INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE INTRINSIC PHARYNGEAL muscles are important in maintaining upper airway (UAW) function during swallowing and airway patency during respiration and preventing UAW narrowing and obstructive sleep apnea (31). Early research identified the pharyngeal dilators, especially the genioglossus (GG), as important regulators of UAW resistance in the oropharynx (reviewed in Refs. 3 and 38). In contrast, less clear are the respiratory-related activity and control of the pharyngeal constrictors, which encase the entire posterior and lateral walls of the naso-, oro-, and hypopharynx (13, 20, 24, 32-34, 36, 37). Nevertheless, the pharyngeal constrictors and an opposing dilator, the stylopharyngeus (SP), have been hypothesized to play an important role in the regulation of UAW patency in patients with sleep apnea (1, 33, 40).

Although reciprocal activation has been observed for mechanically opposing laryngeal adductors and abductors under conditions of increased and decreased respiratory drive (18), it has not been shown for pharyngeal muscles under awake, nonsedated conditions. Similarly, endoscopy of the laryngeal airway has shown the effect of laryngeal adductors and abductors on the glottic aperture (9, 10), but not the effect of mechanically opposing activation of hypopharyngeal muscles on airway dimensions. The importance of the reciprocal activation of mechanically opposing pharyngeal muscles was suggested by Badr et al. (1), who found pharyngeal narrowing during spontaneous and induced central apneas during sleep. They hypothesized that a lack of inspiratory drive during these apneas resulted in a loss of reciprocal inhibition to increase pharyngeal muscle activity and result in the observed UAW narrowing. Although Guilleminault et al. (12) did not find increased superior pharyngeal constrictor activation during spontaneous apneas in subjects, activation of the middle and inferior pharyngeal constrictors may still play a role in the narrowing. Whether the subject is awake or asleep, reciprocal control of pharyngeal and laryngeal muscles may have a role in the neuromuscular component of the "balance of forces" acting to maintain UAW patency and function (7, 17, 31).

The aim of this research was to establish in awake goats the pattern of activation and the effect of altering chemical drive on the pharyngeal constrictors in an intact airway preparation and then, in an isolated airway preparation, to examine the mechanical effect of pharyngeal muscle activation on the supraglottic (SG) space in the UAW (UAW_SG). Our primary focus was on the inferior pharyngeal constrictor [thyropharyngeus (TP)], which encloses the lateral and posterior walls of the hypopharynx, and, to a lesser extent, the middle pharyngeal constrictor [hypopharyngeus (HP)], which lies more rostral. We hypothesized that the TP and HP are reciprocally activated relative to the diaphragm (Dia) and that their activity would be reciprocally activated with increases and decreases in respiratory drive to the Dia. In addition, we hypothesized that the SP, a putative pharyngeal dilator, would be activated in parallel to the Dia but reciprocally activated to its mechanically opposing TP and HP. To examine the potential mechanical consequence of TP and SP activity on the SG during induced central apneas, we simultaneously monitored the electrical activity of these muscles while endoscopically viewing the hypopharynx in awake goats with isolated UAW. We hypothesized that 1) the increase in TP activity and the decrease in SP activity with mechanically induced central apneas are associated with narrowing in the retroglossal space (SG) of the UAW and 2) if the pharyngeal constrictors and SP are anatomically mechanically positioned as agonist-antagonist paired muscles, then the activation of the SP is required for increasing UAW_SG during the resumption of breathing. For comparative purposes, we also studied two reciprocally activated and mechanically opposing laryngeal muscles, the posterior cricoarytenoid (PCA) and the thyroarytenoid (TYA).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Study Protocol 1

Surgical preparation. Surgery was performed to implant chronic electromyographic (EMG) electrodes in the Dia (EMG_Dia), TP (EMG_TP), PCA (EMG_PCA), and TYA (EMG_TYA). In three of nine goats, the HP (EMG_HP) and SP (EMG_SP) electrodes were also chronically implanted. Before surgery the animals received an intravenous injection of ketamine (Ketaset) and xylazine (12:1, 15 mg/kg) for induction of anesthesia before intubation. Anesthesia was maintained with 1.5% halothane (in oxygen).

