- Email: [email protected]
Laryngeal and Pharyngeal Movements During Inner Singing: A Cross-Sectional Study
ARTICLE IN PRESS
Laryngeal and Pharyngeal Movements During Inner Singing: A Cross-Sectional Study € lfler, ‡Naglaa Mansour, and *,†Carmen Unterhofer, *,†Anna Maria Stefanie Buchberger, †Olivia Jeleff-Wo †Simone Graf, yzMunich, Germany SUMMARY: Introduction. Laryngeal and pharyngeal activity during inner singing is discussed in the context of vocal hygiene. Inner singing is defined as imagined singing, reading music silently, and listening to vocal music. When vocal rest is prescribed, doctors, speech therapists, and voice pedagogues recommend avoiding listening to music or reading music silently, since it is suggested that inner singing unconsciously influences the glottis, and thus moves the vocal folds involuntarily. The aim of this study was to compare the degree to which involuntary laryngeal and/or pharyngeal activity occur during inner singing, inner speech, and at rest, and to evaluate if current recommendations concerning vocal hygiene are still reasonable. Material and method. Thirty vocally healthy participants were examined transnasally with a flexible videoendoscope. The sample consisted of 10 nonsingers, 10 lay singers, and 10 professional singers. Participants were examined during five tasks including rest, silent reading, imagining a melody, listening to music, and reading music. Two medical doctors specializing in phoniatrics analyzed the videos both qualitatively and quantitatively. Results. During the endoscopic examination, the raters identified movements at the base of the tongue, the posterior and lateral pharynx wall, the arytenoid cartilage, and the vocal folds. The inner singing tasks showed significantly more laryngeal movements as well as significantly more glottal closures than the control tasks (at rest, silent reading). Pharyngeal structures did not show an increase in activity during inner singing. These findings were independent of the level of proficiency in singing. Conclusion. When total vocal rest is prescribed, patients should also be advised to avoid music imagination. Still, further research is needed to survey in detail the actual effects of these involuntary movements during inner singing on the regeneration process of vocal fold healing. Key Words: Inner Singing−Vocal hygiene−Vocal Folds−Singer−Vocal rest−Voice rest.
INTRODUCTION Vocal hygiene Several specialists recommend vocal rest or voice conservation during acute vocal complaints,1−5 although there is some controversy concerning the effectiveness of these measures.6 At this juncture, it is important to define vocal rest. Richter and Sandel4 and Sataloff7 distinguish absolute and relative voice rest. Absolute voice rest means total silence, both silence at work and in everyday life. The patients should not speak, sing, or whisper. Furthermore, they should not listen to music or read musical scores silently, because the vocal folds are assumed to move involuntarily. Relative voice rest means activating voice rest. The patients should speak as little as possible. They should speak or sing with technical awareness and should practice only clearly defined exercises. They should not actively make music, listen to music, or read music. Instead of absolute voice rest, active voice care should follow. Accepted for publication February 19, 2019. * These authors contributed equally. From the yDepartment of Otorhinolaryngology/Phoniatrics, Klinikum rechts der Isar, Technical University Munich, Munich, Germany; and the zDepartment of Otorhinolaryngology, Klinikum rechts der Isar, Technical University Munich, Munich, Germany. Address correspondence and reprint request to Carmen Unterhofer, Department of Otorhinolaryngology/Phoniatrics, Klinikum rechts der Isar, Technical University Munich, Street 22 81675, Munich, Germany. E-mail: [email protected] Journal of Voice, Vol. &&, No. &&, pp. &&−&& 0892-1997 © 2019 Published by Elsevier Inc. on behalf of The Voice Foundation. https://doi.org/10.1016/j.jvoice.2019.02.011
Vocal rest is prescribed by vocal specialists (ENT specialists, phoniatrists) for dysphonia caused by acute infections such as laryngitis,1,6 excessive vocal strain, mucosal swelling,4 and after laryngeal surgery1,6,8 or laryngeal trauma.6 The multiple causes for dysphonia do not have specific types of therapeutic vocal rest associated with them. There are no precise recommendations regarding the kind of vocal rest that should be prescribed in any given situation, and there is no agreement on the level of safe vocal activity after specific traumas or surgeries.2,3,6,8 One reason could be the large range between absolute vocal rest and normal, unrestricted vocal activity. This complicates the medical practitioner's recommendations and prescriptions.7 Furthermore, there are no guidelines concerning the duration of prescribed vocal rest. Different studies recommend between three to seven days, depending on the type of diagnosis and its degree of severity. This lack of clarity is associated with the vague definition of vocal rest in different papers.9
Neurological aspects In the human brain, the processing of action (the planning or execution of a movement) and perception (the pure awareness of the action) happen partially in the same cortical areas.10 Kleber et al11 found similar activity in some cortical and subcortical areas during inner and active singing. However, some specific areas were active only during inner singing.
ARTICLE IN PRESS 2
Journal of Voice, Vol. &&, No. &&, 2019
Processes governing the imagination of music and processes governing music perception seem to be handled by similar neuronal mechanisms.12−15 This observation is an important foundation for our study: if action and perception are processed cortically in the same brain areas, it might be assumed that movements of the larynx should be observable not only during action, but also during perception.
Inner singing Little literature exists on the issue of inner singing.16 There are more than 60 similar terms, which are closely linked in content17 since most studies describe different empirical findings using the same terms. Although many studies describe inner singing, there are, to our knowledge, no studies that examined it endoscopically with a group of subjects. Looking at the current research situation, several authors have suggested that unconscious movements of the larynx occur during inner singing (cf.4,16−19). Several authors have stated the opinion that only professional musicians, and not lay singers (LS) or nonsingers (NS), are capable of inner singing. It could be that inner singing becomes refined during their musical education.16,18,20 Kleber et al11 examined 16 singers with surface electromyography (EMG) and could not see increased activity during inner singing compared to rest. There was no significant difference between vocal tract movement during normal breathing and imagined singing. This fact stands in contrast to the widely published hypothesis that involuntary laryngeal movement occurs during inner singing. It indicates the necessity for further research in this area.
Issues The aim of this study was (1) to find out which kinds of movements could be seen during inner singing and which anatomical structures were involved. (2) We also examined how these movements differed depending on the tasks given to the participants, and whether professional singers(PS) moved differently from LSs or NS. Finally, (3) we examined the degree to which participants were aware of their movements during the experiment. MATERIALS AND METHODS Sample Ten NS, 10 LS, and 10 PS participated in the study. Each subgroup consisted of five women and five men. The age of the participants ranged from 22 to 38 years. Inclusion criteria were the following: All participants should be between the ages of 18 and 40. NS should not sing regularly and should not have undergone a singing education or vocal studies. LS should sing regularly but not as a full-time occupation, nor should they have studied singing professionally. The PS should have studied voice and should sing regularly. Their main job should be as a singer.
