Impact of Expiratory Muscle Strength Training on Voluntary Cough and Swallow Function in Parkinson Disease

Abstract

Background

Cough provides high expiratory airflows to aerosolize and remove material that cannot be adequately removed by ciliary action. Cough is particularly important for clearing foreign particles from the airway in those with dysphagia who may be at risk for penetration/aspiration (P/A). Expiratory muscle strength training (EMST) was tested to improve cough and swallow function.

Methods

Ten male participants, diagnosed with Parkinson disease (PD), with videofluorographic evidence of penetration or with evidence for aspiration of material during swallow of a thin 30-mL bolus, completed 4 weeks of an EMST program to test the hypothesis that EMST would improve cough and/or swallow function. Measured parameters from an airflow waveform produced during voluntary cough, pre-EMST and post-EMST, included inspiration phase duration, compression phase duration (CPD), expiratory phase peak flow (EPPF), expiratory phase rise time (EPRT), and cough volume acceleration (VA) [ie, the EPPF/EPRT ratio]. The swallow outcome measure was the degree of P/A during the swallow task.

Results

There was a significant decrease in the duration of the CPD and EPRT; the decrease in EPRT resulted in a significant increase in cough VA. Significant decreases in the P/A scores were found posttraining.

Conclusions

The results demonstrate that EMST is a viable treatment modality for a population of participants with PD at risk of aspiration.

Cough is a mechanism of airway clearance that adds to normal ciliary function.^{1, 2} Composed of three events, cough production contains an inspiratory effort that is followed by a rapid vocal fold adduction and a contraction of the expiratory muscles, including all abdominal muscles, with a majority of force production from the internal and external oblique muscles.^{3, 4, 5, 6, 7} The dynamic narrowing of the airways and ballistic vocal fold adduction (via contraction of the thyroarytenoid and interarytenoid muscles) allows for the production of high expiratory airflow velocity. The high airflow velocity provides the force to aerosolize material and safely remove it from the lungs.^{4, 8} These events are followed by vocal fold opening to widen the glottis, releasing the high subglottic pressure.

Current treatments that have been promoted to assist airway clearance include postural drainage,^{9, 10} manually assisted cough,^{2, 11} incentive spirometry,^{9, 12} percussion and vibration (chest clapping/shaking),^{2, 13, 14, 15} a forced expiratory technique also known as huffing, ² and an active cycle of breathing techniques.^{9, 16} Most of these training paradigms include forced expiratory maneuvers. Van Den Eeden et al ¹⁷ concluded that this forced expiratory maneuver is an effective component for airway clearance.

In those with Parkinson disease (PD), aspiration may occur during swallowing, potentially causing pneumonia,^{3, 18} a leading cause of death in the population with PD.^{19, 20, 21, 22} Given the high prevalence of morbidity and mortality due to aspiration in PD, treatments focusing on airway protection from aspiration while improving cough effectiveness are ideal.

Fontana et al ⁵ and Ebihara et al ³ reported significant decrement in cough function with PD, including decreased peak electromyogram amplitude of abdominal muscles during both reflexive and voluntary cough, and decreases in cough sensitivity necessary for activation of a reflexively induced cough. These changes represent the difficulty that the pulmonary and/or laryngeal systems may have reacting to and eventually removing foreign material from the airway. Furthermore, Pitts et al ²³ demonstrated significant differences in the voluntary induced cough patterns of participants with PD who penetrated/aspirated compared with those with PD who did not. The cough pattern changes also were related significantly to the level of penetration/aspiration (P/A) during a 30-mL sequential swallow task.

