Spatial distribution of external and internal intercostal activity in dogs

Abstract

1. The observation that the external and internal interosseous intercostal muscles in the dog show marked regional differences in mechanical advantage has prompted us to re-examine the topographic distribution of electrical activity among these muscles during spontaneous breathing.

2. Inspiratory activity was recorded only from the areas of the external intercostals with an inspiratory mechanical advantage, and expiratory activity was recorded only from the areas of the internal intercostals with an expiratory mechanical advantage. The expiratory discharges previously recorded from the caudal external intercostals and the inspiratory discharges recorded from the rostral internal intercostals were probably due to cross-contamination.

3. Activity in each muscle area was also quantified relative to the activity measured during tetanic, supramaximal nerve stimulation (maximal activity). External intercostal inspiratory activity was consistently greater in the areas with a greater inspiratory advantage (i.e. the dorsal aspect of the rostral segments) than in the areas with a smaller inspiratory advantage, and internal intercostal expiratory activity was invariably greatest in the areas with the greatest expiratory advantage (i.e. the dorsal aspect of the caudal segments).

4. This topographic distribution of neural drive confers to the external intercostal muscles an inspiratory action on the lung during breathing and to the internal interosseous intercostals an expiratory action.

The current conventional view of intercostal muscle actions is based on the theory of Hamberger (1749) and maintains that as a result of the orientation of the muscle fibres, the external intercostals have an inspiratory action on the lung whereas the internal interosseous intercostals have an expiratory action. In the preceding paper, however, we have shown that in supine dogs, these muscles show marked topographic differences in mechanical advantage (De Troyer et al. 1999). Specifically, the external intercostals in the dorsal third of the rostral interspaces were found to have a large inspiratory mechanical advantage (i.e. a great ability to cause lung inflation), but this inspiratory mechanical advantage decreases rapidly toward the costochondral junctions and toward the base of the rib cage. Consequently, the muscles in the ventral portion of the caudal segments have an expiratory, rather than inspiratory mechanical advantage. The internal intercostal muscles in the dorsal portion of the caudal interspaces have a large expiratory mechanical advantage, but this advantage decreases ventrally and cranially such that in the most rostral interspaces, it is reversed into an inspiratory mechanical advantage (De Troyer et al. 1999). These results imply that the actions of these muscles on the lung during breathing are largely determined by the topographic distribution of activity among them, rather than the orientation of the muscle fibres.

A number of electrical recordings from intercostal muscles and nerves in anaesthetized cats (Sears, 1964; Bainton et al. 1978; Kirkwood et al. 1982, 1984; Greer & Martin, 1990) and dogs (De Troyer & Ninane, 1986) have shown that the external intercostals are active during inspiration. The muscles also appeared to display greater inspiratory activity in the rostral than in the caudal segments and greater inspiratory activity in the dorsal than in the ventral portion of the rib cage. In contrast, the internal intercostals were electrically active during expiration and displayed greater activity in the caudal than the rostral segments (Bainton et al. 1978; De Troyer & Ninane, 1986; Greer & Martin, 1990). In view of the distributions of mechanical advantage, such distributions of activity would suggest that the external intercostals have an inspiratory action on the lung during breathing and that the internal intercostals have an expiratory action. However, these descriptions of electrical activity are qualitative, and the sites where the recordings were made were not standardized. Consequently, the correspondence between the distributions of activity and the distributions of mechanical advantage can only be approximate, and no estimates can be made of the pressures contributed by the different muscle areas during breathing. More importantly, efferent discharges to the external intercostals in the caudal segments during expiration and to the internal intercostal in the second interspace during inspiration have also been recorded in decerebrate cats (Le Bars & Duron, 1984). As these discharges were recorded in animals performing forceful respiratory efforts against an occluded trachea, the possibility exists that they were the result of cross-contamination between the two muscle layers (De Troyer & Ninane, 1986). Yet such discharges would match, respectively, the expiratory mechanical advantage of the external intercostal muscles in the caudal segments and the inspiratory mechanical advantage of the internal intercostals in the most rostral segments, and this raises the possibility that these muscles may not have distinct effects on the lung during breathing.

