PHYSIOLOGY
Respiration: control of
ventilation
Emrys Kirkman
Abstract
Rhythmic ventilation is an automatic process controlled by the central
nervous system. Groups of cells in the brainstem, predominantly the
ventral and dorsal respiratory groups, are responsible for generating
basic respiratory rhythm. This basic rhythm is subject to modulation
by both conscious and reflex actions. In normal individuals the respira-
tory minute volume is set to closely regulate arterial carbon dioxide ten-
sion (PaCO2) at approximately 5.3 kPa, predominantly via a negative
feedback reflex involving the central chemoreceptors. A separate
group of chemoreceptors, the arterial chemoreceptors, are responsible
for initiating the increased ventilatory response to counter arterial hyp-
oxia, but a brisk response is not seen until PaO2 levels fall to approxi-
mately 8.0 kPa from the normal 13.3 kPa. Combined hypercarbia and
hypoxia (asphyxia) is a very powerful stimulus to breathe as the two in-
puts interact in a synergistic manner. The chemoreceptor reflexes can
be modified when the need arises (e.g. blockade of the respiratory part
of the arterial chemoreflex by the trigeminal reflex as part of the diving
response). Other reflexes such as the HeringeBreuer reflex contribute
to setting the balance between tidal volume and respiratory rate to
attain a given minute volume, although this reflex does not appear to
play a major role in humans at resting tidal volumes. Superimposed
on this ‘tonic’ control, additional protective reflexes (e.g. from receptors
in the upper airways) are recruited to protect the lungs and airways with
responses such as coughs and sneezes when required.
Keywords Aortic bodies; arterial chemoreceptors; carotid bodies;
central chemoreceptors; diving response; HeringeBreuer reflex; res-
piratory rhythm
Royal College of Anaesthetists CPD Matrix: 1A01
Ventilation is an automatic process that is controlled by the
central nervous system. The brainstem is a key area responsible
for the basic respiratory rhythm. This in turn is modulated by a
number of influences ranging from conscious alterations (origi-
nating form ‘higher centres’), tonic reflexes such as the chemo-
receptor reflexes regulating carbon dioxide levels of the blood
and preventing dangerous hypoxia to episodic reflexes such as
coughing and sneezing.
Origin of respiratory rhythm
The search for the area(s) of the brain and underlying mecha-
nisms responsible for setting the basic respiratory rhythm has
Emrys Kirkman PhD is a Principal Scientist (Physiologist) at Dstl,
Porton Down, and is an Honorary Senior Lecturer in Physiology at the
University of Durham, UK. Conflicts of interest: none declared.
ANAESTHESIA AND INTENSIVE CARE MEDICINE --:-
Please cite this article in press as: Kirkman E, Respiration: control of ventilation, Anaesthesia and intensive care medicine (2017), https://doi.org/
10.1016/j.mpaic.2017.09.006
Learning objectives
After reading this article you should be able to describe the:
C neural arrangement that produces the basic respiratory rhythm
C short-term control of blood carbon dioxide, oxygen, pH and
combined changes of these variables
C role of the peripheral chemoreceptors in controlling these var-
iables, and of the central chemoreceptors when chronic hypoxia
is present (at altitude)
C role of neural influences in regulating the chemoreceptor re-
flexes, as well as in regulating the rate and depth of ventilation
taken many years and is still continuing. Earlier ideas, based on
experiments where gross changes in patterns of breathing were
seen following transactions and focal lesions in the brainstem led
to the concept of multiple respiratory ‘centres’ (e.g. apneustic
and pneumotaxic centres), which interacted to provide respira-
tory rhythm. This type of concept is now outdated and the altered
