Respiration Physiology 109 (1997) 197 – 204

Frontiers Review
Brain, breathing and breathlessness

A. Guz *
Charing Cross and Westminster Medical School, London, W6 8RF, UK

Accepted 17 June 1997

Abstract

This review attempts to summarize: (i) evidence on how man voluntarily or behaviourally (as in speech) alters
breathing; and (ii) evidence on how the breathlessness induced by CO2 inhalation, is perceived. The application of
new methods to study these problems, e.g. functional brain imaging and transcranial focal brain stimulation, is
summarized. Studies of patients with specific neurological lesions have shed considerable light in this area. The key
requirement for the ponto-medullary respiratory oscillator to be both ‘intact’ and ‘responsive’ for the perception of
CO2-induced air hunger is emphasized. We are ignorant as to how the voluntary/behavioural control system interacts
with the automatic system at any site above the final common pathway of the respiratory anterior horn cells in the
cervical and thoracic spinal cord. The opportunities for further work are outlined. © 1997 Elsevier Science B.V.

Keywords: Brain; Higher centers; Perception; Control of breathing; Breathlessness; Mammals; Man; Speech;
Breathing

1. The neurology of automatic breathing Automatic breathing originates in the ponto-

The objective of this presentation is to review
recent experimental evidence in man on two sepa-
rate but related subjects: (i) mechanisms underly-
ing voluntary or behavioural control of breathing;
and (ii) mechanisms underlying the sensation of
breathlessness. In order to understand the story, it
is necessary to examine Fig. 1.
* Tel.: +44 181 8467337; fax: + 44 181 8467326; e-mail:
a.guz@cxwms.ac.uk
medullary respiratory oscillator from which a de-
scending bulbo-spinal projection synapses with
the anterior horn cells in the cervical and thoracic
spinal cord with projections to the respiratory
muscles to cause rhythmic breathing. We now
know from studies in the isolated brain-stem/high
cervical cord preparation from the guinea-pig that
the ponto-medullary respiratory oscillator gen-
uinely oscillates automatically and can function
without any peripheral feedback but only in the

0034-5687/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved.

PII S 0 0 3 4 - 5 6 8 7 ( 9 7 ) 0 0 0 5 0 - 9
198 A. Guz / Respiration Physiology 109 (1997) 197–204

presence of a normal PCO2 and pH, provided the Fig. 1 shows a descending cortico-spinal tract
preparation is kept well oxygenated. This auto- on the left side making synapses with the anterior
matic breathing is what we see in man asleep horn cells of the right diaphragm and intercostals.
(especially in the deep non-rapid-eye-movement Evidence for this tract has been obtained in ani-
(REM) state) or anaesthetized. Proprioceptive af- mals since 1958 (Colle and Massion, 1958;
ferent input from the lungs via the vagus nerve Aminoff and Sears, 1971; Planche, 1972) by stim-
with a synapse in the nucleus of the tractus soli- ulating the frontal lobes and recording an output
tarius can either inhibit or excite the respiratory
oscillator, but there is very little evidence that this
system, thoroughly studied over the last century
(Ullmann, 1970), controls breathing in man with
normal lungs and at rest. Vagal input does be-
come significant in driving breathing, particularly
respiratory frequency, when the lobes of the lungs
are the seat of either inflammation at alveolar
level, collapsed or water-logged. There is also
little evidence that afferent input from the di-
aphragm and intercostal muscles has any influ-
ence on breathing in normal humans at rest.
Afferent input from the carotid body (and to a
minimal extent, the aortic body) excited by hy-
poxia, hypercapnia or acidosis acts via synapses in
the nucleus of the tractus solitarius to excite or
inhibit the respiratory oscillator. Significant chem-
ical control of the oscillator is also exerted at the
medulla via its ventral surfaces or deep within its
structure.

