Incidence of Supraglottic Activity in Males and Females: A Preliminary Report *Sheila V. Stager, †Rebecca Neubert, *Susan Miller, ‡Joan Roddy Regnell, and *Steven A. Bielamowicz Washington, D.C. and Charlottesville, Virginia
Summary: Supraglottic activity was rated from flexible endoscopic video recordings of subjects with normal laryngeal structure and function as they sustained vowels and repeated syllables and sentences. Judges rated these recordings for false vocal fold (FVF) adduction and anterior-to-posterior (A– P) compression at the initiation of the speech task, throughout the whole speech task (static supraglottic activity), and as brief individual adductions within a speech task (dynamic supraglottic activity). Significant differences in A–P ( p ⬍ 0.0003) and FVF (p ⬍ 0.0000001) compression were found between tasks. Dynamic FVF activity was associated with glottal stops. Static A–P and FVF activities were present in males significantly more ( p ⬍ 0.0001) than females. FVF activity associated with speech initiation was found in females significantly more ( p ⫽ 0.0256) than males. Supraglottic activity plays a role in normal speech production, and should not necessarily be considered suggestive of a voice use pattern with excessive muscle tension. Key Words: Supraglottic—Voice—Larynx—Flexible fiberoptic.
INTRODUCTION The term supraglottic activity describes movement of the structures immediately above the level of the true vocal folds. When present, it is typically associated with temporary obstruction of part or all of the view of the true vocal folds from a trans– oral or trans–nasal approach. Supraglottic activity during voice production may be separated into two types. Anterior-to-posterior supraglottic activity occurs when the arytenoid cartilages appear to approach the petiole of the epiglottis (A–P compression). Medial supraglottic activity is characterized by adduction of the false vocal folds (FVF compression). These two types of supraglottic activities may be observed individually, or in combination within one person.
Accepted for publication September 16, 2002. Presented at The Voice Foundation’s 28th Annual symposium: Care of the Professional Voice; Philadelphia, Pennsylvania, June 2–6, 1999. From *The Voice Treatment Center, The George Washington University, Washington, D.C.; †The University of Virginia Medical Center, Charlottesville, Virginia; and the ‡Department of Speech and Hearing, The George Washington University, Washington, D.C. Address correspondence and reprint requests to Sheila V. Stager, The George Washington University Voice Treatment Center, 2150 Pennsylvania Avenue N.W., Suite 6-301, Washington, D.C. 20037. E-mail:
[email protected] Journal of Voice, Vol. 17, No. 3, pp. 395–402 쑕 2003 The Voice Foundation 0892-1997/2003 $30.00⫹0 doi:10.1067/S0892-1997(03)00034-1
395
396
SUPRAGLOTTIC ACTIVITY IN MALES AND FEMALES
Stager, Bielamowicz, Gupta, Regnell, and Barkmeier reported that A-P compression is best considered a static activity, while FVF compression has both static and dynamic components.1 The static component refers to a position or configuration taken by the supraglottic structures during connected speech. Supraglottic structures move to a position at the beginning of the speech task, maintain the position throughout the speech task, and then move back to the rest position when the speech task is completed. The dynamic component refers to discrete, quick adductory movements of the false vocal folds (FVF) that can occur several times during connected speech. Traditionally, supraglottic activity has been considered to be either a compensatory behavior when the true vocal folds do not completely close2–4 or a behavior associated with a vocal use pattern of excessive muscular tension.5,6 Two previous studies have attempted to assess the structure and function of supraglottic structures in normal larynges using the flexible fiberoptic examination.7,8 Both studies reported on the variability in structure and movement in male and female speakers with no vocal training. Casper et al reported movement differences in the FVF, but not in the A–P dimension.7 Pemberton et al reported constriction in the A–P laryngeal dimension in 24% of their subjects, and slight compression of the FVF in 16% of their subjects.8 Although the tasks involved in these studies included quiet breathing, deep inspiration, sustaining a vowel on exhalation and inhalation, gliding from high to low, whistling, counting to 10 on inspiration and repeating sentences, both studies reported that supraglottic activity could be present in individuals with a normal larynx. Stager et al investigated the incidence of supraglottic activity across different speech tasks in three populations: (1) a control group of individuals with normal laryngeal mucosa, normal voice quality, and no voice complaints; (2) subjects with vocal fold nodules; and (3) subjects with complaints of dysphonia without visible vocal fold lesions, glottal incompetence, or impairment of arytenoid cartilage motion.1 The control group was found to exhibit the lowest incidence of static A–P and FVF compression compared to the voice-disordered groups. The Journal of Voice, Vol. 17, No. 3, 2003
