LABORATORY INVESTIGATION

PHYSIOLOGY

Mechanisms mediating the heart rate response

to hypoxemia
HITOSHI KATO, M.D., ANIL S. MENON, PH.D., AND ARTHUR S. SLUTSKY, M.D.

ABSTRACT We studied the effect ofphasic pulmonary afferent information on heart rate (HR) during
a progressive reduction in oxygen saturation (SaO2). The Hering-Breuer reflex was evaluated with the
use of the ratio of apnea duration after lung inflation to the preceding expiratory time (dT). Phasic
afferent activity was stopped in anesthetized, paralyzed dogs by constant-flow ventilation (CFV), a
technique that removes cyclic changes in lung volume. During normocapnic (Paco2 = 36.4 + 1.1
mm Hg) spontaneous breathing, there was a wide variability in HR response, with a mean AHR/ASaO2
( ± SE) of 0.62 0.27 beats/min/% (values greater than 0 indicate a tachycardiac response). There
Downloaded from http://circ.ahajournals.org/ by guest on June 8, 2018

was a good correlation between AHRIASaO2 and dT (r = .79). Mean AHR/ASaO2 for the combined
normocapnic and hypercapnic studies during CFV was lower ( - 1.32 ± 0.19 bpm/%) than that during
spontaneous breathing (0.23 + 0.19, p < .0001). We suggest that the HR response to hypoxemia is
strongly related to the strength of the Hering-Breuer reflex, which may explain the large interdog
variability in HR responses.
Circulation 77, No. 2, 407-414, 1988.
ARTERIAL HYPOXEMIA is a common accompani-
ment of many disorders of the cardiorespiratory sys-
tem. Although the heart rate response to acute hypo-
xemia has been studied for many years, the
mechanisms mediating this response are incompletely
understood. ' Hypoxemia in conscious humans and in
dogs generally results in a tachycardia. Until the work
of Bernthal et al.2 it was believed that this cardioac-
celerator response was due to carotid body chemore-
ceptor stimulation. Bernthal showed that in most arti-
ficially ventilated dogs, hypoxic stimulation of the
carotid bodies produced a decrease in heart rate, dem-
onstrating that the carotid bodies were not responsible
for the tachycardia. These results were further corrob-
orated by a number of investigators,3-5 and it is now
generally accepted that in artificially ventilated dogs,
stimulation of the isolated carotid body chemorecep-
tors results in bradycardia.
By contrast, spontaneously breathing dogs show a
wide and varied response to hypoxia of the isolated
From the Department of Medicine, Mount Sinai Hospital Research
Institute, University of Toronto, Toronto, Ontario, Canada.
Supported in part by grant AN 743 from the Heart and Stroke Foun-
dation of Canada.
Address for correspondence: A. S. Slutsky, M.D., Department of
Medicine, Mount Sinai Hospital, 600 University Ave., #656A,
Toronto, Ontario, Canada M5G 1X5.
Received Aug. 31, 1987; accepted Oct. 22, 1987.
Dr. A. S. Menon is a fellow of the Medical Research Council of
Canada. Dr. A. S. Slutsky is an MRC (Canada) Scientist.
Presented in abstract form at the 1986 meeting of the Federation of
American Societies for Experimental Biology (Fed Proc 45: 298, 1986).
Vol. 77, No. 2, February 1988
carotid body, with studies describing tachycardia,
bradycardia, or no change in heart rate.2 3 Daly and
Scott6 suggested that respiratory alkalosis secondary to
increased ventilation was responsible for the varied
response and masked the bradycardiac response of the
carotid body. Prevention of hypocapnia significantly
diminishes the tachycardia but a large variability in
heart rate response among animals still exists. Further,
Daly and Scott3' 6 showed that when hypoxic stimu-
lation of the carotid bodies was performed after pul-
monary denervation, the cardioaccelerator response
was reduced or converted to a bradycardia. These data
support the hypothesis that the heart rate response to
hypoxia is mediated by a complex interaction between
carotid body stimulation which tends to cause brady-
cardia, and increased ventilation, which tends to cause
tachycardia.
The purpose of the present study was to examine the
effect of pulmonary reflexes on the heart rate response
to hypoxemia. Earlier studies had eliminated pulmo-
nary reflexes either by denervating the lungs or by use
of artificial ventilation with constant tidal vol-
ume.3 5' 7, 8 The former model suffers from its inva-
siveness; the latter fails to abolish phasic afferent infor-
mation but maintains it at a constant level. In the present
study, we attempted to overcome these problems by
using two approaches: (1) airway anesthesia and (2) the
technique of constant-flow ventilation (CFV),9 by
which normal blood gases can be maintained in the
407
 KATO et al.
absence of cyclic lung stretch. Using these techniques
we examined the mechanisms responsible for the large
variability in heart rate response to hypoxemia among
spontaneously breathing animals. Our hypothesis was
that if phasic pulmonary afferent information is impor-
tant for the tachycardiac response, then the animals
with greater Hering-Breuer reflexes will exhibit greater
tachycardiac responses.
Methods
Experimental preparation and protocol.Twenty-five dogs
of mixed breed weighing 14 to 36 kg were premedicated with
fentanyl and droperidol and anesthesia was induced with an
intravenous loading dose of a 0.8 ml/kg chloralose-urethane
solution (2.5 g chloralose plus 25 g urethane in 100 ml saline).
Anesthesia was sustained with supplementary doses of approx-
imately 0.3 ml/kg/hr. Pedal and corneal reflexes were examined
frequently to monitor the adequacy of anesthesia.
Femoral artery and venous catheters were introduced to record
arterial blood pressure (Gould-Statham PD 23), to withdraw
Downloaded from http://circ.ahajournals.org/ by guest on June 8, 2018
blood for blood gas analyses (Corning model 168), and to
administer drugs and fluid. Beat-by-beat heart rate was obtained
from the pulse interval of the at terial pressure waveform (Gould
Biotach amplifier). The animals were intubated via a midcer-
vical tracheostomy and an esophageal balloon was used to
measure esophageal pressure (Pes) (Validyne MP 45-30). Oxy-
gen saturation (Sa02) was monitored with a Biox H oximeter