EMG electrode insertion. Teflon-coated 32-gauge stainless steel bipolar microelectrodes (no. AS637, Cooner Wire) were inserted into the muscles defined above. Two wires ~20 cm long were tied in a square knot at one end, cut, and then covered with a dental cement to form a bead to fix and electrically isolate the cut ends. One-millimeter cuts into the Teflon coat were made within 1 mm of the bead for use as recording sites in a differential bipolar alternating-current amplifier-recording setup. For implants into airway muscles, a midline incision was made on the ventral surface of the neck from the hyoid bone to 4 cm below the thyroid cartilage to expose the lateral aspect of the pharynx and the TP, HP, and SP. The EMG_TP electrode was sewn into the TP midway between the posterior midline of the pharynx and the insertion on the thyroid cartilage 0.5 cm below the cranial laryngeal nerve. To implant the EMG_PCA, a small window was made at the inferior edge of the thyroid cartilage, and the tissue was dissected between the anterior wall of the esophagus and the posterior wall of cricoid cartilage, to expose the PCA. A U-shaped window was made in the lateral wall of the thyroid cartilage to expose the TYA for placement of an EMG electrode. In three animals the EMG_HP electrode was placed ~1 cm above the cranial laryngeal nerve. The SP was located at the cranial tip of the parotid gland, just caudal to the insertion of the styloglossus, and followed to where it descends below the HP muscle. The electrode was sewn to the SP close to its disappearance under the HP. All the wires were looped in the subcutaneous layer of connective tissue and exited on the lateral ventral surface of the neck. For the EMG_Dia, a lateral thoracotomy was performed between the 9th and 10th ribs midway between the sternum and spine. Dia electrodes were implanted in the costal portion of the Dia and exteriorized next to the incision.

Methods of measurement. A custom face mask for each animal was connected to a one-way breathing circuit for measurements of airflow. The one-way breathing circuit had two variations. The first circuit was used in the hypercapnic challenges in which a two-way breathing valve permitted breathing of room air or hypercapnic gases. The second circuit was used for hyperventilation of the animals and consisted of a standard home ventilation circuit (B & F Medical Products) connected to an Ohio ventilator (model 300) or a Bear-2 ventilator. Pneumotachographs were connected in-line on the inspiratory side of both ventilatory circuits and connected to a differential pressure transducer to measure inspiratory airflow.

Experimental design. Before and 2 days after surgery, the goats were acclimated to experimental protocols and were in the awake state to accept passive hyperventilation to or near a hypocapnic apnea. All goats were studied in the prone recumbent position. One week after surgery, 10 hyperventilation studies and 2 hypercapnic studies were performed over a 7- to 10-day period. If hypercapnic studies were performed on the same day, studies were separated by 2 h. For the hypercapnic studies the goats were studied while breathing for 5 min at 3, 5, and 7% inspired CO₂ fraction (FI_CO2). The hyperventilation studies were performed 30 min before and/or after the hypercapnia studies. In addition, spontaneous pauses in breathing with and without a preceding augmented breath were investigated.

Data analysis. Respiratory airflow and EMG signals were processed and analyzed (WinDaq, DATAQ Instruments). Raw EMG data were full-wave rectified and passed through a moving time averager to obtain an integrated EMG signal. The integrated signals from the Grass recorder or from the raw EMG data were analyzed to obtain the total activity (tonic + phasic) for each muscle for individual breaths. The EMG_Dia signal was analyzed for phasic activity only. The EMG_TP, EMG_HP, and EMG_TYA signals were analyzed from the peak of EMG_Dia to the peak of the next EMG_Dia. The mean activity of the EMG_SP and EMG_PCA was analyzed from the beginning to the end of EMG_Dia activity.

Statistical analysis. For hyperventilation studies, the percentage of control EMG_TP (%EMG_TP) and EMG_SP (%EMG_SP) activity for Max and 95% was calculated under conditions of NAFL, AFL, and Obstr for each animal and analyzed for statistical significance by repeated-measures two-way ANOVA. Posttest analysis was performed to compare change in EMG activity at each period with zero and to compare pairs of group means by Bonferroni's test (P < 0.02, adjusted for multiple comparisons).

Study Protocol 2

Surgical preparation. Two surgeries were performed ~6-8 wk apart. The first surgery was performed to implant EMG electrodes, as described in Study Protocol 1. A second surgery to isolate the UAW was performed to create a tracheostoma at the level of the eighth tracheal ring. Postsurgical care of the animal was the same as described above.