small incidence of supraglottic activity in these subjects may have been caused by laryngopharyngeal reflux induced symptoms, as these “controls” complained of sensory changes within their pharynx. However, they were included because no previous data had confirmed the influence of pharyngeal symptoms of foreign body sensation or excess mucus on the presence of supraglottic activity. These findings should be confirmed for individuals with no complaints of sensory changes within their pharynx, and who have normal laryngeal mucosa, normal voice quality, and no voice complaints. Stager et al found that the control group and both voice-disordered groups exhibited FVF compression in speech tasks that included glottal stops, suggesting that supraglottic activity might assist laryngeal articulation and speech intelligibility.1 The current study should be considered as a preliminary study in which we examined the association of supraglottic activity with other linguistic contexts, and with respect to possible differences in incidence between males and females. The purpose of this study was a preliminary examination of the effects of the specific linguistic context and gender on the incidence of each type of supraglottic activity, A–P and FVF compression, in male and female subjects with no sensory pharyngeal symptoms, no complaints of dysphonia, laryngeal pain, vocal fatigue, or effortful voice use, voice quality judged as being within normal limits, and normal vocal fold mucosa and laryngeal structure. The presence of supraglottic activity was assessed during initiation of the speech task, throughout the whole speech task (static), and within a speech task (dynamic).
METHOD Subjects Thirty-seven subjects (23 females, 14 males), ages 18 to 71 years participated in this study. All subjects signed an informed consent form approved by the George Washington University Institutional Review Board. The mean subject age was 42 years for females and 37 years for males. For initial inclusion into the study, subjects were interviewed by a speech–language pathologist (SLP) and met the
SHEILA V. STAGER ET AL following criteria: native speaker of American English, no voice complaints or history of a voice disorder, no history of asthma or other respiratory disease, no acute respiratory or laryngeal infections, no complaints of heartburn or acid indigestion (symptoms of gastroesophageal reflux), no complaints of pharyngeal pain, excessive pharyngeal mucus or throat clearing (symptoms of laryngopharyngeal reflux), no smoking for the previous 5 years, and no professional voice training. Spontaneous conversation was recorded from each subject. Approximately 15 seconds of the recorded conversation was randomized onto a single audiotape. Two experienced SLPs rated overall voice quality on a scale from 0 (normal) to 4 (severely disordered voice). Subjects were excluded from the study if they received a rating of 1 or more by at least one judge. Based upon these criteria, one subject was eliminated from the study. The remaining subjects underwent a 70-⬚ videostroboscopic examination. During this examination, subjects were asked to produce a sustained vowel /i/ at three different pitches, (comfortable pitch, low pitch and high pitch), and comfortable pitch at loud and soft intensities. A board-certified otolaryngologist and certified SLP with experience interpreting endoscopic images rated these examinations for the presence of vocal fold lesions. Subjects were excluded if an hourglass vocal fold configuration was observed during phonation at any pitch or if a laryngeal structural abnormality was present. Based upon the laryngeal examination criteria, four additional subjects were eliminated from the study. The final number of subjects meeting all criteria for inclusion in this study was 32 individuals (21 females, 11 males). Procedures Subjects were enrolled in the current study and underwent a separate voice recording and a videostroboscopic examination using both rigid and flexible scopes. Voice samples were recorded using a Tascam DA-88 digital tape recorder (Montebello, CA) with an AKG 420 head-mounted microphone (Vienna, Austria) placed 1 inch from the subject’s mouth. The speech sample consisted of the following tasks: sustaining the vowels /i/ and /a/, and the consonants /s/ and /z/; reading the first paragraph of the
