INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Fatiguing exercise and metabolic and respiratory acidosis have long been recognized to reduce force production in skeletal muscle. The relationship between muscle contractility, fatigue, and pH has been mostly investigated using in vitro muscle strip preparations (1) or in the intact animal (4) but there is little information regarding the interaction of these factors in humans and especially in the diaphragm.

It is well established that diaphragm force production can be significantly reduced by fatigue induced by periods of high-intensity voluntary isocapnic ventilation (5). During a 2-min maximal voluntary ventilation (MVV) maneuver there is a progressive reduction in ventilation and transdiaphragmatic pressure (Pdi) generation associated with the development of fatigue of the diaphragm (6, 8). Low-frequency fatigue, the reduction in the force response to low-frequency stimulation after severe muscular activity, can be demonstrated in the diaphragm for a prolonged period after MVV (8, 9).

Vianna and coworkers (10) demonstrated in limb muscle that acute hypercapnic acidosis caused a reduction in force production at rest and increased fatigability. Juan and colleagues (11), by examining the relationship between diaphragm electromyographic (EMG) activity and Pdi, showed acute hypercapnia reduced diaphragm contractility in the unfatigued state and also reduced endurance time. It is difficult, however, to accurately quantify diaphragm EMG responses owing to the effect of diaphragm shortening. Vianna and coworkers (12) were unable to demonstrate any effect of acute hypercapnia on diaphragm contractility in humans and most recently Mador and associates (13) showed that acute moderate hypercapnia reduced limb muscle contractility but had no effect on the diaphragm.

Clinical investigations have suggested that patients with chronic obstructive pulmonary disease and hypercapnia may have respiratory muscle weakness (14). It is not clear, however, whether hypercapnia itself contributes to the reduction in respiratory muscle strength. Studies in dogs (15) have demonstrated that hypercapnia reduces diaphragm contractility but does not increase fatigue. Schnader and coworkers (15) suggested that although both fatigue and hypercapnia decrease Pdi, they appeared to be operating by separate mechanisms.

Cervical magnetic stimulation of the phrenic nerve roots provides a nonvolitional, easily applied, reproducible and well tolerated technique to examine changes in diaphragm contractility (16, 17). The aim of the study was to examine the effect on the human diaphragm of simultaneous exposure to fatigue, induced by maximal voluntary ventilation, and acute hypercapnia by studying the changes in transdiaphragmatic twitch pressure (TwPdi) elicited by magnetic stimulation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Seven normal subjects (4 female, 3 male), mean age 30.4 yr (2.3 SD) were studied. Permission was obtained from the ethical committee of King's College Hospital, and informed consent was obtained in accordance with the guidelines laid down by the committee.

Equipment

Force production from the diaphragm was assessed as Pdi recorded from two 10-cm latex balloons placed in the midesophagus and stomach (P. K. Morgan, Gillingham, Kent, UK). Correct positioning was checked using the standard occlusion test (18). Pressure was measured by differential pressure transducers (MP45-1; Validyne, Northridge, CA) and carrier amplifiers (CD 280; Validyne). The pressure signals were analyzed and displayed in real time by a Macintosh Quadra 650 computer (Apple Computer Inc., Cupertino, CA) running Labview software (National Instruments, Austin, TX) with analogue-to-digital sampling at 100 Hz (12-bit NB-MIO-16; National Instruments, Austin, TX). Pdi was obtained on line by subtraction of esophageal pressure (Pes) from gastric pressure (Pgas).

Magnetic Stimulation

A circular 90-mm magnetic coil (P/N 8443) powered by a Magstim 200 (high-power) magnetic stimulator (Magstim Co., Whitland, Dyfed, UK) was used to bilaterally stimulate the cervical phrenic nerve roots. To find the optimal position for magnetic stimulation, the neck was flexed and the coil was placed over the spinous processes. A series of stimulations were performed in the midline between C5 and C7 until the site of maximal response was located. This spot was then marked and used for all subsequent stimulations. All stimulations were performed at the maximal power output of the stimulator.

