Coordination of cough and swallow: A meta-behavioral response to aspiration

Abstract

Airway protection is the prevention and/or removal of material by behaviors, such as cough and swallow. We tested the hypothesis that cough and swallow, in response to aspiration, are a “meta-behavior” and thus are coordinated and have alterations in excitability to respond to aspiration risk and maintain homeostasis. Anesthetized animals were challenged with a protocol that simulated ongoing aspiration and induced both coughing and swallowing. Electromyograms of the mylohyoid, geniohyoid, thyrohyoid, thyroarytenoid, thyropharyngeus, cricopharyngeus, parasternal, rectus abdominis muscles together with esophageal pressure were recorded to identify and evaluate cough and swallow. During simulated aspiration, both cough and swallow intensity increased and swallow duration decreased consistent with a more rapid pharyngeal clearance. A phase restriction between cough and swallow was also observed; swallow was restricted to the E2 phase of cough during chest wall and abdominal motor quiescence. These results support the conclusion that the cough and swallow pattern generators are an airway protective meta-behavior. The resulting alterations in swallow drive during the simulated aspiration protocol also supports the conclusion that the trachea provides feedback on swallow quality, informing the brainstem about aspiration incidences. The overall coordination of cough and swallow led to the additional conclusion that mechanically the larynx and upper esophageal sphincter act as two separate valves controlling the direction of positive and negative pressures from the upper airway into the thorax.

1.0 Introduction

Airway protection is the coordination of several behaviors to prevent and/or correct the aspiration of material into the lungs. Two important behaviors in airway protection are swallow and cough. Swallowing is a coordinated behavior that is dependent upon afferent feedback for initiation and modulation. Touch, pressure, and/or liquid on the tongue, faucial pillars, soft palate, uvula, epiglottis, pharyngeal wall, and/or junction of the pharynx/esophagus can induce swallowing (Miller & Scheeington, 1916; Pommerenke, 1928; Storey, 1968; Miller, 1982). Cough is a reflex which responds to material entering the airway by producing high velocity airflows creating shearing forces in larger airways and squeezing actions in smaller airways to remove mucus and foreign matters (Ross et al., 1955; Fontana & Lavorini, 2006; Widdicombe & Chung, 2007).

Disordered airway protection, is clinically defined as intrusion of material below the level of the vocal folds during swallowing (dysphagia), (DePippo et al., 1992; Aviv et al., 1996; Rosenbek et al., 1996a; Robbins et al., 1999; McCullough et al., 2001a; Kalia, 2003; Robbins et al., 2008; Cichero & Altman, 2012), and/or an impaired/lack of cough response to aspiration (dystussia) (Muz et al., 1989; Martin et al., 1994; Smith Hammond et al., 2001; Kelly et al., 2007). Cough is the most noticeable response to aspiration; however there are a host of responses including swallowing, expiration reflex, increased mucous secretions, and/or alterations contractions of the smooth muscle lining the airway (Bolser et al., 1995; Belvisi & Bolser, 2002; Bolser & Davenport, 2007; Vovk et al., 2007). The patient may also exhibit other clinical indicators such as postural changes and changes in voice quality (McCullough et al., 2001b; McCullough et al., 2005; Logemann et al., 2008).

Cough and swallow can both be elicited in experimental models. Cough can be initiated by mechanical stimulation of the trachea or larynx, (Bolser & DeGennaro, 1994; Bolser et al., 2006; Wang et al., 2009; Poliacek et al., 2011) or inhalation of an irritant aerosol (Bolser et al., 1995); and swallow by injection of water into the oropharynx, mechanical stimulation of the pharynx, and/or electrical stimulation of the superior laryngeal nerve (Miller & Sherrington, 1915). Swallow is proposed to be generated by a dorsal and ventral medullary network that may share upper airway motor outputs with that of the respiratory pattern generator (Jean, 2001). The central initiation and rhythmogenesis of swallow is thought to be restricted to the dorsal swallow group and not controlled by the ventral respiratory pattern generator (Jean, 2001). On the other hand, the available evidence supports reconfiguration of existing elements of the respiratory pattern generator in the production of coughing (Shannon et al., 1996; Shannon et al., 1998; Shannon et al., 2004), although some control functions for cough are mediated by brainstem systems that are not required for breathing (Bolser & Davenport, 2002). As such, preclinical data support some sharing of neural elements between the pattern generators for swallow, cough, and breathing but the core network for swallow appears to be anatomically separate within the brainstem from that for cough and breathing (Dick et al., 1993; Oku et al., 1994; Shannon et al., 1996; Baekey et al., 2001).

