Abstract
Objective
To measure the latency of laryngeal adductor reflex (LAR) motion onset at 2 laryngopharyngeal subsites using calibrated aesthesiometers.
Study Design
Cross-sectional.
Setting
Academic institution.
Methods
Twenty-one asymptomatic, healthy subjects (11 male, 10 female) underwent laryngopharyngeal sensory testing with tactile stimuli delivered to the aryepiglottic fold and medial pyriform sinus using 30-mm Cheung-Bearelly monofilaments (4-0 and 5-0 nylon sutures) via channeled flexible laryngoscope. The LAR onset latency, defined as the first visual detection of ipsilateral vocal fold adduction following tactile stimulation, was measured with frame-by-frame analysis of video recordings.
Results
The overall mean LAR latency across both subsites and stimulation forces was 176.6 (95% CI, 170.3-183.0) ms, without significant difference between subsites or forces. The critical value for LAR response latency prolongation at the .01 significance level was 244 ms. At 30 frames/s video capture resolution, LAR response latency ≥8 frame intervals would indicate abnormal prolongation.
Conclusion
Aesthesiometer-triggered LAR latency appears to be invariant over an 8.7-dB force range and between the aryepiglottic fold and medial pyriform sinus subsites in controls. Laryngeal adductor reflex latency incongruences between stimulation forces or laryngopharyngeal subsites may serve as pathophysiological features to dissect mechanisms of upper aerodigestive tract disorders.
Level of Evidence
Level 3B.
The laryngeal adductor reflex (LAR), rapid bilateral vocal fold closure in response laryngopharyngeal stimulation, is an important brainstem-mediated airway-protective mechanism. 1 During swallowing, the vocal folds adduct in coordination with epiglottic inversion and laryngeal vestibule closure to protect the airway from food and salivary contamination. Laryngopharyngeal sensory impairment severity has traditionally been determined by the sensory stimulus amplitude required to trigger the LAR response within a go/no-go evaluative framework.2-4 This dichotomized classifier approach has been a useful clinical tool to inform treatment for a variety of upper aerodigestive tract disorders, including aspiration in adults5,6 and children, 7 infant apnea,8,9 and dysphagia in the aging,10,11 poststroke,12,13 and amyotrophic lateral sclerosis patient populations.14,15 While this single feature clinical tool is simple and deployable, it may fail to detect other clinically significant aspects of laryngopharyngeal dysfunction. For example, temporal analysis of laryngeal closure events has shown delay in motoric adductor initiation is a significant predictor of aspiration in poststroke patients.16,17 A more comprehensive assessment of LAR integrity should move beyond simply determining the presence or absence of a triggered reflex. Timing measurements of observable triggered LAR motoric events may be useful in evaluating laryngopharyngeal sensorimotor function.
Stimulation of sensory receptors in the laryngopharynx generates afferent signals via the internal branch of the superior laryngeal nerve (iSLN). This signal is sent to the interstitial subnucleus of the nucleus tractus solitarius and subsequently relayed to the nucleus ambiguous bilaterally. In turn, efferent signals are carried by both recurrent laryngeal nerves to intrinsic laryngeal adductor muscles for coordinated contraction to effect endolaryngeal closure.18,19 The bilateral actions of thyroarytenoid (TA) and lateral cricoarytenoid (LCA) muscles (TA-LCA complex) and contraction of the interarytenoid (IA) muscle bring the vocal folds together. In laryngeal electromyographic (LEMG) studies of the TA muscle, 2 distinct responses to electrical stimulation of iSLN have been identified: an early, brief R1 component and a later, bilateral, prolonged, and centrally modulated R2 component for completion of vocal fold adduction.4,20,21
An alternative to iSLN electrical stimulation for studying triggered LAR dynamics is the use of laryngopharyngeal tactile stimulation in awake human participants. Cheung-Bearelly buckling force aesthesiometers have been used in the oropharynx to measure sensory thresholds and perceptual strength, as well as adapted for use in the laryngopharynx to measure LAR response thresholds.22-24 The Cheung-Bearelly monofilament is passed through a channeled laryngoscope, and a calibrated stimulus force is applied at distinct laryngopharyngeal subsites to elicit the LAR. The rate of triggered LAR response increases monotonically with greater tactile stimulus force delivery to the aryepiglottic fold, medial pyriform sinus, and lateral pyriform sinus subsites of the laryngopharynx in healthy, asymptomatic adults. 24
The main goal of this study is to characterize a novel measure of LAR response dynamics beyond the traditional dichotomized triggering threshold. Latency of the triggered LAR, defined as the time interval from mucosal stimulation to onset of ipsilateral vocal fold adduction, interrogates an early component of integrative laryngopharyngeal sensorimotor function. We report LAR response latency profiles from tactile stimulation of the aryepiglottic folds and medial pyriform sinuses in a cohort of healthy subjects using Cheung-Bearelly buckling force aesthesiometers.