These issues prompted us to re-examine in detail the spatial distribution of activity among the canine external and internal intercostal muscles. The pattern of activation of the external intercostals in the caudal segments and of the internal intercostals in the most rostral segments was studied first. Selective muscle denervations were performed so that any cross-contamination could be identified. The distributions of external and internal intercostal activity along the rostrocaudal and dorsoventral axes of the rib cage were next assessed quantitatively by comparing for each muscle area the amount of activity recorded during breathing with that recorded during supramaximal, tetanic stimulation of the motor nerve. As the recordings were made in muscle areas with well-defined mechanical advantages (De Troyer et al. 1999), the respiratory function of each muscle could therefore be definitely established and the pressures contributed by the various areas could be estimated.

METHODS

The studies were carried out on twenty-three adult mongrel dogs (12-29 kg body weight) anaesthetized with pentobarbitone sodium (initial dose, 25 mg kg⁻¹i.v.). The animals were placed in the supine posture and intubated with a cuffed endotracheal tube, and a venous cannula was inserted in the forelimb to give maintenance doses of anaesthetic (1-2 mg kg⁻¹i.v.). A cathether was also inserted in a femoral artery to monitor blood pressure and sample arterial blood periodically for blood gas analysis, after which the rib cage and intercostal muscles were exposed on the right side of the chest from the first to the tenth rib by deflection of the skin and underlying muscle layers. Three experimental protocols were followed, as approved by the Animal Ethics and Welfare Committee of the Brussels School of Medicine.

Experiment 1

Six animals were studied to assess the activation of the external intercostals with an expiratory mechanical advantage. In each animal, two pairs of stainless-steel hook electrodes spaced 3-4 mm apart were thus placed in the ventral third of the eighth intercostal space; one pair was inserted into the external intercostal muscle, and the other was inserted 1 cm more ventral into the exposed internal intercostal. Each pair of electrodes was placed in parallel fibres. The two electromyographic (EMG) signals were processed by using amplifiers (model 830/1, CWE Inc., Ardmore, PA, USA), bandpass filtered below 100 Hz and above 2000 Hz, and rectified before their passage through leaky integrators with a time constant of 0.2 s. In addition, changes in lung volume were measured by electronic integration of the flow signal derived from a heated Fleisch pneumotachograph and a Validyne differential pressure transducer (Validyne, Northridge, CA, USA), and airway pressure was measured with another differential pressure transducer connected to a side port of the endotracheal tube.

Recordings of EMG activity, lung volume and airway pressure were obtained first during resting, room air breathing. The animal was then connected to a Hans-Rudolph valve and given a gas mixture containing 6 % CO₂ in 50 % O₂, 44 % N₂. After returning to room air, the expiratory port of the valve was placed under water at various depths so as to apply positive end-expiratory pressures (PEEP) of 5, 10, 15 and 20 cmH₂O. Phasic expiratory activity was recorded from the external intercostal muscle in all animals, in agreement with the previous observation of Le Bars & Duron (1984). After completion of the control procedure, therefore, the external intercostal nerve in the eighth interspace was exposed and sectioned at the rib angle, leaving the internal intercostal nerve intact. A second set of recordings was then made. A last set of measurements was obtained after section of the internal intercostal nerve.

The animal was subsequently made apnoeic by mechanical hyperventilation, and the peripheral end of the cut internal intercostal nerve was laid over a bipolar stimulating electrode. Pulses of 0.2 ms duration were delivered at intervals of 1 s, and stimulus intensity was progressively increased until it was 20-30 % greater than that required to produce a compound muscle action potential (CMAP) of maximal amplitude. As CMAPs were recorded from both muscles, the portion of external intercostal muscle being studied was carefully incised over 2-3 cm along its caudal insertions, so that it could be lifted off the plane of the internal intercostal. A second series of CMAPs was recorded in this condition. The entire procedure was repeated in the ventral third of the sixth intercostal space in three animals and in the ventral third of the tenth intercostal space in two animals. For the six animals studied, a total of eleven areas where the external intercostals have a clear-cut expiratory mechanical advantage (De Troyer et al. 1999) was thus investigated.

