JORNAL OF APPLIED PHYSIOLOGY Vol. 41, No . 6, December 1976. Printed
in U.S.A.
Effect of beta-adrenergic exercise on ventilation
blockade during and gas exchange
HARVEY V. BROWN, KARLMAN WASSERMAN, AND BRIAN J. WHIPP Division of Respiratory Physiology and Medicine, Department of Medicine, Harbor General Hospital-UCLA School of Medicine, Torrance, California 90509
BROWN, HARVEY V., KARLMAN WHIPP. Effect of beta-adrenergic
WASSERMAN, AND BRIAN J. bZockade during exercise on ventilation and gas exchange. J. Appl. Physiol. 41(6): 886892. 1976. -The ventilatory effects of beta-adrenergic blockade during steady-state exercise were studied in eight normal subjects using intravenous propranolol hydrochloride (0.2 mg/ kg). Heart rate decreased in all subjects by an average of 17%. Coincident with the phase of decreasing heart rate was a significant decrease in both minute ventilation (VE) and CO, output (VCO,), averaging 9.6 and 9.2%, respectively. Both functions returned to prepropranolol levels after heart rate had reached its reduced steady-state value. The change in VE was significantly correlated with the change in ho2 (r = 0.85, P < O.OOS>,and was associated with negligible changes in endtidal CO, tensions and ventilatory equivalents for CO,. We interpret these studies as showing that the transient isocapnic hypopnea concomitant with an acute reduction in cardiac output was secondary to a transient decrease in CO, flux (cardiac output x mixed venous CO, content). This decrease in VE appears to be induced by the acute decrease in cardiac output (“cardiodynamic hypopnea”), in fashion similar to the previously described cardiodynamic hyperpnea (J. Appl. Physiol. 36: 457-464, 1974).
respiratory hyperpnea;
control; cardiodynamic hypopnea; cardiodynamic isocapnic hyperpnea; exercise ventilation
stancy of arterial Pco2 despite widely ranging work rates and alveolar ventilations (19). These observations suggest the existence of a fine regulatory mechanism which controls the level of alveolar ventilation in order to minimize change in arterial Pco2 despite considerable change in CO, production. To further test the mechanism, we induced hemodynamic changes in exercising man with intravenous propranolol and followed the changes in VE and end-tidal Pco2 (PET~~J. Epstein et al. (5) have shown that propranolol hydrochloride, a beta-adrenergic blocker, administered intravenously to normal human subjects during submaximal exercise, causes decreases in heart rate, cardiac output, and pulmonary artery oxygen saturation which are sustained, while oxygen consumption and arterial oxygen saturation remain unchanged. The present studies in man demonstrate the effects of similar doses of the same drug administered during steadystate exercise, on ventilation and gas exchange. Our results indicate that following the hemodynamic alterations induced by propranolol there are changes in VE consistent with a ventilatory control mechanism which functions to minimize changes in arterial Pcoz. METHODS
THE
PRECISE
REGULATION
of arterial
carbon
dioxide
ten-
sion (Pco~) during moderate intensity exercise in normal man suggests a tight coupling between minute ventilation (VE) and CO, output #co,> (13, 17). If there is a causal relationship between the constancy of arterial PCO* and the associated close relationship between vcoZ and VE, then changes in vcoZ should cause secondary changes in VE. Since vco2 is the product of cardiac output (Q) and the difference in venous and arterial CO, content (CVcoZ- Caco,), it follows that changes in ho2 would be induced by acute alterations in Q, Cijco2 Ca coZYor both. This scheme has been tested experimentally in the dog. Altering the CO, flux (Q x CV,,,) to the lungs, by increasing cardiac output, results in ventilatory adjustments which maintain end-tidal (and presumably arterial) Pcoz constant (16). Also, increasing CO, flux to the lungs by intravenous CO, loading in rats and dogs has been shown to produce an hyperpnea with little, if any, increase in arterial Pcoz (15, 20). Recent studies in man, using sinusoidal forcing work functions below the anaerobic threshold (17), confirm the close coupling of J?co2 and VE (2), and demonstrate the con-
Eight normal subjects exercised on a cycle ergometer at a constant work rate for a duration of 16-20 min. The work rate selected for each subject was equivalent to approximately 80% of his anaerobic threshold (AT), as determined by an incremental exercise test (17) performed on a previous day. The subjects were blindfolded and music was played at a low volume to encourage relaxation and minimize random breathing fluctuations. After a minimum of 6 min of exercise, propranolol hydrochloride (0.2 mg/kg) was injected intravenously, over a l-min period through an indwelling venous catheter. The catheter was continuously infused with 0.9% sodium chloride solution at a rate of approximately 1 ml/min. This infusion was deliberately increased for short periods during the study to obscure the actual time of drug injection. Questioning after the study revealed that the subjects were unable to discern the time of actual administration of propranolol. Exercise was continued for 10 min after injection of the drug. Subjects breathed through a low-resistance valve (Otis-NIcKerrow), with a 210-ml dead space. Expiratory flow was measured with a Fleisch pneumotachograph
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
BETA-ADRENERGIC
BLOCKADE,
VENTILATION,
AND
GAS
(no. 3). Tidal volume (VT) was calculated breath by breath, by digital integration of the expiratory flow signal, and minute ventilation (ATPS) was derived from VT and the interbreath period (1). Calibration of the pneumotachograph was done with a 1.5liter syringe using a range of pulsed flow rates to encompass the values encountered during exercise. Respired gases were sampled continuously from the mouthpiece and analyzed for oxygen and CO, with a mass spectrometer (Perkin-Elmer MGA 1100). This instrument was calibrated each day with standard gas mixtures previously analyzed by the Scholander technique (11). On-line calculations of VE, oxygen consumption (h), VCOZ, PETco2, ventilatory equivalents for oxygen and carbon dioxide (TE/~o~ and ~E/%%o~), and heart rate (HR) were performed breath by breath by a minicomputer and stored on digital tape, as previously described (1). In addition, the ECG was monitored continuously on an oscilloscope from a modified standard lead I. For purposes of analysis, the study was divided into three experimental periods with respect to the drug infusion: the preinfusion phase, the response phase and the postinfusion phase (Fig. 1). The preinfusion phase comprised the 2 min immediately prior to the time of propranolol injection. To select the time when cardiac output was decreasing, the response phase was defined from the time that heart rate started to decrease after the injection of propranolol to the time where the heart rate had undergone 80% of its total change to a new reduced “steady-state” value. Since the injection was made over a 1-min period, and then flushed with saline through the intravenous catheter, a delay of approximately 1.5 min occurred between the time of the start of the injection and the onset of the fall in heart rate. The postinfusion phase comprised the 6th and 7th min following injection, by which time the heart rate had stabilized to it new value. Mean values for each respiratory variable during each phase were determined by averaging individual breath-by-breath values. Since the differences between the pre- and postinfusion phases were minor, their values were averaged to minimize the effects of any nonsteady-state conditions. Hereafter, this value is termed “exercise control,” and compared with the response phase using a two-sample t-test. Differences were judged significant at P c 0.05. The laboratory is air conditioned and its temperature averaged 22 k 0.5OC (range) during the study. This study was approved by the institutional human use committee and signed informed consent was obtained from each subject. RESULTS
Heart rate decreased in all subjects following propranolol injection (Table 1). The magnitude of the fall in heart rate was variable between individuals, but was related to the preinfusion heart rate, in that subjects with higher heart rates during exercise tended to show the largest decrease after propranolol (r = 0.86, P < 0.01). The mean preinfusion heart rate was 132 beats/ min, decreasing to a mean of 109 beats/min (Table l), an average decrease of 17%.
