Effect of position on the mechanical interaction between the rib cage and abdomen in preterm infants MARLA JULIAN
R. WOLFSON, JAY S. GREENSPAN, KIRAN L. ALLEN, AND THOMAS H. SHAFFER
S. DEORAS,
Departments of Physiology and Pediatrics, Temple University School of Medicine, and St. Christopher’s Hospital for Children, Philadelphia, Pennsylvania 19140 WOLFSON, MARLA R., JAY S. GREENSPAN, KIRAN S. Studies of the adult have demonstrated that changes DEORAS,JULIAN L. ALLEN,ANDTHOMAS H. SHAFFER.IZ"~~~~~~in the orientation of the body cavity relative to gravity position on the mechanicalinteraction betweenthe rib cageand alter the pressure-volume characteristics of the chest abdomenin preterm infants. J. Appl. Physiol. 72(3): 1032-1038, wall, increase passive tension and tonic activity of the 1992.-To determine the influence of body position on chest respiratory muscles, and increase the curvature of the wall and pulmonary function, we studied the ventilatory, pul- dome of the diaphragm in the direction of gravity (37). monary mechanics,and thoracoabdominal motion profiles in 20 preterm infants recovering from respiratory diseasewho Although previous studies of the infant have related the in chest wall asynchrony to were positioned in both the supine and prone position. Thora- severity of and improvement pulmonary mechanics (2, l4), it is unclear coabdominal motion was assessed from measurementsof rela- underlying changes alter this relationship. Postive rib cage and abdominal movement and the calculated whether positional phaseangle (an index of thoracoabdominal synchrony) of the turally, related alterations in musculoskeletal mechanics rib and abdomenLissajousfigures. The ventilatory and pulmo- would be particularly advantageous to the preterm infant nary function profiles were assessedfrom simultaneous mea- in whom immature respiratory muscles are presented surementsof transpulmonary pressure,airflow, and tidal vol- with high work loads, force development is dissipated in ume. The infants were studied in quiet sleep,and the order of distorting the highly compliant rib cage (RC), and a relapositioning was randomized acrosspatients. The results dem- tively small area of apposition limits mechanical couonstrated no significant difference in ventilatory and pulmopling between the abdomen (ABD) and RC (1, 9, 29). nary function measurementsas a function of position. In conWith this in mind, the purpose of the present study was trast, there was a significant reduction (-49%) in the phase angle of the Lissajous figures and an increase (+66%) in rib to determine the effect of posture (supine and prone) on mechanics and thoracoabdominal motion cage motion in prone compared with the supine position. In pulmonary addition, the degree of improvement in phase angle in the and on the relationship between these indexes of pulmoprone position was correlated to the severity of asynchrony in nary function in preterm infants recovering from respirathe supine position. We speculate that the improvement in tory disease. thoracoabdominal synchrony in the prone position is related to alterations of chest wall mechanics and respiratory muscle tone mediatedby a posturally related shift in the area of appo- METHODS sition of the diaphragm to the anterior inner rib cagewall and Subjects. Twenty-four preterm infants were selected increasein passivetension of the musclesof the rib cage. This study suggeststhat the mechanical advantage associatedwith from the neonatal intensive care units of Temple UniverHospital for Children prone positioning may confer a useful alternative breathing sity Hospital and St. Christopher’s pattern to the preterm infant in whom elevated respiratory in Philadelphia, PA, after approval by the Institutional work loads and respiratory musculoskeletalimmaturity may Review Boards. Informed consent was obtained for all predisposeto respiratory failure. studies. Patient selection was on the basis of the following: 1) in the recovery stages of respiratory distress; 2) thoracoabdominal motion; chest wall synchrony; positioning had received mechanical ventilation and supplemental THERE ISREASONABLE PRECEDENT to suggestthatpos-
tural changes may play a significant physiological role in infant recovery from respiratory disease. In this regard, studies of adults and infants have reported that relative to the supine position, lung volume, tidal volume, and compliance are greater; respiration is more regular; and infants are calmer in the prone position (3, 7, 31, 38). Increases in arterial oxygen tension have been related to an increase in lung volume as well as to improved ventilation-perfusion matching and chest wall synchrony (25, 27, 38). 1032
oxygen for a minimum of 2 days; 3) free of cardiac involvement and surgical intervention; and 4) spontaneously breathing room air or supplemental oxygen not exceeding an inspiratory oxygen fraction of 0.40. The population was born at 28 t 0.50 (SE) wk gestational age and at LO2 t 0.06 (SE) kg birth weight. The infants were studied at 33 t 0.80 (SE) wk postconceptional age at which time the weight was 1.45 t 0.02 (SE) kg. Measurement of ventilatory mechanics. Ventilatory mechanics were studied by simultaneously measuring transpulmonary pressure, airflow, and volume with esophageal and airway manometry and pneumotachography, respectively.
0161-7567/92 $2.00 Copyright0 1992 the AmericanPhysiological Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
POSTURAL
CHANGES
IN
CHEST
Air was withdrawn from an esophageal balloon catheter (Mallinckrodt, 8-Fr, St. Louis, MO) that was then inserted through the mouth into the esophagus. The balloon (4 cm length) was then inflated with air (0.6-0.9 ml), the catheter was attached to the positive side of a differential pressure transducer (Celesco P7D, Canoga Park, CA), and the pressure signal was monitored on a video screen. The catheter was advanced for intragastric balloon placement, which was confirmed by positive pressure deflection during inspiration. The catheter was then withdrawn to clear the sphincter and was positioned in the distal esophagus as confirmed by negative pressure deflection during inspiration. Adjustments of balloon placement and inflation were performed to obtain pressure waves that were reproducible and free of cardiac artifact. The esophageal catheter was taped to the infant’s nose and cheek to maintain appropriate positioning throughout the protocol. The esophageal balloon catheter was tested in vitro wherein the operating range of the balloon was found to be within 0.6-1.7 ml and the frequency response was flat to 5 Hz, consistant with appropriate esophageal measurement devices (4). Airflow was measured with a heated pneumotachometer (Fleisch no. 00) and differential pressure transducer (Validyne MP45, North Ridge, CA). The pneumotachometer was attached to a facemask (Vital Signs, Totowa, NJ) by a low-volume adapter. A tube from the sideport of the adapter was connected to the negative port of the Celesco differential pressure transducer to measure airway pressure. Potential artifact associated with esophageal and airway manometry assessment of transpulmonary pressure was eliminated by on-line monitoring of the pressure tracings and the occlusion technique (4). The facemask was gently applied over the infant’s mouth and nose while flow and volume tracings were monitored on the video screen to prevent leakage. The resistance and dead space of the pneumotachometer assembly were 13.2 cm H,O 1-l s-l and 1.7 ml, respectively; the pneumotachometer was linear for gas flows ranging from 0 to 0.15 l/s. Simultaneous measurements of transpulmonary pressure, airflow, and tidal volume were processed and stored (PeDS, Medical Associated Services, Hatfield, PA) for least-mean-square microprocessor analysis (5, 26, 33) and subsequent graphic presentation. Signals of airflow were recorded additionally on a polygraph recorder (Grass model 7CPB), amplified (Grass model 7DAC), and integrated (Grass model 7PlOCD) to yield tidal volume. Measurement of chest wall motion. Patterns of thoracoabdominal motion were assessed by the use of respiratory inductive plethysmography (RIP; Respitrace, Ardsley, NY). Bands containing inductive coils were placed around the upper RC at the level of the axillae and around the ABD at the level of the umbilicus, below the lower border of the RC. Voltage changes in response to changes in band inductance and underlying cross-sectional area were recorded on a polygraph recorder (Grass model 7PCPB). The RIP was used in the uncalibrated mode because signals were used solely as indexes of relative timing and magnitude of RC and ABD motion, rather than volumetric contribution to tidal volume (2, 30). To record signals of reasonable magnitude, the l
l
WALL
MOTION
IN
INFANTS
1033
