| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
HIGHLIGHTED TOPICS
Physiology and Pathophysiology of Sleep Apnea
1Ludwig Engel Centre for Respiratory Research, Department of Respiratory Research, Westmead Hospital, Westmead; and 2University of Sydney, Camperdown, New South Wales, Australia
Submitted 12 May 2005 ; accepted in final form 17 August 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
0.1 cmH2O/mm MA (n = 7–11; all P < 0.0005). However, the fall in Rua (measured at 20 ml/s) with MA was greater for MA at 75° (0.03 mmH2O·ml–1·s·mm–1) than at 100° (0.01 mmH2O·ml–1·s·mm–1; P < 0.02). From these findings, we conclude that MA decreases ETP and is more effective in reducing Rua without mouth opening. upper airway extraluminal tissue pressure; upper airway patency
37% of OSA patients (8, 20). MA has been shown to increase pharyngeal cross-sectional area in anesthetized (12) and awake human subjects (9), as well as to decrease upper airway resistance (Rua) in awake human subjects (17) and to decrease upper airway closing pressures in sleeping OSA patients (21). However, there has been poor understanding of the mechanisms by which MA improves upper airway patency or, at least, reduces upper airway collapsibility. Previous workers have suggested that MA may improve upper airway patency via an increase in genioglossus muscle activity (1), a decrease in upper airway extraluminal tissue pressure (ETP) (12), or traction on the lateral pharyngeal walls caused by anterior displacement of the tongue (12). However, because MA can improve upper airway patency during muscle paralysis (12), increased upper airway muscle activity is unlikely to be an exclusive mechanism.
Although it is widely understood that the patency of the pharyngeal airway is determined by the ability of the airway walls to resist the collapsing forces generated by negative intraluminal pressures during inspiration, it has been appreciated for more than two decades (27) that pharyngeal airway patency is also impacted by the pressure existing in the peripharyngeal tissues. Despite this awareness, there have been few studies that have measured ETP. The first measurements of ETP were made in pigs by Winter et al. (38, 39) almost a decade ago. More recently, our laboratory has measured ETP in an anesthetized rabbit model and shown that ETP is usually positive (thus exerting a collapsing pressure on the pharyngeal airway), fluctuates with respiration, and is increased by mechanical influences such as head and neck flexion (14). The precise determinants of ETP remain unclear, but it has been suggested that the boundary conditions that define the upper airway extraluminal tissue space may play a role (34, 37).
Although usually modeled as a floppy-walled tube, the upper airway can also be considered as a luminal shape formed by the surrounding tissues. In turn, these tissues are partially enclosed within a bony box subtended by the mandible, skull, and cervical spine. In this model, the ETP may be influenced by the amount of enclosed tissue relative to the size of the bony box (13, 34, 37). Similar models are used to describe determinants of brain tissue pressure after head injury (5) or muscle tissue pressure in acute compartment syndromes (33). When influences on upper airway patency are viewed via this framework, it suggests that mandibular position may be an important determinant of the bony box volume and, hence, may alter the transmural pressure operating across the pharyngeal wall through an effect on ETP.
To explore the hypothesis that MA may improve upper airway patency, at least in part, via a reduction in ETP, we utilized an established animal model to examine relationships between MA, pharyngeal ETP, and upper airway patency.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Anesthesia. Anesthesia was induced with an intramuscular injection of ketamine (35 mg/kg) and xylazine (5 mg/kg) and maintained using intravenous ketamine (15 mg·kg–1·h–1) and xylazine (4.5 mg·kg–1·h–1).
Upper airway airflow, tidal volume, and tracheal pressure.
