Você está na página 1de 23

Am J Physiol Regul Integr Comp Physiol 296: R1473R1495, 2009.

First published February 11, 2009; doi:10.1152/ajpregu.91008.2008.

Invited Review

Integration of cerebrovascular CO2 reactivity and chemoreflex control


of breathing: mechanisms of regulation, measurement, and interpretation
Philip N. Ainslie1,2 and James Duffin3
1

Department of Physiology, University of Otago, Dunedin, New Zealand; 2Department of Human Kinetics, Faculty of Health
and Social Development, University of British Columbia, Okanagan, Kelowna, Canada; and 3Departments of Anaesthesia and
Physiology, University of Toronto, Toronto, Ontario, Canada
Ainslie PN, Duffin J. Integration of cerebrovascular CO2 reactivity and chemoreflex control of breathing: mechanisms of regulation, measurement, and interpretation. Am J Physiol Regul Integr Comp Physiol 296: R1473R1495, 2009. First
published February 11, 2009; doi:10.1152/ajpregu.91008.2008.Cerebral blood
flow (CBF) and its distribution are highly sensitive to changes in the partial
pressure of arterial CO2 (PaCO2). This physiological response, termed cerebrovascular CO2 reactivity, is a vital homeostatic function that helps regulate and maintain
central pH and, therefore, affects the respiratory central chemoreceptor stimulus.
CBF increases with hypercapnia to wash out CO2 from brain tissue, thereby
attenuating the rise in central PCO2, whereas hypocapnia causes cerebral
vasoconstriction, which reduces CBF and attenuates the fall of brain tissue
PCO2. Cerebrovascular reactivity and ventilatory response to PaCO2 are therefore
tightly linked, so that the regulation of CBF has an important role in stabilizing
breathing during fluctuating levels of chemical stimuli. Indeed, recent reports
indicate that cerebrovascular responsiveness to CO2, primarily via its effects at
the level of the central chemoreceptors, is an important determinant of eupneic
and hypercapnic ventilatory responsiveness in otherwise healthy humans during
wakefulness, sleep, and exercise and at high altitude. In particular, reductions
in cerebrovascular responsiveness to CO2 that provoke an increase in the gain
of the chemoreflex control of breathing may underpin breathing instability
during central sleep apnea in patients with congestive heart failure and on
ascent to high altitude. In this review, we summarize the major factors that
regulate CBF to emphasize the integrated mechanisms, in addition to PaCO2, that
control CBF. We discuss in detail the assessment and interpretation of cerebrovascular reactivity to CO2. Next, we provide a detailed update on the
integration of the role of cerebrovascular CO2 reactivity and CBF in regulation
of chemoreflex control of breathing in health and disease. Finally, we describe
the use of a newly developed steady-state modeling approach to examine the
effects of changes in CBF on the chemoreflex control of breathing and suggest avenues
for future research.
cerebral blood flow; ventilation; reactivity

RESPIRATORY-INDUCED CHANGES

in the partial pressure of arterial


CO2 (PaCO2) play a major role in cerebral blood flow (CBF)
regulation (30). Elevations in PaCO2 (hypercapnia) lead to
vasodilation of cerebral arterioles in the downstream bed and a
subsequent increase in CBF, whereas a reduction in PaCO2
(hypocapnia) leads to vasoconstriction and a subsequent decrease in CBF (97, 212). The highly sensitive distribution of
CBF in response to changes in PaCO2, termed cerebrovascular
CO2 reactivity, is a vital homeostatic function that helps
regulate and maintain central pH and, therefore, affects the
respiratory central chemoreceptor stimulus. In this sense, elevations in CBF with hypercapnia wash out CO2 from brain
tissue, thereby attenuating the rise in central PCO2, whereas
hypocapnia causes cerebral vasoconstriction, which reduces
CBF and attenuates the fall of brain tissue PCO2. An emerging
Address for reprint requests and other correspondence: P. N. Ainslie, Dept.
of Physiology, Univ. of Otago, Dunedin, New Zealand (e-mail: philip.ainslie
@otago.ac.nz).
http://www.ajpregu.org

concept, therefore, is that cerebrovascular reactivity and ventilatory response to PaCO2 are tightly linked, so that regulation
of CBF has an important role in stabilizing breathing during
fluctuating levels of chemical stimuli.
This review comprises four main parts. 1) We present the
major factors that regulate CBF to emphasize the integrated
mechanisms, in addition PaCO2, that control CBF. 2) We assess
and interpret cerebrovascular reactivity to CO2, with particular
focus on the integrative mechanisms linking cerebrovascular
function with ventilatory control in health and disease. The
related literature is reviewed and discussed, with particular
focus on the advantages and disadvantages of the steady-state
and rebreathing techniques for the assessment of cerebrovascular and ventilatory responsiveness to CO2; the physiological
interpretation of these different tests are critically examined,
and practical guidelines are provided, so that test results may
be collected to enhance comparisons, reproducibility, and accuracy. 3) We provide a detailed update on the integration of
the role of cerebrovascular CO2 reactivity and CBF in regula-

0363-6119/09 $8.00 Copyright 2009 the American Physiological Society

R1473

Invited Review
R1474

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

tion of the chemoreflex control of breathing in health and


disease. 4) We describe the use of a newly developed steadystate modeling approach to describe the effects of changes in
CBF on the chemoreflex control of breathing and suggest
avenues for future research.
REGULATION OF CBF WITH PARTICULAR FOCUS
ON THE INFLUENCE OF PaCO2

The relative constancy of the overall craniospinal volume


(203), the cerebrovascular anatomy, and the positional and
delay differences of the cerebral circulation relative to the heart
contribute to the complexity of intracranial hemodynamics (58,
119, 150, 151, 203) (Fig. 1). Control of cerebral perfusion is
linked closely to regulation of intracranial volume; it comprises
the arterial cerebrovascular bed, the large cerebral veins, and
the processes associated with production and reabsorption of
cerebrospinal fluid (70). According to Poiseuilles law (122)
(Eq. 1), CBF is determined by the cerebral perfusion pressure
and the cerebrovascular conductance or its reciprocal, cerebrovascular resistance (2). Cerebral perfusion pressure is the
difference between blood pressure at the level of the circle of
Willis and intracranial pressure, and intracranial pressure, in
turn, encompasses central venous pressure and the pressures
within the cerebrospinal fluid (Fig. 1). The variable resistance
to flow is encountered mostly in the cerebral arteriolar and
capillary bed (58). In contrast, the large cerebral arteries and
veins are dedicated to blood distribution and storage of blood
volume and are, in principle, noncompliant and act merely as
a conduit for the pulsatile arterial flow from the aorta to the
brain (145). CBF is dynamically adjusted to changes in perfusion pressure, metabolic activity of the brain, humoral factors,
and autonomic nerve activity (33). Figure 1 is a simplified
diagram of these main regulatory or active mechanisms that
interact to maintain O2 supply matched to cerebral metabolic
demand and protect the brain from changes in cerebral perfusion. This integrative model highlights the interdependence of
CBF and many other variables through complex and likely
nonlinear relationships. The highlighted areas represent the
potential areas, direct and indirect, in which PaCO2 may influence CBF. The major factors that affect CBF, with particular
focus on the modulatory effects of CO2, are described below

and illustrated in Fig. 2 to highlight the directional influence of


these factors on CBF.
Influence of PaCO2 on CBF: mechanisms of action. It is well
established that the cerebral vasculature is profoundly affected
by PaCO2 (30). The teleological relevance of the exquisite
sensitivity of CBF to changes in PaCO2 seems to be a vital
homeostatic function that helps regulate and maintain central
pH (38) and, therefore, affects the respiratory central chemoreceptor stimulus. The interaction between PaCO2 and vasodilation/vasoconstriction is normally ascribed to occur at the
level of the arterioles and the precapillary sphincters (18, 33,
58). Increased CO2 results in a relaxation of the vascular
smooth muscle of all cerebral vessels, although the small
vessels are the most responsive (213). By contrast, the vasoconstrictor effect of hypocapnia is unaffected by vessel size
(213). The mechanism(s) behind these actions of CO2 has not
been entirely elucidated. Some evidence indicates that elevations in CO2 and the concomitant change in pH activate K
channels in the vascular smooth muscle. In support of this, the
cerebral endothelial cells express four classes of K channels:
inward rectifying K channels, Ca2-activated K channels,
ATP-sensitive K channels, and voltage-gated K channels (99, 134). Of these, ATP-sensitive K and voltage-gated
K channels are activated by a reduction in pH (21, 224),
suggesting that their contribution to acidosis induced cerebral
vasodilation (29, 99). Increases in the open probability of K
channels result in K efflux and endothelial cell hyperpolarization (64, 89). Hyperpolarizing currents may then be transmitted to the underlying smooth muscle, via myoendothelial
junctions (69), where they may elicit vasodilation secondary to
closure of voltage-gated Ca2 channels, a reduction in intracellular Ca2, and vascular relaxation (89, 101, 134). Hyperpolarization can also be transmitted to neighboring endothelial
cells by way of gap junctions, which allow the conduction of a
hyperpolarizing current for long distances along the endothelium (89). Thus, K channels, through their impact on the
endothelium and vascular smooth muscle, play a role in coordinating vascular tone in upstream and downstream vessels.
An alternative, or possible complementary, means by which
cerebral vasodilation may be mediated is via CO2/pH-induced
alterations in vasoactive factors. For example, CO2-mediated

Fig. 1. Simplified block diagram of the main mechanisms


responsible for control of cerebral blood flow (CBF). Vascular bed corresponds to passive properties of blood vessels, usually modeled as a combination of resistances and
compliances. Red highlighted areas represent potential areas, direct and indirect, in which arterial PCO2 (PaCO2) may
influence CBF. ABP, arterial blood pressure; ICP, intracranial pressure; CSF, cerebrospinal fluid; CPP, cerebral perfusion pressure, CBV, cerebral blood volume. [Modified
from Panerai (150).]

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

R1475

Fig. 2. Directional influence of major factors


involved in regulation of CBF. BP, blood pressure; SNA, sympathetic nerve activity.

cerebral vasodilation and a concomitant increase in CBF may


be evoked by the shear stress-mediated release of vasodilatory
agents such as nitric oxide (NO) and prostaglandins (107, 153).
Evidence for a potential involvement of several other vasodilators, including adrenomedullin and C-natriuretic peptide, a
potential candidate for the unknown endothelium-derived hyperpolarizing factor (37, 39), is indirect (157). A recent study
in our laboratory examined the cerebral release of a number of
these potential candidates in otherwise healthy humans. The
findings supported a role for NO and a lesser role for Cnatriuretic peptide. Interestingly, the role of NO-mediated
cerebral vasodilation in response to hypercapnia, but not hypocapnia, is in agreement with findings from studies of other
animals and humans, although this was the first study to
directly quantify NO release (as indexed by plasma nitrate) in
the human brain. The extent to which NO acts in an obligatory
or permissive manner is not known, but it is clear that the role
of vasoactive factors is highly complex (82, 189).
Regardless of the underlying mechanism, the time course of
the response of the cerebral pial (resistance) vessels to alterations in pH is relatively rapid, with changes in diameter
occurring within 10 s, independent of the resting vessel diameter (58). Early reports in humans (181) indicated that the CBF
response started within 30 s of the beginning of CO2 inhalation
and that 2 min were required to reach peak values (60). More
recent studies (165, 166), consistent with direct microscopic
observation of vessels by a cranial window technique, incorporating transcranial Doppler (TCD) and more sophisticated
methods to manipulate end-tidal gases, demonstrated that the
CBF response to step changes in CO2 in humans was much
faster (6-s delay) than that documented in the early reports.
Arterial PO2. The role of arterial PO2 (PaO2) in the day-to-day
regulation of CBF seems to be minor, reflected in the findings
that, depending on the prevailing PaCO2 (15), a drop in PaO2
below a certain threshold (40 mmHg) is required for cerebral
vasodilation (73). However, although hypoxia per se is a
cerebral vasodilator, reflected in a rise in CBF in proportion to
the severity of isocapnic hypoxia (41, 225), under normal
conditions the hypoxia-induced activation of peripheral chemoreceptor activity leads to hyperventilation-induced lowering
of PaCO2 and subsequent cerebral vasoconstriction. Therefore,
the cerebrovascular bed receives conflicting signals during
exposure to acute hypoxia. The role of PaO2 becomes important
on ascent to high altitude (8, 9) and possibly during hypoxemia
associated with chronic lung disease (22).
Cerebral autoregulation. Cerebral autoregulation (CA) adjusts cerebral arteriolar caliber, or cerebrovascular resistance,
to ensure that CBF levels are matched to metabolic needs, and
it comprises two main components: static and dynamic. Static
CA keeps CBF constant over gradual and progressive changes
in cerebral perfusion (155). Dynamic CA refers to the rapid
regulation of CBF in response to changes in arterial blood

pressure (BP) that occur in a few seconds (229). Control


mechanisms for dynamic CA may differ from those for static
CA (45), and data indicate that neural control of cerebral
circulation may be more effective under dynamic than steadystate conditions (229). Although the actual range of BP in
which CBF is maintained is variable between subjects (104)
and was originally based on steady-state CBF data points under
several different conditions presented in previous publications
(104), there have been a number of reports of the influence of
CO2 on CA. For example, whereas mild hypercapnia has been
shown to impair dynamic CA as estimated by the thigh cuff
deflation method (3) and transfer function analysis (10, 24,
150, 228), relative hypocapnia has variable effects on dynamic
CA (10, 57, 130). The important point is that if hypercapnia
does impair CA, hypercapnia-induced elevations in BP may
influence CBF and thereby mask the true interpretation of
cerebrovascular CO2 reactivity (see Analysis and curve-fitting
considerations).
Sympathetic nerve activity. Although the cerebral circulation
is richly innervated with sympathetic nerve fibers, the effect of
sympathetic nerve activity (SNA) on the regulation of CBF
remains controversial (206). The traditional view is that increases in sympathetic activity appear to have a limited effect
on the cerebral vasculature of humans, particularly at rest (16,
76). It seems likely that any potential influence of SNA on CBF
regulation is masked by the other more powerful regulatory
influences on CBF: autoregulation, cerebrovascular CO2 reac ) (110, 195, 206).
tivity, and, potentially, cardiac output (Q
Recently, however, elegant animal studies incorporating continuous recording of SNA in the superior cervical ganglion (35)
concluded that SNA directed to cerebral vessels increases with
acute hypertension, but not hypotension, suggesting that it
serves a protective function for the cerebral microcirculation,
and not a regulatory role for maintenance of systemic arterial
pressure. In addition to the influence of SNA on CA (143, 229),
there are reports that sympathetic activity influences the reactivity of CBF to PaCO2 (44, 94). Thus, although the influence
might be regarded as minor, CO2-induced elevations in SNA
have the potential to affect CBF by direct and indirect mechanisms (Figs. 1 and 2).
Cerebral metabolism. The O2 supply to the brain depends on
the arterial O2 content and the CBF. Cerebral O2 consumption in normal, conscious, young humans is 3.5 ml100 g brain1 min1
(172). The rate of O2 consumption of the entire brain of
average weight (1,400 g) is therefore 49 ml O2/min. The
magnitude of this rate can be appreciated more fully when it is
compared with the average metabolic rate of the whole body,
250 ml O2/min in the basal state for a 70-kg man. Therefore,
the brain, which represents only 2% of total body weight,
accounts for 20% of the resting total body O2 consumption.
Previous studies in humans have shown that cerebral O2
consumption was unchanged with alterations in PaCO2 in the

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
R1476

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

range comparable with those measured during cerebrovascular


reactivity testing (97, 114, 212). The unchanged cerebral metabolic O2 consumption can be maintained via the increases in
O2 extraction.
Systemic factors. Although the classical notion is that CBF is
maintained over a range of BP, it has more recently been
(85, 142, 205; for
established that CBF is also dependent on Q
review see Ref. 177). Thus, because SNA and heart rate (HR)
are altered during CO2 reactivity testing, subsequent changes
may theoretically influence CBF. It should be acknowlin Q

edged, however, that it is not clear whether these changes in Q


would have direct effects on CBF independent of changes in
cerebral perfusion pressure. Although this idea is supported
indirectly, it remains counterintuitive, because, according to
Poiseuilles law (Eq. 1) (122), blood flow is the quotient of

perfusion pressure and vascular resistance, and changes in Q


are not variables of the equation (202)
F P1 P2r4/8L

(1)

where F is flow; P1 is inflow pressure, P2 is outflow pressure,


r is radius, is viscosity of the fluid, and L is length (122).
Previous studies, however, have reported a positive relation and CBF at normothermia during rest and
ship between Q

exercise (83, 142). For example, increases or decreases in Q


with volume expansion or application of lower body negative
pressure, respectively, altered CBF velocity in a linear fashion
when mean BP and PaCO2 were kept constant. Moreover, in
patients with acute ischemic stroke and severe head injury,
close correlation was shown between CBF in the affected brain
during intravascular volume expanregions and changes in Q
sion, alone or in combination with induced hypertension (202).
affects CBF
Although the proposed mechanism(s) by which Q
are unknown, nonneural, flow-mediated mechanisms have
been suggested (227). For instance, sheer stress related to
increasing pulsatile pressure and/or blood flow may modulate
the steady-state pressure-flow relationship of the cerebral circulation (173). This may occur via the release of sheer stress
response cerebrovascular endothelial cells, which in turn results in a decrease in cerebrovascular resistance in the arterioles and elevates CBF (173, 202). Together, although acknowledging that the mechanisms have not been clearly delin
eated, these findings suggest a direct effect of changes in Q
on CBF.
CEREBROVASCULAR REACTIVITY TO CO2