bands were secured to ensure voltage change throughout the entire tidal volume excursion and the gain and zero offset were adjusted with the RIP in the alternating current-coupled mode; measurements used to calculate the phase angle were recorded with the RIP in the direct current-coupled mode. RC and ABD signals were displayed additionally on an X-Y recorder (Hewlett-Packard 7035B) to form Lissajous figures. As shown in Fig. 1, the Lissajous figures were used to quantitate relative timing or synchrony and magnitude of RC and ABD motion. The width of the Lissajous figure is an index of timing and/or synchrony of motion between the two compartments and is quantified by calculating a phase angle (0) from the ratio of ABD excursion at mid-RC excursion (m) to maximum ABD excursion (s) (1). For 8 < 90”, sin 8 = m/s; for 90° < 8 < 180”, 8 = 180 - p, where sin p = m/s. The lower the phase angle the more synchronous is the RC to ABD compartment motion. For example, if the RC and ABD are moving simultaneously, the Lissajous figure appears as a closed loop with a positive slope and 8 = 0. As the RC and ABD begin to move at different times, the loop opens, becomes wider, and 8 increases toward 90’. As the compartments move with increasing asynchrony, the loop may begin to narrow, the slope becomes negative, and 8 approaches 180°. The Lissajous figure resulting from frank paradox appears as a closed loop, with a negative slope and 8 of 180’. This methodology for calculating phase angles can be applied to both elliptical and nonelliptical Lissajous figures (2). In addition to timing, the Lissajous figures were used to assess qualitatively the relative magnitude of motion between the RC and ABD compartments. Magnitudes of ABD and RC excursion were expressed in arbitrary units (Fig. 1) (30). Provided the gains on the plethysmographic unit and recorders remain constant throughout a protocol, one can assess the effect of a perturbation on the relative synchrony and magnitude of thoracoabdominal excursion. For this reason, the gains of the RC and ABD channels on the RIP, polygraph recorder, and X-Y recorder were set equal, maintained constant, and a cotton jersey was used to maintain band position constant throughout the protocol. The effect of position on the band gains was assessed in five infants. In two infants, the relative displacement of the RC to ABD was assessed during a minimum of three airway occlusions in each the supine and prone positions. RC/ABD was calculated from recorded pen displacement in millimeters. In three infants, tidal volume was assessed from integration of pneumotachometer signals and was recorded simultaneously with the recorded ABD motion during 10 breaths in both supine and prone positions. ABD motion in millimeters of pen displacement was plotted against tidal volume, and a least-mean squares linear-regression equation was determined for each patient in each position. As described by Pascucci et al. (30), displacement of the ABD band as a function of position was obtained by the percentage difference in millimeter pen displacement at the average tidal volume in each position. Four infants were instrumented with the balloon catheter placed in the ABD to assess the effect of position on
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
1034
POSTURAL
CHANGES
IN CHEST
WALL
MOTION
IN INFANTS
MAGNITUDE l
Excunlon
of At30 and RC compartmentt
In arbitrary
unb
FIG. 1. Rib cage (RC) and abdomen (ABD) motion plotted to form Lissajous figure. Phase angle (e) is calculated from ratio of width at mid-RC excursion to width at maximum ABD excursion (m/S) and used index of thoracoabdominal synchrony or relative timing between RC and ABD excursion. Maximum excursions of RC (R) and ABD (s) are expressed in arbitrary units and are used as indexes of magnitude of excursion of compartments.
intragastric pressure. The catheter was advanced into the ABD as pressure swings were monitored; intragastric placement was confirmed by a positive deflection on inspiration The catheter was secured as previously described for the esophageal pressure monitoring. Protocol. Ventilatory mechanics and chest wall motion were assessed simultaneously in each infant in both supine and prone positions. The infants were facilitated to assume developmentally appropriate positions in both supine and prone, and the sequence of positions was randomized across infants (6). To eliminate the influence of neck position on ventilatory mechanics, we turned the head to a consistent direction side between positions. The studies were performed 30-60 min after the infants were fed. All measurements were recorded -1 min after the facemask-pneumotachograph assembly was applied to allow the infant to adjust to the measurement apparatus (13) before data collection was initiated. The infants were studied for a minimum of 10 min in each position. The facemask was applied for a maximum of 3-min intervals. Data collection was performed during quiet sleep, as defined by clinical criteria (35,36), to eliminate the influence of sleep state on respiratory muscle activity (11, 22, 23), and phase angle measurements were reassessed in the initial test position in eight of the infants. Data analysis. Analyses of ventilatory mechanics and thoracoabdominal motion were based on a minimum of 10 breaths and were expressed as means t SE. Intrasubject variability was assessed with the use of the coefficient of variation (CV). Statistical differences as a function of position were evaluated by the two-tailed Student’s t test for paired data and Bonferroni correction, where significance was accepted at the P < 0.05 level. RESULTS
The summarized profile (means t SE) of ventilatory and nulmonarv mechanics nrofile recorded in the supine and &prone position is shown in Table 1. Tidal volume modestly increased in 15 of 20 infants in the prone position; the mean 16% increase from supine values did not reach statistical significance. This trend in tidal volume
occurred without significant difference in peak to peak esophageal driving pressures between supine and prone positions. There was no significant difference in breathing frequency in supine compared with the prone position. A small increase in minute ventilation was recorded in 14 of 20 infants in prone position compared with supine values; however, the resultant 11% increase from supine values did not reach statistical significance. There were no statistically significant differences noted between supine values compared with values in the prone positions for pulmonary resistance or compliance. Table 1 also displays the values (means t SE) of the CV of the ventilatory parameters and pulmonary mechanics calculated in each infant in both supine and prone positions. Comparison of the CV of these data did not reveal marked differences in intrasubject variability as a function of position. Figure 2 displays representative Lissajous figures for one infant obtained in the supine and prone position. During quiet steady-state breathing, the gross morpholTABLE
1. Summarized ventilator-y and pulmonary
mechanics profize Supine Absolute
VT ml/kg ’ PTPes, cmH,O
fybreaths’min iTE,
ml. min-’
l
kg-’
RL, cmH,Ol 1-l. s
cL9 m1’ cmH20-1 kg-’ l
4.69 to.44 8.93 t1.24 80.50 t3.94 374.60 k28.30 119.50 k20.60 1.20 kO.22
Prone cov,
%
18 k3.5 17 k2 12 t1.3 18 t1.9 27 t3.7 16.4 t3.3
Absolute
5.29 kO.39 9.44 21.41 82.10 k4.33 418.15 k30.83 100.15 H5.30 1.23 k0.23
cov,
%
15.8 t2.0 16 kl.5 12.7 t1.2 19 t3.4 25 t4.3 12.80 kO.79
Values are means t SE. VT, tidal volume; PTPes, peak-to-peak esophageal pressure; f, breathing frequency; VE, minute ventilation; RL, pulmonary resistance; CL, pulmonary compliance; COV, coefficient of variation. No significant differences (P < 0.05 level),--in anvM parameter were noted as”a function of position.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
POSTURAL
CHANGES
IN CHEST
SUPINE
63’
PRONE
/+a -
a/v 29’
42’
39O
27’
34O
2. Representative Lissajous figures and respective phase angles recorded from same infant in supine and prone positions. FIG.
WALL
MOTION
IN INFANTS
1035
12.3 t 7.8% (SE)] in the prone position was significantly lower (P < 0.05) than in the supine position [CV; RC = 21 t 8.3% (SE)]. Airway occlusions of breaths in which the phase angles ranged between 176 and 180* resulted in RC/ABD of 0.96 in the supine position and 1.11 in the prone position in one infant and 1.90 in the supine position and 2.15 in the prone position in the other infant, yielding a 15.6 and 13% change as a function of position. These changes in band gain were not sufficiently large to explain the +66% increase in RC excursion in prone compared with supine position during tidal breathing. Furthermore, comparison of millimeter pen displacement for the ABD at the average tidal volume using developed least-meansquares linear-regression equations resulted in 4.80-16% differences as a function of position; the percentage differences were not associated consistently with a change to a specific position. Varying effects of position on intragastric pressure were observed, and statistical difference could not be demonstrated. Pressure did not change in two infants, increased in one, and decreased in one, resulting in values of 3.85 t 1.37 (SE) cmH,O in supine positions and 3.45 t 0.78 cmH,O in prone positions.
