Healthy Volunteers Immediately Adapt to Submental Stimulation During Swallowing

Abstract

Objectives:

Transcutaneous stimulation above and below the hyoid is used to assist patients with swallowing disorders (dysphagia) but has shown different effects. Previously, infrahyoid transcutaneous stimulation lowered the hyoid and larynx resisting swallowing movement while suprahyoid stimulation had no effects on hyolaryngeal movement either at rest or during swallowing. More recently AMPCARE electrodes, covering the submental region, are used for swallowing therapy in combination with resistance therapy. To gain insight into the effects of AMPCARE electrodes on movement during swallowing, we studied healthy volunteers using video-fluoroscopy. We hypothesized that AMPCARE stimulation might elevate the hyoid but not the larynx increasing vestibular opening potentially reducing swallowing safety.

Materials and Methods:

While undergoing video-fluoroscopy, seven healthy volunteers (mean age 51, 5 males) swallowed 5 ml of liquid barium on at least 10 trials randomly ordered across three conditions: stimulation at rest, swallowing without stimulation and swallowing with stimulation.

Results:

During stimulation at rest, significant (one tailed p< 0.05) anterior movement occurred in the hyoid and larynx, no superior hyoid and laryngeal movement and an increase in the distance between the hyoid and larynx. When comparing swallowing with and without submental stimulation, during stimulation volunteers significantly reduced anterior hyoid motion (p = 0.028), and increased hyoid elevation (p= 0.043) without changing anterior or superior laryngeal movement or the distance between the hyoid and larynx.

Conclusions:

The healthy volunteers immediately corrected for the effects of submental stimulation by reducing hyoid anterior motion and increasing superior hyoid motion without changing laryngeal motion to prevent increased vestibule opening with stimulation. This suggests that healthy volunteers had an internal schema for swallowing movement patterning with feedforward correction for the effects of stimulation.

INTRODUCTION

During normal swallowing, to prevent aspiration of food or liquid into the airway, several movements must be coordinated including: hyoid and laryngeal elevation, tongue propulsion, epiglottic inversion, closure of the laryngeal vestibule, vocal fold adduction, and squeezing the hypopharynx to move the bolus through the upper esophageal sphincter (1). Normally the larynx must elevate to a greater degree than the hyoid to reduce the hyoid to laryngeal distance and close the laryngeal vestibule (2). In patients with central nervous system disorders such as stroke, the timing, coordination, and extent of hyoid and laryngeal movement is often altered resulting in more frequent aspiration which can lead to pneumonia (3, 4).

Transcutaneous electrical stimulation (TES) has received attention as a treatment for swallowing disorders (dysphagia) although its effectiveness is controversial (5, 6). Initially bipolar electrodes were positioned on the skin over two locations to benefit movement during swallowing (7):a submental (suprahyoid) midline location between the mandible and the hyoid and an infrahyoid placement between the hyoid and cricoid of the larynx (7). The aim was to facilitate hyo-laryngeal elevation during swallowing (7, 8). Video-fluoroscopic studies in healthy volunteers measured hyoid and laryngeal movement induced by TES both at rest and during swallowing (9). When submental electrodes were placed horizontally over the platysma, geniohyoid, mylohyoid and anterior belly of the digastric, with the mouth closed, stimulation did not alter the position of the hyoid or larynx in healthy volunteers at rest (9). On the other hand, when electrodes were in the infrahyoid region over the platysma, the sternohyoid, omohyoid and the thyrohyoid muscles, stimulation at rest depressed both the hyoid and larynx (9). However, when suprahyoid and infrahyoid stimulation placements were combined in patients with severe chronic dysphagia, the severity of penetration and aspiration were reduced in relation to the extent of hyoid lowering during stimulation at rest (10). This suggested that patients adapted to the effects of stimulation; that is, when combined suprahyoid-infrahyoid TES resisted hyo-laryngeal elevation the patients may have used higher levels of hyo-laryngeal muscle activation to protect their airway during swallowing (10).

