Breathing changes, such as reduction in ventilation (10,22,23,26,30), decreases in both inspiratory and expiratory flows (10,29), expiratory flow limitation and dynamic hyperinflation (10,20,30), increases in transpulmonary pressure (10,24,30), and related increases in dyspnea sensation (3,24,27,30) have been described in subjects exercising at high ambient pressure. The primary cause of these changes is the increase in respiratory resistance attributable to increased air density (9,16,23).
The few observations concerning dyspnea sensation during exercise in a hyperbaric environment are not consistent (10,23,24,26,29), and no study has been specifically devoted to the investigation of dyspnea mechanisms in these conditions. During loaded exercise, dyspnea depends to a great extent on the power developed by respiratory muscles (7) and on the degree of dynamic hyperinflation (14,19). The power is the product of length-tension and force-velocity characteristics of the respiratory muscles; the higher the tension and velocity of contraction, the greater the intensity of dyspnea (7). During loaded breathing exercise, the increase in dyspnea that is caused by increased tension (pleural pressure) is modulated by the reduced velocity of shortening (7). On the other hand, dynamic hyperinflation contributes importantly to generation of dyspnea (14) by increasing the pressure per breath and decreasing the maximal pressure-generating ability of the inspiratory muscles (7). However, end-expiratory lung volume tends to decrease during exercise in a hyperbaric environment (20). Moreover, it has recently been shown that dyspnea intensity is not different between normal subjects who hyperinflate and those who do not hyperinflate during expiratory flow-limited exercise (11).
On this basis, we hypothesized that i) hyperinflation might not play an important role in the genesis of dyspnea during exercise in hyperbaric conditions and that ii) the lower velocity of shortening of the respiratory muscles modulates the intensity of dyspnea to a lesser extent for any developed pleural pressure.
We carried out this study to verify this hypothesis by evaluating pulmonary volumes, lung mechanics, respiratory muscle function, and dyspnea sensation in a group of normal subjects exercising under hyperbaric conditions.
Five healthy untrained male nonsmoking volunteers participated in the study after giving their written informed consent. The study was approved by the local ethics committee. All subjects were trained in respiratory maneuvers (pulmonary function testing, inspiratory maneuvers during exercise, sniff maneuvers), in the use of the Borg scale, and in cardiopulmonary exercise testing. Before the study, all subjects were tested about their tolerance to the hyperbaric conditions (4 ATA for 20-30 min at rest, on at least two occasions). No fitness/training sessions were conducted under normal or hyperbaric conditions in the current study. A clinical assessment was conducted before the study, and all subjects were determined to be healthy. Anthropometric and baseline pulmonary function data are presented in Table 1.
Subjects attended the laboratory on three different days. On the first day, routine spirometry and FRC measurement were accomplished. On the second and third days, subjects performed a cycloergometer exercise at 1 (sea level, SL) and 4 ATA in random order. Except for the compression and decompression periods in the hyperbaric experiments, the general procedure and experimental protocol were the same for the two pressures. The two exercise tests were separated by a 7- to 10-d period. During the entire session, subjects were seated on the cycle. The protocol started with the measurement of pleural pressure during 10-15 sniff maneuvers, and then the subject was connected to the pneumotachograph by a mouthpiece and was asked to breath quietly for 10-12 min. At the end of this period, the subject was instructed to make a maximal inspiration to determine the inspiratory capacity (IC). The last 3 min of quiet breathing and the two IC maneuvers were recorded. Then, during the appropriate sessions, the air inside the chamber was compressed to an ambient pressure of 4 ATA. Before starting the exercise, subjects again performed the sniff maneuvers, and the quiet breathing period was followed by IC. During exercise, the workload was increased by 25 W·min-1. At the end of each minute of exercise, subjects performed an IC maneuver and were asked to rate dyspnea.
Simulated dives were carried out in a multiplace double-lock hyperbaric chamber (Sistemi Iperbarici Integrati Ltd., Rome, Italy) at four absolute atmospheres (equivalent depth of 30 m of sea water). To avoid problems caused by the large increase in ambient temperature that develops during the compression phase, fast compression of volunteers and inside observers was performed in the equilibrium lock; soon afterwards, personnel were transferred into the main compartment already under pressure and with a stable temperature value (about 27°C).
