Download JAP March 88/3

J. Appl. Physiol.
88: 1084–1092, 2000.
Effects of arousal and sleep state on systemic
and pulmonary hemodynamics in obstructive apnea
H. SCHNEIDER, C. D. SCHAUB, C. A. CHEN, K. A. ANDREONI,
A. R. SCHWARTZ, P. L. SMITH, J. L. ROBOTHAM, AND C. P. O’DONNELL
Department of Medicine, Division of Pulmonary and Critical Care Medicine,
Department of Anesthesiology and Critical Care Medicine, and Department
of Surgery, Johns Hopkins University, Baltimore, Maryland 21224
airway obstruction; arterial pressure; canine; heart rate;
hypoxia; non-rapid-eye-movement sleep; rapid-eye-movement sleep; stroke volume
inspiration (18, 24), peak arterial blood gas disturbances (23), and arousal from sleep (14, 19). A primary
purpose of this study is to examine the effect of eliminating arousal on the cardiovascular response to obstructive apneas that are otherwise matched for respiratory
characteristics and arterial Hb desaturation.
The cardiovascular responses to apnea and arousal
may be dependent on the sleep state before arousal.
Consider that our previous work in dogs during normal
unobstructed sleep has demonstrated a differential
sleep state effect on systemic and pulmonary hemodynamics (21). Specifically, Psa and total peripheral resistance were significantly lower in rapid-eye-movement
(REM) sleep compared with non-rapid-eye-movement
(NREM) sleep. The Ppa and vascular resistance during
normal unobstructed sleep, however, were the same
during NREM and REM sleep. Thus it is possible that,
during OSA, sleep state may have a differential effect
on Psa and Ppa responses to apnea and arousal.
In this study, we examine the effects of arousal and
sleep state on Psa, Ppa, and ventricular SV in a canine
model of OSA. Specifically, we 1) examine the effects of
arousal on Psa, Ppa, and ventricular SV responses to
obstructive apnea and 2) determine what role the
parasympathetic nervous system plays in these responses to arousal. Furthermore, we 3) examine
whether the cardiovascular response to obstructive
apnea and arousal differs between NREM and REM
sleep.
METHODS
Surgical Procedures
(OSA) is associated with
cyclical cardiovascular changes that are tightly linked
to the periodicity of the apnea (3, 27). Maximal changes
in several cardiovascular factors occur in the immediate postapneic period. Systemic (Psa) and pulmonary
(Ppa) arterial pressures peak after apnea termination,
whereas left and right ventricular stroke volumes (SV)
reach a nadir (2, 20, 21, 26). These cardiovascular
responses are associated with changes in lung volume
during the transition from obstructed to unobstructed
OBSTRUCTIVE SLEEP APNEA
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
1084
Experiments were performed on six mongrel dogs (3 male
and 3 female) in the weight range of 20–25 kg. The animals
were pretreated with fentanyl (0.4 mg im) and droperidol (20
mg im) and anesthetized with pentobarbital sodium (30
mg/kg iv), and mechanical ventilation was initiated. A chronic
tracheal stoma was created by removing the ventral portion
of five tracheal rings. Tygon catheters (0.05 in. ID, 0.09 in.
OD) were introduced into the right femoral artery and vein
and advanced rostrally to the thoracic aorta and inferior vena
cava, respectively. A left thoracotomy (fourth interspace) was
performed, and electromagnetic flow probes (Zepeda Instruments, Seattle, WA) were placed around the ascending aorta
in two dogs, around the pulmonary artery in two dogs, and
around both the ascending aorta and the pulmonary artery in
two dogs as previously described (21). The size of each
circular flow probe was chosen to closely match the external
8750-7587/00 $5.00 Copyright r 2000 the American Physiological Society
http://www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.33.6 on September 17, 2016
Schneider, H., C. D. Schaub, C. A. Chen, K. A. Andreoni, A. R. Schwartz, P. L. Smith, J. L. Robotham, and
C. P. O’Donnell. Effects of arousal and sleep state on
systemic and pulmonary hemodynamics in obstructive apnea. J. Appl. Physiol. 88: 1084–1092, 2000.—During obstructive sleep apnea (OSA), systemic (Psa) and pulmonary (Ppa)
arterial pressures acutely increase after apnea termination,
whereas left and right ventricular stroke volumes (SV) reach
a nadir. In a canine model (n ⫽ 6), we examined the effects of
arousal, parasympathetic blockade (atropine 1 mg/kg iv), and
sleep state on cardiovascular responses to OSA. In the
absence of arousal, SV remained constant after apnea termination, compared with a 4.4 ⫾ 1.7% decrease after apnea with
arousal (P ⬍ 0.025). The rise in transmural Ppa was independent of arousal (4.5 ⫾ 1.0 vs. 4.1 ⫾ 1.2 mmHg with and
without arousal, respectively), whereas Psa increased more
after apnea termination in apneas with arousal compared
with apneas without arousal. Parasympathetic blockade abolished the arousal-induced increase in Psa, indicating that
arousal is associated with a vagal withdrawal of the parasympathetic tone to the heart. Rapid-eye-movement (REM) sleep
blunted the increase in Psa (pre- to end-apnea: 5.6 ⫾ 2.3
mmHg vs. 10.3 ⫾ 1.6 mmHg, REM vs. non-REM, respectively,
P ⬍ 0.025), but not transmural Ppa, during an obstructive
apnea. We conclude that arousal and sleep state both have
differential effects on the systemic and pulmonary circulation
in OSA, indicating that, in patients with underlying cardiovascular disease, the hemodynamic consequences of OSA may be
different for the right or the left side of the circulation.