397
California passage; conversing for approximately 15 seconds on a topic of their choice; and performing pitch glides from high to low and low to high. The speech sample was recorded following the endoscopic examinations, so that the normalcy of laryngeal structure and function could be documented first. The Kay Elemetrics Stroboscopy unit (Model RLS 9100; Kay Elemetrics, Lincoln Park, NJ) was used for both rigid and flexible examinations. Generally, the rigid videostroboscopy was performed first, followed by trans-nasal flexible endoscopy. Topical oral 10% lidocaine was sprayed on the oropharynx prior to the rigid examination if the subject reported having a strong gag response. During the trans-nasal fiberoptic evaluation, a mixture of 4% lidocaine and oxymetazolene was administered into the nasal passage. A Pentax ENF-P3 (Hamburg, Germany) flexible scope was used for all trans–nasal flexible laryngeal examinations. The flexible scope was introduced into the nostril with the best airway and the scope was passed into the hypopharynx so that the tip of the epiglottis was not visible during the examination. After the flexible scope was situated so that the tip of the epiglottis was not visible, the laryngologist rotated the patient’s head until the image was centered, thus minimizing the skew introduced by the off-center trans–nasal introduction of the scope. All laryngeal examinations were performed by one of the authors (SAB), a board certified otolaryngologist with more than 10 years experience using this technique for laryngeal examination. Only well-centered images were selected for this study. The reliability of measures throughout a flexible examination has been established.9 During the flexible examination, speech and nonspeech tasks were recorded onto the videotape of the Kay Elemetrics video system: (1) a sniff gesture; (2) pitch glides from low to high and from high to low; (3) sustained /i/ at a comfortable pitch and loudness; (4) repetition of the syllables /i/, /si/, /mi/ and /i-si/; and (5) repetition of the sentences, “we eat eels every day,” “she speaks pleasingly,” and “Peter will keep at the peak.” Data analysis Using Kay Elemetrics Visipitch and MDVP programs, the following measures were obtained from Journal of Voice, Vol. 17, No. 3, 2003
398
SUPRAGLOTTIC ACTIVITY IN MALES AND FEMALES TABLE 1. Means and ranges for the acoustic measures for both male and female subjects in the current study as well as norms from Kent (1994) Males
Fundamental Frequency (Hz) SD of F0 (Hz) Range (semitones) Range (Hz) MPT (sec) Sustained /s/ (sec)
Norms
Mean
Range
Norms
Mean
Range
100 – 37 490 21 18
92 12 33 439 22 20
79–113 5–21 27–40 288–687 12–34 8–45
221 – 34 760 21 18
176 29 38 775 20 21
134–211 16–60 20–50 283–1363 10–32 8–34
the acoustic recordings sampled at 25,000 Hz: average speaking fundamental frequency of the voiced phonemes within the reading passage consisting of the first paragraph from the “California” passage; average speaking fundamental frequency of the voiced phonemes within a 15 second sample of spontaneous conversation; standard deviation of fundamental frequency during reading; standard deviation of fundamental frequency during conversation; pitch range in Hz for the pitch glides; jitter, shimmer and noise to harmonic ratio for the middle one second of an approximately 3 to 4 second sustained vowel /a/; and maximum phonation time (MPT). The trans–nasal fiberoptic examinations were rated by two SLPs. Prior to performing the ratings within this study, the two SLPs together rated a series of laryngeal examinations and discussed both what was agreed upon and what was disagreed upon in rating the presence of various types of supraglottic activity. The number of laryngeal examinations that were discussed to reach acceptable agreement between raters (ie, 90% regardless of speech task) was 10. Each of these training laryngeal examinations included all tasks included in the present study. Each rater was allowed to review the video segments as often as necessary to arrive at a rating, as is performed in the clinical setting. Intra- and interjudge reliability was computed. The rating form listed four choices for both FVF compression and A–P compression. Raters were asked to select as many types of supraglottic activity as they observed during each speech task. The following categories of supraglottic activity were used: Journal of Voice, Vol. 17, No. 3, 2003
