We tested the hypothesis that the mechanical arrangement of costal (COS) and crural (CRU) diaphragms can be changed from parallel to series when direct or indirect transmission of tension occurs. Ratio of rib cage to abdominal displacement (RC/AB) resulting from separate COS and CRU stimulations were used to measure RC expanding action. Hyperinflation in six dogs caused RC/AB with COS and CRU stimulations to change progressively from 0.53 +/- 0.07 (SE) and 0.03 +/- 0.05 at functional residual capacity (FRC) to -0.48 +/- 0.08 and -0.46 +/- 0.05 at 68% inspiratory capacity, respectively. Liquid substitution of abdominal contents in six other dogs equalized abdominal pressure swings (delta Pab), without changing chest wall elastic properties or geometry, or costal RC/AB (0.35 +/- 0.07 before and 0.33 +/- 0.06 after) but caused crural RC/AB to change from 0.01 +/- 0.05 to 0.31 +/- 0.01. We conclude that hyperinflation changes fiber orientation, allowing direct transmission of tension between COS and CRU, which become linked mechanically in series (the diaphragm acts as a unit with RC deflating action); and equalization of delta Pab causes indirect transmission of tension between COS and CRU, which become linked in series (the diaphragm acts as a unit with RC inflating action).

We examined the relationship between changes in abdominal cross-sectional area, measured by respiratory inductive plethysmography, and changes in length in the costal and crural parts of the diaphragm, measured by sonomicrometry, in nine supine, anesthetized dogs. During passive inflation, both parts of the diaphragm shortened and abdominal cross-sectional area increased. During passive deflation, both parts of the diaphragm lengthened and abdominal cross-sectional area decreased. We subsequently used the relationship between costal and crural diaphragmatic length, respectively, and abdominal cross-sectional area during passive inflation-deflation to predict the length changes in the costal and crural diaphragm during quiet breathing before and after bilateral phrenicotomy. In the intact animal the inspiratory shortening in the crural diaphragm was almost invariably greater than predicted from the relationship during passive inflation. During inspiration after phrenicotomy the crural diaphragm invariably lengthened, whereas the costal diaphragm often shortened. In general there was a good correlation between the measured and predicted length change for the crural diaphragm (r = 0.72 before and 0.79 after phrenicotomy) and a poor one for the costal diaphragm (r = 0.05 before and 0.19 after phrenicotomy).

It is conventionally considered that because of their fiber orientations, the external intercostal muscles elevate the ribs, whereas the internal interosseous intercostals lower the ribs. The mechanical action of the intercostal muscles, however, has never been studied directly, and the electromyographic observations supporting this conventional thinking must be interpreted with caution. In the present studies, the external and internal interosseous intercostal muscles have been separately stimulated in different interspaces at, above, and below end-expiratory rib cage volume in anesthetized dogs. The axial (cephalo-caudal) displacements of the ribs were measured using linear displacement transducers. The results indicate that when contracting in a single interspace and other muscles are relaxed, both the external and internal intercostals have a net rib elevating action at end-expiratory rib cage volume. This action increases as rib cage volume decreases, but it progressively decreases as rib cage volume increases such that at high rib cage volumes, both the external and internal intercostals lower the ribs. Stimulating the intercostal muscles in three adjacent intercostal spaces simultaneously produced similar directional rib motion results. We conclude that (a) in contrast with the conventional thinking, the external and internal interosseous intercostals acting alone have by and large a similar effect on the ribs into which they insert; (b) this effect is very much dependent on rib cage (lung) volume; and (c) intercostal muscle action is primarily determined by the resistance of the upper ribs to caudad displacement relative to the resistance of the lower ribs to cephalad displacement. The lateral intercostals, however, might be more involved in postural movements than in respiration. Their primary involvement in rotations of the trunk might account for the presence of two differently oriented muscle layers between the ribs.

The pressure swings under the costal (Pcos) and crural diaphragms (Pcru) and between the intestinal loops (Pint) were compared with the swings in gastric pressure (Pga) in 13 supine anesthetized dogs. Pcos, Pcru, and Pint were measured with air-filled latex balloons in eight dogs and saline-filled catheters in five. Pga was measured with an air-filled balloon in all dogs. During quiet breathing differences were often present, the directions of which were variable from animal to animal. During mechanical ventilation, all pressures increased, but both Pcos and Pcru increased more than Pga, whereas only a small change was observed in Pint. During bilateral stimulation of the costal diaphragm, Pcos invariably increased more than Pga and Pint, whereas almost no change was observed in Pcru. During bilateral stimulation of the crural diaphragm, Pcru invariably increased more than Pga, Pint, and Pcos. During abdominal muscle stimulation as during external abdominal compression, Pint always increased more than Pcos and Pcru. During lower rib cage compression, Pga, Pcos, and Pcru increased more than Pint. During sternocleidomastoid stimulation, all pressure swings were negative, but the change in Pint was always smaller than in Pcos, Pcru, or Pga. Inhomogeneities observed with balloons and saline-filled catheters were similar. After the abdomen was filled with 2 liters of saline all pressure swings became much more homogeneous.

To assess the mechanical arrangement of the costal and crural parts of the diaphragm, we studied changes in diaphragmatic length with piezoelectric crystals in 17 supine anesthetized dogs. During control resting inspiration, the crural part usually shortened more and earlier than the costal part. After phrenicotomy, the crural part always lengthened during inspiration, whereas the costal part shortened or lengthened. These interanimal differences disappeared after opening of the abdomen; the costal part then always lengthened during inspiration. During stimulation of one part, the relaxed nonstimulated part always lengthened. However, when compared with the relationship between length and transdiaphragmatic pressure (Pdi) obtained during passive deflation, the lengthening of the relaxed part during stimulation of either part was small. This difference between predicted and measured Pdi-length relationship decreased in magnitude as lung volume increased above functional residual capacity (FRC) and increased as residual volume was approached. These results indicate that 1) even during quiet breathing the diaphragm in the dog is not a single functional entity; 2) at FRC the costal and crural portions of the diaphragm behave as if they were mechanically arranged partly in parallel and partly in series; and 3) they gradually move into a pure mechanical series arrangement as lung volume increases.