INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Although most patients with chronic obstructive pulmonary disease (COPD) do not use the sternocleidomastoid muscles when breathing at rest (1), recent electromyographic studies have shown that these patients have increased firing frequencies in the parasternal intercostal, scalene, and diaphragmatic motor units as compared with control subjects (2, 3). In contrast to healthy individuals, many resting patients with severe COPD also contract the abdominal muscles, in particular the transversus abdominis, during expiration (4, 5). Thus, such patients have a greater than normal neural drive to the rib cage inspiratory muscles, the diaphragm, and the abdominal muscles, but the mechanism of this overall increase in drive remains uncertain.

It is well established that the inspiratory muscles of the rib cage are not severely affected by hyperinflation, and essentially maintain their ability to shorten (6). The increased activation of the parasternal intercostal and scalene muscles in COPD therefore results in a greater than normal elevation of the ribs and expansion of the rib cage (7). On the other hand, the diaphragm in such patients is low and flat. Consequently, irrespective of the degree of neural activation, the ability of the diaphragmatic dome to descend and to produce an increase in lung volume is impaired, and indeed, measurements of thoracoabdominal motion during breathing have repeatedly shown that many patients with severe COPD have a reduced outward displacement or a paradoxical inward displacement of the ventral abdominal wall during inspiration (7). Similarly, many patients with severe COPD are flow-limited at rest (8), and it therefore appears unlikely that an expiratory contraction of the transversus abdominis may cause a significant increase in expiratory flow or a significant reduction in end-expiratory lung volume (4). In severe COPD, the increased activation of the diaphragm and abdominal muscles would thus have limited beneficial effects on the act of breathing, and this has led to the suggestion that the overall increase in neural drive is simply an automatic response of the respiratory system to a greater than resting stimulation (4).

If this hypothesis were correct, it would imply that the pattern of respiratory-muscle activation in humans with chronic respiratory diseases would be relatively uniform and essentially independent of the nature of the disease. With this in mind, we examined the pattern of respiratory-muscle activation in a group of patients with thoracic scoliosis. Such patients resemble patients with COPD in that they also breathe against an increased load and have impaired diaphragmatic function (9). However, whereas COPD is primarily characterized by increased airflow resistance and hyperinflation, patients with scoliosis have a severe restrictive ventilatory impairment, with marked reductions in lung and chest-wall compliance.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The study was performed with eight patients (five men and three women) who were 37 to 60 yr of age (mean ± SD: 48.4 ± 7.1 yr) and had severe thoracic scoliosis. Their principal anthropometric characteristics are listed in Table 1. The scoliosis was congenital in two patients, idiopathic in three patients, and traumatic in three others; no patient, therefore, had any neuromuscular disease. Because the deformity in one patient was exclusively in the sagittal plane, the Lipmann- Cobb angle could be measured on anteroposterior (AP) chest X-ray films in only seven patients; in all of these, this angle was 100 ° or greater. Four patients were treated with nocturnal assisted ventilation, either noninvasively (Patients 2 and 4) or through a tracheostomy (Patients 3 and 8), and one patient (Patient 1) had undergone corrective surgery by the Harrington instrumentation method. All patients gave verbal informed consent to the study procedures as approved by the human studies committee of our institution.

Although all patients had been in a clinically stable condition for at least 4 wk before the study, they had a severe restrictive ventilatory defect. As shown in Table 2, their vital capacity (VC) and plethysmographic total lung capacity (TLC) were reduced to 25.3 ± 10.1% (mean ± SD) and 44.0 ± 8.4% of the predicted values (10), as calculated for body height corrected from arm span (11). The patients' functional residual capacity (FRC) and residual volume (RV) were also decreased to 60.1 ± 10.8% and 82.3 ± 19.8%, respectively of the predicted values. Their Pa_O2 during breathing of room air ranged from 48 to 75 mm Hg, and their Pa_CO2 was between 40 and 67 mm Hg. Maximal inspiratory mouth pressure (PI_max) measured during maximal static inspiratory efforts at FRC ranged from 38 to 60 cm H₂O, which corresponded to 61.1% to 91.6% of the predicted normal value (12).