EXCHANGE TABLE
887
Subj
Identifying
1
exerczse
data
and
C haracteristics Heart
Sex
Rate
Pre
t
Post I
HM BB MW DH PF JJ GA LD Mean *SD
M M M M M M F F
173 188 198 183 194 183 166 161
100 86 86 93 81 70 64 55
21 19 24 37 25 24 18 22
75 90 130 90 105 110 75 90
1.54 1.46 1.76 1.46 1.64 1.38 1.18 1.21
138 118 124 138 114 119 158 149
181 t13.1
79 215.2
24 k5.9
96 k18.6
1.45 kO.20
132 f 15.7
* VO, (SS) = steady-state 0, consumption during rate pre and post refer to the steady state heart rates propranolol injection.
116 94 101 110 1 104 111 117 118 109 k8.6
study. t Heart before and after
An example of a record of the breath-by-breath changes in VE and gas exchange before, during, and after propranolol infusion is shown in Fig. 1. VE and ho2 started to decrease simultaneously with the start of the heart rate decrease and reached a nadir when heart rate reached its lowest value, following which both increased back to prepropranolol levels. A similar change of lesser magnitude occurred in \io,. The parallel changes in I?E and ho2 are reflected by the failure for the ventilatory equivalent for CO, to change during the response period. End-tidal CO, tension increased less than 0.5 mmHg in this subject during the response period suggesting close coupling of TE and vco.,. All subjects showed significant decreases in VE during the response phase, averaging 9.6 t 1.2 (SEM) % of their exercise control values (Fig. 2). Respiratory frequency (f) was not significantly changed in six of eight subjects; the remaining two decreased f by an average of 2.0 breaths/min. Thus, the changes in TE were predominantly due to changes in tidal volume. Comparisons of pre- and postinfusion values of VE (Table 2), showed no significant differences in six of eight subjects. The remaining two subjects (MW and JJ) had slightly higher postinfusion values, presumably reflecting nonsteadystate conditions at the time of propranolol injection (i.e., VE was continuing to rise slightly prior to the infusion). However, despite the tendency for VE to continue to increase at a slow rate in these two subj ects at the time of the start of the propranolol injection, both had a significant decrease in VE during the response phase. In all instances, the fall in VE was coincident with the subject’s fall in heart rate and decrease in VCO~, similar to the example shown in Fig. 1. After the nadir was reached, both VE and ho2 returned to preinfusion levels despite the persistent reduction in heart rate. The duration of the transient fall in I?E averaged 170 t 20 (SEM) s. End-tidal CO, tension was unchanged during the response phase in three subjects (Fig. 3). In the remaining five, end-tidal CO, tension increased; however, the magnitude of this increase for the group was small and averaged 0.71 mmHg. Four subjects had end-tidal CO, tensions averaging in excess of 45 mmHg during exercise, although all were normal at rest (s 42 mmHg).
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
888
BROWN,
WASSERMAN,
AND
WHIPP
FIG. 1. Breath-by-breath measurement of heart rate (HR), minute ventilation (VE), CO, output (%ko,), oxygen uptake (VOW), end-tidal Cq, tenSiOn(PETCO,), and ventilatory equivalent for CO, (~E/VCO~) in subject HM during steady-state exercise. Bar indicates time during which propranolol was injected intravenously. Exercise period shown is divided into 1) preinfusion phase (A), 2) response phase (B), and 3) postin-
WOp (Llmin)
rWyrlrsi, Is7 1 ]!Jlh 1.3
M
fusiOn
48
(0.
creases in VE (Fig. 5). In contrast, VO, showed no significant change in five subjects and diminished slightly in the remaining three (Fig. 4). There was a poorer correTABLE
I
-d2I-
Phase
ljll
2. Gas exchange I N
VE
during
exercise
ho*
PETco,
VO,
Subj
1
HM BB MW DH PF JJ GA LD
I 404
Pre Resp tI 38 24 24 40 38 22 39 19 38 17 38 16 29 16 40 31
Posi 45 44 47 43 36 39 37 59
N = no. of breaths postinfusion phase.
54
h
Pre
28-
analyzed;
1
RESPONSE
Resp
Post
Pre
Resp
Post
Pre
Resp
Post
46.99 48.46 46.14 40.34 41.02 52.20 45.11 40.55
46.71 48.43 44.62 38.81 40.36 51.34 42.70 38.47
1.43 1.35 1.43 1.32 1.65 1.45 1.56 1.41 1.48 1.32 1.48 1.26 1.16 0.99 1.29 1.13
1.40 1.40 1.62 1.46 1.49 1.49 1.11 1.23
1.54 1.46 1.76 1.46 1.64 1.38 1.18 1.21
1.47 1.46 1.62 1.45 1.54 1.30 1.08 1.18
1.48 1.51 1.69 1.49 1.64 1.47 1.13 1.21
Pre = preinfusion
e*
i
:*
i
f
phase;
Resp = response
phase;
Post =
I
cu s
2. Change in ventilation during period of decreasing heart rate (response phase) compared with exercise control. Points shown are mean values of total number of breaths averaged during each period (see Table 2). Vertical bars are + 1 SEM. All of the decreases during response phase are significant (P < 0.05). FIG.
Those with the higher values during exercise tended to be “fitter” subjects with higher AT’s, and were therefore exercising at a higher steady-state work rate with consequently higher levels of vcoz. ho2 demonstrated consistent decreases in each subject during the response phase (Fig. 4), which correlated well (r = 0.85, P < 0.005) with the corresponding de-
Pre
50
z -zz CONTROL
Post
-I I
; E
Resp
35.4633.65 35.37 46.56 34.94 33.09 35.45 48.82 40.43 37.08 42.78 46.45 46.49 42.30 45.5939.88 41.26 36.65 42.22 39.71 35.6932.40 38.09 53.11 29.24 25.6930.82 45.61 42.39 37.50 42.06 39.08
tl,
24l
variables I
2 i= 42
38-’
CONTROL
RESPONSE
3. Change in end-tidal CO, tension during period of decreasing heart rate (response phase) compared with exercise controls. Points have same significance as defined in legend for Fig. 2. Asterisk denotes nonsignificant changes (P > 0.05). FIG.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
BETA-ADRENERGIC
BLOCKADE,
VENTILATION,
AND
GAS
lation (r = 0.67, 0.05 > P > 0.01) between the decreases in VE and voz (Fig. 5). The magnitude of the decrease in vcoz was consistently greater than the decrease in vo, for each subject. There was close agreement between the magnitude of the average group decrease in VE (9.6%) and vco, (9.2%) during the response phase, whereas vo2 decreased an average of 4.2%. ~E/%o~ changed from a mean of 27.51 to 27.33 units for the group. In six subjects the change was not significant while each of the remaining two decreased ~E/%o, by only 0.64 units. In contrast, VE/~O~ decreased significantly in seven subjects, from a mean of 27.01 to 25.36 units.
ventilation in normal man at rest have been previously investigated, specifically with respect to the inhibition of the stimulating effects of various catecholamines. Keltz et al. (9) demonstrated a significant increase in minute ventilation (1.3 l/min) during isoproterenol infusion associated with a 25beatlmin increase in heart rate. Administration of intravenous propranolol, while these subjects continued to receive the isoproterenol, caused a return of both ventilation and heart rate to base-line values, whereas minute ventilation was unchanged during infusion of propranolol alone, at rest. Neither the catecholamine hyperpnea nor the subsequent decrease in J?E following the propranolol infusion was associated with a significant change in Pace,, although other investigators have reported hypocapnia as a consequence of catecholamine-induced hyperpnea (4,
DISCUSSION
The effects of beta-adrenergic
889
EXCHANGE
blockade on minute
Qco,
I .70-
: 1.505; I
FIG. 4. Change in CO, output (Vco,) and O2 uptake (\jo2) during the period of decreasing heart rate (response phase) compared with the exercise control. Points and vertical bars have the same significance as defined in legend of Fig. 2. Asterisk denotes nonsignificant changes.
