Lack of causal link between muscle [H⁺] and ventilation during exercise

The following is the abstract of the article discussed in the subsequent letter:

	ABSTRACT

Oelberg, David A., Allison B. Evans, Mirko I. Hrovat, Paul P. Pappagianopoulos, Samuel Patz, and David M. Systrom. Skeletal muscle chemoreflex and pH_i in exercise ventilatory control. J. Appl. Physiol. 84(2): 676-682, 1998.To determine whether skeletal muscle hydrogen ion mediates ventilatory drive in humans during exercise, 12 healthy subjects performed three bouts of isotonic submaximal quadriceps exercise on each of 2 days in a 1.5-T magnet for ³¹P-magnetic resonance spectroscopy (³¹P-MRS). Bilateral lower extremity positive pressure cuffs were inflated to 45 Torr during exercise (BLPP_ex) or recovery (BLPP_rec) in a randomized order to accentuate a muscle chemoreflex. Simultaneous measurements were made of breath-by-breath expired gases and minute ventilation, arterialized venous blood, and by ³¹P-MRS of the vastus medialis, acquired from the average of 12 radio-frequency pulses at a repetition time of 2.5 s. With BLPP_ex, end-exercise minute ventilation was higher (53.3 ± 3.8 vs. 37.3 ± 2.2 l/min; P < 0.0001), arterialized PCO₂ lower (33 ± 1 vs. 36 ± 1 Torr; P = 0.0009), and quadriceps intracellular pH (pH_i) more acid (6.44 ± 0.07 vs. 6.62 ± 0.07; P = 0.004), compared with BLPP_rec. Blood lactate was modestly increased with BLPP_ex but without a change in arterialized pH. For each subject, pH_i was linearly related to minute ventilation during exercise but not to arterialized pH. These data suggest that skeletal muscle hydrogen ion contributes to the exercise ventilatory response.

	LETTER

Lack of causal link between muscle [H⁺] and ventilation during exercise

To the Editor: We have read with a lot of interest the paper by Oelberg et al. (9) in which the ventilatory effects of inflating proximal thigh cuffs to 45 Torr during isotonic quadriceps exercise were analyzed in 12 healthy subjects. In the last sentence of this paper, the authors conclude: "Our data suggest, however, that exercise hyperventilation is induced by a chemoreflex stimulated by muscle acidosis." Because such a conclusion was solely based on the contention that ventilation was linearly related to the reduction in the quadriceps intracellular pH (pH_i), we decided to replot the available data displayed in Figs. 1 and 3 in Ref. 9 to gain an insight into the minute ventilation (E) vs. pH_i relationship. Consequently, we obtained Fig. 1 which presents a dramatic hysteresis when both exercise and recovery are taken into account. In keeping with conclusions by Oelberg et al., it is difficult to explain that at the cessation of exercise ventilation decreased by 35 l/min in 5 min (from 55 to 20 l/min) without any change in local pH (which remained at the same level as at the E peak), whereas after 5 min pH_i increased with trivial E changes. Therefore, the dissociation between E and pH_i is quite evident. Incidentally, the authors logically dismissed the role of blood lactate in ventilatory stimulation on the basis of a nearly analogous dissociation (see Ref. 9, p. 679, right-hand column). Additionally, we have superimposed in Fig. 1 the mean ventilatory values at rest and at the peak of exercise, computed from all the individual regression lines given by Oelberg et al. in their Table 1. Clearly, the mean levels of ventilation obtained from the regression analysis data are much higher than mean E values reported in Fig. 1 in Ref. 9 (e.g., resting E = 25 l/min vs. 9 l/min), suggesting a computational error. Finally, we have attempted to take a closer look at the E vs. intracellular H⁺ concentration ([H⁺]_i) relationship during the contractions. Figure 2A reveals that the difference in E between unobstructed exercise and partial occlusion is actually independent of [H⁺]_i. Indeed, if pH_i was the factor stimulating ventilation during the partial vascular obstruction, one should expect occluded E data to be, at best, on the same iso [H⁺]_i line as the control; i.e., when pH_i decreased "spontaneously." Obviously, they are not: for a given [H⁺]_i, E is much higher during occlusion than during control. Similar conclusions may be drawn from the example shown by Oelberg et al. in their Fig. 4 (top) where all the E occlusion values are higher than those obtained during control at any given pH_i. Any correlation between E and pH_i obtained from both occluded and nonoccluded data is therefore questionable. In addition, the "slope" of the E-[H⁺]_i relationship is not constant throughout the range of [H⁺]_i changes (Fig. 2B), and, perhaps more importantly, the gain of the E response decreased dramatically as [H⁺]_i increased. The opposite is known to take place in higher regions of exercise intensity.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Relationship between ventilation and intracellular pH (pHi) during contractions (solid arrows) and recovery (open arrows) in control () and occlusion () conditions. Note the large hysteresis of this relationship. All data were obtained from Figs. 1 and 3 of paper by Oelberg et al. (9). Lozenges and dotted line represent the averaged data obtained from individual regression analysis displayed in Table 1 (9). Minute ventilation (E) values obtained from this computation are way too high, compared to the averaged level of ventilation given by Oelberg et al. in their Fig. 1.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   A: relationship between ventilation and intracellular H+ concentration ([H+]i) during muscular contractions. Note that during occlusion ventilation was much higher than in control at any given level of acidity. The "normal" relationship between E and [H+]i was shifted upward during occlusion, leading to the conclusion that factors other than the local acidity must account for stimulation in E during vascular impediment. B: relationship between slope of the E/[H+]i relationship and [H+]i during muscular contractions. Note that 1) this slope was higher during occlusion than during control contractions at any given [H+]i and 2) the sensibility of the response decreased when [H+]i increased.