All subjects were studied while seated in a comfortable high-backed armchair. Airflow was measured with a heated Fleisch pneumotachograph connected to a Validyne differential pressure transducer (Validyne Corp., Northridge, CA), and volume was obtained by electrical integration of the flow signal. After all garments were removed, a pair of linearized magnetometers (N.H. Peterson, Boston, MA) was attached to the midline 2 cm above the umbilicus to measure the respiratory changes in the AP diameter of the abdomen. Esophageal (Pes) and gastric (Pga) pressures were measured with conventional balloon-catheter systems placed in the midesophagus and the stomach, respectively; the esophageal balloon was filled with 0.5 ml of air and the gastric balloon contained 2.0 ml of air. The transdiaphragmatic pressure (Pdi = Pga  Pes) generated during maximum sniffs at FRC was measured in four patients (Patients 1 and 4 to 6), and the values ranged between 49 cm H₂O and 66 cm H₂O (normal values > 90 cm H₂O [13]).

When the magnetometers and balloon-catheter systems were in place, concentric needle electrodes were implanted into the sternocleidomastoid muscle, midway between the angle of the jaw and the clavicle, and into the scalene muscle, 0.5 to 1.0 cm above the clavicle, just behind the clavicular fibers of the sternocleidomastoid. The side selected for investigation was usually the one on which the distance between the angle of the jaw and the clavicle was the largest (making the muscles more accessible), but in two patients, recordings of sternocleidomastoid and scalene electromyographic activity were obtained on both sides of the neck. Concentric needle electrodes were also inserted in the rectus abdominis, external oblique, and transversus abdominis muscles with the aid of a high-resolution, 5-MHz linear ultrasound probe (Acuson 128 computed sonography system; Acuson, Inc., Mountain View, CA). This technique has previously been described in detail (14). The probe was first placed 2 to 3 cm above the umbilicus, 2 to 3 cm from the midline, to insert an electrode into the rectus abdominis muscle. When this electrode was in place, the probe was moved to the right or left anterior axillary line to locate and visualize the three muscle layers of the lateral abdomen. Under continuous sonographic monitoring, two electrodes were then inserted perpendicular to the skin, 2 to 3 cm apart, and advanced progressively until their tips were embedded in the external oblique and transversus abdominis muscles, respectively. All electromyographic signals were processed with amplifiers (PA 63; Medelec, Surrey, UK), and filtered below 80 Hz and above 800 Hz.

None of the patients suffered discomfort attributable to the needles. The patients were nevertheless allowed to recover for 10 to 15 min after instrumentation so that variations in their breathing patterns could stabilize. The respiratory changes in abdominal AP diameter, Pes, Pga, and electromyographic activity during resting breathing were then recorded for 3 to 4 min. The patient was subsequently connected to the mouthpiece assembly, and another set of measurements was obtained. The patient was then engaged in a quiet conversation in the hope that his or her attention would be diverted from the study. He or she was then asked to remain silent for 2 to 3 min, after which another period of resting breathing was recorded. At least five periods with and five periods without the mouthpiece were recorded for each subject.

Tidal volume (VT), breathing frequency (f), inspiratory duration (TI), inspiratory duty cycle (TI/Ttot), and dynamic pulmonary compliance (Cdyn) were averaged over at least 10 consecutive breaths from each run, and the results were compared with the predicted normal values reported by Sorli and colleagues (15). In addition, the change in Pga resulting from contraction of the abdominal muscles during expiration was assessed by measuring the pressure difference between expiration and the initial part of inspiration, as described in our previous study (5). The respiratory changes in Pga and abdominal AP diameter were also displayed as X-Y plots on a Tektronix 5111 storage oscilloscope (Tektronix Corp., Beaverton, OR) and photographed with a Polaroid camera.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pattern of Breathing and Pulmonary Compliance

The patients had a rapid, shallow breathing pattern with a VT of 0.367 ± 0.110 L (mean ± SD) (predicted: 0.778 ± 0.084 L) and a f of 24.2 ± 6.4 min¹ (predicted: 17.2 ± 2.7 min¹). TI/ Ttot was within normal limits (0.40 ± 0.05) because the increased f resulted from a proportional decrease in TI (1.07 ± 0.33 s) and Te (1.57 ± 0.42 s). However, Cdyn was markedly reduced, and amounted to 0.040 ± 0.026 L/cm H₂O.