Recently, the authors examined the use of expiratory muscle strength training (EMST) as a treatment for increasing maximal expiratory pressure (Pemax) generation. Evidence of its benefits following training comes from numerous studies, including persons with PD,^{24, 25, 26, 27} the sedentary elderly, ²⁸ those with multiple sclerosis,^{29, 30} instrumentalists, ³¹ professional voice users, ³² and young healthy adults. ³³

EMST (EMST 150; Aspire Products; Gainesville, FL) uses a calibrated, one-way, spring-loaded valve to overload the expiratory muscles mechanically. The valve blocks the flow of air until a sufficient expiratory pressure is produced. Once the targeted pressure is produced, the valve opens, and air begins to flow through the device. The physiologic load on the targeted muscles can be increased or decreased depending on the device setting. When calibrated to a person's Pemax generation, the load can create a condition that results in peripheral adaptations to the muscle.^{34, 35, 36}

By using the device to facilitate the development of greater Pemax values following training, the functional ability to develop higher expiratory airflows during cough should result. Preliminary evidence supporting this hypothesis comes from Chiara et al.^{29, 30} Compared with other positive expiratory pressure devices, the EMST device may provide additional benefits.^{9, 37} To achieve the preset pressure level and open the valve on the EMST device, the users must produce an isometric muscle contraction. With most other positive expiratory devices, the respiratory load is less, and the manner of the resistance allows users simply to alter their breathing in a compensatory manner to reach the target. The authors hypothesize that EMST operates by allowing task-specific training directed to the “ballistic nature” of voluntary and reflexive cough. This training activity may provide a major advantage in populations with diseases like PD who have difficulty performing high-velocity tasks.

It was hypothesized that the voluntary cough airflow pattern in those with PD with known P/A would improve significantly from pre-EMST to post-EMST. Specifically, the authors predicted that with training, there would be a decrease in the duration of the inspiratory and compression phases along with the expiratory phase rise time (EPRT) as measured from the cough airflow waveform. Additionally, the authors hypothesized that with training, there would be an increase in the peak expiratory flow and cough volume acceleration (VA) [a measure relating to the “shearing force” potential]. Third, it was hypothesized there would be a significant decrease in the P/A score, as measured from the videofluorographic examination of the sequential swallow (3 mL) task.

Materials and Methods

Ten male participants (60 to 82 years of age) with PD were included in the study. A neurologist who specializes in movement disorders, affiliated with the University of Florida Movement Disorders Center, provided the diagnosis of PD by UK Brain Bank Criteria, and he further provided the evaluation disease stage (Hoehn and Yahr scale, 1967). Only participants with midstage PD were used with Hoehn and Yahr scores between 2 and 3 (Table 1), and these participants had demonstrated videofluorographic evidence of P/A into the laryngeal vestibule during a thin 30-mL sequential swallow task. Participants were all oriented to person, place, and time, able to follow two- to three-step directions, and scored at least a 24 on the Mini Mental State Examination. ³⁸

Table 1.

Participant Demographics and P/A Scores Pretraining and Posttraining^*

				P/A
Participant	Sex	Age, yr	H & Y	Pretraining	Posttraining
1	M	72	3	5	3
2	M	77	2.5	5	2
3	M	72	3	3	3
4	M	74	3	5	5
5	M	78	2.5	8	1
6	M	70	3	3	2
7	M	77	3	2	1
8	M	82	3	7	5
9	M	67	3	3	3
10	M	60	3	5	2

^*H & Y = Hoehn and Yahr score; M = male.

All participants reported no history of being treated for pulmonary disease, stroke, tobacco use within the last 5 years, or a diagnosis of dementia as confirmed by neuropsychological testing. Review of the participants' self-reported medical history at the start of the study determined that all were on standard medications for treatment of PD, and none were taking any medications with a potential influence on cough production, such as codeine. All videofluorographic examinations and voluntary cough productions were sampled in a medication “on” phase. Medication “on” was defined as 60 min following the ingestion of the participants' medications. The institutional review board at the University of Florida approved the study (IRB No. 154–2003).

The participants completed one baseline session, trained with the EMST 150 device (Fig 1) for 4 weeks, and then they returned for a visit 1 week following completion of training. During the 4-week training period, the participants used the device 5 days per week at home, performing five sets of 5 breaths through the device for a total of 25 breaths per day. The sets were performed sequentially and at approximately the same time each training day for 4 weeks. The trainer was set at 75% of the participant's Pemax (discussed in task 2). The participants were provided with verbal and written instructions for the task.

Figure 1.

EMST device.