In each animal, the two pairs of recording electrodes were then transferred to the external and internal intercostal muscles in the middle or the ventral third of the second interspace, that is in an area where the internal intercostals have an inspiratory, rather than expiratory, mechanical advantage (De Troyer et al. 1999). As in the caudal segments, the internal intercostal electrode was placed ∼1 cm ventral to the external intercostal electrode, midway between the two ribs making up the interspace. Recordings were made first during resting breathing, after which the animal was given a succession of increased inspiratory airflow resistances; the four resistors thus used were 1-cm-long Plexiglass cylinders through which holes of 4, 3, 1.75 and 1 mm diameter had been bored. As phasic inspiratory EMG activity was recorded from both the external and the internal intercostal muscles (see Results), recordings were repeated after denervation of the external intercostals. All measurements were obtained during steady-state conditions (i.e. 10 min after the gas mixture had been given or after the change in airway pressure had stabilized), and the animals were maintained at a reasonable depth of anaesthesia throughout. Thus they had no spontaneous movement of the fore- or hindlimbs, and although the corneal reflex was kept present, they had no pupillary light reflex and no flexor withdrawal of the forelimbs. Rectal temperature was also maintained constant between 36 and 38°C with infrared lamps. At the end of the experiment, the animals were given an overdose of pentobarbitone sodium (40-50 mg kg⁻¹i.v.).

Experiment 2

The distribution of external intercostal activity along the rostrocaudal and dorsoventral axes of the rib cage was subsequently studied in eight animals. In each animal, the length of the second, fourth, sixth and eighth intercostal spaces was measured from the angle of the rib dorsally to the costochondral junctions ventrally, and four pairs of hook electrodes were implanted into the external intercostal muscle in the middle portion of the dorsal third of each interspace. Two additional electrode pairs were placed in the middle portion of the middle third and the ventral third of the fourth interspace.

As in Expt 1, measurements were obtained first during resting, room air breathing in the supine posture. Three periods of resting breathing were recorded in each animal. The animal was then repositioned on the board in the prone posture, and two additional periods of resting breathing were recorded. In the supine posture, the animal was subsequently given a succession of CO₂-enriched gas mixtures (0 % CO₂ in 50 % O₂, 50 % N₂; 2 % CO₂ in 50 % O₂, 48 % N₂; 4 % CO₂ in 50 % O₂, 46 % N₂; and 6 % CO₂ in 50 % O₂, 44 % N₂) and a series of inspiratory airflow resistances (four resistors), after which it was made apnoeic by supplemental anaesthesia (pentobarbitone sodium, 5-8 mg kg⁻¹i.v.) and mechanical hyperventilation. The external intercostal nerve in the second, fourth, sixth and eighth interspace was then exposed carefully in the vicinity of the rib angle. Each nerve was positioned across a bipolar stimulating electrode, and pulses of 0.2 ms duration and supramaximal voltage were delivered at intervals at a frequency of 50 impulses s⁻¹ so as to produce maximal, simultaneous contraction of all external intercostal muscle fibres. The animals were subsequently given an overdose of pentobarbitone sodium (40-50 mg kg⁻¹i.v.).

Phasic inspiratory EMG activity in each area of external intercostal muscle was quantified by measuring the peak height of the integrated EMG signal in arbitrary units. Activity in each condition was averaged over ten consecutive breaths from each run, and it was then expressed as a percentage of the activity recorded during tetanic, supramaximal stimulation of the corresponding external intercostal nerve (maximal activity). The data were finally averaged for the animal group, and they are presented as means ±s.e.m. Statistical comparisons between the amounts of inspiratory EMG activity recorded in the four interspaces studied and between the dorsal third, the middle third, and the ventral third of the fourth interspace were made by analysis of variance (ANOVA) with repeated measures, and multiple comparison testing of the mean values was performed, when appropriate, using Student-Newman-Keuls tests. The criterion for statistical significance was taken as P < 0.05.

Experiment 3

The topographic distribution of internal intercostal activity was assessed in nine animals. As in Expt 2, the length of the second, fourth, sixth and eighth intercostal spaces was measured from the angle of the ribs to the costochondral junctions in each animal. Small segments of external intercostal muscle in the dorsal third of each interspace and in the middle third and the ventral third of the eighth interspace were then sectioned along their caudal insertions, and six pairs of hook electrodes were implanted into the internal intercostals; each pair of electrodes was also placed in parallel fibres.