pattern of breathing seen in these experiments is now viewed as
the result of general damage rather than disconnecting specific
centres.
Control of ventilation is affected by clusters of neurones that
are grouped in several distinct areas of the brainstem. Two such
areas in the medulla are called the ventral respiratory group
(VRG) and the dorsal respiratory group (DRG). Current thinking
is that rhythm generation is the domain of a group of tightly
clustered neurones in the VRG (specifically in the rostral
ventrolateral medulla) called the pre-Botzinger complex (PBC).
Although the PBC is itself a complex area containing a hetero-
geneous population of neurones, the PBC neurones thought to be
responsible for respiratory rhythmogenesis are all interneurones
in the sense that they have no axons projecting out of the
brainstem. PBC neurones then ‘drive’ groups of inspiratory
neurones. A strong piece of evidence suggesting that the PBC
neurones are indeed the respiratory rhythm generator is the
finding that a small focal lesion in this area produces an imme-
diate fatal apnoea. In addition, the VRG is viewed as being more
important than the DRG in the generation of respiratory rhythm
because the DRG has not been found in a number of species. The
inspiratory neurones have a major output into the spinal cord
and relevant cranial nuclei to influence the activity of relevant
motorneurones. Other areas in the brain that influence the VRG
(and DRG) are thought to modify the pattern of breathing,
especially depth. One of these areas is the apneustic centre or
Botzinger complex (Kolliker-fuse area) of the pons. Experimental
evidence suggests that disruption of any of these areas (other
than the PBC) will modify the depth or pattern of breathing, but
will not abolish rhythmogenesis.
Control of carbon dioxide tension in arterial blood
Carbon dioxide tension in arterial blood (PaCO2) is very closely
regulated around a set point (5.3 kPa or 40 mmHg in a healthy
individual) by a negative feedback system. A rise in PaCO2
stimulates a reflex increase in ventilation, excreting excess CO2
and returning PaCO2 to the set point. Conversely, a fall in CO2
attenuates tonic respiratory drive, resulting in a fall in ventilation
1 Ó 2017 Published by Elsevier Ltd.
 PHYSIOLOGY
and accumulation of CO2 until PaCO2 has again reached normal
levels (Figure 1). It is important to control PaCO2 within such
narrow limits because any alteration in PaCO2 will in turn modify
pH: hypercapnia causes a respiratory acidosis while hypocapnia
causes a respiratory alkalosis. Significant changes in extracellular
pH can lead to conformational changes in proteins such as en-
zymes and hence gross disturbances of body function.
The primary mechanism responsible for controlling PaCO2 is a
reflex originating from the central chemoreceptors. These che-
moreceptors comprise a group of cells in the medulla of the
brainstem, near the floor of the fourth cerebral ventricle
(Figure 1b). The central chemoreceptors are separate and distinct
from the cells generating respiratory rhythm, but the central
chemoreceptors do normally provide a tonic respiratory ‘drive’.
The predominant view is that the central chemoreceptors
respond primarily to hydrogen ions (Hþ) rather than directly to
CO2. In this scheme the CO2 in the cerebrospinal fluid (CSF) leads
to the generation of Hþ (Figure 1b) and stimulation of the che-
moreceptors. Because the CSF contains few buffering proteins a
change in CO2 levels here causes greater alterations in Hþ levels
than in the blood. Since CO2 can cross the bloodebrain barrier
with ease, any change in blood CO2 is fairly quickly reflected by
an alteration in CSF CO2, and ultimately a change in chemore-
ceptor activity. Although the same reaction (resulting in eleva-
tion of Hþ levels when CO2 levels rise) takes place in the blood,
a
b