ogy of the Lissajous figures was relatively consistent within each position. However, marked differences in these figures between positions can be seen. In general, the width of the figures at mid-RC excursion (m) relative to maximum ABD excursion (s), the m/s ratio, was greater in supine than in the prone position. Intrasubject variability of the calculated phase angle was generally greater in the supine position than in prone. Calculated phase angles (8) of the Lissajous figures for all infants were analyzed statistically, and the summarized data are shown in Fig. 3A. In comparison with the calculated 8 of the Lissajous figures recorded in the su- DISCUSSION pine position [65.20 t 9.76 (SE) degrees], the 6 of the This study has demonstrated that thoracoabdominal Lissajous figures recorded in prone [32.90 t 3.93 (SE) degrees] were significantly lower (P < 0.001). The abso- motion in preterm infants recovering from respiratory lute magnitude of this difference was independent of the distress is more synchronous in the prone than in the initial testing position. On returning eight infants to the 85 original test position, the phase angle measurements reA 0 Supine r turned to within 10% initial values. Comparison with m Prone the CV of the 8 obtained in all infants indicated that the intrasubject variability of 8 in the prone position [ 19.60 t 3.5% (SE)] was significantly lower (P < 0.05) than that obtained in the supine position [27.3 t 6.30% (SE)]. In addition, as shown in Fig. 3B, there was a significant correlation (r = -0.58; P < 0.05) between the percent difference in 8 from supine to prone position and the absolute 0 in supine (% difference,,,,, to prone= -0.47 8 . - 5.84); wherein, infants with the highest 8 in the suz!G position tended to demonstrate the greatest decrease in 8 in the prone position. Positionally related differences in the relative magni60 TB tude of motion of the RC and ABD compartments can I! also be seen in Fig. 2. In both positions, there was a via sually observed greater outward motion of the ABD as compared with the RC. Whereas ABD excursion in the prone position was comparable with that in the supine position, there was a marked increase in RC excursion in the prone position compared with the supine position. Statistical analysis of RC excursion as a function of position for all infants demonstrated a significant increase in RC excursion in the prone position compared with the supine position [ +66 t 14% (SE) change from the supine 8? -100 ! II I1 II I1 11 I position; P < 0.011. Comparison with the CV of these 0 25 50 75 100 125 150 parameters obtained in all infants indicated that the inPhose Angle in Supine (degrees) trasubject variability of ABD excursion in the prone poFIG. 3. A: summarized phase angle profile (means t SE) calculated sition [CV 10 t 4.8% (SE)] was not markedly different from Lissajous figures recorded in all infants in supine (0) and prone from that in the supine position [CV 14 t 5.9% (SE)], (@ positions; n = 20. *P < 0.001. B: relationship between change in % whereas intrasubiect variability of RC excursion lCV difference in nhase angle (sunine to Drone) and-nhase angle in &nine.
1
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
1036
POSTURAL
CHANGES
IN CHEST
supine position. The improved synchrony, as reflected by a reduction in the phase angle of the Liss lajous figures, was associated with an observed increase in the excursion of the RC and RC relative to ABD motion. In contrast, the results of this study demonstrated no posturally related alterations in the ventilatory and pulmonary mechanics profile. Previous studies of adults and infants have employed indexes of thoracoabdominal motion as indirect assessment of respiratory function (1, 2, 10, 18, 22, 24, 34). In the mature respiratory system, thoracoabdominal motion is largely synchronous (34). During inspiration, the downward motion of the diaphragm increases intra-abdominal pressure and the abdominal wall moves outward. Simultaneously, outward motion of the RC is affected by contractile activity of the intercostal muscles and by the action of the diaphragm through the area of apposition to the RC. Age-related changes in chest wall motion have been linked to musculoskeletal maturation of the chest wall (9, 16). In this regard, the greater synchrony of thoracoabdominal motion of the full-term compared with the preterm infant may be associated with an age-related increase in RC ossification, intercostal muscle activity, and area of apposition of the diaphragm to the RC. Functionally, these developmental changes would support greater RC stability and outward movement on inspiration. Chest wall motion during early development has also been discussed relative to underlying pulmonary mechanics. Heldt and McIlroy (22) have indicated that the timing and relative amount of chest wall asynchrony during inspiration were unrelated to the pulmonary mechanics in infants recovering from hyaline membrane disease. Allen et al. (2) demonstrated improved chest wall synchrony in infants with airflow obstruction after bronchodilator treatment, wherein decreased resistance and increased compliance reduced intrapleural swings that reduced chest wall distortion. Carlo et al. (10) noted differences in chest wall motion between infants with and without respiratory distress syndrome and speculated that a lower incidence of chest wall asynchrony in infants with respiratory distress syndrome may be associated with a vagally mediated increase in respiratory drive@ Within this context, the effect of negative intrapleural pressure swings to draw the highly compliant neonatal RC inward is counterbalanced by the chest wall stabilization afforded by an increase in intercostal muscle activity. Finally, Pascucci et al. (30) recently have demonstrated that spinal anesthesia reduces inspiratory RC movement in full-term infants and associated this change in breathing pattern to a reduction in abdominal and intercostal muscle activity. Collectively, the cited studies present age, alterations in pulmonary and chest wall mechanics, and respiratory muscle activity as potential explanations for the effect of position on thoracoabdominal motion demonstrated in the present study. Because each infant in the present study was evaluated in both supine and prone positions during quiet sleep on the same day, developmental alterations in factors such as pulmonary mechanics, respiratory drive, influence and duration of sleep states, muscle activity, and chest wall compliance mav be eliminated as an explanation for
WALL
MOTION
IN INFANTS
SUPINE
-w
SPECULATIONS: 0 shift in area of apposition l
direct anterior
0 increase
//////////I
chest wall stabilization
in anterior
intercostal
tension
-
FIG. 4. Schematic representation of proposed mechanisms contributing to posturally related changes in thoracoabdominal motion. Shaded area, area of apposition of diaphragm to RC. Dotted line, inward motion of RC during inspiration in supine.
the effect of ’position on thoracoabdominal motion. The ventilatory, pulmona *rY mechanics, and ch .est wall motion profile of the supine infants in the present study is comparable with previously studied age-matched infants (12, 19). The present profile of low pulmonary compliance, high pulmonary resistance and breathing frequency, and thoracoabdominal asynchrony is characteristic of the supine preterm infant (12,19,%0,22,27). The similarity in ventilatory and pulmonary mechanics profile of infants in the supine and prone position in the present study is consistent with previous studies of the preterm infant (27, 38). Our findings of a decreased phase angle in the prone position agree with those of Martin et al. (27), who reported that prone- positioned infants de monstrated improved oxygenation and a decrease in the a.mount of time that chest w rail motion was asynchro nous (qualitative assessment). Finally, there were no consistent relationships between pulmonary mechanics and chest wall motion in the infants in thepresent study. On these bases, it appears that the observed posturally related altera tions in chest wall motion cannot be explained on the basis of the underlying ventilatory or pulmonary mechanics profile. We speculate that the improvement in thoracoabdominal synchron .y in the prone position is related to alterations in chest wal l mechanics and respiratory muscle tone. The mechanisms involved are illustrated in Fig. 4. First, we speculate that there is a shift in the area of apposition accompanying the change from the supine to prone position. As shown in . Fig. 4 (top), in the supine position, the greatest area of apposition and greater curvature of the dome of the diaphragm is located posteriorly. In the presence of incomplete RC ossification and respiratory muscle immaturity, inspiratory efforts present an imbalance across the chest wall such that negative intrapleural pressure swings draw the RC inward. In ad-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
POSTURAL
CHANGES
IN
CHEST