Others have found that submental stimulation may not benefit hyoid and laryngeal movement during swallowing. When a pair of bipolar electrodes were placed in the submental region far apart on each side in healthy volunteers (11), the effects of stimulation at rest were compared with movement that normally occurred during swallowing. The anterior motion of the hyoid and larynx induced by submental stimulation at rest was approximately half of that occurring during swallowing. However, the hyoid was elevated more than the larynx which could open the laryngeal vestibule and increase the risk of aspiration (11). Others using pairs of bipolar submental electrodes placed bilaterally compared pulse durations from 300 to 700 μs and found that the shorter pulses induced greater hyoid movement. However, when submental stimulation was applied during swallowing, the anterior hyoid motion was reduced compared with swallowing without stimulation. The authors concluded that short pulse stimulation may have penetrated deeper into submental muscles moving the hyoid more anteriorly prior to swallowing. It seemed that when submental stimulation began prior to swallowing, the extent of anterior hyoid motion during swallowing was reduced.

Electrode arrangement can alter the effects of stimulation. Two studies (11, 12) used a pair of bipolar electrodes on each side of the submental area and found significant anterior hyoid excursion with stimulation at rest. In contrast, a previous study that used horizontally placed pairs of electrodes across the two sides, failed to find movement with stimulation at rest (9).

The AMPCARE electrode was designed to stimulate the entire submental region using one pair of electrodes, each covering one side of the submental region to activate nerve endings to the geniohyoid, mylohyoid and anterior belly of the digastric muscles (https://swallowtherapy.com/). Submental stimulation at rest using the AMPCARE electrode induced only a portion of the movement occurring during swallowing: 22.5 % of anterior hyoid motion, 17.5% of hyoid superior motion and only 6 % of laryngeal superior motion (13). However, compared with unstimulated swallowing, no changes in movement occurred with continued AMPCARE stimulation during swallowing. The only difference was the hyoid position was more anterior prior to swallowing with AMPCARE stimulation (13).

Two studies examined whether repeated AMPCARE TES stimulation during swallowing results in changes in subsequent swallowing without stimulation. After 50 trials of swallowing with submental stimulation, no evidence of adaptation of hyo-laryngeal kinematics were found during swallowing after stimulation was withdrawn (13). In the other study, after 10 trials with submental stimulation were administered, and stimulation was turned off, laryngeal vestibule closure was more rapid indicating some adaptation learning (14).

We wanted to address several issues using the AMPCARE electrode. First, the AMPCARE electrode pair is placed horizontally in the submental area similar to the horizontal pairs of electrodes used by Humbert et al (9) who found no effect of submental stimulation on hyoid or laryngeal position at rest. In contrast, when two electrode pairs were placed on each side in the submental area (11, 12), considerable anterior hyoid motion occurred during stimulation at rest. Perhaps separate bipolar electrode stimulation is needed bilaterally to induce hyoid and laryngeal motion and the AMPCARE electrode might not be effective. Further, as one study found reduced anterior motion with bilateral submental stimulation during swallowing (12), we wanted to examine this using the AMPCARE electrode.

Our purpose was to examine the effects of submental TES stimulation with the AMPCARE electrode both at rest and during swallowing in healthy volunteers to determine if use of this electrode might aid or interfere with swallowing kinematics. Two hypotheses were examined: first, if submental muscle stimulation at rest, expected to elevate and pull the hyoid forwards, would alter laryngeal position, and second, if applying submental stimulation during swallowing would increase hyoid elevation but not laryngeal elevation, thus increasing the distance between hyoid and larynx and opening the vestibule during swallowing.

MATERIALS AND METHODS

Institutional Review Board approval was obtained prior to commencing the study at the National Institute of Neurological Disorders and Stroke at the National Institutes of Health (NIH) (# 99-N-0178). Healthy volunteers were recruited from the community via the NIH research volunteer program. Informed consent was obtained with all participants. Selection criteria included no cardiac, pulmonary, neurological, otolaryngological, psychiatric or speech, swallowing, and hearing problems as determined by medical history and examination by a physician. All participants were screened by a Board-certified Otolaryngologist and a licensed Speech Language Pathologist for normal swallowing, and language and speech abilities. The data collection sessions involved a certified Speech Language Pathologist in the NIH Clinical Center.