The increased tissue-nitrogen uptake in subjects performing heavy physical exercise is a recognized risk factor for decompression illness. A set of simulated dives was produced with the demo version of the Advanced Diving Planning Software (Abysmal Diving Inc., Boulder, CO), which allows for the introduction of different levels of workload into the simulation. Dive profiles for 4 ATA were planned with a total maximum bottom time of 27 min, two extra-deep stops at −18 and −12 m (1 min each), and a decompression stop of 15 min in oxygen at −3 m. At the end of the dive, the subjects who had performed the exercise were kept under observation for 1 h near the hyperbaric facility.
Symptom-limited incremental exercise tests were performed on an electrically braked cycle ergometer (Corival Honeywell) at two different ambient pressures. After a 3-min warm-up, the exercise load was increased by 25 W·min−1, with the subjects pedalling at 50-60 rpm. The choice of increasing the load by 25 W·min−1 was suggested by several considerations: i) this type of increment allowed us to stress the respiratory system with regard to respiratory muscles, thus eliciting an adequate response in terms of breathing discomfort; ii) with an increment of 25 W·min−1, the duration of the test did not exceed 10 min, allowing for maintainence of the permanence in the hyperbaric chamber for up to 27 min. A permanence in the hyperbaric chamber longer than 27 min would have required a longer decompression time. The test ended when the imposed load could no longer be sustained.
Baseline lung function was determined by a Vmax 22 device (Sensor Medics, Yorba Linda, CA). Spirometry was conducted according to the recommendations of the ATS (2). Functional residual capacity (FRC) was measured by nitrogen washout. Residual volume (RV) was computed by subtracting expiratory volume reserve from FRC, and total lung capacity was computed (TLC) by adding vital capacity to RV. Reference equations were taken from Quanjer et al. (21). Inspiratory capacity was measured during quiet breathing and at the end of each exercise step (see Protocol section); end-expiratory (EELV) and end-inspiratory (EILV) lung volumes were calculated as TLC-IC and EELV + VT, respectively.
For ventilation measurements, patients breathed through a Fleisch #3 pneumotachograph connected to a flow transducer. Volume was obtained by electrical integration of the flow signal. From the spirogram, we derived inspiratory time (TI), expiratory time (TE), total time of the respiratory cycle (TTOT), and tidal volume (VT). Mean inspiratory flow (VT/TI), duty cycle (TI/TTOT), respiratory frequency (Rf = 1/TTOT × 60), and minute ventilation (VE = VT × Rf) were also calculated.
Esophageal (Pes) pressure was measured using a conventional balloon-catheter system connected to a differential pressure transducer (Validyne Corp., Northridge, CA). Pes was used as an index of pleural pressure. From the pressure signal, we measured esophageal pressure swing (Pessw; the difference between the pressure measured at beginning of inspiration and the maximal inspiratory pressure), and ΔPes (i.e., the difference between the most negative and the less negative pleural pressure during tidal breathing). Total lung resistance was obtained using the isovolume method of Frank et al. (8). To measure the pressure developed by the inspiratory (Pm,i) and expiratory (Pm,e) muscles, V-P loops were constructed by plotting VT against Pes; the chest wall-relaxation line is from Aliverti et al. (1). Inspiratory and expiratory muscle pressures were measured as the horizontal distance along the pressure axis between the dynamic V-P loop and the relaxation curve (1).
Inspiratory muscle strength was assessed by measuring minimal (i.e., the greatest negative) inspiratory pleural pressure at FRC during sniff maneuvers (18). The patients were repeatedly encouraged to try as hard as possible. The maneuvers were repeated until three measurements with less than 5% variability had been recorded. The lowest pleural pressure value obtained was used for analysis. Mean inspiratory flow was used as an index of velocity of contraction of inspiratory muscles, and ΔPes was used as an index of the tension developed by respiratory muscles.