AROUSAL AND SLEEP STATE IN OSA
Apparatus and Methods of Measurement
A custom-designed endotracheal tube was used to control
airway patency, measure arterial Hb saturation (averaged
every 1 s; minimum detectable change of 1%), and allow
sampling of end-tidal carbon dioxide and measurement of
tracheal pressure (Ptr) from side ports (16). After atropine
administration, dilation of the trachea precluded reliable
contact of the oximeter probe with the mucosal lining and
prevented measurement of arterial Hb saturation, as previously discussed (14). The resistance of the endotracheal tube
and the time constants for obstructing and restoring airway
patency have been previously described (16). The connections
from the endotracheal tube, the polysomnographic extension
leads, and the vascular lines were placed in a 40-in.-long
flexible tube that attached to the back of the animal’s jacket.
The animals slept in a specially constructed box with a
clear Plexiglas front panel that could be monitored from an
adjacent room with a shortwave closed-circuit television. The
flexible tube containing the recording wires and catheters
exited through a hole in the top of the sleep box, passed under
a communicating door, and attached to the recording equipment in the adjacent room. Intravascular and airway pressures measurements were made with pressure transducers
(Cobe, Lakewood, CO) referenced to zero at midthoracic level
with the animal lying prone. Calibrations were checked at
30-min intervals throughout experiments. A pen recorder
(Grass Instruments, Quincy, MA) was used to record EEG
and EMG activity and Psa, Ppa, Pla, Ppl, and Ptr traces. The
SV of the left and right ventricles were measured with a
model SWF-5RD electromagnetic flowmeter (Zepeda Instruments). A Nellcor N-200 pulse oximeter (Haywood, CA)
measured arterial Hb saturation, and a Beckman analyzer
(Anaheim, CA) sampled end-tidal carbon dioxide. Both instru-
ments were connected to the pen recorder. Data from the pen
recorder were sampled at 300 Hz, converted to digital format
(DI-200 data acquisition board; Dataq Instruments, Akron,
OH), and acquired to optical disk for storage with Windaq/200
acquisition software (Dataq Instruments). The velocity signal
from the electromagnetic flow probe was digitally integrated
to determine SV. Transmural pulmonary arterial pressure
(Ppa ⫺ Ppl; Ppa,tm) and transmural left atrial pressure
(Pla ⫺ Ppl; Pla,tm) were derived digitally for analysis during
periods of obstructive apnea.
Experimental Protocol
We performed two separate experiments at least 1 wk
apart in each animal. Before each experiment, the animals
were sleep deprived for a period of 24 h as previously
described (14, 15, 21). Experiments were conducted between
1400 and 2000. In the first experiment, a series of apneas was
induced in six animals during both NREM and REM sleep.
The duration of airway obstruction was controlled such that
restoration of airway patency was associated with either the
absence or the presence of an EEG arousal (as previously
described; Ref. 14). In the second experiment, a series of
airway obstructions with arousal was induced in six animals
during NREM sleep before (control) and after pharmacological blockade of the parasympathetic nervous system with
atropine (intravenous infusion of 1 mg/kg). The efficacy of the
blockade was verified by bolus administration of ACh (40
µg/kg iv) and by the marked increase in heart rate (HR) from
72 ⫾ 3 to 164 ⫾ 9 beats/min (bpm). All data in the second
experiment were collected within 60 min before and 60 min
after atropine administration.
Data Analyses
Sleep/wake states were characterized as either awake (lowamplitude and high-frequency central EEG activity and
high-frequency nuchal EMG activity), NREM sleep (highamplitude and low-frequency central EEG activity and decreased nuchal EMG activity compared with awake), and
REM sleep (low-amplitude and high-frequency central
EEG activity, similar to awake, and reduced nuchal EMG
activity compared with awake and NREM sleep), as previously described for the canine model (16). Arousal was
classified by the changes in EEG and EMG activity at apnea
termination according to American Sleep Disorders Association criteria (1).
Apnea duration was defined as the time from onset of an
apnea to the nadir of the last inspiratory effort during the
apnea. For each apnea, the number of respiratory efforts, the
duration, and the concomitant oxyhemoglobin saturation
were determined. In each animal, apneas with and without
arousal were matched for length by selecting the five longest
apneas without arousal and comparing them to five apneas of
equal duration with arousal. In this way, the duration,
number of respiratory efforts, and degree of oxygen desaturation were matched in apneas that were terminated either
with or without arousal. Ventricular output was calculated as
SV ⫻ HR. There were no differences between left and right
ventricular SV as previously described (21), so data were then
pooled to provide a single value for ventricular SV.
The following definitions and criteria were used in the
measurement of preapnea, end-apnea, and postapnea values
for Psa, Ppa,tm, Pla,tm, and HR: preapnea is the three
respiratory cycles immediately before apnea, end-apnea is
the last two ‘‘inspiratory and expiratory cycles’’ of the apnea,
and postapnea is the first three respiratory cycles after the
restoration of airway patency.