Procedures

All male participants were clean-shaven. The skin in the submental region was cleaned with alcohol and wiped with a TENS Clean-Cote Skin Wipe to increase electrode adherence (Uni-Patch Model UP220; Tyco). A pair of AMPCARE E series electrodes (AMPCARE; Restorative AQ3 Medical Inc.) was placed on the skin, one electrode per side between the hyoid bone inferiorly and the rim of the mandible superiorly overlying the region of the mylohyoid, geniohyoid, anterior digastric, and stylohyoid muscles. The two electrodes did not overlap in the midline (Figure 1). A self-adherent bandaging tape (3M Vetrap bandaging tape; St Paul, Minnesota, USA; 3 mol/l) was fitted over the electrodes to maintain good skin contact without discomfort.

Figure 1.

AMPCARE E series submental electrode placements.

Stimulation parameters were: 30 Hertz (Hz) pulse rate, 250 microseconds pulse duration, and current ranged from 0 to 30 milliamperes (mA). During initial stimulation, a ramp up of 1 second per mA was used while familiarizing the participant with stimulation. The stimulation intensity was gradually increased until the participant experienced a tingling sensation, designated as sensory only. Then, the stimulation level was gradually increased until the participant reported a tugging sensation, designated as motor level. To achieve the maximum displacement, the level was increased again until the participant indicated the stimulation was becoming slightly uncomfortable, the maximum tolerated level. This level was used to determine the effects of stimulation on displacement of the hyoid and larynx both at rest and during swallowing. For sham trials, stimulation was only at the sensory level.

During videofluoroscopy (VF) recordings, surface electromyographic (EMG) electrodes (KayPENTAX, Lincoln Park, NJ, USA) were placed on the neck lateral to the submental electrodes to record the interference from electrical stimulation to confirm when stimulation was applied during the videofluoroscopic recording. A radio-opaque 19-millimeter (mm) diameter sphere was taped to the side of the subject’s neck for converting pixels into mms. A digital Siemens fluoroscope captured a lateral view of the anterior neck from the 6^th cervical vertebra (C6) to C1 and the floor of the nasal cavity. Digital recordings at 30 full frames per second were stored for each trial.

A minimum of ten trials (Figure 2), were randomly ordered within each participant to include: three trials of submental stimulation at rest without swallowing (stimulation at rest); three trials of swallowing 5 milliliters (ml) of liquid barium with submental motor level stimulation; three trials of swallowing 5 m l liquid without stimulation (swallowing with no stimulation), and one sham stimulation swallow with sensory level stimulation.

Figure 2.

The procedures followed for each participant in the study

Several minutes occurred between each trial. Video fluoroscopy was turned off after each trial and the participant was instructed for the next trial. If swallowing was required on the next trial, the bolus was presented orally by syringe containing 5 ml thin liquid barium (Varibar viscosity of <15 centipoise International Dysphagia Diet Standardization Initiative level 0) that could be fully ingested on a single swallow. The participant was instructed to hold the barium in their mouth until after fluoroscopy was turned on. The fluoroscopy camera was then turned on. During trials with stimulation, the stimulation was applied and then the participant was cued to swallow.

Data Processing

VF recordings were transferred to Peak Motus 8.5 software (Vicon, Denver, CO, USA) for analysis. To control for bias, the investigator who marked the points was blinded to whether stimulation occurred during data analyses of the videotaped samples. Pixels were converted into mm in Peak Motus using the standardized sphere on the side of the participant’s neck. Reference points were marked on each VF frame in a recording: C2, the anterior-inferior corner of the second cervical vertebra and C4, the anterior-inferior corner of the fourth cervical vertebra (Figure 3). The superior-inferior plane (y-axis) was defined by a line from the anterior inferior point of C2 to the anterior inferior point of C4. The anterior-inferior point of C4 served as 0 for both the y axis and the anterior-posterior (x axis) planes. The x axis was at 90 degrees to the y axis and intersected at the anterior-inferior corner of C4. Two points were tracked in x and y space on each frame throughout each trial: the anterior inferior portion of the hyoid bone and the posterior superior top of the sub-glottal air column. Nine or more trials were recorded per volunteer (stimulation at rest, swallowing with and without stimulation) and included in the motion and data analysis. The sensory sham trial was not included in the final analysis.

Figure 3.

Point marked on each VF frame. C2 represents the second cervical vertebra, C4 represents the fourth cervical vertebra. The y axis was on a line connecting the anterior inferior point of C2 to C4, the x axis was at right angle to the y axis at the anterior inferior point of C4.