There are clear clinical associations between dysphagia (disordered swallow) and dystussia (disordered cough) in those with Parkinson’s disease and stroke (Smith Hammond et al., 2001; Pitts et al., 2008; Pitts et al., 2009; Smith Hammond et al., 2009). Training paradigms to influence or prevent episodes of aspiration have been intensely studied (Rosenbek et al., 1991; Schmidt et al., 1994; Ali et al., 1996b; Rosenbek et al., 1996a; Rosenbek et al., 1996b; Aviv et al., 1997; Rosenbek et al., 1998; Aviv et al., 2002; Miller et al., 2006; Clave et al., 2008; Miller, 2008; Troche et al., 2010; Voytas & Al Rifai, 2012), however, few treatments have demonstrated therapeutic effectiveness in modifying or preventing aspiration across patient populations. This may be because the primary treatment is for dysphagia with little intervention for dystussia (Bath et al., 1999; Foley et al., 2008; Wheeler-Hegland et al., 2009).

The clinical association between dysphagia and dystussia could be explained simply by the fact that the minimal neural elements for both cough and swallow are located in the brainstem. In this scenario, disease processes such as stroke, would be expected to affect each behavior similarly because of this anatomical association. However, dystussia and dysphagia can occur in patients with neurologic diseases that do not directly affect the brainstem. Alternatively, a more complex, but not mutually exclusive, hypothesis could account for co-depression of cough and swallow in neurological diseases. A central control system could exist that that coordinates the expression of these behaviors to optimize airway protection. . Additionally, the coordinated expression of several behaviors, each with unique regulation, to achieve a common goal – such as cough and swallow - is consistent with the hypothesis that response to aspiration is a “meta-behavior.” This is analogous to the behavior of autonomous agents used to schedule responses when two or more components are combined to react to incoming stimuli (Guessoum & Briot, 1999). Features of the behavior include “precedence” in which the actions that have little to no central processing take precedence over actions which require additional processing, and “blocking” in which any of the components (behaviors) can block any other action until it is completed. An additional assumption is that the gain or excitability of the components (behaviors) can also be altered without sacrificing homeostasis (Fibla et al., 2010). If these hypotheses are true, this system may be affected and/or impaired by multiple neurologic disease states, which may account for the known clinical associations between disordered cough and swallow. However, the evidence for a coordinating mechanism between reflexive swallow and cough is based solely on inferences from clinical observations.

The aims of this study were to determine if the cough and swallow motor patterns are coordinated and, if so, identify operational principles which govern their interactions following an aspiration event. We hypothesized that during a simulated aspiration, there will be minimal overlap of the cough and swallow behaviors.. Furthermore, we speculated that the behaviors interact spatially to optimize mechanical effectiveness during aspiration.

2.0 Methods

Experiments were performed on 17 spontaneously breathing adult male cats. Ethical approval of the protocol was confirmed by the University of Florida Intuitional Animal Care and Use Committee (IACUC). The animals were initially anesthetized with sodium pentobarbital (35-40 mg/kg i.v.); supplementary doses were administered as needed (1-3 mg/kg i.v.). A dose of atropine sulfate (0.1-0.2 mg/kg, i.v.) was given at the beginning of the experiment to reduce secretions from repeated tracheal stimulation. Cannulas were placed in the femoral artery, femoral vein, and trachea. An esophageal balloon was placed via an oral approach to measure pressure in the midthoracic esophagus. Arterial blood pressure and end-tidal CO₂ were continuously monitored. Body temperature was monitored and maintained at 37.5 ± 0.5 °C using a heating lamp and pad. Arterial blood samples were periodically removed for blood gas analysis. PO₂ was maintained using air mixtures with enriched oxygen (25-60%) to maintain values above 100 mm Hg.

