BREAKING POINT OF BREATH HOLDING AND TOLERANCE TIME IN REBREATHING*

The Japanese Journal of Physiology 17, pp.43-56, 1967 BREAKING POINT OF BREATH HOLDING AND TOLERANCE TIME IN REBREATHING* Syoiti KOBAYASI AND Chieko SASAKI The 2nd Department of Physiology, Niigata University School of Medicine Many studies have stated that the break of breath holding is brought on when hypercapnic and hypoxic stimulations that urge respiration overcome voluntary efforts to resist them, and several diagrams have been published to indicate the relationship between alveolar tensions of oxygen and carbon dioxide at the breaking point1,6,12,17). This relationship has been described to be modified by neural factors, such as pulmonary vagal afferents3,15) and proprioceptive a fferents from respiratory muscles7,18). The Po2-Pco2 relationship at the breaking point markedly varies if the breath is held with varied concentrations of O2, N2 and CO2 in the inspired gas, even though the initial neural conditions, such as body posture and lung volume, are kept unchanged. From this finding KOBAYASI and TAKAGI14) and TAKAGI and KOBAYASI19) derived the importance of the time factor in CO2 stimulation, which concerns the determination of the breaking point. On the other hand, there have been reports which emphasize a close relationship between respiratory distress sensation and movements of the thorax and lungs. One example is the fact reported by FOWLER7) that if a subject, after a voluntary breath holding of maximal duration, rebreathes several times a gas mixture which does not improve the alveolar gas composition, a sufficient relief from distress is afforded, so that he can hold his breath for two additional periods despite a progressive decrease in the alveolar O2 tension and a progressive increase in the alveolar CO2 tension. Another example is the observation that rebreathing can be continued longer than breath holding and alveolar gas tensions at the breaking point are lower for O2 and higher for CO2 than those in breath holding10,16). It is the purpose of this paper to present characteristics of the time factor in CO2 stimulation on respiration and the effects of respiratory muscle activities and thorax movements on the termination of breath holding and rebreathing. Received for publication April 25, 1966

S. KOBAYASI AND C. SASAKI Fourteen men and three women, aged 18-40 years, acted as subjects. All of them were either laboratory workers or medical students who were untrained for breath holding and rebreathing experiments. In breath holding experiments, the subject sat on a straight chair, breathing room air through a mouthpiece on one arm of a T-way stopcock. After a full expiration, the stopcock was turned to connect the subject to a spirometer containing pure O2 or 2-CO2 mixture. He inspired maximally and the volume was read. The breath was O then held as long as possible, and at the breaking point he fully exhaled into an open rubber tube, 1 m in length and 2. 5 cm in inside diameter. An end-expiratory alveolar gas sample was thus immediately taken from the tube and volumetrically analysed. Occasionally an infrared-co2-analyser was inserted between the mouthpiece and the stopcock. The time course of changes in PA,CO2 during breath holding and PA,CO2 at the onset of diaphragm activity were estimated with the theoretical formula of KOBA- YASI13) and were occasionally checked by repeating breath holdings of predetermined durations varied by 10 or 20 sec steps. In rebreathing experiments, the subject, sitting in a chair, was connected through a mouthpiece and an infrared-co2-analyser to an arm of a fourway stopcock. Two other arms were connected to rubber bags, and the remaining was opened to room air. After a control period of room air breathing, the subject was instructed to make a full expiration, and the stopcock was immediately turned to connect him to a rubber bag containing an O2 or O2-CO2 mixture of the volume not less than his vital capacity. Then he inspired maximally, and the stopcock was turned to open to an empty rubber bag of 0.1-0.51 in capacity. The subject thus rebreathed with a confined volume and at a given rate until he could no longer continue it. This uncontrollable point was called" the breaking point of rebreathing." Rebreathing without a tidal volume limitation was also performed in some cases. The CO2 concentration in the rebreathing air was continuously recorded by the infrared-co2-analyser and occasionally checked by a Scholander apparatus. Breath holding as well as rebreathing was started after a full inspiration of a high concentration of O2 for the purpose of keeping the initial lung volume constant and for the elimination of anoxic effects, but neither was done with prior hyperventilation nor with prior oxygen breathing. Ventilatory movement was simply recorded by an electric perimetrical pneumograph, usually at the epigastric level. Intensity-duration relationship of CO2-stimulation for the breaking of, and onset of diaphragm activity during breath holding: When the breath was held, the thoracogram showed at gradual decrease in thoracic volume, and after a while contractions of the diaphragm appeared, which progressively increased in intensity and rate up to the breaking point (FIGS. 1 and 7). From the onset of diaphragm activity up to the breaking point, the air flow to and from the outer environment was sufficiently impeded by intensive voluntary closure of the glottis. When the breath was held with O2, the period from the beginning up to the onset of rhythmic diaphragm activity, which was called by AGOSTONID