The term cerebrovascular CO2 reactivity reflects an index


of the ability of the cerebrovascular bed to dilate or constrict in
response to changes in PaCO2. For more than 20 years, TCD has
been used extensively to study CBF regulation and cerebrovascular CO2 reactivity in otherwise healthy subjects as well as
in patients with various forms of cerebrovascular disease. For
example, the measurement of cerebrovascular reactivity to
CO2 has been applied in clinical practice to evaluate cerebrovascular function [e.g., in patients with carotid artery stenosis
(216), hypertension (179), stroke (217), and heart failure
(221)], and a related impairment has been linked to cerebral
ischemic events (217). Interestingly, links between systemic
endothelial function and cerebrovascular CO2 reactivity have
been reported (14, 81, 105), indicating a common pathway
between these responses. Since TCD measures flow velocity,

and not CBF per se, only assessment of changes in flow, rather
than absolute values, can be made. Nevertheless, research
indicates that middle cerebral artery (MCA) blood flow velocity (MCAv) is a reliable and valid index of CBF (68, 100, 156,
178, 204). Moreover, since determinations of cerebrovascular
reactivity are based on stimulus-response principles, absolute
CBF values are not as important as reliable and repeatable
recordings with short (beat-to-beat) time resolution. For this
purpose, therefore, TCD is a well-suited technique in the
experimental research and clinical settings. Before we review
the evidence linking cerebrovascular CO2 reactivity with ventilatory control, we first provide a discussion of the relevant
aspects of cerebrovascular CO2 reactivity to clarify its use,
measurement, and related interpretation.
Regional differences in cerebrovascular CO2 reactivity. Although there is marked variability in the literature for normal
cerebrovascular CO2 reactivity (see below), when derived as
the average reactivity across the majority of the selected
experimental studies, the global CBF reactivity to changes in
PaCO2 is 3.8%/mmHg (72, 77, 95, 170) within the PaCO2
range of 3555 mmHg. On closer examination, however, this
reactivity is normally higher in the hypercapnic than hypocapnic range (43, 84, 221, 222). The mechanisms underlying this
greater reactivity to hypercapnia than hypocapnia may be
related to a greater influence of vasodilator than vasoconstrictive mediators on intracranial vascular tone (157, 199). From a
teleological perspective, the normal lower reactivity in the
hypocapnic than hypercapnic range might be a protective
mechanism to prevent cerebral ischemia during transient drops
in PaCO2, which are known to occur in a range of physiological
(postural change and exercise) and pathophysiological (asthma, syncope, sleep apnea, congestive heart failure, and anxiety attacks)
situations. In other words, maintaining oxygenation might be
more important than pH disturbances. In addition, it is important to note that alterations in CBF reactivity to hypocapnia
may be as important as hypercapnic reactivity for breathing
stability, especially during sleep, when the wakefulness dive is
absent and PaCO2 becomes the critical factor in maintaining
rhythmic breathing (219).
It should be noted that, partly dependent on the method used
to estimate CBF, there is marked inhomogeneity of cerebrovascular CO2 reactivity (137, 160). For example, CBF reactivity is much higher in the gray matter than in the white matter.
Thus, although TCD has many advantages because of the
noninvasive nature of measurement and continuous recordings
with high temporal resolution, TCD normally measures blood
velocity in the MCA, an area that transports blood to large
brain volumes, including gray and white matter, whereas sophisticated approaches, such as pulsed arterial spin labeling
MRI, measure the reactivity of small vessels and capillaries
within a purely cortical gray matter area, where hypercapnic
CO2 reactivity is much higher (114); potential regional differences in hypocapnic CO2 reactivity have not been reported.
Thus, because it is believed that changes in CBF are transmitted accurately to the MCA (1), it seems that the use of TCD
may be described as more of an average indicator of CBF,
whereas greater local resolution can be obtained using MRI.
In addition to the noted differences in cerebrovascular CO2
reactivity between white and gray matter, there are marked
regional differences of gray matter reactivity. A typical composite blood oxygenation level-dependent (BOLD) MR cere-

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

brovascular reactivity map in Fig. 3A shows several bilaterally


symmetrical regions of negative reactivity (depicted in blue).
Moreover, negative reactivity in cigar-shaped bands is apparent in the deep white matter of the centrum semiovale, in
the white matter about the occipital horns of the lateral ventricles and, to a lesser extent, frontal horns in the genu and
splenium of the corpus callosum, and in the posterior limb of
the internal capsule. Comparable regional distribution during
reactivity was confirmed with the use of composite arterial spin
labeling MR (114) (Fig. 3B), indicating that the related changes
are not confounded by CO2-induced changes in cerebral blood
volume and cerebral metabolic rate of O2 consumption. Thus,
an emerging concept is that, even in healthy young subjects,
the cerebrovascular CO2 reactivity may reduce blood flow to
these regions in its attempt to maintain flow elsewhere in the
brain (114). This differential response has aptly been called a
steal phenomenon. Although this steal phenomenon occurs
in those regions where elderly patients most frequently develop
white matter disease, or leukoaraiosis (74, 114), and is a strong
and independent risk factor for ischemic stroke, it is not clear
whether this response might affect ventilatory control.
Regardless of the aforementioned regional difference, it is
clear that cerebrovascular CO2 reactivity is influenced by the
degree of wakefulness and subsequent neuronal activity. For
example, compared with wakeful rest, cerebrovascular CO2
reactivity is reduced during sleep (124) and elevated during
exercise (144, 168). These changes are illustrated in Fig. 4.
Moreover, recent reports indicate that CBF (14, 42, 43) and
cerebrovascular CO2 reactivity show a pronounced circadian
rhythm, with low CBF and cerebrovascular CO2 reactivity
evident in the early morning compared with late afternoon. The
potential implications of these sleep- and circadian-related
alterations in CBF reactivity may be underscored by the common (10 40%) nighttime occurrence of strokes (20, 191)
and the marked peak of these events in the early morning
(115, 118).

R1477

Local regulation of cerebrovascular CO2 reactivity. The


bulk of cerebrovascular CO2 reactivity occurs within the small
arterioles and capillaries, whereas large arteries and veins are
dedicated to blood distribution and storage of blood volume.
This concept reflects the general notion that arterioles control
the majority of cerebrovascular resistance but hold only a
minority of total cerebral blood volume. In reality, however,
there are no strict borderlines between vessel types based on
size, structure, or function, which change smoothly between
arterial and venous macro- and microcirculation. It should also
be noted that reactivity is not necessarily an active phenomenon (18, 158), either in reactive or counterreactive fashion, but
may be combined with possible passive vasodilation due to
changes in downstream transmission of intravascular pressure.
Also, changes in the number of perfused and nonperfused
vessels within the capillary compartment, known as capillary
recruitment, have been hypothesized.
METHODOLOGICAL CONSIDERATIONS IN THE ASSESSMENT
OF CEREBROVASCULAR REACTIVITY TO CO2

For meaningful physiological interpretation, there are a


number of critical considerations about the different means by
which cerebrovascular (and ventilatory) responsiveness to CO2
can be assessed. The following is not intended to mandate the
details of a particular experimental setup and the testing conditions; rather, we emphasize selected principles of testing that
should be followed to enhance comparisons, reproducibility,
and accuracy of data collection.
Steady-state vs. rebreathing techniques. A critical consideration in the assessment of cerebrovascular CO2 reactivity is
whether to use steady-state or rebreathing techniques (Figs. 57). We
believe that the rationale for using a steady-state or rebreathing
method (or both) should firmly reflect the experimental question that
is to be addressed. For example, the incremental change in brain
tissue PCO2 is less during a steady-state change in PaCO2 than during
rebreathing (65, 157); conversely, during rebreathing, any PCO2 gra-

Fig. 3. A: composite blood oxygenation leveldependent (BOLD) MR map of cerebrovascular


reactivity in young, healthy subjects (percent
change in MR signal intensity per mmHg change
in end-tidal PCO2) overlaid on anatomic images.
B: composite arterial spin labeling MR map of
cerebrovascular reactivity in young, healthy subjects [percent flow-sensitive alternating-inversion
recovery (FAIR) MR signal change per mmHg
change in end-tidal PCO2] overlaid on anatomic
images. [Modified from Mandell et al. (114).]

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
R1478

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

Fig. 4. A: characteristics of cerebrovascular CO2 reactivity at rest and during


exercise (EX), with cerebrovascular CO2 reactivity characterized by an exponential function. Arrows show leftward shift of the operating point with
exercise. MCAv, middle cerebral artery blood flow velocity. B: changes in
cerebrovascular reactivity, assessed using linear regression, in each individual
from wakefulness to sleep. Cerebrovascular reactivity to CO2 from wakefulness to stage 2 sleep is reduced. [Modified from Meadows et al. (124) and
Ogoh et al. (144).]

dients between arterial, venous, and brain tissue compartments are


abolished. As a result, since the respiratory central chemoreceptors
respond to medullary H concentration ([H]) set by medullary
PCO2, rebreathing excludes the effects of cerebrovascular reactivity
from the ventilatory response to PCO2, whereas steady-state techniques include it. Similar considerations apply to the measurement of
cerebrovascular reactivity; in this case, however, the stimulus and its
location are uncertain. For example, if the stimulus is arterial [H] as
well as capillary [H], then the rebreathing technique applies the
same stimulus to each location, whereas the steady-state technique
does not. If part of the cerebrovascular stimulus-response dynamics is
slowed (independent of any dynamics associated with the stimulus),
then the full response may not be achieved during rebreathing. This
latter consideration may account for the observation that cerebrovascular CO2 reactivity is greater when assessed with steady-state than
with rebreathing procedures (cf. Fig. 5 with Fig. 7). Until the physiology of cerebrovascular control is better understood, a clear choice
of method is problematic. For example, when the goal is to assess

ventilatory sensitivity to CO2, the change in medullary [H] will


depend on the extent of cerebrovascular reactivity, an unpredictable
response. Unfortunately, at least in humans, it is not possible to make
continuous measurements of the medullary (or venous) [H]; therefore, marked variability will also exist. There are a number of
different ways in which steady-state and rebreathing techniques are
assessed (see above).
Practical considerations. In addition to consideration of the
most appropriate change in PaCO2 and related means to best
fit the relationship with CBF (see Analysis and curve-fitting
considerations), various factors are known to affect CBF and
CBF-CO2 reactivity; therefore, consideration of such influencing factors is critical to improve reproducibility and accuracy
in experimental design. Established influencing factors include
hydration (63), temperature, diet, caffeine (159), time of day
(14, 42, 43), alcohol (25), prior exercise (113, 131), posture
(121), age and sex (12), and extent of collaterals with completeness of the circle of Willis (208). Although it is established that changes in PaCO2 per se will not alter cerebral O2
consumption (97, 114, 212), reports indicate that visual stimulation may profoundly influence CBF and cerebral O2 consumption (51, 95); therefore, care should be taken to ensure
that subjects receive no visual stimulation during cerebrovascular CO2 reactivity tests. Various medications such as angiotensin-converting enzyme inhibitors and statin may influence
normal cerebrovascular responsiveness to changes in PaCO2
(175, 209).
End-tidal, arterial, or brain PCO2? In most instances, CBF
reactivity is expressed as the percent change in CBF per mmHg
change in PaCO2, or end-tidal PCO2, obviating the more invasive
PaCO2 measurement. The tight correlation between the percent
change in MCAv (54, 84, 117) measured by TCD ultrasonography during end-tidal PCO2 variations has encouraged the use
of TCD ultrasonography to measure CO2 cerebrovascular reactivity. However, several considerations are important when
PaCO2 (or end-tidal PCO2) is used to investigate CBF reactivity.
First, end-tidal PCO2 has been shown to underestimate PaCO2 at
rest and to overestimate PaCO2 during hypercapnia (156) and
exercise (93). Such alterations in the PaCO2-end-tidal PCO2
relationship may have implications for the true representation
and physiological interpretation of cerebrovascular reactivity
to CO2. To address these inaccuracies, regression equations
using end-tidal PCO2 and tidal volume have been formulated to
provide an estimate of PaCO2 during exercise (93) and during
hypercapnia (156) to compensate for the overestimation of
PaCO2 by end-tidal PCO2. Results from studies using end-tidal
forcing or fixed fraction of inspired CO2 (FICO2) change without correction of end-tidal PCO2 to PaCO2 should therefore be
treated with caution. Recently, however, prospective targeting
of end-tidal PCO2 with low gas flows, via a sequential gas
delivery circuit, has been shown to adequately reflect PaCO2 at
rest and during various changes in end-tidal PCO2 (86).
Another consideration is that, at least during steady-state
testing, cerebrovascular reactivity to CO2 might be even better
expressed as a percentage of brain tissue PCO2. For example,
early work by Shapiro and colleagues (181) showed a higher
correlation of jugular venous PCO2 than PaCO2 with the associated changes in CBF during elevations in CO2, indicating that
brain tissue PCO2, rather than PaCO2, provides a better correlate
of the physiological stimulus. Conversely, during hypocapnia,
experimental data indicate that PaCO2 (or arterial wall PCO2)

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

R1479

Fig. 5. Polygraph record of a single trial during 5% CO2 exposure. CBF velocity (CBFv) progressively increased to a peak with the increase in end-tidal PCO2
(PETCO2) and then was maintained at a high level throughout hypercapnia. At termination of CO2 exposure, CBFv quickly decreased with reduction of end-tidal
PCO2. Time delay was measured from beginning of CO2 inhalation to onset of CBFv increase (T1), from establishment of the new level of end-tidal PCO2 to the
peak of CBFv (T2), and from off-stimulation to the start of the decline of CBFv (T3) as well as to the nadir of CBFv (T4). SaO2, O2 saturation of arterial blood.
[Modified from Xie et al. (221)]. Note relative drop in end-tidal PCO2 from T2 to T3 with hypercapnia-induced elevations in airflow; this observation indicates
that CBF reactivity during such steady-state testing is somewhat ventilatory dependent.

may have a more important role than jugular venous PCO2


(180). Recently, we reported that cerebrovascular CO2 reactivity in the hypercapnic and hypocapnic ranges was higher when
presented against jugular venous PCO2 than PaCO2. This increase in jugular venous PCO2-CBF reactivity was higher in the
hypercapnic than hypocapnic range (97% vs. 24%). It is

Fig. 6. Representative trace highlights complexity of MCAv response to a step


changes in end-tidal PCO2. Note step change in end-tidal PCO2 within 35 s and
fast and slow components of the elevation in MCAv. Such changes occurred
independent of any obvious changes in BP. Changes in PCO2 were measured
using a RespirAct (184), and MCAv was measured with an instrument from
Spence Technologies (ST3). TCD, transcranial Doppler. (Unpublished data
courtesy of Rosemary Regan and Dr. Joseph Fisher.)

tempting to speculate that these hypercapnic-hypocapnic reactivity differences may reflect a differential control of CBF via
brain tissue PCO2 or PaCO2 during hypercapnia (65, 156, 222)
and hypocapnia (135); however, although jugular venous PCO2
is likely to be a closer index of brain tissue PCO2 than PaCO2
(10, 65, 156, 182, 222), it is difficult to clarify the mechanisms
involved with knowledge of only arterial-venous PCO2 gradients and no information related to the CO2 flux across the
brain. Thus, considerations regarding the differences in endtidal, arterial, and brain tissue PCO2 have important implications for the true representation and physiological interpretation of cerebrovascular CO2 reactivity.
Appropriate experimental change in PaCO2. In healthy subjects, an acute lowering of CBF is associated with mild symptoms of cerebral hypoperfusion. Mental confusion becomes
prominent with 50 60% reduction, and cerebral oxygenation
becomes affected (135); thus, for a safe range for human
experimentation, hypocapnia should be 20 mmHg. Similar,
marked subject discomfort is apparent with administration of
8% CO2 (15 mmHg PaCO2 above rest); thus, changes in
PaCO2 under this range are recommended for safety and related
subject comfort.
Analysis and curve-fitting considerations. Steady-state cerebrovascular CO2 reactivity testing (Figs. 5, 6, and 8) has been
assessed by least squares linear regression (165, 166, 207),
biasymptotic curve analysis (i.e., a tangent-hyperbolic function) (117), a dynamic single-compartment model (165, 166,
207), and an exponential function (56, 59, 60, 77, 125, 152). In
addition, with a steady-state change in end-tidal PCO2, the
speed (or onset response) of CBF velocity at the transition
period of on- and off-CO2 administration can be assessed (165,

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
R1480

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

arterial BP on CBF, leading to a more precise estimation of


cerebrovascular CO2 reactivity based on the cerebrovascular
conductance-end-tidal PCO2 relationship (40). Despite the
mixed reports of the influence of BP, the direct influence of

SNA or indirect influence of increases in HR on elevating Q


may also influence cerebrovascular CO2 reactivity (see above).
For example, even when cerebrovascular CO2 reactivity is
expressed using cerebrovascular conductance, the potential
influence of SNA is important. Some studies have suggested
that increases in SNA attenuate the rise in CBF associated with
hypercapnia (44, 94), whereas others have shown no modulating effect (108, 167). More recently, in the highly controlled
lamb model, it has been reported that cerebral SNA is withdrawn during rapid-eye-movement (REM) sleep, and, therefore, the CBF response to hypercapnia is augmented (34).
Thus, although the influence might be regarded as slight,
CO2-induced elevations in SNA have the potential to affect
CBF by direct and indirect mechanisms. Moreover, although
the modified rebreathing technique may be advantageous in
revealing the effects of BP on cerebrovascular reactivity, the
lack of sustained (15-s) periods of hypocapnia may preclude
useful information gained from assessment of reactivity in the
hypocapnic range.
Fig. 7. Typical recording of baseline, hyperventilation, rebreathing, and recovery in 1 subject. HR, heart rate; MSNA, muscle SNA. Note significant
increases in ABP and MSNA, concurrent with increase in CBFv during
rebreathing. Arrows indicate start and end of hyperventilation and rebreathing
periods. Note 1) equilibrium between fraction of inspired CO2 and end-tidal
PCO2 (PETCO2) during rebreathing and progressive rise in PETCO2 and 2) progressive
increase in MSNA and mean ABP during rebreathing. [Modified from Claassen
et al. (40).]