.-: \ g.- 1.30A
bYs
I .lO-
1 CONTROL
O-
RESPONSE
CONTROL
RESPONSE
5. Correlation of percent decreases in and lkoz (lef? panel) and VE and vo2 (right panel) in each subject during response phase. Correlation coefficient for the h02-vE plot is 0.85 (P c 0.005) and for the v02-vE plot is 0.67 (P < 0.05). Solid line is least-squares regression line for data. Dashed line is line of identity. Standard error of the estimate for %o~-~E plot is 1.85 and for vO&E plot is 2.63. FIG.
VE
/
0 0 /
/ /
/
/
/
/
/
y=O.Sl x + 1.24 /
/
/
/
/
/
/
/
/
/
/’
/ y = 1.05x +5.24
/ /
I I
II
I I
4
II
I I
8 -PER
CENT CHANGE
I I
12
I I
1 I
I
I
I
I
4 vO, -PER
I I
8 CENT
I I
I2 CHANGE
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
890
BROWN,
WASSERMAN,
AND
WHIPP
h (L/mid
ko* (L/mid
VO* (L/mid
pETco2
(mm
Hg) R 0.8 VT
(L) f (breothdmin) FIG. 6. Breath-by-breath (VE), CO, output (vco,),
(PETE&
gas exchange
I 5J measurements of minute ventilation oxygen uptake (vo,), end-tidal CO, tension ratio (R), tidal volume (VT), and breathing
12, 18). Similar results using propranolol to block the
effects of isoproterenol were obtained by Heistad et al. (7). Keltz et al. (9) and Heistad et al. (7) concluded that the isoproterenol-induced hyperpnea was mediated by beta-adrenergic receptors. They failed however, to consider the effects of changing cardiac output on the CO, flux to the lungs and its possible role in the regulation of ventilation (16). Considerable breath-to-breath variability in measurements of minute ventilation at rest can occur secondary to psychogenic factors or swallowing. In addition, changes in VE as large as 10% may be dificult to detect because of the high noise-to-signal ratio of normal breathing at rest (8). For these reasons, transient changes may go undetected or may not be appreciated unless measurements can be made during the process of hemodynamic change. In this study, we used exercise to increase the level of VE and &02. Thus, changing hemodynamics resulted in greater changes in CO2 output, and consequently TE, than could have been observed if the study were done at rest. However, mixed venous CO, must increase if the reduction in cardiac output is sustained and the CO, output from the lungs must again equal the CO, produced at the cellular level. The fact that VE returns to the prepropranolol level after %oZ increases back to the steady-state level is re-affirmation that the cardiodynamic hypopnea is a CO*-linked phenomenon. During steady-state exercise at work rates below the anaerobic threshold, the constant load of metabolic CO, generated is accompanied by a level of alveolar ventilation which maintains Pace, near resting levels (13). VE
frequency (f) in a normal subject during steady-state exercise W. Horizontal bar indicates period of voluntary hypoventilation.
at 30
is therefore uniquely determined for each individual by the relationship among %ko2, VD/VT, and PacOzat the particular work rate (14). Given these relationships, and assuming the constancy of Pace, and VD/VT during exercise, a change in ho2 must be accompanied by a change in VE. The occurrence of a transient hypopnea after an exercise steady state was established can be explained by our hypothesis that a change in CO, flux to the lungs will be associated with ventilatory adjustments resulting in the maintenance of isocapnia. A sudden decrease in cardiac output during steady-state exercise would have the immediate effect of increasing the exercising muscle and mixed venous CO, stores by creating a transient discrepancy between the amount of CO2 produced metabolically and the amount of CO, delivered to the lungs. Farhi and Rahn (6) predicted that under these circumstances, if alveolar ventilation were kept constant by artificial ventilation, there would be a decrease in arterial (and alveolar) Pco2, the magnitude of change being constrained by the arteriovenous PCO* difference. The present studies indicate that, unlike the artificially ventilated animal, human subjects will not decrease PETITE, when breathing spontaneously. The transient nature of the ventilatory response can be explained by the restoration of CO, flux which occurs when mixed venous Pco2 rises and CO, stores return to a steady state. Of course, a widening of the arteriovenous 0, content should also occur when cardiac output is reduced. However, the reserve in tissue oxygen stores is small (3) and mixed venous 0, content would decrease rapidly. For
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
BETA-ADRENERGIC
BLOCKADE,
VENTILATION,
AND
GAS
EXCHANGE
891
this reason a significant decrease in Vo, was not measgroup was only 0.71 mmHg during the response phase, urable in most of our subjects. The smaller tissue 0, which itself was approximately one minute in duration. reserve, relative to its ability to increase in CO, content, Second, following cessation of the period of voluntary is reflected by the greater decrease in ho2 relative to hypoventilation (Fig. 6), there is an immediate compenvo, (Fig. 6), at the time of the decrease in VE. satory increase in VCO* and VE above control values as Although cardiac output was not directly measured in chemoreceptor control of VE responds to the blood gas our study, previous work (5) has demonstrated that effects of the reduced VE. In contrast, there was no propranolol, given intravenously during steady-state overshoot in ho2 and VE following the propranolol exercise, uniformly causes significant decreases in both injection in any subject (Table 2>, thus providing further heart rate and cardiac output which occur simultanesupport that the decrement in VE was proportional to ously. Thus, changes in heart rate will accurately re- the reduction in &. VE and ko2 returned to the preproflect the period during which cardiac output is decreas- pranolol level, not greater, in all subjects during the ing in exercising subjects given intravenous propranoposthypoventilation period (Table 2). This can only be 101. Our studies indicate that the hypopnea which ac- explained by less CO, being delivered to the lungs durcompanies this sudden decrease in cardiac output reing the period of hypopnea. sults in negligible changes in end-tidal Pco~. This obserThird, the transient decrease in ho2 and VOW,withvation suggests that the decrease in VE found in these out subsequent overshoot, are precisely what should be studies was appropriate to the change in CO, output. expected with a sustained decrease in cardiac output Figure 5 demonstrates that the proportional change in which has been shown to accompany beta-adrenergic the transient reduction in VE is, in fact, similar to that blockade. Thus, propranolol is not acting in any unexof the CO, output. The observation that changes in VE pected way with respect to gas exchange. The unique and \jco2 occurred coincidently, as evidenced by an feature of this control process is that VE transiently unchanging ~E/%%o~, and that both decreased only durfollows vcoz rather than remaining at the prepropranoing the period in which heart rate was decreasing sug- 101level with consequent hypocapnia. gests a causal relationship between these ventilatory Fourth, we know of no evidence for direct action of changes and the