The experimental and analytical procedure was essentially similar to that described in Expt 2. Thus measurements of EMG activity, lung volume and airway pressure were obtained: (1) during resting breathing in the supine posture (three periods); (2) during resting breathing in the prone posture (two periods); (3) during hyperoxic hypercapnia (four gas mixtures); and (4) during breathing against PEEP of 5, 10, 15 and 20 cmH₂O. The internal intercostal nerves in the four interspaces studied were subsequently exposed at the rib angle and tetanically stimulated during hyperventilation-induced apnoea (maximal activity), after which the animals received an overdose of pentobarbitone sodium.

RESULTS

External intercostal activity during expiration (Expt 1)

Representative records of external and internal intercostal activity in the ventral third of the eighth interspace are shown in Fig. 1. The six animals studied had phasic expiratory activity in the internal intercostal muscle during resting breathing (arterial O₂ pressure, P_a,O2: 86.2 ± 2.4 Torr; arterial CO₂ pressure, P_a,CO2: 41.7 ± 1.8 Torr), and four of them also showed phasic expiratory EMG activity in the external intercostal (Fig. 1A). The two remaining animals developed external intercostal expiratory activity during hyperoxic hypercapnia and during graded expiratory threshold loading, in agreement with the previous observation in cats by Le Bars & Duron (1984). However, this external intercostal expiratory activity remained unchanged after section of the external intercostal nerve (Fig. 1B) and disappeared only after section of the internal intercostal nerve and abolition of expiratory activity in the internal intercostal muscle (Fig. 1C). Essentially similar recordings were obtained from the sixth and tenth interspaces, thus indicating that the expiratory discharges detected in the external intercostal muscles of the caudal segments were due to impulses travelling along the internal, rather than the external, intercostal nerves.

Figure 1. Traces of EMG activity recorded from the external and internal intercostal muscles in the ventral third of the eighth interspace in a representative animal.

Records obtained during resting, room air breathing. A, in the control condition; B, after section of the external intercostal nerve; C, after section of the internal intercostal nerve. Note the persistence of phasic expiratory activity in the external intercostal after section of the external intercostal nerve.

The recordings obtained during stimulation of the internal intercostal nerves confirmed this conclusion, and they further eliminated the possibility that these nerves might provide motor supply to the external intercostal muscles in the caudal segments. Indeed, when the muscles were intact, stimulating the peripheral end of the cut internal intercostal nerve elicited a CMAP in both intercostal muscles (Fig. 2A and B), but when the external intercostal was lifted a few millimetres off the plane of the internal intercostal, the external intercostal CMAP was abolished (Fig. 2C). The CMAP in this muscle was restored as soon as the muscle was replaced in its initial, anatomical position (Fig. 2D). The expiratory discharges detected in the external intercostal muscles of the caudal segments of the rib cage, therefore, were exclusively the result of a cross-contamination from the underlying internal intercostals.

Figure 2. Compound muscle action potentials recorded during stimulation of the internal intercostal nerve in the eighth interspace.

The action potentials in A and B were recorded, respectively, from the internal and external intercostal muscles in the control condition. Those in C and D were recorded from the external intercostal after the muscle was separated from the internal intercostal (C) and after it was replaced in its initial position (D). Four consecutive traces have been superimposed in each panel.

Internal intercostal activity during inspiration

Phasic inspiratory EMG activity was recorded in all animals from both the external intercostal and the internal intercostal muscle in the middle third of the second interspace. Activity in both muscles was present already during resting breathing and increased after the addition of increased inspiratory airflow resistances (Fig. 3A). However, monitoring of these activities on a loudspeaker suggested that the discharges recorded in the internal intercostal originated from a more distant source than those recorded in the external intercostal, and indeed, when the external intercostal nerve in the second interspace was sectioned, inspiratory activity in both muscles was suppressed or markedly reduced. The low-voltage inspiratory activity that persisted during resistive breathing in four animals was eliminated after the external intercostals in the first and third interspaces were also denervated (Fig. 3B). Yet the internal intercostal nerve in this condition was still intact, and its stimulation continued to elicit a CMAP. Thus, this muscle was not contracting at all during breathing, and the inspiratory discharges that were initially recorded were entirely due to a cross-contamination from the external intercostals.

