Quantification of the Esophageal Diaphragm Electromyogram with Magnetic Phrenic Nerve Stimulation YUAN M. LUO, REBECCA A. LYALL, M. LOU HARRIS, GERRARD F. RAFFERTY, MICHAEL I. POLKEY, and JOHN MOXHAM Department of Respiratory Medicine, King’s College Hospital, London, United Kingdom

Measurement of the amplitude of the esophageal diaphragm compound muscle action potential (CMAP) could be useful for the diagnosis and monitoring of neuromuscular disease. However, quantification of the diaphragm CMAP has been hampered by difficulty in positioning the esophageal electrode at the diaphragm’s electrically active center and many investigators report arbitrary units rather than voltage. To quantify the esophageal diaphragm CMAP we designed a multipair electrode which we evaluated during unilateral magnetic stimulation. The esophageal catheter consisted of four sequential electrode pairs. Overall the electrode spanned 17 cm and covered the entire electrically active region of the diaphragm. The diaphragm CMAP was simultaneously recorded from the four pairs at distances of 40, 39, 38, and 37 cm from the nose to the proximal electrode pair. Studies were undertaken in 10 normal subjects and 10 patients with diaphragm dysfunction. The amplitude of the CMAP (peak to peak) was defined as the average of five twitches recorded from the optimal pair of electrodes. The amplitude of the diaphragm CMAP elicited by unilateral maximal magnetic stimulation was 1.45 6 0.35 mV (mean 6 SD) for the right side and 1.68 6 0.47 mV for the left. When measured on different occasions the coefficient of variation (CV) was 8.6%. The amplitude of the CMAP measured from dysfunctional hemidiaphragms was much less than that measured from normal subjects. This study suggests that the diaphragm CMAP can be quantified using an appropriate esophageal electrode. Luo YM, Lyall RA, Harris ML, Rafferty GF, Polkey MI, Moxham J. Quantification of the esophageal diaphragm electromyogram with magnetic phrenic nerve AM J RESPIR CRIT CARE MED 1999;160:1629–1634. stimulation.

The diaphragm compound muscle action potential (CMAP) elicited by phrenic nerve stimulation is often examined to obtain the conduction time but the amplitude of the CMAP is not normally used in clinical practice. Many investigators report arbitrary units rather than voltage (1–3). Some neuromuscular diseases (e.g., axonal neuropathies) which affect respiratory function may cause a reduction in the amplitude of the CMAP, without a change of phrenic nerve conduction time as the surviving axons conduct the action potential normally (4, 5). Therefore quantification of amplitude of the CMAP is sometimes more relevant than measurement of the phrenic nerve conduction time in assessing diaphragm function. Previous studies have used the combination of chest wall surface electrodes and electrical stimulation (ES) of the phrenic nerve to quantify diaphragm CMAP (5–8). However, it may be technically demanding to achieve and sustain supramaximal stimulation of the phrenic nerve with ES. This is particularly the case in obese subjects with short necks or in patients with severe diaphragmatic weakness (9). Unilateral

magnetic stimulation (UMS) is easier to perform (3) but the diaphragm CMAP elicited by UMS recorded from chest wall electrodes sometimes differs from that elicited by ES (3). The esophageal diaphragm electromyogram (EMG) is relatively free of contamination from other muscles (7) and less influenced by obesity. The esophageal electrode might therefore have advantages for the quantification of the diaphragm CMAP. However, a single pair electrode can be difficult to accurately position at the center of the electrically active region of the diaphragm (EARdi) and the CMAP recorded may therefore not represent the maximal amplitude. To overcome this problem we designed an esophageal electrode with a multipair electrode array which spanned the entire electrically active region of the diaphragm and used it with UMS to quantify the diaphragm CMAP. We also observed whether the amplitude of the CMAP in patients with neuromuscular disease was lower than that measured in normal subjects.

METHODS Subjects

(Received in original form September 14, 1998 and in revised form May 17, 1999) Correspondence and requests for reprints should be addressed to Dr. Y. M. Luo, Respiratory Muscle Laboratory, King’s College Hospital, Bessemer Rd., London SE5 9PJ, UK. Am J Respir Crit Care Med Vol 160. pp 1629–1634, 1999 Internet address: www.atsjournals.org

Ten healthy volunteers (5 men and 5 women) age 29 to 54 yr (mean age 36) and 10 patients with neuromuscular disease participated in the study. Five of the 10 patients had amyotrophic lateral sclerosis (ALS), three neuralgic amyotrophy, and two myopathy. The study was approved by the ethics committee of King’s College Hospital and all subjects gave their informed consent.