Methods
Study Design
This cross-sectional study was approved by the University of California, San Francisco Committee on Human Research (19-27065). Written informed consent was obtained prior to subject enrollment. The Cheung-Bearelly monofilaments 24 used for this study to perform tactile stimulation of laryngopharyngeal subsites received a nonsignificant risk determination from the Food and Drug Administration.
Participants
Healthy adult subjects of both sexes aged between 18 and 60 years and without upper aerodigestive tract symptoms were recruited and enrolled. 24 Subjects completed surveys that queried throat symptoms (Reflux Symptom Index [RSI]), 25 voice and swallowing problems (Voice Handicap Index–10 [VHI-10], 26 Eating Assessment Tool–10 [EAT-10]), 27 globus sensation (Glasgow Edinburgh Throat Scale [GETS]), 28 stress and mood (Perceived Stress Scale–10 [PSS-10], 29 Patient Health Questionnaire–9 [PHQ-9], 30 Generalized Anxiety Disorder–7 [GAD-7]), 31 and general health (12-item Short Form [SF-12 physical and mental]). 32
Twenty-one asymptomatic subjects (11 male, 10 female) contributed to the omnibus analyzable data set. Spatial-focused analyses for triggered LAR response rates and perceptual strength gradients were recently reported. 24 Temporal-focused analyses to examine LAR response latency profiles using this data set were performed for this investigation. The mean ages and male and female subjects were well matched at ~36 years (P = .75). All subjects displayed normal scores in each of the 9 survey instruments. Demographic data were consistent with a cohort of normal control subjects ( Table 1 ).
Cohort Demographic Data. a
All values are presented as mean (SD). All survey scores are within normal ranges of instruments.
Laryngopharyngeal Tactile Stimulation
Laryngopharyngeal aesthesiometer-based stimulation procedure and video data capture (nCare v10, 30 frames/s, MPEG-4 format; Image Stream Medical) details have been published elsewhere. 24 The test target laryngopharyngeal subsites were the aryepiglottic fold and medial pyriform sinus bilaterally, and 4-0 and 5-0 monofilaments delivered 0.30-g and 0.11-g stimulation forces, respectively, which differed by 8.7 dB. 24
Data Analysis
We analyzed stimulation events where the LAR with full glottic closure was triggered by monofilament tactile stimulation to extract timing of first and last aesthesiometer contact with the test target and first movement of the ipsilateral vocal fold toward adduction. The time stamps were marked using frame-by-frame analysis of video recordings (QuickTime 7, 30 frames/s with interframe interval ~33.3 ms; Apple, Inc). Stimulus duration was defined by the number of intervals from the first to last frames of monofilament contact with the test target. Motoric event duration of vocal fold adduction, complete glottic closure, and vocal fold abduction was defined by a similar frame-by-frame analysis procedure and in accordance with the LAR motoric sequence intervals of Shock et al. 15 LAR response latency was defined by the number of intervals from first frame of monofilament contact with the test target to the first frame of ipsilateral vocal fold adduction motion.