166, 221); the physiological meaning of this response, however, is unclear. Indeed, recent experimental results from our
laboratory indicate that the initial CBF response to a step
change in PaCO2 is a highly complex response with marked
intrasubject variability (Fig. 6).
Numerous studies have also used linear regression to assess
CO2 reactivity during modified rebreathing tests (5, 6, 8, 15,
19, 43, 63, 96, 105, 121, 124, 157, 171, 181, 221, 222) (Figs.
5 and 8). However, more research exploring the physiological
characteristics of cerebrovascular CO2 reactivity has demonstrated that the relationship between PaCO2 and CBF is nonlinear and might be affected by CO2-induced changes in arterial
BP (Fig. 7) (71, 94). On closer examination, during hypercapnia, increases in SNA and HR are well documented, whereas
variable changes in BP have been reported. Some studies
report an increase (40, 56, 125, 152), while others report no
change (71, 94). Methodological differences in duration and
intensity of hypercapnia and in the measurement of BP may
partly explain the discrepancies between studies.
In addition, the control mechanisms that govern the responses of CBF and arterial BP to CO2 have dynamic properties (40, 125, 126). These studies, among many others, have
quantified cerebrovascular CO2 reactivity using linear regression of steady-state responses of CBF to changes in CO2,
without incorporation of the effects of BP. The likely reason
for this simplicity is that well-controlled experiments and
complex modeling methods may not be practical for clinical
use (40). To account for the nonlinear CBF-end-tidal PCO2
relationship, the use of cerebrovascular conductance index
during this process may reveal direct effects of changes in

CHEMOREFLEX CONTROL OF BREATHING: INFLUENCE


OF CBF

General organization. Breathing is stimulated via a chemoreflex arc, which includes the central and peripheral chemoreceptors, the central nervous system, and the respiratory
muscles. The chemoreflexes form the feedback part of a
negative-feedback loop (Fig. 9). The loop is completed by the
forward part, in which alveolar ventilation controls the uptake
of O2 and elimination of CO2. The forward part of the loop
relates the dependence of PCO2 and PO2 on ventilation, and, as
the alveolar air equation states, the product of ventilation and
alveolar PCO2 is proportional to the amount of CO2 eliminated;
hence, in any steady metabolic state, PCO2 depends on CO2
production divided by alveolar ventilation (52). This rectangular hyperbolic relationship is referred to as the metabolic
hyperbola. PCO2 is the major factor controlling the [H] sensed
by the chemoreceptors, but acid-base status is also involved
(Fig. 8). Stimulation of the chemoreceptors results in an increased alveolar ventilation, which results in an increased
elimination of CO2 and a fall in PCO2 and [H] stimulus at the
chemoreceptors, hence, the negative-feedback designation of
the system.
Central chemoreflex. The central chemoreceptors are stimulated by the [H] of their local environment in the medulla
(scattered within and near the ventrolateral surface of the
medulla) (129, 132). Since [H] is directly dependent on the
PCO2 of chemoreceptor tissue, they are often thought of as CO2
receptors, but this simplification can be misleading. Although
CO2 passes freely across the blood-brain barrier, H ions do
not; therefore, central [H] may differ from arterial [H]. As
a consequence, medullary acid-base conditions determine the
PCO2-[H] relationship at the central chemoreceptors, and the central chemoreceptors are isolated from arterial acid-base disturbances, except as they involve changes in PaCO2. The control of
central acid-base status to keep central [H] at normal values
may result in differences in the PCO2-central [H] relationship

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

R1481

Fig. 8. Representative recordings from 1 subject during


isocapnia (A) and steady-state hypercapnia (B; PETCO2
10 mmHg above resting baseline value). End-tidal forcing was used. Note 1) tight control of end-tidal gases
independent of changes in ventilation [expired and inspired tidal volumes (VTE and VTI)] and 2) progressive
increase in MSNA and mean arterial blood pressure
(MAP) during hypercapnia. PETO2, end-tidal PO2; VP,
index of MCAv. [Modified from Ainslie (5).]

that shift the central chemoreflex ventilatory response to


CO2 (52).
Central PCO2 is determined by three factors, PaCO2, medullary CO2 production, and medullary blood flow, varying directly with PaCO2 and inversely with medullary blood flow. The
time constant for change in central PCO2 is 100 s (medullary
volume/blood flow), so a full response to any alterations in
PaCO2 and medullary blood flow by central PCO2 requires 5
min (3 time constants). Models of the drive to breathe from the
central chemoreceptors (Fig. 10) assume that central neural
drive increases linearly with [H] above a chemoreceptor
threshold [H] with a constant slope (sensitivity); therefore,
sensitivity and threshold determine the response to any particular stimulus.
Peripheral chemoreflex. The peripheral chemoreceptors, located in the carotid bodies at the bifurcation of the carotid
arteries, are stimulated by arterial [H] and hypoxia. They

taste the blood approaching the brain, signaling the respiratory controller in the medulla via the carotid sinus nerve (a
branch of the glossopharyngeal or IXth cranial nerve) (87,
88). In the same way as for the central chemoreceptors,
acid-base conditions, in this case in the arterial blood,
determine the PaCO2-[H] relationship at the peripheral
chemoreceptors.
Models of the drive to breathe from the peripheral chemoreceptors (Fig. 11) assume that the neural drive to breathe from
peripheral chemoreceptors increases linearly with [H] above
a chemoreceptor threshold [H] with a constant slope (sensitivity); therefore, sensitivity and threshold determine the response to any particular stimulus. However, in contrast to the
central chemoreceptors, the sensitivity of the peripheral chemoreflex response to [H] also depends on PaO2. Thus, hypoxia acts by increasing the peripheral chemoreceptor response
to [H]; therefore, the receptors are maximally stimulated by a

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
R1482

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

Fig. 9. Respiratory chemoreflex control of ventilation. Pulmonary ventilation


controls PCO2 and PO2, the forward part of the loop. Modified Stewart model
equations are used to determine arterial and central H concentration ([H])
by their respective PCO2, strong ion difference (SID), albumin concentration

([Alb]), and phosphate concentration ([PO3


4 ]). [H ] and PO2 control ventilation via the respiratory chemoreflexes in the feedback part of the loop.
Ventilation is also affected by drives to breathe independent of chemoreflexes,
here called the wakefulness drive.

simultaneous increase in [H ] and decrease in PO2, in other


words, by asphyxia (201).
Traditionally, the peripheral chemoreceptors have been assumed to be hypoxia sensors, driving ventilation independent
of [H]. This assumption is misleading for acute changes;
hypoxia affects the peripheral chemoreceptor drives to breathe,
but its ultimate influence is strongly dependent on the prevailing [H] (103). This aspect of the peripheral chemoreflex has
two important implications: 1) hyperoxia reduces the peripheral chemoreflex sensitivity to [H] to negligible values in
many individuals, and 2) if [H] is below the ventilatory
recruitment threshold, there is little peripheral chemoreflex
response to hypoxia (192). However, in the presence of pro-

Fig. 10. Drive to breathe from central chemoreceptors assumes that central
drive increases linearly with [H] above a chemoreceptor threshold [H] with
a constant slope (sensitivity); therefore, sensitivity (DcS) and threshold (Tc)
determine the response to any particular stimulus.

Fig. 11. Models of the drive to breathe from peripheral chemoreceptors


assume that the drive to breathe from peripheral chemoreceptors increases
linearly with [H] above a chemoreceptor threshold [H] with a constant slope
[sensitivity (DpS)]; therefore, sensitivity and threshold (Tp) determine the
response to any particular stimulus. Inset: sensitivity depends on PO2 in a
rectangular hyperbolic relation, with asymptotes for PO2 (usually 30 mmHg)
and DpS (usually zero).

longed continuous or intermittent hypoxia, the chemoreflex


may undergo modifications (53).
Chemoreflex control system. The central and peripheral
chemoreceptor neural drives to breathe sum within the medulla
to provide the chemoreflex neural drive to breathe in humans
(66, 127, 146, 183), but perhaps not in rats (46). Pulmonary
ventilation is unaffected until the total neural drive exceeds a
neural drive threshold; therefore, an increase in PCO2 does not
increase ventilation until a threshold PCO2 has been exceeded,
referred to as the ventilatory recruitment threshold (Fig. 12).

Fig. 12. Complete chemoreflex model. Ventilation, dependent on PCO2 and


PO2, is shown as a series of isoxic lines. Central and peripheral ventilatory
neural drives (dashed and dotted lines, respectively) are added to provide total
chemoreflex neural drive to breathe (solid lines) but only increase ventilation
once a neural drive threshold (TD) is exceeded (heavy dashed line), establishing ventilatory recruitment PCO2 thresholds at each isoxic PO2. Central chemoreceptor threshold PCO2, when plotted against PaCO2, is less than that for the
peripheral chemoreceptors because of PaCO2-central chemoreceptor PCO2 difference (a-c diff).

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

Above the ventilatory recruitment threshold, the ventilatory


response to PCO2 is usually linear, with a slope (sensitivity)
varying with PO2 (52, 127).
When PO2 is high, the ventilatory response to CO2 is due
almost entirely to the central chemoreflex in many individuals,
although for some a reduced peripheral response remains.
When PCO2 is below the ventilatory recruitment threshold,
ventilation is maintained by the waking neural drive independent of the chemoreflexes, which reflects the state of the
subject. It is referred to as the waking neural drive, because
it is withdrawn during sleep (66, 146, 183). When this drive is
lost during sleep, PCO2 below the ventilatory recruitment
threshold results in apnea; the apnea threshold and the ventilatory recruitment threshold are equal.
The graphic representation of the chemoreflex control of
breathing is useful for predicting ventilation and PCO2 under a
variety of conditions. Figure 13 shows one example, that of an
individual during sleep at altitude. In this case, the waking
neural drive is absent, the peripheral chemoreflex drive is
enhanced because of low PO2, and the resting PCO2 is decreased
because of the lowering of the chemoreflex threshold with
adaptation to altitude (10, 190). As a result, the individuals
equilibrium point is close to the apnea threshold, and, under
these conditions, a small decrease in PCO2 results in a loss of
ventilation (52).
Stability. The stability of the chemoreflex control system is
governed by the sensitivity or gain of the entire control loop,
both forward and feedback (chemoreflex) portions (Fig. 9).
Increases in loop gain are associated with instability (98, 226).
As illustrated by a sleeping individual at altitude (Fig. 13, see
Fig. 18), sleep also moves the equilibrium point to increasing
PCO2, which increases the ability of ventilation to change PCO2;
i.e., the sensitivity or gain of the forward part of the loop is
increased. The chemoreflex sensitivity (feedback portion of the
loop) is also increased due to the hypoxia, so that overall the
loop gain, which is the combination of the forward and feed-

apnea

Fig. 13. Chemoreflex control of breathing for prediction of resting ventilation


and PCO2 under a variety of conditions. Dashed line, metabolic hyperbola;
dotted line, addition of normal waking (W) and chemoreflex drives to breathe
in an individual at altitude. Note decreased equilibrium PCO2 and increased
chemoreflex sensitivity due to hypoxic environment. Solid line, drive to
breathe during sleep, where disappearance of waking drive produces an apnea
threshold. During sleep, with resting PCO2 and apnea threshold so close
together (difference CO2 reserve) because of high chemoreflex sensitivity,
a slight increase in ventilation will decrease PCO2 below apnea threshold, and
breathing stops.

R1483

back gains, is increased, thereby reducing control system


stability. It should also be noted that peripheral chemosensitivity plays a key role in breathing stability (50, 98, 147, 186,
220). Moreover, recent evidence from animal experiments
emphasizes that brain tissue PCO2 (46) may also affect peripheral chemosensitivity in a hypoadditive manner (47). This
possible link between central sleep apnea and CBF control is
discussed further (see Role of the carotid bodies in breathing
stability).
Influence of CBF. The effect of CBF control by PCO2 should
also be considered in an overview of the chemoreflex control of
breathing; as PaCO2 increases, CBF also increases and washes
out CO2 from brain tissue, reducing the central chemoreceptor
stimulus. Any reduction in the sensitivity of CBF to PCO2,
therefore, increases the overall sensitivity of the central chemoreflex response to changes in PaCO2 (222). On the other
hand, if the sensitivity is high in an individual and CBF
increases markedly with PCO2, then the difference between
central chemoreceptor PCO2 and PaCO2 declines as PaCO2 increases. Measurement of sensitivity using a steady-state technique (128, 149), which measures the central chemoreflex
ventilatory response and the effect of the CBF response, may
provide a lower sensitivity of the ventilatory response of the
central chemoreflex than the rebreathing technique (52), which
measures only the central chemoreflex ventilatory response
unaffected by CBF. Nevertheless, in many individuals, this
effect is negligible (149).
Methodological considerations. Several components of the
chemoreflex control system described above must be considered when tests are designed to assess the ventilatory responses
to CO2 and hypoxia. 1) The location of the central chemoreceptors in medullary tissue means that changes in PaCO2 will
not be fully expressed at the central chemoreceptors for 5
min (3 time constants of 100 s); however, at least on the basis
of two-compartment modeling data, central and peripheral
chemoreceptors have a relatively short delay (1224 and
6 8 s, respectively) (43a). 2) Because hypoxia affects ventilation by changing the sensitivity of the peripheral chemoreflex to PCO2, the ventilatory response to hypoxia depends on the
choice of isocapnia; which is problematic for comparisons
(185). 3) Because CBF changes with PCO2 and hypoxia, the
central chemoreceptor PCO2, as well as its associated time
constant, also changes; as a result, the ventilatory response
to PaCO2 may differ from the ventilatory response to brain tissue PCO2.
These considerations and the fundamental issue of the experimental question dictate the protocol for testing. With consideration 2 in mind, evaluation of the chemoreflexes should
consist of measuring the ventilatory responses to PCO2 at
several isoxic PO2 levels. Two techniques may be used: steady
state and rebreathing.
The steady-state ventilation response relates end-tidal PCO2
and ventilation and, therefore, provides a measure of the
chemoreflex sensitivity, which includes the effect of CBF in
mitigating the central stimulus. Techniques such as end-tidal
forcing with high gas flows (165, 166, 207) and prospective
targeting with low gas flows (185) can be used to change
end-tidal PCO2 while maintaining isoxia. Expression of the full
ventilatory response takes several minutes at each end-tidal
PCO2, and, for the isoxic hypoxic responses, care must be taken
to limit the accumulated hypoxic exposure time to avoid