d&ion, a reduced area of apposition of the diaphragm to the RC in the infant relative to the adult would limit the effect of the diaphragm to facilitate outward motion of the RC during inspiration. Therefore, in the supine position, while the ABD moves outward during inspiration, the immature intercostal muscles cannot support the RC against the intrapleural pressure gradient; thus the highly compliant RC is drawn inward. Within this context and as noted in the present study, thoracoabdominal motion is relatively asynchronous and outward RC motion is small relative to the ABD. In the prone position (Fig. 4, bottom), the effect of gravity would shift the curvature of the dome of the diaphragm downward, increase the curvature of the diaphragm, and increase the area of apposition along the anterior chest wall (37). Functionally, this should make the contraction of the diaphragm more effective, and in accordance with the LaPlacian relationship, this would facilitate diaphragmatic pressure generation along the anterior chest wall; this region, because it is the farthest from the vertebral column, has the greatest mechanical advantage. This would improve mechanical coupling between the ABD and RC; therefore, for the same abdominal excursion, theoretically, RC excursion would increase. Our experimental results support this theoretical premise. Second, we postulate that in the prone position the effect of gravity would provide direct anterior chest wall stabilization during inspiration. Although the intercostal muscles oppose the downward forces of gravity in the supine position, gravity would assist the intercostal muscles in the prone position to effect outward RC motion during inspiration. Finally, we propose that the prone position increases anterior intercostal muscle tension. Axial displacement of the chest wall in the prone position would passively distend the lower RC (28, 37), favor lengthening, and, thus, an increase in the passive tension of the RC muscles. As such, the reduction in phase angle and variability of the phase angle in the prone position may reflect axial displacement of the chest wall and gravity-assisted alterations in the mechanical properties of the muscles, which would help prevent chest wall retractions during inspiration. Previous studies have suggested that abdominal loading associated with gavage feeding (21) or abdominal cuff inflation (15) 1) increases the area of apposition and chest wall stability, 2) decreases chest wall distortion, 3) improves the pressure-volume efficiency of the diaphragm, and 4) decreases the amplitude of the diaphragmatic electromyogram. These studies raise the questions of whether abdominal loading occurred in the prone position and whether the improved synchrony in the prone position may be the result of an increased area of apposition and stabilization of the lower RC as a passive consequence of increased abdominal pressure. Although this mechanism cannot be discounted on the basis of the few infants assessed for this purpose, the data indicate that the prone position did not increase significantly gastric pressure or load the ABD. It is possible that the developmentally appropriate prone position of hip and knee flexion assumed by the infants in this study acted to sling the ABD off of the surface and limit an increase in ab-
WALL
MOTION
IN
INFANTS
1037
dominal pressure. With respect to diaphragmatic work, it is possible that in the prone position the posterior portion of the diaphragm is not subjected to the potential afterloading effect of hydrostatic pressures of the abdominal contents (8, 32) that may occur in the supine position. Within this context, for the same abdominal excursion, diaphragmatic work would in fact be decreased. Summarily, the finding of improved synchrony, indicative of decreased distortion, may have therapeutic implications for decreasing diaphragmatic work (20). The potential influence of methodological concerns, such as using an uncalibrated RIP to quantitatively evaluate phase angle and qualitatively-assess compartmental excursion, was also addressed. The methodology for calculating a phase angle traditionally is based on the assessment of sinusoidal wave forms producing elliptical loops; however, because the chest wall of the infant and that of adults with lung disease does not demonstrate precise sinusoidal patterns, the use of the term phase angle as an index of thoracoabdominal motion must be viewed as an approximation. This terminology has been used previously to quantitate chest wall motion in adults (1, 34). In addition, we previously have demonstrated in infants that the margin of error associated with this methodology is relatively small (510%) (2). To the degree that the mean percent difference (49.5%) in phase angle between positions in the present study far exceeds this margin of error, it is reasonable to believe that the limitations associated with this methodology are acceptable within the context of this study design. In addition, we evaluated the effect of position on the gains of the bands in a subpopulation of infants to ascertain the margin of error associated with qualitative assessment of compartmental excursion. We found no greater than a 16% difference in the band gains between positions when we employed either airway occlusion or tidal volume methods. Because of the small differences in gain noted by either technique, it is unlikely that these small errors could account for the relatively large differences noted in chest wall excursion between positions. In summary, this study demonstrates that prone positioning reduces thoracoabdominal asynchrony and chest wall distortion in the preterm infant. We suggest that this was accomplished by increasing the mechanical coupling between the ABD and RC, thereby facilitating the respiratory muscles to expand the relatively highly compliant chest wall. The effect of prone positioning to improve thoracoabdominal synchrony may be related to the severity of asynchrony in supine. As shown in Fig. 3B, although most of the infants demonstrated improvement in synchrony, those with the highest phase angles in the supine position demonstrated the greatest reduction in phase angles in the prone position. As indicated by the lower CV in the prone compared with the supine position, these infants seemed to be mechanically more stable and able to maintain a more consistent breathing pattern in the prone position. In addition, it is particularly noteworthy that previously reported phase angles for full-term infants ranged between 0 and 18* (2). Within this context, the improvement in synchrony noted in the prone position in the present study may represent a breathing pattern approaching that of the
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
1038
POSTURAL
CHANGES
IN CHEST
full-term infant. Furthermore, because paradoxical inward movement of the RC necessitates greater inspiratory muscle shortening and work to achieve the same tidal volume (17), asynchronous thoracoabdominal motion may be another contributing factor to respiratory muscle fatigue in the preterm infant. As such, the mechanical advantage associated with prone positioning may confer a useful alternative breathing pattern to the preterm infant in whom respiratory musculoskeletal immaturity may predispose to respiratory failure. This study was supported in part by the Foundation for Physical Therapy (M. R. Wolfson) and National Heart, Lung, and Blood Institute First Independent Research Support and Transition Award HLlR29-41132 (J. L. Allen). This study was presented in part at the American Thoracic Society Annual Meeting, Boston, MA, May 1990. Address for reprint requests: M. R. Wolfson, Dept. of Physiology, Temple University School of Medicine, 3420 N. Broad St., Philadelphia, PA 19140. Received 23 July 1990; accepted in final form 13 September 1991. REFERENCES 1. AGOSTONI, E., AND P. MOGNONI. Deformation of the chest wall during breathing efforts. J. Appl. Physiol. 21: 1827-1832, 1966. 2. ALLEN, J. A., M. R. WOLFSON, K. MCDOWELL, AND T. H. SHAFFER. Thoracoabdominal asynchrony in infants with airflow obstruction. Am. Reu. Respir. Dis. 141: 337-342, 1990. 3. ATTINGER, E. O., R. G. MONROE, AND M. S. SEGAL. The mechanics of breathing in different body positions I. In normal subjects. J. Clin. Invest. 35: 904-907, 1956. 4. BEARDSMORE, C. S., P. HELMS, J. STOCKS, D. J. HATCH, AND M. SILVERMAN. Improved esophageal balloon techniques for use in infants. J. Appl. Physiol. 49: 735-742, 1980. 5. BHUTANI, V. K., E. M. SMERI, S. ABBASI, AND T. H. SHAFFER. Evaluation of neonatal pulmonary mechanics and energetics: a two factor least mean square analysis. Pediatr. Pulmonol. 4: 150-158, 1988. 6. BLYE, L. Components of normal movement during first year of life. In: Proceedings. Chapel Hill: Univ. North Carolina, 1981, p. 85-123. 7. BRACKBILL, Y. R., T. DOUTHITI-, AND H. WEST. Psychophysiologic effects in the neonate of prone versus supine placement. J. Pediatr. 82: 82-84,1973. 8. BRYAN, A. C. Comments of a devils’ advocate. Am. Rev. Respir. Dis. 110: 143,1974. 9. BRYAN, A. C., AND M. E. B. WOHL. Respiratory mechanics in children. In: Handbook of Physiology. The Respiratory System. Mechanics of Breathing. Bethesda, MD: Am. Physiol. Sot., 1986, sect. 3, vol. III, p. 179-191. 10. CARLO, W. A., R. J. MARTIN, G. A. VERSTEEGH, M. D. GOLDMAN, S. S. ROBERTSON, AND A. A. FANAROFF. The effect of respiratory distress syndrome on chest wall movements and respiratory pauses in preterm infants. Am. Rev. Respir. Dis. 126: 103-107, 1982. 11. DAVI, M., K. SANKARAN, M. MACCALLUM, D. CATES, AND H. RIGATTO. Effect of sleep state of chest distortion and on ventilatory response to CO, in neonates. Pediatr. Res. 13: 982-986, 1976. 12. DEORAS, K., E. KEKLIKIAN, M. WOLFSON, J. GREENSPAN, T. SHAFFER, AND J. ALLEN. Assessment of asynchronous rib cage and abdominal motion in infants (Abstract). Am. Rev. Respir. Dis. 139: A341,1989. 13 DOLFIN, T., P. DUFFTY, D. WILKES, S. ENGLAND, AND H. RYAN. Effects of a face mask and pneumotachograph on breathing in sleeping infants. Am. Rev. Respir. Dis. 128: 977-979, 1983. 14 DUAFW, S., AND K. K. BESSARD. Duplication of the chest lag in REM sleep by inspiratory flow-resistive loading in NREM sleep (Abstract). Pediatr. Res. 18: 390A, 1984. 15. FLEMING, P. J., N. L. MIJLLER, H. BRYAN, AND A. C. BRYAN. The