Females
1. No visual obstruction of the true vocal folds by false vocal folds during the speech task (NO FVF) 2. Visual obstruction of the true vocal folds by the false vocal folds at initiation of the speech task (INIT FVF) 3. Visual obstruction of the true vocal folds by the false vocal folds during the entire speech task (STATIC FVF) 4. Visual obstruction of the true vocal folds by brief single adductions of the false vocal folds within the speech task (DYNAMIC FVF) 5. No visual obstruction of the true vocal folds by apposition of the arytenoid cartilages and petiole of the epiglottis during the speech task (NO A–P) 6. Visual obstruction of the true vocal folds by arytenoid-to-petiole apposition at initiation of speech task (INIT A–P) 7. Visual obstruction of the true vocal folds by arytenoid-to-petiole apposition during the entire speech task (STATIC A–P) 8. Visual obstruction of the true vocal folds by brief single appositions of the arytenoid and petiole within the speech task (DYNAMIC A–P) RESULTS The acoustic measures of the males and females in the study were within normal limits according to previous studies (Table 1).10 Glottic closure was assessed using rigid endoscopy during sustained /i/ at comfortable pitch and
SHEILA V. STAGER ET AL
399
TABLE 2. Percent occurrences of supraglottic activity by task
PROLONGED /i/ REPEATED /i/ REPEATED /mi/ REPEATED /si/ REPEATED /i-si/ EELS SHE PETER
PROLONGED /i/ REPEATED /i/ REPEATED /mi/ REPEATED /si/ REPEATED /i-si/ EELS SHE PETER
N
NO A–P
INIT A–P
STATIC A–P
DYNAMIC A–P
31 32 31 31 32 32 32 32
61% 56% 74% 61% 50% 56% 59% 56%
3% 0% 0% 0% 0% 0% 0% 0%
36% 44% 26% 32% 13% 22% 28% 31%
0% 0% 0% 7% 38% 25% 16% 13%
N
NO FVF
INIT FVF
STATIC FVF
DYNAMIC FVF
30 32 30 30 32 31 32 32
45% 3% 83% 83% 6% 7% 81% 84%
52% 38% 0% 0% 13% 0% 0% 0%
21% 10% 17% 17% 9% 13% 16% 16%
0% 97% 0% 0% 91% 94% 3% 0%
More than 1 type of supraglottic activity could be present for each task; thus, percentages do not always sum to 100%. N ⫽ total number of occurrences; A–P ⫽ anterior–posterior compression; FVF ⫽ false vocal fold compression; INIT ⫽ speech initiation
loudness. The frame representing the most closed portion of the glottal cycle was selected from the rigid videostroboscopy tape and was rated using the Hirano and Bless classification system.11 Glottic closure could not be assessed for 7 females and 3 males because of an obstructed view of the vocal folds during comfortable pitch and loudness by the free edge of the epiglottis. For the remaining subjects, complete glottic closure was identified for 50% (4/8) of males, but only 7% (1/14) of females. A posterior glottal gap involving only the cartilaginous larynx was identified in 38% (3/8) of males and 85% (12/14) of females. One male and one female demonstrated incomplete glottic closure of the membranous vocal folds at comfortable pitch and loudness. Overall exact agreement between the two raters for the type of supraglottic activity observed was 74% compared to 80% in our previous supraglottic rating study.1 Disagreements primarily arose in recognizing more than one type of supraglottic activity within a speech task (ie, if supraglottic activity was present at speech initiation as well as during the speech task), rather than whether any type of supraglottic activity was present for a speech task. Twenty percent of the original ratings were remeasured by both judges
to assess within-judge reliability. Within-judge reliability was 83% for one rater, and 81% for the other. Incidence of supraglottic activity By task Chi-square tests revealed significant differences between tasks for A–P activity ( p ⬍ 0.0003) and FVF activity ( p ⬍ 0.0000001). Table 2 summarizes the percent occurrence of each type of supraglottic activity by task. The linguistic components of speech were evaluated as a possible explanation for the patterns of supraglottic activity seen in Table 2. Ninety-nine percent of the instances of dynamic FVF activity occurred during repeated /i/, repeated /i-si/ and the sentence “we eat eels every day.” An average of 94% of individuals demonstrated dynamic FVF activity. The linguistic component shared by these speech tasks was the presence of glottal stops. All instances of FVF activity at speech initiation occurred during prolonged /i/, repeated /i/ and repeated /i-si/ tasks. Thirty-four percent of individuals demonstrated this activity at speech initiation. Speech initiation with a vowel was the linguistic component shared by these speech tasks. All instances of dynamic A–P activity occurred during Journal of Voice, Vol. 17, No. 3, 2003