Dispersion about the difference of timing values for 2 events was computed by treating event occurrence and event detection as independent random variables and computing total variance by adding contributions from the sample population and measurement error. While the former was derived from experimental data, the latter was estimated by assuming measurement error was distributed uniformly from 0 to 33.3 ms. Extraction of timing values required subtraction of 2 marked frames, which cancelled associated random errors that preceded each marked frame, but variance associated with each preceding frame was added to compute the measurement error standard deviation of 13.6 ms. 33
LAR responses to 4-0 and 5-0 monofilament stimulation of the aryepiglottic folds and medial pyriform sinuses constituted the data set (n = 84). In 22 cases where the test subsite was stimulated twice, the average stimulus duration and average LAR response latency were used to prevent double counting. Four LAR onset times were 3 SDs outside the population mean and thus rejected from further analysis. Those data originated from 2 different subjects, in which each subject contributed one 4-0 and one 5-0 monofilament stimulation event. All outlier latencies were 266.9 ms or 8 frame intervals (see Results). The final data set (n = 80) included the following: (1) 4-0 monofilament, aryepiglottic fold (left, n = 12; right, n = 10); (2) 4-0 monofilament, medial pyriform sinus (left, n = 16; right, n = 11); (3) 5-0 monofilament, aryepiglottic fold (left, n = 9; right, n = 7); and (4) 5-0 monofilament, medial pyriform sinus (left, n = 10; right, n = 5). The lower number of 5-0 triggered LAR responses is due to the lower force delivered to the 2 laryngopharyngeal subsites. 24
The primary outcome measure was LAR response latency. Population data were transformed to a z-score standard normal distribution to determine the LAR response latency critical value for abnormal prolongation. The critical value was chosen at the .01 significance level. Pairwise comparisons were performed for stimulation force, laryngopharyngeal subsite, sex (<35 vs ≥35 years), age, and laterality. Descriptive statistics of centrality and dispersion were applied to the data. The unpaired and paired t tests and the χ2 test were used to assess statistical significance of differences between groups.
The secondary outcome measures were duration of vocal fold adduction, complete glottic closure, and vocal fold abduction to characterize the entire timing sequence of aesthesiometer-triggered LAR motoric events. Only cases with uninterrupted visualization of both vocal folds from initial tactile stimulation to the return of complete abduction were included. Whiteout, partial laryngeal visualization, saliva obscuration of the vocal folds, and cough with excessive motion were the main disqualifying factors. Thirteen subjects (8 male, 5 female) with suitable video capture quality enabled this subpopulation analysis (n = 24) and included the following: (1) 4-0 monofilament, aryepiglottic fold (n = 6); (2) 4-0 monofilament, medial pyriform sinus (n = 10); (3) 5-0 monofilament, aryepiglottic fold (n = 6); and (4) 5-0 monofilament, medial pyriform sinus (n = 2). All data were pooled and displayed using descriptive statistics.
Results
The overall mean LAR response latency of ipsilateral vocal fold motion was 176.6 (95% CI, 170.3-183.0) ms, corresponding to 5.3 video frame intervals. The triggering tactile stimuli were delivered to the aryepiglottic folds and medial pyriform sinuses using 4-0 (0.30-g) and 5-0 (0.11-g) monofilaments. In all cases, the minimum stimulus duration extended beyond LAR response latency. Mean stimulus durations for 4-0 monofilament (625.8 ms; 95% CI, 580.9-670.7 ms) and 5-0 monofilament (655.0 ms; 95% CI, 609.8-700.1 ms) were indistinguishable (P = .37). Variations in LAR response latency contingent on stimulation force and laryngopharyngeal subsite were statistically insignificant (
Time to Laryngeal Adductor Reflex Motion Following Tactile Stimulation. a
Time is in milliseconds.
Stimulation force of the 4-0 monofilament is 8.7 dB higher than the 5-0 monofilament.
Latency comparison by stimulation force (n = 49, 4-0 monofilament; n = 31, 5-0 monofilament) for both subsites together, P = .61.
Stimulus duration comparison by stimulation force for both subsites together, P = .37.
Minimum stimulus duration exceeds maximum vocal fold adduction onset latency in all cases.
Population LAR response latency data were transformed to a z-score standard normal distribution to determine the critical value for LAR latency prolongation. Choosing z = 2.33 as the cutoff value that corresponded to 1% of the population at the distribution tail, the latency critical value was 243.8 ms or 7.3 frame intervals. At 30 frames/s video resolution, the conservative critical value for LAR response latency prolongation would be 8 video frame intervals.