Methods of measurement. For subsequent mechanical ventilation and measurements of airflow, an 8-Fr cuffed tracheostomy tube was inserted into the trachea, inflated, and connected to a one-way breathing circuit. The circuit consisted of a standard home ventilation circuit connected to an Ohio ventilator or a Bear-2 ventilator. A pneumotachograph was connected in-line on the inspiratory side of both ventilatory circuits and connected to a differential pressure transducer to measure inspiratory airflow. A custom face mask was made for each animal to stabilize the placement of the endoscope. The connection of the EMG wires for signal processing and recording was the same as described above.

Experimental design. The goats were familiarized with the experimental protocol and trained in the awake state to accept passive hyperventilation to achieve a hypocapnic apnea. Each animal was then studied endoscopically on 2 separate days. Before endoscopy the goats were sedated with an intravenous injection of ketamine and xylazine. The endoscope was then passed, with 4% viscous lidocaine used as a lubricant, through an opening in the mask into one of the nasal passages until the hypopharynx was visualized. The scope was then stabilized using a foam plug in the opening of the mask, and the head of the scope with the camera was mounted on a stand. After the goats recovered from the sedation (30-45 min), they were mechanically hyperventilated through their tracheotomy tube to an apnea at least four times over a 2- to 3-h period. In addition, the video and EMG data during SPA and PSA were recorded. All goats were studied in the prone recumbent position.

Data analysis. Corresponding episodes of mechanical ventilation and induced apneas that were collected simultaneously by the CODAS system and by the custom data-acquisition system were identified to acquire EMG data and video images. Control data from video and EMG data were collected 30 s before a mechanical hyperventilation. After the cessation of hyperventilation, EMG_TP and EMG_SP activities were obtained 1) immediately after cessation of mechanical ventilation, 2) at maximal activity just before the onset of Dia activity (Max), and 3) during the second breath after the return of Dia activity. Video images corresponding to the EMG data were identified by their time code from the custom video data-acquisition system. From the videotape recording, the selected video images were then converted to digital data by use of video analog-to-digital hardware and software. Although the endoscope was moved in each animal during a study, only images with the endoscope at the same location were used in the data analysis for an experimental run. Cross-sectional area (in pixels) of the SG space in the hypopharynx from each video image was analyzed using image analysis software (MetaMorph, version 4.01). The EMG values were expressed as a percentage of the control period, as described above, as well as UAW_SG (%UAW_SG). PSA and SPA were enumerated and classified as described above.

Statistical analysis. The %EMG_TP, %EMG_SP (n = 4), and %UAW_SG were calculated immediately after cessation of mechanical ventilation and at maximal activity just before the onset of Dia activity for each goat and analyzed for statistical significance with repeated-measures ANOVA. Posttest analysis was performed to compare change in %EMG activity and %UAW_SG with control (100%) and to compare pairs of group means with Bonferroni's test (P < 0.02, adjusted for multiple comparisons). To analyze the relationship between UAW_SG and TP and SP activation, a ratio of %EMG_SP activity during the second inspiration to the preceding %EMG_TP expiratory activity (%EMG_SP/%EMG_TP) was calculated and then correlated to UAW_SG for each goat by the Pearson product moment correlation analysis.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Intact Airways

Eupnea. Phasic expiratory activity of the TP and HP during eupneic breathing, observed in all goats with EMG_TP (n = 9) and EMG_HP (n = 3) electrodes, was characterized by a sustained electrical activity during expiration (Fig. 1A). HP activity differed from TP activity only in the amount and discharge pattern. The mean electrical activity of the TP and HP as a percentage of the average peak activity during a swallow was 10-25 and 10-55%, respectively. In all three goats with EMG_SP electrodes, SP activity (Fig. 1A) exhibited an increase in frequency and then in amplitude just before the onset of EMG_Dia activity. EMG_SP activity remained increased throughout inspiration until peak EMG_Dia activity and then briefly decreased to minimal activity after inspiration before the resumption of tonic activity during expiration.