Electromyograms (EMG) were recorded using bipolar insulated fine wire electrodes. Seven muscles were used to evaluate cough and/or swallow function: mylohyoid, geniohyoid, thyrohyoid, thyropharyngeus, thyroarytenoid, cricopharyngeus, parasternal, and rectus abdominis. The digastric muscles were dissected away from the surface of the mylohyoid and electrodes were placed on the left mylohyoid. A small horizontal incision was made at the rostral end of the right mylohyoid followed by an incision following the midline for approximately 1cm to reveal the geniohyoid underneath. Electrodes were placed 1cm from the caudal insertion of the geniohyoid muscle. The thyroarytenoid electrodes were inserted through the cricothyroid window into the anterior portion of the vocal folds, which were visually inspected post-mortem. Rotation of the larynx and pharynx counterclockwise revealed the superior laryngeal nerve, which facilitated placement of the thyropharyngeus muscle electrodes. The thyropharyngeus is a fan shaped muscle with the smallest portion attached to the thyroid cartilage; electrodes were placed in the ventral, caudal portion of the muscle overlaying thyroid cartilage within 5 mm of the rostral insertion of the muscle. To place the electrodes within the cricopharyngeus muscle, the larynx and pharynx were rotated counterclockwise to reveal the posterior aspect of the larynx. The tissue was palpated for the edge of the cricoid cartilage and electrodes were placed just cranial to the edge of this structure. Thyrohyoid electrodes were inserted approximately one cm rostral to the attachment to the thyroid cartilage; those for the parasternal muscle were placed in the third intercostal space, just adjacent to the sternum. The rectus abdominis electrodes were located approximately two cm caudal to the xiphoid process just medial to the margin of the rectus abdominis. The positions of all electrodes were confirmed by visual inspection and EMG activity patterns during breathing, cough and swallow.

Cough was induced by mechanical stimulation of the extra and intra-thoracic trachea using a thin polyethylene catheter (diameter 1.27mm). The catheter was manually rotated along the length of the intrathoracic trachea. Cough was defined as a burst of activity in the parasternal EMG, followed by (and partially overlapping) a burst in the thyroarytenoid and rectus abdominis, along with a negative to positive change in esophageal pressure. To initiate swallowing, a one-inch long, thin polyethylene catheter (diameter 2.37 mm), attached to a 6cc syringe was placed into the oropharynx. Water was injected into the pharynx via a syringe (3 cc’s). Swallowing was defined as a quiescence of the cricopharyngeus with overlapping activity in the mylohyoid, geniohyoid, thyropharyngeus, thyrohyoid, thyroarytenoid and the parasternal (representing the schluckatmung or swallow breath) (Wilson et al., 1981; Gestreau et al., 2000; Saito et al., 2002; Bonis et al., 2011).

The protocol included non-overlapping stimulus intervals for sequential induction of cough and swallow behaviors followed by temporally overlapped stimulation trials. Mechanical stimulation of the trachea mimics aspiration of material into the trachea and readily provokes vigorous coughing in this model (Bolser et al., 2001; Belvisi & Bolser, 2002; Poliacek et al., 2007; Wang et al., 2009; Poliacek et al., 2011). Injection of water into the oropharynx is a reliable stimulus for swallow. The purpose for this temporal overlap was to simulate aspiration in the presence of a water filled pharyngeal airway. As such, this protocol closely approximated not just a single aspiration event, but the risk of further ingestion of material into the tracheal airway from the pharynx. This process is similar to that experienced by humans with pathologies that predispose them to significant risk of aspiration during a meal (Horner & Massey, 1988; Bushmann et al., 1989; Coates & Bakheit, 1997; Potulska et al., 2003; Prosiegel et al., 2004; Daniels et al., 2006; Miller, 2008; Cabre et al., 2010). The following sequential protocol was used: two trials with mechanical stimulation of the trachea for 20s each, two trials with pharyngeal injection of water, two combined stimulation trials with mechanical stimulation of the trachea for 20s and water injection into the pharynx 5s after the onset of the cough stimulus.

All EMG signals were amplified, filtered (200-5000 Hz), rectified, and integrated (time constant 50 ms). EMG amplitude measures were normalized to the largest cough or swallow respectively. Cough phase durations were measured using the definitions from Wang et al (2009) . The inspiratory phase (CT_I), the expiratory phase with active muscle activity (CT_E1), passive expiratory phase (CT_E2), and total cough (CTtot) durations were measured. CT_I was defined as the onset of parasternal activity to the maximum burst of the parasternal EMG, CT_E1 was defined as the maximum burst of the parasternal EMG to the end of the abdominal EMG activity, and CT_E2 was defined as the end of the abdominal motor burst to the onset of the parasternal EMG activity for the next cough in the epoch. Swallow duration measures were defined as laryngeal elevation: the onset of the mylohyoid to the end of the EMG burst in the geniohyoid; upper esophageal opening: from the sharp decrease in cricopharyngeus activity to its resumption; and total swallow duration: the onset of the mylohyoid activity to the resumption of the cricopharyngeus activity.