BREATH-HOLDING FIG.1. Pneumogram wards. Slow slow unaware difficult observed "the to AND during breath fluctuations at the thoraco-abdominal hold the breath. in untrained first part of breath early part restlessness Such 45 REBREATHING an holding. Inspiration of breath seeking unaware down- holding indicate a position least movement is often subjects. holding", was 80 sec in mean, and the PA,co, at the onset of diaphragm activity was estimated at 48.8 mmhg in average by means of the equation of KOBAYASI13). The rhythmic diaphragm activity appeared earlier and the PA,co, at its onset point mostly higher, if the breath was held with a higher concentration of CO2 in O2 (TABLE 1). The linear Q-t relationship and the hyperbolic P-t relationship in respiratory stimulation of CO2, at the breaking point of breath holding, which were reported by KOBAYASI and TAKAGI14) and TAKAGI and KOBAYAsi19),seem to hold true also at the onset of diaphragm activity. The relationships are expressed as Q=H+St and P=(H/t)+S. Here S corresponds to the rheobase in electrical stimulation FIG.2. P-t point and is called a" basic tolerable curves and and for the basic onset tolerable CO2 levels of diaphragm CO2 level", for activity. breaking and H, a

S.KOBAYASI AND C.SASAKI

BREATH-HOLDING AND REBIZEATHING hyperbolic component, represents a time factor, and t corresponds to the utilization time. In TABLE 1, values of S and H obtained from the present experiments are presented. The basic tolerable CO2 level (S) for the onset of diaphragm activity was 30.8 mmhg, and the time factor (H) 805 mmhg Esec. The former was slightly less and the latter markedly less than those for the breaking point, which were 35.9 mmhg and 1908 mmhg Esec, respectively. The mean P-t curves for the breaking point and for the onset of diaphragm activity computed from the data in TABLE 1 are illustrated in FIG. 2. The period up to the onset of diaphragm activity increased with repetitive daily trials, but the increase was far less marked and less stable as compared with the increase in breath holding time; e. g., in subject RK breath holding time increased from 80 sec to 181 sec in 8 days, while the time up to the onset of diaphragm activity increased only from 46 sec to 65 sec in the same period. The tolerance time in rebreathing: The rebreathing was performed with or without a tidal volume limitation and rate definition. In cases with the limitations, after a single maximal inspiration of O2 or air following a full expiration, the subject was connected to an empty rubber bag of 0.1-0.51 in capacity. Then he was instructed to start rebreathing, adjusting his tidal volume voluntarily, not forcibly by a limited bag volume, to a definite volume of the rebreathing bag and to keep his breathing rate which was beaten by a metronome, at the rate of his prior normal breathing (mostly 15-20/min). The tolerance time in such restricted rebreathing was longer than breath holding time, both in cases with oxygen and with air. A substantial increase in tolerance time appeared in rebreathing with as small a tidal volume as 0.11, which was far less than bodily and instrumental dead space volume although this does not mean an absence of alveolar ventilation (FIG. 3). Rebreathing time further increased, if the rebreathing rate was not forced at a definite rate and/or the tidal volume was not confined (FIG. 3). Since all rebreathing and breath holding experiments were performed with the volume derived from a preceding single full inspiration, the initial gas volume held in the closed lung-bag system or in the lungs was the same among all rebreathing and breath holding experiments in the same subject. The rising rate of PA,CO2, therefore, showed little difference between breath holding and rebreathing in the same subject (FIG. 4). In the period from the onset of diaphragm activity up to the breaking point-the second part of breath holding in AGOSTONI'S classification-movements of the thorax and the lungs are being involuntarily performed, even though the air flow is voluntarily completely obstructed. Accordingly, the second part of breath holding may be considered to be a type of rebreathing with an extremely small tidal movement. In FIG. 5, the relationship between