placed on the tongue. Rectal telmperature (YSI Instruments) was
maintained at 37.5 ± 10 C by a heated surgical table.
Series 1. In 19 spontaneously breathing dogs, a tracheost-
omy tube was connected via a T piece to a circuit through which
a constant flow of a gas mixture containing 02, N2, and CO2
was passed. Progressive hypoxia was induced over an 8 to 10
min period by continuously decreasing the 02 and increasing
the N2 flow with a Matheson flow controller until a Sa02 of
approximately 60% was reached. End-tidal CO2 (Sensor-
Medics LB2) was held constant by adding carbon dioxide to
the inspired gas. Arterial blood gases were drawn anaerobically
after every 5% change in Sa02. Results of experiments in which
the Paco2 varied by more than ± 2 mm Hg within the run were
discarded.
In nine dogs, the Hering-Breuer inflation reflex was measured
as the duration of apnea after airway occlusion at peak inspi-
ration while the animals were normocapnic and hyperoxic. As
a measure of the Hering-Breuer reflex, during steady-state con-
ditions, we used the ratio (dT) of apnea duration to the expiratory
duration (Te) of the preceding breath. This method is similar to
that used by Widdicombel0 and Angell-James et al." In these
experiments a pneumotachograph (Fleisch No. 1; Validyne DP45
transducer) was used to measure tidal volume.
Fourteen dogs were then prepared for CFV. The system has
been described elsewhere,9' 12 and only a brief description will
be provided here. It comprises a pair of polyethylene supporting
tubes (2.67 mm inside diameter) cradling the carina and extend-
ing 0.5 cm into the mainstem bronchi. The CFV delivery cath-
eters (1.67 mm ID) are placed concentrically within the sup-
porting tubes and usually located at a depth of 3.5 cm distal to
the carina, one in each lung. The tracheostomy tube is then
inserted.
A gas mixture of 02, N2, and CO2 was insufflated into the
delivery catheters at a constant flow to maintain hyperoxic
normocapnia (usually about 20 liters/min in each catheter). The
dogs were then paralyzed by slow infusion of metocurine iodide
(loading dose 0.2 mg/kg) and paralysis was maintained with 0.2
408
mg/kg/hr. Anesthesia was maintained with the dose and schedule
determined during the spontaneous breathing runs.
Progressive hypoxia down to a Sa02 - 60% was induced by
varying the composition of the insufflated gases (02, N2, and
C02) while maintaining the total flow of the three gases constant.
This ensured a constant lung volume, as monitored by Pes. Due
to the leftward shift in the CO2 dissociation curve induced by
hypoxia, CO2 flow was increased to maintain a constant Paco2.
The heart rate responses to progressive hypoxia during spon-
taneous breathing and CFV were studied during normocapnia
(Paco2 37 mm Hg) and hypercapnia (Paco2 - 50 mm Hg).
The protocols for the hypercapnic and normocapnic experiments
were the same.
In four of 14 dogs, during CFV we injected atropine sulphate
(0.1 mg/kg) intravenously to block efferent parasympathetic
activity. The progressive hypoxic run under normocapnic con-
ditions was then repeated. After control conditions were rein-
stated in each dog, bilateral vagotomy was performed and a
progressive normocapnic hypoxia experiment was repeated.
Series 2. In six additional dogs, we set out to confirm the
specific role of pulmonary receptors on the heart rate response
to hypoxemia by selectively decreasing pulmonary afferent
activity using airway anesthesia.'3 14 Since airway anesthesia
alters the baseline tidal volume,'4 in these experiments we
evaluated the strength of the Hering-Breuer reflex by inflating
the lungs at end-expiration with a volume of 500 ml under
normocapnic, hyperoxic conditions. We calculated the ratio of
apnea duration to Te averaged over the preceding five breaths
(dT500). In these experiments, after a normocapnic, hypoxic run
according to the protocol described previously, 5% bupivacaine
aerosol was generated by the technique of Cross et al. 14 After
30 to 50 inflations with bupivacaine, progressive normocapnic
hypoxia was induced. The Hering-Breuer reflex (dTs5) was
checked before and after this run. We took the average of dTs5
before and after this run as the representative dTH5,. We then
repeated the administration of bupivacaine to depress the Her-
ing-Breuer reflex further and the same protocol was carried out.
About 1 hr after the bupivacaine experiments, a repeat control
run was conducted after ensuring that dTs5o had returned to
near-control values. After these spontaneous breathing runs the
dogs were prepared for CFV and a normocapnic, hypoxic run
was performed as described earlier.
Data analysis. In 10 dogs we compared the measured SaO2
with the SaO2 calculated from the blood gases with the equation
of Reeves et al. '5 Since the oximeter slightly underestimated the
calculated SaO2, the measured saturations were corrected with
use of the regression equations obtained from the measured vs
calculated SaO2.
The recorded data (Gould ES 1000) were analyzed with use of
a digitizer (Summagraphics Super-grid) connected to a com-
puter (IBM-PCXT). The heart rate was measured for every 1%
to 1.5% decrease in saturation and averaged over each breath in
the cases in which there was respiratory sinus arrhythmia. In the
experiments in which the tidal volume was measured, tidal
volume at the same SaO2 levels was also calculated.
Since the duration of apnea after airway occlusion can depend
on the animal's ventilatory response to hypercapnia, 6 we cal-
culated the change in baseline ventilation (AVE) divided by the
change in Paco2 under hyperoxic conditions at the two levels of
Paco2 studied. We used this value (AVE/APacO2) as an index of
the animal's sensitivity to CO2.
The data were then averaged over 5% SaO2 intervals and the
results were plotted as heart rate vs SaO2. The responses were
described as slopes (AHR/ASaO2) calculated with the use of
linear regression analysis. All values are summarized as mean
+ SE. The effect of Pco2 and mode of ventilation were examined
with a 2 x 2 factorial design. A two-way analysis of variance
CIRCULATION