Results are expressed as means ± standard error. For statistical analysis Student’s paired t-tests were used to identify differences. Results were corrected for multiple comparison by controlling false discovery rate to 0.05 (Benjamini & Hochberg, 1995). Relationships between normalized burst amplitudes of the cricopharyngeus, parasternal, abdominal, and esophageal pressure during cough was evaluated by linear regression analysis. Relationships between swallow and cough phase were analyzed for randomness using the runs test. A difference was considered significant if the p-value was less than 0.05.

3.0 Results

Injection of a water bolus into the oropharynx elicited an average of 2.1± 0.2 swallows with water alone and significantly more swallows 2.9± 0.4 when combined with mechanical stimulation of the trachea (p< 0.01). Mechanical stimulation of the trachea alone elicited an average of 7.5± 0.9 coughs per trial. Raw EMG traces for cough and swallow are represented in Figure 1. Water alone elicited swallow during breathing, and 3% of swallows (2 of 73) were in the inspiratory phase of breathing, 76% of swallows (56 of 73) were during the expiratory phase of breathing, 11% of swallows (8 of 73) were in the transition from inspiration to expiration, and 10% of swallows (7 of 73) were in the transition of expiration to inspiration. The combined stimulation modality produced swallows during sequential cough efforts and 95% of the swallows (87 of 92) were completed during the E₂ cough phase. There were four instances of a swallow occurring from the transition of E₂ to a cough I phase and one instance of swallow occurring from the transition of E₂ to a eupneic inspiration. Figure 2 is a histogram of swallow initiation and termination within the cough phases during the aspiration protocol. Each cough phase (I, E1, and E2) duration measures were normalized to 100, and each phase was segmented into 10 bins (30 bins total over the three phases). The runs test for swallow onset (p < 0.001) and swallow termination (p < 0.001) during the cough phases was non-linear.

Figure 1.

Raw EMG traces of the coordination of cough and swallow. Swallow is denoted by circles and cough by arrows. The first panel is injection of water into the pharynx resulting in four swallows, and the second panel is coordinated coughs and swallows resulting from the aspiration protocol.

Figure 2.

Histogram of swallow initiation and termination within the cough phases during the aspiration protocol. The occurrence and termination of swallows were plotted across cough phases that were segmented into quartiles. Swallows were executed primarily in the E2 cough phase. The solid line is swallow initiation as demarked by elevation of the hyoid and relaxation of the upper esophageal sphincter, and swallow termination was identified by increased tone to the upper esophageal sphincter following relaxation. There is one occurrence of a swallow being initiated during the E1 phase and four occurrences swallows being completed during the I phase of the subsequent cough (n=73 swallows from 17 animals).

The swallows which occurred during the combined stimulus modality had significantly greater EMG amplitudes for the parasternal (p < 0.04), geniohyoid (p < 0.01), thyrohyoid (p< 0.01), thyropharyngeus (p < 0.001), and cricopharyngeus (p ≤ 0.01) (see Table 1). There was also a significant decrease in the parasternal burst duration (p< 0.04). The burst duration of the laryngeal, pharyngeal, and submental muscles did not change significantly in the combined stimulus modality, however the duration of the laryngeal elevation (p ≤ 0.01), the opening of the upper esophageal sphincter (p < 0.01), and the total swallow duration (P < 0.01) were significantly decreased.

Table 1.

Effect of water and water plus tracheal stimulation (TS + Water) on normalized EMG amplitudes and durations (ms) of selected swallow-related muscles.

Amplitude (% of maximum)^	Water	TS + Water
Parasternal	51 ± 6	67 ± 4*
Mylohyoid	68 ± 5	78 ± 5
Geniohyoid	69 ± 4	82 ± 3**
Thyroarytenoid	75 ± 3	80 ± 3
Thyrohyoid	78 ± 3	88 ± 1**
Thyropharyngeus	61 ± 4	74 ± 2**
Cricopharyngeus (post-relaxation burst)	52 ± 6	70 ± 4**
Duration (ms)	Water	TS ± Water
Laryngeal Elevation□	507 ± 38	430 ± 30**
Total Swallow	572 ± 43	508 ± 36**
Upper Esophageal Sphincter Open	455 ± 25	386 ± 14**
Parasternal	419 ± 46	331 ± 43*
Mylohyoid	468 ± 39	420 ± 31
Geniohyoid	366 ± 33	360 ± 32
Thyroarytenoid	440 ± 25	418 ± 34
Thyrohyoid	398 ± 39	392 ± 24
Thyropharyngeus	182 ± 24	185 ± 25

^*p ≤ 0.05

^**P ≤ 0.01

^□Onset of mylohyoid activity to offset of geniohyoid activity.