decrease in cardiac output (i.e., cardipropranolol on respiratory chemoreceptors or the respiodynamic hypopnea). These changes therefore are the ratory center which affects breathing. opposite of those observed in dogs given int.ravenous Fifth, the effect of propranolol on respiratory control isoproterenol (i.e., cardiodynamic hyperpnea (16)), and is transient and only associated with the period during may therefore be mediated by the same ventilatory which heart rate is decreasing (avg of 3 min). Thus, one control process. would have to postulate that the independent action of If propranolol produced a primary reduction in VE, propranolol on respiratory control is only transient while the effect on heart rate is sustained. The half-life unrelated to cardiac output, we would also expect to find a decrease in Jko2. However, we believe that a primary of the action of propranolol is known to be several hours. reduction in VE produced by the drug is an unlikely and The study also addresses itself to the importance of an unreasonable explanation for the hypopnea observed muscle receptors which might affect VE in man. When in these studies for the following reasons. cardiac output decreases during constant load exercise, First, a primary reduction in VE would be acmmpatissue PO, decreases and Pco2 and H+ increase, while nied by an immediate increase in PETITE in excess of CO, flux to the lungs decreases. Thus, the reduction in that observed during our studies (mean increase = 0.71 cardiac output transiently uncouples tissue metabolism mmHg) (see Table 2 for individual values). The magnifrom CO, flux to the lungs. The observation that VE tude of this increase, once a steady state among producchanges in a direction opposite to that which would be tion, stores, and output has been achieved, would be expected if tissue metabolite concentrations were important stimuli for the mechanism of the exercise hyperpproportional to the magnitude of the decrease in J?E. The time required to attain such a steady state, and nea argues against the importance of postulated tissue hence the full change in PETITE, would be dependent on metabolism receptors (10) in normal respiratory control. The proportional change of VE and vco,! reinforces the the rate of change of the lung and tissue CO, stores. At rest, the calculated time constant for the lung CO, concept that receptors for respiratory control “sense” CO, flux to the lungs. stores is generally short (co.5 min). The time constants We have demonstrated that ventilation diminishes and sizes of the tissue CO, compartments vary widely depending on the tissue (3). Blood would equilibrate transiently in response to acute decreases in cardiac rapidly and has a capacitance of 4.5 ml/l per mmHg output in man and that respiratory control mechanisms function to regulate Pco2 at a constant level. These while the time constant of bone is so long that it would not be measurably involved in the exchange for the adjustments in ventilation appear to depend on receptors affected by changes in CO, flux to the lungs, rather period of hypopnea observed in these studies. Indeed, approximately 50% of the total expected steady-state than oxygen consumption or Pco2 and PO, tensions in change in PET~-~ occurs within a few breaths when a the muscle cells, per se. normal subject voluntarily hypoventilates after reachThis study was supported by National Institutes of Health Grants ing a steady state of exercise at 90 W (Fig. 6). For our experimental subjects with a mean decrease in J?E of HL-05916 and HL-11907. B. J. Whipp is an Established Investigator of the American Heart 9%, we would therefore expect at least 50% of the 9% Association. expected rise in PETITEto occur or a mean increase of 2.2 mmHg. In fact, the mean increase observed for the Received for publication 3 March 1976. Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
892
BROWN,
WASSERMAN,
AND
WHIPP
REFERENCES 1. BEAVER, W. L., K. WASSERMAN, AND B. J. WHIPP. On-line computer analysis and breath-by-breath graphical display of exercise function tests. J. Appl. Physiol. 34: 128-132, 1973. 2. CASABURI, R., B. J. WHIPP, S. N. KOYAL, AND K. WASSERMAN. Ventilatory and gas exchange dynamics in response to sinusoidal work (Abstract). Federation Proc. 34: 431, 1975. 3. CHERNIACK, N. S., AND G. S. LONGOBARDO. Oxygen and carbon dioxide gas stores of the body. Physiol. Reu. 50: 196-243, 1970. 4. CUNNINGHAM, D. J. C., E. N. HEY, J. M. PATRICK, AND B. B. LLOYD. The effect of noradrenaline infusion on the relationship between pulmonary ventilation and the alveolar Pop and PCO* in man. Ann. N.Y. Acad. Sci. 109: 756-770, 1963. 5. EPSTEIN, S. E., B. F. ROBINSON, R. L. KAHLER, AND E. BRAUNWALD. Effects of beta-adrenergic blockade on the cardiac response to maximal and submaximal exercise in man. J. CZin. Invest. 44: 1745-1753, 1965. 6. FARHI, L. E., AND H. RAHN. Gas stores of the body and the unsteady state. J. AppZ. Physiol. 7: 472-484, 1955. 7. HEISTAD, D. D., R. C. WHEELER, A. L. MARK, P. G. SCHMID, AND F. M. ABBOUD. Effects of adrenergic stimulation on ventilation in man. J. CZin. Inuest. 51: 1469-1475, 1972. 8. HLASTALA, M. P., B. WRANNE, AND C. J. LENFANT. Cyclical variations in FRC and other respiratory variables in resting man. J. AppZ. Physiol. 34: 670-676, 1973. 9. KELTZ, H., T. SAMORTIN, AND D. J. STONE. Hyperventilation: a manifestation of exogenous beta-adrenergic stimulation. Am. Rev. Respirat. Diseases 105: 637-640, 1972. 10. LEVINE, S., AND W. E. HUCKABEE. Ventilatory response to druginduced hypermetabolism. J. AppZ. Physiol. 38: 827-833, 1975.
11. SCHOLANDER, P. F. Analyzer for accurate estimating of respiratory gases in one-half cubic centimeter samples. J. BioZ. Chew 167: 235-257, 1947. 12. STONE, D. J., H. KELTZ, T. K. SARKAR, AND J. SINGZON. Ventilatory response to alpha-adrenergic stimulation and inhibition. J. AppZ. Physiol. 34: 619-623, 1973. 13. WASSERMAN, K., A. L. VAN KESSEL, AND G. G. BURTON. Interaction of physiologic mechanisms during exercise. J. AppZ. PhysioZ. 22: 71-85, 1967. 14. WASSERMAN, K., AND B. J. WHIPP. Exercise physiology in health and disease. Am. Rev. Respirat. Diseases 112: 219-249, 1975. 15. WASSERMAN, K., B. J. WHIPP, R. CASABURI, D. J. HUNTSMAN, J. CASTAGNA, AND R. LUGLIANI. Regulation of arterial Pco2 during intravenous CO, loading. J. AppZ. PhysioZ. 38: 651-656, 1975. 16. WASSERMAN, K., B. J. WHIPP, AND J. CASTAGNA. Cardiodynamic hyperpnea; hyperpnea secondary to cardiac output increase. J. AppZ. PhysioZ. 36: 457-464, 1974. 17. WASSERMAN, K., B. J. WHIPP, S. N. KOYAL, AND W. L. BEAVER. Anaerobic threshold and respiratory gas exchange during exercise. J. AppZ. Physiol. 35: 236-243, 1973. 18. WHELAN, R. F., and I. M. YOUNG. The effect of adrenaline and noradrenaline infusions on respiration in man. Brit. J. PharmacoZ. 8: 98-102, 1953. 19. WHIPP, B. J., R. CASABURI, S. N. KOYAL, AND K. WASSERMAN. Regulation of arterial Pco2 during sinusoidal exercise (Abstract). Federation Proc. 34: 431, 1975. 20. YAMOMOTO, W. S., AND M. W. EDWARDS, JR. Homeostasis of carbon dioxide during intravenous infusion of carbon dioxide. J. AppZ. Physiol. 15: 807-818, 1960.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
in U.S.A.