Tasks

• 1.Videofluorographic examination of swallow. Participants were seated in an upright position and asked to swallow a 30-mL thin bolus (Varibar; E-Z-EM; Lake Success, NY) in a continuous manner. A qualified speech-language pathologist, blinded to the experimental condition, measured degree of P/A from the videofluorographic examination of the sequential swallow task using the P/A scale (Table 2). ³⁹ The P/A scale is a standard measure developed and used by speech pathologists for the evaluation of pharyngeal dysphagia.^{39, 40, 41, 42, 43, 44} The judge had > 20 years of experience in dysphagia. All of the videofluorographic examinations were archived using the Digital Swallow Station (model 7200; Kay Elemetrics Corp; Lincoln Park, NJ).
• 2.Pemax. As an indirect measure of expiratory muscle strength, Pemax was determined for each participant. Pemax was measured using a pressure manometer (FLUKE 713-30G; Fluke Corporation; Everett, WA) coupled to a mouthpiece by a 50 cm × 2 mm inner diameter tubing, with an air leak achieved with a 14-gauge needle. Participants were asked to stand, a nose clip was used to occlude the nose, and they were then asked to breathe in to total lung capacity and blow hard into the tubing. Three measures of Pemax were made until three values were obtained within ± 5% of one another. The average of the three values was used as the average measure of Pemax.
• 3.Voluntary cough measures. Airflow produced during voluntary cough production was sampled using an oral pneumotachograph (MLT 1000; ADInstruments, Inc; Colorado Springs, CO), connected to a spirometer (ML141, ADInstruments, Inc.). A nose clip was used to occlude nasal airflow during the cough maneuver. The airflow signal was measured and digitized at 1 KHz and displayed using appropriate software (Chart, version 5, for Windows; Microsoft Corp; Redmond, WA). Each sample was low-pass filtered at 150 Hz within the program (ie, Chart software).

Table 2.

P/A Scale Developed by Rosenbek et al ³⁹

Score	Contrast	P/A
1	Contrast does not enter the airway	No penetration
2	Contrast enters the airway; remains above the vocal folds	Penetration
3	Contrast remains above the vocal folds with visible residue	Penetration
4	Contrast contacts vocal folds; no residue	Penetration
5	Contrast contacts vocal folds; visible residue	Penetration
6	Contrast passes glottis; no subglottic residue	Aspiration
7	Contrast passes glottis; visible subglottic residue despite response	Aspiration
8	Contrast passes glottis; visible subglottic residue; absence of response	Aspiration

The instruction given to each participant included the following (1) relaxing and breathing into the pneumotachograph tube (held by the researcher); (2) following three tidal volume breaths, breathing deeply and then coughing hard; and (3) completing three trials of the voluntary cough.

The following measures were made from the cough flow waveform (Fig. 2):

• 1.Inspiration phase duration (IPD): the onset of inspiration, following tidal volume breathing, to the end of inspiration prior to the compression phase;
• 2.Compression phase duration (CPD): the time from the end of the inspiratory phase to the beginning of the expiratory phase;
• 3.Expiratory rise time (EPRT): the time from the beginning of the expiratory phase to the peak expiratory flow;
• 4.Expiratory phase peak flow (EPPF): the peak airflow during the expiratory phase of the cough; and
• 5.Cough VA: expiratory peak flow/EPRT.

Figure 2.

Examples of airflow waveforms during a voluntary cough task from group 1 and group 2. The vertical lines within the group 2 waveform denote the phases of the cough waveform as described in the text. CVA = cough VA.

Means and SDs were calculated from the three trials of the cough. Intrameasurer reliability was calculated on 100% of the data set for the assessment of P/A and 20% of the data set for the measures of the voluntary cough parameters. Pretraining and posttraining differences for P/A scores, Pemax, and the cough measures were tested using the Wilcoxon signed rank test. Significance was set at p = 0.05 for P/A score and Pemax but adjusted for the number of comparisons for the cough airflow waveform parameters using the Bonferroni correction procedure because significant relationships existed between the dependent variables (p = 0.05/5 = 0.01).