Figure 1a is reproduced from Pocock G, Richards CD. Human Physiology:
the basis of medicine. Oxford: Oxford University Press, 2006.
Figure 1
ANAESTHESIA AND INTENSIVE CARE MEDICINE --:-
Please cite this article in press as: Kirkman E, Respiration: control of ventilation, Anaesthesia and intensive care medicine (2017), https://doi.org/
10.1016/j.mpaic.2017.09.006
the resulting Hþ in the blood is irrelevant for the
central chemoreceptors because Hþ cannot diffuse across the
bloodebrain barrier in healthy individuals.
Having given the detailed description of the response of the
central chemoreceptors to extracellular CSF Hþ, it should be
noted that some authors take a different view and argue that the
central chemoreceptors respond to a change in intracellular Hþ
levels resulting from an alteration in extracellular CSF CO2 levels.
Control of oxygen tension in arterial blood
Oxygen tension in arterial blood (PaO2) is also controlled by a
negative feedback reflex. However, PaO2 is not regulated as
sensitively as PaCO2 until there has been a considerable fall in
PaO2 from the normal levels of approximately 13.3 kPa to about
8.0 kPa, beyond which there is a powerful increase in ventilation.
Increases in PaO2 above 13.3 kPa have little effect on ventilation
(Figure 2), presumably because humans have not evolved with a
threat of hyperbaric oxygen. One reason why this sort of control
is appropriate for oxygen relates to the oxyhaemoglobin disso-
ciation curve. Most of the oxygen in normal blood is bound to
haemoglobin; hence, it is the amount and degree of saturation of
haemoglobin that dictates the amount of oxygen carried in blood
(CaO2). The oxyhaemoglobin dissociation curve is sigmoidal
(Figure 2). At normal arterial oxygen tensions haemoglobin is
about 98.5% saturated and we are on the upper flat part of the
curve. Consequently, falls in PaO2 to about 8.0 kPa have rela-
tively little effect on CaO2. However, falls below this level results
in our descent down the steep part of the oxyhaemoglobin curve
and consequently large, dangerous, falls in CaO2. Hence, the
need for a strong stimulus to breathing is only necessary if PaO2
falls below approximately 8.0 kPa.
The receptors responsible for monitoring arterial oxygen ten-
sion are the peripheral or arterial chemoreceptors found in the
carotid and aortic bodies. These organs receive an enormous blood
flow in relation to their mass of tissue. Estimations of the total ca-
rotid body blood flow range from 700 to 2000 ml/minute/100 g,
which far exceeds the organ’s metabolic demand and hence the
venous effluent from the carotid bodies normally have arterial-like
oxygen levels. The situation is likely to be similar in the aortic
bodies. As a consequence of this high blood flow the tissue derives
all of the oxygen needed for metabolic purposes from the plasma
rather than needing to extract oxygen from haemoglobin. As a
result they effectively monitor the tension of oxygen in plasma
(PaO2) rather than the content of oxygen in arterial blood (CaO2).
Normally, this arrangement is adequate to monitor blood oxygen
levels since the concentration of functional haemoglobin does not
usually vary acutely. However, in circumstances where functional
haemoglobin levels do drop (e.g. anaemia or carbon monoxide
poisoning) CaO2 can drop to dangerously low levels without
detection by the arterial chemoreceptors.
The arterial chemoreceptors are sensitive to alterations in
their blood flow. This is probably the reason why significant
haemorrhage quickly leads to increased ventilation as an intense
sympathetically-derived vasoconstriction in the arterial chemo-
receptors, coupled in severe haemorrhage with reduced arterial
blood pressure, causes a marked reduction in chemoreceptor
blood flow and hence stagnant hypoxia locally in the
chemoreceptors.
2 Ó 2017 Published by Elsevier Ltd.
 PHYSIOLOGY