Downloaded from http://jap.physiology.org/ by 10.220.33.6 on September 17, 2016
diameter of the vessel. A minimum of 2 wk was allowed for
local fibrosis to occur (confirmed at autopsy) before data
collection, assuring that all flow measurements occurred
through a constant cross-sectional area. In those animals
with pulmonary artery flow probes, Tygon catheters (0.05 in.
ID, 0.09 in. OD) were also placed in the pulmonary artery and
left atrium to monitor Ppa and left atrial pressure (Pla),
respectively. A balloon catheter was anchored to the inside of
the chest wall at the level of the fourth interspace on the right
side of the chest to estimate pleural pressure (Ppl) as
previously reported (21). The chest was then closed, and
negative intrapleural pressure was established with a chest
tube to completely reinflate the lungs. The chest tube was
removed, and spontaneous ventilation was allowed to resume. All catheters and electrical leads were tunneled subcutaneously and exteriorized between the scapulae and were
protected in the pocket of a jacket. Polysomnographic leads
for measurement of electroencephalographic (EEG) and nuchal electromyographic (EMG) activity were attached with
needle electrodes only during data collection periods. Patency
and sterility of the vascular catheters were maintained by
filling them with a mixture of heparin (1,000 U/ml) and
penicillin G potassium (20,000 U/ml). All catheters were
flushed and had the dead space fluid replaced at a minimum
of every 72 h. The animals were allowed at least 2 wk to
recover from surgery, during which time they were monitored
daily and acclimated to the laboratory environment. All
animals were treated with a broad-spectrum antibiotic (30
mg · kg⫺1 · day⫺1 of trimethoprim/sulfadiazine), beginning at
the time of surgery and continuing for 7–10 days postoperatively. The study was approved by the Johns Hopkins University Animal Use and Care Committee and complied with the
American Physiological Society guidelines.
1085
1086
AROUSAL AND SLEEP STATE IN OSA
Table 1. Duration of apnea, arterial oxygen
desaturation, number of respiratory efforts, and
number of apneas analyzed per dog comparing
obstructive apneas that terminate either
with or without arousal
Duration, s
Oxygen desaturation, %
Number of efforts
Number of apneas per dog
Arousal
No Arousal
P
28.2 ⫾ 2.9
5.9 ⫾ 0.6
6.4 ⫾ 0.9
5.0 ⫾ 0.4
27.5 ⫾ 3.8
5.6 ⫾ 0.6
6.2 ⫾ 0.9
5.0 ⫾ 0
NS
NS
NS
NS
Values are means ⫾ SE; n ⫽ 6 dogs. NS, not significant. P,
differences between no arousal and arousal conditions determined by
paired t-test (two-tailed).
lated as the maximum expiratory value during each period.
The HR was measured from the pulsatile arterial blood
pressure trace by counting the number of systolic peaks over
each period. Summaries of the apnea characteristics and dogs
used for each analysis are presented in Tables 1–3. For each
animal, the values for pre-, end-, and postapnea were averaged over three to five apneas. The averaged data from each
animal were then pooled across all the animals and analyzed
by paired t-test. Differences were considered significant if P ⬍
0.05. Data were analyzed using Crunch 4 (Crunch Software,
Oakland, CA) and are reported as means ⫾ SE.
RESULTS
Effect of Apnea-Induced Arousal
SV. The SV, HR, Psa, and Ppa,tm were examined
during apneas that were terminated either with or
without an arousal during NREM sleep. Respiratory
characteristics of the apneas are presented in Table 1.
Figure 1 shows a trace from one animal during an
apnea with an arousal (A) and without an arousal (B).
In both apneas, Psa, Ppa,tm, and left ventricular SV at
Fig. 1. Computer-generated sample traces from one animal during an obstructive apnea with arousal (A) and an
obstructive apnea without arousal (B) showing electroencephalographic (EEG) and nuchal electromyographic
(EMG) activity, expired CO2, tracheal pressure (Ptr), arterial Hb saturation (SaO2), systemic arterial pressure (Psa),
transmural pulmonary arterial pressure (Ppa,tm), and left ventricular stroke volume (LVSV) during non-rapid-eyemovement (NREM) sleep. Note that, in comparison to apnea without arousal, apnea with arousal is associated with
a greater end- to postapnea increase in heart rate (HR) and in Psa but not in Ppa,tm.
Downloaded from http://jap.physiology.org/ by 10.220.33.6 on September 17, 2016
Before data analyses, the vascular pressure signals were
filtered digitally (Windaq, filter factor 100, Dataq Instruments) to provide a mean pressure. The filtered traces were
then used to assess the mean vascular pressures at various
points throughout the apnea cycle. At preapnea, vascular
pressures were calculated as the maximum expiratory level
averaged for the three respiratory cycles during this period.
At end-apnea and postapnea, vascular pressures were calcu-
AROUSAL AND SLEEP STATE IN OSA
1087
pre- and end-apnea were comparable. In contrast, from
end-apnea to postapnea the Psa and HR increased
more and the SV decreased more in the apnea that was
terminated by arousal compared with the apnea without arousal.
The pooled data (n ⫽ 6) comparing cardiovascular
responses to obstructive apneas with and without
arousal are shown in Fig. 2. From end-apnea to postapnea, HR increased 20.7 ⫾ 2.6 bpm with arousal but
only 4.6 ⫾ 2.1 bpm without arousal (P ⬍ 0.004).