(ANOVA) with repeated measures was applied. Two-sided
paired t tests, applying the Bonferroni correction, were used for
specific comparisons of AHR/ASaO2 between conditions if
ANOVA indicated a significant difference (p < .05).
Results
An example of the raw data from one experiment for
two ranges of SaO2 are presented in figure 1. During
spontaneous breathing, this dog showed a progressive
increase in heart rate as SaO2 was reduced, whereas
during CFV, the same decrease in SaO2 produced a
modest decrease in heart rate. The moving time-
averaged phrenic neurogram was obtained in this prep-
aration and was taken as an index of central respiratory
drive. During CFV, coincident with each phrenic neural
burst, the dog exhibited increases in heart rate (respi-
ratory sinus arrhythmia) despite a lack of change in lung
volume (constant Pes).
Figure 2 depicts data obtained from one dog that
Downloaded from http://circ.ahajournals.org/ by guest on June 8, 2018
was qualitatively representative of the patterns
observed in 60% of the dogs studied. Individual
results for all dogs are shown in figure 3 and data for
the 14 dogs of series 1 in which both ventilatory
methods at both Paco2s were studied are summarized
in table 1. Values of AHR/ASaO2 greater than zero
indicate a tachycardiac heart rate response to hypox-
emia. AHR/ASaO2 during normocapnic spontaneous
breathing was quite variable among dogs and in gen-
eral was greater than that during hypercapnia. During
CFV, 13 of the 14 dogs exhibited a bradycardiac
response to progressive hypoxemia (figure 3).
ANOVA indicated a significant effect of mode of
ventilation (CFV vs spontaneous breathing) on the
mean AHR/ASaO2 averaged over the normocapnic
and hypercapnic experiments (CFV - 1.32 ± 0.19
beats/min/%, spontaneous breathing 0.23 ± 0.19
beats/min/%; F(1,13) = 48.8, p < .0001). Similarly,
the level of CO2 (normocapnic vs hypercapnic) aver-
aged over both modes of ventilation had a significant
effect on AHR/ASaO2 (normocapnia -0.32 ± 0.19
beats/min/%, hypercapnia - 0.77 ± 0.15 beats/
min/%; F (1,13) = 7.12, p = .02). The heart rate
response during spontaneous breathing was signifi-
cantly greater under normocapnic conditions than
under hypercapnic conditions (p = .012). During
CFV, we could detect no significant difference
between normocapnic and hypercapnic conditions
(p = .66).
The heart rate response during normocapnic spon-
taneous breathing correlated well with dT ( r = .79,
p < .01). Since the net pulmonary afferent activity
depends on lung volume and the strength of Hering-
Breuer reflex, we plotted the heart rate response against
Vol. 77, No. 2, February 1988
LABORATORY INVESTIGATION-PHYSIOLOGY
the product of dT and the change in tidal volume
expressed as a percentage of the normocapnic control
value (figure 4). There was a good correlation (r = . 89,
p < .01) between the two variables. While the baseline
tidal volume under hypercapnic conditions (24.0 ±
2.7 ml/kg) was greater than that under normocapnic
conditions (15.3 ± 1.6 ml/kg); p < .01), tidal volume
at the end of the hypoxic run (SaO2 = 60%) was not
different under these two conditions. Among animals
there was no relationship between dT and AVEJAPaCO2
(r=.37, p=.42).
Figure 5 depicts the data for a single study during
CFV in which atropine was used to block parasympa-
thetic efferent activity, followed by bilateral vagotomy.
These result are representative of those in all four dogs.
These interventions yielded an abolition of the brady-
cardiac response that normally accompanied progres-
sive hypoxia during CFV.
Heart rate responses (AHRIASaO2) for the experi-
ments after airway anesthesia (series 2) are presented
in figure 6. Mean Paco2 for the control and two post-
bupivacaine studies (mean = 38 mm Hg) were within
± 1 mm Hg. AHR/ASaO2 during CFV (mean Paco2 =
37 mm Hg) were similar to values obtained during
spontaneous breathing at the lowest dT5so (figure 6).
There were no significant differences in the ventilatory
responses observed during control experiments and those
in experiments conducted after airway anesthesia.
Discussion
In the present study we investigated mechanisms
controlling the cardioaccelerator response to hypoxe-
mia, with a specific focus on the contribution of phasic
pulmonary afferent feedback. Our main findings are
that (1) complete elimination of cyclic lung stretch
during hypoxemia results in a bradycardiac response,
(2) the heart rate response to hypoxemia is greater
during normocapnia than hypercapnia during sponta-
neous breathing, but not during CFV, and (3) the wide
variability in heart rate response among normocapnic
spontaneously breathing animals correlates with the
magnitude of pulmonary afferent feedback (as quan-
tified by dT or dT x % tidal volume).
These studies are unique in that they employed a
nonconventional method of ventilation - CFV9 - a
technique that can maintain normal blood gases and yet
completely remove the phasic pulmonary afferent
activity that accompanies tidal ventilation. To accom-
plish this, we used an anesthetized, paralyzed canine
preparation. Although the paralytic agent, metocurine
iodide, has relatively minor effects on heart rate, the
mean baseline heart rate during CFV for some dogs was
409
 O
KATO et al.