Tracheal pressure (Ptr) and upper airway airflow (
) were monitored using a pressure transducer (±10 cmH2O; Celesco, Chatsworth, CA) and heated pneumotachograph (model 8300A, Hans Rudolph, Kansas, MO), respectively, both inserted in series between the third and tenth cartilage rings of the intact trachea (Fig. 1). Tidal volume (VT) was obtained via electrical integration of the signal.
|
MA. Head and neck position were controlled at 50° to the horizontal relative to a line drawn from the tragus to the external nares. The lower incisor teeth were removed, and an 0 silk suture was passed around the mandibular symphysis. The suture thread was attached to a specially designed external MA device fitted with a screw via which graded tension could be applied to the suture thread and the mandible advanced. MA (0–5 mm in 1-mm increments from baseline) and angle of MA [referenced to the horizontal; 75° (n = 11), 85° (n = 7) and 100° (n = 7)] were quantified using an incorporated scale and angle finder. Mouth opening dimension (distance between upper incisor teeth and lower jaw) was also measured at MAs of 0 and 5 mm (Fig. 1).
Data analysis.
Mean and 
(the size of the respiratory fluctuations, maximum – minimum) values were obtained for three runs of five steady-state breaths for each condition. Power functions (Ptr = ab + c, where a, b, and c are constants) were fitted to the inspiratory limb of Ptr/ curves between = 0 and 20 ml/s (Figs. 2 and 3), and Rua (Ptr/ at inspiratory = 20 ml/s) was calculated. The presence or absence of limitation was quantified by visual inspection of the Ptr vs. plots, and it was said to be present if there was an increase in Ptr in the absence of an increase in inspiratory .
|
|
A linear mixed-effects model (25) was used to examine the individual rabbit (random effect) interactions between MA (fixed effect), advancement angle (fixed effect), Rua (fixed effect), and ETP values anteriorly and laterally (fixed effect). Relationships between the change in mean ETP from baseline with both MA (fixed effect) and Rua (expressed as a percentage of baseline Rua, fixed effect), at each angle of MA (fixed effect), were also examined within individual rabbits (random effect) at each catheter position. P < 0.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
0.7 cmH2O (P < 0.003; Table 1). For the group, MA at all angles resulted in a fall in ETPlat and ETPant (Fig. 2), and it was associated with a reduction in Rua (Fig. 3).
|
Effect of MA on VT. When the mandible was advanced to 5 mm, at 75° the VT was 16.7 ± 2.2 ml, at 85° it was 16.1 ± 4.0 ml, and at 100° it was 15.7 ± 3.6 ml (all not significantly different from baseline value, P > 0.4).
Effect of MA on mouth opening. When 5 mm of MA at 75° were applied, there was no significant increase in the measured mouth-opening dimension from the resting value of 4.0 ± 0.2 mm. However, mouth opening increased to 5.7 ± 0.6 mm at 85° and 6.6 ± 0.4 mm at 100° (both P < 0.004).
Effect of MA on limitation.
For a MA angle of 75°, from inspection of the Ptr vs. plots, inspiratory limitation (see Fig. 3) was present at baseline in 4 of 11 rabbits, was present at 1 and 2 mm of advancement in 3 of 11 rabbits, and was absent at all levels of MA >2 mm. For MA angles of both 85 and 100°, flow limitation was present in one of seven rabbits at both baseline and at 1 mm, and it was absent at all levels of advancement >1 mm.
Effect of MA on Rua. Graded MA, across all angles, resulted in a 0.01- to 0.03-mm H2O·ml–1·s·mm–1 of MA decrease in Rua (Figs. 4A and 5A). However, the rate of fall for Rua was significantly greater at an MA of 75° than at 85° or 100° (P < 0.01 compared with MA 85° and 100°; Fig. 5A).
|
|
ETPlat all fell progressively at all MA angles [Table 2, Figs. 4 (B and C) and 5 (B and C)]. When the data were examined using the linear mixed-effects model, for every 1 mm of MA, both group mean ETPlat and mean ETPant fell by 0.1 cmH2O (both P < 0.01; Table 2, Fig. 5B). The only change in ETP with MA was a decrease in ETPlat at an MA angle of 75° (Table 2, Fig. 5C).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Critique of methods. Anesthetized rabbits have been used previously in a number of physiological studies of upper airway mechanics (15, 19, 24), and our laboratory has also employed this model to investigate the mechanics of the upper airway extraluminal tissue space (14). General anesthesia depresses upper airway muscle activity (7, 22, 23), producing a physiological model similar to sleep (6). In the present study, upper airway muscle activity was not monitored; thus it is possible our results may have been influenced by upper airway muscle recruitment. However, using the same anesthetic regime in a feline upper airway model, other workers have demonstrated quiescent upper airway muscles during tidal breathing (36). Although this remains a potential mechanism linking MA with changes in Rua and/or ETP, the main aim of our study was to determine whether ETP changed with MA and, if so, whether there was a relationship between change in ETP and change in Rua.