Effect of beta-adrenergic exercise on ventilation
blockade during and gas exchange
HARVEY V. BROWN, KARLMAN WASSERMAN, AND BRIAN J. WHIPP Division of Respiratory Physiology and Medicine, Department of Medicine, Harbor General Hospital-UCLA School of Medicine, Torrance, California 90509
BROWN, HARVEY V., KARLMAN WHIPP. Effect of beta-adrenergic
WASSERMAN, AND BRIAN J. bZockade during exercise on ventilation and gas exchange. J. Appl. Physiol. 41(6): 886892. 1976. -The ventilatory effects of beta-adrenergic blockade during steady-state exercise were studied in eight normal subjects using intravenous propranolol hydrochloride (0.2 mg/ kg). Heart rate decreased in all subjects by an average of 17%. Coincident with the phase of decreasing heart rate was a significant decrease in both minute ventilation (VE) and CO, output (VCO,), averaging 9.6 and 9.2%, respectively. Both functions returned to prepropranolol levels after heart rate had reached its reduced steady-state value. The change in VE was significantly correlated with the change in ho2 (r = 0.85, P < O.OOS>,and was associated with negligible changes in endtidal CO, tensions and ventilatory equivalents for CO,. We interpret these studies as showing that the transient isocapnic hypopnea concomitant with an acute reduction in cardiac output was secondary to a transient decrease in CO, flux (cardiac output x mixed venous CO, content). This decrease in VE appears to be induced by the acute decrease in cardiac output (“cardiodynamic hypopnea”), in fashion similar to the previously described cardiodynamic hyperpnea (J. Appl. Physiol. 36: 457-464, 1974).
respiratory hyperpnea;
control; cardiodynamic hypopnea; cardiodynamic isocapnic hyperpnea; exercise ventilation
stancy of arterial Pco2 despite widely ranging work rates and alveolar ventilations (19). These observations suggest the existence of a fine regulatory mechanism which controls the level of alveolar ventilation in order to minimize change in arterial Pco2 despite considerable change in CO, production. To further test the mechanism, we induced hemodynamic changes in exercising man with intravenous propranolol and followed the changes in VE and end-tidal Pco2 (PET~~J. Epstein et al. (5) have shown that propranolol hydrochloride, a beta-adrenergic blocker, administered intravenously to normal human subjects during submaximal exercise, causes decreases in heart rate, cardiac output, and pulmonary artery oxygen saturation which are sustained, while oxygen consumption and arterial oxygen saturation remain unchanged. The present studies in man demonstrate the effects of similar doses of the same drug administered during steadystate exercise, on ventilation and gas exchange. Our results indicate that following the hemodynamic alterations induced by propranolol there are changes in VE consistent with a ventilatory control mechanism which functions to minimize changes in arterial Pcoz. METHODS
THE
PRECISE
REGULATION
of arterial
carbon
dioxide
ten-
sion (Pco~) during moderate intensity exercise in normal man suggests a tight coupling between minute ventilation (VE) and CO, output #co,> (13, 17). If there is a causal relationship between the constancy of arterial PCO* and the associated close relationship between vcoZ and VE, then changes in vcoZ should cause secondary changes in VE. Since vco2 is the product of cardiac output (Q) and the difference in venous and arterial CO, content (CVcoZ- Caco,), it follows that changes in ho2 would be induced by acute alterations in Q, Cijco2 Ca coZYor both. This scheme has been tested experimentally in the dog. Altering the CO, flux (Q x CV,,,) to the lungs, by increasing cardiac output, results in ventilatory adjustments which maintain end-tidal (and presumably arterial) Pcoz constant (16). Also, increasing CO, flux to the lungs by intravenous CO, loading in rats and dogs has been shown to produce an hyperpnea with little, if any, increase in arterial Pcoz (15, 20). Recent studies in man, using sinusoidal forcing work functions below the anaerobic threshold (17), confirm the close coupling of J?co2 and VE (2), and demonstrate the con-
Eight normal subjects exercised on a cycle ergometer at a constant work rate for a duration of 16-20 min. The work rate selected for each subject was equivalent to approximately 80% of his anaerobic threshold (AT), as determined by an incremental exercise test (17) performed on a previous day. The subjects were blindfolded and music was played at a low volume to encourage relaxation and minimize random breathing fluctuations. After a minimum of 6 min of exercise, propranolol hydrochloride (0.2 mg/kg) was injected intravenously, over a l-min period through an indwelling venous catheter. The catheter was continuously infused with 0.9% sodium chloride solution at a rate of approximately 1 ml/min. This infusion was deliberately increased for short periods during the study to obscure the actual time of drug injection. Questioning after the study revealed that the subjects were unable to discern the time of actual administration of propranolol. Exercise was continued for 10 min after injection of the drug. Subjects breathed through a low-resistance valve (Otis-NIcKerrow), with a 210-ml dead space. Expiratory flow was measured with a Fleisch pneumotachograph
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
BETA-ADRENERGIC
BLOCKADE,
VENTILATION,
AND
GAS
(no. 3). Tidal volume (VT) was calculated breath by breath, by digital integration of the expiratory flow signal, and minute ventilation (ATPS) was derived from VT and the interbreath period (1). Calibration of the pneumotachograph was done with a 1.5liter syringe using a range of pulsed flow rates to encompass the values encountered during exercise. Respired gases were sampled continuously from the mouthpiece and analyzed for oxygen and CO, with a mass spectrometer (Perkin-Elmer MGA 1100). This instrument was calibrated each day with standard gas mixtures previously analyzed by the Scholander technique (11). On-line calculations of VE, oxygen consumption (h), VCOZ, PETco2, ventilatory equivalents for oxygen and carbon dioxide (TE/~o~ and ~E/%%o~), and heart rate (HR) were performed breath by breath by a minicomputer and stored on digital tape, as previously described (1). In addition, the ECG was monitored continuously on an oscilloscope from a modified standard lead I. For purposes of analysis, the study was divided into three experimental periods with respect to the drug infusion: the preinfusion phase, the response phase and the postinfusion phase (Fig. 1). The preinfusion phase comprised the 2 min immediately prior to the time of propranolol injection. To select the time when cardiac output was decreasing, the response phase was defined from the time that heart rate started to decrease after the injection of propranolol to the time where the heart rate had undergone 80% of its total change to a new reduced “steady-state” value. Since the injection was made over a 1-min period, and then flushed with saline through the intravenous catheter, a delay of approximately 1.5 min occurred between the time of the start of the injection and the onset of the fall in heart rate. The postinfusion phase comprised the 6th and 7th min following injection, by which time the heart rate had stabilized to it new value. Mean values for each respiratory variable during each phase were determined by averaging individual breath-by-breath values. Since the differences between the pre- and postinfusion phases were minor, their values were averaged to minimize the effects of any nonsteady-state conditions. Hereafter, this value is termed “exercise control,” and compared with the response phase using a two-sample t-test. Differences were judged significant at P c 0.05. The laboratory is air conditioned and its temperature averaged 22 k 0.5OC (range) during the study. This study was approved by the institutional human use committee and signed informed consent was obtained from each subject. RESULTS
Heart rate decreased in all subjects following propranolol injection (Table 1). The magnitude of the fall in heart rate was variable between individuals, but was related to the preinfusion heart rate, in that subjects with higher heart rates during exercise tended to show the largest decrease after propranolol (r = 0.86, P < 0.01). The mean preinfusion heart rate was 132 beats/ min, decreasing to a mean of 109 beats/min (Table l), an average decrease of 17%.