chemoreceptors. These individuals are especially troublesome

The effect of acute hypoxia for the anaesthetist because they can stop breathing when some
8 anaesthetic agents (at quite low levels) blunt or abolish the
response to hypoxia. In addition, administration of oxygen, by
transiently correcting the hypoxia, can also remove the drive to
6
Impulses/second

breathe in these individuals.

4 Effects of acidosis and carbon dioxide on arterial

chemoreceptors and the whole body response to
metabolic acidosis
2
The arterial chemoreceptors can also respond to Hþ ions and
hence indirectly to CO2 via the Hþ ions it produces (Figure 1b).
0 10 20 30 40 kPa The effect of Hþ on the arterial chemoreceptors forms the basis of
0 100 200 300 mmHg the partial respiratory compensation for a metabolic acidosis.
PO2 The increase in arterial Hþ concentration (fall in arterial pH) due
50 100 to the addition of non-volatile acids from metabolic processes
leads to the stimulation of arterial chemoreceptors. The resultant
Ventilation (litres/minute)

Oxyhaemoglobin increase in respiratory drive elevates ventilation and causes the

% Saturation of haemoglobin
40 80
dissociation curve
excretion of more CO2 and hence a lowering of PaCO2. The
30 60 equation depicted in Figure 1b drifts to the left and Hþ levels fall,
restoring arterial pH towards normal. However, the compensa-
20 40 tion is never complete since the fall in PaCO2 also unloads the
central chemoreceptor and hence reduces the respiratory drive
10 Ventilation 20 from these receptors. The end result is a partial compensation,
with some degree of respiratory stimulation from the residual
0 0 acidosis acting on the arterial chemoreceptors to offset the
0 4 8 12 16 kPa reduced respiratory drive from the central chemoreceptors
because of the lowered PaCO2.
0 20 40 60 80 100 120 mmHg
PO2
Asphyxia: simultaneous elevation in PaCO2 and fall in PaO2
The effect of acute hypoxia on the discharge rate of arterial
chemoreceptor afferent nerve fibres (upper graph) and Asphyxia is a very powerful stimulus to increase ventilation. This
pulmonary ventilation compared with the oxyhaemoglobin
dissociation curve (lower graph).
is because both arterial and central chemoreceptors are activated.
The resulting interaction leads to a very marked increase in
Reproduced from Pocock G, Richards CD. Human Physiology:
ventilation (multiplicative effect due to simultaneous activation
the basis of medicine. Oxford: Oxford University Press, 2006.
of the two reflexes; Figure 1a).
Figure 2
Modulation of the arterial chemoreflex by trigeminal and
Longer term hypoxia laryngeal reflexes

When hypoxia persists the initial ventilatory stimulation (driven Stimulation of receptors in the face, nose or pharynx, which relay
by the arterial chemoreceptors) wanes for 3e5 minutes (a phe- in the trigeminal nerve, or receptors in the larynx that relay in the
nomenon called hypoxic ventilatory decline), partly due to the superior laryngeal nerve (SLN) cause a reflex apnoea. The reader
resulting hypocapnia. However, the level of ventilation does may be aware of this reflex when standing in the shower and
persist above the pre-hypoxic level. Continued hypoxia then re- turning on the cold water by mistake; when the cold water hits
sults in a period of acclimatization whereby CSF Hþ levels are the head and face it often feels very difficult to breathe. The re-
returned to normal by active transport of Hþ across the blood flex apnoea resulting from strong stimulation of the trigeminal
ebrain barrier and the kidneys excreting additional bicarbonate nerve and SLN is very powerful and can completely block the
ions (HCO3) to restore blood pH. The detail of these processes is increase in ventilation initiated by the arterial chemoreceptors
given in the article on the effects of altitude (see volume 8, issue (inhibitory interaction). When this occurs the primary cardio-
11) and the end result is a level of ventilation sufficient to pre- vascular response to peripheral chemoreceptor stimulation is
vent hypoxia. It is important to recognize that one group of pa- unmasked: a powerful vagally mediated bradycardia and an in-
tients with chronic pulmonary disease, sometimes rather crease in sympathetic tone to the vasculature causing an increase
unkindly describes as ‘blue bloaters’, do not make this in peripheral resistance. This forms the basis of the diving
compensation. (Blue bloaters are so called because they are blue response. The survival advantage of the diving response is
due to the cyanosis of hypoxia and bloated because of congestive obvious. When an air-breathing animal submerges its head
heart failure.) ‘Blue bloaters’ adapt to a high PaCO2 and, unlike under water it cannot (must not) breathe. It would be disastrous
other individuals, the majority of the drive to breathe in these should the activation of the arterial chemoreflex succeed in
patients comes from hypoxia acting on the arterial increasing ventilation under this circumstance, hence the

ANAESTHESIA AND INTENSIVE CARE MEDICINE --:- 3 Ó 2017 Published by Elsevier Ltd.