WALL
MOTION
IN INFANTS
effects of abdominal loading on rib cage distortion in premature infants. Pediatrics 64: 425-428, 1979. GERHARDT, T., AND BANCALARI. Chest wall compliance in full“0 term and premature E.infants. Acta Paediatr. Stand. 69: 359-364, 1980. 17, GOLDMAN, M. D., A. GRIMBY, AND J. MEAD. Mechanical work of breathing derived from rib age and abdominal V-P partitioning. J. Appl. Physiol. 41: 752-763, 1976. M. D., AND J. MEAD. Mechanical interaction between 18# GOLDMAN, the diaphragm and rib cage. J. Appl. Physiol. 35: 197-204,1973. 19. GREENSPAN, J. S., S. ABBASI, AND V. K. BHUTANI. Sequential changes in pulmonary mechanics in the very low birth weight (
R. WOLFSON, JAY S. GREENSPAN, KIRAN L. ALLEN, AND THOMAS H. SHAFFER
S. DEORAS,
Departments of Physiology and Pediatrics, Temple University School of Medicine, and St. Christopher’s Hospital for Children, Philadelphia, Pennsylvania 19140 WOLFSON, MARLA R., JAY S. GREENSPAN, KIRAN S. Studies of the adult have demonstrated that changes DEORAS,JULIAN L. ALLEN,ANDTHOMAS H. SHAFFER.IZ"~~~~~~in the orientation of the body cavity relative to gravity position on the mechanicalinteraction betweenthe rib cageand alter the pressure-volume characteristics of the chest abdomenin preterm infants. J. Appl. Physiol. 72(3): 1032-1038, wall, increase passive tension and tonic activity of the 1992.-To determine the influence of body position on chest respiratory muscles, and increase the curvature of the wall and pulmonary function, we studied the ventilatory, pul- dome of the diaphragm in the direction of gravity (37). monary mechanics,and thoracoabdominal motion profiles in 20 preterm infants recovering from respiratory diseasewho Although previous studies of the infant have related the in chest wall asynchrony to were positioned in both the supine and prone position. Thora- severity of and improvement pulmonary mechanics (2, l4), it is unclear coabdominal motion was assessed from measurementsof rela- underlying changes alter this relationship. Postive rib cage and abdominal movement and the calculated whether positional phaseangle (an index of thoracoabdominal synchrony) of the turally, related alterations in musculoskeletal mechanics rib and abdomenLissajousfigures. The ventilatory and pulmo- would be particularly advantageous to the preterm infant nary function profiles were assessedfrom simultaneous mea- in whom immature respiratory muscles are presented surementsof transpulmonary pressure,airflow, and tidal vol- with high work loads, force development is dissipated in ume. The infants were studied in quiet sleep,and the order of distorting the highly compliant rib cage (RC), and a relapositioning was randomized acrosspatients. The results dem- tively small area of apposition limits mechanical couonstrated no significant difference in ventilatory and pulmopling between the abdomen (ABD) and RC (1, 9, 29). nary function measurementsas a function of position. In conWith this in mind, the purpose of the present study was trast, there was a significant reduction (-49%) in the phase angle of the Lissajous figures and an increase (+66%) in rib to determine the effect of posture (supine and prone) on mechanics and thoracoabdominal motion cage motion in prone compared with the supine position. In pulmonary addition, the degree of improvement in phase angle in the and on the relationship between these indexes of pulmoprone position was correlated to the severity of asynchrony in nary function in preterm infants recovering from respirathe supine position. We speculate that the improvement in tory disease. thoracoabdominal synchrony in the prone position is related to alterations of chest wall mechanics and respiratory muscle tone mediatedby a posturally related shift in the area of appo- METHODS sition of the diaphragm to the anterior inner rib cagewall and Subjects. Twenty-four preterm infants were selected increasein passivetension of the musclesof the rib cage. This study suggeststhat the mechanical advantage associatedwith from the neonatal intensive care units of Temple UniverHospital for Children prone positioning may confer a useful alternative breathing sity Hospital and St. Christopher’s pattern to the preterm infant in whom elevated respiratory in Philadelphia, PA, after approval by the Institutional work loads and respiratory musculoskeletalimmaturity may Review Boards. Informed consent was obtained for all predisposeto respiratory failure. studies. Patient selection was on the basis of the following: 1) in the recovery stages of respiratory distress; 2) thoracoabdominal motion; chest wall synchrony; positioning had received mechanical ventilation and supplemental THERE ISREASONABLE PRECEDENT to suggestthatpos-
tural changes may play a significant physiological role in infant recovery from respiratory disease. In this regard, studies of adults and infants have reported that relative to the supine position, lung volume, tidal volume, and compliance are greater; respiration is more regular; and infants are calmer in the prone position (3, 7, 31, 38). Increases in arterial oxygen tension have been related to an increase in lung volume as well as to improved ventilation-perfusion matching and chest wall synchrony (25, 27, 38). 1032
oxygen for a minimum of 2 days; 3) free of cardiac involvement and surgical intervention; and 4) spontaneously breathing room air or supplemental oxygen not exceeding an inspiratory oxygen fraction of 0.40. The population was born at 28 t 0.50 (SE) wk gestational age and at LO2 t 0.06 (SE) kg birth weight. The infants were studied at 33 t 0.80 (SE) wk postconceptional age at which time the weight was 1.45 t 0.02 (SE) kg. Measurement of ventilatory mechanics. Ventilatory mechanics were studied by simultaneously measuring transpulmonary pressure, airflow, and volume with esophageal and airway manometry and pneumotachography, respectively.
0161-7567/92 $2.00 Copyright0 1992 the AmericanPhysiological Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
POSTURAL
CHANGES
IN
CHEST
Air was withdrawn from an esophageal balloon catheter (Mallinckrodt, 8-Fr, St. Louis, MO) that was then inserted through the mouth into the esophagus. The balloon (4 cm length) was then inflated with air (0.6-0.9 ml), the catheter was attached to the positive side of a differential pressure transducer (Celesco P7D, Canoga Park, CA), and the pressure signal was monitored on a video screen. The catheter was advanced for intragastric balloon placement, which was confirmed by positive pressure deflection during inspiration. The catheter was then withdrawn to clear the sphincter and was positioned in the distal esophagus as confirmed by negative pressure deflection during inspiration. Adjustments of balloon placement and inflation were performed to obtain pressure waves that were reproducible and free of cardiac artifact. The esophageal catheter was taped to the infant’s nose and cheek to maintain appropriate positioning throughout the protocol. The esophageal balloon catheter was tested in vitro wherein the operating range of the balloon was found to be within 0.6-1.7 ml and the frequency response was flat to 5 Hz, consistant with appropriate esophageal measurement devices (4). Airflow was measured with a heated pneumotachometer (Fleisch no. 00) and differential pressure transducer (Validyne MP45, North Ridge, CA). The pneumotachometer was attached to a facemask (Vital Signs, Totowa, NJ) by a low-volume adapter. A tube from the sideport of the adapter was connected to the negative port of the Celesco differential pressure transducer to measure airway pressure. Potential artifact associated with esophageal and airway manometry assessment of transpulmonary pressure was eliminated by on-line monitoring of the pressure tracings and the occlusion technique (4). The facemask was gently applied over the infant’s mouth and nose while flow and volume tracings were monitored on the video screen to prevent leakage. The resistance and dead space of the pneumotachometer assembly were 13.2 cm H,O 1-l s-l and 1.7 ml, respectively; the pneumotachometer was linear for gas flows ranging from 0 to 0.15 l/s. Simultaneous measurements of transpulmonary pressure, airflow, and tidal volume were processed and stored (PeDS, Medical Associated Services, Hatfield, PA) for least-mean-square microprocessor analysis (5, 26, 33) and subsequent graphic presentation. Signals of airflow were recorded additionally on a polygraph recorder (Grass model 7CPB), amplified (Grass model 7DAC), and integrated (Grass model 7PlOCD) to yield tidal volume. Measurement of chest wall motion. Patterns of thoracoabdominal motion were assessed by the use of respiratory inductive plethysmography (RIP; Respitrace, Ardsley, NY). Bands containing inductive coils were placed around the upper RC at the level of the axillae and around the ABD at the level of the umbilicus, below the lower border of the RC. Voltage changes in response to changes in band inductance and underlying cross-sectional area were recorded on a polygraph recorder (Grass model 7PCPB). The RIP was used in the uncalibrated mode because signals were used solely as indexes of relative timing and magnitude of RC and ABD motion, rather than volumetric contribution to tidal volume (2, 30). To record signals of reasonable magnitude, the l