Only pneumotachograph and pressure transducers were placed inside the hyperbaric chamber, whereas all amplifiers and recorders were put outside. Outputs from transducers were hard wired through bulkhead fittings. Pressure and flow signals were recorded onto a personal computer by a 16-channel analog/digital board at sampling rate of 100Hz (National Instrument DAQCard 6024E).
Subjects were asked to rate the respiratory-effort sensation during cycling using a modified Borg scale (5), with 0 indicating no effort and 10 indicating the maximum tolerable level. The scale is a continuous vertical linear display associated with 10 verbal descriptors of the extent of the symptom.
An analysis of variance (ANOVA) for repeated measurements was used to test for the differences between variables at the two different ambient pressures during the exercise tests; post hoc comparisons were done by means of student t-test, applying the Bonferroni's adjustment (0.05/no. of tests) for multiple testing. The Student's t-test for paired samples was employed to test differences between exercise peak values. Pearson's correlation analysis was used to assess the significance of relationships between variables. Power and β value were calculated for all comparisons with a two-sided significance level of α = 0.05. Values of P < 0.05 were considered statistically significant. Data are presented as mean ± SD. Statistical procedures were carried out by SPSS 13.0 for Windows (SPSS inc, Chicago IL) and by Intercooled Stata 8.0 for Windows (Stata Corporation, Lakeway Drive, TX).
Compared with SL during quiet breathing at 4 ATA, forced expiratory flows decreased (P < 0.007); total lung resistance (P < 0.009), pleural pressure swings (P < 0.005), and ΔPes (P < 0.005) increased; and the minimal pleural pressure developed during a sniff maneuver did not change (Table 2).
The maximal workloads that subjects sustained at SL and 4 ATA were 205 ± 37.1 and 195 ± 27.4 W, respectively. These values represented 89.6 ± 12.3 and 85.7 ± 12.5% of predicted values (12), respectively. There were no differences between the maximal workloads at SL and 4 ATA when expressed as absolute values or as percentages of predicted values (Student's t-test: t = 1, P = NS for both comparisons).
Operational lung volumes.
Compared with SL at 4 ATA, both end-expiratory lung volume and end-inspiratory lung volume were greater during quiet breathing (Table 2). Thus, subjects started the exercise test at a higher end-expiratory lung volume, but despite a progressive decrease during the exercise, the difference within SL remained significant for the entire test (Fig. 1, left panel; two-way ANOVA for repeated measures, F = 157.56, P < 0.0000). Compared with SL during the test, end-inspiratory lung volume was significantly higher at 4 ATA (Fig. 1, left panel; two-way ANOVA for repeated measures, F = 36.80, P < 0.0000), with the difference being no longer significant at peak exercise (Table 3).
Compared with SL during the exercise at 4 ATA, ventilation (VE) increased less (two-way ANOVA for repeated measures, F = 31.07, P < 0.0000) (Fig. 1, right panel); peak ventilation and mean inspiratory flow (VT/TI) were significantly lower; and the ratio of inspiratory time to total time of respiratory cycle (TI/TTOT) was not different (Table 3). Lower values of both tidal volume and respiratory frequency contributed to lower ventilation values (Table 3).
Respiratory muscle pressure.
At peak exercise, the maximal pressure developed by the inspiratory muscles during tidal breathing (Pm,i) was similar at SL and 4 ATA (Table 4), whereas the maximal pressure developed by expiratory muscles (Pm,e) was significantly higher at increased ambient pressure (Table 4). When maximal pressures were compared at the same ventilation level, both inspiratory and expiratory muscle pressures were significantly higher at high ambient pressure. At 60 L·min−1 of VE at 4 ATA, the maximal pressure developed by the inspiratory muscles was 49% higher than at SL; for expiratory muscles, the difference was even greater, with values at 4 ATA being 145% higher than at SL (Fig. 2, right lower panel).