Similarly, ventricular output increased 21.4 ⫾ 2.6%
with arousal compared with 6.0 ⫾ 3.0% without arousal
from end-apnea to postapnea (P ⬍ 0.002). The SV
decreased from end-apnea to postapnea by 4.4 ⫾ 1.7%
with arousal (P ⬍ 0.025) and remained constant without arousal. Thus apneas with arousal produced an
Fig. 3. Psa (n ⫽ 6) (A) and Ppa,tm (n ⫽ 4) (B)
responses (mean ⫾ SE) to obstructive apneas
with arousal (k) and without arousal (j) during
NREM sleep. Differences from end-apnea to postapnea between the arousal and nonarousal condition were determined by paired t-test (twotailed).
increase in HR and ventricular output but a decrease in
SV compared with apneas without arousal.
Psa and Ppa. Pooled data comparing Psa and Ppa,tm
responses to apneas with and without arousal during
NREM are shown in Fig. 3. From end-apnea to postapnea, mean Psa increased 10.9 ⫾ 2.8 mmHg with
arousal compared with 3.3 ⫾ 1.4 mmHg without arousal
(P ⬍ 0.025). In contrast, the presence of arousal did not
affect the increase in the Ppa,tm from end-apnea to
postapnea (Fig. 3). In the three animals in which Pla
was measured, comparable changes in Pla,tm occurred
from preapnea to postapnea between the arousal (0.3 ⫾
0.1 mmHg) and no arousal (0.6 ⫾ 0.3 mmHg) conditions. Thus arousal increased the systemic, but not the
pulmonary, arterial pressure responses to an obstructive apnea.
Downloaded from http://jap.physiology.org/ by 10.220.33.6 on September 17, 2016
Fig. 2. HR (A), ventricular output (VO; B), and stroke volume (SV; C) responses (mean ⫾ SE; n ⫽ 6) to obstructive
apneas with arousal (k) and without arousal (j) during NREM sleep. Pre, 3 respiratory cycles immediately before
apnea; End, last 2 ‘‘inspiratory and expiratory cycles’’ of apnea; Post, first 3 respiratory cycles after restoration of
airway patency; bpm, beats per minute. Differences from end-apnea to postapnea between the arousal and
nonarousal condition were determined by paired t-test (two-tailed).
1088
AROUSAL AND SLEEP STATE IN OSA
Table 2. Duration of apnea, number of respiratory
efforts, and number of obstructive apneas analyzed
per dog before (control) and after parasympathetic
blockade (1 mg/kg atropine)
Duration, s
Number of efforts
Number of apneas per dog
Control
Atropine
P
28.2 ⫾ 2.9
6.4 ⫾ 0.9
5.0 ⫾ 0.4
26.3 ⫾ 2.4
6.8 ⫾ 1.6
3.3 ⫾ 0.3
NS
NS
⬍0.05
Table 3. Duration of apnea, arterial oxygen
desaturation, number of respiratory efforts, and
number of apneas analyzed per dog during obstructive
apneas in NREM and REM sleep
Duration, s
Oxygen desaturation, %
Number of efforts
Number of apneas per dog
NREM
REM
P
25.8 ⫾ 1.1
6.4 ⫾ 0.8
6.9 ⫾ 1.3
5.0 ⫾ 0.5
27.9 ⫾ 3.0
9.0 ⫾ 0.7
6.1 ⫾ 1.0
3.0 ⫾ 0.5
NS
⬍0.05
NS
⬍0.05
Values are means ⫾ SE; n ⫽ 4 dogs. P, differences between
non-rapid-eye-movement (NREM) and rapid-eye-movement (REM)
conditions determined by paired t-test (two-tailed).
Effect of Parasympathetic Blockade
Effect of Sleep State
Atropine administration increased the baseline, or
preapnea, HR (⫹92 ⫾ 7.7 bpm) and Psa (⫹16.8 ⫾ 2.0
mmHg) and decreased the SV (⫺38.5 ⫾ 4.2%). Therefore, the cardiovascular responses to apnea and arousal
were compared between the control and atropine conditions as a percent change from preapnea. The respiratory characteristics of the obstructive apneas are presented in Table 2. The effect of parasympathetic
blockade on HR, Psa, and ventricular output responses
to obstructive apneas with arousal is shown in Fig. 4.
Compared with control, the effect of parasympathetic
blockade was to abolish the HR and Psa increase from
end- to postapnea. During parasympathetic blockade,
SV (Fig. 4) and also ventricular output remained constant between end- and postapnea. In four of the six
animals, the Ppa,tm response to obstructive apnea was
measured. Before atropine administration, the Ppa,tm
was 10.1 ⫾ 0.5 mmHg at preapnea, 11.6 ⫾ 0.8 mmHg at
end-apnea, and 16.2 ⫾ 1.7 mmHg at postapnea. The comparable values after atropine administration were 11.2 ⫾
1.6, 12.9 ⫾ 2.2, and 14.6 ⫾ 2.1 mmHg, respectively, and
these were not different from the control response.