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Representative traces of eupneic breathing in 2 goats. A: trace from goat 2, which shows inspiratory airflow () and raw and moving time average (MTA) signals of stylopharyngeus (SP), hypopharyngeus (HP), thyropharyngeus (TP), and diaphragm (Dia). Dashed blocks below SP trace, short (~100-ms) postinspiratory minimal activities frequently seen in SP signals in 3 goats. B: trace from goat 5, which shows and MTA signals of TP, thyroarytenoid (TYA), posterior cricoarytenoid (PCA), genioglossus (GG), and Dia. A swallow can be seen as a large sharp increase in TP, TYA, PCA, and GG signals interrupting inspiratory flow.

Hyperventilation and induced central apnea. In seven of nine goats, mechanically induced central apneas (122 total, Table 1) increased EMG_TP as EMG_Dia decreased. With the cessation of mechanical ventilation before the resumption of breathing, tonic EMG_TP activity remained elevated or increased further (Fig. 2C). With the resumption of breathing, the slightest EMG_Dia activity markedly attenuated the elevated EMG_TP activity. However, with the return of EMG_Dia activity, a reduced or absent inspiratory airflow was frequently observed in all goats (Table 1), indicating airway narrowing (Fig. 2). The average maximal %EMG_TP in response to mechanical hyperventilation was significant (262.8 ± 120.2%, P < 0.02), with the mean individual response ranging from no change to 480% of control levels (Table 1). Similarly, even when the Dia had returned to 95% of its control, %EMG_TP activity remained significantly increased (135.2 ± 33.6%, P < 0.02). On average, baseline levels of inspiratory airflow and the activation of the Dia and TP did not return for 30 s after mechanical ventilation.

View larger version (45K): [in this window] [in a new window]   Fig. 2.   Representative traces of , TP, TPMTA, Dia, and DiaMTA response to mechanical hyperventilation in goat 2 in awake state. A: breaths before and after start of mechanical ventilation. B: reduced respiratory drive with sustained central apneas of 10 s at ~1 min after start of mechanical ventilation. C: termination of mechanical ventilation and 90 s of recovery. Dotted line below DiaMTA, periods of obstructions or airflow limitations.

View larger version (42K): [in this window] [in a new window]   Fig. 3.   Representative traces of , HPMTA, TPMTA, TYAMTA, SPMTA, GGMTA, and DiaMTA response to mechanical hyperventilation from goat 3. A: breaths just before and after start of mechanical ventilation. B: response at end of mechanical ventilation and resumption of breathing to control levels. Dotted line below DiaMTA, periods of obstructions or airflow limitations.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Representative traces of , TPMTA, TYAMTA, PCAMTA, and DiaMTA response to mechanical ventilation from goat 9. A: control breathing and immediate response to mechanical ventilation. B: 25 s of induced apnea during ventilation. C: response to cessation of ventilation with a decreased tonic TPMTA compared with control and an increase in TYAMTA. Dotted line below DiaMTA, periods of obstructions or airflow limitations.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Percentage of electromyographic activity of TP and SP [%EMGTP (A) and %EMGSP (B)] at maximal response (Max) during induced apneas and when Dia activity returned to 95% of control activity (95%) during conditions of no airflow limitation (NAFL), airflow limitation (AFL), and obstruction (Obstr). EMG activity is expressed as a percentage of control. Each symbol represents average response from a goat. Mean EMGTP responses are as follows: 182.6 ± 54.2% (Max) and 114.1 ± 9.3% (95%) for NAFL, 304.2 ± 114.0% (Max) and 132.9 ± 15.3% (95%) for AFL, and 306.2 ± 140.5% (Max) and 158.4 ± 46.1% (95%) for Obstr. Mean EMGSP responses are as follows: 18.50 ± 3.3% (Max) and 64.3 ± 5.0% (95%) for NAFL, 6.98 ± 1.2% (Max) and 60.2 ± 11.8% (95%) for AFL, and 3.2 ± 1.2% (Max) and 44.4 ± 2.6% (95%) for Obstr.  Significantly different from control (100%), P < 0.02.  Significantly different from NAFL, P < 0.05.