MVV

MVV was used to fatigue the diaphragm. MVV was performed using the technique previously described by this laboratory (6). In brief, an open circuit was used. Subjects breathed via a mouthpiece from a 6-L anesthetic bag through a low-resistance, two-way nonrebreathing valve (Hans Rudolph, Kansas City, MO). The gas in the bag was constantly replenished such that it was always fully inflated at the start of inspiration, any excess gas being vented through an exhaust valve attached to the bag. Two flow generators (Downs No. 9200; Medic-Aid Ltd, Pagham, Sussex, UK) connected to the air cylinders were used to achieve the high air flow required during the MVV by entrainment of additional airflow from the atmosphere. The inspired air in the bag could be enriched with varying concentrations of oxygen (O₂) and carbon dioxide (CO₂), the flow of both gases being controlled by rotameters. PCO₂ and PO₂ were sampled directly and continuously from the mouthpiece and gas reservoir respectively and measured by a dual capnograph/O₂ monitor (Model No. 455; P. K. Morgan). The inspired (PI_CO2) and end-tidal (PET_CO2) PCO₂ were displayed in real time on the computer screen, PET_CO2 giving an accurate noninvasive measurement of arterial PCO₂ (19). Expiratory flow was measured using a pneumotachograph (Fleisch No. 4) attached to the expiratory port of the nonrebreathing valve via a brass tube (41 cm × 3.5 cm interior diameter) which acted to smooth flow. Expiratory flow from the pneumotachograph was measured by a Mercury CS6 electrospirometer (G.M. Engineering, Kilwinning, Scotland, UK) and analyzed in real time by the computer to give an on-line display of expired tidal volume.

Protocol

All measurements were performed with the subject seated and wearing a nose clip. Esophageal and gastric balloons were positioned as described previously and the site for optimal cervical stimulation located. Subjects were then instructed to breathe quietly and remain silent for 20 min to avoid twitch potentiation (17). A set of 10 cervical stimulations were recorded to give a measure of baseline diaphragm contractility (TwPdi) prior to the fatiguing protocol, each stimulation being at functional residual capacity and applied at 30-s intervals to avoid twitch on twitch potentiation. The subjects then performed a 2-min MVV breathing air during which they were vigorously encouraged to produce a maximal ventilatory effort. Phrenic nerve stimulation was repeated immediately after MVV and at 20, 40, 60, and 90 min.

The study was repeated on a separate occasion, at least 7 d apart, with baseline TwPdi being measured as before (subject breathing room air). On this occasion, however, after the baseline TwPdi measurements the subjects inspired 8% CO₂ for 12 min before and during the 2-min MVV. Immediately after the MVV the subjects returned to breathing room air and phrenic nerve stimulation was repeated. As before, phrenic nerve stimulation was also performed at 20, 40, 60, and 90 min after the MVV. The order in which the CO₂ and control runs were performed was randomized.

Due to the design of the study, the TwPdi measurements made immediately after both MVV maneuvers will be potentiated, although provided the levels of ventilation are equal during both MVV maneuvers, any twitch potentiation present immediately after each MVV would be expected to be similar. However, it is not clear whether CO₂ itself has an effect on twitch potentiation. To address this issue, an additional study was performed in five subjects (2 female, 3 male) with a mean age of 30.6 yr (2.4 SD) in which the direct effect of CO₂ on twitch potentiation was examined. A 5-s maximal quasi-static inspiratory effort against a closed airway (PI_max) was used to potentiate the diaphragm (20) and the degree of potentiation assessed using bilateral anterior magnetic phrenic nerve stimulation performed immediately after each maneuver. Two 43-mm figure-of-eight coils (P/N 8456) powered by two Magstim 200 magnetic stimulators (Magstim Co., Whitland, Dyfed, UK) were used to bilaterally stimulate the phrenic nerves on the anterior aspects of the neck (21). As before, the site of maximal response was marked and used for all subsequent stimulation. The supramaximality of the stimuli was assessed by examining TwPdi at varying stimulator outputs imposed in random order (80, 85, 90, 95, and 100%). The study was conducted under the same conditions as described previously. After a 20-min rest period, 10 unpotentiated and 10 potentiated stimulations were performed to obtain a measure of baseline diaphragm contractility while breathing air. The subjects then inspired 8% CO₂ for 14 min, matching the hypercapnic conditions in the earlier study. Immediately after, with the subjects returned to breathing room air, 10 potentiated twitches were recorded. After a second 20-min rest period a further 10 potentiated stimulations were recorded.