Neck Muscle Activity

The electromyographic data obtained from the eight patients are summarized in Table 3. When breathing at rest, all patients invariably had phasic inspiratory activity in the scalene muscles. As shown by the records of two representative patients in Figure 1, this activity started together with the onset of inspiration, involved many motor units, and reached its peak at the end of inspiration. In contrast, only one patient showed a similar pattern in the sternocleidomastoid muscles (Figure 1, top). Indeed, four patients did not show any electromyographic activity at all in the sternocleidomastoids during breathing (Figure 1, bottom), and although some inspiratory electromyographic activity was detected in the other three patients, this activity was intermittent, being present only when the patient was connected to the mouthpiece assembly. Even then, sternocleidomastoid activity consisted of only a few discharges originating from one or two motor units. No difference was seen between the right and left side of the neck in the two patients studied.

View larger version (14K): [in this window] [in a new window]   Figure 1.   Records of Pes, abdominal AP diameter (increase upward), and electromyographic activity of the scalene (SCA) and sternocleidomastoid (SM) muscles during unencumbered breathing in Patient 8 (top panel ) and Patient 4 (bottom panel ).

Abdominal Muscle Activity

The rectus abdominis and external oblique muscles were electrically silent throughout the study in four patients (Table 3). Intermittent phasic expiratory activity was detected in the rectus abdominis in four patients and in the external oblique in three patients; only one patient showed invariable phasic expiratory activity in the external oblique. On the other hand, five patients had invariable phasic expiratory activity in the transversus abdominis. As illustrated by the records of one representative patient in Figure 2, this activity persisted throughout the expiratory phase of the breathing cycle, and involved many motor units, and was detected whether the patient was breathing freely or through the mouthpiece, although it tended to increase in magnitude in the latter condition. The other three patients also showed phasic expiratory activity in the transversus abdominis, but only after introduction of the mouthpiece.

View larger version (11K): [in this window] [in a new window]   Figure 2.   Records of Pes, Pga, abdominal AP diameter, and electromyographic activity of the transversus abdominis (TA), rectus abdominis (RA), and external oblique (EO) muscles obtained in Patient 8 during breathing on the mouthpiece.

Gastric Pressure and Abdominal Motion

Most patients showed an increase in Pga during expiration (Table 3), and Figure 3 shows plots of abdominal AP diameter versus Pga during resting breathing in two patients. In Patient 7 (left panel ), the abdominal muscles were electrically silent during unencumbered breathing, and the increase in Pga during inspiration was associated with an increase in abdominal AP diameter, whereas the decrease in Pga during expiration occurred together with a decrease in abdominal AP diameter. Patients 2 and 5 behaved similarly. In contrast, Patient 8 had clear-cut phasic expiratory activity in the transversus (Table 3), and most of the increase in Pga occurred during the expiratory, rather than during the inspiratory, phase of the breathing cycle. As a result, the plot of abdominal AP diameter versus Pga had a figure-eight pattern (Figure 3, right panel ). A qualitatively similar pattern was seen in Patients 2, 5, and 7 when they were connected to the mouthpiece, and in the other five patients irrespective of the presence or absence of the mouthpiece. Although these plots reflected well the pattern of abdominal muscle activity, they were not related to the severity of the patients' thoracic deformity.

View larger version (8K): [in this window] [in a new window]   Figure 3.   Plots of the respiratory changes in abdominal AP diameter versus Pga in Patient 7 (left panel ) and Patient 8 (right panel ). In each panel, the open circle corresponds to end-expiration; the arrow indicates the direction of the loop and marks the inspiratory phase of the breathing cycle. Calibration on Pga = 5 cm H2O.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All patients in this study had a severe thoracic deformity, and in the seven patients in whom it could be measured, the Cobb angle was 100° or greater. In addition, the patients had a marked restrictive ventilatory impairment, with an important reduction in dynamic pulmonary compliance and a rapid, shallow breathing pattern. Even though none of the patients had any neuromuscular disease, values of maximum static inspiratory pressures and sniff Pdi were moderately reduced. Six patients also had chronic hypercapnia, and four of them were treated with nocturnal assisted ventilation. These patients therefore had the clinical and functional characteristics of patients with advanced thoracic scoliosis (9, 16).