Exclusion criteria were any lung-related, laryngeal, neurologic, or oncologic diseases in the prior medical history as well as an inability to participate in the study because of mental or linguistic incomprehension. Using these questionnaires, which are described further down, the participants could be categorized in three groups and inclusion and exclusion criteria could be checked. Materials Two questionnaires were handed out to the participants of the study: The Voice Handicap Index (VHI)21 and the Singing Voice Handicap Index (SVHI)22 were used to gather more information about the sample and to find out if any singer or NS reported having a voice disorder. Also, the participants were instructed to give a subjective assessment of larynx and pharynx movements during the examination. The transnasal videoendoscopical stroboscopy recorded movements of the larynx and pharynx. The advantage of this procedure is that it uses a flexible endoscope, so that the tongue and the oral zone are not obstructed as they would be with a rigid endoscope while performing a stroboscopy for the vibrations of the vocal folds. Additionally, when using the transnasal method there are fewer or no interruptions due to the gag reflex. This examination technique is also commonly used as a diagnostic instrument in ambulant surgery.23,24 Implementation After the participants had completed all the questionnaires they were examined endoscopically. To reduce nasal mucosal swelling and for nasal local anaesthesia, medical cotton with Lidocain 5%, and Xylomethazolin was put into the nasal passage for 5 minutes. Then the 3.4 mm thin flexible videoendoscope with microphone (XION Video Nasopharyngoscope, Germany) was inserted through the nose. The side of the nasal passage was chosen depending on the individual anatomy of each participant. The microphone was used to document and exclude any phonation from the participants during the examination. The endoscope was positioned directly above the uvula. This enabled a good view over the oro-, and hypopharynx as well as the larynx.24−26 With this position (swallow position) neither the swallowing procedure nor other laryngeal movements were affected.25 The subjects then produced the syllable [hi], so the phoniatric specialist could exclude any movement disorders. The study included five tasks: at rest (AR), silent reading (SR), imagining a melody (IM), listening to music (LM) and reading music (RM)—three musical tasks, and two nonmusical tasks. During the examination, a phoniatric specialist created video and audio recordings of the larynx, pharynx, and the base of the tongue. To avoid possible biases from the tasks influencing each other, the order of the tasks was randomized for each participant. The instructions remained the same. During the first task, the participants were AR (nonmusical task), and were instructed to stay
ARTICLE IN PRESS Carmen Unterhofer, et al
Laryngeal and Pharyngeal Movements During Inner Singing
quiet. In the second task, the attendants were asked to read a text (Der Nordwind und die Sonne) silently (nonmusical task). The third task was to silently sight-sing an unknown melody (musical task). During the fourth task, the participants were asked to imagine singing a famous children's song (Alle V€ ogel sind schon da, musical task). The fifth task had the participants listen to a well-known Opera aria (La donna e mobile, musical task). Every task was to be carried out without using the voice. Each participant was recorded for 2.5 minutes, 30 seconds per task. At the end of the examination, the attendants were asked to give a subjective assessment of larynx and pharynx movements during the different situations. They were asked to indicate if they had felt movements in the larynx or not. Analysis The 150 videos were shuffled and anonymized. An ENT specialist and a phoniatric specialist analyzed the videos. Neither the task nor the group of subjects was familiar to the raters. The answer sheet (Figure A.1) was developed specifically for this study. The number of pharyngeal and laryngeal movements and type of laryngeal movements were to be filled in: The rater indicated the number of the movements of the base of the tongue, the pharyngeal wall, the vocal folds, and the arytenoid cartilage. The rater also indicated if the vocal folds stood lateral, intermediary, paramedian or median, and how often the vocal folds closed or touched each other. The number of movements was scaled in four categories. More than 10 movements were categorized as often, six to ten movements as sometimes, one to five as rarely, and zero as never. Statistical analysis The statistics software SPSS Statistics 24 by IBM was used to create a statistical analysis of the gathered data. The reliability of the results was evaluated using Cohen's kappa (K). Range, mean (M), and standard deviation (SD) were used for descriptive statistics. For inferential statistics, a dependent t test with paired samples was calculated to compare the mean difference between two observations. The sample with the mean values of the measurements were tested for their distribution prior to the decision which test was used. Since the sample was normally distributed and paired measurement were conducted, the paired sample t test was chosen. RESULTS The evaluation of the questionnaires showed that no singer reported having an acute voice disorder. But when calculating the (S)VHI score, in the SVHI two participants from the PS group showed an increased figure (41/144 points), which does not show up as abnormal. In the VHI one of the NS showed a mild dysphonia (21/120 points) and one of the PS a moderate dysphonia (33/120 points). The degree of severity was made based on Nawka27 and Lorenz et al28. One out of 30 participants noted, that he smoked.
3
Reliability Intra- and inter-rater reliability were verified through the agreement between the raters within the four rating categories (often, sometimes, rarely, and never). Both raters analyzed all videos. A moderate agreement was found for movements of the vocal folds and the arytenoid cartilage (K = 0.47) and the median movements of the vocal folds (K = 0.52) (Table A.1). The raters achieved a fair agreement for movements of the base of the tongue (K = 0.34) and a slight agreement for the movements of the pharyngeal wall (K = 0.20). All data were statistically significant (level of significance P = 0.1). To ensure intra-rater reliability, one rater analyzed a statistically sufficient segment of the recordings a second time. As seen in Table A.1, the rater achieved a substantial agreement for movements of the base of the tongue (K = 0.61) and median movements of the vocal folds (K = 0.79), a moderate agreement for the movements of the vocal folds and the arytenoid cartilage (K = 0.59), and a slight agreement for the movements on the pharyngeal wall (K = 0.16). All the data except movements of the pharyngeal wall were statistically significant (level of significance P = 0.1). (1) Which kinds of movements could be seen during inner singing and which anatomical structures were involved? The recordings showed movement at the base of the tongue, the posterior and lateral pharynx wall, the arytenoid cartilage, the vocal folds, and the epiglottis. With the given examination method, an assessment of laryngeal raising and lowering was not possible. The vestibular folds did not show an increase in activity. Observed movements were a mixture of slow movements with sustained tone, faster and shorter, more repetitive movements as well as combinations of the two. There were some participants who showed a great deal laryngeal and pharyngeal activity during the whole examination. Others showed little to no movement during all tasks. Most vocal fold movements were intermediary (Table 1) (range: 0−27; M = 3.74; SD = 4.57). The rater also observed paramedian (range: 0−17; M = 1.71; SD = 2.91) as well as lateral (range: 0−13; M = 2.51; SD = 3.48) movements. Glottal closures were the rarest of the observed movements (range: 0−8; M = 0.35; SD = 1.15). Among the 90 videos taken during musical tasks, rater 1 noticed glottal closures in 18.9% of the cases, rater 2 in 21.1%. However, none of the participants showed vibrations of the vocal folds. (2) What were the differences between the observed movements regarding task and the grade of professionalism of singing? During both inner singing (RMc, LM, and IM) and inner speech (SR), laryngeal and partially pharyngeal movements were most notable. However, these movements also occurred AR. Clear differences between the groups of participants within each task could not be identified. Instead, the groups showed differences with respect to which anatomical structures moved the most. NS and LS displayed the most laryngeal movement. The most movement at the base of the tongue was recorded for PS while NS showed the most pharyngeal movements.
ARTICLE IN PRESS 4
Journal of Voice, Vol. &&, No. &&, 2019
TABLE 1. Number of Vocal Folds Movements: Lateral, Intermediary, Paramedian, Median
Lateral movements Intermediary movements Paramedian movements Median movements
N
Range
Sum
Mean
Standard Deviation
147 147 147 147
0−13 0−27 0−17 0−8
369 550 251 52
2.51 3.74 1.71 0.35
3.48 4.57 2.91 1.15
Figure 1 shows the frequency of movements of the arytenoid cartilages and the vocal folds in different tasks and group of subjects. In every group, laryngeal movements were less frequent during AR and SR. Both rater 1 and rater 2 observed this. Looking more closely at the group of subjects, PS showed the least laryngeal movements, NS (rater 1) and LS (rater 2) showed the most. Figure 2 shows the frequency of glottal closures in different tasks and group of subjects. Independently from the group of subjects glottal closures where seen more often during the musical tasks (IM, LM, and RM) from both rater 1 and rater 2. This can be seen also in Figure 3. While during rest the vocal folds are open, during RM without using phonation the vocal folds are closed with a small dorsal gap. Since there were no significant differences between the three musical tasks (IM, LM, and RM), these tasks were summarized for the following analysis (Figure 1). This way a comparison from the average movements during inner singing (IM, LM, and RM) and AR or SR was possible. As shown in Table 2 during the inner singing task, there were significantly more laryngeal movements as well as significantly more glottal closures than during the control tasks (AR, SR). This observation did not include the pharyngeal structures.
The study revealed a significant average difference in laryngeal movements between the scenarios inner singing and AR (rater 1: t28 = 4.94, P < 0.001; rater 2: t29 = 3.70, P < 0.01). On average, during inner singing, 2.28 (rater 1) and 2.56 (rater 2) more movements were observed in comparison to the situation AR (rater 1: CI = 95% [1.33−3.22]; rater 2: CI = 95% [1.14−3.97]) (Table 2). Also, the difference between inner singing and SR was significant. The average difference between these tasks showed 2.63 (rater 1) and 2.46 (rater 2) more movements during inner singing (rater 1: t29 = 5.43, P < 0.001; rater 2: t29 = 5.10, P < 0.001). On the other hand, the comparison of SR and AR yielded no significant results (Table A.2). Also, more glottal closures were observed during the musical tasks than were observed during the nonmusical tasks as seen in Figure 2. Comparing the average number of median vocal fold movements for inner singing and AR showed that there were more movements during inner singing for both raters. This observation was significant for rater 2 (t26 = 2.06, P < 0.05) while it was not for rater 1 (Table 2). In the SR task, there were also significantly fewer glottal closures on average than during inner singing (rater 1: t29 = 2.96, P < 0.01; rater 2: t29 = 2.75, P < 0.05) (Table A.2).