EXCHANGE TABLE
887
Subj
Identifying
1
exerczse
data
and
C haracteristics Heart
Sex
Rate
Pre
t
Post I
HM BB MW DH PF JJ GA LD Mean *SD
M M M M M M F F
173 188 198 183 194 183 166 161
100 86 86 93 81 70 64 55
21 19 24 37 25 24 18 22
75 90 130 90 105 110 75 90
1.54 1.46 1.76 1.46 1.64 1.38 1.18 1.21
138 118 124 138 114 119 158 149
181 t13.1
79 215.2
24 k5.9
96 k18.6
1.45 kO.20
132 f 15.7
* VO, (SS) = steady-state 0, consumption during rate pre and post refer to the steady state heart rates propranolol injection.
116 94 101 110 1 104 111 117 118 109 k8.6
study. t Heart before and after
An example of a record of the breath-by-breath changes in VE and gas exchange before, during, and after propranolol infusion is shown in Fig. 1. VE and ho2 started to decrease simultaneously with the start of the heart rate decrease and reached a nadir when heart rate reached its lowest value, following which both increased back to prepropranolol levels. A similar change of lesser magnitude occurred in \io,. The parallel changes in I?E and ho2 are reflected by the failure for the ventilatory equivalent for CO, to change during the response period. End-tidal CO, tension increased less than 0.5 mmHg in this subject during the response period suggesting close coupling of TE and vco.,. All subjects showed significant decreases in VE during the response phase, averaging 9.6 t 1.2 (SEM) % of their exercise control values (Fig. 2). Respiratory frequency (f) was not significantly changed in six of eight subjects; the remaining two decreased f by an average of 2.0 breaths/min. Thus, the changes in TE were predominantly due to changes in tidal volume. Comparisons of pre- and postinfusion values of VE (Table 2), showed no significant differences in six of eight subjects. The remaining two subjects (MW and JJ) had slightly higher postinfusion values, presumably reflecting nonsteadystate conditions at the time of propranolol injection (i.e., VE was continuing to rise slightly prior to the infusion). However, despite the tendency for VE to continue to increase at a slow rate in these two subj ects at the time of the start of the propranolol injection, both had a significant decrease in VE during the response phase. In all instances, the fall in VE was coincident with the subject’s fall in heart rate and decrease in VCO~, similar to the example shown in Fig. 1. After the nadir was reached, both VE and ho2 returned to preinfusion levels despite the persistent reduction in heart rate. The duration of the transient fall in I?E averaged 170 t 20 (SEM) s. End-tidal CO, tension was unchanged during the response phase in three subjects (Fig. 3). In the remaining five, end-tidal CO, tension increased; however, the magnitude of this increase for the group was small and averaged 0.71 mmHg. Four subjects had end-tidal CO, tensions averaging in excess of 45 mmHg during exercise, although all were normal at rest (s 42 mmHg).
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
888
BROWN,
WASSERMAN,
AND
WHIPP
FIG. 1. Breath-by-breath measurement of heart rate (HR), minute ventilation (VE), CO, output (%ko,), oxygen uptake (VOW), end-tidal Cq, tenSiOn(PETCO,), and ventilatory equivalent for CO, (~E/VCO~) in subject HM during steady-state exercise. Bar indicates time during which propranolol was injected intravenously. Exercise period shown is divided into 1) preinfusion phase (A), 2) response phase (B), and 3) postin-
WOp (Llmin)
rWyrlrsi, Is7 1 ]!Jlh 1.3
M
fusiOn
48
(0.
creases in VE (Fig. 5). In contrast, VO, showed no significant change in five subjects and diminished slightly in the remaining three (Fig. 4). There was a poorer correTABLE
I
-d2I-
Phase
ljll
2. Gas exchange I N
VE
during
exercise
ho*
PETco,
VO,
Subj
1
HM BB MW DH PF JJ GA LD
I 404
Pre Resp tI 38 24 24 40 38 22 39 19 38 17 38 16 29 16 40 31
Posi 45 44 47 43 36 39 37 59
N = no. of breaths postinfusion phase.
54
h
Pre
28-
analyzed;
1
RESPONSE
Resp
Post
Pre
Resp
Post
Pre
Resp
Post
46.99 48.46 46.14 40.34 41.02 52.20 45.11 40.55
46.71 48.43 44.62 38.81 40.36 51.34 42.70 38.47
1.43 1.35 1.43 1.32 1.65 1.45 1.56 1.41 1.48 1.32 1.48 1.26 1.16 0.99 1.29 1.13
1.40 1.40 1.62 1.46 1.49 1.49 1.11 1.23
1.54 1.46 1.76 1.46 1.64 1.38 1.18 1.21
1.47 1.46 1.62 1.45 1.54 1.30 1.08 1.18
1.48 1.51 1.69 1.49 1.64 1.47 1.13 1.21
Pre = preinfusion
e*
i
:*
i
f
phase;
Resp = response
phase;
Post =
I
cu s
2. Change in ventilation during period of decreasing heart rate (response phase) compared with exercise control. Points shown are mean values of total number of breaths averaged during each period (see Table 2). Vertical bars are + 1 SEM. All of the decreases during response phase are significant (P < 0.05). FIG.
Those with the higher values during exercise tended to be “fitter” subjects with higher AT’s, and were therefore exercising at a higher steady-state work rate with consequently higher levels of vcoz. ho2 demonstrated consistent decreases in each subject during the response phase (Fig. 4), which correlated well (r = 0.85, P < 0.005) with the corresponding de-
Pre
50
z -zz CONTROL
Post
-I I
; E
Resp
35.4633.65 35.37 46.56 34.94 33.09 35.45 48.82 40.43 37.08 42.78 46.45 46.49 42.30 45.5939.88 41.26 36.65 42.22 39.71 35.6932.40 38.09 53.11 29.24 25.6930.82 45.61 42.39 37.50 42.06 39.08
tl,
24l
variables I
2 i= 42
38-’
CONTROL
RESPONSE
3. Change in end-tidal CO, tension during period of decreasing heart rate (response phase) compared with exercise controls. Points have same significance as defined in legend for Fig. 2. Asterisk denotes nonsignificant changes (P > 0.05). FIG.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
BETA-ADRENERGIC
BLOCKADE,
VENTILATION,
AND
GAS
lation (r = 0.67, 0.05 > P > 0.01) between the decreases in VE and voz (Fig. 5). The magnitude of the decrease in vcoz was consistently greater than the decrease in vo, for each subject. There was close agreement between the magnitude of the average group decrease in VE (9.6%) and vco, (9.2%) during the response phase, whereas vo2 decreased an average of 4.2%. ~E/%o~ changed from a mean of 27.51 to 27.33 units for the group. In six subjects the change was not significant while each of the remaining two decreased ~E/%o, by only 0.64 units. In contrast, VE/~O~ decreased significantly in seven subjects, from a mean of 27.01 to 25.36 units.
ventilation in normal man at rest have been previously investigated, specifically with respect to the inhibition of the stimulating effects of various catecholamines. Keltz et al. (9) demonstrated a significant increase in minute ventilation (1.3 l/min) during isoproterenol infusion associated with a 25beatlmin increase in heart rate. Administration of intravenous propranolol, while these subjects continued to receive the isoproterenol, caused a return of both ventilation and heart rate to base-line values, whereas minute ventilation was unchanged during infusion of propranolol alone, at rest. Neither the catecholamine hyperpnea nor the subsequent decrease in J?E following the propranolol infusion was associated with a significant change in Pace,, although other investigators have reported hypocapnia as a consequence of catecholamine-induced hyperpnea (4,
DISCUSSION
The effects of beta-adrenergic
889
EXCHANGE
blockade on minute
Qco,
I .70-
: 1.505; I
FIG. 4. Change in CO, output (Vco,) and O2 uptake (\jo2) during the period of decreasing heart rate (response phase) compared with the exercise control. Points and vertical bars have the same significance as defined in legend of Fig. 2. Asterisk denotes nonsignificant changes.