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
R1484

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

hypoxic ventilatory decline. Obtaining several points on the


response, therefore, takes time, and since only a few points on
the response are measured, care must be taken to ensure that
undetected nonlinearities, such as thresholds, do not affect the
measurement of sensitivity (55, 128). With the range of PCO2
limited to hypercapnia, the ventilatory recruitment threshold is
not measured, unless special techniques, such as assisted ventilation, are employed (220).
With the modified rebreathing method (52, 128), ventilation
response relates mixed venous PCO2 and ventilation and, therefore, provides a measure of the chemoreflex sensitivity and the
ventilatory recruitment threshold without the effect of CBF in
mitigating the central stimulus. This method is based on the
rebreathing method of Read and Leigh (169) with two modifications. 1) A 5-min period of hyperventilation before rebreathing reduces the arterial and central PCO2 at the start, so
that as PCO2 rises during rebreathing, the ventilatory recruitment threshold is discernible. 2) A controlled flow of O2 into
the rebreathing bag maintains isoxia, and since the method
measures the response within 5 min of hypoxia, it avoids
hypoxic ventilatory decline.
The modified rebreathing method rapidly equilibrates arterial, alveolar, and rebreathing bag PCO2 to mixed venous PCO2
at the start of rebreathing and relies on the resting metabolic
production of CO2 to slowly increase PCO2 over time; thus
tissue PCO2, rather than inspired PCO2, drives the increase, and
the slow dynamics of the central response in the latter case are
providing the mixing, end-tidal
absent. With breathing and Q
PCO2, therefore, estimates PaCO2 and central PCO2, and the
arterial-venous PCO2 difference is reduced sufficiently to negate any effects of CBF changes. However, the rebreathing
method may overestimate the CO2 sensitivity if brain CO2
production is greater than whole body CO2 production and
central CO2 increases faster than mixed venous CO2. Although
modeling shows that this factor is negligible (169), direct
experimental evidence for this assumption is lacking. For
testing the sensitivity of the central chemoreceptors (i.e., to
ensure a defined stimulus), the rebreathing technique is likely
to be superior to the steady-state method; however, the latter is
arguably less physiological in nature when blood gas perturbations are considered (i.e., during sleep apnea). Clearly, there
are limitations with rebreathing and steady-state methods that
should be considered in the design of the experimental question.
Methodological considerations for the integrated assessment of cerebrovascular and ventilatory control. The range of
steady-state and rebreathing techniques has been reviewed for
the assessment of cerebrovascular reactivity; many of the same
issues apply for the assessment of ventilatory reactivity. It is
clear that there are many advantages and disadvantages to the
different testing regimens, and attempts at consensus have
produced no uniform agreement. Furthermore, the nature of the
integrative cerebrovascular and ventilatory responses requires
additional comment. For example, as mentioned, with steadystate testing, there is always a PCO2 gradient between the brain
tissue, arterial, and venous compartments; this gradient, in part,
will be determined by the cerebrovascular and ventilatory
reactivity. Therefore, results from steady-state tests of cerebrovascular reactivity that use a fixed change in FICO2 may be
interpreted as ventilatory dependent. Since end-tidal clamping methods using end-tidal forcing (165, 166, 207) or pro-

spective targeting with low gas flows (185) result in a fixed


change in PCO2 independent of ventilation, the cerebrovascular
reactivity is likely not to be influenced by related changes in
ventilation, and the cerebrovascular reactivity results may be
interpreted as ventilatory independent. During rebreathing,
PCO2 gradients between arterial, venous, and brain tissue compartments are abolished; therefore, assessment of cerebrovascular reactivity with this method may be also be interpreted as
ventilatory independent.
INTEGRATION OF CEREBROVASCULAR REACTIVITY TO CO2
AND THE CHEMOREFLEX CONTROL OF BREATHING

A close relationship between changes in CBF and ventilatory output has been shown in animal studies. Results of the
first major study on the interrelationship between CBF and
ventilation in anesthetized dogs reported in 1928 (176) showed
that if CBF was increased from low levels, ventilation increased if the animals were hypoxic but decreased if the
animals were breathing room air. It was suggested that the
former response was indicative of removal of brain hypoxic
depression by the increasing CBF, and the latter response
(ventilatory depression) was a consequence of the removal of
accumulated brain CO2 and acid metabolites. Other studies
investigated the ventilatory responses to acute reductions of
CBF by clamping the vertebral artery and/or the common
carotid artery (78, 161). After total or partial reduction of CBF,
ventilation is usually increased and, depending on the degree of
CBF reduction, apnea may develop. However, although brain
hypoxia and brain acidosis were evident in these studies,
because the animals were anesthetized and the contribution of
the carotid chemoreceptors to the ventilatory response was
unknown, interpretation of these data is difficult.
Evidence for the increased sensitivity of ventilation to a
change in PaCO2 by reduced CBF and cerebrovascular CO2
reactivity was first provided in highly controlled studies in the
unanesthetized goat (36) (Fig. 14). This study demonstrated
that when CBF was reduced by 30%, the cerebrovascular
response was attenuated, whereas the ventilatory responsiveness to CO2 was increased. This observation suggests that CBF
affects ventilation through modification of tissue PCO2 in the
region of the medullary chemoreceptor area. With poor brain
perfusion, the central chemoreceptor will sense a relatively
higher [H] for a given change of PaCO2 and will correspondingly trigger a more vigorous ventilatory response. However,
when CBF was further reduced to 50%, which nearly abolished
CBF-CO2 reactivity, there was a marked blunting of the minute
E)-PaCO2 response, possibly as a consequence of
ventilation (V
hypoxic depression of the respiratory neurons (36). Thus, on
the basis of these animal experiments, there seems to be a
threshold at which changes in CBF and cerebrovascular CO2
reactivity may lead to differential alterations in the ventilatory
responsiveness to CO2.
Experimental evidence for the close relationship between
changes in CBF and ventilatory control in humans is mostly
based on early studies in a range of pathological conditions.
For example, elevations in ventilatory CO2 responsiveness
have been found primarily in clinical studies of patients with
moderate cerebrovascular disease. Heyman et al. (79) used the
steady-state method to determine ventilatory responsiveness to
CO2 in a group of patients with chronic bilateral forebrain

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

E) and CBF responses to CO2 at 100, 70, and 50% of


Fig. 14. Ventilatory (V
control CBF in goats. Ventilatory response to CO2 increased at 70% CBF but
was depressed at 50% CBF, indicating that severe brain ischemia blunts the
sensitivity of the ventilatory chemoreflex. [Modified from Chapman et al. (36).]

dysfunction secondary to occlusive cerebrovascular disease


and found that responsiveness was enhanced. This finding has
been confirmed by other investigators using the same method
to determine the CO2 response (162, 163). These studies
inferred that the normal inhibitory influence of the forebrain
was abolished by ischemic damage; the absence of this inhibition resulted in augmentation of the ventilatory response to
CO2. In contrast, North and Jennett (136), using a rebreathing
method, found that the ventilatory response to CO2 was depressed in patients with acute ischemic injury of the forebrain.
At least part of this discrepancy might be accounted for by
differences in the technique used to assess CO2 responsiveness
(steady state vs. rebreathing), because the steady-state method
is more subject to the potential artifact induced by impairment
of the cerebrovascular response to CO2.
In summary, the finding in goats that, at 50% CBF, the
ventilatory response to CO2 was reduced is broadly analogous
to the finding that ventilatory responsiveness to CO2 is de-

R1485

pressed in patients with severe cerebrovascular disease (136,


163, 215). Such changes likely reflect impairment of metabolism of respiratory neurons, as reflected in a marked decline in
whole brain O2 consumption in patients with severe brain
ischemia (104). Collectively, on the basis of animal and human
pathological studies, there seems to be a threshold at which
changes in CBF and cerebrovascular CO2 reactivity may lead
to differential alterations in the ventilatory responsiveness
to CO2.
A number of related experiments utilizing TCD have speculated that a reduction of CBF-CO2 reactivity may lead to
subsequent alterations in ventilatory reactivity to CO2 (7, 8, 14,
43, 63, 144, 156, 221). Specifically, these studies indicated that
a reduced CBF-CO2 reactivity results in an enhanced ventilatory sensitivity to CO2 (measured using steady-state methods)
via a CBF-related attenuation of brain [H] washout and
greater stimulus at the central chemoreceptors. Three important
considerations arise. 1) Is there direct evidence for a role of
cerebrovascular CO2 reactivity in the ventilatory response to
PaCO2? 2) Does a change in ventilatory response to CO2 affect
cerebrovascular CO2 reactivity? 3) Does the commonly used
TCD-measured blood flow velocity in the MCA reflect blood
flow in the vicinity of the central chemoreceptors?
Is there direct evidence of a role for cerebrovascular CO2
reactivity in the ventilatory response to PaCO2? Recent elegant
studies have utilized powerful pharmacological manipulation
of CBF-CO2 reactivity to determine the influence of CBF
reactivity per se on steady-state ventilatory responsiveness.
The pharmacological intervention used in these studies was
indomethacin, a potent reversible cyclooxygenase inhibitor
that effectively decreases CBF and attenuates the cerebrovascular sensitivity to hypercapnia and hypocapnia (49, 61, 116,
193, 210, 214, 219, 222) without concomitant changes in
metabolic rate (222) or plasma catecholamines (31, 61, 116,
193). Such acute reductions in CBF-CO2 sensitivity result in an
E-CO2 sensitivity under steady-state testing conelevation in V
ditions (221). Moreover, as recently discussed elsewhere (219,
222), the direct influence of indomethacin on ventilation (32)
(via inhibition of prostaglandins) or the carotid body (90, 123,
218) seems negligible, and recent findings at sea level indicate
that indomethacin has no measurable effect on peripheral
chemosensitivity in response to transient hypoxia, hypercapnia,
or hyperoxia (P. N. Ainslie, unpublished data). This feature
makes indomethacin an ideal tool for investigating the effect of
CBF on the control of breathing in humans (see Fig. 18; also
see INSIGHT FROM MATHEMATICAL MODELING). Together, the aforementioned findings support the notion that cerebrovascular
CO2 reactivity plays a direct role in regulation of the ventilatory response to PaCO2.
Does a heightened ventilatory response to CO2 affect cerebrovascular CO2 reactivity? Although it has not been clearly
established, it seems reasonable to suggest that those individuals with a heightened ventilatory response to a step change in
PaCO2 (especially if the carotid bodies are sensitive to CO2) will
limit the increase in arterial and brain tissue PCO2; therefore,
E
the CO2 stimulus should be somewhat less than a low-V
responder. The influence of alterations in ventilatory responsiveness on cerebrovascular CO2 reactivity are likely to be
most obvious when tested under conditions of a steady-state
change in CO2, rather than by the rebreathing method (see

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
R1486

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

Methodological considerations for the integrated assessment


of cerebrovascular and ventilatory control).
Does the commonly used TCD blood flow velocity in the
MCA reflect blood flow in the vicinity of the central chemoreceptors? Important in the use of TCD is the issue that changes
in MCAv reflect changes in the posterior circulation supplying
the medulla (i.e., changes in CBF during CO2 reactivity are
likely to cause changes in central chemoreceptor stimulation).
As highlighted throughout this review, numerous efforts have
been made to establish normal criteria for CO2 reactivity of the
cerebral circulation by the use of TCD. Although there has
been ample research involving normal values for the anterior
half of the circle of Willis, mainly for the MCA, we have found
few studies involving these values for the basilar artery (19, 67,
80, 96, 139, 148, 154). Nonetheless, these studies seem to
support the notion that MCAv measurement is likely to reflect
blood flow in the region(s) of the central chemoreceptors.
In summary, evidence from animal and human studies indicates that the effects of CO2 on CBF provide an important
counterregulatory mechanism that serves to minimize changes
in brain [H], thereby stabilizing the breathing pattern in the
face of perturbations in PaCO2 (27, 50). An inappropriate CBF
response to a given change in CO2, in the hypercapnic or
hypocapnic range, may lead to breathing instability, in particular, development of central sleep apnea (CSA).
CSA IN CONGESTIVE HEART FAILURE AND AT
HIGH ALTITUDE

The pathogenesis of CSA, also referred to as Cheyne-Stokes


respiration, in congestive heart failure and at high altitude
remains incompletely understood (26, 28, 50, 109, 111). CSA
is an abnormal periodic breathing pattern in which central
apneas and hypopneas alternate with periods of hyperventilation that have a waxing-waning pattern of tidal volume that
classically has been associated with severe decompensated

heart failure (8, 174, 230). The key pathophysiological mechanism(s) leading to Cheyne-Stokes respiration is a fluctuation
of PaCO2 above and below the so-called apneic threshold (50,
221). In addition to the common occurrence of CSA in patients
with congestive heart failure and in healthy individuals after
ascent to high altitude, periodic breathing during sleep is
almost universal, occurring in 90% of humans (5). A reduction in cerebrovascular CO2 reactivity during wakefulness has
been identified in patients with congestive heart failure and
CSA; this reduction may affect stability of the breathing
pattern (26, 28, 50, 109, 111). More recently, a comparable
finding has been shown at high altitude (8). Such reductions in
cerebrovascular responsiveness to CO2 that provoke an increase in the sensitivity or gain of the chemoreflex control of
breathing may underpin breathing instability during CSA observed in patients with congestive heart failure and on ascent to
high altitude (Fig. 15).
In patients with congestive heart failure, however, a number
of destabilizing factors contribute to fluctuations in PaCO2.
First, a low PaCO2 close to the apneic threshold predisposes to
the development of central apneas (50); however, this will only
occur if the resting PCO2 is decreased because the chemoreflex
gain has increased. In support of this, a number of studies have
reported that PaCO2 is lower in patients with congestive heart
failure and Cheyne-Stokes respiration (75, 92, 112, 133, 198,
223). Naughton et al. (133) and Hanly et al. (75) found lower
PaCO2 during wakefulness and non-REM (NREM) sleep in
patients of comparable age, left ventricular ejection fraction,
PaO2, and lung-to-chemoreceptor circulatory delay with congestive heart failure with Cheyne-Stokes respiration than in
patients without congestive heart failure with Cheyne-Stokes
respiration. All episodes of Cheyne-Stokes respiration starting
during stage 2 sleep were almost always precipitated by hyperventilation in association with arousals from sleep. Furthermore, during Cheyne-Stokes episodes in stage 2 sleep, PaCO2

Fig. 15. Mechanisms involved in the genesis of CheyneStokes respiration and related overlap between congestive heart failure and high altitude. Alterations in cerebrovascular reactivity (in hyper- and hypocapnic range)
to CO2 as a consequence of congestive heart failure or
exposure to high altitude may affect chemosensitivity
and lead to elevations in total loop gain of the respiratory chemoreflex and breathing instability. Dashed lines
denote an additional influence, e.g., augmented gain of
chemoreceptors, secondary to pulmonary edema (as a
consequence of exposure to high altitude or congestive
heart failure), that destabilizes the respiratory control
system by making it more prone to ventilatory overshoot; or if inappropriate elevations in SNA occur
during development of congestive heart failure or at
high altitude, then its potential influence on constraining
cerebrovascular reactivity may affect breathing stability.
Balance between sleep and arousals also affect ventilation
and related arterial blood gases and stability. LF, left
ventricular. , Additive factors that may propagate development of central sleep apnea, e.g., development of hypocapnia as a consequence of congestive heart failure and
exposure to high altitude, acting to dive PaCO2 below the
apneic threshold during sleep; arousals, through promotion of hyperventilation, appear to facilitate, rather than
provoke, periodic breathing directly. Sleep-wakefulness
transitions may also produce respiratory control instability
via related alterations in upper airway resistance (for review
see Refs. 120 and 200).
AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

E
fell on average by 1.5 mmHg, which mirrored a 23% rise in V
(223). A lower resting PCO2 can be the result of chemoreflex
feedback gain increases and, also, chemoreflex threshold decreases; total loop gain (and thus stability) is determined in part
by the balance of feedback and plant gain (50). For example, if
feedback gain alone decreases the resting PCO2, the decrease in
plant gain will offset the increase in feedback gain, so that loop
gain increases only if the feedback gain increases more than
plant gain decreases; if loop gain increases, then stability will
be decreased (Fig. 15). Under this condition, a relatively small
increase in ventilation would drive PaCO2 below threshold and
trigger a central apnea (75). It is unknown how CBF and
CBF-reactivity might be altered during sleep in congestive
heart failure patients with and without CSA.
Patients with congestive heart failure and Cheyne-Stokes
respiration have an unusual response to sleep, in that their
PaCO2 levels do not increase in the progression from wakefulness to sleep (112), and, as a consequence, PaCO2 is closer
to their apneic threshold during sleep (112). Lorenzi-Filho
et al. (111, 112) abolished Cheyne-Stokes respiration in
congestive heart failure patients by inhalation treatment
with small concentrations of CO2 delivered by a mask.
Stabilization of breathing was achieved by a small (2
mmHg), but significant, increase in transcutaneous PCO2. In
contrast, supplemental O2 was not able to raise transcutaneous PCO2 and had no significant impact on the frequency
of Cheyne-Stokes respiration. This study reveals the critical
importance of small variations in PaCO2: a decline in PaCO2
below the apneic threshold triggers central apneas and
Cheyne-Stokes respiration. Arousals, through promotion of
hyperventilation, appear to facilitate, rather than provoke,
periodic breathing directly. More recently, we reported a
reduction of cerebrovascular CO2 reactivity at high altitude
compared with low altitude and a greater decrease in CBF
and loss of CA during sleep (8). The lowered CBF and
impaired CA during sleep at high altitude may cause
changes in central chemoreceptor stimulation and, thus, the
apneic threshold, which potentially could promote the
breathing instability at high altitude. This would seem
especially likely if the reduced CBF-CO2 reactivity, especially reductions in the CBF response to hypocapnia, during
wakefulness also occurs during sleep. The critical role of the
CBF response to hypocapnia during sleep has recently been
reported (219). In support of this notion, a strong correlation
was found between the severity of CSA and the change in
CBF (8). In other words, at high altitude, subjects with the
largest decrease in CBF during sleep had the highest number of
central events per hour (Fig. 16), suggesting a link between central
chemoreceptor stimulation and CBF. More recently, we extended these findings in a placebo-controlled, randomized
design (n 12), where indomethacin (100 mg po) and acetazolamide (10 mg/kg iv) were used during wakefulness and
before sleep to alter CBF and cerebrovascular reactivity to CO2
(11). These measures and blood gases were obtained before
and during each intervention at sea level and high altitude
(5,050 m, 6 days of acclimatization). Sleep, including CBF
monitoring, was studied using full polysomnography. Acetazolamide, relative to indomethacin, halved the frequency of
CSA (54 46 vs. 101 28 events/h, P 0.01) and elevated
MCAv more during NREM sleep. These results indicate that
reductions in CBF and CBF-CO2 reactivity play a key role in