l
WALL
MOTION
IN
INFANTS
1033
bands were secured to ensure voltage change throughout the entire tidal volume excursion and the gain and zero offset were adjusted with the RIP in the alternating current-coupled mode; measurements used to calculate the phase angle were recorded with the RIP in the direct current-coupled mode. RC and ABD signals were displayed additionally on an X-Y recorder (Hewlett-Packard 7035B) to form Lissajous figures. As shown in Fig. 1, the Lissajous figures were used to quantitate relative timing or synchrony and magnitude of RC and ABD motion. The width of the Lissajous figure is an index of timing and/or synchrony of motion between the two compartments and is quantified by calculating a phase angle (0) from the ratio of ABD excursion at mid-RC excursion (m) to maximum ABD excursion (s) (1). For 8 < 90”, sin 8 = m/s; for 90° < 8 < 180”, 8 = 180 - p, where sin p = m/s. The lower the phase angle the more synchronous is the RC to ABD compartment motion. For example, if the RC and ABD are moving simultaneously, the Lissajous figure appears as a closed loop with a positive slope and 8 = 0. As the RC and ABD begin to move at different times, the loop opens, becomes wider, and 8 increases toward 90’. As the compartments move with increasing asynchrony, the loop may begin to narrow, the slope becomes negative, and 8 approaches 180°. The Lissajous figure resulting from frank paradox appears as a closed loop, with a negative slope and 8 of 180’. This methodology for calculating phase angles can be applied to both elliptical and nonelliptical Lissajous figures (2). In addition to timing, the Lissajous figures were used to assess qualitatively the relative magnitude of motion between the RC and ABD compartments. Magnitudes of ABD and RC excursion were expressed in arbitrary units (Fig. 1) (30). Provided the gains on the plethysmographic unit and recorders remain constant throughout a protocol, one can assess the effect of a perturbation on the relative synchrony and magnitude of thoracoabdominal excursion. For this reason, the gains of the RC and ABD channels on the RIP, polygraph recorder, and X-Y recorder were set equal, maintained constant, and a cotton jersey was used to maintain band position constant throughout the protocol. The effect of position on the band gains was assessed in five infants. In two infants, the relative displacement of the RC to ABD was assessed during a minimum of three airway occlusions in each the supine and prone positions. RC/ABD was calculated from recorded pen displacement in millimeters. In three infants, tidal volume was assessed from integration of pneumotachometer signals and was recorded simultaneously with the recorded ABD motion during 10 breaths in both supine and prone positions. ABD motion in millimeters of pen displacement was plotted against tidal volume, and a least-mean squares linear-regression equation was determined for each patient in each position. As described by Pascucci et al. (30), displacement of the ABD band as a function of position was obtained by the percentage difference in millimeter pen displacement at the average tidal volume in each position. Four infants were instrumented with the balloon catheter placed in the ABD to assess the effect of position on
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
1034
POSTURAL
CHANGES
IN CHEST
WALL
MOTION
IN INFANTS
MAGNITUDE l
Excunlon
of At30 and RC compartmentt
In arbitrary
unb
FIG. 1. Rib cage (RC) and abdomen (ABD) motion plotted to form Lissajous figure. Phase angle (e) is calculated from ratio of width at mid-RC excursion to width at maximum ABD excursion (m/S) and used index of thoracoabdominal synchrony or relative timing between RC and ABD excursion. Maximum excursions of RC (R) and ABD (s) are expressed in arbitrary units and are used as indexes of magnitude of excursion of compartments.
intragastric pressure. The catheter was advanced into the ABD as pressure swings were monitored; intragastric placement was confirmed by a positive deflection on inspiration The catheter was secured as previously described for the esophageal pressure monitoring. Protocol. Ventilatory mechanics and chest wall motion were assessed simultaneously in each infant in both supine and prone positions. The infants were facilitated to assume developmentally appropriate positions in both supine and prone, and the sequence of positions was randomized across infants (6). To eliminate the influence of neck position on ventilatory mechanics, we turned the head to a consistent direction side between positions. The studies were performed 30-60 min after the infants were fed. All measurements were recorded -1 min after the facemask-pneumotachograph assembly was applied to allow the infant to adjust to the measurement apparatus (13) before data collection was initiated. The infants were studied for a minimum of 10 min in each position. The facemask was applied for a maximum of 3-min intervals. Data collection was performed during quiet sleep, as defined by clinical criteria (35,36), to eliminate the influence of sleep state on respiratory muscle activity (11, 22, 23), and phase angle measurements were reassessed in the initial test position in eight of the infants. Data analysis. Analyses of ventilatory mechanics and thoracoabdominal motion were based on a minimum of 10 breaths and were expressed as means t SE. Intrasubject variability was assessed with the use of the coefficient of variation (CV). Statistical differences as a function of position were evaluated by the two-tailed Student’s t test for paired data and Bonferroni correction, where significance was accepted at the P < 0.05 level. RESULTS
The summarized profile (means t SE) of ventilatory and nulmonarv mechanics nrofile recorded in the supine and &prone position is shown in Table 1. Tidal volume modestly increased in 15 of 20 infants in the prone position; the mean 16% increase from supine values did not reach statistical significance. This trend in tidal volume
occurred without significant difference in peak to peak esophageal driving pressures between supine and prone positions. There was no significant difference in breathing frequency in supine compared with the prone position. A small increase in minute ventilation was recorded in 14 of 20 infants in prone position compared with supine values; however, the resultant 11% increase from supine values did not reach statistical significance. There were no statistically significant differences noted between supine values compared with values in the prone positions for pulmonary resistance or compliance. Table 1 also displays the values (means t SE) of the CV of the ventilatory parameters and pulmonary mechanics calculated in each infant in both supine and prone positions. Comparison of the CV of these data did not reveal marked differences in intrasubject variability as a function of position. Figure 2 displays representative Lissajous figures for one infant obtained in the supine and prone position. During quiet steady-state breathing, the gross morpholTABLE
1. Summarized ventilator-y and pulmonary
mechanics profize Supine Absolute
VT ml/kg ’ PTPes, cmH,O
fybreaths’min iTE,
ml. min-’
l
kg-’
RL, cmH,Ol 1-l. s
cL9 m1’ cmH20-1 kg-’ l
4.69 to.44 8.93 t1.24 80.50 t3.94 374.60 k28.30 119.50 k20.60 1.20 kO.22
Prone cov,
%
18 k3.5 17 k2 12 t1.3 18 t1.9 27 t3.7 16.4 t3.3
Absolute
5.29 kO.39 9.44 21.41 82.10 k4.33 418.15 k30.83 100.15 H5.30 1.23 k0.23
cov,
%
15.8 t2.0 16 kl.5 12.7 t1.2 19 t3.4 25 t4.3 12.80 kO.79
Values are means t SE. VT, tidal volume; PTPes, peak-to-peak esophageal pressure; f, breathing frequency; VE, minute ventilation; RL, pulmonary resistance; CL, pulmonary compliance; COV, coefficient of variation. No significant differences (P < 0.05 level),--in anvM parameter were noted as”a function of position.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
POSTURAL
CHANGES
IN CHEST
SUPINE
63’
PRONE
/+a -
a/v 29’
42’
39O
27’
34O
2. Representative Lissajous figures and respective phase angles recorded from same infant in supine and prone positions. FIG.
WALL
MOTION
IN INFANTS
1035
12.3 t 7.8% (SE)] in the prone position was significantly lower (P < 0.05) than in the supine position [CV; RC = 21 t 8.3% (SE)]. Airway occlusions of breaths in which the phase angles ranged between 176 and 180* resulted in RC/ABD of 0.96 in the supine position and 1.11 in the prone position in one infant and 1.90 in the supine position and 2.15 in the prone position in the other infant, yielding a 15.6 and 13% change as a function of position. These changes in band gain were not sufficiently large to explain the +66% increase in RC excursion in prone compared with supine position during tidal breathing. Furthermore, comparison of millimeter pen displacement for the ABD at the average tidal volume using developed least-meansquares linear-regression equations resulted in 4.80-16% differences as a function of position; the percentage differences were not associated consistently with a change to a specific position. Varying effects of position on intragastric pressure were observed, and statistical difference could not be demonstrated. Pressure did not change in two infants, increased in one, and decreased in one, resulting in values of 3.85 t 1.37 (SE) cmH,O in supine positions and 3.45 t 0.78 cmH,O in prone positions.