A SPONTANEOUS BREATHING
NORMOXIA HYPOXIA
1-4-10 sec -*-
t 200
Idhua h a-d-JL A AAh-ALA &L 1
-
100 +
A

W-w~w-v- T-1q
Part [mmHg]
0 SaO2
-ir
Pes
10cm H20
1
T
if ,zr8 SaO2 [0%]

F11
J~~SQ
Downloaded from http://circ.ahajournals.org/ by guest on June 8, 2018

iK
HR [bpm]

v o0 50 +

B CONSTANT FLOW VENTILATION

NORMOXIA HYPOXIA
14-10 sec-*-1
A ---~~~~
200

Part [mmHg] O1
SaO

Pes Pes1
l0cm H20
T MTA IA A A
4 250
SaO2 [%/

111 SaO2
r.
1, ." 1,
HR [bpm] L,
'I-T"Cl

0 50

FIGURE 1. Experimental record showing the heart rate response to progressive hypoxemia in one dog. Heart rate (HR), arterial
blood pressure (Part), SaO2, and Pes, during normoxia (left) and hypoxia (right) in one animal breathing spontaneously (A) and
one ventilated via CFV (B). The moving time-averaged (MTA) phrenic neurogram is shown for the CFV run. The mean Paco2
during spontaneous breathing was 38 mm Hg and during CFV it was 37 mm Hg.
410 CIRCULATION
 LABORATORY INVESTIGATION-PHYSIOLOGY
170 1c3dF D O SB PaCO2 = 38mmHg
0 0 SB
0 PaCO2 = 50mmHg
v CFV: PaCO2 = 38mmHg
150 0b,WD1cl JO OCFV: PaCO2 = 5OmmHg