Spontaneous apneas. Most goats exhibited SPA or pauses in breathing after augmented breaths (PSA). For SPA, EMG_TP activity sharply increased by twofold or more at the onset of the apnea (Fig. 6A), and there was a reciprocal decrease in EMG_SP activity. EMG_HP activity remained the same as expiratory levels during the apnea or decreased and then increased slightly through the apnea (not shown). In two goats, TP activity did not increase but remained at eupneic expiratory levels, and EMG_TP activity did not increase above the expiratory level before and after the PSA, whereas EMG_HP activity and EMG_SP activity decreased after the PSA (Fig. 6B).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   , TPMTA, SPMTA, and DiaMTA during spontaneous apneas (SPA) and apneas after a sigh or an augmented breath (PSA). A: SPA in goat 3. B: PSA in goat 7.

Hypercapnia. When FI_CO2 was increased to 3, 5, and 7% to progressively increase ventilation, the phasic expiratory EMG_TP activity progressively decreased in seven of the nine goats (Figs. 7 and 8). The average %EMG_TP decrease from control was statistically significant at 3, 5, and 7% FI_CO2 (Fig. 8). In the remaining two goats, EMG_TP activity decreased and then increased at the highest level of FI_CO2. The EMG_HP response to CO₂ challenge varied. At some times HP activity mirrored the decreases in TP activity; at other times HP activity increased and then decreased with progressive increases in FI_CO2 (not shown). In two of the three goats with SP electrodes, EMG_SP activity increased during increased inspiratory drive (Fig. 8). In the third goat, EMG_SP activity increased for the first two levels of FI_CO2 but then decreased at the highest level (Fig. 7). %EMG_SP activities at 3 and 5% FI_CO2 were significantly different from control.

View larger version (52K): [in this window] [in a new window]   Fig. 7.   Representative traces of effect of progressive hypercapnia on , HP, HPMTA, TP, TPMTA, SP, SPMTA, Dia, and DiaMTA in goat 2. A: end of 5-min control period. B: end of period of breathing at 3% inspired CO2 fraction (FICO2). C: end of period of breathing at 5% FICO2. D: end of period of breathing at 7% FICO2.

View larger version (14K): [in this window] [in a new window]   Fig. 8.   TP and SP activity at 3, 5, and 7% FICO2 in goats with intact upper airways. TP and SP activity is expressed as percentage of control [%EMGTP () and %EMGSP ()]. Each symbol represents an average of %EMGTP and %EMGSP from 2 CO2 studies at each level of FICO2 from each goat. Mean decrease in %EMGTP was 81.8 ± 11.3, 57.2 ± 10.7, and 49.8 ± 51.0% at 3, 5, and 7% FICO2, respectively. Mean increase in %EMGSP was 113.5 ± 4.6, 137.5 ± 10.1, and 129.5 ± 39.7% at 3, 5, and 7% FICO2, respectively.  All mean %EMGTP values and %EMGSP values at 3 and 5% were significantly different from control (P < 0.02).

Isolated Airways With Endoscopy

Eupnea. Similar to intact airway studies, phasic activity was observed in EMG_TP (expiratory) and EMG_SP (inspiratory) during isolated airway studies. Correspondingly, there was a phasic decrease and then an increase in the UAW_SG during expiration and inspiration (Fig. 9, images a-g). The average change in UAW_SG from inspiration to expiration as a percentage of expiratory UAW_SG was significant (46 ± 14%, P < 0.01). At the beginning (image b) and midway (image c) through inspiration, an increase in UAW_SG was associated with a decrease in TP and an increase in SP discharge. At the beginning and through expiration, a decrease in UAW_SG was associated with a reciprocal increase in TP activity and a decrease in SP activity (images d-g). Image d shows that the lateral retroglossal walls move medially as TP activity increases and SP activity decreases. In image e, minimum SG activity was reached just after TP activity had reached a peak, and SP activity was at its lowest level. From images f to g, the UAW_SG remains unchanged as the tonic expiratory activity of the TP and SP remains steady. In addition, in image e the epiglottis has moved dorsally to further reduce UAW_SG. A dotted line drawn in the midline between the aryepiglottic folds in images c-e indicates that the lateral wall moves medially over the aryepiglottic folds as the epiglottis moves dorsally.

View larger version (47K): [in this window] [in a new window]   Fig. 9.   Changes in supraglottic upper airway (UAWSG) area during a single respiratory cycle in an isolated upper airway. Video images a-g were taken from video recording of UAWSG area with use of a fiber-optic endoscope at moments indicated by arrows within respiratory cycle shown by . Phasic changes in UAWSG area correlate with phasic changes in TP and SP.