FIGURE 1. Frequency of movements of the arytenoid cartilage and the vocal folds in the three different groups (each N = 10) and settings. Abbreviations: AR, at rest; IM, imagining a melody; LM, listening to music; LS, lay singers; NS, nonsingers; PS, professional singers; RM, reading music; SR, silent reading.
ARTICLE IN PRESS Carmen Unterhofer, et al
Laryngeal and Pharyngeal Movements During Inner Singing
5
FIGURE 2. Frequency of glottal closures in the three different groups (each N = 10) and settings. Abbreviations: AR, at rest; IM, imagining a melody; LM, listening to music; LS, lay singers; NS, nonsingers; PS, professional singers; RM, reading music; SR, silent reading.
Rater 1 observed the most median vocal fold movements during the RM task while rater 2 noticed the most during the LM task. For both raters, the least glottal closures happened in the PS group, the most closures in the NS group. This lies in accordance with the finding that NS showed the most laryngeal movements while PS displayed the least. (3) Are the participants aware of their movements? Eighteen attendants reported to have felt movements in the larynx (three NS, nine PS, and six LS). None of these happened AR and only a few during SR (four PS). PS felt the most movements across all tasks except for LM. Five of 10 LS also gave account of feeling movements during that task (one NS, five LS, and four PS). NS was the group with the least subjectively felt movement. For example, during the RM task, eight out of 10 PS reported laryngeal activity, while only four LS and two NS reported the same.
DISCUSSION Our results are in agreement with some results previously published by other researchers. They inferred laryngeal activity during the imagination of music (cf.4,16−19), an inference we could prove endoscopically. However, we were also able to observe laryngeal and pharyngeal movement not only during musical tasks, but also AR, albeit to a significantly lesser extent. In this study, there were significantly more laryngeal movements during inner singing (RM, LM, IM) than during the control tasks (AR, SR). Because there were no significant differences between the three musical tasks, these tasks were summarized. There were also significantly more glottal closures during inner singing than AR. This is important if total vocal rest is prescribed by phoniatrists or ENT specialists. The specific consideration of median movements
FIGURE 3. Example of the larynx at rest (left) and the larynx during reading music (right) (Department of Otorhinolaryngology/ Phoniatrics, Klinikum rechts der Isar, Dr. S. Graf).
ARTICLE IN PRESS 6
Journal of Voice, Vol. &&, No. &&, 2019
TABLE 2. Paired Samples t test—Inner Singing vs At Rest Laryngeal Movements: Inner Singing vs at Rest Paired Differences 95% CI
Rater 1 Rater 2
Laryngeal movements: inner singing—at rest Laryngeal movements: inner singing—at rest
M
SD
SEM
Lower
Upper
t
df
2.28
2.48
0.46
1.33
3.22
4.94***
28
2.56
3.79
0.69
1.14
3.97
3.70**
29
Glottal Closure: Inner Singing vs at Rest Paired Differences 95% CI
Rater 1 Rater 2
glottal closure: inner singing—at rest glottal closure: inner singing—at rest
M
SD
SEM
Lower
Upper
t
df
0.32
1.09
0.20
¡0.09
0.74
1.60 n.s.
28
0.34
0.86
0.17
0.00
0.68
2.06*
26
* P < 0.05. ** P < 0.01. *** P < 0.001. Abbreviations: M, mean; SD, standard deviation; SEM, Standard Error of Mean; CI, confidence interval; t, t-value; df, degrees of freedom; n.s., non-significant.
of the vocal folds is necessary because frequent glottal closures could influence wound healing. During our experiment, no movement of the pharynx was observed. Movements at the base of the tongue and the posterior and lateral pharynx wall did not show an increase in activity during inner singing, and moved only in a random manner. This might be due to the fact that pharyngeal structures are less involved in the process of phonation. It has been suggested that only well-trained singers are capable of inner singing, and that during inner singing their larynx move more than the laryngeal structures of the NS (cf.16,18,20). This hypothesis cannot be confirmed. On the contrary, NS showed tendentially the most laryngeal movement during all tasks, and PS showed the least. This might be explained by the ability of trained singers (PS and LS) to suppress or activate movements more consciously, having developed a greater awareness of the vocal tract. This possibility must be considered while interpreting the results. Interestingly, the participants’ subjective reports did not match the results the raters produced. According to the raters, NS displayed the most laryngeal activity while PS showed the least. PS was the group with the most subjectively felt movement, and only a few NS felt laryngeal activity during the musical tasks. Movements happened during all tasks for all participants. This discrepancy might be due to not only by the awareness of PS of the vocal tract (they feel more, and move less), but also to the expectations of the individual participants. With their knowledge of the aim of the study, none of the participants assumed to feel laryngeal
or pharyngeal movement AR, and yet, the rater observed movement also during this task. It remains unclear if there is a correlation between the subjective movements and objectively observed laryngeal movements. Furthermore, outliers within each group should be considered. In general, our study suggested that PS moved their glottises less during inner singing than did LS, but within all three groups there were individuals who did not fit this trend. It should not be assumed, for instance, that any given PS can safely listen to music without involuntary glottis movement, and this indicates the importance of individualized recommendations concerning vocal rest. We investigated endoscopically the influence of inner singing on laryngeal and pharyngeal movement and were able to show involuntary activity in the vocal tract. One might call this phenomenon senso-motoric inner singing.
Methodological reflections and view In addition to transnasal videoendoscopical stroboscopy, intramuscular EMG could be used to obtain more specific and clear information. It has already been shown that surface EMG detects no significant differences between breathing and inner singing as previously published by Kleber.11 While intramuscular EMG could yield more helpful results, this procedure is difficult to support ethically because needle electrodes would have to be placed into the muscles around the larynx. This technique will lead to at least some
ARTICLE IN PRESS Carmen Unterhofer, et al
7
Laryngeal and Pharyngeal Movements During Inner Singing
discomfort for the participants and incur a higher risk of unwanted side effects due to the invasiveness of the method. Another possibility for future research could be to modify or extend the tasks, especially with regard to wound healing, since the effect on the healing process remains uncertain. In this scenario, a cohort of patients needing medical vocal rest could be assembled and the observed rate of healing correlated with the participants’ grade of unconscious laryngeal activity. It could be assumed that inner singing has no effect on wound healing because we observed no vibrations of the vocal folds, only glottal closures which also occur during swallowing and coughing. To our knowledge, there exists no literature on the fact, if swallowing affects the wound healing process. It could be assumed that during inner singing there is a higher muscles tension in the entire laryngeal area besides the glottal closures that might occur. This fact could influence the wound healing. Yet, the exact effect is not clear. Furthermore, a clarification of inner singing is important. Doctors, specialists, speech therapists, and vocal pedagogues should be aware of inner singing, because at the moment the recommendations are diverse. We are left with the need for a clear definition of vocal rest. Vocal specialists should make clear to their patients that listening to singing, studying singing, or even imagining singing might engage their vocal mechanisms and impede their healing. Our findings suggested that PS or LS were able to suppress or control movements consciously and thus show less movement during inner singing. However, statistical outliers existed in each group; there were NS who barely moved during all tasks, and PS who moved a great deal. This complexity should be considered in further studies and in clinical settings. During clinical diagnostics, the larynx should be examined in various situations (phonation, AR, imagined singing) to give corresponding individual recommendations concerning vocal hygiene. CONCLUSION In this study, we investigated 10 vocally healthy NS, 10 LS, and 10 PS transnasally with a flexible videoendoscope. The aim of this study was to find out if pharyngeal and laryngeal
movements could be observed during inner singing, and if so, to describe the nature of those movements. During two nonmusical tasks (AR, SR) and three musical tasks (IM, LM, RM) the participants were examined, and two specialists analyzed the videos qualitatively and quantitatively. During inner singing (musical tasks), there were significantly more laryngeal movements, as well as significantly more glottal closures, observed in our participants than while AR or during SR. These movements could possibly impair healing processes. Differences in the level of activity were observed based on the grade of professionalism of singing, but outliers existed in all three groups. Pharyngeal movements where observed only in a random manner. Vocal fold vibrations were not observed at all. In summary, based on our findings we recommend that music imagination (LM or practicing musical scores silently) should be avoided when vocal rest is prescribed. That said, the extent of vocal rest could be adapted based on the individual patient's awareness of movement and level of activity of the vocal tract. These would have to be examined endoscopically. Further studies are needed concerning the effects of inner singing on the healing process when vocal rest has been prescribed for laryngeal disease. ACKNOWLEDGMENTS We are very grateful for the statistical help and advisement from Thomas Maierhofer with the Statistisches Beratungslabor LMU in Munich (StaBLab). This study was supported by the Lehre@Forschung LMU. We are very thankful for the independent financial support. SUPPLEMENTARY MATERIALS Supplementary material associated with this article can be found in the online version at https://doi.org/10.1016/j. jvoice.2019.02.011. APPENDIX Table A.1, Table A.2.