At peak exercise, the swing of pleural pressure showed a small but significant increase at 4 ATA compared with SL (Table 4). ΔPes was higher at high ambient pressure than at SL during the exercise test (two-way ANOVA for repeated measures, F = 16.22, P ≤ 0.0003) and at peak exercise (Table 4 and Fig. 2, left upper panel). The difference was even more striking when the values of ΔPes were compared at the same ventilation: at a ventilation level of 60 L·min−1, ΔPes was 84.8% higher at 4 ATA than at 1 ATA (Fig. 2, lower left panel).
The increase in Borg score during exercise and at peak exercise was not different at the two ambient pressures (two-way ANOVA for repeated measures, F = 3.45, P = 0.07 and Student's t-test, t = −0.8, P = NS, respectively) (Fig. 2, upper right panel).
Compared with SL at 4 ATA, the slope of the ΔBorg/ΔVE relationship was significantly higher (Table 5 and Fig. 3, left panel), whereas the slope of the ΔBorg/Δ(ΔPes) relationship was similar, with its intercept being significantly lower (Table 5 and Fig. 3, middle panel). The slopes of the ΔVE/Δ(ΔPes) and ΔVT/TI/ΔPessw relationships were significantly lower at 4 ATA compared with at SL, with no significant difference in the intercept (Table 5 and Fig 3, right panel).
Finally, the differences in end-expiratory lung volume between SL and 4 ATA did not relate to the concurrent differences in Borg score throughout the test.
The novel findings of this study during exercise in hyperbaric conditions are that 1) the increase in dyspnea attributable to the greater tension developed by the respiratory muscles is modulated by the decrease in their velocity of shortening, and 2) hyperinflation does not play a major role in determining dyspnea intensity.
Only a few reports concerning dyspnea perception during exercise in the hyperbaric environment are available in the literature, and none of them have examined in detail the possible mechanisms of symptoms in these conditions. Subjects exercising under hyperbaric conditions feel intense dyspnea, requiring premature termination of exercise only when working at the lowest depths (10,23,24,30), whereas at more shallow depths, dyspnea sensation is commonly not different from normobaric conditions (10,23,26,30). Based on the changes in respiratory function observed in hyperbaric conditions (10,20,22-24,26,29,30), the following factors are likely to be involved in determining the intensity of dyspnea: i) increase in respiratory muscle effort, ii) decrease in respiratory flow, and iii) dynamic hyperinflation.
Increase in respiratory muscle effort.
The increase in respiratory muscle effort-that is, the breathing pressure developed by the respiratory muscles in proportion to the total pressure available-is the main mechanism generating dyspnea during exercise in normal subjects (7). The higher pleural pressures developed by our subjects both at rest and while exercising indicated that respiratory muscle effort was increased at increased ambient pressure. This observation is in line with the results of previous studies showing that in conditions of hyperbarism, transpulmonary pressure increases considerably because of the increased resistive load determined by gas density (10,24,30). This is evident from Figure 2, which shows that higher values of pleural pressure were reached at high ambient pressure.
Compared with control conditions, both inspiratory and expiratory muscle pressures increased at high ambient pressure, but the increase in expiratory muscle pressure was twice the increase in inspiratory muscle pressure, suggesting a major role of expiratory load in the increased ΔPes. Our data showing greater drive on the expiratory muscles than on the inspiratory muscles are in line with the results of a previous study (13), indicating the major role played by the expiratory muscles in increasing dyspnea sensation in exercising flow-limited healthy humans.
Decrease in expiratory flow.
We found that the respiratory flow decreased significantly in hyperbarism and that, for any given ΔPes, the flow was significantly lower, indicating a reduction in the velocity of shortening of the respiratory muscles. The role of velocity of shortening of the respiratory muscles in the genesis of dyspnea sensation has been investigated by El-Manshawi et al (7). These authors showed that during exercise with respiratory resistive loading, the intensity of breathlessness was independently predicted by pleural pressure, inspiratory flow, duty cycle, and respiratory frequency. In our patients, pleural pressure greatly increased, whereas inspiratory flow and respiratory frequency decreased, and duty cycle remained unchanged. In turn, hyperbarism shifted the dyspnea score down for any given ΔPes (Fig. 3), indicating that dyspnea was lower than expected on a pressure basis.
Expiratory flow limitation and dynamic hyperinflation.