C000
0 ma ° f
D

-E 130- 0C
091 l1

1 000 0o%$ 00 0 UCO FIGURE 2. Heart rate vs SaO2 in one dog. SB =

W
< 90
v v
0n spontaneous breathing.
w v

701 )0 0cxxico
V7

Difrti F -10
_
1 l
60 70 80 90 100
SaO2 [%]

greater than that observed during spontaneous breath- below this SaO2 yields conclusions similar to those
Downloaded from http://circ.ahajournals.org/ by guest on June 8, 2018

ing. However, we do not think this factor affects our obtained by considering the response over the 60% to
major conclusions. When we examined a subgroup 100% range of SaO2.
(n = 7) of dogs that had matched baseline heart rates As a measure of the strength of the Hering-Breuer
(CFV 1 18 ± 7.3 beats/min, spontaneous breathing 1 17 reflex, we used the duration of apnea after airway
± 7.5 beats/min), the results under normocapnic con- occlusion. We were concerned that this apnea duration
ditions (CFV AHR/ASaO2 = - 1.29 ± 0.28 beats/ might be influenced by chemical factors, in addition to
min/%, spontaneous breathing AHR/ASaO2 = 0.47 + pulmonary reflexes, and hence in series 1 we measured
0.34 beats/min/%; p < .005) were similar to those the duration of apnea at end-inspiration rather than at
obtained in all 14 dogs in series 1. Furthermore, in greater lung volumes. Younes et all. 16 have shown that
virtually all of the experimental runs in which baseline the apneic period after airway occlusion is not influ-
heart rates were higher on CFV, the heart rate vs SaO2 enced by chemical factors when the inflation volumes
plots during CFV and spontaneous breathing crossed are as small as those we used. Furthermore, they
(figure 7). If one now views this crossover heart rate showed that the major nonreflex factor affecting dura-
as the baseline, then analysis of the heart rate responses tion of apnea was the animal's ventilatory response to

2
3
0
1.5
0
0

&5
E
)-
n
0 0 0
U')

m 0

-2-

-.5-

00
1

100 300 500 700 900 1100 1300

°/ VT x dT [%]
-31
SB(n 19) CFV(n- 14) CFV(n- 14) SB(n 17)
FIGURE 4. Relationship between the heart rate response and the
NORMOCAPNIA HYPERCAPNIA product of % tidal volume (% VT) x dT during normocapnic sponta-
FIGURE 3. Results of heart rate response to progressive hypoxemia neous breathing. % VT is the percentage change in VT derived from the
during spontaneous breathing (SB) and during CFV in all dogs of series quotient of the average tidal volume at the end of a run divided by
1. AHR/ASaO2 was calculated by linear regression analysis. Positive baseline tidal volume. The correlation coefficient by linear regression
values of AHRIASaO2 denote an increase in heart rate with hypoxemia. analysis is .89 (p < .01).
Vol. 77, No. 2, February 1988 411
 KATO et al.

TABLE 1
Summary of results in the 14 dogs of series 1 in which all conditions were studied

A1HRIA1SaO2
Paco2 (mm Hg) bHR (bpm) bMAP (mm Hg) MAP (mm Hg) (bpm/%)
Spontaneous breathing
Normocapnia 36.4-+ 1.1 114.1 + 6.7 119.5 +-4.3 16.9+ 1.6 0.62 + 0.27
Hypercapnia 50.4±0.6 116.5 ±4.4 119.5+5.1 13.3± 1.5 -0.16±0.19
CFV
Normocapnia 37.2 + 1.1 135.3 + 8.0 120.3 + 5.2 17.1+ 2.4 - 1.26 + 0.26
Hypercapnia 52.9+ 1.1 123.1 +5.6 117.4+5.1 15.2±4.2 - 1.39+0.16
Values for AHRIAtSaO2 were calculated by linear regression analysis as described in the text. Positive values for AHR/ASa0,
indicate an increase in heart rate with hypoxemia. All values are the mean + SE. There were no significant differences in either
bMAP or MAP in the comparable groups. bHR during normocapnic spontaneous breathing was marginally less than that during
normnocapnic CFV (p = .06).
bHR = baseline heart rate; bMAP = baseline mean arterial pressure: MAP = change in mean arterial pressure during
hypoxemia.
Ap < .0012; Bp = .66.
Downloaded from http://circ.ahajournals.org/ by guest on June 8, 2018

Pco2. We found no correlation between dT and
AVE/APaco2, indicating that chemical factors were
unlikely to have had a major effect on apnea duration.
Our results are in accord with those of Daly and
Scott3 and Kontos et al.7 and support Daly and Scott's
hypothesis that pulmonary reflexes play an important
role in the tachycardiac response to hypoxia.3 6 To
minimize phasic pulmonary afferent feedback, previ-
ous studies have used either very invasive preparations
such as pulmonary denervation, which requires an
open-chest preparation with extensive perihilar sur-
gery, or have resorted to conventional ventilation with
a fixed tidal volume.3 5 7, 8 The latter preparation
does not eliminate phasic afferent feedback from the
170 -
lungs. Using CFV we were able to ventilate our animals
at a constant lung volume just above functional residual
capacity.17 Hence, we can assume that we maintained
tonic stretch receptor activity relatively low and near-
baseline levels. To ensure that the mechanism causing
the tachycardia was related to pulmonary rather than
nonpulmonary factors (e.g., feedback from muscle
afferents,18 19 oscillations in blood gases),20 we used
airway anesthesia to achieve a graded reduction in the
strength of the Hering-Breuer reflex (figure 6). Since
airway anesthesia by this technique has no systemic
effects and no direct cardiac effects,14' 21, 22 the sim-
ilarity in results during CFV and at the lowest value of
dT500 (figure 6) provides strong evidence that the major
3-