S. KOBAYASI AND C. SASAKI FIG. 3. Tolerance time of breath holding and rebreathing with restricted tidal volume. Solid columns represent breath holding times, white bars indicate the onset point of diaphragm activity. Open columns represent tolerance times of rebreathing with the restricted tidal volume indicated in each column, and at a difinite rate. Hatched column represents rebreathing time of free tidal volume and free rate. Columns enclosed with broken line indicate times of rebreathing without rate specification, but with volume restriction. FIG. 4. Time course of FA,CO2 increase during rebreathing, simple breath holding, and successive breath holding interrupted by 3 rebreathing acts (Subj. RK).

BREATH-HOLDING AND REBREATHING the tolerance time in rebreathing and the rebreathing tidal volume is illustrated as a percentage of the period of the first part of breath holding. In this figure the breath holding is tentatively regarded as a rebreathing with a tidal movement of less than 0.11. This diagram shows that the tolerance time increases with increased tidal movement, but at a diminishing rate. FIG. 5. Increase in tolerance time with increase in tidal movement. DA: Period before the onset of diaphragm activity. BH: Breath holding time. Successive breath holding interrupted by a deep rebreathing: In the breath holding experiments, if at the breaking point the subject exhaled his first breath into an empty bag and immediately re-inhaled his own expired air from the bag, he experienced a great relief from distress and came to be able to hold his breath for a further period. The same maneuver at the breaking point of the second holding made the third holding possible, and so forth. This reveals that only one performance of the respiratory movement is sufficient for a transient relief from distress and the immediate resumption of breath holding. The holding time of each successive breath holding, however, decreased successively in a nearly exponential manner (FIG. 6), and it could not be extended, even if three rebreathings instead of single rebreathing were done at each breaking point. The A,m, increased progressively during successive holdings (FIG. 7). The infrared-capnogram in FIG. 7 shows that despite intensive diaphragm activities, the air flow is completely obstructed by an active closure of the glottis. The time course of the alveolar CO2, increase was almost the same among simple breath holding, successive breath holding, and rebreathing with various tidal volumes (FIG. 4). It should be noticed here that the resumed breath holdings, after the interposition of a single rebreathing, is devoid of the first part of breath holding1)-the period of absence of rhythmic respiratory activity (FIG. 7).

S. KOBAYASI AND C. SASAKI FIG. 6. Relative holding time of successive breath holding interrupted by a single deep rebreathing. Roman numerals on abscissa indicate the order of successive breath holdings. FIG. 7. Infrared capnogram (upper one) and pneumogram (lower one, inspiration downwards). Numerals on CO2 tracing indicate the breaking points of successive breath holdings. Despite intensive respiratory activity, no evidence of outflow of air is recognized in the infrared-capnogram. Several irregular deflections of the pneumogram at the early part of breath holding are artefacts due to involuntary postural movements. CRAIG and CAIN5) proposed an index of stimulus intensity at the breaking point of breath holding in consideration of the time factor. It was evaluated in terms of a pre-hold ventilation minute volume, multiplied by the holding