Hyperventilation-induced central apnea. All the mechanically induced central apneas (3 per goat, 18 total) resulted in an increased tonic TP activity, reduced SP activity, and retroglossal narrowing. During the mechanical hyperventilation, 1) Dia activity was abolished, 2) the phasic decreases in TP during inspiration disappeared, 3) the tonic TP activity increased, 4) SP activity was abolished (only 4 of 6 goats had SP electrodes), and 5) UAW_SG was reduced compared with control expiration (Fig. 10, part 2, image a). After mechanical ventilation, the induced central apnea resulted in a further increase in tonic TP activity and a cessation of tonic and phasic SP activity (Fig. 10, part 2). Despite intermittent efforts with phasic SP activity and sharp reciprocal reductions in TP, the progressive increases in TP activity are associated with an absence of tonic SP activity. Correspondingly, with increases in TP activity, there was progressive narrowing of UAW_SG (Fig. 10, part 2, images a and b). When SP activity increased and tonic TP activity was abolished (3rd respiratory effort) for several seconds, UAW_SG increased (image c). Subsequently, TP activity increased and UAW_SG decreased (not shown). There was a minimal increase in UAW_SG during an inspiration (image d) with phasic increase in SP activity while expiratory levels of TP activity remained elevated. UAW_SG was the smallest (image e) at the peak of TP activity with an absence of SP activity. Thereafter, TP activity abruptly decreased, and phasic and tonic levels of SP activity increased toward control levels. With the return of TP and SP activity to control levels, UAW_SG also returned to normal phasic respiratory changes in dimensions.

View larger version (42K): [in this window] [in a new window]   Fig. 10.   UAWSP, TPMTA, SPMTA, and DiaMTA from a mechanically induced central apnea in goat 3. 1: Prehyperventilation traces of TPMTA, SPMTA, and DiaMTA with increases in UAWSG area during inspiration (image a) and decreases in area with expiration (image b). 2: Posthyperventilation response in TPMTA, SPMTA, DiaMTA, and UAWSG area during induced apnea and return to breathing. Images a-e, changes in UAWSG area that correlate with decreases in TPMTA and increases in SPMTA activity.

View larger version (41K): [in this window] [in a new window]   Fig. 11.   During an induced apnea and in initial breath of recovery, UAWSG area is correlated with reciprocal changes in TP and SP activity. A: traces and video images during control eupneic breathing (1) and after cessation of mechanical ventilation (2). Decreases in tonic and phasic activity of SP and increases in tonic TP activity during mechanical ventilation are associated with a decrease in UAWSG area (image a). Further increases in TP and SP together (image b) or a decrease in SP (2, image c) correlates with an increase and then a decrease in UAWSG area. Progressive return of TP and SP activity toward control parallels increases in UAWSG area. B: relationship between UAWSG area and ratio of SP to TP activity during 2nd breath in resumption of breathing after an induced apnea. Separate symbols are shown for each of 4 goats with TP and SP electrodes for each recovery after mechanical ventilation. Individual correlation coefficient for each goat ranged from 0.86 to 0.96. %UAWSG, percentage of peak UAWSG area during inspiration normalized by average inspiratory UAWSG area during control; %EMGSP/%EMGTP, ratio of %EMGSP activity during 2nd inspiration in relation to preceding %EMGTP expiratory activity. , Goat 3; , goat 7; , goat 10; , goat 11.

Spontaneous Central Apnea During Endoscopy

View larger version (38K): [in this window] [in a new window]   Fig. 12.   TP, SP, and Dia before and after an augmented breath leading to an apnea in goat 5. a: Control inspiration; b: control expiration; c: peak inspiration during an augmented breath when UAWSG area was greatly increased at a time when TP was minimum and SP and probably other laryngopharyngeal dilator muscles are activated; d: peak TP activity corresponding with minimum SP activity and UAWSG area; e: increased UAWSG area with decrease in TP activity; f: end of apnea, where UAWSG area has increased with increased tonic SP activity and a minimum TP activity; g and h: return of phasic inspiratory and expiratory variation, respectively, of TP, SP, and UAWSG area. Range in UAWSG area that was associated with range of TP and SP activity found with spontaneous apneas is shown.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Overall, our data in awake goats show that mechanically opposing pharyngeal constrictor and dilator muscles are reciprocally activated relative to phases of respiration and the degree of activation changes with changes in respiratory drive, and mechanically their activation is associated with changes in HP area. In addition, during the resumption of breathing after induced central apneas, UAW obstructions can result from elevated pharyngeal constrictor and reduced dilator activity. These data indicate that a balance of neuromuscular forces by opposing pharyngeal muscles can greatly influence UAW patency.