^{^}EMG’s were normalized to the maximum EMG amplitude.

Repetitive coughs were compared before and after the introduction of the swallow stimulus into the oropharynx. The coughs which occurred after the introduction of the swallow stimulus had significantly greater EMG amplitudes for the parasternal (p < 0.05), rectus abdominis (p < 0.001), thyropharyngeus (p < 0.01), and the positive component of esophageal pressure (p < 0.02) significantly increased (Table 2 and Figure 3). There were no significant changes in CT_I and CT_E1 phases of cough. Additionally, CT_E2 durations with and without a swallow present were compared. Following the injection of water into the oropharynx, CT_E2 with a swallow was significantly longer (p < 0.05) than those without a swallow.

Table 2.

Effect of tracheal stimulation (TS) and water plus tracheal stimulation (TS + Water) on normalized EMG amplitudes and durations (ms) of selected cough-related muscles and pressure. Change in CT_E2 duration when no swallow was present versus swallow present.

Amplitude (% of maximum)^	TS	TS + Water
Parasternal	58 ± 4	71 ± 3*
Rectus Abdominis	48 ± 5	69 ± 3***
Thyroarytenoid	52 ± 5	62 ± 5
Cricopharyngeus	71 ± 3	73 ± 3
Thyrohyoid	54 ± 6	65 ± 5**
Esophageal Pressure (cm H20)	19 ± 3	31 ± 5**
Duration (ms)	TS	TS ± Water
Inspiratory	823 ± 93	835 ± 115
Compression	139 ± 19	147 ± 16
Expiratory 1 (E1)	437 ± 38	408 ± 28
Total Cough Cycle	5286 ± 1169	4338 ± 1006
Duration (ms)	No swallow	With swallow
Expiratory 2 (E2)□	1497 ± 890	3097 ± 2783*

^*p ≤ 0.05

^**p ≤ 0.01

^***p ≤ 0.001

^□E2 durations after water injected into the pharynx

^{^}EMG’s were normalized to the maximum EMG amplitude.

Figure 3.

Change in cough motor drive with injection of water into the oropharynx. Triangle denotes swallow. Note the increased rectus abdominis and parasternal electromyographic activity and expiratory esophageal pressure in the second cough.

3.1 Additional observations on pharyngeal muscle activities during eupnea and cough

The cricopharyngeus (upper esophageal sphincter) was active during repetitive cough (Figure 1, 3, and 4). It had an augmenting pattern over the inspiratory phase, peaked at the transition from the inspiratory to the expiratory phase, and declined in magnitude during the expiratory phase. The cricopharyngeus EMG amplitude was not correlated with the peak in the positive esophageal pressure (r² = 0.005), rectus abdominis amplitude (r² = 0.08), or parasternal amplitude (r² = 0.04) during coughing. The thyropharyngeus muscle had expiratory phasic activity during eupnea which decreased during the cough stimulus and repetitive coughing (Figure 4).

Figure 4.

Cannula was inserted into the trachea at the downward arrow. Cough cycles are noted with stars. Expiratory phasic thyropharyngeus EMG activity was suppressed, and cricopharyngeus EMG activity was increased at the onset of the tracheal stimulation. The cricopharyngeus EMG has dynamic activity during cough with the peak during the transition from the inspiratory to the expiratory phase.

4.0 Discussion

This is the first study to examine the coordination of cough and swallow during the aspiration response. The aspiration protocol elicited significantly more swallows than the water bolus without concurrent tracheal stimulation. Most (95%) of the aspiration swallows occurred during the E2 phase of cough. The total swallow duration was decreased without decreasing the burst duration of any pharyngeal/laryngeal/submental muscles, and there was an increase in the EMG magnitude of the pharyngeal muscles, inspiratory muscle activity for schluckatmung production, and hyoid elevators (geniohyoid and thyrohyoid) for swallows that occurred during repetitive coughing episodes. The duration of cough E2 phases which contained a swallow were significantly longer than in control trials. Moreover, chest wall and abdominal (inspiratory and expiratory) and thyrohyoid muscle electromyographic activity increased during coughs following the injection of the water into the pharynx as did expiratory esophageal pressures. Cricopharyngeus activity was also elevated during coughing.