Results

Intrameasurer reliability for the P/A scores and the measurements made from the cough flow signals was assessed using intraclass correlation coefficients. The ratings were significant for reliability for the P/A scores (α = 0.56; p = 0.001) and the measurements from the cough flow signals (α = 0.70; p < 0.001).

Videofluorographic Evaluation of Swallow

P/A scores before and after training significantly decreased (Z = 2.388; p = 0.01) [Table 1].

Pemax

There was a significant increase in Pemax (Z = 2.803; p = 0.005) due to training. The mean Pemax before training was 108.2 ± 23.2, and the mean Pemax following training was 135.9 ± 37.5.

Voluntary Cough

Pretraining and posttraining differences were found for particular parameters of the cough waveform. Table 3 ¹ shows the means and SDs. There was a reduction (insignificant) in the IPD (Z = 2.090; p = 0.04). There was a significant reduction in the CPD (Z = 2.803; p = 0.005) and EPRT (Z = 2.492; p = 0.01) following the EMST training (Fig 3). Due to the decrease in EPRT, there was a significant increase in cough VA (Z = 2.497; p = 0.01) [Figure 3, Figure 4]. There was no significant training effect for IPPF (Z = 1.376; p = 0.17) or EPPF (Z = 0.459; p = 0.65).

Table 3.

Voluntary Cough Measures Pretraining and Posttraining^*

Measure	Pretraining	Posttraining
IPD	1.02 (0.30)	0.75 (0.27)
CPD	0.32 (0.03)	0.15 (0.01)
EPRT	0.10 (0.04)	0.06 (0.03)
EPPF	6.54 (0.84)	6.77 (1.09)
Cough VA	80.63 (28.3)	165.35 (87.7)

^*Values are given as the mean and (SD).

Figure 3.

Portions of the cough waveform demonstrating differences in CPD and EPRT from before training to after training in one participant.

Figure 4.

Cough VA differences from before training to after training.

Discussion

This study examined the effects of 4 weeks of EMST on voluntary cough function and the occurrence of P/A in a group of persons with PD. The overall effectiveness of the participants' voluntary cough increased, as indicated by the increase in cough VA, which relates to the ability of the cough to create shearing forces and remove unwanted material from the airway. ⁴⁵ Specifically, CPD and EPRT, as measured from the voluntary cough airflow waveform, from before training to after training, significantly decreased. The significant decrease in the EPRT led to a significant increase in the cough VA. This may be due to the nature of the training task, which includes a short duration and isometric contraction of the expiratory muscles to generate the maximum pressure to open the pressure release valve on the device. Specifically, to complete a trial with the EMST device, participants were required to achieve 75% Pemax repeatedly on the trainer by producing an expiratory maneuver forcefully. If the task was performed at slower speeds, producing decreased force generation, an inadequate amount of flow would result and the device's valve would not release. The release of the valve with air moving through the trainer is the signal that the trial was successful.

There was no significant increase in EPPF. Rather than simply examining cough magnitude, analysis of the cough pattern provides more information as to the viability of airway clearance because EPPF is highly dependent on pulmonary function ⁴⁶ and not entirely on the participant's effort or strength. ⁴⁷ Moreover, a productive cough relies on all three phases to generate the necessary pressures and an acceleration of the gases within the pulmonary system to achieve shearing forces. Pulmonary function testing revealed that 80% of the participants in this study presented with restrictive lung disease, which is well known in PD.^{48, 49} A post hoc comparison of pulmonary function before to after training revealed no significant training effect on the measures of pulmonary function FEV₁ (t = −1.115; p = 0.29) or FVC (t = 1.702; p = 0.12). Thus potential change in EPPF is limited by the restrictions of the pulmonary system regardless of the participant's effort.

The P/A scores, obtained from the videofluorographic examination of participant's 3-mL sequential swallow task, significantly decreased. P/A scores reflect a clinician's evaluation of the depth of material penetrated into the airway, and whether there is a cough or throat clearing response to assist in the removal of the material. Table 2 describes the defined criteria used for judging the degree of penetration or aspiration. For example, when P/A scores reach a value of 8, it indicates a severe degree of threat to the airway because the material has passed below the level of the vocal folds and was not removed with a reflexive cough. A decrease in the score from before to after training indicates a decrease in the severity of the material entering the laryngeal vestibule or airway, and this improvement hypothetically decreases participant risk.