R1487

Fig. 16. Relationship between degree of central sleep apnea and change in
MCAv from wakefulness to sleep at high altitude. Each point represents an
individual subject. Since there was no central sleep apnea at low altitude, there
was no relationship with the change in MCAv between wakefulness and sleep
(r 0.05). [Modified from Ainslie et al. (8).]

the pathogenesis of CSA at altitude (unpublished observations). Although acetazolamide-induced elevations in MCAv
may also provide additional protection against development of
CSA, we cannot dissociate this effect from its other known
prophylactic benefits (91, 106, 197).
Interestingly, our ongoing studies have shown a potential
difference in the influence of indomethacin between sea level
and high altitude. For example, in contrast to the findings at sea
level, administration of indomethacin (100 mg po) at 5,050 m
resulted in a greater reduction in MCAv-CO2 reactivity
(47 27% vs. 88 62%, P 0.01), led to periodic
breathing in 10 subjects during wakefulness (62) (see supplementary video in the online version of this article), and increased occurrence of CSA (Ainslie et al., unpublished observations).
In summary, reductions in CBF and cerebrovascular responsiveness to CO2, which provoke an increase in the sensitivity
or gain of the chemoreflex control of breathing, may underpin
breathing instability during CSA in patients with congestive
heart failure and on ascent to high altitude. However, many
other factors are also important in the development of CSA.
The integration of these key factors and the related comparison
between CSA in congestive heart failure and CSA at high
altitude are illustrated in Fig. 15, which shows considerable
overlap between the pathogenesis of CSA in congestive heart
failure and CSA that develops as a consequence of exposure to
high altitude. For example, the reduction in CO2 reactivity
results in larger changes in the [H] presented to the central
chemoreceptors (50, 186, 222). The potential changes in CBFCO2 reactivity are not exclusive to hypercapnia; rather, a
reduction in the CBF response to hypocapnia seems to be a
critical response. For example, a ventilatory overshoot would
lead to a greater fall in PaCO2, so the CBF response within the
hypocapnic range (in the context of CSA) is probably more
important (or at least equally important) during sleep than the
CBF response within the hypercapnic range (219). However, it
would seem likely that an inappropriate CBF response in either

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
R1488

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

range would promote breathing instability. Therefore, a reduced cerebrovascular response to CO2 (to hyper- or hypocapnia) might directly increase the susceptibility for periodic
breathing by increasing the central chemoreceptors contribution to loop gain. In support of this concept is the aforementioned study indicating that an indomethacin-induced alteration
in the cerebrovascular response to CO2 is sufficient to change
the ventilatory response to CO2 (Ainslie et al., unpublished
observations; 62, 219, 222). However, changes in cerebral
responsiveness to CO2 may only be partly sufficient to destabilize breathing, inasmuch as the carotid bodies play a significant role in the rapid ventilatory response to CO2 challenge
(184, 187, 188) (see Role of the carotid bodies in breathing
stability: interactions with central chemoreceptor stimulation
and CBF). Finally, within the hypocapnic range, input from the
central chemoreceptors would actually be reduced owing to a
maintained CBF (from a reduced CBF sensitivity) (219).
Role of the carotid bodies in breathing stability: interactions
with central chemoreceptor stimulation and CBF. Although we
have emphasized the influence of CBF in determining the PCO2
stimulus at central chemoreceptors, it is the degree of hypocapnia sensed at the carotid bodies that seems critical for the
development of a central apnea, especially during NREM sleep
(184, 187, 188). Moreover, the carotid bodies are an important
determinant in breathing stability at rest through tonic inputs
and, possibly, through compensation for a reduced central CO2
sensitivity (50, 98, 147, 186, 220). However, there may be a
marked degree of interaction between the peripheral and central chemoreceptors (46 48, 219). For example, elegant animal studies using an in situ dual-perfused preparation demonstrated that a single bout of hypoxia or hypercapnia elicited a
larger phrenic response when the brain stem was held at 25
mmHg PCO2 than at 50 mmHg PCO2. More recently, this study
was extended to show that the peripheral chemoreceptors were
more responsive to the lower brain stem PCO2, whether the
peripheral chemoreceptors received stimuli that increased (hypercapnia or hypoxia) or decreased activation (hyperoxia or
hypocapnia). These findings clearly demonstrate a negative
interaction between brain stem and peripheral chemosensitivity
in animals, and if such an interaction occurs in humans, it may
contribute to increased controller gain associated with sleeprelated breathing disorders. Moreover, because brain stem PCO2
affects the degree of peripheral chemosensitivity (48), these
findings again highlight the critical importance of CBF reactivity in determining the brain PCO2. Thus it seems possible that
the interactions of CBF reactivity that determine brain PCO2 at
the level of the central chemoreceptors might have a significant
influence on peripheral chemosensitivity.
It is tempting to speculate that any conditions that result in
chronic hypocapnia (e.g., congestive heart failure and high
altitude) and related alterations in cerebrovascular CO2 reactivity may result in different degrees of hypocapnia in the central
and peripheral compartments and, subsequently, breathing instability. Consistent with this notion, Xie et al. (219) recently suggested that the predominant role of peripheral vs. central chemoreceptors in causing apnea may be explained, in part, by preservation of PCO2 and [H] at the central chemoreceptors by
hypocapnia-induced cerebrovascular constriction. In their study,
indomethacin-induced reduction of cerebrovascular CO2 led to
higher CBF during hypocapnia, leading to instability and related
apnea (219). Another interpretation of these novel findings is that,

after indomethacin administration, brain PCO2 would likely be


higher, and, therefore, on the basis of the aforementioned animal
studies, the tonic input from the peripheral chemoreceptors would
be less and, therefore, may affect breathing stability (48).
EXERCISE

The mechanism(s) subserving ventilatory control during


exercise remain(s) controversial (17, 23, 164, 211); however,
traditionally, these mechanisms are proposed to include elements of proportional feedback, central and carotid chemosensory, and feedforward systems, central command, and muscle
reflex (211). Exercise increases ventilation via the respiratory
chemoreflex and also modifies the ventilatory response to CO2
(144); however, it remains unclear how exercise-induced alterations in the respiratory chemoreflex might influence dynamic CBF regulation, in particular, cerebrovascular reactivity
to CO2. This intriguing question has been examined in only
two studies, which showed that, during exercise, cerebrovascular CO2 reactivity at the operating point is enhanced (144,
168) (Fig. 4A). During such exercise, cerebrovascular CO2
reactivity to hypercapnia has been shown to be elevated during
exercise, whereas it was unchanged during hypocapnia (144).
Meadows et al. (124) reported that sleep decreased cerebral
CO2 reactivity, suggesting that the level of cerebral activation
influences the cerebrovascular reactivity to CO2. Exerciseinduced physiological changes (e.g., autonomic neural control)
may also modify the cerebral CO2 reactivity (4, 140). Ogoh
et al. (144) reported that, under conditions of hypercapnia and
exercise, the total respiratory loop gain was markedly reduced,
whereas cerebrovascular CO2 reactivity was increased. More
recently (141), using exponential regression modeling, we
extended these initial findings and examined the interaction
between the onset responses of the respiratory chemoreflex and
CBF at rest and during dynamic exercise. Our findings indicate
that, at rest, the faster washout of CO2 via hypercapnia-induced
cerebral vasodilation results in a reduced activation of the central
E. In
chemoreflex and subsequent slower onset response of V
contrast, during exercise, despite higher rates of increasing PaCO2,
there was no change in the onset response of CBF, and, therefore,
reduced washout of CO2 may act to augment the onset response
E (141). Together, these findings indicate that, despite an
of V
attenuated chemoreflex system controlling ventilation, elevations
in cerebrovascular reactivity might help maintain CO2 homeostasis in the brain during exercise. The extent to which this relationship might be influenced by factors such as exercise intensity,
duration, fitness, and aging is not known.
The major implications of alterations in CBF-CO2 reactivity
during exercise might well be related to cerebral oxygenation
and muscular energetics: if central [H] can be mitigated using
increased CBF over an increase in ventilation, energy might be
spared for muscular exercise. For example, cerebral hypoxia
has been proposed as a critical factor limiting exercise performance (138, 196). This recent report (196) demonstrated that
the hypoxia-induced peripheral chemoreflex activity led to
marked reductions in PaCO2, which were related to subsequent
reductions in MCAv and cerebral oxygenation. Interestingly,
the influence of the reductions in PaCO2 were much stronger
from moderate to high (75100% of maximal intensity) work
rates (196), indicating a distinct point from rising end-tidal
PCO2 and MCAv (during moderate-intensity exercise) to falling

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

R1489

K is slope of the linearized CO2 dissociation curve (0.005


mlml1 mmHg1).
For cerebrovascular reactivity
Q
rQ
s*PaCO PaCO
Q
(3)
2
2r
r is resting central blood flow (55 mlmin1 100 g1),
where Q
s is central blood flow sensitivity (2 mlmin1 100 g1 mmHg1),
Q
and Pa CO2(r) is resting PaCO2.
As demonstrated by Eqs. 2 and 3, central-arterial PCO2
difference (normally 10 mmHg) varies with medullary blood
flow. For example, an increase in PaCO2 of 10 mmHg above 40
mmHg results in an 8-mmHg reduction in central-arterial PCO2
difference because of an increase in medullary blood flow (Eq.
3). With a normal central chemoreflex sensitivity of 3
lmin1 mmHg1, ventilation response is reduced by 6 l/min.
Combining these equations

CO2/ K*Q
rQ
s*PaCO PaCO
PbCO2 PaCO2 V
2
2r
Fig. 17. Isoxic hyperoxic (PO2 150 mmHg) and hypoxic (PO2 50 mmHg)
ventilatory responses to PCO2 for modified rebreathing and steady-state (SS)
testing methods. Steady-state method responses show effect of cerebrovascular
reactivity (see model equation). Modified rebreathing method responses show
the effect of equalizing central-arterial PCO2 difference to mixed venous PCO2.
Chemo, responses to eliminating the effect of cerebrovascular reactivity by
assuming a constant 10-mmHg central-arterial PCO2 difference.

end-tidal PCO2 and MCAv (at maximal intensity). These findings are consistent with another recent study (13) that showed
a greater lowering of MCAv and cerebral oxygenation during
hypoxic exercise after acclimation to intermittent or continuous
hypoxia; these changes were mediated by a greater chemoreflex-induced hypocapnia, rather than a compromise in dynamic
cerebral autoregulation or alterations in MCAv-CO2 reactivity
(13). Thus, during strenuous or hypoxic exercise, hyperventilation lowers the PaCO2 and blunts the increase in CBF, which
can lead to an inadequate O2 delivery to the brain and contribute to the development of fatigue (for review see Ref. 138).
Further research is required to clearly establish whether the
modulation of CBF during exercise may influence central
motor drive and potentially limit peripheral muscle fatigue.

(4)
and with resting numerical values
s*PaCO 40
PbCO2 40 3/0.005*55 Q
2

(5)

Equation 5 was incorporated into the chemoreflex model and


used to predict the effect of cerebrovascular reactivity on the
isoxic ventilatory responses to PCO2; in effect, the model
compares steady-state with modified rebreathing responses
(Figs. 17 and 18, Tables 1 and 2). As illustrated in Fig. 17,
during isoxic conditions of hyperoxia and hypoxia, the steadystate method responses show the effect of cerebrovascular
reactivity (Eq. 5). In contrast, the modified rebreathing method
responses show the effect of equalizing the central-arterial
PCO2 difference to the mixed venous PCO2 and, therefore,

INSIGHT FROM MATHEMATICAL MODELING

Steady-state chemoreflex model with CBF effects. To further


systematically examine the effects of changes in CBF on the
chemoreflex control of breathing, we developed a new steadystate mathematical model. The chemoreflex model, including use
of modified Stewart equations (194) to convert PCO2 at the
chemoreceptors to [H], is described in detail elsewhere (52) and
was modified to include the influence of CBF. The effect of
changes in CBF with PaCO2 is modeled as a linear response of CBF
to changes in PaCO2, and the CBF determines the gradient between
medullary or brain PCO2 at the central chemoreceptors and PaCO2.
Equation 2 defines the steady-state relationship for the brain
compartment
CO2/K*Q

PbCO2 PaCO2 V

(2)

where PbCO2 is brain or central PCO2 (mmHg), PaCO2 is 40 mmHg


CO2 is central CO2 production (3 mlmin1 100 g1 at
at rest, V

rest), Q is central blood flow (55 mlmin1 100 g1 at rest), and

Fig. 18. Hyperoxic ventilatory responses to PCO2 for modified rebreathing


(Rebr) and steady-state (SS) testing methods for a sleeping subject with
(indo) and without (ctrl) indomethacin. Central strong ion difference was
changed to maintain central [H]. Note difference between hypocapnic and
hypercapnic sensitivities in control condition and increase in steady-state
sensitivity with indomethacin. Inset: predicted reduction of CBF reactivity
by indomethacin. Table 1 lists model parameters and Table 2 lists predicted
values.

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
R1490

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

Table 1. Predicted values for hyperoxic sleeping subject


with and without indomethacin
H , nmol/l
Central
Arterial
Resting PCO2, mmHg
Resting ventilation, l/min
Apnea threshold, mmHg
CO2 reserve, mmHg
Hypocapnic gain, l min1 mmHg1
Plant gain, mmHg l1 min1
Loop gain

Control

Indomethacin

42.0
41.0
46.1
6.6
40.0
6.1
1.1
6.9
7.5

42.2
39.8
44.2
7.0
40.0
4.2
1.7
6.3
10.6

H, H concentration.

E-venous PCO2
subsequently greater elevations in rebreathing V

sensitivity than the steady-state VE-PaCO2 sensitivity. The


Chemo responses show the effect of eliminating the cerebrovascular reactivity and assuming a constant 10-mmHg central-arterial PCO2 difference. Next, we considered the influence
of indomethacin on cerebrovascular CO2 reactivity during
sleep (Fig. 18). These findings from our mathematical model
support recent experimental data during wakefulness (222) and
sleep (219, 222). Moreover, the related elevations in hypocap E sensitivities underscore the potential importance of a
nic V
stable CBF reactivity in the hypocapnic and hypercapnia
ranges. An inappropriate CBF reactivity in either range may be
a critical factor leading to breathing instability. The preliminary findings from our newly described steady-state modeling
approach have significant potential to provide unique and
complementary insight into the effects of changes in CBF on
the chemoreflex control of breathing.
FUTURE DIRECTIONS

Clinging to the view that these systems operate as separate


physiological identities, as is still common practice in the
literature, can have damaging consequences for progress in this
area. Future research should concentrate on describing and
clarifying the integrative mechanisms operating in the cerebral
circulation and the role of these mechanisms in the chemoreflex control of breathing. The conceptual importance of this
direction for future research is a widening of our perception of
the complexity of the cerebrovascular bed and the intimate
relationship with breathing and systemic vascular control.
Additional methodological studies are need for the further
assessment of cerebrovascular CO2 reactivity using steadystate and rebreathing procedures, especially after interventional
investigations. Moreover, testing CO2 reactivity under ventilatory-dependent (i.e., after a fixed change in FICO2) and
ventilatory-independent (i.e., end-tidal clamping or rebreathing) is highly encouraged. Similarly, because of the established
interactions between the central and peripheral chemoreceptors
(46, 47, 219), the influence of CBF reactivity on this complex
relationship needs to be considered in human studies and in
intact anesthetized preparations. The combination of beat-tobeat assessment of MCAv (and basilar artery velocity) and
appropriate manipulation of end-tidal PCO2 with and without
pharmacological (e.g., indomethacin) interventions offers an
exciting means to examine such interrelationships between
CBF and cerebrovascular and ventilatory responsiveness to
CO2 in health and disease. The further utilization, development, and refinement of our newly described steady-state
modeling approach to quantify the effects of changes in CBF
on the chemoreflex control of breathing may also provide
unique and complementary insights.
ACKNOWLEDGMENTS

Over the last century, extensive progress has been made in


the study of CBF, respiratory, and cardiovascular control.

