our diaphragm or one half of our larynx. The origin of this control is probably the motor cortex, and the voluntary pathways bypass the PRG and rhythm generator in the brainstem to descend in the pyramidal tracts (Fig. 10.7).

These voluntary pathways can be destroyed independently of the involuntary pathways, for example by a stroke. Patients with such a stroke breathe normally, respond to reflex and chemical stimuli, but cannot voluntarily change their pattern of breathing; they will cough if their larynx is stimulated, but cannot cough on command. Very rarely, the opposite situation is seen, where the automatic pathways have been destroyed but the pyramidal tracts are left intact. The patient can breathe deliberately but not automatically, as in sleep. This condition is called ‘Ondine’s curse’ after the folk-tale of the mortal man who formed a liaison with an ondine or water-nymph. Her father, the king of the ondines, objected to this relationship (the way fathers do) and put a curse on the man that if he did not remember to keep his vital functions, such as heartbeat and breathing, going, they would stop. Of course, when he fell asleep he died. Fortunately, this condition, and ondines, are very rare.

The pathways for automatic and voluntary breathing are also separate in the spinal cord. The automatic pathways from the brainstem run in the anterior region near the outlets of the ventral roots, whereas the voluntary pathways pass down in more lateral areas.

Motor innervation of the diaphragm is very unusual. Unlike most other skeletal muscle the diaphragm is controlled almost entirely directly by motor neurons from the cervical region (‘C 3, 4 and 5 keep the diaphragm alive’). These motor neurons are also unusual in lacking feedback to Renshaw cells which, in other motor neurons, controlling other muscles of the body, cut short the afterdischarge of activity. Furthermore, as if to prevent fatigue of the diaphragm, the motor neurons ‘take turns’ to be active, and so stimulate different populations of muscle fibres in the diaphragm, which themselves ‘take turns’ to bring about successive inspirations.

The inspiratory motor neurons of the spinal cord are inhibited during expiration and the expiratory motor neurons are inhibited during inspiration. This reciprocal inhibition prevents the opposing inspiratory and expiratory muscles contracting simultaneously. Unlike opposing skeletal muscle groups this inhibition is not brought about by muscle-spindle reflexes but is the direct result of the dorsal and ventral respiratory groups activating specific motor neurons sequentially. This type of direct control of opposing inspiratory and expiratory muscle groups is necessary because the diaphragm is very poorly supplied with muscle spindles, which carry out this reciprocal inhibition in other opposing muscle groups. This absence of spindles probably explains why we do not feel fatigue of the diaphragm in the same way as we feel fatigue in other muscles. Some activity in the diaphragm continues into expiration, causing a ‘brake’ on the rate of expiration and extending its duration.

The larynx exhibits movements synchronized with breathing. These movements are brought about by the muscles of the larynx, which are innervated by the superior laryngeal and recurrent laryngeal nerves. These are branches of the vagus nerve. During inspiration the vocal folds are pulled apart and this reduces the resistance to airflow into the lungs. During expiration the vocal folds come together, slowing down expiratory flow. These movements are automatic – we are not conscious of them, and the nerve impulses that cause them arise from the rostral part of the ventral respiratory group in the nucleus ambiguus. We can, of course, consciously control our larynx, as in vocalization. The larynx is therefore almost a model of the control of respiration, with an automatic rhythm that can be consciously overridden.

The major expiratory muscles are those of the abdomen and the internal intercostal muscles. Unlike the diaphragm, these muscles are well supplied with spindles and behave more like other voluntary muscles, in that their contraction is caused by a combination of direct activation of extrafusal muscle fibres plus indirect activation via stimulation of the intrafusal fibres of muscle spindles which produces reflex contraction. Both internal and external intercostal muscles are active, but to only an insignificant degree, in quiet breathing.

The abdominal muscles only become involved when forced expiration is required during exercise or coughing. As exercise becomes more intense, or breathing becomes more laboured, as in disease, more accessory muscles, for example those of the shoulder girdle, are recruited. The muscles of the abdomen and chest are also involved in posture, locomotion, and in the movement of the arms in lifting heavy weights. During almost all of these activities, if not too extreme, it is possible to breathe and vocalize, which involves well controlled modification of airflow. This ability reaches its peak in opera singers, who seem to be able to sing in the most unusual positions.

Breathing is initiated by several steps, from the respiratory neurones in the brain via the spinal cord through peripheral nerves to the respiratory muscles. At each of these steps and the junctions between them the process is susceptible to disorder.

With respiratory muscle weakness comes a reduction in lung volume, particularly vital capacity and its components. This leads to a reduction in ventilation consequent on the weakness of inspiratory muscles. Equally, if not more important, is the weakness of expiratory muscles which reduces the efficiency of cough. This can result in inefficient clearance of mucus and frequent pulmonary infections.

Perhaps because of our unique power of speech, reflex control of breathing seems to be very different in human beings from that in other animals, and is not yet satisfactorily explained (at least not to the satisfaction of your author).

The control of breathing pattern in most mammals is profoundly influenced by inputs travelling in the tenth cranial nerves (the vagus nerves). These are two very large nerves which run in the neck, one either side and parallel to the trachea. These nerves carry information from other parts of the body, but the information coming from the lungs is of particular importance in the control of breathing. It originates in three types of receptor:

The stretch and rapidly adapting receptors send their information to the brain in large- and small-diameter myelinated fibres, receptively.

The C-fibre receptors, as their name suggests, send their information in small-diameter non-myelinated (C) fibres.

The two types of receptor that have been most investigated are slowly adapting (PSR) and rapidly adapting (RAR).

Rate of adaptation describes the way a receptor (in the lungs, or anywhere else in the body) responds to a stimulus.

If a constant stimulus is applied to a receptor it ‘gets used to it’ and the frequency of the receptor discharge (in action potentials per second) decreases even though the stimulus does not change (Fig. 10.8).

The rate at which this happens defines receptors as slowly adapting – i.e. the frequency of discharge returns only slowly to the rest frequency, and rapidly adapting – i.e. the frequency rapidly returns to normal.

Often the only or major symptom patients with lung disease complain of is dyspnoea. This condition is difficult to define, but is usually described as ‘difficult breathing’ or ‘air hunger’. The sensation of dyspnoea arises when there is a disproportion between the demand for ventilation and its supply. The sense of respiratory effort, detected as respiratory muscle tension, is not matched by changes in respiratory muscle length and therefore ventilation.

This history of dyspnoea is important in diagnosis. Dyspnoea of sudden onset is associated with an acute cardiopulmonary event – pulmonary embolism or left ventricular failure, for example. Chronic dyspnoea of slow onset is usually associated with respiratory disease – COPD, asthma, pulmonary fibrosis – but can be the result of cardiac diseases, particularly those that cause pulmonary venous congestion.

Reflex changes in pattern of breathing produced by non-pulmonary sources are important in humans and other animals. Those that transmit emotion, amicable or otherwise, to our peers are important in non-verbal communication, for example. The changes in breathing that occur when one stubs one’s bare toe, and the hilarity of such an event, changes the pattern of breathing in callous observers. Thus emotion and pain can change pattern of breathing via nervous pathways from higher regions in the central nervous system.