A file of positions of each structure on the x and y axis for each Peak Motus frame from a VF recording for a volunteer was converted into MATLAB. The initial position when the fluoroscopy was turned on was transposed to zero for each structure on the x and y axes. The x and y peak displacements of the hyoid and subglottal air column and the reduction in distance between the inferior anterior hyoid and the posterior superior subglottal air column were computed. A negative value indicated laryngeal or hyoid descent on the y-axis or posterior movement of the hyoid or larynx on the x-axis. A reduction in distance between the hyoid and subglottal air column indicated closing of the vestibule.

For statistical analyses, a mean of the peak displacements for three trials of each type (stimulation at rest, unstimulated swallows and stimulated swallows) were computed for the hyoid and subglottal air column and the change in distance from the hyoid to the top of the subglottal air column for each participant. The full data set included one mean value for each trial type per participant of total excursion for hyoid anterior-posterior, hyoid superior-inferior, subglottal air column anterior-posterior, subglottal air column superior-inferior, and the distance from the hyoid to the subglottal air column in mm.

For stimulation at rest trials, the mean displacement of the hyoid and subglottal air column on the x and y axes were computed for each participant. Wilcoxon signed-rank tests were computed to determine if the group showed a similar change in position with submental stimulation. For stimulation at rest, one-tailed tests at a probability of 0.05 were computed to determine if a similar change from the initial baseline position occurred in the group.

To examine whether stimulation levels used across participants were related to the amount of total displacement occurring during stimulation at rest, Spearman Rank correlation coefficients were computed between the stimulation amplitude in mA and the total displacement. This was examined for each structure on the two axes.

To contrast displacements of the hyoid and superior subglottal air columns during swallowing with and without submental stimulation, within subject comparisons were conducted between mean total displacement on each axis using the Wilcoxon signed-rank tests. To examine for significant within group changes, two-tailed tests with type I error probability of 0.05 were used. All statistical analyses were conducted using Systat 13 (Systat Software, Inc., San Jose, CA 95131 USA).

RESULTS

Seven healthy (5 male) participants with mean age 51 ± 5.6 years were included for study. The total number of the recorded trials was 82 over three conditions; all participants received at least ten trials distributed among four conditions: at least three stimulation trials at rest, three swallows without stimulation, three swallows with stimulation and one sham sensory swallow. The level of stimulation varied from 40 to 65 mA with a mean of 55.7 mA for the group. No adverse events occurred.

Hyolaryngeal displacement During Submental Stimulation at Rest

The rest position was designated as zero and compared with the peak position of the hyoid and subglottal air column (larynx) on the anterior-posterior (x axis) and the superior-posterior (y axis) during stimulation within each participant (Figure 4A–D). Similarly, the distance between the hyoid and subglottal air column (larynx) was also computed as zero for the initial rest position for comparison with the maximal change (positive or negative) with stimulation within each participant (Figure 4E).

Figure 4.

Line graphs showing the individual participants measures of position at rest and maximal change in position during submental stimulation at rest for A. anterior hyoid, B. superior hyoid, C. anterior laryngeal D. superior laryngeal movement and E. the distance between the hyoid and larynx in millimeters. Values were positive in millimeters in the anterior direction on the x axis and in the superior direction on the y axis while increases in the distance between the hyoid and subglottal air column (larynx) are shown as positive changes.

Within participant changes were examined using Wilcoxon signed-rank tests. Z values computed were the ((Sum of signed ranks)/Square root (sum of squared ranks). For the hyoid anterior position, Wilcoxon Z= −2.028; p=0.022 with the range of anterior motion from −0.738 to 6.172 mm and a mean group change of 2.381 mm. For anterior laryngeal motion, Wilcoxon Z=−2.197 p=0.014 although the range of change in position was minimal from 0.00 to 1.562 mm with a mean of 0.154 mm, showing negligible change in position with submental stimulation. No significant changes occurred in the superior position of the hyoid with stimulation (Z= −1.014; p=0.15) and the superior laryngeal position was unchanged (Z= −0.269, p= 0.433). The distance between the hyoid and larynx, increased by a mean of 1.43 mm indicating vestibule opening, (Z=−1.690; p=0.045) for the group during stimulation at rest.