2. Anatomy and physiology underlying

voluntary/behavioural breathing

The schematic outline so far does not explain

how man can take a breath at will, nor how
expiratory airflow can be perfectly controlled to
produce phonation. It also does not explain how
emotions can affect breathing. Most importantly,
it does not explain how breathing increases the
‘right amount’ during exercise where there would
appear to be no error signal to regulate breathing,
in the absence of metabolic acidosis with more
severe exercise. Such failure to explain essentially
error-free regulation of a bodily function such as
the cardio-respiratory response to exercise is now
clear after intensive investigation over the last 25
years (Somjen, 1992). As a result, concepts such as
‘central command’ in physiological studies or ‘feed
forward’ in bioengineering studies, have arisen.
Fig. 1. Neurology of breathing. The figure is not drawn to
scale. R, right; L, left; GP, glossopharyngeal nerve; CB,
carotid body; AB, aortic body; V, vagus nerve; P, phrenic
nerve; I, intercostal nerves; D, diaphragm; LM, leg muscle; pH
CO2, medullary chemosensitivity to H + and CO2; IN and EN,
inspiratory and expiratory neuronal pools constituting pon-
tomedullary respiratory oscillator; M, medulla; NTS, nucleus
of tractus solitarius; PRC, pontine respiratory complex. Two
corticospinal tracts are shown. Right, descending tract crosses
pyramidal decussation to synapse in anterior horn cells of left
L1 − 5 controlling leg movement. Dotted lines to medullary
respiratory oscillator symbolically represent irradiation from
impulse traffic in the corticospinal tract. Left: descending
corticospinal track decussates to synapse on to anterior horn
cells of C3 − 5 and T1 − 12 controlling voluntary or behavioural
inspiration and expiration; dotted line to the medullary res-
piratory oscillator represents a probable input from this corti-
cospinal tract.
 A. Guz / Respiration Physiology 109 (1997) 197–204
from the contralateral phrenic motoneurones with
a short latency. Evidence in humans was obtained
as long ago as 1936 by Foerster (1936); stimula-
tion of the exposed motor cortex in humans dur-
ing neurosurgery under local anaesthesia localised
a site anterior to the ‘chest’ area of Foerster’s
homunculus where stimulation resulted in a hic-
cup! More recently transcranial electrical stimula-
tion at the vertex in conscious humans (Gandevia
and Rothwell, 1987) elicited diaphragmatic con-
tractions with compound action potentials occur-
ring 14 ms after stimulation.
Further studies by our own group (Maskill et
al., 1991) with transcranial magnetic stimulation
using a coil with some focussing facility over one
motor cortex have shown that compound action
potentials recorded over the contralateral di-
aphragm were maximal when the centre of the
coil (where the highest magnetic flux was present)
was placed 2–3 cm anterior to the inter-auricular
plane and 3 cm to one side of the midline. This is
approximately where Foerster found stimulation
effective. The compound action potential was pre-
dominantly contralateral, although a small re-
sponse was seen on the ipsilateral side; ultrasonic
imaging confirmed that the diaphragmatic twitch
occurred contralaterally and not ipsilaterally. This
result raises the question of how the two di-
aphragms move together? Does the integration
occur within the cervical cord or via the corpus
callosum? Another question that can be asked is
how the voluntary/behavioural system interacts (if
at all) with the automatic ponto-medullary oscilla-
tor?
3. Imaging the brain to elucidate neuronal areas
concerned with voluntary/behavioural breathing
Positron emission tomography (PET) allows the
non-invasive measurement of regional cerebral
blood flow and has been used to define areas of
increased neural activity associated with specific
motor tasks in conscious humans. Regional blood
flow increases when regional O2 consumption in-
creases, particularly in association with local in-
creases in synaptic activity. This increase in blood
199
flow can be ‘marked’ by the use of isotopes that
accumulate in such regions, following the distri-
bution of blood flow; H15 2 O with a half life of 2.1