Previous electromyographic studies have shown that there is a marked difference in the recruitment of the scalene and sternocleidomastoid muscles in humans. Campbell (21) first reported that normal subjects breathing at rest frequently use the scalene but not the sternocleidomastoids during inspiration. Delhez (22) and Raper and colleagues (23) have also shown that the scalenes are active during eupnea in virtually all healthy individuals, whereas the sternocleidomastoids are invariably silent. In fact the sternocleidomastoids were recruited only at the end of a maximal inspiration and during maximal voluntary ventilation (22, 23). More recent studies with a group of 40 stable patients with severe COPD have shown a similar difference; all patients had strong phasic inspiratory activity in the scalenes during resting breathing, but only four also had inspiratory activity in the sternocleidomastoids (1). On the basis of these observations, one would thus conclude that the threshold of activation of the sternocleidomastoids in humans is much higher than that of the scalenes, and one would therefore speculate that patients with chronic restrictive ventilatory disorders, including patients with thoracic scoliosis, would also recruit the scalenes much earlier than the sternocleidomastoids. As shown in Figure 1, this is exactly what we observed; whereas all patients had high-amplitude phasic inspiratory activity in the scalenes, only one had invariable phasic inspiratory activity in the sternocleidomastoids.

Although normal humans do not use the muscles of the anterolateral wall of the abdomen during resting breathing, these muscles also show differences in their recruitment threshold. When healthy subjects increase their ventilation, such as during hyperoxic hypercapnia, they recruit the transversus abdominis during expiration well before activity can be recorded from the rectus abdominis or external oblique (14, 24). Expiratory activation of the transversus abdominis with little or no activity in the rectus or the external oblique also occurs when normal subjects breathe against increased inspiratory mechanical loads (14). In the absence of expiratory flow limitation, contracting the transversus abdominis during expiration is an appropriate response to these challenges because the associated reduction in end-expiratory lung volume allows the increased work of breathing to be shared between the inspiratory and expiratory muscles. However, many patients with severe COPD also show isolated contraction of the transversus abdominis during expiration (4, 5). Because such patients are flow-limited at rest (8), this contraction is unlikely to have a significant impact on expiratory flow and end-expiratory lung volume (4). Thus, the transversus abdominis has a lower threshold of activation than the rectus or the external oblique in humans, and its recruitment during expiration appears to be independent of its potential benefit to the act of breathing.

As for the inspiratory muscles of the neck, the observed behavior of the abdominal muscles in the patients in the present study agreed well with this conclusion. Indeed, many patients showed phasic expiratory activity in the transversus abdominis without any activity in the rectus and external oblique (Figure 2), and this contraction was mechanically significant, since the decrease in abdominal AP diameter during expiration was commonly associated with an increase in gastric pressure (Figure 3). However, because of their thoracic deformity, patients with severe thoracic scoliosis have a marked reduction in FRC and a small expiratory reserve volume; this volume was only 0.24 L for the eight patients in this study. In addition, the restrictive ventilatory impairment in these patients is such that their breathing frequency is markedly increased and their expiratory time substantially reduced. Consequently, the expiratory branch of the tidal flow-volume loop in such patients is very close to, or superimposed on, the maximal flow-volume curve, and there is little or no possibility of increasing expiratory flow (25) and reducing the end-expiratory lung volume below the neutral position of the respiratory system.

In conclusion, the present study has established that the pattern of respiratory-muscle activation during resting breathing in patients with severe thoracic scoliosis is similar to that seen in normal subjects with increased ventilatory drive and in patients with advanced COPD. This pattern of activation is therefore neither load- nor disease-specific. The current observations thus provide strong support for the concept that this pattern of respiratory-muscle activation is essentially an automatic response of the central respiratory controller.