The authors thank Karen Peebles, Kate Thomas, and Dr. Sam Lucas for
skilled input and feedback.

Table 2. Model parameters used in simulations


Simulation
Sleep
Parameter

Symbol

Wakefulness

Control

Indomethacin

Peripheral sensitivity, mmHg


DpS vs. PO2, l min1 mmHg1 mmHg1
Rectangular hyperbola, mmHg l min1
Arterial SID, mmol/l
Arterial albumin concn, g/dl
Arterial phosphate concn, mmol/l
Central (brain) SID, mmol/l
Central albumin concn, g/dl
Central phosphate concn, mmol/l
Isoxic PO2, mmHg
Isoxic PO2, mmHg
Peripheral threshold, nmol/l
Central threshold, nmol l1 mmHg
Central sensitivity, l min1 mmHg1
Wakefulness drive, l/min
Drive threshold, l/min
CO2 production, ml/min
CBF sensitivity, ml min1 100 g1 mmHg1
CBF at rest, ml min1 100 g1
PCO2 at resting CBF, mmHg

PO2 asymptote
DpS asymptote
Area const
SIDa
Alba
PO
4 a
SIDc
Albc
PO
4 c
PO2 (hyperoxic)
PO2 (hypoxic)
Tp
Tc
DcS
W
TD
CO2
V
s
Q
r
Q
PaCO2(r)

30
0
17.8
38
4.2
1.16
31
1.25
0.61
150
50
34
34
1.8
7
18.5
300
2
55
40

30
0
17.8
38
4.2
1.16
38
1.25
0.61
150
50
34
34
3
0
18.5
230
2
42
40

30
0
17.8
38
4.2
1.16
39.2
1.25
0.61
150
50
34
34
3
0
18.5
230
0.3
38
40

, cardiac output.
See Ref. 52 for a detailed description of model equations. CBF, cerebral blood flow; SID, strong ion difference; Q
AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL
GRANTS
This study was supported by the New Zealand Lottery Grants Board, a
University of Otago Research Grant, Otago Medical Research Funding, the
Health Research Council of New Zealand, and the Marsden Fund.
REFERENCES
1. Aaslid R. Cerebral autoregulation and vasomotor reactivity. Front Neurol Neurosci 21: 216 228, 2006.
2. Aaslid R. Cerebral hemodynamics. In: Transcranial Doppler, edited by
Newell DW and Aaslid R. New York: Raven, 1992, p. 49 55.
3. Aaslid R, Lindegaard KF, Sorteberg W, Nornes H. Cerebral autoregulation dynamics in humans. Stroke 20: 4552, 1989.
4. Ainslie PN. Cardiorespiratory fitness: a simple, cheap and nonpharmocological means to prevent stroke? Stroke 40: 6, 2009.
5. Ainslie PN. Sympathetic activity does/does not influence cerebral blood
flow. J Appl Physiol 105: 1370, 2008.
6. Ainslie PN, Ashmead JC, Ide K, Morgan BJ, Poulin MJ. Differential
responses to CO2 and sympathetic stimulation in the cerebral and femoral
circulations in humans. J Physiol 566: 613 624, 2005.
7. Ainslie PN, Barach A, Cummings KJ, Murrell C, Hamlin M, Hellemans J. Cardiorespiratory and cerebrovascular responses to acute
poikilocapnic hypoxia following intermittent and continuous exposure to
hypoxia in humans. J Appl Physiol 102: 19531961, 2007.
8. Ainslie PN, Burgess K, Subedi P, Burgess KR. Alterations in cerebral
dynamics at high altitude following partial acclimatization in humans:
wakefulness and sleep. J Appl Physiol 102: 658 664, 2007.
9. Ainslie PN, Burgess KR. Cardiorespiratory and cerebrovascular responses to hyperoxic and hypoxic rebreathing: effects of acclimatization
to high altitude. Respir Physiol Neurobiol 161: 201209, 2008.
10. Ainslie PN, Celi L, McGrattan K, Peebles K, Ogoh S. Dynamic
cerebral autoregulation and baroreflex sensitivity during modest and
severe step changes in arterial PCO2. Brain Res 1230: 115124, 2008.
11. Ainslie PN, Cotter JD, Dawson A, Fan JL, Lucas RAI, Lucas SJE,
Peebles KN, Bilson K, Swart M, Thomas KN, and K. R. Burgess.
Influence of cerebral blood flow on central sleep apnea at high altitude
(Abstract). High Alt Med Biol. In press.
12. Ainslie PN, Cotter JD, George KP, Lucas S, Murrell C, Shave R,
Thomas KN, Williams JA, Atkinson G. Elevation in cerebral blood
flow velocity with aerobic fitness throughout healthy human ageing.
J Physiol 586: 4005 4010, 2008.
13. Ainslie PN, Hamlin M, Hellemans J, Rasmussen P, Ogoh S. Cerebral
hypoperfusion during hypoxic exercise following two different hypoxic
exposures: independence from changes in dynamic autoregulation and
reactivity. Am J Physiol Regul Integr Comp Physiol 295: R1613R1622,
2008.
14. Ainslie PN, Murrell C, Peebles K, Swart M, Skinner MA, Williams
MJ, Taylor RD. Early morning impairment in cerebral autoregulation
and cerebrovascular CO2 reactivity in healthy humans: relation to endothelial function. Exp Physiol 92: 769 777, 2007.
15. Ainslie PN, Poulin MJ. Ventilatory, cerebrovascular, and cardiovascular
interactions in acute hypoxia: regulation by carbon dioxide. J Appl
Physiol 97: 149 159, 2004.
16. Alm A, Bill A. The effect of stimulation of the cervical sympathetic
chain on retinal oxygen tension and on uveal, retinal and cerebral blood
flow in cats. Acta Physiol Scand 88: 84 94, 1973.
17. Asmussen E, Nielsen M. Ventilatory response to CO2 during work at
normal and at low oxygen tensions. Acta Physiol Scand 39: 2735, 1957.
18. Atkinson JL, Anderson RE, Sundt TM Jr. The effect of carbon
dioxide on the diameter of brain capillaries. Brain Res 517: 333340,
1990.
19. Barrett KM, Ackerman RH, Gahn G, Romero JM, Candia M.
Basilar and middle cerebral artery reserve: a comparative study using
transcranial Doppler and breath-holding techniques. Stroke 32: 2793
2796, 2001.
20. Bassetti C, Aldrich M. Night time versus daytime transient ischaemic
attack and ischaemic stroke: a prospective study of 110 patients. J Neurol
Neurosurg Psychiatry 67: 463 467, 1999.
21. Berger MG, Vandier C, Bonnet P, Jackson WF, Rusch NJ. Intracellular acidosis differentially regulates KV channels in coronary and pulmonary vascular muscle. Am J Physiol Heart Circ Physiol 275: H1351
H1359, 1998.
22. Bernardi L, Casucci G, Haider T, Brandstatter E, Pocecco E, Ehrenbourg I, Burtscher M. Autonomic and cerebrovascular abnormalities in
AJP-Regul Integr Comp Physiol VOL

R1491

mild COPD are worsened by chronic smoking. Eur Respir J 32: 1458
1465, 2008.
23. Bhattacharyya NK, Cunningham DJ, Goode RC, Howson MG,
Lloyd BB. The effects of hypoxia and light exercise on the respiratory
response to CO2 in man. J Physiol 194: 14P16P, 1968.
24. Birch AA, Dirnhuber MJ, Hartley-Davies R, Iannotti F, Neil-Dwyer
G. Assessment of autoregulation by means of periodic changes in blood
pressure. Stroke 26: 834 837, 1995.
25. Blaha M, Aaslid R, Douville CM, Correra R, Newell DW. Cerebral
blood flow and dynamic cerebral autoregulation during ethanol intoxication and hypercapnia. J Clin Neurosci 10: 195198, 2003.
26. Bradley TD, Floras JS. Pathophysiologic and therapeutic implications
of sleep apnea in congestive heart failure. J Card Fail 2: 223240, 1996.
27. Bradley TD, Floras JS. Sleep apnea and heart failure. II. Central sleep
apnea. Clin Med 107: 18221826, 2003.
28. Bradley TD, Lorenzi-Filho G, Floras JS. Pathophysiological interactions between sleep apnea and the heart. In: Respiratory-Circulatory
Interactions in Health and Disease, edited by Sharf SM, Pinsky MR, and
Magder S. New York: Informa Health Care, 2001, p. 577 611.
29. Brian JE Jr. Carbon dioxide and the cerebral circulation. Anesthesiology
88: 13651386, 1998.
30. Brian JE Jr, Faraci FM, Heistad DD. Recent insights into the regulation of cerebral circulation. Clin Exp Pharmacol Physiol 23: 449 457,
1996.
31. Bruhn H, Fransson P, Frahm J. Modulation of cerebral blood oxygenation by indomethacin: MRI at rest and functional brain activation. J
Magn Reson Imaging 13: 325334, 2001.
32. Bulow K. Respiration and wakefulness in man. Acta Physiol Scand Suppl
1110, 1963.
33. Busija DW, Heistad DD. Factors involved in the physiological regulation of the cerebral circulation. Rev Physiol Biochem Pharmacol 101:
161211, 1984.
34. Cassaglia PAGR, Walker AM. Sympathetic withdrawal augments cerebral blood flow during acute hypercapnia in sleeping lambs. Sleep 31:
1729 1734, 2008.
35. Cassaglia PA, Griffiths RI, Walker AM. Sympathetic nerve activity in
the superior cervical ganglia increases in response to imposed increases
in arterial pressure. Am J Physiol Regul Integr Comp Physiol 294:
R1255R1261, 2008.
36. Chapman RW, Santiago TV, Edelman NH. Effects of graded reduction of brain blood flow on chemical control of breathing. J Appl Physiol
47: 1289 1294, 1979.
37. Chauhan SD, Hobbs AJ, Ahluwalia A. C-type natriuretic peptide: new
candidate for endothelium-derived hyperpolarising factor. Int J Biochem
Cell Biol 36: 1878 1881, 2004.
38. Chesler M. Regulation and modulation of pH in the brain. Physiol Rev
83: 11831221, 2003.
39. Chun TH, Itoh H, Ogawa Y, Tamura N, Takaya K, Igaki T,
Yamashita J, Doi K, Inoue M, Masatsugu K, Korenaga R, Ando J,
Nakao K. Shear stress augments expression of C-type natriuretic peptide
and adrenomedullin. Hypertension 29: 1296 1302, 1997.
40. Claassen JA, Zhang R, Fu Q, Witkowski S, Levine BD. Transcranial
Doppler estimation of cerebral blood flow and cerebrovascular conductance during modified rebreathing. J Appl Physiol 102: 870 877, 2007.
41. Cohen PJ, Alexander SC, Smith TC, Reivich M, Wollman H. Effects
of hypoxia and normocarbia on cerebral blood flow and metabolism in
conscious man. J Appl Physiol 23: 183189, 1967.
42. Conroy D, Spielman A, Scott R. Daily rhythm of cerebral blood flow
velocity. J Circadian Rhythms 3: 3, 2005.
43. Cummings KJ, Swart M, Ainslie PN. Morning attenuation in cerebrovascular CO2 reactivity in humans is associated with a lowered cerebral
oxygenation and an augmented ventilatory response to CO2. J Appl
Physiol 102: 18911898, 2007.
43a.Dahan A, Nieuwenhuijs D, Teppema L. Plasticity of central chemoreceptors: effect of bilateral carotid body reaction on central CO2 sensitivity. PLoS Med 4: 11951203, 2007.
44. DAlecy LG, Rose CJ, Sellers SA. Sympathetic modulation of hypercapnic cerebral vasodilation in dogs. Circ Res 45: 771785, 1979.
45. Dawson SL, Blake MJ, Panerai RB, Potter JF. Dynamic but not static
cerebral autoregulation is impaired in acute ischaemic stroke. Cerebrovasc Dis 10: 126 132, 2000.
46. Day TA, Wilson RJ. Brainstem PCO2 modulates phrenic responses to
specific carotid body hypoxia in an in situ dual perfused rat preparation.
J Physiol 578: 843 857, 2007.
296 MAY 2009

www.ajpregu.org

Invited Review
R1492

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

47. Day TA, Wilson RJ. A negative interaction between brainstem and
peripheral respiratory chemoreceptors modulates peripheral chemoreflex
magnitude. J Physiol 587: 883 896, 2009.
48. Day TA, Wilson RJ. A negative interaction between central and peripheral respiratory chemoreceptors may underlie sleep-induced respiratory
instability: a novel hypothesis. Adv Exp Med Biol 605: 447 451, 2008.
49. DeGiulio PA, Roth RA, Mishra OP, Delivoria-Papadopoulos M,
Wagerle LC. Effect of indomethacin on the regulation of cerebral blood
flow during respiratory alkalosis in newborn piglets. Pediatr Res 26:
593597, 1989.
50. Dempsey JA. Crossing the apnoeic threshold: causes and consequences.
Exp Physiol 90: 1324, 2005.
51. Donahue MJ, Stevens RD, de Boorder M, Pekar JJ, Hendrikse J, van
Zijl PC. Hemodynamic changes after visual stimulation and breath
holding provide evidence for an uncoupling of cerebral blood flow and
volume from oxygen metabolism. J Cereb Blood Flow Metab 29:
176 185, 2009.
52. Duffin J. Role of acid-base balance in the chemoreflex control of
breathing. J Appl Physiol 99: 22552265, 2005.
53. Duffin J, Mahamed S. Adaptation in the respiratory control system. Can
J Physiol Pharmacol 81: 765773, 2003.
54. Duffin J, McAvoy GV. The peripheral-chemoreceptor threshold to
carbon dioxide in man. J Physiol 406: 1526, 1988.
55. Duffin J, Mohan RM, Vasiliou P, Stephenson R, Mahamed S. A
model of the chemoreflex control of breathing in humans: model parameters measurement. Respir Physiol 120: 1326, 2000.
56. Dumville J, Panerai RB, Lennard NS, Naylor AR, Evans DH. Can
cerebrovascular reactivity be assessed without measuring blood pressure
in patients with carotid artery disease? Stroke 29: 968 974, 1998.
57. Eames PJ, Potter JF, Panerai RB. Influence of controlled breathing
patterns on cerebrovascular autoregulation and cardiac baroreceptor sensitivity. Clin Sci (Lond) 106: 155162, 2004.
58. Edvinsson L, Krause DN. Cerebral Blood Flow and Metabolism.
Philadelphia: Lippincott, Williams & Wilkins, 2002.
59. Edwards M, Topor Z, Hughson R. A new two-breath technique for
extracting the cerebrovascular response to arterial carbon dioxide. Am J
Physiol Regul Integr Comp Physiol 284: R853R859, 2003.
60. Ellingsen I, Hauge A, Nicolaysen G, Thoresen M, Walloe L. Changes
in human cerebral blood flow due to step changes in PaO2 and PaCO2. Acta
Physiol Scand 129: 157163, 1987.
61. Eriksson S, Hagenfeldt L, Law D, Patrono C, Pinca E, Wennmalm A.
Effect of prostaglandin synthesis inhibitors on basal and carbon dioxide
stimulated cerebral blood flow in man. Acta Physiol Scand 117: 203211,
1983.
62. Fan JL, Burgess KR, Thomas KN, Peebles KC, Lucas RAI, Cotter
JD, Lucas SJE, Ainslie PN. Alterations in cerebral blood flow and
ventilatory reactivity to CO2 at high altitude: implications in the development of periodic breathing (Abstract). Proc NZ Physiol Soc Conf,
Dunedin, NZ, 2008, p. A91.
63. Fan JL, Cotter JD, Lucas RA, Thomas K, Wilson L, Ainslie PN.
Human cardiorespiratory and cerebrovascular function during severe
passive hyperthermia: effects of mild hypohydration. J Appl Physiol 105:
433 445, 2008.
64. Faraci FM, Sobey CG. Potassium channels and the cerebral circulation.
Clin Exp Pharmacol Physiol 23: 10911095, 1996.
65. Fencl V. Acid-Base Balance in Cerebral Fluids. Baltimore, MD: Waverly, 1986, p. 115140.
66. Fink BR. Influence of cerebral activity in wakefulness on regulation of
breathing. J Appl Physiol 16: 1520, 1961.
67. Garbin L, Habetswallner F, Clivati A. Vascular reactivity in middle
cerebral artery and basilar artery by transcranial Doppler in normal
subjects during hypoxia. Ital J Neurol Sci 18: 135137, 1997.
68. Giller CA, Bowman G, Dyer H, Mootz L, Krippner W. Cerebral
arterial diameters during changes in blood pressure and carbon dioxide
during craniotomy. Neurosurgery 32: 737741, 1993.
69. Golding EM, Marrelli SP, You J, Bryan RM Jr. Endothelium-derived
hyperpolarizing factor in the brain: a new regulator of cerebral blood
flow? Stroke 33: 661 663, 2002.
70. Grant DA, Franzini C, Wild J, Walker AM. Continuous measurement
of blood flow in the superior sagittal sinus of the lamb. Am J Physiol
Regul Integr Comp Physiol 269: R274 R279, 1995.
71. Greenberg HE, Sica A, Batson D, Scharf SM. Chronic intermittent
hypoxia increases sympathetic responsiveness to hypoxia and hypercapnia. J Appl Physiol 86: 298 305, 1999.
AJP-Regul Integr Comp Physiol VOL