min, given intravenously is an effective ‘marker’.
Positrons are generated in a known configuration
that permits anatomical localization with an ap-
propriate scanning system. Problems encountered
include the requirement to normalize for any as-
sociated changes in global blood flow. Success has
depended on designing an experimental paradigm
with appropriate controls, so that images of con-
trasts (between experiment and control) can be
generated that eliminate factors including ‘noise’
common to both.
The studies to be summarized in this presenta-
tion have all been done in this way and have also
averaged the contrasts within a subject and across
subjects. Recently, with improved technology, it is
becoming possible to rely on single-subject stud-
ies. It is difficult to be certain about the sensitivity
of the technique. Furthermore, because of the
lack of good animal studies, it is impossible to be
certain about any relationship between the num-
ber and firing rate distribution in any group of
motor or sensory neurones involved and the de-
gree of ‘activation’ seen in a contrast.
Our group (Colebatch et al., 1991) has used this
technique to identify areas of neuronal activation
associated with volitional inspiration. Such areas
were found: (i) bilaterally in the primary motor
cortex—superolaterally (where Foerster, 1936
and we, Maskill et al., 1991) had previously found
optimal sites of stimulation for the diaphragm);
(ii) in the right premotor cortex and in the supple-
mentary motor area; and (iii) in the cerebellum.
These results are analogous to what has been
found with similar studies in voluntary limb
movement.
We have further defined areas of neuronal acti-
vation with volitional expiration against a
threshold load together with passive inspiration
from a ventilator (Ramsay et al., 1993). Similar
areas were activated, as was found in the study on
volitional inspiration, with one notable addition:
the right and left primary motor cortices showed
increases in regional cerebral blood flow over
large areas ventro-laterally well down towards the
200 A. Guz / Respiration Physiology 109 (1997) 197–204
Sylvian fissure. It is of extraordinary interest that
vocalization (or speech arrest) was obtained by
Penfield and Rasmussen (1950) in these areas
when stimulating the exposed motor cortex in
conscious subjects during neurosurgery.
One cannot speak without control of expiratory
airflow together with appropriate regulation of
the laryngeal aperture. The sophisticated nature
of this control in man has been demonstrated by
Draper et al. (1959) during a continuous utterance
(counting) when coordinated activity of expira-
tory and inspiratory muscles resulted in the pro-
duction of a constant subglottic pressure while the
thorax and lung were becoming smaller. There
has been a great deal of confusion about the areas
of the brain involved in ‘speaking’, presumably
because the act of speaking demands simulta-
neous control of respiratory, laryngeal and articu-
latory systems, together with the involvement of
language centres in the dominant hemisphere. Our
group (Murphy et al., 1996) have designed a
paradigm that minimizes language processing.
Any such processing, if present, is common to all
the experimental conditions. The phrase ‘Buy
Bobby a Poppy’ has been continuously repeated
in four ways while scanning the brain: (A) spoken
aloud, (B) mouthed silently, (C) vocalized without
articulation (’ah-ahah-ah-ahah’ — vocalized to the
same rhythm), and (D) thought silently. Contrast-
ing images from conditions (A) versus (C) and (B)
versus (D) highlighted areas associated with artic-
ulation alone, since control of breathing for
speech was common in the contrasts. Contrast of
images from conditions (A) versus (B) and (C)
versus (D) highlighted areas associated with the
control of breathing for speech and vocalization
since articulation was common in the contrasts.
Bilateral symmetrical activations were found lat-
erally in the sensorimotor cortex to be concerned
with articulation. Similar bilateral activations
were found in the sensorimotor and motor cor-
tices concerned with control of breathing for
speech and vocalization. They were close to, but