ogy of the Lissajous figures was relatively consistent within each position. However, marked differences in these figures between positions can be seen. In general, the width of the figures at mid-RC excursion (m) relative to maximum ABD excursion (s), the m/s ratio, was greater in supine than in the prone position. Intrasubject variability of the calculated phase angle was generally greater in the supine position than in prone. Calculated phase angles (8) of the Lissajous figures for all infants were analyzed statistically, and the summarized data are shown in Fig. 3A. In comparison with the calculated 8 of the Lissajous figures recorded in the su- DISCUSSION pine position [65.20 t 9.76 (SE) degrees], the 6 of the This study has demonstrated that thoracoabdominal Lissajous figures recorded in prone [32.90 t 3.93 (SE) degrees] were significantly lower (P < 0.001). The abso- motion in preterm infants recovering from respiratory lute magnitude of this difference was independent of the distress is more synchronous in the prone than in the initial testing position. On returning eight infants to the 85 original test position, the phase angle measurements reA 0 Supine r turned to within 10% initial values. Comparison with m Prone the CV of the 8 obtained in all infants indicated that the intrasubject variability of 8 in the prone position [ 19.60 t 3.5% (SE)] was significantly lower (P < 0.05) than that obtained in the supine position [27.3 t 6.30% (SE)]. In addition, as shown in Fig. 3B, there was a significant correlation (r = -0.58; P < 0.05) between the percent difference in 8 from supine to prone position and the absolute 0 in supine (% difference,,,,, to prone= -0.47 8 . - 5.84); wherein, infants with the highest 8 in the suz!G position tended to demonstrate the greatest decrease in 8 in the prone position. Positionally related differences in the relative magni60 TB tude of motion of the RC and ABD compartments can I! also be seen in Fig. 2. In both positions, there was a via sually observed greater outward motion of the ABD as compared with the RC. Whereas ABD excursion in the prone position was comparable with that in the supine position, there was a marked increase in RC excursion in the prone position compared with the supine position. Statistical analysis of RC excursion as a function of position for all infants demonstrated a significant increase in RC excursion in the prone position compared with the supine position [ +66 t 14% (SE) change from the supine 8? -100 ! II I1 II I1 11 I position; P < 0.011. Comparison with the CV of these 0 25 50 75 100 125 150 parameters obtained in all infants indicated that the inPhose Angle in Supine (degrees) trasubject variability of ABD excursion in the prone poFIG. 3. A: summarized phase angle profile (means t SE) calculated sition [CV 10 t 4.8% (SE)] was not markedly different from Lissajous figures recorded in all infants in supine (0) and prone from that in the supine position [CV 14 t 5.9% (SE)], (@ positions; n = 20. *P < 0.001. B: relationship between change in % whereas intrasubiect variability of RC excursion lCV difference in nhase angle (sunine to Drone) and-nhase angle in &nine.
1
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
1036
POSTURAL
CHANGES
IN CHEST
supine position. The improved synchrony, as reflected by a reduction in the phase angle of the Liss lajous figures, was associated with an observed increase in the excursion of the RC and RC relative to ABD motion. In contrast, the results of this study demonstrated no posturally related alterations in the ventilatory and pulmonary mechanics profile. Previous studies of adults and infants have employed indexes of thoracoabdominal motion as indirect assessment of respiratory function (1, 2, 10, 18, 22, 24, 34). In the mature respiratory system, thoracoabdominal motion is largely synchronous (34). During inspiration, the downward motion of the diaphragm increases intra-abdominal pressure and the abdominal wall moves outward. Simultaneously, outward motion of the RC is affected by contractile activity of the intercostal muscles and by the action of the diaphragm through the area of apposition to the RC. Age-related changes in chest wall motion have been linked to musculoskeletal maturation of the chest wall (9, 16). In this regard, the greater synchrony of thoracoabdominal motion of the full-term compared with the preterm infant may be associated with an age-related increase in RC ossification, intercostal muscle activity, and area of apposition of the diaphragm to the RC. Functionally, these developmental changes would support greater RC stability and outward movement on inspiration. Chest wall motion during early development has also been discussed relative to underlying pulmonary mechanics. Heldt and McIlroy (22) have indicated that the timing and relative amount of chest wall asynchrony during inspiration were unrelated to the pulmonary mechanics in infants recovering from hyaline membrane disease. Allen et al. (2) demonstrated improved chest wall synchrony in infants with airflow obstruction after bronchodilator treatment, wherein decreased resistance and increased compliance reduced intrapleural swings that reduced chest wall distortion. Carlo et al. (10) noted differences in chest wall motion between infants with and without respiratory distress syndrome and speculated that a lower incidence of chest wall asynchrony in infants with respiratory distress syndrome may be associated with a vagally mediated increase in respiratory drive@ Within this context, the effect of negative intrapleural pressure swings to draw the highly compliant neonatal RC inward is counterbalanced by the chest wall stabilization afforded by an increase in intercostal muscle activity. Finally, Pascucci et al. (30) recently have demonstrated that spinal anesthesia reduces inspiratory RC movement in full-term infants and associated this change in breathing pattern to a reduction in abdominal and intercostal muscle activity. Collectively, the cited studies present age, alterations in pulmonary and chest wall mechanics, and respiratory muscle activity as potential explanations for the effect of position on thoracoabdominal motion demonstrated in the present study. Because each infant in the present study was evaluated in both supine and prone positions during quiet sleep on the same day, developmental alterations in factors such as pulmonary mechanics, respiratory drive, influence and duration of sleep states, muscle activity, and chest wall compliance mav be eliminated as an explanation for
WALL
MOTION
IN INFANTS
SUPINE
-w
SPECULATIONS: 0 shift in area of apposition l
direct anterior
0 increase
//////////I
chest wall stabilization
in anterior
intercostal
tension
-
FIG. 4. Schematic representation of proposed mechanisms contributing to posturally related changes in thoracoabdominal motion. Shaded area, area of apposition of diaphragm to RC. Dotted line, inward motion of RC during inspiration in supine.
the effect of ’position on thoracoabdominal motion. The ventilatory, pulmona *rY mechanics, and ch .est wall motion profile of the supine infants in the present study is comparable with previously studied age-matched infants (12, 19). The present profile of low pulmonary compliance, high pulmonary resistance and breathing frequency, and thoracoabdominal asynchrony is characteristic of the supine preterm infant (12,19,%0,22,27). The similarity in ventilatory and pulmonary mechanics profile of infants in the supine and prone position in the present study is consistent with previous studies of the preterm infant (27, 38). Our findings of a decreased phase angle in the prone position agree with those of Martin et al. (27), who reported that prone- positioned infants de monstrated improved oxygenation and a decrease in the a.mount of time that chest w rail motion was asynchro nous (qualitative assessment). Finally, there were no consistent relationships between pulmonary mechanics and chest wall motion in the infants in thepresent study. On these bases, it appears that the observed posturally related altera tions in chest wall motion cannot be explained on the basis of the underlying ventilatory or pulmonary mechanics profile. We speculate that the improvement in thoracoabdominal synchron .y in the prone position is related to alterations in chest wal l mechanics and respiratory muscle tone. The mechanisms involved are illustrated in Fig. 4. First, we speculate that there is a shift in the area of apposition accompanying the change from the supine to prone position. As shown in . Fig. 4 (top), in the supine position, the greatest area of apposition and greater curvature of the dome of the diaphragm is located posteriorly. In the presence of incomplete RC ossification and respiratory muscle immaturity, inspiratory efforts present an imbalance across the chest wall such that negative intrapleural pressure swings draw the RC inward. In ad-
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
POSTURAL
CHANGES
IN
CHEST