72. Grubb RL Jr, Raichle ME, Eichling JO, ter-Pogossian MM. The
effects of changes in PaCO2 on cerebral blood volume, blood flow, and
vascular mean transit time. Stroke 5: 630 639, 1974.
73. Gupta AK, Menon DK, Czosnyka M, Smielewski P, Jones JG.
Thresholds for hypoxic cerebral vasodilation in volunteers. Anesth Analg
85: 817 820, 1997.
74. Hachinski VC, Potter P, Merskey H. Leuko-araiosis: an ancient term
for a new problem. Can J Neurol Sci 13: 533534, 1986.
75. Hanly P, Zuberi N, Gray R. Pathogenesis of Cheyne-Stokes respiration
in patients with congestive heart failure. Relationship to arterial PCO2.
Chest 104: 1079 1084, 1993.
76. Harper AM, Deshmukh VD, Rowan JO, Jennett WB. The influence of
sympathetic nervous activity on cerebral blood flow. Arch Neurol 27:
1 6, 1972.
77. Harper AM, Glass HI. Effect of alterations in the arterial carbon dioxide
tension on the blood flow through the cerebral cortex at normal and low
arterial blood pressures. J Neurol Neurosurg Psychiatry 28: 449 452,
1965.
78. Hastings AB. The changes in the concentration of carbon dioxide
resulting from changes in the volume of blood flowing through the
medulla oblongata. Am J Physiol 57: 104 109, 1921.
79. Heyman A, Birchfield RI, Sieker HO. Effects of bilateral cerebral
infarction on respiratory center sensitivity. Neurology 8: 694 700, 1958.
80. Hida W, Kikuchi Y, Okabe S, Miki H, Kurosawa H, Shirato K. CO2
response for the brain stem artery blood flow velocity in man. Respir
Physiol 104: 7175, 1996.
81. Hoth KF, Tate DF, Poppas A, Forman DE, Gunstad J, Moser DJ,
Paul RH, Jefferson AL, Haley AP, Cohen RA. Endothelial function
and white matter hyperintensities in older adults with cardiovascular
disease. Stroke 38: 308 312, 2007.
82. Iadecola C, Zhang F. Permissive and obligatory roles of NO in cerebrovascular responses to hypercapnia and acetylcholine. Am J Physiol
Regul Integr Comp Physiol 271: R990 R1001, 1996.
83. Ide K, Boushel R, Sorensen HM, Fernandes A, Cai Y, Pott F, Secher
NH. Middle cerebral artery blood velocity during exercise with 1adrenergic and unilateral stellate ganglion blockade in humans. Acta
Physiol Scand 170: 3338, 2000.
84. Ide K, Eliasziw M, Poulin MJ. Relationship between middle cerebral
artery blood velocity and end-tidal PCO2 in the hypocapnic-hypercapnic
range in humans. J Appl Physiol 95: 129 137, 2003.
85. Ide K, Pott F, Van Lieshout JJ, Secher NH. Middle cerebral artery
blood velocity depends on cardiac output during exercise with a large
muscle mass. Acta Physiol Scand 162: 1320, 1998.
86. Ito S, Mardimae A, Han J, Duffin J, Wells G, Fedorko L, Minkovich
L, Katznelson R, Meineri M, Arenovich T, Kessler C, Fisher JA.
Non-invasive prospective targeting of arterial PCO2 in subjects at rest.
J Physiol 586: 36753682, 2008.
87. Iturriaga R, Alcayaga J. Neurotransmission in the carotid body: transmitters and modulators between glomus cells and petrosal ganglion nerve
terminals. Brain Res Brain Res Rev 47: 46 53, 2004.
88. Iturriaga R, Varas R, Alcayaga J. Electrical and pharmacological
properties of petrosal ganglion neurons that innervate the carotid body.
Respir Physiol Neurobiol 157: 130 139, 2007.
89. Jackson WF. Potassium channels in the peripheral microcirculation.
Microcirculation 12: 113127, 2005.
90. Jansen AH, De Boeck C, Ioffe S, Chernick V. Indomethacin-induced
fetal breathing: mechanism and site of action. J Appl Physiol 57:
360 365, 1984.
91. Javaheri S. Acetazolamide improves central sleep apnea in heart failure:
a double-blind, prospective study. Am J Respir Crit Care Med 173:
234 237, 2006.
92. Javaheri S, Corbett WS. Association of low PaCO2 with central sleep
apnea and ventricular arrhythmias in ambulatory patients with stable
heart failure. Ann Intern Med 128: 204 207, 1998.
93. Jones NL, Robertson DG, Kane JW. Difference between end-tidal and
arterial PCO2 in exercise. J Appl Physiol 47: 954 960, 1979.
94. Jordan J, Shannon JR, Diedrich A, Black B, Costa F, Robertson D,
Biaggioni I. Interaction of carbon dioxide and sympathetic nervous
system activity in the regulation of cerebral perfusion in humans. Hypertension 36: 383388, 2000.
95. Kemna LJ, Posse S, Tellmann L, Schmitz T, Herzog H. Interdependence of regional and global cerebral blood flow during visual stimulation: an O-15-butanol positron emission tomography study. J Cereb
Blood Flow Metab 21: 664 670, 2001.
296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL
96. Kessler C, von Maravic M, Albrecht DM, von Maravic C, Muller M,
Schmidt P, Kompf D. Doppler CO2 test in patients with vertebrobasilar
ischemia. Acta Neurol Scand 84: 519 522, 1991.
97. Kety SS, Schmidt CF. The effects of altered arterial tensions of carbon
dioxide and oxygen on cerebral blood flow and cerebral oxygen consumption of normal young men. J Clin Invest 27: 484 492, 1948.
98. Khoo MC. Determinants of ventilatory instability and variability. Respir
Physiol 122: 167182, 2000.
99. Kinoshita H, Katusic ZS. Role of potassium channels in relaxations of
isolated canine basilar arteries to acidosis. Stroke 28: 433 438, 1997.
100. Kirkham FJ, Padayachee TS, Parsons S, Seargeant LS, House FR,
Gosling RG. Transcranial measurement of blood velocities in the basal
cerebral arteries using pulsed Doppler ultrasound: velocity as an index of
flow. Ultrasound Med Biol 12: 1521, 1986.
101. Kitazono T, Faraci FM, Taguchi H, Heistad DD. Role of potassium
channels in cerebral blood vessels. Stroke 26: 17131723, 1995.
103. Lahiri S, DeLaney RG. Stimulus interaction in the responses of carotid
body chemoreceptor single afferent fibers. Respir Physiol 24: 249 266,
1975.
104. Lassen NA. Cerebral blood flow and oxygen consumption in man.
Physiol Rev 39: 183238, 1959.
105. Lavi S, Gaitini D, Milloul V, Jacob G. Impaired cerebral CO2 vasoreactivity: association with endothelial dysfunction. Am J Physiol Heart
Circ Physiol 291: H1856 H1861, 2006.
106. Leaf DE, Goldfarb DS. Mechanisms of action of acetazolamide in the
prophylaxis and treatment of acute mountain sickness. J Appl Physiol
102: 13131322, 2007.
107. Leffler CW, Mirro R, Pharris LJ, Shibata M. Permissive role of
prostacyclin in cerebral vasodilation to hypercapnia in newborn pigs.
Am J Physiol Heart Circ Physiol 267: H285H291, 1994.
108. LeMarbre G, Stauber S, Khayat RN, Puleo DS, Skatrud JB, Morgan
BJ. Baroreflex-induced sympathetic activation does not alter cerebrovascular CO2 responsiveness in humans. J Physiol 551: 609 616, 2003.
109. Leung RS, Bradley TD. Sleep apnea and cardiovascular disease. Am J
Respir Crit Care Med 164: 21472165, 2001.
110. Levine BD, Zhang R. Sympathetic activity does/does not influence
cerebral blood flow. Autonomic control of the cerebral circulation is most
important for dynamic cerebral autoregulation. J Appl Physiol 105:
1369 1373, 2008.
111. Lorenzi-Filho G, Bradley TD. Cardiac function in sleep apnea. In:
Sleep Apnea: Pathogenesis, Diagnosis and Treatment, edited by Pack AI.
New York: Informa Health Care, 2002, p. 377 410.
112. Lorenzi-Filho G, Rankin F, Bies I, Douglas Bradley T. Effects of
inhaled carbon dioxide and oxygen on Cheyne-Stokes respiration in
patients with heart failure. Am J Respir Crit Care Med 159: 1490 1498,
1999.
113. Lucas SJ, Cotter JD, Murrell C, Wilson L, Anson JG, Gaze D,
George KP, Ainslie PN. Mechanisms of orthostatic intolerance following very prolonged exercise. J Appl Physiol 105: 213225, 2008.
114. Mandell DM, Han JS, Poublanc J, Crawley AP, Kassner A, Fisher
JA, Mikulis DJ. Selective reduction of blood flow to white matter during
hypercapnia corresponds with leukoaraiosis. Stroke 39: 19931998,
2008.
115. Manfredini R, Boari B, Smolensky MH, Salmi R, la Cecilia O, Maria
Malagoni A, Haus E, Manfredini F. Circadian variation in stroke onset:
identical temporal pattern in ischemic and hemorrhagic events. Chronobiol Int 22: 417 453, 2005.
116. Markus HS, Vallance P, Brown MM. Differential effect of three
cyclooxygenase inhibitors on human cerebral blood flow velocity and
carbon dioxide reactivity. Stroke 25: 1760 1764, 1994.
117. Markwalder TM, Grolimund P, Seiler RW, Roth F, Aaslid R.
Dependency of blood flow velocity in the middle cerebral artery on
end-tidal carbon dioxide partial pressurea transcranial ultrasound
Doppler study. J Cereb Blood Flow Metab 4: 368 372, 1984.
118. Marler JR, Price TR, Clark GL, Muller JE, Robertson T, Mohr JP,
Hier DB, Wolf PA, Caplan LR, Foulkes MA. Morning increase in
onset of ischemic stroke. Stroke 20: 473 476, 1989.
119. Marmarou A, Shulman K, LaMorgese J. Compartmental analysis of
compliance and outflow resistance of the cerebrospinal fluid system.
J Neurosurg 43: 523534, 1975.
120. Mateika JH, Narwani G. Intermittent hypoxia and respiratory plasticity
in animals and humans: does exposure to intermittent hypoxia promote or
mitigate sleep apnea? Exp Physiol 94: 279 296, 2009.

R1493

121. Mayberg TS, Lam AM, Matta BF, Visco E. The variability of cerebrovascular reactivity with posture and time. J Neurosurg Anesthesiol 8:
268 272, 1996.
122. McDonald DA. Blood Flow in Arteries. London: Arnold, 1974, p. 450.
123. McQueen DS, Belmonte C. The effects of prostaglandins E2, A2 and F2
on carotid baroreceptors and chemoreceptors. Q J Exp Physiol Cogn Med
Sci 59: 6371, 1974.
124. Meadows GE, Dunroy HMA, Morrell MJ, Corfield DR. Hypercapnic
cerebral vascular reactivity is decreased, in humans, during sleep compared with wakefulness. J Appl Physiol 94: 21972202, 2003.
125. Mitsis GD, Ainslie PN, Poulin MJ, Robbins PA, Marmarelis VZ.
Nonlinear modeling of the dynamic effects of arterial pressure and blood
gas variations on cerebral blood flow in healthy humans. Adv Exp Med
Biol 551: 259 265, 2004.
126. Mitsis GD, Zhang R, Levine BD, Marmarelis VZ. Cerebral hemodynamics during orthostatic stress assessed by nonlinear modeling. J Appl
Physiol 101: 354 366, 2006.
127. Mohan R, Duffin J. The effect of hypoxia on the ventilatory response to
carbon dioxide in man. Respir Physiol 108: 101115, 1997.
128. Mohan RM, Amara CE, Cunningham DA, Duffin J. Measuring
central-chemoreflex sensitivity in man: rebreathing and steady-state
methods compared. Respir Physiol 115: 2333, 1999.
129. Morrell MJ, Heywood P, Moosavi SH, Stevens J, Guz A. Central
chemosensitivity and breathing asleep in unilateral medullary lesion
patients: comparisons to animal data. Respir Physiol 129: 269 277,
2001.
130. Muller M, Marziniak M. The linear behavior of the system middle
cerebral artery flow velocity and blood pressure in patients with migraine: lack of autonomic control? Stroke 36: 1886 1890, 2005.
131. Murrell C, Wilson L, Cotter JD, Lucas S, Ogoh S, George K, Ainslie
PN. Alterations in autonomic function and cerebral hemodynamics to
orthostatic challenge following a mountain marathon. J Appl Physiol
103: 88 96, 2007.
132. Nattie EE, Li A. Central chemoreception is a complex system function
that involves multiple brain stem sites. J Appl Physiol. May 2008;
doi:10.1152/japplphysiol.00112.2008.
133. Naughton M, Benard D, Tam A, Rutherford R, Bradley TD. Role of
hyperventilation in the pathogenesis of central sleep apneas in patients
with congestive heart failure. Am Rev Respir Dis 148: 330 338, 1993.
134. Nelson MT, Quayle JM. Physiological roles and properties of potassium
channels in arterial smooth muscle. Am J Physiol Cell Physiol 268:
C799 C822, 1995.
135. Njemanze PC. Critical limits of pressure-flow relation in the human
brain. Stroke 23: 17431747, 1992.
136. North JB, Jennett S. Response of ventilation and of intracranial pressure
during rebreathing of carbon dioxide in patients with acute brain damage.
Brain 99: 169 182, 1976.
137. Noth U, Kotajima F, Deichmann R, Turner R, Corfield DR. Mapping
of the cerebral vascular response to hypoxia and hypercapnia using
quantitative perfusion MRI at 3 T. NMR Biomed 21: 464 472, 2008.
138. Nybo L, Rasmussen P. Inadequate cerebral oxygen delivery and central
fatigue during strenuous exercise. Exerc Sport Sci Rev 35: 110 118,
2007.
139. Ogawa S, Handa N, Matsumoto M, Etani H, Yoneda S, Kimura K,
Kamada T. Carbon dioxide reactivity of the blood flow in human basilar
artery estimated by the transcranial Doppler method in normal men: a
comparison with that of the middle cerebral artery. Ultrasound Med Biol
14: 479 483, 1988.
140. Ogoh S, Ainslie P. Regulatory mechanisms of cerebral blood flow during
exercise: new concepts. Exerc Sport Sci Rev. In press.
141. Ogoh S, Ainslie PN, Miyamoto T. Onset responses of ventilation and
cerebral blood flow to hypercapnia in humans: rest and exercise. J Appl
Physiol 106: 880 886, 2009.
142. Ogoh S, Brothers RM, Barnes Q, Eubank WL, Hawkins MN, Purkayastha S, O-Yurvati A, Raven PB. The effect of changes in cardiac
output on middle cerebral artery mean blood velocity at rest and during
exercise. J Physiol 569: 697704, 2005.
143. Ogoh S, Brothers S, Raven PB. Autonomic neural control of the
cerebral vasculature: acute hypotension. Stroke 39: 1979 1987, 2008.
144. Ogoh S, Hayashi N, Inagaki M, Ainslie P, Miyamoto T. Interaction
between the ventilatory and cerebrovascular responses to hypo- and
hypercapnia at rest and during exercise. J Physiol 586: 4327 4338, 2008.