TABLE A.1. Intrarater and Interrater Reliability
Base of the tongue (never, rarely, sometimes, often) Pharyngeal wall (never, rarely, sometimes, often) Vocal folds and the arytenoid cartilage (never, rarely, sometimes, often) Median movements of the vocal folds (yes, no) *** P < 0.001.
Intrarater Reliability Kappa
Inter-rater Reliability Kappa
0.61***
0.34***
0.16
0.20***
0.59***
0.47***
0.79***
0.52***
ARTICLE IN PRESS 8
Journal of Voice, Vol. &&, No. &&, 2019
TABLE A.2. Paired Samples t test—Inner Singing vs Silent Reading Laryngeal Movements: Inner Singing vs Silent Reading Paired Differences 95% CI
Rater 1 Rater 2
Laryngeal movements: inner singing—silent reading Laryngeal movements: inner singing—silent reading
M
SD
SEM
Lower
Upper
t
df
2.63
2.66
0.49
1.64
3.63
5.43***
29
2.46
2.64
0.48
1.47
3.44
5.10***
29
Glottal Closure: Inner Singing vs Silent Reading Paired Differences 95% CI
Rater 1 Rater 2
Glottal closure: inner singing—silent reading Glottal closure: inner singing—silent reading
M
SD
SEM
Lower
Upper
t
df
0.54
1.001
0.18
0.17
0.92
2.96**
29
0.41
0.81
0.15
0.10
0.71
2.75*
29
* P < 0.05. ** P < 0.01. *** P < 0.001. Abbreviations: CI, confidence interval; df, degrees of freedom; M, mean; SD, standard deviation; SEM, Standard Error of Mean; t, t value.
FIGURE A.1. Answer sheet (example). REFERENCES 1. D€ ohne E. Bedeutet Fl€ ustern Stimmruhe oder Stimmschonung? J Mol Med. 1942;21:692–693. 2. Behrman A, Sulica L. Voice rest after microlaryngoscopy: current opinion and practice. Laryngoscope. 2003;113:2182–2186. 3. Ishikawa K, Thibeault S. Voice rest versus exercise: a review of the literature. J Voice. 2010;24:379–387. 4. Richter B, Sandel M. Gesunderhaltung, hygiene, pr€avention. In: Richter B, Echternach M, eds. Die Stimme: Grundlagen, k€ unstlerische Praxis, Gesunderhaltung. 2nd ed.Henschel: Leipzig; 2014:197–218. 5. Fedrowitz H. Stimmhygiene—ein wichtiger Baustein zur Pr€avention von Berufsdysphonien bei Lehrern. Praxis Sprache. 2016;61: 124–128. 6. Behlau M, Oliveira G. Vocal hygiene for the voice professional. Curr Opin Otolaryngol Head Neck Surg. 2009;17:149–154.
7. Sataloff RT. Voice Rest. In: Sataloff RT, ed. Professional Voice: The Science and Art of Clinical Care. New York: Raven Press; 1991. 8. Koufman JA, Blalock DP. Is voice rest never indicated? J Voice. 1989;3:87–91. 9. Kiagiadaki D, Remacle M, Lawson G, Bachy V, van der Vorst S. The effect of voice rest on the outcome of phonosurgery for benign laryngeal lesions: preliminary results of a prospective randomized study. Ann Otol Rhinol Laryngol. 2015;124:407–412. 10. Maes P-J, Leman M, Palmer C, Wanderley MM. Action-based effects on music perception. Front Psychol. 2014;4:1–14. 11. Kleber B, Birbaumer N, Veit R, Trevorrow T, Lotze M. Overt and imagined singing of an Italian aria. NeuroImage. 2007;36:889–900. 12. Zatorre RJ, Halpern AR, Perry DW, Meyer E, Evans AC. Hearing in the mind's ear: a PET investigation of musical imagery and perception. J Cognit Neurosci. 1996;8:29–46.
ARTICLE IN PRESS Carmen Unterhofer, et al
Laryngeal and Pharyngeal Movements During Inner Singing
13. Zatorre RJ. Brain imaging studies of musical perception and musical imagery. J New Music Res. 1999;28:229–236. 14. Aleman A, Nieuwenstein MR, B€ ocker KB, de Haan EHF. Music training and mental imagery ability. Neuropsychologia. 2000;38:1664– 1668. 15. Herholz SC, Halpern AR, Zatorre RJ. Neuronal correlates of perception, imagery, and memory for familiar tunes. J Cognit Neurosci. 2012;24:1382–1397. 16. Habermann G. Stimme und Mensch: Beobachtungen und Betrachtungen. Heidelberg: Median-Verl; 1996. 17. N. Saintilan, The use of imagery during the performance of memorized music, psychomusicology: music, mind, and brain 24 (2014) 309−315. 18. Frank F. Bemerkung zum inneren Singen. Sprache Stimme Geh€or. 1989;13:8–10. 19. Nawka T, Wirth G, Anders L-C. Stimmst€orungen: F€ur Ärzte, Logop€ aden, Sprachheilp€ adagogen und Sprechwissenschaftler; mit 30 Tabellen. 5th ed. K€ oln: Dt. Ärzte-Verl.; 2008. 20. Nadolecƶny M. Untersuchungen €uber den Kunstgesang: I. Atem- und Kehlkopfbewegungen. Berlin, Heidelberg: Springer Berlin Heidelberg; 1923. 21. Deutsche Gesellschaft f€ ur Phoniatrie und P€adaudiologie e.V., Voice Handicap Index, deutsche Fassung, available at http://www.dgpp.de/ Profi/Sources/VHI-Bogen-Konsensfassung_(C)DGPP%202003-2.pdf (accessed on February 16, 2018). 22. A. Lorenz, Singing Voice Handicap Index, deutsche Fassung, available at http://www.dgpp.de/cms/media/download_gallery/SVHI-Testanleitung. pdf (accessed on February 16, 2018).
9
23. Brockmann-Bauser M, Bohlender JE. Praktische Stimmdiagnostik: Theoretischer und praktischer Leitfaden. 1st ed. Stuttgart: Georg Thieme Verlag KG; 2014. 24. Fleischer S, Hess M. Systematischer Untersuchungsablauf bei der Videolaryngoskopie und Stroboskopie. HNO. 2003;51:160–167. 25. Schr€ oter-Morasch H. Klinische Untersuchung des Oropharynx und videoendoskopische Untersuchung der Schluckfunktion. In: Bartolome G, Schr€ oter-Morasch H, Buchholz D, Feussner H, Graf S, Holzapfel K, Neumann S, Pehl C, Prosiegel M, eds. Schluckst€orungen: Diagnostik und Rehabilitation; [mit dem Plus im Web; Zugangscode im Buch]. 5th ed. M€ unchen: Elsevier Urban & Fischer; 2014:167–204. 26. Schr€ oter-Morasch H, Graf S. Dysphagiediagnostik durch den HNOArzt. HNO. 2014;62:324–334. 27. Nawka T. Voice Handicap Index, deutsche Fassung, available at http://www.dgpp.de/Profi/Sources/vhi-dt_2006.pdf (accessed on February 16, 2018). 28. Lorenz A, Gonnermann U, Nawka T, Graduierung der deutschen Fassung des Singing Voice Handicap Index (SVHI), in: Deutsche Gesellschaft f€ ur Phoniatrie und P€adaudiologie e.V. (DGPP) (Eds.). 31. Wissenschaftliche Jahrestagung der Deutschen Gesellschaft f€ ur Phoniatrie und P€adaudiologie (DGPP) zusammen mit dem 5. P€adakustiker-Symposium der Akademie f€ ur H€ orger€ate-Akustik, L€ ubeck, 2014, available at http://www.egms.de/static/en/meetings/ dgpp2014/14dgpp52.shtml (accessed on February 16, 2018).