.-: \ g.- 1.30A
bYs
I .lO-
1 CONTROL
O-
RESPONSE
CONTROL
RESPONSE
5. Correlation of percent decreases in and lkoz (lef? panel) and VE and vo2 (right panel) in each subject during response phase. Correlation coefficient for the h02-vE plot is 0.85 (P c 0.005) and for the v02-vE plot is 0.67 (P < 0.05). Solid line is least-squares regression line for data. Dashed line is line of identity. Standard error of the estimate for %o~-~E plot is 1.85 and for vO&E plot is 2.63. FIG.
VE
/
0 0 /
/ /
/
/
/
/
/
y=O.Sl x + 1.24 /
/
/
/
/
/
/
/
/
/
/’
/ y = 1.05x +5.24
/ /
I I
II
I I
4
II
I I
8 -PER
CENT CHANGE
I I
12
I I
1 I
I
I
I
I
4 vO, -PER
I I
8 CENT
I I
I2 CHANGE
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
890
BROWN,
WASSERMAN,
AND
WHIPP
h (L/mid
ko* (L/mid
VO* (L/mid
pETco2
(mm
Hg) R 0.8 VT
(L) f (breothdmin) FIG. 6. Breath-by-breath (VE), CO, output (vco,),
(PETE&
gas exchange
I 5J measurements of minute ventilation oxygen uptake (vo,), end-tidal CO, tension ratio (R), tidal volume (VT), and breathing
12, 18). Similar results using propranolol to block the
effects of isoproterenol were obtained by Heistad et al. (7). Keltz et al. (9) and Heistad et al. (7) concluded that the isoproterenol-induced hyperpnea was mediated by beta-adrenergic receptors. They failed however, to consider the effects of changing cardiac output on the CO, flux to the lungs and its possible role in the regulation of ventilation (16). Considerable breath-to-breath variability in measurements of minute ventilation at rest can occur secondary to psychogenic factors or swallowing. In addition, changes in VE as large as 10% may be dificult to detect because of the high noise-to-signal ratio of normal breathing at rest (8). For these reasons, transient changes may go undetected or may not be appreciated unless measurements can be made during the process of hemodynamic change. In this study, we used exercise to increase the level of VE and &02. Thus, changing hemodynamics resulted in greater changes in CO2 output, and consequently TE, than could have been observed if the study were done at rest. However, mixed venous CO, must increase if the reduction in cardiac output is sustained and the CO, output from the lungs must again equal the CO, produced at the cellular level. The fact that VE returns to the prepropranolol level after %oZ increases back to the steady-state level is re-affirmation that the cardiodynamic hypopnea is a CO*-linked phenomenon. During steady-state exercise at work rates below the anaerobic threshold, the constant load of metabolic CO, generated is accompanied by a level of alveolar ventilation which maintains Pace, near resting levels (13). VE
frequency (f) in a normal subject during steady-state exercise W. Horizontal bar indicates period of voluntary hypoventilation.
at 30
is therefore uniquely determined for each individual by the relationship among %ko2, VD/VT, and PacOzat the particular work rate (14). Given these relationships, and assuming the constancy of Pace, and VD/VT during exercise, a change in ho2 must be accompanied by a change in VE. The occurrence of a transient hypopnea after an exercise steady state was established can be explained by our hypothesis that a change in CO, flux to the lungs will be associated with ventilatory adjustments resulting in the maintenance of isocapnia. A sudden decrease in cardiac output during steady-state exercise would have the immediate effect of increasing the exercising muscle and mixed venous CO, stores by creating a transient discrepancy between the amount of CO2 produced metabolically and the amount of CO, delivered to the lungs. Farhi and Rahn (6) predicted that under these circumstances, if alveolar ventilation were kept constant by artificial ventilation, there would be a decrease in arterial (and alveolar) Pco2, the magnitude of change being constrained by the arteriovenous PCO* difference. The present studies indicate that, unlike the artificially ventilated animal, human subjects will not decrease PETITE, when breathing spontaneously. The transient nature of the ventilatory response can be explained by the restoration of CO, flux which occurs when mixed venous Pco2 rises and CO, stores return to a steady state. Of course, a widening of the arteriovenous 0, content should also occur when cardiac output is reduced. However, the reserve in tissue oxygen stores is small (3) and mixed venous 0, content would decrease rapidly. For
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
BETA-ADRENERGIC
BLOCKADE,
VENTILATION,
AND
GAS
EXCHANGE
891
this reason a significant decrease in Vo, was not measgroup was only 0.71 mmHg during the response phase, urable in most of our subjects. The smaller tissue 0, which itself was approximately one minute in duration. reserve, relative to its ability to increase in CO, content, Second, following cessation of the period of voluntary is reflected by the greater decrease in ho2 relative to hypoventilation (Fig. 6), there is an immediate compenvo, (Fig. 6), at the time of the decrease in VE. satory increase in VCO* and VE above control values as Although cardiac output was not directly measured in chemoreceptor control of VE responds to the blood gas our study, previous work (5) has demonstrated that effects of the reduced VE. In contrast, there was no propranolol, given intravenously during steady-state overshoot in ho2 and VE following the propranolol exercise, uniformly causes significant decreases in both injection in any subject (Table 2>, thus providing further heart rate and cardiac output which occur simultanesupport that the decrement in VE was proportional to ously. Thus, changes in heart rate will accurately re- the reduction in &. VE and ko2 returned to the preproflect the period during which cardiac output is decreas- pranolol level, not greater, in all subjects during the ing in exercising subjects given intravenous propranoposthypoventilation period (Table 2). This can only be 101. Our studies indicate that the hypopnea which ac- explained by less CO, being delivered to the lungs durcompanies this sudden decrease in cardiac output reing the period of hypopnea. sults in negligible changes in end-tidal Pco~. This obserThird, the transient decrease in ho2 and VOW,withvation suggests that the decrease in VE found in these out subsequent overshoot, are precisely what should be studies was appropriate to the change in CO, output. expected with a sustained decrease in cardiac output Figure 5 demonstrates that the proportional change in which has been shown to accompany beta-adrenergic the transient reduction in VE is, in fact, similar to that blockade. Thus, propranolol is not acting in any unexof the CO, output. The observation that changes in VE pected way with respect to gas exchange. The unique and \jco2 occurred coincidently, as evidenced by an feature of this control process is that VE transiently unchanging ~E/%%o~, and that both decreased only durfollows vcoz rather than remaining at the prepropranoing the period in which heart rate was decreasing sug- 101level with consequent hypocapnia. gests a causal relationship between these ventilatory Fourth, we know of no evidence for direct action of changes and the decrease in cardiac