BREATH-HOLDING AND REBREATHING 51 time and divided by the lung volume during the holding. Regarding the time factor, KOBAYASI and TAKAGI14) and TAKAGI and KOBAYASI19) have arrived at a conception for determining the quantity of CO, stimulation which affects during breath holding. This quantity can be estimated by an integration of their theoretical equation, which gives the time course of the change in the alveolar CO, tension during breath holding from the beginning up to the breaking point. The relationship between the quantity of CO, stimulation (Q) thus estimated and the holding time (t) is fairly linear and the relationship between the CO, tension (P) and the holding time is represented by a rectangular hyperbola whose equation is P=(H/t)+S. The S corresponds to the rheobase and is called a " basic tolerable CO, level The basic tolerable CO, level shows only a small individual difference and remains fairly unchanged while the holding time increases with training14). It was estimated at 39.1 mmhg in [mean for the breaking point19). The values obtained in the present series were 35.9 mmhg for the breaking point and 30.8 mmhg for the onset of diaphragm activity. The basic tolerable CO, level for the breaking point nearly coincide with the normal resting This suggests that the basic tolerable CO, level may relate to the resting alveolar or arterial PCO2, before the breath holding, which depends on the chemosensibility of the respiratory central structures. So it may be surmised that the basic tolerable level might be shifted due to changes in the resting P resting,co2, PA caused by altered metabolic conditions, such as muscular exercises, thermal stresses, high altitude, and so on. This supposition, however, requires further investigation. The increase in holding time due to repetitive daily trials was not accompanied by either a significant change of the time course of the CO2 accumulation9) or by a change of the basic tolerable CO2 level, but by an increase in the time factor (H) in the P-t relationship. This suggests that the training effect on holding time may have resulted less from physical adaptations than from an increase in psychological tolerance or will power. If so, the hyperbolic component may be in greater part related to the will power or the intensity of voluntary central inhibitions. The difference in holding time among subjects may also be dependent largely on the individual difference of will power. GAENSLER, RAYL and DONNELLY8) reported that breath holding time could be increased by informing the sub ject of the passage of time during the holding and furthermore by a deceptive call on time, e. g., calling "5 seconds " every 6 or 7 seconds. This finding clearly shows the important role of the psychological factor in the expiration of breath holding. The hyperbolic intensity-duration relationship has been also reported in ether anaesthesia11) and in heat tolerance). It is of interest that an i-t relationship similar to that seen in the electrical stimulation of peripheral exciting tissues was also recognized in the responses of more composite integral

52 S. KOBAYASI AND C. SASAKI functions from chemical and thermal stimulations. AGOSTONI1) proposed that the period of breath holding should be divided by the onset of involuntary diaphragm activity into two parts : the first, characterized by voluntary inhibition of the respiratory activity, and the second, by involuntary respiratory effort impeded by the voluntary closure of glottis. In the present experiments, the period of the first part of the breath holding, after a single maximum inspiration of O2, was 80 sec in mean (TABLE 1), while AGOSTONI1) obtained a mean value of 42.5 sec with breath holding at a resting pulmonary volume after O2 breathing. The difference of the time may be attributed to the difference of the initial pulmonary volume. The mean value of PA,co, at the point of the onset of diaphragm activity was estimated at 48.8 mmhg (TABLE 1), which coincided with the result of AGOSTONI (48.3 mmhg). Concerning the onset of diaphragm activity also, the Q-t relationship was linear and the P-t relationship hyperbolic, but the values of H and S were lower than those at the breaking point (TABLE 1, FIG. 2). As we have already discussed, the basic tolerable CO2 level is dependent mostly on the chemosensibility of the respiratory central structures, while the hyperbolic component-time factor-is in a greater part related to voluntary effort. In consequence, the smaller time factor for the onset of diaphragm activity suggests that the onset point may be determined largely by physical factors and less by mental or psychological factors. A smaller increase in the period up to the onset of diaphragm activity by training also suggests a minimum of mental factors in the onset of diaphragm activity. This agrees with the opinion of AGOSTONI1) that the onset of diaphragm activity during breath holding is mainly attributed to chemical stimulation of the respiratory central structures, and little to such neurogenic factors as those related to lung volume and respiratory movements. He further suggested the difference in nature and in control level between the onset of diaphragm activity and the breaking point. AGOSTONII) stated that the PA.co, necessary to release diaphragm activity should be represented by the value of 9 sec before its onset because the mean circulation time from the lung to the brain is about 9 sec. This correction, however, did not cause any essential change in the Q-t or P-t relationship in our experiments. At present, no further comment on this problem seems worthwhile because of the lack of knowledge of the kinetics and of the site of CO2 stimulation on ventilation. In the second part of breath holding-the period after the onset of diaphragm activity-ventilatory movement should be considered to be not only impeded mechanically by a voluntary closure of the glottis, but also suppressed neurally by the voluntary inhibition of the activity of the respiratory center, though the inhibition is incomplete. In rebreathing with a limited tidal volume and at a definite rate, similar voluntary partial inhibition on respira-