Laryngeal and Pharyngeal Movements During Inner Singing: A Cross-Sectional Study € lfler, ‡Naglaa Mansour, and *,†Carmen Unterhofer, *,†Anna Maria Stefanie Buchberger, †Olivia Jeleff-Wo †Simone Graf, yzMunich, Germany SUMMARY: Introduction. Laryngeal and pharyngeal activity during inner singing is discussed in the context of vocal hygiene. Inner singing is defined as imagined singing, reading music silently, and listening to vocal music. When vocal rest is prescribed, doctors, speech therapists, and voice pedagogues recommend avoiding listening to music or reading music silently, since it is suggested that inner singing unconsciously influences the glottis, and thus moves the vocal folds involuntarily. The aim of this study was to compare the degree to which involuntary laryngeal and/or pharyngeal activity occur during inner singing, inner speech, and at rest, and to evaluate if current recommendations concerning vocal hygiene are still reasonable. Material and method. Thirty vocally healthy participants were examined transnasally with a flexible videoendoscope. The sample consisted of 10 nonsingers, 10 lay singers, and 10 professional singers. Participants were examined during five tasks including rest, silent reading, imagining a melody, listening to music, and reading music. Two medical doctors specializing in phoniatrics analyzed the videos both qualitatively and quantitatively. Results. During the endoscopic examination, the raters identified movements at the base of the tongue, the posterior and lateral pharynx wall, the arytenoid cartilage, and the vocal folds. The inner singing tasks showed significantly more laryngeal movements as well as significantly more glottal closures than the control tasks (at rest, silent reading). Pharyngeal structures did not show an increase in activity during inner singing. These findings were independent of the level of proficiency in singing. Conclusion. When total vocal rest is prescribed, patients should also be advised to avoid music imagination. Still, further research is needed to survey in detail the actual effects of these involuntary movements during inner singing on the regeneration process of vocal fold healing. Key Words: Inner Singing−Vocal hygiene−Vocal Folds−Singer−Vocal rest−Voice rest.
INTRODUCTION Vocal hygiene Several specialists recommend vocal rest or voice conservation during acute vocal complaints,1−5 although there is some controversy concerning the effectiveness of these measures.6 At this juncture, it is important to define vocal rest. Richter and Sandel4 and Sataloff7 distinguish absolute and relative voice rest. Absolute voice rest means total silence, both silence at work and in everyday life. The patients should not speak, sing, or whisper. Furthermore, they should not listen to music or read musical scores silently, because the vocal folds are assumed to move involuntarily. Relative voice rest means activating voice rest. The patients should speak as little as possible. They should speak or sing with technical awareness and should practice only clearly defined exercises. They should not actively make music, listen to music, or read music. Instead of absolute voice rest, active voice care should follow. Accepted for publication February 19, 2019. * These authors contributed equally. From the yDepartment of Otorhinolaryngology/Phoniatrics, Klinikum rechts der Isar, Technical University Munich, Munich, Germany; and the zDepartment of Otorhinolaryngology, Klinikum rechts der Isar, Technical University Munich, Munich, Germany. Address correspondence and reprint request to Carmen Unterhofer, Department of Otorhinolaryngology/Phoniatrics, Klinikum rechts der Isar, Technical University Munich, Street 22 81675, Munich, Germany. E-mail: [email protected] Journal of Voice, Vol. &&, No. &&, pp. &&−&& 0892-1997 © 2019 Published by Elsevier Inc. on behalf of The Voice Foundation. https://doi.org/10.1016/j.jvoice.2019.02.011
Vocal rest is prescribed by vocal specialists (ENT specialists, phoniatrists) for dysphonia caused by acute infections such as laryngitis,1,6 excessive vocal strain, mucosal swelling,4 and after laryngeal surgery1,6,8 or laryngeal trauma.6 The multiple causes for dysphonia do not have specific types of therapeutic vocal rest associated with them. There are no precise recommendations regarding the kind of vocal rest that should be prescribed in any given situation, and there is no agreement on the level of safe vocal activity after specific traumas or surgeries.2,3,6,8 One reason could be the large range between absolute vocal rest and normal, unrestricted vocal activity. This complicates the medical practitioner's recommendations and prescriptions.7 Furthermore, there are no guidelines concerning the duration of prescribed vocal rest. Different studies recommend between three to seven days, depending on the type of diagnosis and its degree of severity. This lack of clarity is associated with the vague definition of vocal rest in different papers.9
Neurological aspects In the human brain, the processing of action (the planning or execution of a movement) and perception (the pure awareness of the action) happen partially in the same cortical areas.10 Kleber et al11 found similar activity in some cortical and subcortical areas during inner and active singing. However, some specific areas were active only during inner singing.
ARTICLE IN PRESS 2
Journal of Voice, Vol. &&, No. &&, 2019
Processes governing the imagination of music and processes governing music perception seem to be handled by similar neuronal mechanisms.12−15 This observation is an important foundation for our study: if action and perception are processed cortically in the same brain areas, it might be assumed that movements of the larynx should be observable not only during action, but also during perception.
Inner singing Little literature exists on the issue of inner singing.16 There are more than 60 similar terms, which are closely linked in content17 since most studies describe different empirical findings using the same terms. Although many studies describe inner singing, there are, to our knowledge, no studies that examined it endoscopically with a group of subjects. Looking at the current research situation, several authors have suggested that unconscious movements of the larynx occur during inner singing (cf.4,16−19). Several authors have stated the opinion that only professional musicians, and not lay singers (LS) or nonsingers (NS), are capable of inner singing. It could be that inner singing becomes refined during their musical education.16,18,20 Kleber et al11 examined 16 singers with surface electromyography (EMG) and could not see increased activity during inner singing compared to rest. There was no significant difference between vocal tract movement during normal breathing and imagined singing. This fact stands in contrast to the widely published hypothesis that involuntary laryngeal movement occurs during inner singing. It indicates the necessity for further research in this area.
Issues The aim of this study was (1) to find out which kinds of movements could be seen during inner singing and which anatomical structures were involved. (2) We also examined how these movements differed depending on the tasks given to the participants, and whether professional singers(PS) moved differently from LSs or NS. Finally, (3) we examined the degree to which participants were aware of their movements during the experiment. MATERIALS AND METHODS Sample Ten NS, 10 LS, and 10 PS participated in the study. Each subgroup consisted of five women and five men. The age of the participants ranged from 22 to 38 years. Inclusion criteria were the following: All participants should be between the ages of 18 and 40. NS should not sing regularly and should not have undergone a singing education or vocal studies. LS should sing regularly but not as a full-time occupation, nor should they have studied singing professionally. The PS should have studied voice and should sing regularly. Their main job should be as a singer.