Wheeler et al ⁵⁰ describe the biomechanical events that occur with EMST beyond the exercise of expiratory muscles that could be contributing to reduced P/A scores, including increased vertical elevation of the hyoid bone via increased activation of the submental muscles. Vertical elevation of the hyoid bone is important for the pharyngeal phase of swallowing, and lack of coordination or muscle weakness in the submental group results in changes in laryngeal elevation and the opening of the upper esophageal sphincter, all leading to P/A of the bolus into the airway.^{50, 51, 52} It is speculated that the decrease in the P/A scores is due in part to a strengthening of the submental muscles that are responsible for elevating the hyoid bone, and thus the larynx, which is necessary to close off the airway during the swallow.

Another possible mechanism for the change in the P/A scores after training is an increase in subglottic air pressure during the swallow. Various studies examining participants with tracheotomy tubes^{53, 54, 55} demonstrate that closure of the tube during swallow results in a general decrease in the depth of penetration and/or aspiration of the bolus compared with a condition in which the tube was open. The current results demonstrated a significant increase in the Pemax from before training to after training, and this increase in respiratory pressure generation capacity could have assisted the participants in generating higher subglottal pressures during swallow. However, the amount of pressure needed during a swallow at tidal volume is very low. Gross et al ⁵⁶ demonstrated (in one healthy participant) that a swallow at tidal volume uses approximately 2 cm H₂O. A swallow at total lung capacity requires approximately 7 to 10 cm H₂O. It may be that an increase in the magnitude of Pemax does not influence the generation of subglottal pressure needed during swallow in these participants but induces better coordination in generating subglottal pressure for the swallow task. Strength training, in general, alters the way in which motor units are recruited, the total number of motor units recruited, and the coordination of recruitment. These changes in neural activation are observed clinically not only as an improvement in force production but also in the coordination and precision of movement.^{35, 57}

In the future, it would be interesting to determine the relationship between P/A scores and timing events such as laryngeal course during swallow and how they relate to changes in voluntary cough pattern with EMST, particularly the relationship between laryngeal closure during swallow and laryngeal compression phase during cough. Wheeler et al ⁵⁸ demonstrated that those with PD who had P/A produced a significantly later maximum laryngeal closure during a single 5-mL swallow. Unfortunately, voluntary cough was not measured in this cohort of participants.

One conclusion from the results could be that the positive effects were due to a practice effect. Early researchers believed that any therapeutic interventions focusing on speech or laryngeal function with PD would fail because of the progressive nature of the disease.^{59, 60} Sarno ⁶⁰ studied > 300 PD participants attempting to develop methods to rehabilitate speech volume, facial mobility, speed of speech, and accuracy of articulation. He was able to demonstrate effects within a treatment session (which lasted 2 h). However, even with 2 h/wk for 6 weeks, he was unable to establish significant carryover effects to any of these areas. Later work done by Ramig et al ⁶¹ into PD demonstrated positive treatment effects when the participants were treated for 50 to 60 min four times a week, citing significant changes in speech including vocal loudness and articulation. Based on this prior literature, the evaluation time (1 h and then repeated 4 weeks later) does not meet the minimum time threshold to elicit a practice effect. Therefore a larger/longer research protocol would be necessary to achieve this effect.

Conclusion

This study demonstrates clear improvement in cough and swallow, as measured by P/A scores, following EMST training, and it shows it is a viable treatment option for participants with PD who are at risk for aspiration. Future studies should examine larger cohorts and also a more diverse group of participants, including those in different disease stages and of a different gender.

In future studies we aim to establish the relationship between P/A scores and timing events. These events may include the laryngeal course during a swallow and the relationship to changes in voluntary cough pattern with EMST. We are particularly interested in the relationship between laryngeal closure during swallow and laryngeal compression phase during cough. If findings from these studies can be related to delaying the morbidity and mortality from aspiration in PD and correlated with validated measures of quality of life, the EMST approach may be an appealing approach for participants with PD who exhibit evidence of aspiration.