The respiratory characteristics for obstructive apneas with arousal during NREM and REM sleep are
shown in Table 3. Sleep state did not affect the changes
from preapnea to end-apnea and end-apnea to postapnea for HR, SV, ventricular output, or Ppa,tm (Table
4). In contrast, the response of the Psa to an obstructive
apnea with arousal was different between the sleep
states. Figure 5 shows a sample trace from one dog
demonstrating that during REM sleep the Ppa,tm, but
not the Psa, increased from preapnea to end-apnea.
The pooled data in Table 4 show that the increase in
Psa from preapnea to end-apnea was reduced in REM
sleep compared with NREM sleep. Conversely, from
end-apnea to postapnea, Psa increased more in REM
than NREM sleep, such that the absolute Psa at
end-apnea was not different between REM (129.4 ⫾ 2.9
mmHg) and NREM (132.8 ⫾ 3.3 mmHg) sleep.
DISCUSSION
This study investigated the influences of arousal and
sleep state on cardiovascular responses to OSA in a
canine model. We present several new findings from
Fig. 4. HR (A), Psa (B), and SV (C) responses (mean ⫾ SE; n ⫽ 6) to obstructive apneas before (control; k) and after
(j) parasympathetic blockade (1 mg/kg atropine) during NREM sleep. Comparisons of cardiovascular parameters
between control and atropine were performed as a percentage of values at preapnea. Differences from end-apnea to
postapnea between control and atropine conditions were determined by paired t-test (two-tailed).
Downloaded from http://jap.physiology.org/ by 10.220.33.6 on September 17, 2016
Values are means ⫾ SE; n ⫽ 6 dogs. After atropine administration,
we were unable to record arterial Hb saturation from the mucosal
lining of the trachea for reasons previously described (14). P, differences between control and atropine conditions determined by paired
t-test (two-tailed).
AROUSAL AND SLEEP STATE IN OSA
Table 4. Effect of sleep state on changes in heart rate,
stroke volume, ventricular output, transmural
pulmonary arterial pressure, and systemic arterial
pressure from preapnea to end-apnea and from
end-apnea to postapnea (with arousal)
Sleep
State
Change from
Pre- to
End-Apnea
NREM
REM
NREM
REM
⫺3.6 ⫾ 1.9
⫺2.0 ⫾ 2.5
⫺4.2 ⫾ 1.0
⫺2.8 ⫾ 1.2
NS
NREM
REM
Ppa,tm, mmHg NREM
REM
Psa, mmHg
NREM
REM
⫺8.7 ⫾ 3.3
⫺5.7 ⫾ 2.6
1.5 ⫾ 0.7
1.0 ⫾ 0.5
10.3 ⫾ 1.6
5.6 ⫾ 2.3
NS
HR, bpm
SV, %
Ventricular
output, %
P
NS
NS
⬍ 0.025
Change from
End- to
Postapnea
P
23.2 ⫾ 2.8
20.0 ⫾ 1.7
⫺6.8 ⫾ 1.2
⫺7.8 ⫾ 1.7
NS
20.7 ⫾ 2.8
17.1 ⫾ 3.2
4.5 ⫾ 1.0
4.4 ⫾ 1.0
13.9 ⫾ 3.3
19.5 ⫾ 2.7
NS
NS
NS
⫽ 0.05
this work. First, we show that the acute increase in Psa
due to arousal from an obstructive apnea is dependent
on a tachycardia mediated by vagal withdrawal of
parasympathetic tone to the heart (Fig. 4). Second, we
demonstrate that a consequence of the arousal-mediated tachycardia is a decrease in ventricular SV immediately after termination of the obstructive apnea
(Fig. 2). Third, we show that arousal from apnea does
not contribute to any of the increase in Ppa after apnea
termination (Fig. 3). This absence of an effect of arousal
on the pulmonary circulation is in contrast to the
systemic circulation in which arousal is the dominant
factor increasing Psa after apnea (Figs. 1 and 3; see Ref.
14). Finally, we show that, in REM sleep, the increase
in Psa that occurs during an obstructive apnea is
blunted compared with the response in NREM sleep
(Fig. 5 and Table 4). In contrast, the pulmonary circulation exhibited no such effect of sleep state in response to
obstructive apnea. Thus both sleep state and arousal
have differential effects on the systemic and pulmonary
circulations in obstructive apnea. In the discussion that
follows, we examine in more detail the relationship of
arousal and sleep state to changes in the systemic and
pulmonary circulation that occur in OSA.
Systemic Circulation
In this study, we confirmed our previous finding that
arousal from NREM sleep increases HR and contributes to the acute systemic hypertension immediately
after apnea termination (14). These increases in HR
and Psa with arousal were blocked after atropine
administration. These data indicate that the cardiovascular effects of apnea-induced arousal are predominantly due to withdrawal of the parasympathetic tone
to the heart. These data are consistent with those of
Horner et al. (11), who showed that spontaneous arousal
from NREM sleep also increased HR through withdrawal of parasympathetic tone. It is possible that
arousal may also affect sympathetic tone to the systemic vasculature, but our data suggest that such a
response is not necessary to explain the increases in
Psa due to arousal. Thus the effects of arousal from
obstructive apnea on the systemic circulation are predominantly dependent on the heart.