Under hyperbaric conditions, subjects showed a higher end-expiratory lung volume during quiet breathing, which is consistent with the results of O'Kroy et al. (20). Although the reasons for this are likely to be complex, O'Kroy et al. (20) hypothesized a new equilibrium between recoil of the respiratory system and airway resistance induced by the increased air density and airway resistance, or, alternatively a behavioral effect of the pressurized chamber (20). Despite the fact that during exercise, the high ambient pressure shifted end-expiratory lung-volume level up at any given workload (a finding also reported in previous studies (10,20,30), end-expiratory lung volume decreased progressively at both SL and 4 ATA (Fig. 1). The differences in end-expiratory lung volume between SL and 4 ATA did not relate with the concurrent differences in Borg score throughout the test. This finding is in agreement with the observation of Iandelli et al. (11) that in normal subjects exercising with expiratory flow limitation, dyspnea intensity was not different between subjects who hyperinflated and subjects who did not hyperflate. This suggests that the role of dynamic hyperinflation on dyspnea may have been overemphasized, regardless of how the increase in airway resistance is being applied, either by increasing air density, as in the present study, or by applying an external flow limitation, as previously found (11). In turn, hyperinflation is not likely to be a determinant factor for dyspnea generation in normal subjects exercising in conditions of airflow limitation.
Arterial blood gases.
At high ambient pressure, partial pressure of oxygen increases proportionally to the increase in ambient pressure, and it is well known that breathing hyperoxic mixtures during exercise can decrease dyspnea (6,17,25,28). In the circumstances of the present study, however, it was not possible to evaluate the role of hyperoxia on dyspnea.
An increased arterial partial pressure of carbon dioxide has been reported during exercise with both external (13) or internal flow limitation (23,30), and hypercapnia has been shown to increase dyspnea per se (4,17). For technical reasons, end-tidal partial pressure of carbon dioxide was not evaluated in our subjects. However, Kayser et al. (13) have shown that the contribution of arterial partial pressure of carbon dioxide to dyspnea is negligible during flow-limited exercise.
In conclusion, the present data confirm our hypothesis that respiratory muscle effort and decreased velocity of shortening of respiratory muscles are important determinants of dyspnea intensity in subjects exercising under hyperbaric conditions. The increased respiratory muscle pressure, necessary to overcome the high respiratory load, caused an increase in dyspnea that was modulated by the lower velocity of shortening of the respiratory muscles. Even if dyspnea was higher for a given ventilation, it was lower for a given pleural pressure. Further investigation is needed to determine whether at very low depths, when the tension required to overcome the respiratory load is very high, the reduction in velocity of shortening of respiratory muscles is able to modulate severe breathing discomfort.
The research was supported by grants from the University of Firenze. The results of the present study do not constitute endorsement of the product by the authors or ACSM.
1. Aliverti, A., S. J. Cala, R. Duranti, et al. Human respiratory muscle actions and control during exercise. J. Appl. Physiol.
2. American Thoracic Society. Lung function testing: selection of reference values and interpretative strategies. Am. Rev. Respir. Dis.
3. Anthonisen, N. R., G. Utz, M. H. Knyger, and J. S. Urbanetti. Exercise tolerance at 4 and 6 ATA. Undersea Biomed. Res.
4. Banzett, R. B., R. W. Lansing, M. B. Reid, L. Adams, and R. Brown. 'Air hunger' arising from increased PCO2 in mechanically ventilated quadriplegics. Respir. Physiol.
5. Borg, G. A. V. Psychophysical basis of perceived exertion. Med. Sci. Sport Exer.
6. Chronos, N., L. Adams, and A. Guz. Effect of hyperoxia and hypoxia on exercise-induced breathlessness in normal subjects. Clin. Sci.
7. El-Manshawi, A., K. J. Killian, E. Summers, and N. L. Jones. Breathlessness during exercise with and without resistive loading. J. Appl. Physiol.
8. Frank, N. R., J. Mead, and B. G. Ferris Jr. The mechanical behaviour of the lungs in healthy elderly persons. J. Clin. Invest.