distinct from, the areas activated with articula-
tion. Broca’s area in the inferior frontal gyrus of
the dominant hemisphere was never seen to be
activated. These findings emphasize the bilateral
nature of the cerebral control of ‘speaking’ with-
out language processing. Furthermore the areas
activated are similar to the lateral areas defined by
Penfield and Rasmussen (1950) and Ramsay et al.
(1993).
4. Is the motor cortex involved in breathing
control during exercise?
We still cannot agree on why breathing in-
creases on exercise! The question takes us back to
the proposal of Krogh and Lindhard (1913) that
motor cortical activation to exercising muscles
might induce increases in breathing via irradiation
of either brainstem respiratory centres (Fig. 1, see
dotted lines from right corticospinal tract to pon-
tomedullary oscillator) or of cortical areas con-
trolling respiratory muscles. Could these cortical
areas, described in Section 3, be shown to be
active during exercise? Recently, we (Fink et al.,
1995) have been able to show, with modest leg
exercise in normal adult subjects lying with their
heads in a PET scanner, that there is indeed
evidence of activation in the left and right supero-
lateral primary motor cortices in areas previously
shown to be associated with volitional breathing.
The activations are still present during the early
post-exercise period when ventilation is still ele-
vated but declining. Such results have suggested
motor cortical involvement in exercise-related hy-
perpnoea. Is this one of the forms of irradia-
tion—at cortical level—suggested by Krogh and
Lindhard (1913)? Or is this a learned response
providing an appropriate amount of central com-
mand?
5. The ‘Curse of Ondine’ and the ‘Locked-in
Syndrome’: clinical lessons
These studies provide some physiological basis
for the voluntary/behavioural aspects of breathing
as opposed to the automatic ponto-medullary
breathing. Fig. 1 shows a dotted line from the left
corticospinal tract for volitional breathing which
represents a probable input into the respiratory
oscillator. Experiments in the cat (Orem and Net-
tick, 1986; Orem, 1988) suggest that voluntary or
 A. Guz / Respiration Physiology 109 (1997) 197–204
behavioural influences are integrated within the
brainstem areas that house the automatic ponto-
medullary oscillator. In humans, evidence comes
from clinical studies of patients with defined
brainstem lesions (Plum and Leigh, 1981), or very
high cervical cord lesions (Lahuerta et al., 1992;
Severinghaus and Mitchell, 1962). The syndrome
of particular interest is that of ‘Ondine’s Curse’;
these patients lack automatic breathing when
drowsy or asleep, but breathe apparently nor-
mally when awake and can breathe normally to
command. Evidence strongly suggests that bulbo-
spinal fibres run separately and more anterolater-
ally in the cord than the relevant corticospinal
fibres. There is, however, no evidence in humans
about integration of corticospinal control with
bulbospinal control at the brainstem level. What
happens to the pontomedullary respiratory oscil-
lator when a person takes a breath without chang-
ing the arterial PCO2?
The congenital Ondine’s curse syndrome has
now been well studied in children. These patients
have ineffective chemical regulation of breathing
and either severely hypoventilate or do not
breathe at all during sleep. Awake, the level of
spontaneous breathing is adequate and during
steady-state moderate exercise — below the level
of lactate accumulation — ventilation is raised ap-
propriately so that the prevailing level of PCO2 is
maintained (Shea et al., 1993b). We do not know
why this happens! What is the meaning of needing
to be awake to breathe? Are these children breath-
ing when awake, at rest or while exercising, with
their corticospinal tracts?
Breathing at rest has a considerable degree of
variability from moment to moment. A lesion in
the ventral pons and lower midbrain involving
motor tracts to the rest of the body will remove
all voluntary control of muscle movement, except
for elevation of the eyes; this is the ‘locked-in
syndrome’ (Heywood et al., 1996; Munschauer et
al., 1990). Sensation is intact. Breathing is normal