4.1 Phase restriction of swallow

Swallowing that is normally executed during eupnea has been intensely studied, and there is a phase preference for swallows to occur during the expiratory phase of breathing; specifically that 80% of swallows occur during the expiratory phase of breathing in humans (Martin-Harris et al., 2003; Wheeler Hegland et al., 2009; Wheeler Hegland et al., 2011), cats (Dick et al., 1993), goats (Feroah et al., 2002a; Feroah et al., 2002b; Bonis et al., 2011), and rats (Saito et al., 2003). Our observations are indicative of a more rigid regulatory control of swallowing during aspiration promoting events, i.e. swallow during repetitive coughing. We propose the concept of phase restriction to explain the fixed occurrence of swallows during the quiescent period following the cough expulsion (cough E2) (Figure 1 and 2). This idea is reinforced by cough E2 prolongation, ensuring adequate time for swallow initiation and completion before the onset of the next inspiratory phase (Figure 1 and 2). This rigid control system is necessary because both cough and swallow share the pharyngeal airway, and the presence of food or liquid in this airway segment represents a significant aspiration risk.

Forssberg, Grillner and Rossignol (1975), Sillar (1991), Watson (1992), and Pearson (1993) all proposed a filtering of afferent information to ensure an appropriate motor response within the context of the ongoing motor activity. This concept includes roles for afferent feedback in the establishment of the temporal order of motor behaviors and in controlling transitions from one behavior to another. Furthermore the decision to produce a behavior, based on afferent information, is dependent on the state of the ongoing motor behaviors, like breathing and/or coughing (Forssberg et al., 1975; Sillar, 1991; Watson, 1992; Pearson, 1993). This effect has been established in other systems including: chick hatching and stepping (Bekoff et al., 1987), hand movements in primates (Sanes et al., 1985), and flight in the locust (Wolf & Pearson, 1987) by using models of deafferentation. Our results support a theory of filtering of pharyngeal and laryngeal afferent information by brainstem networks to inhibit the swallow pattern generator during the inspiratory and active expiration phases of cough.

4.2 Dual valve system

Our results also suggest a highly coordinated control of both the laryngeal airway and upper esophageal sphincter such that they may represent a dual valve system regulating pressure between upper airway and the thoracic cavity (Figure 5). To our knowledge, prior work on cough or swallowing did not observe signatures of a unified control system for the larynx and upper esophageal sphincter. The laryngeal adductor/abductor and upper esophageal sphincter control air/bolus flow into or out-of the lungs or esophagus. For example, during swallowing there is maximal activity of the laryngeal adductor muscles and maximum relaxation of the upper esophageal sphincter, and thus pressures move the bolus into the esophagus and not into the larynx/trachea. Our data support the presence of a reciprocal relationship for cough as well. During our experiments the EMG activity of the cricopharyngeus muscle was very sensitive to mechanical stimulation in the trachea (Figure 4). EMG activity of this muscle steadily increased over the inspiratory period, peaking during the transition from inspiration to E1 and decreased during the E1 and E2 phases. This mechanism mechanically “seals” the upper esophageal sphincter during cough to prevent loss of intra-thoracic pressure and thus maximizes cough effectiveness. This increased cricopharyngeus muscle activity also reduces the risk of esophageal reflux into the pharynx.

Figure 5.

Dual valve system hypothesis. The side by side circles represent the two valves, the larynx and the upper esophageal sphincter. Closure of the two valves is controlled by the thyroarytenoid and the cricopharyngeus muscles, respectively. The valve aperture is represented on a scale from white (maximum opening) to black (maximum closure). Note: during the expiratory phase of eupnea there is some thyroarytenoid activity. ** Thyroarytenoid and cricopharyngeal electromyogram waveform averages across multiple behavior occurrences during cough and swallow. The vertical gray line in the second and third panels is the midline marker of the behavior execution, and during cough this represents the transition from inspiration to active expiration.

Given this dual valve system (Figure 5), swallowing can occur during the inspiratory phase of breathing because trans-laryngeal flows and intra-thoracic pressure are relatively low. We propose that swallow-related laryngeal adductor activity (glottic and supra-glottic) is sufficient to close the airway during a eupneic breathing cycle; however the much higher inspiratory and expiratory flows and pressure during coughing would make it mechanically difficult for a successful bolus transfer across the esophageal sphincter. Even during the compression phase (a time of little or no trans-laryngeal flow) the pressure in the thoracic cavity is high and hindering bolus movement.