BREATH-HOLDING AND REBREATHING 53 tory activity is needed. The subject has to ad just his ventilation voluntarily to a definite tidal volume and rate (0.1-0.5 1 and 15-20/min), which are less than his proper demand. Consequently, in such cases the tolerance time is shorter than that in rebreathing without limitation of depth and rate, when no voluntary suppression is needed at all (FIG. 3). But, the tolerance time of rebreathing, either with or without depth and rate confinement, is always longer than the breath holding time (FIG. 3). It is of much interest that a substantial increase already appeared in rebreathing with such a small tidal volume as 0.1 1, which was far less than the total bodily and instrumental dead space (FIGS. 3 and 5). One or three deep rebreathings at the breaking point of breath holding greatly relieved the distress and made a successive holding possible, though the rebreathing did not improve the alveolar gas composition (FIGS. 6 and 7). But, in the successive breath holdings the period of absence of rhythmic diaphragm activity was hardly recognized (FIG. 7). These findings suggest that a single rebreathing at the breaking point can hardly prohibit the development of forcible involuntary activity of the respiratory muscles, though it definitely relieves the distress sensation. As the forcible respiratory activity depends mostly on chemical events, it can hardly be eliminated merely by mental relief, but must wait until increased chemical stimulation is removed by a renewal of alveolar air. FOWLER7) stated that when a sub ject rebreathed air which had been expired into a small bag, and the volume of the expired air was not sufficient for the succeeding inspiration, it was intolerable and the subject could not proceed to the second breath holding. Our experimental conditions for the successive breath holding were just the same as that of FOWLER, but our subjects could easily proceed with the succeeding breath holding. The reason for this discrepancy is still obscure. The time course of the alveolar CO, increase was almost the same among the simple breath holding, successive breath holding, and rebreathing (FIG. 4). It follows, then that the difference in tolerance time among these three procedures can not be accounted for by chemical events, but by mental and neural factors3,4,7,15,18). Concerning factors responsible for the respiratory distress sensation, NHzimA16) and FOWLER7) stated that unpleasant sensations during breath holding are associated with the increased voluntary effort required to inhibit and oppose the respiratory activity stimulated by hypercapnia. The voluntary effort referred to here, should include the effort for an active closure of the glottis to impede an outflow of air due to an increased intrapulmonary pressure, as well as for neural inhibition of the respiratory centre. Furthermore, for the development of distress, this effort should be continued for some time, because a short interruption of the effort, such as an interposition of a single rebreathing, causes a great relief from distress, A reversed view was pro-

54 S. KOBAYASI AND C. SASAKI posed by WRIGHT and BRANSCOMB20) that the breathlessness may be caused by undue intensity and prolongation of the discharges of the medullary respiratory neurons. A continuous excitation of the inspiratory centre ultimately showers the centres of consciousness with impulses that cause dyspnoea. FOWLER7) suggested that during breath holding the afferent barrage from various muscles is probably large in intensity and disorganized, and the resumption of respiratory movement would reorganize an appropriate relationship between central activity and peripheral muscular activity. CAMPBELL and HOWELL4) stated that the imbalance between the demand for, and the effort of, breathing, or the disturbance of the normal relationship between information about muscular length and information about muscular tension in the control system should be ultimately responsible for the sensation of difficulty. Each of these neural mechanisms may be related to the sensation of respiratory distress, and possibly to the time factor in the tolerance time, but none of them offer a sufficient unitary explanation of all types of dyspnoea. SUMMARY The physical and mental factors which concern the determination of the breaking point of breath holding and of rebreathing were investigated in normal human subjects. The hyperbolic intensity-duration relationship of ventilatory stimulation of CO2, which has been revealed about the breaking point of breath holding, applies also in the onset of diaphragm activity during breath holding. The basic tolerable CO, level, which corresponds to the rheobase in electrical stimulation, was almost the same as, or a little lower than, the normal resting P A,CO, - level for the breaking point as well as for the onset of diaphragm activity, and showed few individual variations and training effects. It is considered that in the intensity-duration relationship the basic tolerable CO, level dominantly related to the chemosensibility of the respiratory central structures, and the time factor may be in larger part related to mental or psychological events. The onset of diaphragm activity during breath holding is mainly attributed to physical factors. The tolerance time of rebreathing with a restricted tidal volume and at a definite rate was longer than breath holding time, even when the tidal volume was as tightly restricted as to 0.11. Single deep rebreathing at the breaking point caused a great relief from distress and made several succssive breath holding possible. In the successive breath holding, however, the period of absence of the rhythmic respiratory activity could hardly be observed. These increasing effects of respiratory movement upon the tolerance time