d&ion, a reduced area of apposition of the diaphragm to the RC in the infant relative to the adult would limit the effect of the diaphragm to facilitate outward motion of the RC during inspiration. Therefore, in the supine position, while the ABD moves outward during inspiration, the immature intercostal muscles cannot support the RC against the intrapleural pressure gradient; thus the highly compliant RC is drawn inward. Within this context and as noted in the present study, thoracoabdominal motion is relatively asynchronous and outward RC motion is small relative to the ABD. In the prone position (Fig. 4, bottom), the effect of gravity would shift the curvature of the dome of the diaphragm downward, increase the curvature of the diaphragm, and increase the area of apposition along the anterior chest wall (37). Functionally, this should make the contraction of the diaphragm more effective, and in accordance with the LaPlacian relationship, this would facilitate diaphragmatic pressure generation along the anterior chest wall; this region, because it is the farthest from the vertebral column, has the greatest mechanical advantage. This would improve mechanical coupling between the ABD and RC; therefore, for the same abdominal excursion, theoretically, RC excursion would increase. Our experimental results support this theoretical premise. Second, we postulate that in the prone position the effect of gravity would provide direct anterior chest wall stabilization during inspiration. Although the intercostal muscles oppose the downward forces of gravity in the supine position, gravity would assist the intercostal muscles in the prone position to effect outward RC motion during inspiration. Finally, we propose that the prone position increases anterior intercostal muscle tension. Axial displacement of the chest wall in the prone position would passively distend the lower RC (28, 37), favor lengthening, and, thus, an increase in the passive tension of the RC muscles. As such, the reduction in phase angle and variability of the phase angle in the prone position may reflect axial displacement of the chest wall and gravity-assisted alterations in the mechanical properties of the muscles, which would help prevent chest wall retractions during inspiration. Previous studies have suggested that abdominal loading associated with gavage feeding (21) or abdominal cuff inflation (15) 1) increases the area of apposition and chest wall stability, 2) decreases chest wall distortion, 3) improves the pressure-volume efficiency of the diaphragm, and 4) decreases the amplitude of the diaphragmatic electromyogram. These studies raise the questions of whether abdominal loading occurred in the prone position and whether the improved synchrony in the prone position may be the result of an increased area of apposition and stabilization of the lower RC as a passive consequence of increased abdominal pressure. Although this mechanism cannot be discounted on the basis of the few infants assessed for this purpose, the data indicate that the prone position did not increase significantly gastric pressure or load the ABD. It is possible that the developmentally appropriate prone position of hip and knee flexion assumed by the infants in this study acted to sling the ABD off of the surface and limit an increase in ab-
WALL
MOTION
IN
INFANTS
1037
dominal pressure. With respect to diaphragmatic work, it is possible that in the prone position the posterior portion of the diaphragm is not subjected to the potential afterloading effect of hydrostatic pressures of the abdominal contents (8, 32) that may occur in the supine position. Within this context, for the same abdominal excursion, diaphragmatic work would in fact be decreased. Summarily, the finding of improved synchrony, indicative of decreased distortion, may have therapeutic implications for decreasing diaphragmatic work (20). The potential influence of methodological concerns, such as using an uncalibrated RIP to quantitatively evaluate phase angle and qualitatively-assess compartmental excursion, was also addressed. The methodology for calculating a phase angle traditionally is based on the assessment of sinusoidal wave forms producing elliptical loops; however, because the chest wall of the infant and that of adults with lung disease does not demonstrate precise sinusoidal patterns, the use of the term phase angle as an index of thoracoabdominal motion must be viewed as an approximation. This terminology has been used previously to quantitate chest wall motion in adults (1, 34). In addition, we previously have demonstrated in infants that the margin of error associated with this methodology is relatively small (510%) (2). To the degree that the mean percent difference (49.5%) in phase angle between positions in the present study far exceeds this margin of error, it is reasonable to believe that the limitations associated with this methodology are acceptable within the context of this study design. In addition, we evaluated the effect of position on the gains of the bands in a subpopulation of infants to ascertain the margin of error associated with qualitative assessment of compartmental excursion. We found no greater than a 16% difference in the band gains between positions when we employed either airway occlusion or tidal volume methods. Because of the small differences in gain noted by either technique, it is unlikely that these small errors could account for the relatively large differences noted in chest wall excursion between positions. In summary, this study demonstrates that prone positioning reduces thoracoabdominal asynchrony and chest wall distortion in the preterm infant. We suggest that this was accomplished by increasing the mechanical coupling between the ABD and RC, thereby facilitating the respiratory muscles to expand the relatively highly compliant chest wall. The effect of prone positioning to improve thoracoabdominal synchrony may be related to the severity of asynchrony in supine. As shown in Fig. 3B, although most of the infants demonstrated improvement in synchrony, those with the highest phase angles in the supine position demonstrated the greatest reduction in phase angles in the prone position. As indicated by the lower CV in the prone compared with the supine position, these infants seemed to be mechanically more stable and able to maintain a more consistent breathing pattern in the prone position. In addition, it is particularly noteworthy that previously reported phase angles for full-term infants ranged between 0 and 18* (2). Within this context, the improvement in synchrony noted in the prone position in the present study may represent a breathing pattern approaching that of the
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.105.215.146) on July 31, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
1038
POSTURAL
CHANGES
IN CHEST
full-term infant. Furthermore, because paradoxical inward movement of the RC necessitates greater inspiratory muscle shortening and work to achieve the same tidal volume (17), asynchronous thoracoabdominal motion may be another contributing factor to respiratory muscle fatigue in the preterm infant. As such, the mechanical advantage associated with prone positioning may confer a useful alternative breathing pattern to the preterm infant in whom respiratory musculoskeletal immaturity may predispose to respiratory failure. This study was supported in part by the Foundation for Physical Therapy (M. R. Wolfson) and National Heart, Lung, and Blood Institute First Independent Research Support and Transition Award HLlR29-41132 (J. L. Allen). This study was presented in part at the American Thoracic Society Annual Meeting, Boston, MA, May 1990. Address for reprint requests: M. R. Wolfson, Dept. of Physiology, Temple University School of Medicine, 3420 N. Broad St., Philadelphia, PA 19140. Received 23 July 1990; accepted in final form 13 September 1991. REFERENCES 1. AGOSTONI, E., AND P. MOGNONI. Deformation of the chest wall during breathing efforts. J. Appl. Physiol. 21: 1827-1832, 1966. 2. ALLEN, J. A., M. R. WOLFSON, K. MCDOWELL, AND T. H. SHAFFER. Thoracoabdominal asynchrony in infants with airflow obstruction. Am. Reu. Respir. Dis. 141: 337-342, 1990. 3. ATTINGER, E. O., R. G. MONROE, AND M. S. SEGAL. The mechanics of breathing in different body positions I. In normal subjects. J. Clin. Invest. 35: 904-907, 1956. 4. BEARDSMORE, C. S., P. HELMS, J. STOCKS, D. J. HATCH, AND M. SILVERMAN. Improved esophageal balloon techniques for use in infants. J. Appl. Physiol. 49: 735-742, 1980. 5. BHUTANI, V. K., E. M. SMERI, S. ABBASI, AND T. H. SHAFFER. Evaluation of neonatal pulmonary mechanics and energetics: a two factor least mean square analysis. Pediatr. Pulmonol. 4: 150-158, 1988. 6. BLYE, L. Components of normal movement during first year of life. In: Proceedings. Chapel Hill: Univ. North Carolina, 1981, p. 85-123. 7. BRACKBILL, Y. R., T. DOUTHITI-, AND H. WEST. Psychophysiologic effects in the neonate of prone versus supine placement. J. Pediatr. 82: 82-84,1973. 8. BRYAN, A. C. Comments of a devils’ advocate. Am. Rev. Respir. Dis. 110: 143,1974. 9. BRYAN, A. C., AND M. E. B. WOHL. Respiratory mechanics in children. In: Handbook of Physiology. The Respiratory System. Mechanics of Breathing. Bethesda, MD: Am. Physiol. Sot., 1986, sect. 3, vol. III, p. 179-191. 10. CARLO, W. A., R. J. MARTIN, G. A. VERSTEEGH, M. D. GOLDMAN, S. S. ROBERTSON, AND A. A. FANAROFF. The effect of respiratory distress syndrome on chest wall movements and respiratory pauses in preterm infants. Am. Rev. Respir. Dis. 126: 103-107, 1982. 11. DAVI, M., K. SANKARAN, M. MACCALLUM, D. CATES, AND H. RIGATTO. Effect of sleep state of chest distortion and on ventilatory response to CO, in neonates. Pediatr. Res. 13: 982-986, 1976. 12. DEORAS, K., E. KEKLIKIAN, M. WOLFSON, J. GREENSPAN, T. SHAFFER, AND J. ALLEN. Assessment of asynchronous rib cage and abdominal motion in infants (Abstract). Am. Rev. Respir. Dis. 139: A341,1989. 13 DOLFIN, T., P. DUFFTY, D. WILKES, S. ENGLAND, AND H. RYAN. Effects of a face mask and pneumotachograph on breathing in sleeping infants. Am. Rev. Respir. Dis. 128: 977-979, 1983. 14 DUAFW, S., AND K. K. BESSARD. Duplication of the chest lag in REM sleep by inspiratory flow-resistive loading in NREM sleep (Abstract). Pediatr. Res. 18: 390A, 1984. 15. FLEMING, P. J., N. L. MIJLLER, H. BRYAN, AND A. C. BRYAN. The
WALL
MOTION
IN INFANTS
effects of abdominal loading on rib cage distortion in premature infants. Pediatrics 64: 425-428, 1979. GERHARDT, T., AND BANCALARI. Chest wall compliance in full“0 term and premature E.infants. Acta Paediatr. Stand. 69: 359-364, 1980. 17, GOLDMAN, M. D., A. GRIMBY, AND J. MEAD. Mechanical work of breathing derived from rib age and abdominal V-P partitioning. J. Appl. Physiol. 41: 752-763, 1976. M. D., AND J. MEAD. Mechanical interaction between 18# GOLDMAN, the diaphragm and rib cage. J. Appl. Physiol. 35: 197-204,1973. 19. GREENSPAN, J. S., S. ABBASI, AND V. K. BHUTANI. Sequential changes in pulmonary mechanics in the very low birth weight (