Shannon, et al (1996; 1998; 2004), and Bolser and Davenport (2002) proposed that the temporal regulation of the expiratory phase of cough is altered by excitability of the expiratory-augmenting late neurons (i.e neurons which fire during the expiratory phase of breathing with significantly more action potentials in the second half of the expiratory period as compared to the first half) within the Bötzinger Complex. We propose that the swallow pattern generator interacts with these neurons, increasing their excitability to prolong the cough E2 phase. These control mechanisms may represent a neural substrate that is essential for airway protection, and may help to explain how a wide range of neurologic diseases results in related dysphagia and dystussia (Smith Hammond et al., 2001; Pitts et al., 2008; Pitts et al., 2009; Smith Hammond et al., 2009; Pitts et al., 2010).

An absent cough response to aspiration is a hallmark of dysphagia, and may be a manifestation of an impaired sensory feedback control system. During a clinical evaluation of swallow, material can penetrate to the level of the vocal folds without a cough or expiration response in young adults during swallowing (Daniels et al., 2004; Daggett et al., 2006). We hypothesize that the trachea, with or without the laryngeal afferent activation, provides feedback on swallow quality, on a cycle by cycle basis. Additionally, dysphagia caused by neurotrauma (Aviv et al., 1996) or damage to the vagus (Halum et al., 2003) can result in a condition known as cricopharyngeal bar and laryngeal dysfunction. Cricopharyngeal bar is defined as hyperactivity of the cricopharyngeus during swallow, resulting in obstruction of the esophageal opening during swallowing leading to residue in the pharynx (Aviv et al., 1996; Halum et al., 2003). These clinical findings are consistent with the dual valve system, representing a control system that when dysfunctional, results in pathological behavior of both valves.

4.3 Pharyngeal clearance

Jean (1984, 2001) discussed oral and pharyngeal afferent feedback as a primary modulator of the swallow motor pattern generator. More specifically the size, texture, taste, temperature of the bolus or pharyngeal distention can alter the swallow pattern (Kahrilas & Logemann, 1993; Logemann et al., 1995; Ali et al., 1996a; Rademaker et al., 1998; Ertekin et al., 2000; Hiss et al., 2001; Jean, 2001; Kendall & Leonard, 2001; Kendall et al., 2001; Butler et al., 2004; Chee et al., 2005; Leow et al., 2007; Troche et al., 2008; Humbert et al., 2009; Thexton et al., 2009; Yamamura et al., 2010). This effect has not been previously demonstrated by stimulation of tracheal afferents. Mechanical stimulation of the trachea activates afferent receptors (c-fibers and rapidly adapting receptors) with axons in the recurrent laryngeal nerve (Kalia & Mesulam, 1980). Our results thus support the hypothesis that tracheal receptors, in addition to the pharyngeal and esophageal receptors proposed by Jean, (1984) modulate the central pattern generator for swallow

Increased swallow EMG activity of the geniohyoid, thyrohyoid, thyropharyngeus, parasternal and post-swallow cricopharyngeus during mechanical stimulation of the trachea is evidence of increased pharyngeal clearance. This is manifested by increased swallow intensity and increased swallow occurrence during the aspiration protocol. These results indicate that mechanical stimulation of the trachea alters activity patterns of submental and pharyngeal muscles. We hypothesized this increased drive would result in increased pharyngeal clearance, because it was also accompanied by decreases in total swallow and laryngeal elevation time. Note these changes were not perpetuated by a decrease in individual muscle activation time, but a faster activation of the oral-pharyngeal-upper esophageal sphincter wave (Table 1).

This is the first report of modulation of the posterior pharyngeal constrictor, the thyropharyngeus muscle during coughing. The implications of this finding extend beyond the activity pattern of a single muscle upper airway muscle during cough or breathing. The activity of this muscle is an additional manifestation of novel coordinating mechanisms between cough and swallow. We believe that effective pharyngeal clearance is also accomplished through a complex interplay of material ejected by cough and subsequently swallowed. Figure 4 shows an example of phasic expiratory activity of the thyropharyngeus muscle. The thyropharyngeus muscle controls the diameter of the pyriform sinus, adjacent to the laryngeal vestibule, which acts as a reservoir for material within the pharynx during swallowing. Accumulated material in the pyriform sinus is similar to bolus accumulation before/during the pharyngeal phase of swallow, mucus ejected from the lower airways by coughing could pool in the pyriform sinus. We hypothesize that the pyriform sinus remains open to accommodate material ejected by coughing by a reduction in phasic activity of the thyropharyngeus muscle. During subsequent swallows contraction of the thyropharyngeus then collapses the sinus, emptying the contents into the esophagus.