150 -
2-
E
n 1- a
E
Cl. 130-
Q 0-
Cl)
w
A
< 110 -
ur
1-
A AA
I

-2 - A
W 90-
3- / T

C1 CONTROL (PaCO2 =r
CFV 5 10 15 20
70 - 38
3mHg)
o ATROPiNE INJ (Pa 38m2 m38mmHg) dT500 [ -
v BIL VAGOTOMY (F3aC02 4mmHg)
50 -
FIGURE 6. Relationship between heart rate responses (AHR/IASaO2)
60 70 80 90 100 and the strength of Hering-Breuer reflex after a 500 ml inflation (dT )0.
Each symbol represents a different dog. The control values (greatest
SaO2 [%]
dTjoo) for AHR/ASaO2 are average values before and after recovery from
FIGURE 5. Heart rate responses (AHR/ASaO2) to hypoxemia under airway anesthesia. The values for the heart rate response during CFV
normocapnic conditions in one dog during CFV. Heart rate was averaged are also presented and are similar to those obtained at the lowest dT5so.
over 5% SaO2 intervals. Both atropine injection and bilateral vagotomy Note that for CFV, two different symbols overlap (filled square and filled
virtually abolished the heart rate response to decre ases in Sa02. circle).
412 CIRCULATION

200 -
w 160-
<
W
140
120 - results are in accord with those of Kontos and Lower,27
100
60 70 80 90 100
SaO2 0/]
FIGURE 7. Heart rate response under normocapnic conditions in one
dog with different baseline heart rates during CFV and spontaneous
breathing (SB). Heart rate was averaged over 5% SaO2 intervals.
mechanism by which CFV altered the heart rate
Downloaded from http://circ.ahajournals.org/ by guest on June 8, 2018
response was by removal of cyclic pulmonary afferent
feedback.
Our results, however, are different from those of
Krasney,8 whose mechanically ventilated dogs exhib-
ited tachycardia in response to hypoxia. Krasney attrib-
uted part of the tachycardia to inhibition of vagal tone.
However, as pointed out by Kontos et al.'23 in Kras-
ney's experiments the dogs were anesthetized with
morphine and chloralose, which usually results in a
preparation with high cardiac parasympathetic activity.
Furthermore, morphine has been shown to depress the
peripheral chemoreceptor response to hypoxemia in
cats.24 Since there is strong evidence that the carotid
body mediates the bradycardiac response to hypox-
emia,1-5 the use of morphine as the anesthetic agent
might abolish or attenuate the bradycardia.
In contrast to CFV, during spontaneous breathing,
we found that the heart rate response to hypercapnic
hypoxia was lower than that observed during normo-
capnic hypoxia (table 1; figures 2 and 3). The chemical
stimuli (i.e., arterial blood gases) during CFV and
spontaneous breathing were virtually the same, and
hence one possible explanation for this influence of
CO2 on heart rate is the effect of carbon dioxide on
slowly adapting receptors, resulting in a decrease in
their activity during spontaneous breathing.25 Since
phasic slowly adapting receptor activity does not play
a role in the heart rate response during CFV, any
modulation of slowly adapting receptor activity by CO2
would also not have an effect. Another possibility is
habituation of slowly adapting receptor activity during
the hypercapnic runs, since mean lung volume was
probably higher due to an increased baseline tidal vol-
ume. We therefore suggest that the effect of CO2 on the
heart rate response during hypoxia is mediated by a
Vol. 77, No. 2, February 1988
LABORATORY INVESTIGATION-PHYSIOLOGY
pulmonary-related reflex and not via central effects or
cathecholamines since CO2 does not affect the heart
rate response during CFV.
Our data also support the view that the heart rate
response to hypoxia is not mediated primarily via sym-
pathetic nerves, via a direct action on the heart, or via
humoral factors, since these effects should be similar
during both spontaneous breathing and CFV. These
who found that, in anesthetized, spontaneously breath-
ing dogs, catecholamine release contributed only 25%
to 35% to the increase in heart rate in response to
hypoxia. In our study, in four dogs we found that
parasympathetic blockade completely abolished the
heart rate response to hypoxia, suggesting that sym-
pathetic efferents play an insignificant role in modu-
lating the heart rate response. Vagotomy yielded results
similar to those observed after parasympathetic block-
ade with atropine.
During spontaneous breathing, progressive hypoxia
yielded a variable heart rate response among dogs.
There was an excellent correlation between the nor-
mocapnic heart rate response and the product of dT and