1630

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

Data Acquisition The esophageal electrode catheter, modified from that used in a previous study (10), consisted of six coils 1 cm in width each separated by a distance of 3 cm. The top coil was connected to earth and the lower five coils formed four sequential pairs of electrodes (Figure 1). The esophageal electrode was passed through the nose and swallowed into the esophagus. Recordings were performed when the proximal electrode pair was 40, 39, 38, and 37 cm from the nostril. Electromyographic signals recorded from all four electrode pairs were amplified and band-pass filtered between 10 Hz and 10 kHz (Magstim Co., Whitland, Dyfed, UK). The signals were then passed to a 12 bit analog-to-digital converter (NB-MIO-16; National Instruments, Austin, TX) and displayed in real time and stored on a Macintosh Centris Computer running Labview 2.2 software (National Instruments). The signals were available in real time to the investigators. Esophageal pressure (Pes) and gastric pressure (Pga) were recorded using conventional balloon catheters (3). The pressures were measured by differential pressure transducers (Validyne MP45; Validyne, Northridge, CA) and were recorded with Labview software. Transdiaphragmatic pressure (Pdi) was measured by electronic subtraction of Pes from Pga. The Pdi elicited by unilateral phrenic nerve stimulation was termed twitch Pdi (Tw Pdi). The Pdi obtained during maximal sniff maneuvers was also recorded and was termed sniff Pdi. UMS of the phrenic nerves was performed using a 43-mm figure-ofeight coil (P/N 9784-00) powered by a Magstim 200 stimulator (Magstim Co. Ltd.). The coil was placed anterolaterally on the neck over the phrenic nerve (3). Stimulation was performed when subjects were seated upright in a chair, at resting end-expiration. Five twitches were obtained using 80% of magnetic stimulator output. Supramaximal ES was performed as described in a previous study (10) and five stimulations were delivered with the same posture as that during UMS.

Data Analysis The diaphragm CMAP was analyzed off-line and the amplitude of the CMAP was measured from peak to peak. To eliminate the influence of the electrocardiogram we only analyzed CMAPs with a constant

VOL 160

1999

shape and a stable baseline before and after stimulation. Unless stated otherwise, data reported were measured from the electrode pair with the greatest CMAP amplitude. The EARdi center was defined as the position of the common electrode of two pairs with opposing signal polarity and similar amplitude. Results were expressed as the mean 6 SD. Nonparametric tests were performed to examine differences and the agreement between ES and UMS was assessed by a Bland and Altman plot (11).

Protocol UMS was performed in all patients and normal subjects. ES was performed only in six of the normal subjects (numbers 1–6). Pes, Pga, Tw Pdi, and sniff Pdi were obtained only from the patient group, to confirm and quantify diaphragm dysfunction. In six of the 10 normal subjects (numbers 1–3, 5, 6, 9) studies were undertaken on 2 to 4 separate occasions, at least 1 mo apart. On one occasion in four subjects the study was performed by another investigator.

RESULTS Diaphragm CMAP Elicited by UMS

The electrode catheter was tolerated by all subjects. The mean amplitude of the CMAP was 1.68 6 0.47 mV and 1.45 6 0.35 mV for the left and right side, respectively (Table 1). Once the proximal electrode pair was placed at 40 cm from the nose, a good quality diaphragm CMAP could always be recorded, although to obtain the maximal amplitude of the CMAP it was sometimes necessary to withdraw the electrode by up to 3 cm in 1-cm steps. Considering all four pairs together reversal of signal polarity was always observed. On the left side there was always a single reversal of polarity between consecutive electrode pairs (Figure 2). This also was the case for the right in most subjects. However, in three subjects there were two reversals of polarity on the right (Figure 3). The shape of the CMAP on the left side usually differed from that on the right side when recordings were made from the same position. The mean distance from the nose to the center of EARdi was 45 cm (range 41 to 49 cm). The level of the EARdi center was usually similar between the left and right sides. In two subjects the level of the EARdi center on the right side was 1 or 2 cm higher than that on the left and in one subject the level on the left was higher than that on the right by 1 cm. The greatest amplitude of the diaphragm CMAP always occurred in one of the two electrode pairs adjacent to the EARdi (Figures 2–4). There