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
R1494

INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

145. Olufsen MS, Nadim A, Lipsitz LA. Dynamics of cerebral blood flow
regulation explained using a lumped parameter model. Am J Physiol
Regul Integr Comp Physiol 282: R611R622, 2002.
146. Orem J, Lovering AT, Dunin-Barkowski W, Vidruk EH. Tonic
activity in the respiratory system in wakefulness, NREM and REM sleep.
Sleep 25: 488 496, 2002.
147. Pan L, Forster HM, Martino P, Strecker P, Beales J, Serra A, Lowry
T, Forster M, Forster AL. Important role of carotid afferents in control
of breathing. J Appl Physiol 85: 1299 1306, 1998.
148. Panczel G, Bonoczk P, Voko Z, Spiegel D, Nagy Z. Impaired vasoreactivity of the basilar artery system in patients with brainstem lacunar
infarcts. Cerebrovasc Dis 9: 218 223, 1999.
149. Pandit JJ, Mohan RM, Paterson ND, Poulin MJ. Cerebral blood flow
sensitivities to CO2 measured with steady-state and modified rebreathing
methods. Respir Physiol Neurobiol 159: 34 44, 2007.
150. Panerai RB. The critical closing pressure of the cerebral circulation.
Med Eng Phys 25: 621 632, 2003.
151. Panerai RB, Eames PJ, Potter JF. Multiple coherence of cerebral blood
flow velocity in humans. Am J Physiol Heart Circ Physiol 291: H251
H259, 2006.
152. Panerai RB, Evans DH, Naylor AR. Influence of arterial blood pressure
on cerebrovascular reactivity. Stroke 30: 12931295, 1999.
153. Parfenova H, Shibata M, Zuckerman S, Leffler CW. CO2 and cerebral
circulation in newborn pigs: cyclic nucleotides and prostanoids in vascular regulation. Am J Physiol Heart Circ Physiol 266: H1494 H1501,
1994.
154. Park CW, Sturzenegger M, Douville CM, Aaslid R, Newell DW.
Autoregulatory response and CO2 reactivity of the basilar artery. Stroke
34: 34 39, 2003.
155. Paulson OB, Strandgaard S, Edvinsson L. Cerebral autoregulation.
Cerebrovasc Brain Metab Rev 2: 161192, 1990.
156. Peebles K, Celi L, McGrattan K, Murrell C, Thomas K, Ainslie PN.
Human cerebrovascular and ventilatory CO2 reactivity to end-tidal,
arterial and internal jugular vein PCO2. J Physiol 584: 347357, 2007.
157. Peebles KC, Richards AM, Celi L, McGrattan K, Murrell CJ, Ainslie
PN. Human cerebral arteriovenous vasoactive exchange during alterations in arterial blood gases. J Appl Physiol 105: 1060 1068, 2008.
158. Peppiatt CM, Howarth C, Mobbs P, Attwell D. Bidirectional control
of CNS capillary diameter by pericytes. Nature 443: 700 704, 2006.
159. Perod AL, Roberts AE, McKinney WM. Caffeine can affect velocity in
the middle cerebral artery during hyperventilation, hypoventilation, and
thinking: a transcranial Doppler study. J Neuroimaging 10: 3338, 2000.
160. Piechnik SK, Chiarelli PA, Jezzard P. Modelling vascular reactivity to
investigate the basis of the relationship between cerebral blood volume
and flow under CO2 manipulation. Neuroimage 39: 107118, 2008.
161. Pike FH. The relation of the rate of blood flow through the medulla
oblongata to the amplitude and frequency of respiratory movements.
Science 48: 121122, 1918.
162. Plum F, Brown HW. The effect on respiration of central nervous system
disease. Ann NY Acad Sci 109: 915931, 1963.
163. Plum F, Brown HW. Hypoxic-hypercapnic interaction in subjects with
bilateral cerebral dysfunction. J Appl Physiol 18: 1139 1145, 1963.
164. Poon CS, Greene JG. Control of exercise hyperpnea during hypercapnia
in humans. J Appl Physiol 59: 792797, 1985.
165. Poulin MJ, Liang PJ, Robbins PA. Dynamics of the cerebral blood
flow response to step changes in end-tidal PCO2 and PO2 in humans.
J Appl Physiol 81: 1084 1095, 1996.
166. Poulin MJ, Liang PJ, Robbins PA. Fast and slow components of
cerebral blood flow response to step decreases in end-tidal PCO2 in
humans. J Appl Physiol 85: 388 397, 1998.
167. Przybylowski T, Fea BM. Mechanisms of the cerebrovascular response
to apnea. J Physiol 31: 1729 1734, 2003.
168. Rasmussen P, Stie H, Nielsen B, Nybo L. Enhanced cerebral CO2
reactivity during strenuous exercise in man. Eur J Appl Physiol 96:
299 304, 2006.
169. Read DJ, Leigh J. Blood-brain tissue PCO2 relationships and ventilation
during rebreathing. J Appl Physiol 23: 5370, 1967.
170. Reivich M. Arterial PCO2 and cerebral hemodynamics. Am J Physiol 206:
2535, 1964.
171. Ringelstein EB, Sievers C, Ecker S, Schneider PA, Otis SM. Noninvasive assessment of CO2-induced cerebral vasomotor response in normal individuals and patients with internal carotid artery occlusions.
Stroke 19: 963969, 1988.

172. Rowell LB. Control of regional blood flow during dynamic exercise. In:
Human Cardiovascular Control, edited by Rowell LB. New York:
Oxford University Press, 1993, p. 204 254.
173. Rubanyi GM, Romero JC, Vanhoutte PM. Flow-induced release of
endothelium-derived relaxing factor. Am J Physiol Heart Circ Physiol
250: H1145H1149, 1986.
174. Salvaggio A, Insalaco G, Marrone O, Romano S, Braghiroli A,
Lanfranchi P, Patruno V, Donner CF, Bonsignore G. Effects of
high-altitude periodic breathing on sleep and arterial oxyhaemoglobin
saturation. Eur Respir J 12: 408 413, 1998.
175. Sander K, Hof U, Poppert H, Conrad B, Sander D. Improved cerebral
vasoreactivity after statin administration in healthy adults. J Neuroimaging 15: 266 270, 2005.
176. Schmidt CF. The influence of cerebral blood flow on respiration. Am J
Physiol 84: 202241, 1928.
177. Secher NH, Seifert T, Van Lieshout JJ. Cerebral blood flow and
metabolism during exercise: implications for fatigue. J Appl Physiol 104:
306 314, 2008.
178. Serrador JM, Picot PA, Rutt BK, Shoemaker JK, Bondar RL. MRI
measures of middle cerebral artery diameter in conscious humans during
simulated orthostasis. Stroke 31: 16721678, 2000.
179. Serrador JM, Sorond FA, Vyas M, Gagnon M, Iloputaife ID, Lipsitz
LA. Cerebral pressure-flow relations in hypertensive elderly humans:
transfer gain in different frequency domains. J Appl Physiol 98: 151159,
2005.
180. Severinghaus JW, Lassen N. Step hypocapnia to separate arterial from
tissue PCO2 in the regulation of cerebral blood flow. Circ Res 20:
272278, 1967.
181. Shapiro E, Wasserman AJ, Patterson JL Jr. Human cerebrovascular
response time to elevation of arterial carbon dioxide tension. Arch Neurol
13: 130 138, 1965.
182. Shapiro W, Wasserman AJ, Patterson JJ. Mechanism and pattern of
human cerebrovascular regulation after rapid changes in blood CO2
tension. J Clin Invest 45: 913922, 1966.
183. Shea SA. Behavioural and arousal-related influences on breathing in
humans. Exp Physiol 81: 126, 1996.
184. Skatrud JB, Dempsey JA. Interaction of sleep state and chemical
stimuli in sustaining rhythmic ventilation. J Appl Physiol 55: 813 822,
1983.
185. Slessarev M, Han J, Mardimae A, Prisman E, Preiss D, Volgyesi G,
Ansel C, Duffin J, Fisher JA. Prospective targeting and control of
end-tidal CO2 and O2 concentrations. J Physiol 581: 12071219, 2007.
186. Smith CA, Rodman JR, Chenuel BJ, Henderson KS, Dempsey JA.
Response time and sensitivity of the ventilatory response to CO2 in
unanesthetized intact dogs: central vs. peripheral chemoreceptors. J Appl
Physiol 100: 1319, 2006.
187. Smith CA, Chenuel BJ, Henderson KS, Dempsey JA. The apneic
threshold during non-REM sleep in dogs: sensitivity of carotid body vs.
central chemoreceptors. J Appl Physiol 103: 578 586, 2007.
188. Smith CA, Nakayama H, Dempsey JA. The essential role of carotid
body chemoreceptors in sleep apnea. Can J Physiol Pharmacol 81:
774 779, 2003.
189. Smith JJ, Lee JG, Hudetz AG, Hillard CJ, Bosnjak ZJ, Kampine JP.
The role of nitric oxide in the cerebrovascular response to hypercapnia.
Anesth Analg 84: 363369, 1997.
190. Somogyi RB, Preiss D, Vesely A, Fisher JA, Duffin J. Changes in
respiratory control after 5 days at altitude. Respir Physiol Neurobiol 145:
4152, 2005.
191. Spengos K, Tsivgoulis G, Manios E, Synetou M, Vassilopoulou S,
Zakopoulos N, Vassilopoulos D, Vemmos KN. Stroke etiology is
associated with symptom onset during sleep. Sleep 28: 233238, 2005.
192. St Croix CM, Cunningham DA, Paterson DH. Nature of the interaction between central and peripheral chemoreceptor drives in human
subjects. Can J Physiol Pharmacol 74: 640 646, 1996.
193. St Lawrence KS, Ye FQ, Lewis BK, Weinberger DR, Frank JA,
McLaughlin AC. Effects of indomethacin on cerebral blood flow at rest
and during hypercapnia: an arterial spin tagging study in humans. J Magn
Reson Imaging 15: 628 635, 2002.
194. Stewart PA. Modern quantitative acid-base chemistry. Can J Physiol
Pharmacol 61: 1444 1461, 1983.
195. Strandgaard S, Sigurdsson ST. Sympathetic nerve activity does not
influence cerebral blood flow. J Appl Physiol 105: 1366 1367, 2008.
196. Subudhi AW, Lorenz MC, Fulco CS, Roach RC. Cerebrovascular
responses to incremental exercise during hypobaric hypoxia: effect of

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org

Invited Review
INFLUENCE OF BRAIN BLOOD FLOW ON VENTILATORY CONTROL

197.
198.

199.
200.
201.
202.
203.
204.

205.
206.
207.
208.

209.
210.

211.
212.
213.

oxygenation on maximal performance. Am J Physiol Heart Circ Physiol


294: H164 H171, 2008.
Swenson ER, Teppema LJ. Prevention of acute mountain sickness by
acetazolamide: as yet an unfinished story. J Appl Physiol 102: 1305
1307, 2007.
Tkacova R, Hall MJ, Liu PP, Fitzgerald FS, Bradley TD. Left
ventricular volume in patients with heart failure and Cheyne-Stokes
respiration during sleep. Am J Respir Crit Care Med 156: 1549 1555,
1997.
Toda N, Okamura T. Cerebral vasodilators. Jpn J Pharmacol 76:
349 367, 1998.
Topor ZL, Vasilakos K, Younes M, Remmers JE. Model-based
analysis of sleep disordered breathing in congestive heart failure. Respir
Physiol Neurobiol 155: 8292, 2007.
Torrance RW. Prolegomena. Chemoreception upstream of transmitters.
Adv Exp Med Biol 410: 1338, 1996.
Treib J, Haass A, Koch D, Grauer MT, Schimrigk K. Influence of
blood pressure and cardiac output on cerebral blood flow and autoregulation in acute stroke measured by TCD. Eur J Neurol 3: 539 543, 1996.
Ursino M, Lodi CA. A simple mathematical model of the interaction
between intracranial pressure and cerebral hemodynamics. J Appl
Physiol 82: 1256 1269, 1997.
Valdueza JM, Balzer JO, Villringer A, Vogl TJ, Kutter R, Einhaupl
KM. Changes in blood flow velocity and diameter of the middle cerebral
artery during hyperventilation: assessment with MR and transcranial
Doppler sonography. AJNR Am J Neuroradiol 18: 1929 1934, 1997.
van Lieshout JJ, Pott F, Madsen PL, van Goudoever J, Secher NH.
Muscle tensing during standing: effects on cerebral tissue oxygenation
and cerebral artery blood velocity. Stroke 32: 1546 1551, 2001.
Van Lieshout JJ, Secher NH. Sympathetic activity does/does not
influence cerebral blood flow. J Appl Physiol 105: 1364 1366, 2008.
Vantanajal JS, Ashmead JC, Anderson TJ, Hepple RT, Poulin MJ.
Differential sensitivities of cerebral and brachial blood flow to hypercapnia in humans. J Appl Physiol 102: 8793, 2007.
Vernieri F, Pasqualetti P, Matteis M, Passarelli F, Troisi E, Rossini
PM, Caltagirone C, Silvestrini M. Effect of collateral blood flow and
cerebral vasomotor reactivity on the outcome of carotid artery occlusion.
Stroke 32: 15521558, 2001.
Walters M, Muir S, Shah I, Lees K. Effect of perindopril on cerebral
vasomotor reactivity in patients with lacunar infarction. Stroke 35:
1899 1902, 2004.
Wang Q, Paulson OB, Lassen NA. Indomethacin abolishes cerebral
blood flow increase in response to acetazolamide-induced extracellular
acidosis: a mechanism for its effect on hypercapnia? J Cereb Blood Flow
Metab 13: 724 727, 1993.
Ward SA. Muscle-energetic and cardio-pulmonary determinants of exercise tolerance in humans. Exp Physiol 92: 321322, 2007.
Wasserman AJ, Patterson JL Jr. The cerebral vascular response to
reduction in arterial carbon dioxide tension. J Clin Invest 40: 12971303,
1961.
Wei EP, Kontos HA, Patterson JL Jr. Dependence of pial arteriolar
response to hypercapnia on vessel size. Am J Physiol Heart Circ Physiol
238: H697H703, 1980.

R1495

214. Wennmalm A, Carlsson I, Edlund A, Eriksson S, Kaijser L, Nowak


J. Central and peripheral haemodynamic effects of non-steroidal antiinflammatory drugs in man. Arch Toxicol Suppl 7: 350 359, 1984.
215. West L, McHenry C, Cherniack NS. Transient responses of cerebral
blood flow and ventilation to changes in PaCO2 in normal subjects and
patients with cerebrovascular disease. Stroke 7: 584 590, 1976.
216. Widder B, Kleiser B, Krapf H. Course of cerebrovascular reactivity in
patients with carotid artery occlusions. Stroke 25: 19631967, 1994.
217. Wijnhoud AD, Koudstaal PJ, Dippel DW. Relationships of transcranial blood flow Doppler parameters with major vascular risk factors:
TCD study in patients with a recent TIA or nondisabling ischemic stroke.
J Clin Ultrasound 34: 70 76, 2006.
218. Wolsink JG, Berkenbosch A, DeGoede J, Olievier CN. The influence
of indomethacin on the ventilatory response to CO2 in newborn anaesthetized piglets. J Physiol 477: 339 345, 1994.
219. Xie A, Skatrud JB, Barczi SR, Reichmuth K, Morgan BJ, Mont S,
Dempsey JA. Influence of cerebral blood flow on breathing stability.
J Appl Physiol 106: 850 856, 2009.
220. Xie A, Skatrud JB, Dempsey JA. Effect of hypoxia on the hypopnoeic
and apnoeic threshold for CO2 in sleeping humans. J Physiol 535:
269 278, 2001.
221. Xie A, Skatrud JB, Khayat R, Dempsey JA, Morgan B, Russell D.
Cerebrovascular response to carbon dioxide in patients with congestive
heart failure. Am J Respir Crit Care Med 172: 371378, 2005.
222. Xie A, Skatrud JB, Morgan B, Chenuel B, Khayat R, Reichmuth K,
Lin J, Dempsey JA. Influence of cerebrovascular function on the
hypercapnic ventilatory response in healthy humans. J Physiol 577:
319 329, 2006.
223. Xie A, Skatrud JB, Puleo DS, Rahko PS, Dempsey JA. Apneahypopnea threshold for CO2 in patients with congestive heart failure.
Am J Respir Crit Care Med 165: 12451250, 2002.
224. Xu H, Cui N, Yang Z, Wu J, Giwa LR, Abdulkadir L, Sharma P,
Jiang C. Direct activation of cloned KATP channels by intracellular
acidosis. J Biol Chem 276: 12898 12902, 2001.
225. Yang SP, Bergo GW, Krasney E, Krasney JA. Cerebral pressure-flow
and metabolic responses to sustained hypoxia: effect of CO2. J Appl
Physiol 76: 303313, 1994.
226. Younes M, Ostrowski M, Atkar R, Laprairie J, Siemens A, Hanly P.
Mechanisms of breathing instability in patients with obstructive sleep
apnea. J Appl Physiol 103: 1929 1941, 2007.
227. Zhang R, Levine BD. Autonomic ganglionic blockade does not prevent
reduction in cerebral blood flow velocity during orthostasis in humans.
Stroke 38: 1238 1244, 2007.
228. Zhang R, Zuckerman JH, Giller CA, Levine BD. Transfer function
analysis of dynamic cerebral autoregulation in humans. Am J Physiol
Heart Circ Physiol 274: H233H241, 1998.
229. Zhang R, Zuckerman JH, Iwasaki K, Wilson TE, Crandall CG,
Levine BD. Autonomic neural control of dynamic cerebral autoregulation in humans. Circulation 106: 1814 1820, 2002.
230. Zielinski J, Koziej M, Mankowski M, Sarybaev AS, Tursalieva JS,
Sabirov IS, Karamuratov AS, Mirrakhimov MM. The quality of sleep
and periodic breathing in healthy subjects at an altitude of 3,200 m. High
Alt Med Biol 1: 331336, 2000.

AJP-Regul Integr Comp Physiol VOL

296 MAY 2009

www.ajpregu.org