Subpopulation mean motoric event duration for the aesthesiometer-triggered LAR sequence was (1) vocal fold adduction, 139.0 ms (95% CI, 125.1-153.0 ms); (2) complete glottic closure, 240.5 ms (95% CI, 209.8-271.2 ms); and (3) vocal fold abduction, 237.7 ms (95% CI, 215.7-259.7 ms). Those data and LAR response latency data, as well as published human data by Shock et al, 15 were plotted in Figure 1 using standard error of the mean to facilitate direct comparisons between aesthesiometer- and air puff–triggered motoric event durations.
Discussion
Latency of the triggered LAR provides an additional measure of laryngopharyngeal function beyond response threshold and perceptual strength. It can be extracted from video analysis of vocal fold motion following delivery of calibrated buckling force stimuli to specific subsites through a channeled laryngoscope. 24 The combined mean value of LAR latency for the aryepiglottic fold and medial pyriform sinus is ~177 ms in asymptomatic, healthy adult controls and appears to be invariant over an 8.7-dB force range. Abnormally prolonged LAR latency at either laryngopharyngeal subsite with 30-mm 4-0 or 5-0 monofilament tactile stimulation is ~244 ms or 7.3 frame intervals. With this finding, we have chosen ≥8 frame intervals to indicate abnormality.
The motoric event duration of vocal fold adduction, complete glottic closure, and vocal fold abduction for aesthesiometer and air puff triggering of the LAR response appeared to be comparable ( Figure 2 ), suggesting posttrigger motoric sequence dynamics is not appreciably dependent on the exact type of tactile stimulation.
Time intervals of triggered laryngeal adductor reflex events. Stimulation at 0 ms. (A) Aesthesiometer in position (t–1). (B) Adduction initiation (t1). (C) Adduction completion (t2). (D) Glottic closure (t3). (E) Abduction completion (t4).
LAR response latency is a measure of integrative sensorimotor function. Significant deviations from the mean or the expectation value may arise from anatomical or functional alterations to peripheral sensors, brainstem circuit elements, central nervous system modulators, and peripheral effectors. This is the first study to report LAR latency by visual detection of ipsilateral vocal fold motion for enhanced evaluation of laryngopharyngeal function.
Electrical stimulation-triggered LAR myogenic signals measured by laryngeal electromyography (LEMG) often differentiate between the R1 unilateral and R2 bilateral response components. Reported latencies of R1 and R2 have been variable, due in part to methodological differences in laryngopharyngeal sensory stimulus delivery (electrical stimulus, air puff, test target) and the LAR response signal capture (needle configuration, hook wire, surface electrode). Electrical stimulation of the iSLN coupled with LEMG recording has been a common technique to measure triggered LAR latencies of the TA-LCA complex and other intrinsic muscles. iSLN stimulation has been performed by inserting bipolar needle electrodes or hooked wires along the inferior edge of the hyoid bone and superior edge of the thyroid cartilage, at the approximate entry location of nerve entry through the thyrohyoid membrane.18,21,34,35 Direct stimulation by bipolar electrodes 36 and transcutaneous stimulation by surface electrodes 37 have also been used, but they are not as widely practiced. Based on a multitude of iSLN electrical stimulation studies in adults, the LEMG R1 response in the ipsilateral TA muscle occurs between 16 and 20 ms,1,18,21,34,35 whereas the bilateral R2 response occurs between 60 and 75 ms.21,34,35 Direct electrical stimulation of laryngeal mucosa is yet another method. Here, a unipolar wire electrode is passed through the working channel of a flexible laryngoscope to stimulate the arytenoid, aryepiglottic fold or ventricular fold. The elicited LEMG response latency of R1 is 16 ms and R2 is 50 ms.38,39
Tactile stimulation of laryngopharyngeal mechanoreceptors to trigger LAR motoric events using pressurized air puffs, pioneered by Aviv and colleagues2,3,10-13 in the 1990s, is a stimulus delivery technique that most closely parallels use of calibrated aesthesiometers. LEMG response latency to air puff stimulation is delayed compared to electrical stimulation, ranging from 150 to 175 ms.40,41 In addition, R1 and R2 responses appear to activate concurrently, suggesting possible efferent mechanism differences between LEMG response to electrical stimulation of the iSLN and to tactile stimulation of laryngeal mucosa. 40 While LAR latency is a novel measure of sensorimotor integration, Shock et al 15 have classified other triggered LAR events (vocal fold adduction phase, glottic closure time, vocal fold adduction phase duration, and total LAR duration) in response to air puff stimulation that may be used to further dissect mechanisms of laryngopharyngeal dysfunction, such as vocal fold abduction impairment in a mouse model of amyotrophic lateral sclerosis. A recent proof-of-principle medical device study using pressurized water microdroplet delivery to arytenoid mucosa with splatter dispersion and high-speed video capture reports LAR vocal fold latency of 90 ms. 42 Our expectation value of 176 ms for triggered ipsilateral vocal fold movement latency to aesthesiometer stimulation of the aryepiglottic fold or medial pyriform sinus is consistent with the 150- to 175-ms range of LAR LEMG response latency values to air puff stimulation of the arytenoid. With the concurrence of latencies for the 2 stimulation methods using different stimulation durations (aesthesiometer, 640 ms; air puff, 50 ms), it would appear tactile-evoked LAR is a phasically driven, rapid motoric response independent of stimulus duration.