The surge in Psa at arousal from an obstructive
apnea was associated with a decrease in SV (Fig. 2). A
fall in SV in response to obstructive apnea has been
demonstrated in both humans (2, 6, 9, 22, 28) and
animals (21, 30). In this study, we show that the arousal
response is an important factor contributing to the fall
in SV immediately after an apnea. This fall in SV is due
to a tachycardia because the presence of a constant HR
during arousal from obstructive apneas following atropine administration eliminated the fall in SV after
apnea termination (Fig. 4). The increase in HR and
decrease in SV with arousal are consistent with a
transient decrease in the preload of the left ventricle,
although direct measurements of ventricular volume
would be necessary to confirm changes in preload.
Nevertheless, in the absence of arousal, SV progressively decreased over the course of an obstructive
apnea (Fig. 4; preapnea to end-apnea), suggesting that
other mechanisms contribute to the fall in SV preceding arousal. In the accompanying study (21a), we
demonstrate that this fall in SV during an obstructive
apnea is produced by an increase in left ventricular
afterload. Thus, in OSA, two different mechanisms
account for the decrease in SV that occurs before and
after arousal.
In the present study, we report that Psa increases
less during an obstructive apnea in REM sleep than in
NREM sleep (Fig. 5 and Table 4). Given that hypoxic
stimulation of the chemoreceptors is the dominant
factor increasing sympathetic nerve activity during
apnea (10, 12, 17), the attenuated systemic blood
pressure response to apnea during REM sleep suggests
that sleep state can inhibit the chemoreceptor-sympathetic nerve activity reflex arc (8, 13). It is known from
human and animal studies that respiratory responses
to chemoreceptor stimulation are blunted during REM
sleep compared with NREM sleep (4, 29). The present
data suggest that a comparable situation may exist
with respect to the cardiovascular system; namely, the
chemoreceptor stimulus associated with an obstructive
apnea is less effective in raising Psa during REM than
during NREM sleep.
Despite a differential effect of sleep state on the Psa
response to apnea, the maximal blood pressure response at arousal was independent of the previous
sleep state. That is, the maximal Psa reached at
arousal was the same whether or not apneas occurred
during NREM sleep or REM sleep. In contrast, the only
comparable study conducted in humans with OSA
indicated that systemic blood pressure surged to higher
levels in REM sleep compared with NREM sleep (7).
However, it was unclear whether this was due to a more
acute increase in blood pressure at arousal from REM
sleep, which would be consistent with our current data,
Downloaded from http://jap.physiology.org/ by 10.220.33.6 on September 17, 2016
Values are means ⫾ SE; n ⫽ 4 dogs. HR, heart rate; SV, stroke
volume; Ppa,tm, transpulmonary arterial pressure; Psa, systemic
arterial pressure. P, differences between NREM and REM conditions
determined by paired t-test (two-tailed).
1089
1090
AROUSAL AND SLEEP STATE IN OSA
or to the fact that baseline blood pressure was higher in
REM sleep than in NREM sleep before apnea.
Pulmonary Circulation
In the present study, the magnitude of the Ppa
response to an obstructive apnea was independent of
sleep state (Table 4). Such a result builds on our
previous demonstration that during normal, unobstructed breathing the Ppa is the same during NREM
and REM sleep (21). This absence of a sleep state effect
on the pulmonary circulation is in contrast to the effect
on the systemic circulation, as discussed above. The
vascular resistance in the pulmonary circulation, therefore, is much less dependent on the autonomic nervous
system (ANS) than is the vascular resistance in the
systemic circulation. Further support for such a hypothesis is provided in the accompanying study (21a), which
demonstrates that the increase in Ppa during apnea is
Downloaded from http://jap.physiology.org/ by 10.220.33.6 on September 17, 2016
Fig. 5. Computer-generated sample
trace from 1 animal during a single
period of induced airway obstruction in
rapid-eye-movement sleep showing EEG
and EMG activity, expired carbon CO2,
Ptr, SaO2, Psa, Ppa,tm, and LVSV. Note
that, from preapnea to end-apnea, Psa
fell whereas Ppa,tm rose, and that the
EEG signal is superimposed by rapid
eye movements.
dependent on hypoxic pulmonary vasoconstriction and
is independent of the ANS. Thus several lines of
evidence suggest that the pulmonary circulation both
has minimal neural regulation and is unresponsive to
changes in sleep state.
In addition to sleep state, the pulmonary circulation
was also unaffected by the arousal response (Figs. 1
and 3). This finding is in contrast to that regarding the
systemic circulation, in which the tachycardia associated with arousal produced a significant increase in
Psa (14) (Fig. 3). One possible explanation for the
absence of a change in Ppa with arousal may be that
the pressure gradient from the pulmonary artery to the
left atrium was higher in the arousal than the nonarousal condition. However, in the three animals in
which Pla was measured, the pressure changes over
the apnea cycle were the same for the arousal and
nonarousal conditions. Thus the pressure gradient
AROUSAL AND SLEEP STATE IN OSA
We acknowledge Nellcor (Haywood, CA) for time in discussing the
measurement of arterial Hb saturation and the generous loan of the
oximetry equipment.
This study was funded by grants from the National Heart, Lung,
and Blood Institute (HL-51292), the American Heart Association, and
Deutsche Forschungsgemeinschaft (SCH543/1–1).
Address for reprint requests and other correspondence: C. P.