Figure 3. Traces of EMG activity recorded from the external and internal intercostal muscles in the middle third of the second interspace.

Records obtained in a representative animal during resting breathing and after the addition of an increased inspiratory airflow resistance (arrow). A, control condition; B, after denervation of the external intercostal muscles in the first, second and third interspaces.

External intercostal inspiratory activity during resting breathing (Expt 2)

Representative records of external intercostal EMG activity in the dorsal third of the second, fourth, sixth and eighth interspaces during resting, room air breathing in the supine posture and during supramaximal, tetanic stimulation of the external intercostal nerves are shown in Fig. 4. When breathing at rest in the supine posture (P_a,O2: 91.2 ± 3.3 Torr; P_a,CO2: 38.4 ± 1.6 Torr), the eight animals had phasic inspiratory EMG activity in the second and fourth interspaces. However, only six animals had external intercostal inspiratory activity in the sixth interspace, and none had inspiratory activity in the eighth interspace. In addition, when the peak heights of the integrated EMG signals recorded during breathing were compared with those recorded during nerve stimulation, the magnitude of inspiratory activity decreased progressively and markedly from the second to the sixth interspace in every animal (Fig. 5A). For the animal group, peak inspiratory EMG activity in the second interspace thus averaged 46.0 ± 10.7 % of maximal activity, whereas activity in the fourth and sixth interspace was only 20.4 ± 6.7 and 6.5 ± 2.8 % of maximal activity, respectively (P < 0.001).

Figure 4. Traces of EMG activity (integrated signals) recorded from the external intercostal muscles in the dorsal third of the second, fourth, sixth and eighth interspaces in a representative animal.

The records in A were obtained during resting, room air breathing in the supine posture, and those in B were obtained during tetanic, supramaximal stimulation of the external intercostal nerves. The numbers in parentheses in B indicate the changes in the amplifier attenuations between records A and B. Note that the amount of inspiratory activity during breathing, relative to the activity recorded during stimulation, decreases progressively from the second to the eighth interspace.

Figure 5. Distribution of external and internal intercostal activity during resting breathing in the supine posture.

Data for the external intercostals (hatched bars) are mean values ±s.e.m. obtained from eight animals; data for the internal intercostals (open bars) are mean values ±s.e.m. obtained from nine animals. A, activity recorded from the dorsal third of the second, fourth, sixth and eighth interspaces; B, activity recorded from the dorsal third, the middle third and the ventral third of the external intercostal in the fourth interspace and the internal intercostal in the eighth interspace. Each activity is expressed as a percentage of the activity recorded during tetanic, supramaximal stimulation of the external or internal intercostal nerve (maximal activity). Note that the amount of external intercostal inspiratory activity decreases gradually from the second to the eighth interspace (A) whereas the amount of internal intercostal expiratory activity progressively increases. Note also that in a given interspace, both the amount of external intercostal inspiratory activity and the amount of internal intercostal expiratory activity decreases from the dorsal to the ventral portion (B).

External intercostal inspiratory activity also decreased gradually (P < 0.005) from the dorsal to the ventral portion of the fourth interspace (Fig. 5B). In fact, whereas inspiratory activity was recorded from the dorsal portion in all animals, it was detected from the middle portion in only six animals and from the ventral portion in only one animal. These rostrocaudal and dorsoventral gradients of inspiratory activity among the external intercostals remained unchanged when the animals were placed in the prone posture.

Internal intercostal expiratory activity during resting breathing (Expt 3)

In the supine posture, the nine animals of Expt 3 had phasic expiratory EMG activity in the internal intercostal muscles in the dorsal third of the sixth and eighth interspaces. In contrast, internal intercostal expiratory activity was recorded from the dorsal third of the fourth interspace in only three animals, and when these activities were compared with those recorded during stimulation of the internal intercostal nerves, activity in every animal showed a gradual decrease from the eighth to the fourth interspace (Fig. 5A; P < 0.001). Internal intercostal expiratory activity decreased also from the dorsal to the ventral portion of the eighth interspace (Fig. 5B; P < 0.005). Expiratory activity was not recorded from the internal intercostal in the second interspace in any animal.