To determine how displacement with stimulation at rest compared with movement that occurred during liquid swallowing, motion during submental stimulation at rest was computed as a percent of motion during normal swallowing (Figure 5).

Figure 5.

Boxplots of the distributions of the percentages of swallowing movement induced by stimulation at rest. The AntHy designates anterior movement of the hyoid, SupHy designates the superior movement of the hyoid. AntLx designates anterior movement of the subglottal air column, SupLx designates the percentage superior movement of the subglottal air column and Hyoid-Laryn designates the change in distance between the hyoid to the subglottal air column.

On the x axis, the group mean anterior hyoid movement was 17.27 percent of that occurring during swallowing and the reduction in distance between the hyoid and larynx was 15.547 percent of the reduction in the distance that occurred during swallowing. None of the other position changes with stimulation at rest were 10 percent or greater when compared to the distance changes that occurred during swallowing without stimulation.

To examine whether displacement on each axis during stimulation at rest was related to stimulation level in mA across participants, Spearman Rank Rs were computed. None of the R values were greater than 0.4 demonstrating that the level of stimulation in mA tolerated by subjects did not predict the amount of displacement induced during stimulation at rest for either the hyoid or larynx or the change in distance between the two.

Comparisons Between Hyolaryngeal Displacement During Swallowing with and Without Submental Stimulation

Within participant differences between movements during swallowing with and without stimulation (Figure 6), were examined using Wilcoxon signed-rank tests (2 tailed probability of p<0.05) to compare position change between swallowing with and without submental stimulation. Anterior hyoid motion decreased from a mean of 13.697 mm during swallowing to a mean of 11.035 during swallowing with stimulation, Z= −2.197 p= 0.028. On the other hand, superior hyoid motion increased with stimulation during swallowing (Z= 2.028, p=0.043) from a mean elevation of 13.26 mm to 15.08 mm during swallowing with submental stimulation. No change in movement with stimulation during swallowing occurred in laryngeal anterior motion (Z=0.676, p=.499), or in superior laryngeal motion (Z=1.69, p= 0.091). No differences were found with stimulation during swallowing in the distance from the hyoid to the larynx (Z=−.507, p=0.612) with a mean reduction of 10.57 mm during swallowing and 10.93 mm during swallowing with stimulation.

Figure 6.

Line graphs showing the individual participant’s measures of maximum change in position during swallowing and during swallowing with submental stimulation for A. anterior hyoid motion, B. superior hyoid motion, C. anterior laryngeal motion, D. superior laryngeal movement and E. change in distance between the hyoid and larynx in millimeters, Positive in millimeters is the anterior direction on the ap axis and in the superior direction on the si axis while decreases in the distance between the hyoid and the subglottal air column (larynx) are shown as negative changes

To examine whether changes in swallowing movement with stimulation occurred immediately, we examined anterior motion of the hyoid on the first swallow trial with stimulation trial in each participant along with a non-stimulation trial that was either just before or after that first stimulation trial (Figure 7). During testing, the bolus was administered to participants and they were told to hold it in their mouth while the fluoroscopy was turned on for the swallowing trial either with or without stimulation. On the non-stimulated swallows, after the fluoroscopy was turned on, the participant was cued to swallow and completed their swallow between 1 and 3 seconds after recording had started. On the stimulated swallows, after the fluoroscopy was turned on, stimulation was turned on for 3 – 4 seconds and then the participant was cued to swallow and completed their swallow within 1 and 3 seconds after cueing which was usually 6 to 8 seconds after recording had started. We examined whether a participant immediately adapted to the presence of stimulation which moved the hyoid anteriorly before they started to swallow. If no adaptation occurred, the anterior hyoid movement would include both movement induced by stimulation in addition to the same anterior motion as occurred during a non-stimulated swallow. Thus, the total anterior excursion during a stimulated swallow would be greater than non-stimulated swallows if participants did not alter their swallowing movement with stimulation.

Figure 7.