In our view, the foregoing analysis of the experimental data by Oelberg et al. does not justify their conclusion that there is a causal link between muscle [H⁺] and ventilation during exercise. Their data simply demonstrate that partial hindrance of venous return from hyperperfused muscles during and immediately after exercise stimulates ventilation, regardless of the existence of a local or systemic acidosis. This, in a variety of ways, has been demonstrated and discussed by others (1-8, 10).

	REFERENCES

1.   Dejours, P., J. C. Mithoefer, and A. Teillac. Essai de mise en évidence de chémorécepteurs veineux de ventilation. J. Physiol. (Paris) 47: 160-163, 1955.

2.   Haouzi, P., J. H. Hill, B. K. Lewis, and M. Kaufman. Skeletal muscles are equipped to sense vascular distension through group III and IV afferent fibers (Abstract). J. Physiol (Lond.) 507: 64P, 1998.

3.   Haouzi, P., J. J. Hirsh, F. Marchal, and A. Huszczuk. Ventilatory and gas-exchange response during walking in severe peripheral vascular disease. Respir. Physiol. 107: 181-190, 1997[Medline].

4.   Haouzi, P., A. Huszczuk, J. P. Gille, B. Chalon, F. Marchal, J. P. Crance, and B. J. Whipp. Vascular distension in muscles contributes to respiratory control in sheep. Respir. Physiol. 99: 41-50, 1995[Medline].

5.   Haouzi, P., A. Huszczuk, J. Porszasz, B. Chalon, K. Wasserman, and B. J. Whipp. Femoral vascular occlusion and ventilation during recovery from heavy exercise. Respir. Physiol. 94: 137-150, 1993[Medline].

6.   Huszczuk, A., E. Yeh, J. A. Innes, I. Solarte, and K. Wasserman. Role of muscular perfusion and baroreception in the hyperpnea following muscle contraction in dog. Respir. Physiol. 91: 207-226, 1993[Medline].

7.   Innes, J. A., I. Solarte, A. Huszczuk, E. Yeh, B. J. Whipp, and K. Wasserman. Respiration during recovery from exercise: effects of trapping and release of femoral blood flow. J. Appl. Physiol. 67: 2608-2613, 1989[Abstract/Free Full Text].