Statistical Analysis

Minute ventilation was calculated for every breath and expressed as the mean value for each 10-s period throughout the MVV maneuvers. Comparison of the MVV maneuver between the control and CO₂ inhalation runs was by two-way analysis of variance (ANOVA) of repeated measures. The mean TwPdi was calculated for each subject for each set of 10 stimulations at baseline and at 0, 20, 40, 60, and 90 min after MVV. Significance of change was investigated by two-way ANOVA of repeated measures of the changes of individual mean data and paired Student's t test comparison for mean group data. Two-way repeated measures ANOVA was used to analyze individual subject data for both the time-dependent changes in TwPdi and the difference between control and 8% CO₂ studies.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All subjects were able to successfully perform the MVV in both the control and CO₂ studies. Full data on ventilation were not obtained in one subject. Ventilation declined rapidly over the first 60 s, reaching a plateau that was sustained for the second minute (Figure 1). Mean ventilation fell significantly (p < 0.01) from 156 L · min¹ and 167 L · min¹ at the start of the MVV to 115 (26% reduction) and 121 (28% reduction) L · min¹ after 60 s during the control and CO₂ run, respectively. All subjects were able to achieve a MVV maneuver with 8% CO₂ within 15% of their control run breathing air. Ventilation was not different between the control and CO₂ run as assessed by repeated-measures ANOVA.

View larger version (12K): [in this window] [in a new window]   Figure 1.   Minute ventilation (mean ± SD) during the normocapnic (closed circles) and hypercapnic (8% CO2, open circles) 2-min MVV. Values given are for six of seven subjects.

Raising PI_CO2 to approximately 8% significantly increased mean PET_CO2 from 5.5% ± 0.26% (mean ± SD) at rest to 8.9% ± 0.35 at 5 min and 8.7% ± 0.34% at 12 min during CO₂ inhalation. PI_CO2 was held at this elevated level throughout the MVV (8.8 ± 0.67%). There was no change in PET_CO2 during the control experiment.

TwPdi was significantly reduced at 20 min after MVV in all subjects in both the control and CO₂ runs (p < 0.01), the group mean (± SD) TwPdi falling from 30.4 ± 7.8 cm H₂O to 27.0 ± 8.1 cm H₂O (11.2% reduction) in the control run and from 30.3 ± 4.1 to 27.3 ± 5.0 cm H₂O (10% reduction) after CO₂ inhalation (Figure 2). Mean TwPdi remained substantially and significantly (p < 0.01) below baseline throughout the recovery period, although it had returned to within 5% of baseline by 90 min. There was no difference between the absolute fall in and subsequent recovery of TwPdi during the control and the CO₂ run at 20, 40, 60, and 90 min. However, there was a greater decrease (15%) (p < 0.05) in TwPdi immediately after MVV in the CO₂ inhalation study than that in the control study (8.1%) (Figure 2) when compared with baseline values.

View larger version (17K): [in this window] [in a new window]   Figure 2.   Reduction and recovery of TwPdi (mean ± SEM) after normocapnic (closed circles) and hypercapnic (8% CO2, open circles) 2-min MVV in seven subjects. Values for TwPdi are expressed as a percentage of baseline. Filled column represents period of MVV. TwPdi is reduced to 92% of baseline in the normocapnic run and to 85% in the hypercapnic run.

The constancy of cervical magnetic stimulation was assessed by measuring the mean within-occasion coefficient of variation (CV) for 10 measurements of TwPdi at baseline values. The CV during the control and CO₂ runs was 5.4% and 5.6% respectively.

Supramaximal stimulation was achieved in the additional study using bilateral anterior magnetic phrenic nerve stimulation as indicated by a plateau in TwPdi with increasing stimulator output (Figure 3). Mean (± SD) unpotentiated baseline TwPdi increased significantly (p < 0.05) from 30.8 ± 7.8 cm H₂O to 41.4 ± 9.2 cm H₂O after potentiation with the 5-s PI_MAX maneuver. As before, increasing PI_CO2 to 8% significantly increased mean (± SD) PET_CO2 from 5.4 ± 0.1% at rest to 8.5 ± 0.3% at 6 min and 8.4 ± 0.3% at 11 min. There was no statistically significant difference in mean (± SD) potentiated TwPdi or PI_MAX before (TwPdi 41.4 [9.2] cm H₂O; PI_MAX 88.8 [21.3] cm H₂O), immediately after (TwPdi 40.3 [8.0] cm H₂O; PI_MAX 88.1 [27.1] cm H₂O), and at 20 min (TwPdi 43.8 [8.2] cm H₂O; PI_MAX 90.4 [21.0] cm H₂O) after 8% CO₂ inhalation (Figure 4).

View larger version (10K): [in this window] [in a new window]   Figure 3.   Mean (± SEM) TwPdi for bilateral anterior phrenic nerve stimulation over a range of stimulus strengths. Stimulus strength is expressed as a percentage of maximal stimulator output.