Exclusion criteria were any lung-related, laryngeal, neurologic, or oncologic diseases in the prior medical history as well as an inability to participate in the study because of mental or linguistic incomprehension. Using these questionnaires, which are described further down, the participants could be categorized in three groups and inclusion and exclusion criteria could be checked. Materials Two questionnaires were handed out to the participants of the study: The Voice Handicap Index (VHI)21 and the Singing Voice Handicap Index (SVHI)22 were used to gather more information about the sample and to find out if any singer or NS reported having a voice disorder. Also, the participants were instructed to give a subjective assessment of larynx and pharynx movements during the examination. The transnasal videoendoscopical stroboscopy recorded movements of the larynx and pharynx. The advantage of this procedure is that it uses a flexible endoscope, so that the tongue and the oral zone are not obstructed as they would be with a rigid endoscope while performing a stroboscopy for the vibrations of the vocal folds. Additionally, when using the transnasal method there are fewer or no interruptions due to the gag reflex. This examination technique is also commonly used as a diagnostic instrument in ambulant surgery.23,24 Implementation After the participants had completed all the questionnaires they were examined endoscopically. To reduce nasal mucosal swelling and for nasal local anaesthesia, medical cotton with Lidocain 5%, and Xylomethazolin was put into the nasal passage for 5 minutes. Then the 3.4 mm thin flexible videoendoscope with microphone (XION Video Nasopharyngoscope, Germany) was inserted through the nose. The side of the nasal passage was chosen depending on the individual anatomy of each participant. The microphone was used to document and exclude any phonation from the participants during the examination. The endoscope was positioned directly above the uvula. This enabled a good view over the oro-, and hypopharynx as well as the larynx.24−26 With this position (swallow position) neither the swallowing procedure nor other laryngeal movements were affected.25 The subjects then produced the syllable [hi], so the phoniatric specialist could exclude any movement disorders. The study included five tasks: at rest (AR), silent reading (SR), imagining a melody (IM), listening to music (LM) and reading music (RM)—three musical tasks, and two nonmusical tasks. During the examination, a phoniatric specialist created video and audio recordings of the larynx, pharynx, and the base of the tongue. To avoid possible biases from the tasks influencing each other, the order of the tasks was randomized for each participant. The instructions remained the same. During the first task, the participants were AR (nonmusical task), and were instructed to stay
ARTICLE IN PRESS Carmen Unterhofer, et al
Laryngeal and Pharyngeal Movements During Inner Singing
quiet. In the second task, the attendants were asked to read a text (Der Nordwind und die Sonne) silently (nonmusical task). The third task was to silently sight-sing an unknown melody (musical task). During the fourth task, the participants were asked to imagine singing a famous children's song (Alle V€ ogel sind schon da, musical task). The fifth task had the participants listen to a well-known Opera aria (La donna e mobile, musical task). Every task was to be carried out without using the voice. Each participant was recorded for 2.5 minutes, 30 seconds per task. At the end of the examination, the attendants were asked to give a subjective assessment of larynx and pharynx movements during the different situations. They were asked to indicate if they had felt movements in the larynx or not. Analysis The 150 videos were shuffled and anonymized. An ENT specialist and a phoniatric specialist analyzed the videos. Neither the task nor the group of subjects was familiar to the raters. The answer sheet (Figure A.1) was developed specifically for this study. The number of pharyngeal and laryngeal movements and type of laryngeal movements were to be filled in: The rater indicated the number of the movements of the base of the tongue, the pharyngeal wall, the vocal folds, and the arytenoid cartilage. The rater also indicated if the vocal folds stood lateral, intermediary, paramedian or median, and how often the vocal folds closed or touched each other. The number of movements was scaled in four categories. More than 10 movements were categorized as often, six to ten movements as sometimes, one to five as rarely, and zero as never. Statistical analysis The statistics software SPSS Statistics 24 by IBM was used to create a statistical analysis of the gathered data. The reliability of the results was evaluated using Cohen's kappa (K). Range, mean (M), and standard deviation (SD) were used for descriptive statistics. For inferential statistics, a dependent t test with paired samples was calculated to compare the mean difference between two observations. The sample with the mean values of the measurements were tested for their distribution prior to the decision which test was used. Since the sample was normally distributed and paired measurement were conducted, the paired sample t test was chosen. RESULTS The evaluation of the questionnaires showed that no singer reported having an acute voice disorder. But when calculating the (S)VHI score, in the SVHI two participants from the PS group showed an increased figure (41/144 points), which does not show up as abnormal. In the VHI one of the NS showed a mild dysphonia (21/120 points) and one of the PS a moderate dysphonia (33/120 points). The degree of severity was made based on Nawka27 and Lorenz et al28. One out of 30 participants noted, that he smoked.
3
Reliability Intra- and inter-rater reliability were verified through the agreement between the raters within the four rating categories (often, sometimes, rarely, and never). Both raters analyzed all videos. A moderate agreement was found for movements of the vocal folds and the arytenoid cartilage (K = 0.47) and the median movements of the vocal folds (K = 0.52) (Table A.1). The raters achieved a fair agreement for movements of the base of the tongue (K = 0.34) and a slight agreement for the movements of the pharyngeal wall (K = 0.20). All data were statistically significant (level of significance P = 0.1). To ensure intra-rater reliability, one rater analyzed a statistically sufficient segment of the recordings a second time. As seen in Table A.1, the rater achieved a substantial agreement for movements of the base of the tongue (K = 0.61) and median movements of the vocal folds (K = 0.79), a moderate agreement for the movements of the vocal folds and the arytenoid cartilage (K = 0.59), and a slight agreement for the movements on the pharyngeal wall (K = 0.16). All the data except movements of the pharyngeal wall were statistically significant (level of significance P = 0.1). (1) Which kinds of movements could be seen during inner singing and which anatomical structures were involved? The recordings showed movement at the base of the tongue, the posterior and lateral pharynx wall, the arytenoid cartilage, the vocal folds, and the epiglottis. With the given examination method, an assessment of laryngeal raising and lowering was not possible. The vestibular folds did not show an increase in activity. Observed movements were a mixture of slow movements with sustained tone, faster and shorter, more repetitive movements as well as combinations of the two. There were some participants who showed a great deal laryngeal and pharyngeal activity during the whole examination. Others showed little to no movement during all tasks. Most vocal fold movements were intermediary (Table 1) (range: 0−27; M = 3.74; SD = 4.57). The rater also observed paramedian (range: 0−17; M = 1.71; SD = 2.91) as well as lateral (range: 0−13; M = 2.51; SD = 3.48) movements. Glottal closures were the rarest of the observed movements (range: 0−8; M = 0.35; SD = 1.15). Among the 90 videos taken during musical tasks, rater 1 noticed glottal closures in 18.9% of the cases, rater 2 in 21.1%. However, none of the participants showed vibrations of the vocal folds. (2) What were the differences between the observed movements regarding task and the grade of professionalism of singing? During both inner singing (RMc, LM, and IM) and inner speech (SR), laryngeal and partially pharyngeal movements were most notable. However, these movements also occurred AR. Clear differences between the groups of participants within each task could not be identified. Instead, the groups showed differences with respect to which anatomical structures moved the most. NS and LS displayed the most laryngeal movement. The most movement at the base of the tongue was recorded for PS while NS showed the most pharyngeal movements.
ARTICLE IN PRESS 4
Journal of Voice, Vol. &&, No. &&, 2019
TABLE 1. Number of Vocal Folds Movements: Lateral, Intermediary, Paramedian, Median
Lateral movements Intermediary movements Paramedian movements Median movements
N
Range
Sum
Mean
Standard Deviation
147 147 147 147
0−13 0−27 0−17 0−8
369 550 251 52
2.51 3.74 1.71 0.35
3.48 4.57 2.91 1.15
Figure 1 shows the frequency of movements of the arytenoid cartilages and the vocal folds in different tasks and group of subjects. In every group, laryngeal movements were less frequent during AR and SR. Both rater 1 and rater 2 observed this. Looking more closely at the group of subjects, PS showed the least laryngeal movements, NS (rater 1) and LS (rater 2) showed the most. Figure 2 shows the frequency of glottal closures in different tasks and group of subjects. Independently from the group of subjects glottal closures where seen more often during the musical tasks (IM, LM, and RM) from both rater 1 and rater 2. This can be seen also in Figure 3. While during rest the vocal folds are open, during RM without using phonation the vocal folds are closed with a small dorsal gap. Since there were no significant differences between the three musical tasks (IM, LM, and RM), these tasks were summarized for the following analysis (Figure 1). This way a comparison from the average movements during inner singing (IM, LM, and RM) and AR or SR was possible. As shown in Table 2 during the inner singing task, there were significantly more laryngeal movements as well as significantly more glottal closures than during the control tasks (AR, SR). This observation did not include the pharyngeal structures.
The study revealed a significant average difference in laryngeal movements between the scenarios inner singing and AR (rater 1: t28 = 4.94, P < 0.001; rater 2: t29 = 3.70, P < 0.01). On average, during inner singing, 2.28 (rater 1) and 2.56 (rater 2) more movements were observed in comparison to the situation AR (rater 1: CI = 95% [1.33−3.22]; rater 2: CI = 95% [1.14−3.97]) (Table 2). Also, the difference between inner singing and SR was significant. The average difference between these tasks showed 2.63 (rater 1) and 2.46 (rater 2) more movements during inner singing (rater 1: t29 = 5.43, P < 0.001; rater 2: t29 = 5.10, P < 0.001). On the other hand, the comparison of SR and AR yielded no significant results (Table A.2). Also, more glottal closures were observed during the musical tasks than were observed during the nonmusical tasks as seen in Figure 2. Comparing the average number of median vocal fold movements for inner singing and AR showed that there were more movements during inner singing for both raters. This observation was significant for rater 2 (t26 = 2.06, P < 0.05) while it was not for rater 1 (Table 2). In the SR task, there were also significantly fewer glottal closures on average than during inner singing (rater 1: t29 = 2.96, P < 0.01; rater 2: t29 = 2.75, P < 0.05) (Table A.2).