8.   McClain, J., C. Hardy, B. Enders, M. Smith, and L. Sinoway. Limb congestion and sympathoexcitation during exercise. J. Clin. Invest. 92: 2353-2359, 1993.

9.   Oelberg, D. A., A. B. Evans, M. I. Hrovat, P. P. Pappagianopoulos, S. Patz, and D. Systrom. Skeletal muscle chemoreflex and pH_i in exercise ventilatory control. J. Appl. Physiol. 84: 676-683, 1998[Abstract/Free Full Text].

10.   Rowell, L. B., L. Hermansen, and J. R. Blackmon. Human cardiovascular and respiratory responses to graded muscle ischemia. J. Appl. Physiol. 41: 693-701, 1976[Abstract/Free Full Text].

Philippe Haouzi, Laboratoire de Physiologie Faculté de Médecine de Nancy 54505 Vandeuvre-les-Nancy Cedex, France Andrew Huszczuk Vacumed Ventura, California 93003

	REPLY

To the Editor: The authors thank Haouzi and Huszczuk for their careful reading of our paper (2). They take issue with our "conclusion," which is quoted in isolation and without any of the discussion that precedes it. Haouzi and Huszczuk seem to miss the very important point that it is the extracellular fluid of muscle, which bathes the group IV afferents and changes in this compartment's pH (pH_e), which are relevant to the ventilatory chemoreflex. They also blur the distinction between exercise (which Fig. 4 and our conclusion allude to) and recovery from exercise, when mechanisms for ventilatory control are very likely different.

As stated in the paper, pH_i was used as a surrogate for pH_e because of the difficulty of measuring the latter in the human noninvasively and because of the relationship between the two during exercise in our rat model (1). Careful examination of those data shows that pH_e recovery begins immediately on cessation of exercise, whereas recovery of pH_i is relatively delayed (1). Thus a role for pH_e in recovery of ventilatory control is entirely plausible and could explain the hysteresis when both epochs are plotted. As indicated in its title, Table 1 depicts data obtained during exercise alone; no resting data were included.

A very similar argument can be made in comparing exercise E during the partially occluded vs. unoccluded state. pH_e, the relevant stimulus, is likely lower during partial circulatory occlusion, although, to our knowledge, it has never been measured under these conditions. As stated in the paper, it is also possible that other factors, either metabolic or physical, could enhance ventilatory gain during partial occlusion.

It is unclear whether the increasing ventilatory gain "in the higher regions of exercise intensity" refers to incremental exercise (our study utilized constant-load work) or the classically described ventilatory drift vs. oxygen uptake at constant loads above the lactate threshold. Decreasing "gain" of E vs. [H⁺]_i, however, is entirely in keeping with the literature and with our previous work in the human (3), although the mechanisms underlying this interesting phenomenon are not yet understood.

	REFERENCES

1.   Evans, A. B., L. W. Tsai, D. A. Oelberg, H. Kazemi, and D. Systrom. Skeletal muscle ECF pH error signal for exercise ventilatory control. J. Appl. Physiol. 84: 90-96, 1998[Abstract/Free Full Text].

2.   Oelberg, D. A., A. B. Evans, M. I. Hrovat, P. P. Pappagianepoules, S. Patz, and D. M. Systrom. Skeletal muscle chemoreflex and pH_i in exercise ventilatory control. J. Appl. Physiol. 84: 676-682, 1998.

3.   Systrom, D. M., S. J. Kohler, D. K. Kanarek, and H. Kazemi. ³¹P nuclear magnetic resonance spectroscopy study of the anaerobic threshold in man. J. Appl. Physiol. 68: 2060-2066, 1990[Abstract/Free Full Text].

David A. Oelberg, David M. Systrom, Pulmonary & Critical Care Unit Massachusetts General Hospital Harvard Medical School Boston, Massachusetts 02114