400
SUPRAGLOTTIC ACTIVITY IN MALES AND FEMALES TABLE 3. Percent occurrences of supraglottic activity for each task by gender N
PROLONGED /i/ REPEATED /i/ REPEATED /mi/ REPEATED /si/ REPEATED /i-si/ EELS SHE PETER
NO A–P
DYNAMIC A–P
M
F
M
F
M
F
M
F
11 11 11 10 11 11 11 11
20 21 20 21 21 21 21 21
46% 36% 55% 40% 10% 36% 27% 27%
70% 67% 85% 71% 57% 67% 76% 71%
55% 64% 46% 50% 50% 46% 64% 64%
25% 33% 15% 24% 5% 10% 10% 14%
0% 0% 0% 10% 40% 27% 18% 9%
0% 0% 0% 5% 38% 24% 14% 14%
M
F
M
F
M
F
M
F
M
F
11 11 11 10 11 11 11 11
19 21 19 20 21 20 21 21
55% 0% 73% 70% 18% 9% 64% 73%
37% 5% 90% 90% 0% 5% 91% 91%
27% 27% 0% 0% 0% 0% 0% 0%
63% 43% 0% 0% 19% 0% 0% 0%
36% 18% 27% 30% 27% 18% 27% 27%
11% 5% 11% 10% 5% 10% 10% 10%
0% 100% 0% 0% 73% 91% 9% 0%
0% 95% 0% 0% 100% 95% 0% 0%
N
PROLONGED /i/ REPEATED /i/ REPEATED /mi/ REPEATED /si/ REPEATED /i-si/ EELS SHE PETER
STATIC A–P
NO FVF
INIT FVF
STATIC FVF
DYNAMIC FVF
N ⫽ total number of occurrences; A–P ⫽ anterior–posterior compression; FVF ⫽ false vocal fold compression; INIT ⫽ speech initiation; M ⫽ male; F ⫽ female
the three sentences and the repeated syllables /si/ and /i-si/. Nineteen percent of individuals demonstrated this dynamic activity. The linguistic component shared by these speech tasks was the presence of consonants requiring the placement of the tongue in the palatal or velar places of articulation. By gender Comparing the total number of instances of each type of supraglottic activity between males and females, Chi-square statistics demonstrated significant differences in the number of occurrences of types of A–P supraglottic activity (p ⫽ 0.00000004) and FVF supraglottic activity (p ⫽ 0.001). Table 3 summarizes the percent occurrence of each type of supraglottic activity for males and females. The difference in A–P static activity between males and females was statistically significant (z ⫽ 5.64, p ⬍ 0.0001),12 with males demonstrating more A–P static activity compared to females (55% vs. 17%). The difference in FVF static activity between males and females was also statistically significant Journal of Voice, Vol. 17, No. 3, 2003
(z ⫽ 3.69, p ⬍ 0.0001),12 with males demonstrating more FVF static activity compared to females (26% vs. 9%). The difference in FVF activity at speech initiation between males and females was also statistically significant (z ⫽ 1.95, p ⫽ 0.0256),12 with males demonstrating less FVF activity (27% vs 42%). By individual If static supraglottic activity is a vocal habit, as has been suggested for individuals whose voice use pattern involves excessive muscle tension, then it should be present for all speech tasks within a speaker. When all speech tasks were considered, 10% (2/ 21) of females demonstrated bilateral static FVF activity, 27% (3/11) of males demonstrated unilateral static FVF activity and 55% (6/11) of males and 10% (2/11) of females demonstrated static A– P activity. DISCUSSION Supraglottic activity was present in individuals with normal vocal fold mucosa, normal laryngeal
SHEILA V. STAGER ET AL structure and perceptually normal voice quality, and its incidence was similar to the finding of Pemberton et al.8 Every individual and every speech task demonstrated some type of supraglottic activity. Repetition of /mi/ revealed the smallest incidence of both FVF and A–P activities. Differences in occurrence of supraglottic activity were dependent upon the individual and the specific linguistic context of the speech task. In general, differences in static supraglottic activity depended upon the speaker, and differences in dynamic supraglottic activity depended upon linguistic context. Significant differences were found between genders for the incidence of static supraglottic activity, and between speech tasks for the incidence of dynamic supraglottic activity. The occurrence of static supraglottic activity was dependent upon gender. Static supraglottic activity is thought to be associated with a voice use pattern with excessive muscle tension. Females are more likely to develop voice use patterns with excessive muscle tension such as nodules or cysts than males.13 However, in this study, the incidence of static supraglottic activity was greater in males than females. This raises questions about the relationship between