output (i.e., cardipropranolol on respiratory chemoreceptors or the respiodynamic hypopnea). These changes therefore are the ratory center which affects breathing. opposite of those observed in dogs given int.ravenous Fifth, the effect of propranolol on respiratory control isoproterenol (i.e., cardiodynamic hyperpnea (16)), and is transient and only associated with the period during may therefore be mediated by the same ventilatory which heart rate is decreasing (avg of 3 min). Thus, one control process. would have to postulate that the independent action of If propranolol produced a primary reduction in VE, propranolol on respiratory control is only transient while the effect on heart rate is sustained. The half-life unrelated to cardiac output, we would also expect to find a decrease in Jko2. However, we believe that a primary of the action of propranolol is known to be several hours. reduction in VE produced by the drug is an unlikely and The study also addresses itself to the importance of an unreasonable explanation for the hypopnea observed muscle receptors which might affect VE in man. When in these studies for the following reasons. cardiac output decreases during constant load exercise, First, a primary reduction in VE would be acmmpatissue PO, decreases and Pco2 and H+ increase, while nied by an immediate increase in PETITE in excess of CO, flux to the lungs decreases. Thus, the reduction in that observed during our studies (mean increase = 0.71 cardiac output transiently uncouples tissue metabolism mmHg) (see Table 2 for individual values). The magnifrom CO, flux to the lungs. The observation that VE tude of this increase, once a steady state among producchanges in a direction opposite to that which would be tion, stores, and output has been achieved, would be expected if tissue metabolite concentrations were important stimuli for the mechanism of the exercise hyperpproportional to the magnitude of the decrease in J?E. The time required to attain such a steady state, and nea argues against the importance of postulated tissue hence the full change in PETITE, would be dependent on metabolism receptors (10) in normal respiratory control. The proportional change of VE and vco,! reinforces the the rate of change of the lung and tissue CO, stores. At rest, the calculated time constant for the lung CO, concept that receptors for respiratory control “sense” CO, flux to the lungs. stores is generally short (co.5 min). The time constants We have demonstrated that ventilation diminishes and sizes of the tissue CO, compartments vary widely depending on the tissue (3). Blood would equilibrate transiently in response to acute decreases in cardiac rapidly and has a capacitance of 4.5 ml/l per mmHg output in man and that respiratory control mechanisms function to regulate Pco2 at a constant level. These while the time constant of bone is so long that it would not be measurably involved in the exchange for the adjustments in ventilation appear to depend on receptors affected by changes in CO, flux to the lungs, rather period of hypopnea observed in these studies. Indeed, approximately 50% of the total expected steady-state than oxygen consumption or Pco2 and PO, tensions in change in PET~-~ occurs within a few breaths when a the muscle cells, per se. normal subject voluntarily hypoventilates after reachThis study was supported by National Institutes of Health Grants ing a steady state of exercise at 90 W (Fig. 6). For our experimental subjects with a mean decrease in J?E of HL-05916 and HL-11907. B. J. Whipp is an Established Investigator of the American Heart 9%, we would therefore expect at least 50% of the 9% Association. expected rise in PETITEto occur or a mean increase of 2.2 mmHg. In fact, the mean increase observed for the Received for publication 3 March 1976. Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
892
BROWN,
WASSERMAN,
AND
WHIPP
REFERENCES 1. BEAVER, W. L., K. WASSERMAN, AND B. J. WHIPP. On-line computer analysis and breath-by-breath graphical display of exercise function tests. J. Appl. Physiol. 34: 128-132, 1973. 2. CASABURI, R., B. J. WHIPP, S. N. KOYAL, AND K. WASSERMAN. Ventilatory and gas exchange dynamics in response to sinusoidal work (Abstract). Federation Proc. 34: 431, 1975. 3. CHERNIACK, N. S., AND G. S. LONGOBARDO. Oxygen and carbon dioxide gas stores of the body. Physiol. Reu. 50: 196-243, 1970. 4. CUNNINGHAM, D. J. C., E. N. HEY, J. M. PATRICK, AND B. B. LLOYD. The effect of noradrenaline infusion on the relationship between pulmonary ventilation and the alveolar Pop and PCO* in man. Ann. N.Y. Acad. Sci. 109: 756-770, 1963. 5. EPSTEIN, S. E., B. F. ROBINSON, R. L. KAHLER, AND E. BRAUNWALD. Effects of beta-adrenergic blockade on the cardiac response to maximal and submaximal exercise in man. J. CZin. Invest. 44: 1745-1753, 1965. 6. FARHI, L. E., AND H. RAHN. Gas stores of the body and the unsteady state. J. AppZ. Physiol. 7: 472-484, 1955. 7. HEISTAD, D. D., R. C. WHEELER, A. L. MARK, P. G. SCHMID, AND F. M. ABBOUD. Effects of adrenergic stimulation on ventilation in man. J. CZin. Inuest. 51: 1469-1475, 1972. 8. HLASTALA, M. P., B. WRANNE, AND C. J. LENFANT. Cyclical variations in FRC and other respiratory variables in resting man. J. AppZ. Physiol. 34: 670-676, 1973. 9. KELTZ, H., T. SAMORTIN, AND D. J. STONE. Hyperventilation: a manifestation of exogenous beta-adrenergic stimulation. Am. Rev. Respirat. Diseases 105: 637-640, 1972. 10. LEVINE, S., AND W. E. HUCKABEE. Ventilatory response to druginduced hypermetabolism. J. AppZ. Physiol. 38: 827-833, 1975.
11. SCHOLANDER, P. F. Analyzer for accurate estimating of respiratory gases in one-half cubic centimeter samples. J. BioZ. Chew 167: 235-257, 1947. 12. STONE, D. J., H. KELTZ, T. K. SARKAR, AND J. SINGZON. Ventilatory response to alpha-adrenergic stimulation and inhibition. J. AppZ. Physiol. 34: 619-623, 1973. 13. WASSERMAN, K., A. L. VAN KESSEL, AND G. G. BURTON. Interaction of physiologic mechanisms during exercise. J. AppZ. PhysioZ. 22: 71-85, 1967. 14. WASSERMAN, K., AND B. J. WHIPP. Exercise physiology in health and disease. Am. Rev. Respirat. Diseases 112: 219-249, 1975. 15. WASSERMAN, K., B. J. WHIPP, R. CASABURI, D. J. HUNTSMAN, J. CASTAGNA, AND R. LUGLIANI. Regulation of arterial Pco2 during intravenous CO, loading. J. AppZ. PhysioZ. 38: 651-656, 1975. 16. WASSERMAN, K., B. J. WHIPP, AND J. CASTAGNA. Cardiodynamic hyperpnea; hyperpnea secondary to cardiac output increase. J. AppZ. PhysioZ. 36: 457-464, 1974. 17. WASSERMAN, K., B. J. WHIPP, S. N. KOYAL, AND W. L. BEAVER. Anaerobic threshold and respiratory gas exchange during exercise. J. AppZ. Physiol. 35: 236-243, 1973. 18. WHELAN, R. F., and I. M. YOUNG. The effect of adrenaline and noradrenaline infusions on respiration in man. Brit. J. PharmacoZ. 8: 98-102, 1953. 19. WHIPP, B. J., R. CASABURI, S. N. KOYAL, AND K. WASSERMAN. Regulation of arterial Pco2 during sinusoidal exercise (Abstract). Federation Proc. 34: 431, 1975. 20. YAMOMOTO, W. S., AND M. W. EDWARDS, JR. Homeostasis of carbon dioxide during intravenous infusion of carbon dioxide. J. AppZ. Physiol. 15: 807-818, 1960.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (132.210.236.020) on January 11, 2019.