Please cite this article in press as: Kirkman E, Respiration: control of ventilation, Anaesthesia and intensive care medicine (2017), https://doi.org/
10.1016/j.mpaic.2017.09.006
 PHYSIOLOGY
advantage of the inhibitory interaction of the respiratory com-
ponents of the two reflexes (trigeminal and arterial chemore-
ceptor). The advantage of the bradycardia and selective
vasoconstriction is thought to be a reduction in oxygen con-
sumption and a preferential direction of blood flow to the organs
most dependent on a constant supply of oxygen.
An interaction between arterial chemoreceptor and trigemi-
nal/laryngeal reflexes can arise clinically when the larynx is
stimulated. However, it is rare to see the profound bradycardia
during intubation of a patient, presumably because patients are
very well oxygenated before the procedure is initiated, hence
there is no chemoreceptor stimulation. However, marked
bradycardia has been reported in tetraplegic patients with high
cervical lesions when they are disconnected from a ventilator for
routine aspiration of the upper airways. Furthermore, it is
possible that this reflex interaction can cause death in extreme
cases (e.g. ‘dry drowning’) when a corpse is recovered from the
water with no evidence of water inhalation on post mortem ex-
amination. Here, it is thought that the bradycardia became so
powerful as to cause cardiac arrest.
Regulating the rate and depth of ventilation
Alveolar ventilation (V_ A , approximately 4 litres/minute at rest) is
total ventilation rate (V_ T ) minus deadspace ventilation.
V_ A ¼ (VT e VD)  respiratory rate
Equation 1: relation between alveolar ventilation rate (V_ A ),
tidal volume (VT), physiological deadspace volume (VD) and
respiratory rate.
Because deadspace volume is relatively constant in an individ-
ual it can be seen from Equation 1 that alveolar ventilation can be
achieved by an infinite number of combinations of tidal volumes
and respiratory rates, ranging from very slow deep breaths to very
rapid shallow breaths. However, the work required to achieve a
given alveolar ventilation varies with the combination of tidal
volume and respiratory rate chosen, with both extreme examples
noted above being associated with increased work of breathing.
With deep breathing much work has to be expended against the
elastic recoil of the lungs. At the other extreme, very rapid shallow
breathing requires much work against airway resistance to gas flow
and is wasteful of effort since a greater proportion of each breath
represents deadspace ventilation. The precise mechanism whereby
an optimal balance is achieved is unclear, but it probably depends
on the interaction of a number of sensory factors.
One reflex, reported to play a part in balancing rate and depth
of ventilation, is the HeringeBreuer reflex. The afferent pathway
originates in slowly adapting mechanoreceptors associated with
airway smooth muscle of the tracheobronchial tree, which are
ANAESTHESIA AND INTENSIVE CARE MEDICINE --:-
Please cite this article in press as: Kirkman E, Respiration: control of ventilation, Anaesthesia and intensive care medicine (2017), https://doi.org/
10.1016/j.mpaic.2017.09.006
found as far peripherally as the respiratory bronchioles and
alveolar ducts. They relay information to the CNS via myelinated
afferent fibres in the vagus. The HeringeBreuer reflex terminates
inspiration when the receptors are sufficiently stimulated. They
are undoubtedly important in controlling depth of resting venti-
lation in some animals such as the cat, since loss of the reflex by
vagotomy leads to deeper, slower breathing. However, vagotomy
does not have this effect in humans and hence the Hering
eBreuer reflex probably does not play an important role in
setting resting tidal volume in people, but this may become
important when breathing at increased tidal volumes.
Other reflexes influencing ventilation
A whole host of other reflexes also modify ventilation. These
include the pulmonary J receptors found near the pulmonary
capillaries and alveoli. These receptors may be responsible for a
reflex apnoea seen immediately after exposure of the chest to a
blast shock wave associated with an explosion. Under less dra-
matic circumstances the pulmonary J receptors can lead to rapid
shallow breathing and possibly dyspnoea associated with pul-
monary oedema. Other protective reflexes such as the irritant
receptors (rapidly adapting mechanoreceptors) in the pulmonary
airways can give rise to both rapid shallow and deep augmented
breaths. The irritant receptors are thought to play a role in
reversing the slow progressive collapse of parts of the lungs
during prolonged periods of quiet breathing by initiating periodic
augmented breaths every 5e20 minutes in resting individuals.
Finally, other types of irritant receptors are found in the nose and
larynx, which give rise to protective reflexes resulting in
coughing and sneezing, respectively. A
FURTHER READING
Angell-James JE, Daly MB. Some aspects of upper respiratory tract
reflexes. Acta Otolaryngol 1975; 79: 242e52.
Gray PA, Rekling JC, Bocchiaro CM, Feldman JL. Modulation of res-
piratory frequency by peptidergic input to rhythmogenic neurons in
the preBotzinger complex. Science 1999; 286: 1566e8.
Mathias CJ. Bradycardia and cardiac arrest during tracheal suction e
mechanisms in tetraplegic patients. Eur J Intensive Care Med 1976;
2: 147e56.
Pandit JJ. Effect of low-dose inhaled anaesthetic agents on the
ventilatory response to carbon dioxide in humans: a quantitative
review. Anaesthesia 2005; 60: 461e9.
Richter DW, Spyer KM. Studying rhythmogenesis of breathing: com-
parison of in vivo and in vitro models. Trends Neurosci 2001; 24:
464e72.
4 Ó 2017 Published by Elsevier Ltd.