static supraglottic activity and development of voice use patterns with excessive muscle tension. It should be remembered that in this study, only 11 male subjects participated. Similar findings using a larger group of males and female subjects over a longer period of time would clarify this relationship. The occurrence of dynamic supraglottic activity was dependent upon specific linguistic contexts. Gender did not appear to affect dynamic supraglottic activity. Dynamic A–P compression was found during tasks that included consonants with palatal and velar places of articulation and was not consistently present. Dynamic FVF compression occurred in speech tasks that included glottal stops and was consistently present in all speakers in the current study. A relationship may also exist for FVF compression with speech initiation for speech tasks beginning with vowel sounds. Many studies have examined the onset of words beginning with vowel sounds. Three major types of voice onsets have been reported—hard, breathy, and soft.14 Hard vocal onsets involve occlusion of airflow prior to voicing.14 During connected speech, glottal stops are
401
usually inserted between a final vowel sound in a word and the next word when it also begins with a vowel, such as “we eat.”15 In addition, glottal stops act as an allophone for the final ‘t sound, especially when followed by a word with an initial vowel sound, such as “eat eels.”15 Physiologically, glottal stops are also characterized as periods of no vocal fold vibration or airflow.16 Thus, the cessation of vocal fold vibration and absent airflow characterize both glottal stops and hard onsets. The adduction of the FVF may be a strategy to allow for quick dampening of true vocal fold vibration as associated with hard onset and glottal stops. In conclusion, this preliminary study demonstrated that dynamic supraglottic activity was present in the normal, healthy larynx, and is not suggestive of a pattern of voice use involving excessive muscle tension. However, the occurrence of static supraglottic activity may be more suggestive of patterns of voice use with excessive muscle tension, especially in females. REFERENCES 1. Stager SV, Bielamowicz S, Gupta A, Regnell JR, Barkmeier J. Supraglottic activity: evidence of vocal hyperfunction or laryngeal articulation? J Speech-Lang Hear Res. 2000;43(1):229–238. 2. D’Antonio LL, Wigley TL, Zimmerman GJ. Quantitative measures of laryngeal function following Teflon injection or thyroplasty type I. Laryngoscope. 1995;105:256–262. 3. Hanson DG, Gerratt BR, Ward PH. Cinegraphic observations of laryngeal function in Parkinson’s disease. Laryngoscope. 1984;94:348–353. 4. Smith M, Ramig L, Dromey C, Perez K, Samandari R. Intensive voice treatment in Parkinson Disease: laryngostroboscopic findings. J Voice. 1995;9:453–459. 5. Koufman JA, Blalock PD. Functional voice disorders. Otolaryngol Clin N Am. 1991;24:1059–1073. 6. Morrison MD, Rammage LA. Muscle misuse voice disorders: description and classification. Acta Otolaryngologica (Stockholm). 1993;113:428–434. 7. Casper JK, Brewer DW, Colton RH. Variations in normal human laryngeal anatomy and physiology as viewed fiberscopically. J Voice. 1987;1:180–185. 8. Pemberton C, Russell A, Priestley J, Havas T, Hooper J, Clark P. Characteristics of normal larynges under flexible fiberscopic and stroboscopic examination: an Australian perspective. J Voice. 1993;7:382–389. 9. Stager SV, Bielamowicz SA, Regnell JR, Marullo S, Gupta A, Barkmeier J. Quantification of static and dynamic supraglottic activity. J Speech, Lang Hear Res. 2001;44:1245– 1256. Journal of Voice, Vol. 17, No. 3, 2003
402
SUPRAGLOTTIC ACTIVITY IN MALES AND FEMALES
10. Kent R. Reference manual for communicative sciences and disorders. Austin: Pro-Ed, Inc; 1994:155–169. 11. Hirano M, Bless D. Videostroboscopic examination of the larynx. San Diego: Singular Publishing Group Inc; 1993:90. 12. Anderson TW, Finn JD. The new statistical analysis of data. New York: Springer; 1996:477–480. 13. Colton RH, Casper JK. Understanding voice problems: a physiological perspective for diagnosis and treatment.
Journal of Voice, Vol. 17, No. 3, 2003
2nd ed. Baltimore: Lippincott Williams & Wilkins; 1996: 99–104. 14. Koike Y, Hirano M, von Leden H. Vocal initiation: Acoustic and aerodynamic investigations of normal subjects. Folia Phoniatrica. 1967;19:173–182. 15. Umeda N. Occurrence of glottal stops in fluent speech. J Acoust Soc Am. 1978;64(1):88–94. 16. Fischer-Jorgensen E. Phonetic analysis of the stød in standard Danish. Phonetica. 1989;46(1–3):1–59.