but very regular, maintaining a normal PCO2; vol-
untary effort to change breathing has no effect
but emotion will disrupt breathing. Maintenance
of emotional influences on breathing in the ab-
sence of voluntary control shows that emotional
(presumably limbic) pathways to the brainstem
201
are anatomically separate from the corticobulbar
pathways. Such patients give us the opportunity
to study the function of the ‘isolated’ pon-
tomedullary oscillator and its intact bulbospinal
tract to the muscles of breathing. Ventilatory
sensitivity to inhaled carbon dioxide is normal
with associated breathlessness. However, if resting
PCO2 is lowered (with mechanical ventilation) by
as little as 1–3 mmHg, breathing ceases (Hey-
wood et al., 1996). This occurs in intact humans
during deep non-REM sleep (Datta et al., 1991).
It does not occur in intact humans awake, al-
though breathing becomes irregular in both rate
and depth while PCO2 returns to normal levels
(Corfield et al., 1995a). Do these findings imply
that higher-centre (cortical) input into the brain-
stem oscillator is occurring normally when awake
at rest to keep breathing stable but irregular? Is
this the wakefulness ‘drive to breathe’ (Fink,
1961)?
6. The sensation of breathlessness; focus on ‘air
hunger’ induced by CO2
An up-to-date review of this subject has been
published as a multi-author volume in 1996
(Adams and Guz, 1996). It would not be appro-
priate to summarize this here. Breathlessness is
not a uniform sensation, it has many different
qualities (Simon et al., 1990). I shall focus on the
sensation of ‘air hunger’ or the ‘urge to breathe’.
This sensation typically occurs in subjects given
carbon dioxide to breathe or made to exercise
when breathing discomfort is further described by
such phrases as ‘could not breathe fast or deep
enough’ or ‘could not get enough air’ or ‘suffocat-
ing’. There would appear to be an enhanced res-
piratory drive normally manifested by an increase
in neural output to the muscles of breathing.
Increasing clinical evidence over the last few
years suggests that it is not the neural output to
the muscles of breathing that is particularly rele-
vant to the genesis of the air hunger sensation.
The key requirement would appear to be an intact
brainstem automatic respiratory controller that
can respond to stimulation. The evidence is shown
below.
202 A. Guz / Respiration Physiology 109 (1997) 197–204
1. Totally curarized unsedated subjects have
been mechanically ventilated while end-tidal (alve-
olar) PCO2 has been varied by changing the in-
spired PCO2 (Banzett et al., 1990). When the
end-tidal PCO2 was raised from a median of 35
mmHg, at which the subjects were comfortable, to
a median of 44 mmHg, all reported severe air
hunger. The subjects reported no fundamental
difference in the sensation during paralysis or, as
a control, before paralysis. Both experiment and
control were done while ventilation was kept con-
stant with a mechanical ventilator.
2. Ventilator-dependent quadriplegics with the
lesion as high as C1 − 2 could sense the typical ‘air
hunger’ when the end-tidal carbon dioxide was
raised as in the paralysis study above (Banzett et
al., 1989). Here, the brainstem respiratory oscilla-
tor is presumably intact, but its activity, at rest or
when stimulated, cannot be transmitted via the
bulbospinal fibres to the anterior horn cells of the
respiratory musculature.
3. Children with the congenital central hy-
poventilation syndrome (curse of Ondine, see Sec-
tion 5) neither have a ventilatory response to
hypercapnia nor experience any air hunger sensa-
tion (Shea et al., 1993a,b). It should be noted that
the entire brain, including the sensory cortex, is
experiencing hypercapnia, but there are no res-
piratory sensations. These children can also vol-
untarily hold their breath for an inordinate length
of time without any apparent urge to breathe.
4. A patient with the locked-in syndrome (Hey-
wood et al., 1996) with intact and regular auto-
matic breathing, has a normal ventilatory
response to the inhalation of carbon dioxide with
the normal associated air hunger. The sensory
pathways are intact.
Although this clinical evidence is weighty, it is