View larger version (26K): [in this window] [in a new window]   Figure 4.   Effect of hypercapnia on twitch potentiation. Potentiated TwPdi (closed circles) and PIMAX (closed squares). Hatched area represents period of CO2 inhalation.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study describes the interaction of acute hypercapnia and fatiguing exercise on the contractility of the human diaphragm. The main findings are that hypercapnia does not reduce maximal ventilatory capacity and does not intensify long lasting fatigue, but may augment the reduction in diaphragm contractility immediately after sustained maximal ventilation. In addition, hypercapnia alone has no effect on twitch potentiation.

Cervical magnetic stimulation was used to stimulate the phrenic nerves. Magnetic stimulation is an ideal technique for use in studies involving human subjects as it provides a nonvolitional measure of contractility which is easily applied, well tolerated, and TwPdi is highly reproducible (16, 17). In contrast, supramaximal bilateral electrical stimulation of the phrenic nerves can be difficult to achieve reproducibly because of problems of accurately locating the phrenic nerves (21, 22). Also, increases in ventilation caused by painful electrical stimuli could cause twitch potentiation (20, 23). Cervical magnetic stimulation has the disadvantage that in some subjects it may not be possible to achieve supramaximal stimulation of the phrenic nerves (21) leading to greater variation in TwPdi. Supramaximality has been demonstrated in normal human subjects (17) but the plateau of the stimulus-response curve occurs close to maximal stimulator output. To minimize variations in stimulus strength, great care was taken during each run to ensure optimal positioning of the coil on the neck at the site of maximal TwPdi response, which was marked and then used during subsequent stimulations. There was no difference in mean baseline TwPdi between the control and CO₂ runs, 30.4 and 30.3 cm H₂O, respectively. To further assess the constancy of the stimulus, we examined the mean within-occasion CV for 10 measurements at baseline. The CV during the control and CO₂ runs compared well with those obtained for supramaximal electrical stimulation (6.7%) (17), supramaximal cervical magnetic stimulation (6.7%) (17), and bilateral anterior magnetic stimulation (5.3%) (21). The low CV for cervical magnetic stimulation in the present study suggests that stimulation was constant. Given that stimulation intensity was constant, the randomized controlled nature of the study makes supramaximality less crucial.

In the additional study, the supramaximality of the stimulus with bilateral anterior magnetic stimulation was assessed directly by examining TwPdi in response to randomly varying stimulator output from 80 to 100%. There was a clear plateau in TwPdi across this range suggesting that the stimulus was supramaximal.

PET_CO2 was used to give an accurate noninvasive measurement of arterial PCO₂ (19) rather than direct sampling of arterial or arterialized venous blood. Although this technique does not give a direct measurement of arterial pH, previous studies have shown reductions in pH and skeletal muscle strength during similar levels of acute hypercapnia. Both Vianna and coworkers (10) and Mador and associates (13) showed significant reductions in adductor pollicis force production after 6 and 8 min of 9% and 8% CO₂, respectively. The results of these studies suggest that the degree of hypercapnia in the present study would reduce pH to a sufficient level to affect skeletal muscle contraction; consistent with this, diaphragm contractility was reduced more immediately after the hypercapnic MVV than after the normocapnic study.

There was no difference between TwPdi at 20 to 90 min after MVV between the control and hypercapnic studies, although there was a greater decrease in TwPdi immediately after the MVV in hypercapnic studies. Although measurements of TwPdi made immediately after the MVV would be potentiated possibly underestimating the severity of fatigue, as the levels of ventilation during the hypercapnic and normocapnic MVV were the same any twitch potentiation present immediately after the MVV would be expected to be similar. It was not clear, however, whether CO₂ itself would have an effect on twitch potentiation and the additional study was performed to address this issue. Diaphragm twitches were potentiated using a 5-s PI_MAX maneuver, and were recorded before and immediately after inhalation of 8% CO₂ for 14 min, the same level and duration of exposure used in the first study. There was no difference in potentiated TwPdi or PI_MAX before or after CO₂ inhalation. The results of this study indicated that hypercapnia had no effect on twitch potentiation. The immediate decrease in TwPdi under hypercapnic conditions would therefore appear to have been caused by a reduction in the underlying diaphragm contractility rather than any effect of CO₂ on the extent of twitch potentiation resulting from the hypercapnic MVV.