O’Donnell, Rm. 4B61, Johns Hopkins Asthma and Allergy Ctr., 5501
Hopkins Bayview Cir., Baltimore, MD 21224 (E-mail: codonnel
@welch.jhu.edu).
Received 3 March 1999; accepted in final form 8 November 1999.
REFERENCES
1. Atlas Task Force. EEG arousals: scoring rules, and examples. A
preliminary report from the Sleep Disorders Atlas Task Force of
the American Sleep Disorders Association. Sleep 15: 174–184,
1992.
2. Bonsignore, M. R., O. Marrone, S. Romano, and D. Pieri.
Time course of right ventricular stroke volume and output in
obstructive sleep apneas. Am. J. Respir. Crit. Care Med. 149:
155–159, 1994.
3. Coccagna, G., M. Mantovani, F. Brignani, C. Parchi, and E.
Lugaresi. Continuous recording of the pulmonary and systemic
arterial pressure during sleep in syndromes of hypersomnia with
periodic breathing. Bull. Physiopathol. Respir. 8: 1159–1172,
1972.
4. Douglas, N. J., D. P. White, J. V. Weil, C. K. Pickett, and
C. W. Zwillich. Hypercapnic ventilatory response in sleeping
adults. Am. Rev. Respir. Dis. 126: 758–762, 1982.
5. Fishman, A. P. Pulmonary circulation. In: Handbook of Physiology. The Respiratory System. Circulation and Nonrespiratory
Function. Bethesda: Am. Physiol. Soc., 1985, sect. 3, chapt. 3, p.
93–165.
6. Garpestad, E., H. Katayama, J. A. Parker, J. Ringler, J.
Lilly, T. Yasuda, R. H. Moore, H. W. Strauss, and J. W. Weiss.
Stroke volume and cardiac output decrease at termination of
obstructive apneas. J. Appl. Physiol. 73: 1743–1748, 1992.
7. Garpestad, E., J. Ringler, J. A. Parker, S. Remsburg, and
J. W. Weiss. Sleep stage influences the hemodynamic response to
obstructive apneas. Am. J. Respir. Crit. Care Med. 152: 199–203,
1995.
8. Guazzi, M., G. Baccelli, and A. Zanchetti. Reflex chemoceptive regulation of arterial pressure during natural sleep in the
cat. Am. J. Physiol. 214: 969–978, 1968.
9. Guilleminault, C., J. Motta, F. Mihm, and K. Melvin. Obstructive sleep apnea and cardiac index. Chest 89: 331–334, 1986.
10. Hardy, J. C., K. Gray, S. Whisler, and U. Leuenberger.
Sympathetic and blood pressure responses to voluntary apnea
are augmented by hypoxemia. J. Appl. Physiol. 77: 2360–2365,
1994.
11. Horner, R. L., D. Brooks, L. F. Kozar, S. Tse, and E. A.
Phillipson. Immediate effects of arousal from sleep on cardiac
autonomic outflow in the absence of breathing in dogs. J. Appl.
Physiol. 79: 151–162, 1995.
12. Leuenberger, U., E. Jacob, L. Sweer, N. Waradekar, C.
Zwillich, and L. Sinoway. Surges of muscle sympathetic nerve
activity during obstructive apnea are linked to hypoxemia. J.
Appl. Physiol. 79: 581–588, 1995.
13. Mancia, G., G. Baccelli, A. B. Adams, and A. Zanchetti.
Vasomotor regulation during sleep in the cat. Am. J. Physiol. 220:
1086–1093, 1971.
14. O’Donnell, C. P., T. Ayuse, E. D. King, A. R. Schwartz, P. L.
Smith, and J. L. Robotham. Airway obstruction during sleep
increases blood pressure without arousal. J. Appl. Physiol. 80:
773–781, 1996.
15. O’Donnell, C. P., E. D. King, A. R. Schwartz, J. L. Robotham, and P. L. Smith. Relationship between blood pressure
and airway obstruction during sleep in the dog. J. Appl. Physiol.
77: 1819–1828, 1994.
16. O’Donnell, C. P., E. D. King, A. R. Schwartz, P. L. Smith, and
J. L. Robotham. Effect of sleep deprivation on responses to
airway obstruction in the sleeping dog. J. Appl. Physiol. 77:
1811–1818, 1994.
17. O’Donnell, C. P., A. R. Schwartz, P. L. Smith, J. L. Robotham, R. S. Fitzgerald, and M. Shirahata. Reflex stimulation of renal sympathetic nerve activity and blood pressure in
response to apnea. Am. J. Respir. Crit. Care Med. 154: 1763–
1770, 1996.
18. Remmers, J. E., W. U. deGroot, E. K. Sauerland, and A. M.
Anch. Interaction of neural and mechanical factors in obstructive sleep apnea. In: Central Nervous Mechanisms in Breathing,
Downloaded from http://jap.physiology.org/ by 10.220.33.6 on September 17, 2016
across the pulmonary circulation was the same with
and without arousal. A second possible explanation
may be that, as argued above, the systemic circulation
is much more responsive to the ANS than is the
pulmonary circulation. Therefore, if arousal increased
vascular resistance through the ANS, then Psa would
be expected to increase more with arousal than does the
Ppa. However, we have shown that it is the tachycardia
of arousal rather than a vascular response that can
account for the increase in systemic vascular pressure.