FIGURE 1. Frequency of movements of the arytenoid cartilage and the vocal folds in the three different groups (each N = 10) and settings. Abbreviations: AR, at rest; IM, imagining a melody; LM, listening to music; LS, lay singers; NS, nonsingers; PS, professional singers; RM, reading music; SR, silent reading.
ARTICLE IN PRESS Carmen Unterhofer, et al
Laryngeal and Pharyngeal Movements During Inner Singing
5
FIGURE 2. Frequency of glottal closures in the three different groups (each N = 10) and settings. Abbreviations: AR, at rest; IM, imagining a melody; LM, listening to music; LS, lay singers; NS, nonsingers; PS, professional singers; RM, reading music; SR, silent reading.
Rater 1 observed the most median vocal fold movements during the RM task while rater 2 noticed the most during the LM task. For both raters, the least glottal closures happened in the PS group, the most closures in the NS group. This lies in accordance with the finding that NS showed the most laryngeal movements while PS displayed the least. (3) Are the participants aware of their movements? Eighteen attendants reported to have felt movements in the larynx (three NS, nine PS, and six LS). None of these happened AR and only a few during SR (four PS). PS felt the most movements across all tasks except for LM. Five of 10 LS also gave account of feeling movements during that task (one NS, five LS, and four PS). NS was the group with the least subjectively felt movement. For example, during the RM task, eight out of 10 PS reported laryngeal activity, while only four LS and two NS reported the same.
DISCUSSION Our results are in agreement with some results previously published by other researchers. They inferred laryngeal activity during the imagination of music (cf.4,16−19), an inference we could prove endoscopically. However, we were also able to observe laryngeal and pharyngeal movement not only during musical tasks, but also AR, albeit to a significantly lesser extent. In this study, there were significantly more laryngeal movements during inner singing (RM, LM, IM) than during the control tasks (AR, SR). Because there were no significant differences between the three musical tasks, these tasks were summarized. There were also significantly more glottal closures during inner singing than AR. This is important if total vocal rest is prescribed by phoniatrists or ENT specialists. The specific consideration of median movements
FIGURE 3. Example of the larynx at rest (left) and the larynx during reading music (right) (Department of Otorhinolaryngology/ Phoniatrics, Klinikum rechts der Isar, Dr. S. Graf).
ARTICLE IN PRESS 6
Journal of Voice, Vol. &&, No. &&, 2019
TABLE 2. Paired Samples t test—Inner Singing vs At Rest Laryngeal Movements: Inner Singing vs at Rest Paired Differences 95% CI
Rater 1 Rater 2
Laryngeal movements: inner singing—at rest Laryngeal movements: inner singing—at rest
M
SD
SEM
Lower
Upper
t
df
2.28
2.48
0.46
1.33
3.22
4.94***
28
2.56
3.79
0.69
1.14
3.97
3.70**
29
Glottal Closure: Inner Singing vs at Rest Paired Differences 95% CI
Rater 1 Rater 2
glottal closure: inner singing—at rest glottal closure: inner singing—at rest
M
SD
SEM
Lower
Upper
t
df
0.32
1.09
0.20
¡0.09
0.74
1.60 n.s.
28
0.34
0.86
0.17
0.00
0.68
2.06*
26
* P < 0.05. ** P < 0.01. *** P < 0.001. Abbreviations: M, mean; SD, standard deviation; SEM, Standard Error of Mean; CI, confidence interval; t, t-value; df, degrees of freedom; n.s., non-significant.
of the vocal folds is necessary because frequent glottal closures could influence wound healing. During our experiment, no movement of the pharynx was observed. Movements at the base of the tongue and the posterior and lateral pharynx wall did not show an increase in activity during inner singing, and moved only in a random manner. This might be due to the fact that pharyngeal structures are less involved in the process of phonation. It has been suggested that only well-trained singers are capable of inner singing, and that during inner singing their larynx move more than the laryngeal structures of the NS (cf.16,18,20). This hypothesis cannot be confirmed. On the contrary, NS showed tendentially the most laryngeal movement during all tasks, and PS showed the least. This might be explained by the ability of trained singers (PS and LS) to suppress or activate movements more consciously, having developed a greater awareness of the vocal tract. This possibility must be considered while interpreting the results. Interestingly, the participants’ subjective reports did not match the results the raters produced. According to the raters, NS displayed the most laryngeal activity while PS showed the least. PS was the group with the most subjectively felt movement, and only a few NS felt laryngeal activity during the musical tasks. Movements happened during all tasks for all participants. This discrepancy might be due to not only by the awareness of PS of the vocal tract (they feel more, and move less), but also to the expectations of the individual participants. With their knowledge of the aim of the study, none of the participants assumed to feel laryngeal
or pharyngeal movement AR, and yet, the rater observed movement also during this task. It remains unclear if there is a correlation between the subjective movements and objectively observed laryngeal movements. Furthermore, outliers within each group should be considered. In general, our study suggested that PS moved their glottises less during inner singing than did LS, but within all three groups there were individuals who did not fit this trend. It should not be assumed, for instance, that any given PS can safely listen to music without involuntary glottis movement, and this indicates the importance of individualized recommendations concerning vocal rest. We investigated endoscopically the influence of inner singing on laryngeal and pharyngeal movement and were able to show involuntary activity in the vocal tract. One might call this phenomenon senso-motoric inner singing.
Methodological reflections and view In addition to transnasal videoendoscopical stroboscopy, intramuscular EMG could be used to obtain more specific and clear information. It has already been shown that surface EMG detects no significant differences between breathing and inner singing as previously published by Kleber.11 While intramuscular EMG could yield more helpful results, this procedure is difficult to support ethically because needle electrodes would have to be placed into the muscles around the larynx. This technique will lead to at least some
ARTICLE IN PRESS Carmen Unterhofer, et al
7
Laryngeal and Pharyngeal Movements During Inner Singing
discomfort for the participants and incur a higher risk of unwanted side effects due to the invasiveness of the method. Another possibility for future research could be to modify or extend the tasks, especially with regard to wound healing, since the effect on the healing process remains uncertain. In this scenario, a cohort of patients needing medical vocal rest could be assembled and the observed rate of healing correlated with the participants’ grade of unconscious laryngeal activity. It could be assumed that inner singing has no effect on wound healing because we observed no vibrations of the vocal folds, only glottal closures which also occur during swallowing and coughing. To our knowledge, there exists no literature on the fact, if swallowing affects the wound healing process. It could be assumed that during inner singing there is a higher muscles tension in the entire laryngeal area besides the glottal closures that might occur. This fact could influence the wound healing. Yet, the exact effect is not clear. Furthermore, a clarification of inner singing is important. Doctors, specialists, speech therapists, and vocal pedagogues should be aware of inner singing, because at the moment the recommendations are diverse. We are left with the need for a clear definition of vocal rest. Vocal specialists should make clear to their patients that listening to singing, studying singing, or even imagining singing might engage their vocal mechanisms and impede their healing. Our findings suggested that PS or LS were able to suppress or control movements consciously and thus show less movement during inner singing. However, statistical outliers existed in each group; there were NS who barely moved during all tasks, and PS who moved a great deal. This complexity should be considered in further studies and in clinical settings. During clinical diagnostics, the larynx should be examined in various situations (phonation, AR, imagined singing) to give corresponding individual recommendations concerning vocal hygiene. CONCLUSION In this study, we investigated 10 vocally healthy NS, 10 LS, and 10 PS transnasally with a flexible videoendoscope. The aim of this study was to find out if pharyngeal and laryngeal
movements could be observed during inner singing, and if so, to describe the nature of those movements. During two nonmusical tasks (AR, SR) and three musical tasks (IM, LM, RM) the participants were examined, and two specialists analyzed the videos qualitatively and quantitatively. During inner singing (musical tasks), there were significantly more laryngeal movements, as well as significantly more glottal closures, observed in our participants than while AR or during SR. These movements could possibly impair healing processes. Differences in the level of activity were observed based on the grade of professionalism of singing, but outliers existed in all three groups. Pharyngeal movements where observed only in a random manner. Vocal fold vibrations were not observed at all. In summary, based on our findings we recommend that music imagination (LM or practicing musical scores silently) should be avoided when vocal rest is prescribed. That said, the extent of vocal rest could be adapted based on the individual patient's awareness of movement and level of activity of the vocal tract. These would have to be examined endoscopically. Further studies are needed concerning the effects of inner singing on the healing process when vocal rest has been prescribed for laryngeal disease. ACKNOWLEDGMENTS We are very grateful for the statistical help and advisement from Thomas Maierhofer with the Statistisches Beratungslabor LMU in Munich (StaBLab). This study was supported by the Lehre@Forschung LMU. We are very thankful for the independent financial support. SUPPLEMENTARY MATERIALS Supplementary material associated with this article can be found in the online version at https://doi.org/10.1016/j. jvoice.2019.02.011. APPENDIX Table A.1, Table A.2.