the percentage change in tidal volume. Although the
Hering-Breuer reflex has been considered to be medi-
ated mainly by slowly adapting receptor activity,28 we
have not conclusively ruled out the contribution of
other receptors, such as the rapidly adapting receptors
(cf. irritant receptors) in mediating the heart rate
response. It is possible that the activity of rapidly
adapting receptors correlates with slowly adapting
receptor activity and hence dT may be an indirect
measure of the former. Since rapidly adapting receptors
are stimulated by rapid changes in lung volume,29 they
would be quiescent during CFV, and this could lead to
the observed bradycardiac response.
Although not the focus of this study, our data clearly
indicate that respiratory sinus arrhythmia can occur
without concomitant changes in lung volume or pleural
pressure (see figure 1). Similar results have been
observed previously30; however, prior studies have
often used an invasive, cross-circulation model. We
suggest that CFV may prove to be ideal for studying
central mechanisms causing respiratory sinus arrhyth-
mia, since it provides a stable, relatively noninvasive
model.
Are the results of the present study applicable to
hypoxemic humans? The Hering-Breuer reflex is much
less sensitive in humans than in dogs,10 and yet the
heart rate response in certain human disease states is
very similar to that observed in the dogs, thus sug-
gesting that the general concepts presented above may
413
 KATO et al.
be more widely applicable. In many patients with
obstructive sleep apnea, the periods of upper airway
obstruction with hypoxemia are associated with a
bradycardiac response,31 which reverts to a tachycar-
diac response with the onset of breathing, even though
the 02 saturation under the two conditions may be
virtually the same. Zwillich et al.32 have shown that
this response is not due to apnea alone, since it is
attenuated when oxygen is administered to the subject,
even though the subsequent periods of apnea are length-
ened. A similar bradycardiac response is also often
observed in apneic premature infants. 33
In summary, we have investigated the effect of cyclic
lung stretch on the heart rate response to hypoxemia.
Our data indicate that cyclic afferent information from
the lungs increases the heart rate response. Our data
also suggest that the afferent receptors responsible for
this effect may be the slowly adapting receptors, but we
Downloaded from http://circ.ahajournals.org/ by guest on June 8, 2018
have not conclusively ruled out roles for other pulmo-
nary receptors.
We are indebted to Mrs. Elena Mates for her technical assis-
tance and to Miss Sheri Hodgson for preparing the manuscript.
We thank Winthrop Laboratories for providing the bupivacaine
hydrochloride.
References
1. Krasney JA, Koehier RC: Neural control of the circulation during
hypoxia. In Hughes MJ, Barnes CD, editors: Neural control of
circulation. New York, 1980, Academic Press, pp 123-147
2. Bernthal T, Greene W Jr, Revzin AM: Role of carotid chemore-
ceptors in hypoxic cardiac acceleration. Proc Soc Exp Biol Med 76:
121, 1951
3. Daly MD, Scott MJ: The effects of stimulation of the carotid body
chemoreceptors on heart rate in the dog. J Physiol (Lond) 144: 148,
1958
4. Daly MD, Scott MJ: An analysis of the primary cardiovascular
reflex effects of stimulation of the carotid body chemoreceptors in
the dog. J Physiol (Lond) 162: 555, 1962
5. Jacobs L, Sampson SR, Comroe JH Jr: Carotid sinus versus carotid
body origin of nicotine and cyanide bradycardia in the dog. Am J
Physiol 220: 472, 1971
6. Daly MD, Scott MJ: The effects of changes in respiration on the
cardiovascular responses to stimulation of the carotid body che-
moreceptors. In Cunningham DJC, Lloyd BB, editors: The regu-
lation of human respiration: John Scott Haldane Centenary Volume.
Oxford, 1963, Blackwell, pp 149-162
7. Kontos HA, Mauck HP Jr, Richardson DW, Patterson JL Jr: Mech-
anism of circulatory responses to systemic hypoxia in the anesthe-
tized dog. Am J Physiol 209: 397, 1965
8. Krasney JA: Efferent components of the cardioaccelerator re-
sponses to oxygen lack and cyanide. Am J Physiol 213:1475, 1967
9. Lehnert BE, Oberdorster G, Slutsky AS: Constant-flow ventilation
of apneic dogs. J Appl Physiol 53: 483, 1982
10. Widdicombe JG: Respiratory reflexes in man and other mammalian
species. Clin Sci 21: 163, 1961
11. Angell-James JE, Elsner R, Daly MD: Lung inflation: effects on