essential to ask how the activation of an intact
pontomedullary respiratory oscillator can be
sensed. Recent work in the cat (Chen et al., 1992)
has shown respiratory-associated firing of both
midbrain, hypothalamic and thalamic neurons re-
lated to respiratory drive and entirely dependent
on an intact connection with the medulla. There
may, therefore, be an afferent pathway from the
oscillator up to the sensory cortex, but this has
not yet been demonstrated. A recent PET study
(Corfield et al., 1995b) has now identified areas of
probable neuronal activation extending from the
upper brainstem, up through the midbrain, hypo-
thalamus and thalamus in humans, while breath-
less as a result of elevated end-tidal PCO2 raised
from 40 to 50 mmHg. In addition, activation of
the limbic system (cingulate gyrus, parahippocam-
pus, hippocampus, fusiform gyrus and insula) was
present; this may be relevant to the unpleasant-
ness of the air hunger sensation. Activation of the
limbic system, particularly the anterior cingulate
gyrus, has been noted when heat pain becomes so
intense as to be unpleasant (Talbot et al., 1991).
7. Focus on sensations induced from the lung
There is clear need to scan the brain for activa-
tions that are likely to occur with afferent vagal
input from the lungs. This could include a study
of the ‘raw’ sensation in the chest provoked by
irritating the airways and also the sensation that
may be evoked by activation of c-fibres from the
lung, as is likely to occur in pulmonary oedema.
The difficulties of doing such studies would ap-
pear to be immense, but were they to be accom-
plished, richly rewarding.
8. Conclusion
We are at a very early stage in trying to under-
stand how higher brain centres can control
breathing and experience breathlessness in hu-
mans. New methods have now made these studies
possible.
Acknowledgements
I am deeply grateful to my colleagues in the
Department of Medicine at Charing Cross and
Westminster Medical School (especially Professor
Lewis Adams) and also to my colleagues at the
MRC Cycloctron Unit, Royal Postgraduate Med-
ical School, London, for their close interaction
with me over many years. Thanks are also ex-
 A. Guz / Respiration Physiology 109 (1997) 197–204
pressed to the Wellcome Trust for their generous
financial support which made this work possible.
References
Adams, L., Guz, A. (Eds.), 1996. Respiratory Sensation. Lung
Biology in Health and Disease, vol. 90. Marcel Dekker,
New York.
Aminoff, M.J., Sears, T.A., 1971. Spinal integration of seg-
mental, cortical and breathing inputs to thoracic respira-
tory motoneurones. J. Physiol. 215, 557–575.
Banzett, R.B., Lansing, R.W., Reid, M.B., Adams, L., Brown,
R., 1989. Air hunger arising from increased PCO2 in me-
chanically ventilated quadriplegics. Resp. Physiol. 76, 53–
67.
Banzett, R.B., Lansing, R.W., Brown, R., Topoulos, P.,
Yager, D., Steele, S.M., Londono, B., Loring, S.H., Reid,
M.B., Adams, L., Nations, C.S., 1990. ‘Air Hunger’ from
increased PC02 persists after complete neuromuscular block
in humans. Respir. Physiol. 80, 1–18.
Chen, Z., Eldridge, F.L., Wagner, P.G., 1992. Respiratory
associated thalamic activity is related to level of respiratory
drive. Respir. Physiol. 90, 99–113.
Colebatch, J.G., Adams, L., Murphy, K., Martin, A.J.,
Lammertsma, A.A., Tochon-Danguy, H.J., Clark, J.C.,
Friston, K.J., Guz, A., 1991. Regional cerebral blood flow
during volitional breathing in man. J. Physiol. 443, 91–
103.
Colle, J., Massion, J., 1958. Effet de la stimulation du cortex
moteur sur l’activite electrique des nerfs phreniques et
medions. Arch. Physiol. Biochem. 66, 496–514.
Corfield, D.R., Morrell, M.J., Guz, A., 1995a. The nature of
breathing during hypocapnia in awake man. Resp. Physiol.
101, 145 – 159.
Corfield, D.R., Fink, J.R., Ramsay, S.C., Murphy, K., Harty,
H.R., Watson, J.D.G., Adams, L., Frackowiak, R.S.J.,
Guz, A., 1995b. Evidence for limbic system activation
during CO2-stimulated breathing in man. J. Physiol. 488,
77– 84.
Datta, A.K., Shea, S.A., Horner, R.L., Guz, A., 1991. The
influence of induced hypocapnoea and sleep on the endoge-
nous respiratory rhythm in humans. J. Physiol. 440, 17–33.