TABLE 1 CHARACTERISTICS OF NORMAL SUBJECTS AND RESULTS OF DIAPHRAGM CMAP MEASURED WITH UNILATERAL MAGNETIC STIMULATION OF THE PHRENIC NERVE Latency of the CMAP (ms)

Height (cm)

EARdi Center to Nose (cm)

Left

Right

Left

Right

1 2 3 4 5 6 7 8 9 10

164 160 160 168 168 170 189 163 188 188

42 41 42 45 43 45 49 42 48 48

2.05 2.10 1.02 2.27 0.89 1.62 1.46 1.58 2.08 1.69

1.71 1.21 1.10 1.82 1.02 1.51 1.42 1.74 1.95 1.02

7.2 7.5 8.1 7.2 6.0 7.6 7.9 8.1 8.7 7.5

6.3 6.1 6.5 6.8 6.0 7.4 7.0 6.3 8.6 8.1

Mean SD

172 12

45 3

1.68 0.47

1.45 0.35

7.6 0.7

6.9 0.9

Subject No.

Figure 1. Multipair esophageal electrode. Each electrode is 1 cm wide and there is a 3-cm distance between electrodes. The top electrode is connected to earth. The lower five electrodes create four consecutive pairs of electrodes. 1 5 electrode connected to positive input terminal of differential amplifier; 2 5 electrode connected to negative input terminal.

Amplitude of the CMAP (mV )

p Value

. 0.05

, 0.05

1631

Luo, Lyall, Harris, et al.: Quantification of the Diaphragm EMG

Figure 2. Diaphragm CMAP elicited by UMS recorded simultaneously from four pairs of electrodes in a normal subject on the left side. Single reversal of signal polarity occurred between consecutive electrode pairs 2 and 3. Maximal amplitude of the CMAP occurred in one of two adjacent pairs with reversed polarity signals. Three traces superimposed.

was a good correlation between subject height and distance from the nose to the EARdi center (r 5 0.97). Phrenic nerve conduction time measured with UMS was 7.6 6 0.7 ms on the left side and 6.9 6 0.9 ms on the right (Table 1). There was a significant difference between the left and the right sides (p , 0.05). For the six subjects studied on more than one occasion the amplitude of the CMAP is shown in Table 2. The coefficient of variation (CV) was 8.6% 6 4.8%. The latency of the CMAP was similar between studies. Comparison of the Diaphragm CMAP Elicited by ES and UMS

A good quality diaphragm CMAP was recorded with ES. The shape, amplitude, and latency of the CMAP elicited by ES were similar to those elicited by UMS (Figure 4). In the six normal subjects in whom both ES and UMS were performed, the mean amplitude of the CMAP was 1.55 6 0.50 mV for ES and 1.66 6 0.59 mV for UMS on the left side, and 1.34 6 0.30 mV for ES and 1.40 6 0.33 mV for UMS on the right. There were no differences in amplitude and latency of the CMAP elicited by ES and UMS (Figure 5). Patient Studies

Patient data including sniff Pdi and Tw Pdi are shown in Table 3. The amplitude of the CMAP in patients (0.42 6 0.51

Figure 3. Diaphragm CMAP elicited by UMS recorded simultaneously from four pairs of electrodes in a normal subject on the right side. Two reversals (pairs 1 and 2, pairs 2 and 3) of initial signal polarity occurred between consecutive electrode pairs. Three traces superimposed.

mV) was much lower than that in normal subjects (1.56 6 0.42 mV) (p , 0.001, pooled left and right data) whereas phrenic nerve conduction time was normal or only slightly longer when CMAP could be recorded (Figure 6). However, no diaphragm CMAP could be recorded in some patients with diaphragm paralysis.