We chose to measure change in Rua, at an not associated with limitation, as a marker of the functional outcome of graded MA. As an index of upper airway patency, Rua is mainly influenced by airway dimensions (35). Thus the outcome tested in the present study is likely to be primarily related to mandibular effects on upper airway geometry, rather than on the collapsibility of upper airway walls. However, some insight into effects on wall collapsibility may be gleaned by an examination of the effect of MA on the prevalence of inspiratory -limited breaths. Inspiratory limitation was present at baseline in some rabbits, but it was abolished by MA in all cases, suggesting reduced upper airway wall collapsibility.
The effect of MA on the rabbit upper airway has not previously been reported. Although our rabbit model is not intended to act as a model of OSA, rabbits are naturally retrognathic, and in humans retrognathia and reduced mandibular length are recognized causes of OSA (4, 11, 18). Consequently, the rabbit may be suited to studies that assess the mechanical consequences of mandibular geometry.
In our experimental design, we have chosen to use an external traction device to advance the mandible, and this differs from intraoral MA splints. It is possible that there are differences in the effect of an intraoral device compared with external traction, due perhaps to stimulation of mechanoreceptors; however, because our rabbits are anesthetized, these reflexes are likely to be suppressed. In addition, MA with intraoral splints is achieved not by graded advancement as performed here but via advancement to one level. This may have a different effect on the upper airway than graded advancement. The external traction device allowed us to vary the angle at which we advanced the mandible. Different angles of advancement have not been reported for MA splints; however, controversy exists over the degree of mouth opening that should occur with MA. Some authors report greater improvements in AHI with MA splints when mouth opening was promoted (3, 16), although a recent randomized crossover trial reported no difference in outcomes between two devices, one with and one without mouth opening (26). In our study, varying the angle of advancement was employed as a methodology for producing varying degrees of mouth opening.
The limitations of the methodology for measuring ETP have been discussed previously (14, 38, 39). Briefly, the surgical insertion of the catheter and the space occupying nature of the catheter will both result in a perturbation of the baseline ETP. For this reason, most of our analyses are based on comparisons of the change from baseline, rather than absolute values. ETP was chosen as an outcome variable rather than the transmural pressure (the pressure acting across the upper airway walls, which is the difference between intraluminal and extraluminal pressures), because the upper airway was intact and there was airflow through the upper airway. Measurement of Ptr, in the presence of airflow, will not be reflective of the luminal pressure acting at the site of the measurement of ETP. Depending on the flow regime, even an intraluminal pressure catheter present at the same level as the catheters measuring ETP may not reflect the intraluminal pressure acting at the luminal mucosa.
Baseline ETP.
Baseline values for all ETP measures were not significantly different for the different advancement angles used and were similar to those our laboratory has reported previously for anesthetized rabbits (14). Values for ETPlat were consistently greater than those for ETPant by 0.7 cmH2O. This finding is also similar to our laboratory's previous data (14) and demonstrates that ETP is not uniformly distributed around the pharyngeal airway. A positive ETP will exert a collapsing pressure on the upper airway. For the airway to remain patent, an "upper airway dilating pressure" (e.g., elastic recoil, dilator muscle activity) equal and opposite to the ETP must be present. When upper airway dilator muscle activity is depressed, this force is presumably associated with the elastic properties of the upper airway wall (outward recoil). Some insight into the effect of MA on this upper airway dilating pressure may be gained by considering the ETP values at the zero-flow points during tidal breathing in the present study. With no airflow, intraluminal pressure in an open upper airway is atmospheric and transmural pressure is equivalent to ETP. Moreover, if the airway wall is not in motion, then upper airway dilating pressure is equal and opposite to ETP. In the present study, ETP (no airflow) fell with progressive MA (see Fig. 2), implying that upper airway dilating pressure also fell. In this scenario, MA lowers ETP, thus reducing transmural compression of the airway and promoting attainment of a new equilibrium point characterized by increased luminal dimensions (as evidenced by reduced Rua) and reduced upper airway-dilating pressure, which in the absence of dilator muscle activity, likely represents reduced outward "wall recoil pressure."