Limitations of Study

The use of the terms "reciprocal activation" and "reciprocal control" in discussion of mechanically opposing muscles is acknowledged to be potentially misleading because of the recognized term "reciprocal inhibition" in spinal segmental control of movement. Bainton and Mitchell (2) and Sears et al. (35) demonstrated a reciprocal relationship between respiratory drive to the external and internal intercostal muscles when inspiratory drive was increased or decreased. Reciprocal inhibition of the bulbospinal motoneurons to the Dia and the cranial motoneurons in the nucleus ambiguus to the TP/HP is unknown, yet unlikely, inasmuch as they are not agonist-antagonist paired muscles. Their reciprocal relationship instead may exist at the level of output from the central respiratory pattern generator. To that end, reciprocal activation and control refer to output of the respiratory center to these cranial motoneurons.

Eupnea

Among awake goats, the intensity of TP and HP discharge varied frequently. This difference is probably not due to a difference in the electrical leak from nearby activated muscles. The sensing portions of our electrodes were electrically isolated, and the mean distance between the sensing portions was consistent (2.5 mm). The EMG_TP and EMG_HP electrodes were visually implanted in the same location and sewn in and not through the muscle to reduce the likelihood of recording electrical activity from nearby muscles. Thus the recording fields were very similar among goats. A second potential cause of the difference between the TP and HP relates to fiber-type composition. The HP and TP exhibit a rostral-to-caudal decrease in type II and an increase in type I muscle fibers of 5% (16) and a functional difference in activation within the HP and TP (24, 33). A third potential cause of TP and HP difference relates to the motoneurons supplying the HP and TP, which are differently distributed in the dorsomedial subgroup of the compact cell group of the nucleus ambiguus (19). Accordingly, the difference in discharge intensity between the TP and HP suggests that the pharyngeal constrictors may not be homogeneous in terms of control and possible function, which has been suggested previously (25).

Decreases in Respiratory Drive and Central Apneas

Irrespective of the elevated TP/HP activity during induced apneas and the resumption of Dia activity, the presence of an inspiratory effort would sharply attenuate TP activity. This suggests that, regardless of the respiratory drive to the TP and HP, the respiratory phasic activity to these muscles is tightly controlled. In contrast, the consistent TP response to mechanical hyperventilation and the variable HP response again suggest that these pharyngeal constrictors may not be homogeneous in muscle fiber-type composition and/or controlled in a parallel manner.

Increases in Respiratory Drive

SG Area and Pharyngeal Muscle Activity

The decrease in UAW_SG area with elevations in TP activity and minimal SP activity within a goat during the recovery from an induced apnea supports the finding from the intact UAW data that had shown AFL and Obstr with similar changes in the TP and SP. The implication of this finding is important, because traditionally it has been thought that induced or central apneas resulted in a lack of neuronal output to the respiratory muscles. Any signs of UAW narrowing, direct or indirect, support the hypothesis that, during central apneas in the awake state, a neuromuscular imbalance can occur with elevated TP activity and minimal SP activity and lead to significant UAW narrowing. The data showing that inspiratory flow did not return to baseline levels for a given level of EMG_Dia until the activity of the pharyngolaryngeal constrictors (TP and TYA) and their agonists (SP and PCA) returned to control levels support the idea that induced apneas can lead to a sustained neuromuscular imbalance in the UAW after the resumption of breathing. In addition, although other pharyngeal dilators were not monitored during this study, it is very conceivable that they may also be reduced for a given level of Dia activity during recovery from an induced apnea and contribute to changes in UAW_SG. An imbalance in the activation of UAW dilators (PCA and GG) relative to the Dia was found in anesthetized and awake cats (7, 14), but not in the GG in awake humans (26, 27, 31).