DISCUSSION In the present study we have shown that the amplitude of the diaphragm CMAP can be quantified with a multipair esophageal electrode and UMS. ES was similar to UMS in depolarizing the phrenic nerve in terms of the diaphragm CMAP shape and amplitude. The amplitude of the CMAP had a good between-occasion reproducibility. The amplitude of the CMAP in patients with neuromuscular disease was substantially reduced in most cases, as was Tw Pdi, and the two were related. It is important that UMS supramaximally stimulates the phrenic nerves. UMS is a relatively new technique and it is useful to further compare this technique with the established technique of ES. Mills and colleagues (12) demonstrated that 80% of magnetic stimulator output produced maximal stimulation of the phrenic nerve as assessed by Tw Pdi. In a previous study we showed that UMS at 80% stimulator output maximally depolarizes the phrenic nerve based on the amplitude of the CMAP recorded from an esophageal electrode (13). In the present study, the shape, amplitude, and latency of

1632

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 160

1999

Figure 4. Diaphragm CMAP elicited by ES and UMS recorded from four pairs of esophageal electrodes in one normal subject for the left hemidiaphragm. The diaphragm CMAP elicited by ES is similar to that elicited by UMS in terms of shape amplitude and latency. Four traces superimposed.

the diaphragm CMAP elicited by UMS was similar to that elicited by supramaximal ES (Figure 4), supporting the contention that UMS was supramaximal. The observation that the latency of the CMAP measured with UMS in the present study was similar to the results reported by McKenzie and Gandevia (7) who used ES and an esophageal electrode, further supports the similarity of ES and UMS in depolarizing the phrenic nerve. Because UMS is easy to perform, and maximal stimulation can be consistently achieved (3, 12) we believe it to be a useful technique for the quantification of the diaphragm CMAP amplitude, although UMS is contraindicated in the presence of implanted metal and cardiac pacemakers. It is essential to locate the electrode at the EARdi center to record a maximal CMAP. This may be difficult with a conventional single pair esophageal electrode. The multipair electrode has a span of 17 cm which is long enough to cover the center of the EARdi. By withdrawing the electrode 3 cm in

1-cm steps CMAPs are obtained between 37 and 57 cm from the nose. Confirmation that the electrode span has covered the EARdi is obtained by the reversed polarity of the CMAP, because the polarity and shape of the diaphragm CMAP are sensitive to electrode position (10, 14). McKenzie and Gandevia (7) reported that the distance from the EARdi center to the nose was 45 to 53 cm and the EARdi center on the left side was usually higher than that on the right. Their data differ slightly from the present results. In the present study the distance of the electrical activity center from the nose was 41 to 49 cm using the criterion of the reversed polarity of the CMAP. This difference could not simply be a result of differences in height because these were similar. We also could not find a consistent difference in the level of EARdi center between the left and right sides. However, in the study of McKenzie and Gandevia (7) the electrode position was determined by a balloon at the distal end of a cathe-

TABLE 2 DIAPHRAGM CMAP AMPLITUDE (MV) OF SIX SUBJECTS MEASURED ON 2 TO 4 OCCASIONS DURING MAGNETIC PHRENIC NERVE STIMULATION ON THE LEFT (L) AND RIGHT (R) SIDES Subject 1

1st 2nd 3rd 4th* Mean SD CV

Subject 2

Subject 3

Subject 5

Subject 6

Subject 9

L

R

L

R

L

R

L

R

L

R

L

R

2.05 1.81 1.99 2.12 1.99 0.13 7

1.71 1.53 1.60 1.93 1.69 0.17 10

2.10 1.82 1.78 1.54 1.81 0.23 13

1.21 1.14 1.05 1.27 1.17 0.09 8

1.02 1.02

1.10 1.15

1.62 1.41 1.29 1.20 1.38 0.18 13

1.51 1.66 1.41 1.75 1.58 0.15 10

1.95 1.91

1.13 0.04 3

1.02 1.19 1.32 1.15 1.17 0.12 11

2.08 2.49

1.02 0.00 0

0.89 1.02 1.18 1.25 1.09 0.16 15

2.29 0.29 13

1.93 0.03 1

Definition of abbreviation : CV 5 coefficient of variation. * Experiment performed by other investigator.