9. Glauser, S. C., E. M. Glauser, and B. F. Rusy. Gas density and the work of breathing. Respir. Physiol.
10. Hesser, C. M., D. Linnarsson, and L. Fagraeus. Pulmonary mechanisms and work of breathing at maximal ventilation and raised air pressure. J. Appl. Physiol.
11. Iandelli, I., A. Aliverti, B. Kayser, et al. Determinants of exercise performance in normal men with externally imposed expiratory flow limitation. J. Appl. Physiol.
12. Jones, N. L., L. Makrides, C. Hitchcock, T. Chypchar, and N. McCartney. Normal standards for an incremental progressive cycle ergometer test. Am. Rev. Respir. Dis.
13. Kayser, B., P. Sliwinski, S. Yan, M. Tobiasz, and P. T. Macklem. Respiratory effort sensation during exercise with induced expiratory-flow limitation in healthy humans. J. Appl. Physiol.
14. Killian, K. J., and J. M. Campbell. Dyspnea. In: The Thorax, Part B: Applied Physiology
, C. Roussos (Ed.). New York: Marcel Dekker, Inc., 1995, pp. 1709-1747.
15. Killian, K. J., S. C. Gandevia, E. Summers, and E. J. Campbell. Effect of increased lung volume on perception of breathlessness, effort, and tension. J. Appl. Physiol.
16. Maio, D. A., and L. E. Farhi. Effect of gas density on mechanics of breathing. J. Appl. Physiol.
17. Manning, H. L., and R. Schwartzstein. Pathophysiology of dyspnea. N. Engl. J. Med.
18. Miller, J. M., J. Moxham, and M. Green. The maximal sniff in the assessment of diaphragm function in man. Clin. Sci.
19. O'Donnell, D. E., J. C. Bertley, L. K. Chau, and K. A. Webb. Qualitative aspects of exertional breathlessness in chronic airflow limitation: pathophysiologic mechanisms. Am. J. Respir. Crit. Care Med.
20. O'Kroy, J. A., J. M. Lawler, J. Stone, and T. G. Babb. Airflow limitation and control of end-expiratory lung volume during exercise. Respir. Physiol.
21. Quanjer, Ph. H., G. J. Tammeling, J. E. Cotes, O. F. Pedersen, R. Peslin, and J.-C. Yernault. Lung volumes and forced ventilatory flows. Standardization of lung function tests. European community for steel and coal. Eur. Respir. J.
6(Suppl. 16):5-40, 1993.
22. Saltzman, H. A., J. V. Salzano, G. D. Benkarn, and J. A. Kylstra. Effects of pressure on ventilation and gas exchange in man. J. Appl. Physiol.
23. Salzano, J. V., E. M. Camporesi, B. W. Stolp, and R. E. Moon. Physiological responses to exercise at 47 and 66 ATA. J. Appl. Physiol.
24. Spaur, W. H., L. W. Raymond, M. M. Knott, et al. Dyspnea in divers at 49.5 ATA: mechanical, not chemical in origin. Undersea Biomed. Res.
25. Swinburn, C. R., J. M. Wakefield, and P. W. Jones. Relationship between ventilation and breathlessness during exercise in chronic obstructive airways disease is not altered by prevention of hypoxaemia. Clin. Sci.
26. Tetzlaff, K., B. Neubauer, C. Buslaps, B. Rummel, and E. Bettinghausen. Respiratory responses to exercise in divers at 0.4 MPa ambient air pressure. Int. Arch. Occup. Environ. Health
27. Van Liew, H. D. Mechanical and physical factors in lung function during work in dense environments. Undersea Biomed. Res.
28. Ward, S., and B. J. Whipp. Effects of peripheral and central chemoreflex activation on the isopneic rating of breathing in exercise humans. J. Physiol.
29. Wood, L. D. H., and A. C. Bryan. Effect of increased ambient pressure on flow-volume curve of the lung. J. Appl. Physiol.
30. Wood, L. D. H., and A. C. Bryan. Exercise ventilatory mechanics at increased ambient pressure. J. Appl. Physiol.