Although tilting the animals from the supine to the prone posture elicited moderate increases in internal intercostal expiratory activity in the dorsal third of the sixth interspace and in the different areas of the eighth interspace (P < 0.05 for each), the spatial distribution of activity was unaffected. Internal intercostal expiratory activity in prone animals still decreased from the dorsal third of the eighth interspace to both the dorsal third of the fourth interspace and the ventral third of the eighth interspace (P < 0.005). As in the supine posture, internal intercostal expiratory activity in the second interspace was not recorded in any prone animal.

Effects of hypercapnia

The effects of hyperoxic hypercapnia on the gradients of external and internal intercostal activity are summarized in Fig. 6. As anticipated, increasing the CO₂ concentration in the inspired air caused progressive increases in both external intercostal inspiratory activity and internal intercostal expiratory activity. However, the amount of external intercostal activity continued invariably to decrease from the second to the sixth interspace (P < 0.005; Fig. 6A) and from the dorsal to the ventral portion of the fourth interspace (P < 0.005; Fig. 6B). The amount of internal intercostal activity also continued to decrease progressively from the dorsal portion of the eighth interspace to both the dorsal portion of the fourth interspace (P < 0.005; Fig. 6C) and the ventral portion of the eighth interspace (P < 0.005; Fig. 6D). As during room air breathing, the external intercostal in the eighth interspace never showed any inspiratory activity, and the internal intercostal in the dorsal portion of the second interspace showed a few expiratory discharges in only one of nine animals.

Figure 6. Effects of hyperoxic hypercapnia on the distributions of external and internal intercostal activity.

Data for the external intercostals (A and B) are mean values ±s.e.m. obtained from eight animals; data for the internal intercostals (C and D) are mean values ±s.e.m. obtained from nine animals. Same conventions as in Fig. 5. Note that the amount of external intercostal inspiratory activity always decreases from the second to the eighth interspace (A) and from the dorsal to the ventral portion of the fourth interspace (B). The amount of internal intercostal expiratory activity also invariably decreases from the dorsal to the ventral portion (D) but increases from the second to the eighth interspace (C).

Effects of increased mechanical loads

Increasing the inspiratory airflow resistance elicited more dramatic increases in external intercostal inspiratory activity than did hyperoxic hypercapnia; with the highest two resistances, phasic inspiratory activity was also recorded from the dorsal portion of the eighth interspace in one of eight animals. Similarly, expiratory threshold loads caused substantial increases in internal intercostal expiratory activity in all interspaces and elicited activity in the dorsal portion of the second interspace in two of nine animals. Yet the gradients of external and internal intercostal activity along the rostrocaudal and the dorsoventral axes were maintained, as shown in Fig. 7 (P < 0.01 or less).

Figure 7. Effects of increased mechanical loads on the distributions of external and internal intercostal activity.

A and B show the effects of graded inspiratory airflow resistance (1 to 4) on the distribution of external intercostal inspiratory activity (mean values ±s.e.m. obtained from eight animals). C and D show the effects of graded expiratory threshold load (from 5 to 20 cmH₂O positive end-expiratory pressure) on the distribution of internal intercostal expiratory activity (mean values ±s.e.m. obtained from nine animals). Same conventions as in Fig. 5.

DISCUSSION

The present studies have provided the quantitative confirmation that neural drive to the external and internal interosseous intercostal muscles is distributed along rostrocaudal and dorsoventral gradients and the demonstration that these gradients mirror specific gradients of mechanical advantage. Whether during resting breathing in the supine or prone posture (Fig. 5), during CO₂-induced hyperpnoea (Fig. 6), or during breathing against increased inspiratory airflow resistance (Fig. 7), external intercostal inspiratory activity was greatest in the dorsal portion of the most rostral segments, that is in the areas of the rib cage where the muscles have the greatest inspiratory mechanical advantage (De Troyer et al. 1999). External intercostal inspiratory activity then declined rapidly and gradually in both the caudal and the ventral direction, as does the muscle inspiratory mechanical advantage. As a result, no inspiratory activity was ever recorded from the ventral portion of the sixth and eighth interspaces where the muscles have an expiratory mechanical advantage. Since it was also demonstrated that the expiratory discharges previously recorded from these areas in the cat (Le Bars & Duron, 1984) were exclusively the result of cross-contamination (Figs 1 and 2), the conclusion must therefore be drawn that the external intercostals have a definite inspiratory action on the lung during breathing.