1633

Luo, Lyall, Harris, et al.: Quantification of the Diaphragm EMG

Figure 5. Bland and Altman plot. There is agreement between ES and UMS in assessment of the diaphragm CMAP. (Left panel) Diaphragm CMAP amplitude. (Right panel) Latency of the CMAP. Pooled data for left and right hemidiaphragms.

ter wedged at the esophageal–gastric junction. Therefore our data may more precisely represent the distance of the EARdi center from the nose. Kim and coworkers (15) found in animal studies with a multipair electrode that the maximal amplitude of the CMAP was usually recorded 4 to 6 cm away from the esophageal hiatus. This differs from the result of the present study where the maximal amplitude was always recorded from the esophageal hiatus, as based on the reversed polarity of the CMAP (16). However, this discrepancy between the two studies may be explained by different methodologies, such as subjects studied, electrode configuration, and stimulating frequency. Quantification of the diaphragm CMAP has previously been performed with chest wall surface electrodes. However, there can be large variation in the amplitude of the CMAP recorded from chest wall electrodes between subjects (6, 7). Swenson and Rubenstein (8) also found that it was difficult to standardize the amplitude of the diaphragm CMAP with chest

wall electrodes. Large variability of the amplitude of the CMAP may be due to anatomic differences, difficulty in accurately positioning the electrodes, and differences in chest wall thickness (6) as well as difficulty in maximally stimulating the phrenic nerve with ES (3, 17). With the esophageal electrode the amplitude of the CMAP was much greater than that with chest wall electrodes (7). However, with a single electrode pair the amplitude of the CMAP is variable with a range of 0.3 to 2.8 mV in the study of McKenzie and Gandevia (7). Similowski and coworkers (17) also reported large variability in amplitude of the CMAP between subjects using a single pair esophageal electrode and cervical magnetic stimulation of the phrenic nerve. This large variability in amplitude of the CMAP makes standardization difficult. In the present study with a multipair esophageal electrode, there was good reproducibility between occasions in terms of maximal CMAP amplitude. The variability of the amplitude of the CMAP (0.89 to 2.27 mV) was also less. Good re-

TABLE 3 TRANSDIAPHRAGMATIC PRESSURE AND THE DIAPHRAGM CMAP ELICITED BY UMS IN 10 PATIENTS Left Side

Right Side

CMAP

CMAP

Patient No.

Sex

Age

Sniff Pdi (cm H2O)

Tw Pdi (cm H2O)

Amplitude (mV )

Latency (ms)

Tw Pdi (cm H2O)

Amplitude (mV )

1 2 3 4 5 6 7 8 9 10

F M M M M F M M F M

38 57 67 58 65 62 36 40 79 59

20 11 22 64 87 38 13 86 23 42

5 5 5 11 6 6 * 12 3 *

0.36 0.20 0.40 1.16 0.71 1.07 0.00 2.10 0.18 0.09

6.3 7.5 11.0 9.8 8.5 7.2

* * * 3 2 4 2 * * *

0.00 0.00 0.20 0.41 0.38 0.36 0.20 0.21 0.00 0.30

8.9 9.8 7.2

Definition of abbreviations : ALS 5 amyotrophic lateral sclerosis; NA 5 neuralgic amyotrophy. * TW Pdi less than 2 cm H2O and difficult to measure accurately.

Latency (ms)

10.5 8.6 8.7 6.4 9.6 13.8 13.2

Diagnosis ALS ALS NA Myopathy ALS Myopathy NA NA ALS ALS

1634

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

VOL 160

1999

plitude (r 5 0.89, p , 0.001) but no relationship between Tw Pdi and phrenic nerve conduction time (r 5 0.09), supporting the potential importance of measurement of the diaphragm CMAP in the evaluation of neuromuscular disease. In conclusion, the diaphragm CMAP amplitude can be measured using a multipair esophageal electrode. UMS combined with the multipair esophageal electrode could be a useful technique in the assessment of diaphragm function. References

Figure 6. Diaphragm CMAP elicited by UMS in patient with ALS. The amplitude of the CMAP was considerably lower than that measured from normal subjects whereas latency was within the normal value. Three traces superimposed.

producibility and reduced variability between subjects may be the result of precise electrode positioning at the EARdi center and reliable stimulation of the phrenic nerve. The amplitude of the CMAP is a useful parameter for evaluating diaphragm function (5) because it reflects the number of diaphragmatic muscle fibers that can be activated by phrenic nerve stimulation. Markand and coworkers (5) found the diaphragm CMAP in patients with ALS to be reduced and further speculated that it was also reduced in patients with muscular dystrophy. In the present study the amplitude of the diaphragm CMAP in patients with ALS or myopathy was substantially smaller than that of normal subjects, without significant prolongation of the phrenic nerve conduction time. Because reduced amplitude of the CMAP usually indicates axonal degeneration and prolonged conduction time reflects demyelination in peripheral nerves, measurement of the amplitude of the diaphragm CMAP has clinical importance. It has been shown that there is a relation between EMG and force for skeletal muscles (18). In the present study we found that there was a relationship between Tw Pdi and CMAP am-