BREATH-HOLDING AND REBREATHING 55 are not ascribed to changes in physical conditions, but probably relate to some unexplained neural mechanisms. The authors are much indebted to Prof. K. TAKAGI, School of Medicine, Nagoya University, for his kind advise and criticism in the preparation of the manuscript. A brief section of this paper was read at the International Symposium on " Breath-Hold Diving and Ama of Japan ", held in Tokyo, August 31-September 1, 1965. REFERENCES 1) AGOSTONI, E. Diaphragm activity during breath holding : Factors: related to its onset. J. Appl. Physiol. 18: 471-477, 1963. 2) BELL, C. R., HELLON, R. F., HIORNS, R. W., NICOL, P. B. AND PROVINS, K. A. Safe exposure of men to severe heat. J. Appl. Physiol. 20 : 288-292, 1965. 3) CAIN, S. M. Breaking point of two breath holds separated by a single inspiration. J. Appl. Physiol. 11 : 87-90, 1957. 4) CAMPBELL, E. J. M. AND HOWELL, J. B. L. The sensation of breathlessness. Brit. Med. Bull. 19 : 36-40, 1963. 5) CRAIG, F. N. AND CAIN, S. M. Breath holding after exercise. J. Appl. Physiol. 10: 19-25, 1957. 6) DOUGLAS, C. G. AND HALDANE, J. S. The regulation of normal breathing. J. Physiol. 38 : 420-440, 1909. 7) FOWLER, W. S. Breaking point of breath-holding. J. Appl. Physiol. 6: 539-545, 1954. 8) GAENSLER, E. A., RAYL, D. F. AND DONNELLY, D. M. The breath holding test in pulmonary insufficiency. Surg. Gyn. Obstetr. 92: 81-90, 1951. 9) HASEGAWA, H. On breath holding. II. Report. Niigata Med. J. 61 : 198-202, 1947. (in Japanese) 10) HILL, L. AND FLACK, M. The effect of excess of CO2 and of want of 02 upon the respiration and the circulation. J. Physiol. 37: 77-111, 1908. 11) KOBAYASI, S. AND KINOSITA, T. Experimentelle Untersuchungen der Athernarkose und iiber die Anwendung eines einfachen Interferometers zur Bestimmung des Athergehaltes in der Atmungsluft. Anaesthesist 7: 1-6, 1958. 12) KOBAYASI, S. On breath holding. I. Report. Niigata Med. J. 61: 176-183, 1947. (in Japanese) 13) KOBAYASI, S. On breath holding. III. Report. Theoreticals. Niigata Med. J. 61 : 421-426, 1947. (in Japanese) 14) KOBAYASI, S. AND TAKAGI, K. On breath holding. IV. Report. Theoreticals. J. Physiol. Soc. Japan. 11: 142-144, 1949. (in Japanese) 15) MITHOEFER, J. C. Lung volume restriction as a ventilatory stimulus during breath holding. J. Appl. Physiol. 14: 701-705, 1959. 16) NIIZIMA, A. On breath holding. VI. Report. Comparison of the breath holding with the rebreathing. J. Physiol. Soc. Japan. 12: 121-127, 1950. (in Japanese) 17) OTIS, A. B., RAHN, H. AND FENN, W. O. Alveolar gas changes during breath holding. Am. J. Physiol. 152: 674-686, 1948. 18) RODBARD, S. The effect of oxygen, altitude and exercise on breath holding time. Am. J. Physiol. 150: 142-148, 1947. 19) TAKAGI, K. AND KOBAYASI, 5, Alveolar gas exchanges during breath holding and

56 S. KOBAYASI AND C. SASAKI considerations on the breaking point. In : RAHN, H. (Ed.) : Physiology of Breath- Hold Diving and Ama of Japan. Washington, D. C., Natl. Acad. of Sciences-Natl. Res. Council, U. S. A., 1965, pp. 147-157. 20) WRIGHT, G. W. AND BRANSCOMB, D. V. Cited from COMROE, J. H. Jr. : Physiology of Respiration. Year Book Medical Publishers Inc., Chicago, U.S. A., 1965, p. 210.