On the other hand, expiratory activity among the internal interosseous intercostals consistently predominated in the dorsal portion of the caudal segments, that is in the areas where the muscles have the greatest expiratory mechanical advantage (De Troyer et al. 1999). As for the expiratory mechanical advantage, internal intercostal expiratory activity then decreased progressively in the cranial and ventral directions, such that no expiratory activity was detected in the middle and ventral portions of the most rostral segments where the muscles have an inspiratory mechanical advantage. These areas, however, did not show any inspiratory activity either, including during breathing against very high inspiratory airflow resistances (Fig. 3). This precise matching between the topographical distribution of expiratory neural drive and the topographical distribution of expiratory mechanical advantage among the internal intercostals must confer to these muscles an expiratory action on the lung during breathing. Although it has long been thought that the external and internal interosseous intercostals have opposite actions on the lung during breathing (Hamberger, 1749), it was not appreciated that these are primarily the result of selective regional activation of the muscles, rather than the orientations of the muscle fibres.

Contribution of the external intercostals to lung expansion

To the extent that the values of mechanical advantage reported in the preceding paper (De Troyer et al. 1999) and the amounts of electrical activity recorded in this study were measured exactly in the same muscle areas, quantitative estimates of the pressures contributed by the different areas of intercostal muscle during breathing can be made. The only significant assumption required for such estimates is that the tension developed by a particular muscle and its effect on the lung is linearly related to the amount of EMG activity.

The values of pressure thus computed for the different areas of external intercostals during resting breathing are shown in Fig. 8. The change in airway pressure (ΔP_ao) generated by a maximal, isolated contraction of the external intercostal muscle in the dorsal third of the second interspace is -0.84 cmH₂O (De Troyer et al. 1999). As the amount of inspiratory EMG activity recorded from this muscle area during resting, room air breathing was 46 % of maximum, the corresponding change in pleural pressure could be estimated to be 0.46 × -0.84 cmH₂O, i.e. -0.39 cmH₂O (Fig. 8A). On the other hand, the maximal ΔP_ao for the external intercostal in the dorsal third of the fourth interspace is -0.54 cmH₂O, and inspiratory activity during resting breathing was only ∼20 % of maximum. The pressure contributed by this muscle area, therefore, would be about 3.5 times smaller than that contributed by the dorsal third of the second interspace. Since the maximal ΔP_ao for the muscle in the dorsal third of the sixth interspace is -0.24 cmH₂O and inspiratory activity during resting breathing was only ∼6 % of maximum, the contribution of this area should be even smaller. Thus, because the topographic distribution of neural drive among the external intercostals in the different segments precisely mirrors the topographic distribution of inspiratory mechanical advantage, the muscle contribution to lung expansion during breathing shows a stronger rostrocaudal gradient than predicted on the basis of the mechanical advantage alone. Similar computations indicate that the contribution to lung expansion of the external intercostal in the dorsal third of the fourth interspace is also much bigger than that of the muscle in the middle third and the ventral third of the same interspace (Fig. 8B).

Figure 8. Contribution of the external intercostal muscles to lung expansion during resting breathing.

The dashed bars are the changes in pressure generated during maximal contraction by the external intercostals in the dorsal third of the second, fourth, sixth and eighth interspace (A) and by the muscles in the dorsal third, the middle third and the ventral third of the fourth interspace (B); these values are from Fig. 3 in De Troyer et al. 1999. The hatched portions are estimates of the pressures contributed by the muscles during resting breathing in the supine posture. Although the pressures show rostrocaudal and dorsoventral gradients during both maximal contraction and resting breathing, the gradients during breathing are stronger.