We also recorded, as has been reported previously by other workers and our laboratory (14, 38, 39), negative values for ETP. If the ETP is negative, it will be exerting a force that will tend to hold the upper airway open. ETP is the total tissue pressure acting at the point, and it is the sum of the pressures exerted by the solid and fluid elements of the tissues, or the solid tissue pressure and the interstitial fluid pressure (10). The presence of a negative pressure in the tissues may be because we have reduced the pressure exerted by the solid elements, through tissue movement as the bony enclosure expands, and are thus predominantly measuring the pressure exerted by the interstitial fluid, which is reported to be negative (10).
MA and ETP.
The present study is the first to combine measurement of ETP with graded MA. Our findings clearly demonstrate that both ETPant and ETPlat are progressively reduced during MA. However, the rate of decrease in all measures of ETP was not significantly influenced by measurement site or angle of advancement. Thus, although ETP is nonuniformly distributed around the pharyngeal airway, the relative effect of MA on these pressures is homogeneous. The sole exception was that ETPlat decreased significantly with increasing MA only when the advancement angle was 75°; i.e., the ETPlat fluctuations associated with the respiratory cycle became increasingly damped, and this effect was influenced by both measurement site and angle of advancement.
Because measured ETP values were mostly above atmospheric pressure, a decrease in ETP reduces the compressive pressure being exerted on the pharyngeal airway walls, i.e., alters transmural pressure. Although the magnitude of the falls in ETP were small in absolute terms (<0.5 cmH2O), the functional effect of this on airway patency will depend on the collapsibility of the airway wall. If airway walls are highly compliant or, inspiratory luminal pressure is close to upper airway closing pressure, a change in ETP of the magnitude demonstrated in the present study may be sufficient to prevent airway closure.
Relationship between Rua and ETP.
During MA, we found significant linear relationships between both ETPlat and ETPant and Rua (Fig. 6). This finding, although not definitive, is consistent with the contention that the effect of MA on upper airway patency may be mediated via its effect on ETP.
However, an intriguing finding emerged for MA at 75°. For this angle alone, the rate of fall in Rua per unit fall in mean ETPlat, but not ETPant, was significantly greater than for the other angles (Fig. 6A). Because the rate of reduction of ETPlat with MA was the same at all angles (Fig. 5), this finding suggests that there is an angle of advancement interaction between the change in ETPlat and the change in Rua and, furthermore, that this does not apply to ETPant; i.e., Rua is more sensitive to changes in ETPlat than ETPant but only when the advancement angle is 75°. Because mouth opening occurred with MA of 85 and 100°, there was potential for leakage of airflow via the mouth (potential low-resistance pathway) to influence the results. However, mouth leak is unlikely to have occurred because overlap of the epiglottis and soft palate is a normal characteristic of rabbit upper airway anatomy. Furthermore, the rate of fall of Rua with MA was greatest at a MA angle (75°) that did not result in mouth opening. A potential explanation may lie in effects of mouth opening on lateral pharyngeal wall compliance, which is the radial displacement that occurs in response to a change in intraluminal pressure.