Display of hyoid anterior position change over time in a non-stimulated swallow (ns) and the first stimulated swallow trial (s) for each participant. The x axis is the time in seconds from the onset of fluoroscopy for both the non-stimulation trial (red) and the stimulated trial (blue). The y axis is the anterior hyoid excursion during swallowing. The horizontal dashed red line is the hyoid position when the fluoroscopy was turned on and the horizontal dashed blue line is the hyoid position with stimulation on before being cued to swallow. The vertical red hatched line shows the extent of anterior hyoid motion during non-stimulated swallow (ns) and the vertical blue hatched line (s) is the extent of anterior motion of the anterior motion of the hyoid during a stimulated swallow. The black arrow shows the expected anterior hyoid motion when the stimulation was on if the participant produced the same hyoid anterior movement for a swallow as they did on on a non-stimulation swallow.

In Figure 7, the starting anterior position of the hyoid is at zero for both the stimulated (s) and non-stimulated (ns) trials. For each participant, the anterior motion of the hyoid during a non-stimulated swallow is the red tracing of anterior hyoid motion labelled “ns”. The vertical red hatched line from the baseline shows the extent of the anterior hyoid excursion during a non-stimulated swallow. For the first stimulated swallow for each participant, a blue line labelled “s”, starts after fluoroscopy was turned on, shows if the hyoid position shifted anteriorly when stimulation was turned on. The hyoid anterior motion as a result of stimulation is shown by the horizontal hatched blue line. Then while the stimulation remained on, the participant was cued to swallow and the blue tracing shows the trajectory of the anterior hyoid motion from the stimulated position. The vertical hatched blue line shows the total anterior displacement of the hyoid that occurred from the original position to after stimulation and swallowing.

The total extent of anterior hyoid motion during the initial stimulated swallow trial in the 7 participants did not equal the sum of their change in anterior position with stimulation plus the change in anterior hyoid motion that occurred on a swallow without stimulation. If this had occurred, the total anterior hyoid excursion that should have occurred is shown by the black arrow in each panel for participants 180, 181, 183, 184, and 188. Each of these participants had less anterior hyoid motion during the stimulated swallow than would be expected if the same hyoid anterior motion had occurred on the stimulated swallow as occurred on the non-stimulated swallow and was added to the effect of the stimulation being turned on before swallowing. The other two participants, 186 and 187, did not have appreciable change in anterior hyoid position with stimulation and did not alter their degree of hyoid anterior motion during the stimulated swallow from that which occurred on the non-stimulated swallow. Thus, immediate adaptation was evident as participants reduced the extent of hyoid anterior movement for swallowing when simulation first moved the hyoid anteriorly before the participant began to swallow.

DISCUSSION

We examined two hypotheses in this study. The first was if submental muscle stimulation at rest, expected to elevate and pull the hyoid forwards, would alter laryngeal position. During stimulation at rest anterior motion of the hyoid averaged 2.4 mm while hyoid elevation was limited to 0.4 mm with stimulation at rest. The changes in position for the larynx with submental stimulation at rest were limited; mean anterior movement was 0.5 mm, while the mean laryngeal elevation was 0.154 mm. Thus, laryngeal elevation was not induced by hyoid motion during submental stimulation at rest.

The second hypothesis was whether applying submental stimulation during swallowing would increase hyoid elevation but not laryngeal elevation thus increasing the distance between the hyoid and larynx and opening the vestibule during swallowing. Contrary to this expectation, we found that the amount of hyoid anterior motion for swallowing during stimulation was reduced in stimulated swallows to correct for anterior motion induced by stimulation alone. As a result, the anterior hyoid motion during stimulated swallows was reduced to adjust for the increase induced by stimulation immediately prior to the swallow (Figure 7). Thus, the distance between the hyoid and larynx was not increased during stimulated swallows as was expected as no increase in hyoid anterior motion occurred during swallowing with stimulation compared to swallowing without stimulation (Figure 6E). In fact, the participants reduced hyoid elevation during swallowing with stimulation to correct for the anterior motion induced by stimulation alone and as a result did not increase the distance between the hyoid and larynx with stimulation during swallowing.