Draper, M.H., Ladefoged, P., Whitteridge, D., 1959. Respira-
tory muscles in speech. J. Speech Hear Res. 2 (1), 16–27.
Fink, B.R., 1961. Influence of cerebral activity in wakefulness
on regulation of breathing. J. Appl. Physiol. 16, 15–21.
Fink, G., Adams, L., Watson, J.D.G., Innes, J.A., Wuyam, B.,
Kobayashi, I., Corfield, D., Murphy, K., Jones, T., Frack-
owiak, R.S.J., Guz, A., 1995. Hyperpnoea during and
immediately after exercise in man: evidence of motor corti-
cal involvement. J. Physiol. 489, 663–675.
Foerster, O., 1936. Motorische Felder und Bahten. In: Bumke,
O., Foerster, O. (Eds.), Handbuch der Neurologie.
Springer, Berlin, pp. 50–51.
203
Gandevia, S.C., Rothwell, J.C., 1987. Activation of the human
diaphragm from the motor cortex. J. Physiol. 384, 109 –
118.
Heywood, P., Murphy, K., Corfield, D.R., Morrell, M.J.,
Howard, R.S., Guz, A., 1996. Control of breathing in man;
insights from the ‘locked-in’ syndrome. Resp. Physiol. 106,
13 – 20.
Krogh, A., Lindhard, J., 1913. The regulation of respiration
and circulation during the initial stages of muscular work.
J. Physiol. 47, 112 – 136.
Lahuerta, J., Buxton, P., Lipton, S., Bowsher, D., 1992. The
location and function of respiratory fibres in the second
cervical spinal cord segment: respiratory dysfunction syn-
drome after cervical cordotomy. J. Neurol. Neurosurg.
Psychiatry 55, 1142 – 1145.
Maskill, D., Murphy, K., Mier, A., Owen, M., Guz, A., 1991.
Motor cortical representation of the diaphragm in man. J.
Physiol. 443, 105 – 121.
Munschauer, F.E., Mador, M.J., Ahuja, A., Jacobs, L., 1990.
Selective paralysis of voluntary but not limbically influ-
enced automatic respiration. Arch. Neurol. 48, 1190 – 1192.
Murphy, K., Corfield, D.R., Guz, A., Fink, G., Harrison, J.,
Wise, R.S.J., Adams, L., 1996. Cerebral areas associated
with the motor control of speech. J. Appl Physiol. (in
press).
Orem, J., Nettick, A., 1986. Behavioural control of breathing
in the cat. Brain Res. 366, 238 – 253.
Orem, J., 1988. Neural basis of behavioural and state-depen-
dent control of breathing. In: Lydic, R., Biebuy, K.J.F.
(Eds.), Clinical Physiology of Sleep, American Physiologi-
cal Society, Bethesda, MD, pp. 78 – 96.
Penfield, W., Rasmussen, T., 1950. Vocalization and speech
arrest. In: The Cerebral Cortex of Man. A Clinical Study
of Localization of Function, Chapter V. Macmillan, New
York, pp. 88 – 108.
Planche, D., 1972. Effets de la stimulation du cortex cerebral
sur l’activite du nerf phrenique. J. Physiol. 64, 51 – 56.
Plum, F., Leigh, R.J., 1981. Abnormalities of central mecha-
nisms. In: Hornbeim, T.F. (Ed.), Regulation of Breathing,
Part II, vol. 17. Lung Biology in Health and Disease.
Marcel Dekker, New York, pp. 989 – 1067.
Ramsay, S.C., Adams, L., Murphy, K., Corfield, D.R.,
Grootoonk, S., Bailey, D.L., Frackowiak, R.S.J., Guz, A.,
1993. Regional cerebral blood flow during volitional expi-
ration in man. J. Physiol. 461, 85 – 101.
Severinghaus, J.W., Mitchell, R.A., 1962. Ondine’s Curse —
failure of respiratory centre automaticity while awake.
Clin. Res. 10, 122.
Shea, S.A., Andres, L.P., Shannon, D.C., Guz, A., Banzett,
R.B., 1993a. Respiratory sensations in subjects who lack a
ventilatory response to CO2. Resp. Physiol. 93, 203 – 219.
Shea, S.A., Andres, L.P., Shannon, D.C., Banzett, R.B.,
1993b. Ventilatory responses to exercise in humans lacking
ventilatory chemosensitivity. J. Physiol. 468, 623 – 640.
Simon, P.M., Schwartzstein, R.M., Weiss, J.W., Fencl, V.,
Teghtsoonian, M., Weinberger, S.E., 1990. Distinguishable
types of dyspnoea in patients with shortness of breath.
Am. Rev. Resp. Dis. 142, 1009 – 1014.
204 A. Guz / Respiration Physiology 109 (1997) 197–204

Somjen, G.G., 1992. The missing error signal-regulation be-
yond negative feedback. News Physiol. Sci. 7, 184–185.
Talbot, J.D., Marrett, S., Evans, A.C., Meyer, E., Bushnell,
M.C., Duncan, G.H., 1991. Multiple representations of
pain in human cerebral cortex. Science 251, 1355–
1358.
Ullmann, E., 1970. The two original papers by Hering and
Breuer submitted by Hering to the K.K. Akademie der
Wissenschaften ZV Wien in 1868. In: Porter, R. (Ed.),
Breathing: Hering-Breuer Centenary Symposium CIBA
Foundation; Churchill Livingstone, Edinburgh, pp. 357 –
394.