TABLE A.1. Intrarater and Interrater Reliability
Base of the tongue (never, rarely, sometimes, often) Pharyngeal wall (never, rarely, sometimes, often) Vocal folds and the arytenoid cartilage (never, rarely, sometimes, often) Median movements of the vocal folds (yes, no) *** P < 0.001.
Intrarater Reliability Kappa
Inter-rater Reliability Kappa
0.61***
0.34***
0.16
0.20***
0.59***
0.47***
0.79***
0.52***
ARTICLE IN PRESS 8
Journal of Voice, Vol. &&, No. &&, 2019
TABLE A.2. Paired Samples t test—Inner Singing vs Silent Reading Laryngeal Movements: Inner Singing vs Silent Reading Paired Differences 95% CI
Rater 1 Rater 2
Laryngeal movements: inner singing—silent reading Laryngeal movements: inner singing—silent reading
M
SD
SEM
Lower
Upper
t
df
2.63
2.66
0.49
1.64
3.63
5.43***
29
2.46
2.64
0.48
1.47
3.44
5.10***
29
Glottal Closure: Inner Singing vs Silent Reading Paired Differences 95% CI
Rater 1 Rater 2
Glottal closure: inner singing—silent reading Glottal closure: inner singing—silent reading
M
SD
SEM
Lower
Upper
t
df
0.54
1.001
0.18
0.17
0.92
2.96**
29
0.41
0.81
0.15
0.10
0.71
2.75*
29
* P < 0.05. ** P < 0.01. *** P < 0.001. Abbreviations: CI, confidence interval; df, degrees of freedom; M, mean; SD, standard deviation; SEM, Standard Error of Mean; t, t value.
FIGURE A.1. Answer sheet (example). REFERENCES 1. D€ ohne E. Bedeutet Fl€ ustern Stimmruhe oder Stimmschonung? J Mol Med. 1942;21:692–693. 2. Behrman A, Sulica L. Voice rest after microlaryngoscopy: current opinion and practice. Laryngoscope. 2003;113:2182–2186. 3. Ishikawa K, Thibeault S. Voice rest versus exercise: a review of the literature. J Voice. 2010;24:379–387. 4. Richter B, Sandel M. Gesunderhaltung, hygiene, pr€avention. In: Richter B, Echternach M, eds. Die Stimme: Grundlagen, k€ unstlerische Praxis, Gesunderhaltung. 2nd ed.Henschel: Leipzig; 2014:197–218. 5. Fedrowitz H. Stimmhygiene—ein wichtiger Baustein zur Pr€avention von Berufsdysphonien bei Lehrern. Praxis Sprache. 2016;61: 124–128. 6. Behlau M, Oliveira G. Vocal hygiene for the voice professional. Curr Opin Otolaryngol Head Neck Surg. 2009;17:149–154.
7. Sataloff RT. Voice Rest. In: Sataloff RT, ed. Professional Voice: The Science and Art of Clinical Care. New York: Raven Press; 1991. 8. Koufman JA, Blalock DP. Is voice rest never indicated? J Voice. 1989;3:87–91. 9. Kiagiadaki D, Remacle M, Lawson G, Bachy V, van der Vorst S. The effect of voice rest on the outcome of phonosurgery for benign laryngeal lesions: preliminary results of a prospective randomized study. Ann Otol Rhinol Laryngol. 2015;124:407–412. 10. Maes P-J, Leman M, Palmer C, Wanderley MM. Action-based effects on music perception. Front Psychol. 2014;4:1–14. 11. Kleber B, Birbaumer N, Veit R, Trevorrow T, Lotze M. Overt and imagined singing of an Italian aria. NeuroImage. 2007;36:889–900. 12. Zatorre RJ, Halpern AR, Perry DW, Meyer E, Evans AC. Hearing in the mind's ear: a PET investigation of musical imagery and perception. J Cognit Neurosci. 1996;8:29–46.
ARTICLE IN PRESS Carmen Unterhofer, et al
Laryngeal and Pharyngeal Movements During Inner Singing
13. Zatorre RJ. Brain imaging studies of musical perception and musical imagery. J New Music Res. 1999;28:229–236. 14. Aleman A, Nieuwenstein MR, B€ ocker KB, de Haan EHF. Music training and mental imagery ability. Neuropsychologia. 2000;38:1664– 1668. 15. Herholz SC, Halpern AR, Zatorre RJ. Neuronal correlates of perception, imagery, and memory for familiar tunes. J Cognit Neurosci. 2012;24:1382–1397. 16. Habermann G. Stimme und Mensch: Beobachtungen und Betrachtungen. Heidelberg: Median-Verl; 1996. 17. N. Saintilan, The use of imagery during the performance of memorized music, psychomusicology: music, mind, and brain 24 (2014) 309−315. 18. Frank F. Bemerkung zum inneren Singen. Sprache Stimme Geh€or. 1989;13:8–10. 19. Nawka T, Wirth G, Anders L-C. Stimmst€orungen: F€ur Ärzte, Logop€ aden, Sprachheilp€ adagogen und Sprechwissenschaftler; mit 30 Tabellen. 5th ed. K€ oln: Dt. Ärzte-Verl.; 2008. 20. Nadolecƶny M. Untersuchungen €uber den Kunstgesang: I. Atem- und Kehlkopfbewegungen. Berlin, Heidelberg: Springer Berlin Heidelberg; 1923. 21. Deutsche Gesellschaft f€ ur Phoniatrie und P€adaudiologie e.V., Voice Handicap Index, deutsche Fassung, available at http://www.dgpp.de/ Profi/Sources/VHI-Bogen-Konsensfassung_(C)DGPP%202003-2.pdf (accessed on February 16, 2018). 22. A. Lorenz, Singing Voice Handicap Index, deutsche Fassung, available at http://www.dgpp.de/cms/media/download_gallery/SVHI-Testanleitung. pdf (accessed on February 16, 2018).
9
23. Brockmann-Bauser M, Bohlender JE. Praktische Stimmdiagnostik: Theoretischer und praktischer Leitfaden. 1st ed. Stuttgart: Georg Thieme Verlag KG; 2014. 24. Fleischer S, Hess M. Systematischer Untersuchungsablauf bei der Videolaryngoskopie und Stroboskopie. HNO. 2003;51:160–167. 25. Schr€ oter-Morasch H. Klinische Untersuchung des Oropharynx und videoendoskopische Untersuchung der Schluckfunktion. In: Bartolome G, Schr€ oter-Morasch H, Buchholz D, Feussner H, Graf S, Holzapfel K, Neumann S, Pehl C, Prosiegel M, eds. Schluckst€orungen: Diagnostik und Rehabilitation; [mit dem Plus im Web; Zugangscode im Buch]. 5th ed. M€ unchen: Elsevier Urban & Fischer; 2014:167–204. 26. Schr€ oter-Morasch H, Graf S. Dysphagiediagnostik durch den HNOArzt. HNO. 2014;62:324–334. 27. Nawka T. Voice Handicap Index, deutsche Fassung, available at http://www.dgpp.de/Profi/Sources/vhi-dt_2006.pdf (accessed on February 16, 2018). 28. Lorenz A, Gonnermann U, Nawka T, Graduierung der deutschen Fassung des Singing Voice Handicap Index (SVHI), in: Deutsche Gesellschaft f€ ur Phoniatrie und P€adaudiologie e.V. (DGPP) (Eds.). 31. Wissenschaftliche Jahrestagung der Deutschen Gesellschaft f€ ur Phoniatrie und P€adaudiologie (DGPP) zusammen mit dem 5. P€adakustiker-Symposium der Akademie f€ ur H€ orger€ate-Akustik, L€ ubeck, 2014, available at http://www.egms.de/static/en/meetings/ dgpp2014/14dgpp52.shtml (accessed on February 16, 2018).