Although we observed a consistent pattern in TP and SP activity to suggest that they may be the cause of the observed signs of UAW narrowing, we have also provided evidence that sometimes an increase in TYA activity without an increase in TP activity was associated with UAW narrowing. This latter observation was also suggested by findings of Kianicka et al. (18), in which an increase in TYA activity and a loss of PCA activity were associated with glottic aperture closing.

Badr et al. (1) hypothesized that the observed narrowing in the oropharynx and hypopharynx in sleeping humans during spontaneous and induced central apneas may in part be a result of an increase in pharyngeal activity due to a loss of inspiratory inhibition. In the awake goat, our findings support this hypothesis by Badr et al. However, intrinsic pharyngeal activity is known to decrease with sleep (3, 15, 20, 31, 34, 38, 39), and therefore our findings during wakefulness may not be applicable during sleep. In a recent study in sleeping humans with an endoscope placed in the UAW, Guilleminault et al. (12) observed narrowing of the UAW during spontaneous central apneas without an increase in activity of the middle and superior pharyngeal constrictors. Although the study of Guilleminault et al. did not support the hypothesis of Badr et al. (1), it also did not define the portion of the UAW under the view of the endoscope, nor did it rule out the involvement by the inferior pharyngeal constrictors.

Balance-of-Forces Theory and UAW Patency

View larger version (16K): [in this window] [in a new window]   Fig. 13.   Balance-of-forces theory in maintaining pharyngeal luminal area with proposed extension of static and dynamic forces and neuromuscular imbalance. A: balance of dynamic and static forces, which maintains pharyngeal patency. B: elevation of neuromuscular closing forces, which results in an imbalance in UAW forces and luminal narrowing. C: shift in neuromuscular opening, which results in pharyngeal luminal airway dilation. A shift in fulcrum to left, which is proposed to have an effect similar to changes in UAW anatomy, to narrow airway would require an increase in neuromuscular opening forces to maintain pharyngeal luminal area.

The reciprocal activation of the mechanically opposing TP and SP muscles found in this study supports a simple model of a balance of neuromuscular forces acting in concert to regulate UAW_SG. For example, during an induced apnea in the awake state, our data support the concept that a loss of tonic SP and the reciprocal increases in TP activity narrowed UAW_SG and shifted the scales of pharyngeal patency of our model toward closure (Fig. 13B). The shifts in tonic TP and SP activity and the subsequent change in UAW_SG within an apnea (Fig. 11A) further support the dynamic effects of these muscles on UAW_SG. Conversely, an increase in SP activity and a decrease in TP activity during an augmented breath swing the scales to increase the neuromuscular opening forces, leading to an increase in UAW_SG area (Fig. 13C). The shift in the neuromuscular forces toward closure in the first example also suggests a subsequent problem of opening. This latter problem was found during the resumption of breathing after an induced apnea in the intact airway. As observed, the increase in TP activity during the apneas led to narrowing of the UAW_SG; however, a simple inactivation of the TP during a breath would not reestablish normal airway dimensions, unless 1) recoil of the tissue was high and/or 2) the tonic or phasic dilator activity was sufficient for airway dimensions. The sustained elevation of TP activity and reduction of SP activity during the resumption of breathing relative to the Dia (Figs. 5 and 11B) further support the concept of neuromuscular imbalance leading to UAW dysfunction.

Despite the extensions presented for this balance-of-forces model of pharyngeal airway patency, several problems exist in using this model to describe pharyngeal patency. About 28 paired muscles that extend from the nasopharynx to the larynx are involved in the intricate control of the many functions of the UAW (3, 17, 39). In addition, the UAW is not a homogeneous tube. It is a complex muscular structure for mastication and swallowing, with two ports of entry for air (nasal and oral). Therefore, the in vivo determination of airway patency at any one point in the UAW is dependent on a pattern of activation and the sum of the vector of forces produced by the group of UAW muscles (39). Therefore, at any one particular point along the UAW, a specific balance of static and dynamic forces would be acting on that location of the UAW. In this study we restricted our focus to the UAW_SG and two agonist-antagonist muscles that have a direct influence on lateral and posterior hypopharyngeal wall movement. In the case of the velopharynx, Launois et al. (22) showed that the palatal muscular group is considerably more complex than the simple model presented here.

Conclusions