414
heart rate, respiration, and vagal afferent activity in seals. Am J
Physiol 240: H190, 1981
12. Slutsky AS, Menon AS: Catheter position and blood gases during
constant flow ventilation. J Appl Physiol 61: 513, 1987
13. Jain SK, Trenchard D, Reynolds F, Noble MIM, Guz A: The effect
of local anesthesia of the airway on respiratory reflexes in the rabbit.
Clin Sci 44: 519, 1973
14. Cross BA, Guz A, Jain SK, Archer S, Stevens J, Reynolds F: The
effect of anaesthesia of the airway in dog and man: a study of
respiratory reflexes, sensations and lung mechanics. Clin Sci Mol
Med 50: 439, 1976
15. Reeves RB, Park JS, Lapennas GN, Olszowka AJ: Oxygen affinity
and Bohr coefficients of dog blood. J Appl Physiol 53: 87, 1982
16. Younes M, Vaillancourt P, Milic-Emili: Interaction between chem-
ical factors and duration of apnea following lung inflation. J Appl
Physiol 36: 190, 1974
17. Menon AS, England SJ, Vallieres E, Rebuck AS, Slutsky AS:
Influence of phasic afferent information on phrenic neural output
during hypercapnia. Fed Proc 44: 1124, 1986(abst)
18. Tibes U: Reflex inputs to the cardiovascular and respiratory centers
from dynamically working canine muscles; some evidence for
involvement of group III or IV nerve fibers. Circ Res 41: 332, 1977
19. Kao FF, Lahiri S, Wang C, Mei SS: Ventilation and cardiac output
in exercise; interaction of chemical and work stimuli. Circ Res
20(suppl I): I-179, 1967
20. Band DM, Wolff CB, Ward J, Cochrane GM, Prior J: Respiratory
oscillations in arterial carbon dioxide tension as a control signal in
exercise. Nature 283: 84, 1980
21. DainDS, Boushey HA, Gold WM: Inhibition of respiratory reflexes
by local anesthetic aerosols in dogs and rabbits. J Appl Physiol 38:
1045, 1975
22. Hamilton RD, Winning AJ, Guz A: Blockade of 'alveolar' and
airway reflexes by local anesthetic aerosol in dogs. Respir Physiol
67: 159, 1987
23. Kontos HA, Vetrovec GW, Richardson DW: Role of carotid che-
moreceptors in circulatory response to hypoxia in dogs. J Appl
Physiol 28: 561, 1970
24. McQueen DS, Ribeiro JA: Inhibitory actions of methionine-enke-
phaline and morphine on the cat carotid chemoreceptors. Br J
Pharmacol 71: 297, 1980
25. Green JF, Schertel ER, Coleridge HM, Coleridge JCG: Effect of
pulmonary arterial PCO2 on slowly adapting pulmonary stretch
receptors. J Appl Physiol 60: 2048, 1986
26. Schoener EP, Frankel HM: Effect of hyperthermia and PaCO2 on
the slowly adapting pulmonary stretch receptor. Am J Physiol 222:
68, 1972
27. Kontos HA, Lower RR: Role of beta-adrenergic receptors in the
circulatory response to hypoxia. Am J Physiol 217: 756, 1969
28. Adrian ED: Afferent impulses in the vagus and their effect on
respiration. J Physiol (Lond) 79: 332, 1933
29. Coleridge HM, Coleridge JCG: Reflexes evoked from tracheobron-
chial tree and lungs. In Fishman AP, Cherniack NS, Widdicombe
JG, and Geiger SR, editors: Handbook of physiology. Section 3:
The respiratory system. Volume II. Control of Breathing, part 1.
Bethesda, 1986, American Physiological Society, pp 395-429
30. Anrep GV, Pascual E, Rossler R: Respiratory variations of the heart
rate. II-The central mechanism of the respiratory arrhythmia and the
inter-relations between the central and the reflex mechanisms. Proc
Roy Soc Lond [Biol] 119: 218, 1936
31. TiLkian AG, MottaJ, GuilleminaultC: Cardiac arrhythmias in sleep
apnea. In Guilleminault C, Dement WC, editors: Sleep apnea
syndromes. New York, 1978, Alan R. Liss, Inc., pp 197-210
32. Zwillich C, Devlin T, White D, Douglas N, Weil J, Martin R:
Bradycardia during sleep apnea: characteristics and mechanism. J
Clin Invest 69: 1286, 1982
33. Daily WJR, Klaus M, Meyer HBP: Apnea in premature infants:
monitoring, incidence, heart rate changes, and an effect of envi-
ronmental temperance. Pediatrics 43: 510, 1969
CIRCULATION
 Mechanisms mediating the heart rate response to hypoxemia.
H Kato, A S Menon and A S Slutsky

Circulation. 1988;77:407-414
doi: 10.1161/01.CIR.77.2.407
Downloaded from http://circ.ahajournals.org/ by guest on June 8, 2018

Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1988 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539

The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/77/2/407

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at:

http://www.lww.com/reprints

Subscriptions: Information about subscribing to Circulation is online at:

http://circ.ahajournals.org//subscriptions/