As such, this tachycardia from arousal should also
affect the pulmonary circulation. We did not observe
any effect of the tachycardia from arousal on the
pulmonary circulation, and we can only speculate that
differences in compliance between the two circulations
may play a role. Consider that the tachycardia associated with arousal is very transient and the vascular
resistance of pulmonary circulation is particularly low
(5, 25). Thus the very brief infusion of blood into the
pulmonary circulation during the tachycardia may be
accommodated without any measurable increase in
Ppa.
The cardiovascular effects of arousal and sleep state
that we have demonstrated in our canine model may
have implications for OSA patients; namely, that the
ANS has distinct effects on the heart and systemic
circulation but not on the pulmonary circulation. Our
results in animals with normal hearts demonstrate
that obstructive apnea produces a neurally mediated
stress that is specific to the left side of the circulation.
This stress may be exacerbated in OSA patients with
underlying left-sided heart disease such that arousal,
for example, may cause significant reductions in SV at
end-apnea and expose the heart or brain to increased
risk of ischemia. Because such studies are extremely
difficult to perform in OSA patients, particularly if
significant comorbidity is present, the need exists to
develop more sophisticated animal models of OSA that
incorporate heart failure, coronary ischemia, hypertension, and other cardiovascular diseases.
In summary, we have demonstrated specific differences in the cardiovascular responses of the systemic
and pulmonary circulations to arousal and sleep state
in a canine model of OSA. Arousal from obstructive
apnea caused an increase in Psa but not in Ppa. The
effect of arousal on the systemic circulation was dependent on a tachycardia, and this tachycardia also contributed to a decrease in ventricular SV immediately after
the apnea. The systemic arterial blood pressure response during an obstructive apnea was attenuated
during REM sleep compared with during NREM sleep.
In contrast, the pulmonary circulation, which is much
less responsive to the ANS, showed similar blood
pressure responses during NREM and REM sleep.
1091
1092
AROUSAL AND SLEEP STATE IN OSA
23. Shepard, J. W., Jr. Gas exchange and hemodynamics during
sleep. Med. Clin. North Am. 69: 1243–1264, 1985.
24. Shepard, J. W. J. Cardiopulmonary consequences of obstructive
sleep apnea. Mayo Clin. Proc. 65: 1250–1259, 1990.
25. Silove, E. D., and R. F. Grover. Effects of alpha adrenergic
blockade and tissue catecholamine depletion on pulmonary vascular responses to hypoxia. J. Clin. Invest. 47: 274–285, 1968.
26. Stoohs, R., and C. Guilleminault. Cardiovascular changes
associated with obstructive sleep apnea syndrome. J. Appl.
Physiol. 72: 583–589, 1992.
27. Tilkian, A. G., C. Guilleminault, J. S. Schroeder, K. L.
Lehrman, F. B. Simmons, and W. C. Dement. Hemodynamics
in sleep-induced apnea: studies during wakefulness and sleep.
Ann. Intern. Med. 85: 714–719, 1976.
28. Tolle, F. A., W. V. Judy, P.-L. Yu, and O. N. Markand. Reduced
stroke volume related to pleural pressure in obstructive sleep
apnea. J. Appl. Physiol. 55: 1718–1724, 1983.
29. White, D. P., N. J. Douglas, C. K. Pickett, C. W. Zwillich, and
J. V. Weil. Sleep deprivation and the control of ventilation. Am.
Rev. Respir. Dis. 128: 984–986, 1983.
30. Zinkovska, S., and D. A. Kirby. Intracerebroventricular propranolol prevented vascular resistance increases on arousal from
sleep apnea. J. Appl. Physiol. 82: 1637–1643, 1997.
Downloaded from http://jap.physiology.org/ by 10.220.33.6 on September 17, 2016
edited by C. V. Euler and H. Lagercrantz. Oxford: Pergamon,
1979, p. 473–482.
19. Ringler, J., R. C. Basner, R. Shannon, R. Schwartzstein, H.
Manning, S. E. Weinberger, and J. W. Weiss. Hypoxemia
alone does not explain blood pressure elevations after obstructive
apneas. J. Appl. Physiol. 69: 2143–2148, 1990.
20. Ringler, J., E. Garpestad, R. C. Basner, and J. W. Weiss.
Systemic blood pressure elevation after airway occlusion during
NREM sleep. Am. J. Respir. Crit. Care Med. 150: 1062–1066,
1994.
21. Schneider, H., C. D. Schaub, K. A. Andreoni, A. R. Schwartz,
P. L. Smith, J. L. Robotham, and C. P. O’Donnell. Systemic
and pulmonary hemodynamic responses to normal and obstructed breathing during sleep. J. Appl. Physiol. 83: 1671–1680,
1997.
21a.Schneider, H., C. D. Schaub, K. A. Andreoni, A. R. Schwartz,
P. L. Smith, J. L. Robotham, and C. P. O’Donnell. Neural and
local effects of hypoxia on cardiovascular responses to obstructive apnea. J. Appl. Physiol. 88: 1093–1102, 2000.
22. Schroeder, J. S., J. Motta, and C. Guilleminault. Hemodynamic studies in sleep apnea. In: Sleep Apnea Syndromes, edited
by C. Guilleminault and E. Lugaresi. New York: Liss, 1978, p.
177–196.