We note that due to the open trachea in this preparation, the alterations in thyropharyngeus muscle EMGs were likely the result of modifications of central mechanisms driving this motoneuron pool rather than sensory feedback from material deposited in the pharynx or pharyngeal airflow. The extent, to which these motor responses of the thyropharyngeus muscle were due to inhibition from the cough pattern generator or due to activation of tracheal afferents, or both, is unknown. Cough motor drive was also increased in the cycles immediately following a swallow occurrence (Figure 2). This observation is explained by one of at least two alternate hypotheses: a) prolonged excitatory relationships exists between the swallow and cough pattern generators in addition to short-term phase restriction, and/or b) prolongation of the preceding cough E2 phase enhances synaptic drive to spinal and upper airway motoneurons. This dynamic interplay between these behaviors and may be a central motor program in anticipation of increased cough-related airflow shear forces in the pharyngeal airway following the swallow.

4.4 Meta-behavior

The results support the idea that the production of cough and swallow in response to aspiration is a “meta behavior” We observed alterations in gain of cough and swallow when the behaviors were induced in the aspiration protocol. In this context, alterations in behavioral gain are consistent with allostasis, or the maintenance of stability through change (Fibla et al., 2010). Allostasis describes a process by which organisms adjust to predictable and unpredictable events. Meta-behavioral responses, such as the coordination of cough and swallow reported in this study, provide tools by which the central nervous system achieves allostasis. Our data also support precedence and blocking as important control mechanisms in airway protection. The specific brainstem mechanisms that underlie these control features are unknown. However, they likely represent a substrate for pathological processes that result in dysphagia and dystussia.

One implication of this designation is that no single behavior is sufficient to protect the airway from aspiration. In patients that are unable to swallow, feeding occurs via a stomach tube to bypass the pharynx (Norton et al., 1996; Britton et al., 1997; Meng et al., 2000; Heffernan et al., 2004). If untreated, these patients will aspirate and are likely to acquire pneumonia (Wada et al., 2000; Kaplan et al., 2002; Cabre et al., 2010). While some of these patients might be able to cough, the presence of coughing does not alter the clinical strategy for their management. Clinical decision-making de facto discounts the sufficiency of coughing alone to prevent pneumonia. It is much more common to encounter patients who have impairment of both cough and swallow [stroke (Smith Hammond et al., 2001; Smith Hammond et al., 2009), PD (Pitts et al., 2008; Pitts et al., 2009; Pitts et al., 2010; Troche et al., 2010; Pitts et al., 2012), etc.], consistent with a linked control system for these behaviors in humans.

4.5 Limitations of the experimental design

A limitation of the experimental design was the use of sodium pentobarbital anesthesia, and its effects on respiratory motor drive. Warner and colleagues (1992), in sodium pentobarbital anesthetized dogs, demonstrated acute suppression of expiratory motor drive during breathing, but the effects were ameliorated over time in spite of constant plasma levels of this anesthetic. Warner et al (1992) concluded that the expiratory depressant effects of this anesthetic were transient. Our laboratory has shown that vigorous cough expiratory motor responses occur in cats anesthetized with sodium pentobarbital (Bolser et al., 2000; Bolser and Davenport, 2000).

An additional limitation was the use of a single stimulus modality to induce each behavior (mechanical for cough and water for swallow). It is not yet known if cough induced by chemical stimuli and/or swallow induced by mechanical/chemical stimuli (e.g., various bolus types of different size, texture, taste, etc.) would be coordinated. However, we hypothesize that these coordinating mechanisms are primarily central in nature, and not dependent on afferent modality. As such, we predict that this meta-behavioral response will be observed regardless of stimulus modality.

5.0 Conclusion

Cough and swallow are highly coordinated through defined excitatory and inhibitory central interactions. This inter-behavior control system minimizes the risk of aspiration and is consistent with the existence of a meta-behavioral control system. These operating principles provide a framework for integrating models of dysphagia and dystussia. Furthermore, increased cough and swallow excitability during simulated aspiration suggests a novel role of tracheal afferent feedback for informing this meta-behavioral control system for airway protection regarding aspiration.

Highlights.

• This manuscript describes novel mechanisms which regulate the coordination of cough and swallow specifically in response to aspiration.

• Our work demonstrates the existence of common system that is sophisticated and exerts control over these behaviors at several different levels.

• This knowledge will stimulate research aimed to understand the control of these behaviors as integrated and coupled.