The mean amount of anterior hyoid motion induced by submental stimulation at rest using AMPCARE electrodes was 2.38 mm which was 17.28% of hyoid anterior motion during swallowing without stimulation. In this and other studies using submental stimulation the maximal tolerable level of stimulation was used. Our result is similar to a previous report of submental stimulation at rest using AMPCARE electrodes (13), which found submental stimulation at rest induced anterior hyoid motion that was 22.5% of anterior movement during swallowing. Both studies found hyoid superior movement induced with stimulation at rest was less than anterior hyoid movement and that laryngeal superior movement induced with stimulation at rest was minimal. The AMPCARE electrode induces stimulation horizontally across the submental region, similar to the bipolar submental arrangement used by Humbert at al 2006 (9) who found no movement induced by two bipolar pairs stimulating horizontally. Thus, the AMPCARE electrode, which stimulates the entire submental area, were superior in inducing movement at rest to bipolar electrodes arranged horizontally in the submental region.

Two other studies (11, 12) used bipolar electrode pairs in the submental region separately on the right and left sides. When a larger submental area was stimulated on each side (11) hyoid anterior movement averaged 5.74 mm, hyoid elevation was 10.5 mm, laryngeal anterior excursion was 1.9 and superior laryngeal movement averaged 9.6 mm. In the more recent study (12), with a smaller area stimulated, hyoid anterior movement was 1.94 mm, hyoid elevation was 1.81 mm, laryngeal anterior movement was 1.48 and laryngeal superior movement averaged 0.82 mm. These differences suggest that when a larger area is covered by stimulation either by bipolar horizontal stimulation or bipolar unilateral stimulation, more movement is induced.

Based on these previous results, it could be expected that if submental stimulation were added to hyoid anterior movement during swallowing, displacement would increase by 17 to 22 percent. We compared hyoid anterior displacement during swallowing with and without submental stimulation and found the opposite, that the extent of anterior hyoid movement decreased during swallowing with submental stimulation in comparison with movement during swallowing without submental stimulation (Figure 6A). The Serel Arslan study observed that when submental stimulation was applied before swallowing onset, the hyoid was moved more anteriorly before the participants started to swallow (13). Similarly, we found in 5 of 7 participants that stimulation raised the hyoid anteriorly off the baseline level prior to the swallow (Figure 7).

The Serel Arslan study was aimed at determining if error-based learning occurred over 50 trials of swallowing in healthy volunteers with either intermittent or continuous stimulation and did not find any changes in measures over 50 trials of stimulation during swallowing. As our study showed that subjects immediately adapted to the effects to stimulation on the first trial of swallowing with stimulation this likely explains why no changes were seen over 50 trials of swallowing with stimulation by Serel Arslan. As shown here, when submental stimulation was turned on and moved the hyoid anteriorly before healthy volunteers were asked to swallow, they immediately reduced their anterior hyoid motion during the subsequent swallow. The mean amount of anterior hyoid motion induced by stimulation at rest in this study was 2.4 mm, while the mean amount anterior hyoid motion during swallowing without stimulation was 13.7 mm. This anterior motion was reduced during swallowing with stimulation to 11.04 mm, a reduction of 2.6 mm. Thus, the mean reduction from swallowing anterior motion without stimulation to that with stimulation was similar to the anterior motion induced by submental stimulation at rest ~ 2.4 mm. As this reduction occurred on the first swallow during stimulation the adaptation to correct for anterior hyoid movement was immediate. This suggests that in the Serel Arslan study, their healthy volunteers were able to adapt their swallowing to changes in hyoid position prior to swallowing immediately on their first swallow with stimulation, similar to our results.

We only found reduced hyoid anterior movement for swallows when stimulation prior to the swallow induced anterior hyoid movement. In two swallows where no appreciable hyoid anterior movement occurred with stimulation in participants 186 and 187, no change in the extent of hyoid anterior movement occurred between the stimulated and non-stimulated swallows (Figure 7). As the effects depended upon the extent the hyoid anterior motion induced by stimulation before the swallow, this demonstrated an automatic feedforward effect on swallowing movement.

We did not find that the current stimulation amplitude, based on each subject’s maximum tolerable levels, was related to movement amplitude induced at rest. However, two previous reports found an inverse relationship between submental stimulation amplitude and movement across tasks; one study reported (r=−0.38) (12) and the other related stimulus amplitude to lingual-palatal pressure (r=−0.24) (15). Perhaps these inverse relationships illustrate adaptation to the effects of higher stimulation levels such that the healthy participants reduced their movement when stimulation levels were higher.