As the pressures generated by the external intercostals or the parasternal intercostals in different segments are essentially additive (Legrand et al. 1998), these data may also be used to estimate the relative contributions of the two muscle groups to lung expansion. Adding to each other the values thus computed for the external intercostals in the dorsal third of the second, fourth, sixth and eighth interspaces during resting breathing (the latter muscle is inactive and therefore does not contribute any pressure) yields a value of -0.51 cmH₂O. On the other hand, based on similar measurements of mechanical advantage and activity, we have previously estimated that the pressure generated in the same condition by the contraction of the medial half of the parasternal intercostals in interspaces 3, 5 and 7 amounts to -1.84 cmH₂O (Legrand et al. 1996). Although these two computed values originate from different sets of data, this pressure difference between the external and parasternal intercostals fully supports the previous observation that in the dog, the contribution of the external intercostals to rib motion during resting breathing is significantly smaller than that of the parasternal intercostals (De Troyer, 1991; De Troyer & Yuehua, 1994).

Contribution of the internal intercostals to lung deflation

The topographic differences in neural drive among the different areas of internal intercostals add similarly to the topographic differences in expiratory mechanical advantage to cause marked inhomogeneities in the contribution of these muscles to lung deflation. However, we have previously emphasized (De Troyer et al. 1999) that the maximal ΔP_ao values for the internal intercostals, in particular in the most caudal segments, are likely to pertain to elevated lung volumes. Consequently, estimates of the contributions of the different muscle areas to breathing can be made only during hyperinflation, as during breathing against elevated positive end-expiratory pressures (Fig. 9). The maximal ΔP_ao for the internal intercostal in the dorsal third of the eighth interspace is +1.32 cmH₂O. Since the amount of expiratory EMG activity recorded during breathing against a 20 cmH₂O PEEP averaged 82 % of maximum, this area would therefore contribute approximately +1.08 cmH₂O. On the other hand, the pressure contributed by the middle third of the same interspace would be +0.56 cmH₂O, and that contributed by the ventral third would be only +0.20 cmH₂O. The pressures contributed by the muscles in the dorsal third of the sixth to second interspaces would be even smaller.

Figure 9. Contribution of the internal intercostal muscles to lung deflation during breathing against a 20 cmH₂O PEEP.

The dashed bars are the changes in pressure generated by different areas of internal intercostal muscle during maximal contraction (from Fig. 3 in De Troyer et al. 1999). The hatched portions are estimates of the pressures contributed by the muscles during breathing against 20 cmH₂O.

Organization of the intercostal muscle pump

Our previous studies have shown that in the dog, the mechanical advantage of the parasternal intercostal in a particular segment of the rib cage is greatest in the muscle bundles situated near the sternum and decreases rapidly toward the costochondral junction (De Troyer & Legrand, 1995). Similarly, inspiratory activity in a given parasternal intercostal decreases from the sternum to the costochondral junction, such that the bundles situated near the junction remain inactive throughout (De Troyer & Legrand, 1995). The inspiratory mechanical advantage of the medial parasternal bundles also decreases progressively from the third to the seventh interspace, and so does inspiratory activity (De Troyer et al. 1996; Legrand et al. 1996). In view of the present finding that the spatial distributions of neural drive among the external and internal interosseous intercostals are matched with the spatial distributions of mechanical advantage, one would tentatively conclude that the matching between mechanical advantage and neural drive is a general characteristic of the respiratory muscle pump.

The benefit of this phenomenon, however, is not immediately apparent. Indeed, activating muscles with a low mechanical advantage should be less effective with respect to lung volume than activating exclusively the muscles with the greatest mechanical advantage. For example, why are the external intercostals in the middle and ventral portions of the fourth interspace and in the dorsal portion of the sixth interspace ever active during inspiration, or, alternatively, why is it that external intercostal inspiratory activity is not entirely concentrated in the dorsal portion of first and second segments? Similarly, why are the internal interosseous intercostals in the fourth interspace and in the ventral two-thirds of the sixth interspace ever active during expiration? The answer to these teleological questions is uncertain at this stage, but it must be pointed out that the respiratory muscles operate to expand (or deflate) both the lung and the chest wall, and that the configuration of the chest wall is a key determinant of the work of breathing. Thus, for a given increase in lung volume, the elastic work of breathing is lowest when the chest wall is displaced along its relaxed configuration (Mead, 1974). Any departure from this configuration requires additional work of distortion, and this leads to the hypothesis that the distribution of activity among the intercostal muscles may help in expanding the rib cage compartment of the chest wall with minimum distortion.