In this study, MA at 75° resulted in a significant reduction in the size of the respiratory-related fluctuations in ETP in the lateral pharyngeal walls (ETPlat), which did not occur at any of the other angles. The magnitude of ETP, as discussed previously (14, 38, 39), may be related either to upper airway muscle activity, tracheal traction, or pressure transmission from the upper airway lumen. Because VT is unchanged by MA, it seems unlikely that there has been a change in tracheal traction. It also seems unlikely, because the degree of mouth opening was small, that different upper airway muscle activation or different vector action of upper muscles has occurred with such small changes in MA; however, this remains a possibility. Alternatively, there may have been a small decrease in the intraluminal pressures transmitted across the lateral pharyngeal walls. This may reflect decreased intraluminal pressure fluctuations associated with a decreased Rua, but that should have influenced both ETPlat and ETPant. Intraluminal pressure transmission is likely to occur through movement of the upper airway walls, which has been demonstrated during tidal breathing in human subjects (31, 30). Thus we speculate that this result may reflect decreased compliance of the lateral pharyngeal walls, perhaps associated with mucosal unfolding during MA at 75°. If this is correct then reduced lateral wall compliance may have contributed to the observed greater rate of fall of Rua with MA at 75°. In support of this, upper airway collapsibility (measured by critical closing pressure) in sedated (midazolam) human subjects increased with increasing mouth opening, with no change in upstream resistance (2). A similar model, where upper airway patency is determined by both wall compliance and ETP, has been proposed to explain the effect of tracheal traction on upper airway patency in the cat (28, 32).
Proposed model of MA. Recently, Watanabe and colleagues (34) proposed a model of the upper airway as an airway lumen surrounded by tissues constrained within a bony box formed by the spine and mandible. These workers suggested that pharyngeal airway patency was influenced by the pressure in the box-enclosed tissues and that this pressure would depend on the relative (unrestrained) volumes of tissues and bony box. Using their model an increase in box volume should be associated with a fall in the pressure in the tissues. The fall in ETP with MA in the present study is consistent with this view. MA may exert an effect on upper airway patency via an ETP-lowering increase in bony box volume.
The single-compartment model of Watanabe and colleagues (34), however, is not consistent with the ETP heterogeneity demonstrated in our present and previous studies. In general we find that, in anesthetized rabbits, ETPlat is 0.7 cmH2O greater than ETPant (14), suggesting structural and/or functional compartmentalization in the peripharyngeal tissues. In Fig. 7, we present a modified version of the model of Watanabe et al. that includes functional anterior and lateral ETP compartments. In the two-dimensional model proposed by Watanabe et al. the bony enclosure is represented as a square or circular container (34). We have represented the bony enclosure as a triangle, because this is more reflective of mandibular anatomy. Anterior movement of the triangular-shaped mandible would result in an increase in both the lateral and anterior dimensions of the bony enclosure of the upper airway (see Fig. 7). In our proposed model, to the extent that the anterior and lateral ETP compartments are functionally independent, MA reduces ETPlat primarily as a result of an increase box lateral dimensions as the wider part of the mandible moves anteriorly, whereas the reduction in ETPant results from the increase in the anteroposterior dimension of the box as the mandible is moved anteriorly (see Fig. 7).
|
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Oliven, N. Tov, M. Odeh, L. Gaitini, U. Steinfeld, A. R. Schwartz, and A. Oliven Interacting effects of genioglossus stimulation and mandibular advancement in sleep apnea J Appl Physiol, May 1, 2009; 106(5): 1668 - 1673. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Kairaitis, L. Howitt, J. R. Wheatley, and T. C. Amis Mass loading of the upper airway extraluminal tissue space in rabbits: effects on tissue pressure and pharyngeal airway lumen geometry J Appl Physiol, March 1, 2009; 106(3): 887 - 892. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. P. Kirkness, A. R. Schwartz, H. Schneider, N. M. Punjabi, J. J. Maly, A. M. Laffan, B. M. McGinley, T. Magnuson, M. Schweitzer, P. L. Smith, et al. Contribution of male sex, age, and obesity to mechanical instability of the upper airway during sleep J Appl Physiol, June 1, 2008; 104(6): 1618 - 1624. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Howitt, K. Kairaitis, J. P. Kirkness, S. R. Garlick, J. R. Wheatley, K. Byth, and T. C. Amis Oscillatory pressure wave transmission from the upper airway to the carotid artery J Appl Physiol, November 1, 2007; 103(5): 1622 - 1627. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. R. Schwartz, J. Kirkness, and P. Smith Extraluminal tissue pressure: what does it mean? J Appl Physiol, January 1, 2006; 100(1): 5 - 6. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