An internal schema for swallowing movements could provide an automatic feedforward adaptation in healthy persons as observed by Wong et al., (2), who found volunteers altered laryngeal elevation magnitude to exceed hyoid elevation based on hyo-laryngeal length before swallowing. Further when their head position was altered before swallowing, individuals immediately adapted their hyo-laryngeal movement during swallowing based on targets required for closing the hyo-laryngeal area for safe swallowing from the new starting position. Our results agree with the concept of an internal schema producing an automatic adaptation of swallowing movements required for safe swallowing. If the participants had not reduced their extent of anterior hyoid motion during swallowing with submental stimulation, the distance from the hyoid to the larynx would have increased, which did not occur (Figure 6 E.). Although we only used three trials each for both stimulation and non-stimulated swallows and the trials of different conditions were randomly ordered, the healthy subjects immediately reduced their hyoid anterior motion during the first stimulated swallow trial showing feedforward adaptation to the stimulation induced anterior hyoid motion. By reducing anterior hyoid motion, the participants prevented an increase in distance between the hyoid and larynx which could have opened the laryngeal vestibule during stimulated swallows putting them at risk of aspiration into the airway.

It is unclear whether patients with dysphagia would similarly adapt to the effects of submental stimulation on hyoid position prior to swallowing. In a recent study of patients with severe long standing (chronic) dysphagia either due to stroke or head and neck cancer, the patients did not adapt their hyolaryngeal movements during swallowing to their initial hyolaryngeal space (16). Further, as the patients’ measures of hyoid or laryngeal peak velocity timing were not synchronized with vestibule closure, it was concluded that the central schema for swallowing patterning seemed disturbed. If the neural mechanisms involved in feedforward adaptation were impaired in patients, then using submental stimulation to increase anterior motion of the hyoid without laryngeal motion, might not be corrected for by patients during swallowing allowing the distance between the hyoid and larynx to increase placing them at increased risk of aspiration.

On the other hand, when submental and infrahyoid stimulation were combined which lower the hyoid and larynx during swallowing of patients with severe chronic dysphagia (10), patients with the greatest hyoid lowering induced by stimulation had reduced aspiration and penetration during swallowing with stimulation. Perhaps the patients with the greatest hyoid lowering by stimulation during swallowing compensated by increasing their hyo-laryngeal elevation to reduce penetration and aspiration on swallows during stimulation. Thus, TES induced stimulation applied during swallowing likely interacted with patients’ swallowing movements. Some patients may be able to adapt their swallowing to reduce risk of aspiration during stimulation while others may not be able to adapt when stimulation is applied. To further complicate this, TES over the submental and infrahyoid regions have different effects on hyoid and laryngeal movements and likely interact differently with swallowing movement patterning in patients.

In conclusion, when considering the effects of transcutaneous stimulation during swallowing, it cannot be assumed that the effects of TES on hyo-laryngeal movement will be simply additive to swallowing movements. Rather, evidence here suggests that healthy participants adapt their movements to accommodate for the effects of stimulation. In a previous study, some patients may have adapted to overcome hyo-laryngeal lowering with stimulation by increasing their elevation (10). However, not all patients may be able to adapt their movement to the effects of TES as they may not have retained an internal schema to adapt their swallowing movement as shown in one study (16). To date it has been assumed that the use of TES augments movement during swallowing and will have the same effects as found using TES at rest. As TES of the muscles affecting hyo-laryngeal motion, may not benefit swallowing movement, kinematic studies of hyo-laryngeal movements during swallowing are needed to determine how patients adapt to the effects of TES on swallowing before using TES for intervention.

This was a small feasibility study aimed at examining the effects of submental stimulation on swallowing in healthy volunteers. A major limitation is that it included only seven participants. The results were unexpected in that the participants adapted their swallowing to the submental stimulation by reducing their hyoid anterior motion to a similar degree as submental muscle stimulation enhanced hyoid anterior position prior to swallow. Clearly, further study is needed to address immediate feedforward adaptation to the effects of submental stimulation on swallowing kinematics in healthy volunteers and in patients with dysphagia. In addition, an examination of different bolus volumes and consistencies are needed to determine if these modify adaptation to TES during swallowing.