Compared to electrical stimulation, tactile stimulation using calibrated aesthesiometers more closely mimics food bolus sensory cuing that initiates swallow events, culminating in airway protection and bolus transit.43,44 Food boluses trigger pharyngeal swallowing with cognitive cuing and engage detector sensing of taste, fluid viscosity, temperature, touch, and pressure to excite several types of sensory fibers throughout the upper aerodigestive tract. 45 As a bolus enters the pharynx, laryngeal vestibule closure and epiglottic inversion divert the food stream away from the airway to the hypopharynx and esophagus, while concurrent vocal fold adduction protects the airway from aspiration.43,46-48 No study has specifically evaluated latency of vocal fold closure during swallowing, but comparison of the timing of laryngeal closure events during swallow using endoscopic and LEMG assessment tools with the timing of laryngeal closure events from air puff stimulation of the arytenoid shows convergent findings. Van Daele et al 47 reported the time interval from initiation of laryngeal adduction to glottic closure is ~250 ms in response to normal swallow. In comparison, Shock et al 15 reported the time interval from vocal fold adduction to glottic closure is ~300 ms in response to air puff stimulation.
Calibrated tactile stimulation of the laryngopharynx by either air puff or buckling force aesthesiometer mimics sensory stimulation by food boluses in a spatially precise manner. Controlled sensory stimulation techniques may be deployed to probe laryngopharyngeal sensorimotor function, where the impact of sensory changes on motoric outcomes may be measured using clinical assessments of swallowing performance. Delay between swallow onset and initiation of vocal fold adduction represents a period of high aspiration risk, where a partially open glottis may be insufficient to protect the airway. The LAR latency metric may be useful to identify upper aerodigestive tract dysfunction in certain patient populations and to dissect mechanisms of dysphagia and aspiration.6,49,50
There are several limitations to this study. First, replication studies in larger cohorts of men and women over a broader age range should be performed to understand better the effects of sex and age on laryngopharyngeal sensation and sensorimotor integration. Second, LAR response latency and motoric event duration centrality and dispersion estimates are dependent on both biological variation and measurement error distribution profile. While much less plausible, an exponential distribution profile would yield a different measurement error estimate. Replication studies enrolling a greater number of subjects and using higher-speed video capture to decrease the interframe interval may refine estimates and lower the critical value of abnormal LAR response latency prolongation. However, the need for high-speed video capture capability would limit widespread clinical adoption of this method.
Conclusion
Cheung-Bearelly buckling force monofilament aesthesiometer-triggered latency of LAR movement onset appears to be invariant over an 8.7-dB force range and between the aryepiglottic fold and medial pyriform sinus subsites in asymptomatic, healthy controls. LAR response latency provides an additional measure of laryngopharyngeal sensory function in addition to response threshold and perceptual strength. Based on those data, LAR response latency incongruences between stimulation forces or laryngopharyngeal subsites may serve as pathophysiological features to dissect mechanisms of upper aerodigestive tract disorders.
Footnotes
Acknowledgements
This research will be presented as an oral presentation at the annual meeting of the AAO-HNSF; October 3-6, 2021; Los Angeles, California.