There are few data in the literature concerning the effect on the diaphragm of combined exposure to hypercapnia and fatigue and conflicting data concerning the effect of hypercapnia alone on diaphragm contractility. Schnader and coworkers (15) reported a reduction in diaphragm contractility and endurance during hypercapnia in dogs. However, exposure to hypercapnia in combination with fatigue did not further reduce diaphragm force production. In humans, Juan and colleagues (11) reported a reduction in diaphragm contractility during hypercapnia alone whereas Mador and associates (13) and Vianna and coworkers (12) both reported no such effects. Although Mador and associates (13) were unable to show a significant change in diaphragm contractility with acute moderate hypercapnia, they did report a decrease of 12.2% in twitch Pdi in an additional experiment when breathing 8% CO₂. The ability of the subjects to relax their respiratory muscles while breathing 8% CO₂ made obtaining adequate diaphragmatic twitches difficult. However, they did suggest that significant changes in TwPdi could have been demonstrated if a larger number of subjects had been studied. The reduction in diaphragm contractility immediately after hypercapnic fatigue in the present study would therefore seem to support the findings of some previous investigators.

Diaphragm muscle activity is an important determinant of diaphragm blood flow. The increase in minute ventilation associated with exercise and CO₂ inhalation have both been shown to increase diaphragm blood flow significantly (24). Kendrick and coworkers (24) demonstrated a linear relationship between hypercapnia and blood flow in the diaphragm but not in limb muscles. They showed that this relationship was abolished after paralysis, suggesting that hypercapnia had no significant effect on respiratory muscle blood flow other than that mediated via the increase in ventilation. Greatly increasing blood flow during exercise acts to maintain diaphragm contractility by preventing the build-up of metabolic by-products in the intra-, or extracellular milieu (28). Therefore, increasing arterial PCO₂ leads to an increase in minute ventilation and hence diaphragm blood flow, which in turn acts to reduce any direct effects of CO₂ on diaphragm contractility, possibly explaining the absence of an effect of acute hypercapnia on the diaphragm in some studies.

Reduced muscle contractility during hypercapnia is thought to be a result of the decrease in intracellular pH (pH_i) (31), which has been demonstrated using ³¹P nuclear magnetic resonance (32). Increased binding of calcium to the sarcoplasmic reticulum, decreased affinity of troponin for calcium, and reduction in the rate of glycolysis and hence ATP resynthesis have been suggested as possible mechanisms underlying the reduction in contractility. Schnader and coworkers demonstrated that respiratory acidosis reduced diaphragmatic contractility (33) and endurance (15) in the dog but hypercapnia did not intensify fatigue. They suggested that while fatigue and hypercapnia both decrease transdiaphragmatic pressure, they appeared to be operating via separate mechanisms. They postulated that decreased force production during hypercapnia was caused by secondary fall in pH_i which would decrease the binding of Ca²⁺ to troponin. The decreases in force production caused by fatigue, however, were operating via a different mechanism as hypercapnia did not alter the degree to which fatigue depressed diaphragm contractility and hypocapnia did not correct the reduced contractility of the fatigued diaphragm as it did in the hypercapnic diaphragm (15).

Sahlin and coworkers (34) reported that intracellular CO₂ was only increased when CO₂ breathing was combined with light bicycle exercise, possibly owing to increased local circulation and increased endogenous CO₂ production in the muscle. In the present study, the MVV maneuver would be analogous to exercise, increasing diaphragm blood flow and endogenous CO₂ production. The combination of an increased CO₂ production with the large hypercapnic load, may have been sufficient to overcome any protective effect of increased diaphragm blood flow, the net effect being a greater decrease in TwPdi after following the hypercapnic MVV.

The clinical implications of these findings are mixed. The absence of any change in the degree of low-frequency fatigue with hypercapnia suggests that the presence of a raised CO₂ during acute respiratory failure, may have no deleterious effects on respiratory muscle function. However, the greater immediate fall in TwPdi after the hypercapnic MVV suggests a direct adverse effect of hypercapnia on diaphragm contractility.

Acute-on-chronic deterioration in the clinical status of patients with lung disease may result in CO₂ retention, which in itself may further compromise respiratory muscle function and lead to a downward spiral of further CO₂ retention, respiratory muscle dysfunction, and possible respiratory failure. In patients with asthma, the work of breathing may no longer be tolerated with the onset of CO₂ retention. The reduction in diaphragm contractility immediately after the hypercapnic MVV suggests that in clinical situations characterized by an acute elevation of CO₂, respiratory muscle dysfunction could contribute to a reduction in ventilatory capacity. Further clinical studies are needed.

In conclusion, hypercapnia does not impair maximal ventilatory capacity and does not lead to more severe long lasting fatigue. Acute hypercapnia during MVV may impair diaphragm contractility but any impairment is rapidly reversed when normocapnia is restored.