1. Laghi, F., M. J. Harrison, and M. J. Tobin. 1996. Comparison of magnetic and electrical phrenic nerve stimulation in assessment of diaphragmatic contractility. J. Appl. Physiol. 80:1731–1742. 2. Gandevia, S. C., and D. K. McKenzie. 1986 Human diaphragmatic EMG: changes with lung volume and posture during supramaximal phrenic stimulation. J. Appl. Physiol. 60:1420–1428. 3. Mills, G. H., D. Kyroussis, C. H. Hamnegard, S. Wragg, J. Moxham, and M. Green. 1995. Unilateral magnetic stimulation of the phrenic nerve. Thorax 50:1162–1172. 4. Lieberman, S. L., R. R. Young, and J. M. Shefner. 1995. Neurological disorders affecting respiration. In C. Roussos, editor. Thorax. Marcel Dekker, New York. 85:2135–2175. 5. Markand, O. N., J. C. Kincaid, R. A. Pourmand, S. S. Moorthy, R. D. King, Y. Mahomed, and J. W. Brom. 1984. Electrophysiologic evaluation of diaphragm by transcutaneous phrenic nerve stimulation. Neurology 34:604–614. 6. Bellemare, F., and B. Bigland-Ritchie. 1984. Assessment of human diaphragm strength and activation using phrenic nerve stimulation. Respir. Physiol. 58:263–277. 7. McKenzie, D. K., and S. C. Gandevia. 1985. Phrenic nerve conduction times and twitch pressures of the human diaphragm. J. Appl. Physiol. 58:1496–1504. 8. Swenson, M. R., and R. S. Rubenstein. 1992. Phrenic nerve conduction studies. Muscle Nerve 15:597–603. 9. Mier, A., C. Brophy, J. Moxham, and M. Green. 1987. Phrenic nerve stimulation in normal subjects and in patients with diaphragmatic weakness. Thorax 42:885–888. 10. Luo, Y. M., M. I. Polkey, L. C. Johnson, R. A. Lyall, M. L. Harris, M. Green, and J. Moxham. 1998. Diaphragm EMG measured by cervical magnetic and electrical phrenic nerve stimulation. J. Appl. Physiol. 85: 2089–2099. 11. Bland, J. M., and D. G. Altman. 1986. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1(8476):307–310. 12. Mills, G. H., D. Kyroussis, C. H. Hamnegard, M. I. Polkey, M. Green, and J. Moxham. 1996. Bilateral magnetic stimulation of the phrenic nerves from an anterolateral approach. Am. J. Respir. Crit. Care Med. 154:1099–1105. 13. Luo, Y. M., L. C. Johnson, M. I. Polkey, M. L. Harris, R. A. Lyall, M. Green, and J. Moxham. 1999. Diaphragm electromyogram measured with unilateral magnetic stimulation. Eur. Respir. J. 13:385–390. 14. Beck, J., C. Sinderby, L. Lindstrom, and A. Grassino. 1997. Diaphragm interference EMG and compound muscle action potentials: effects of chest wall configuration. J. Appl. Physiol. 82:520–530. 15. Kim, M. J., W. S. Druz, J. Danon, W. Machnach, and J. T. Sharp. 1978. Effects of lung volume and electrode position on the esophageal diaphragmatic EMG. J. Appl. Physiol. 45:392–398. 16. Beck, J., C. Sinderby, L. Lindstrom, and A. Grassino. 1996. Influence of bipolar esophageal electrode positioning on measurements of human crural diaphragm electromyogram. J. Appl. Physiol. 81:1434–1449. 17. Similowski, T., B. Fleury, S. Launois, H. P. Cathala, P. Bouche, and J. P. Derenne. 1989. Cervical magnetic stimulation: a new painless method for bilateral phrenic nerve stimulation in conscious humans. J. Appl. Physiol. 67:1311–1318. 18. Bigland, B., and O.-C. J. Lippold. 1954. The relation between force, velocity and integrated electrical activity in human muscles. J. Physiol. (Lond.) 123:214–224.