- Email: [email protected]
Aerobic Exercise as a Warm-Up for Singing: Aerodynamic Changes
Aerobic Exercise as a Warm-Up for Singing: Aerodynamic Changes *Monica McHenry and †Joseph Evans, *Valhalla, New York, and yHouston, Texas Summary: Objectives. This study was designed to determine the impact of aerobic exercise on vocal warm-up. Study design. This is a cohort experimental study. Methods. Sixteen graduate and six undergraduate students in an academic vocal performance program participated. They completed a 30-minute treadmill workout in their target aerobic heart range. Aerodynamic data during singing were acquired before and after the treadmill workout. In full voice, participants sang the first seven notes of the Star Spangled Banner on ‘‘pah,’’ repeating the seventh note seven times, at 1.5 syllables/s after an inhalation. The key was determined by voice type, with the target note within the range of passaggio for men, and in head voice for women. Results. Paired t tests were performed on the data from 17 singers who maintained or increased sound pressure level (SPL) after exercise. Significant pre- to post-exercise increases were found for mean SPL and mean airflow during voicing, although increased estimated subglottal pressure approached significance. These measures were essentially unchanged in individuals who decreased SPL after exercise. There was no significant difference in vocal efficiency after the aerobic exercise, primarily due to large standard deviations within the pre- and post-exercise conditions. Conclusions. Most participants demonstrated favorable aerodynamic changes during singing after aerobic exercise. It is possible that in certain situations, a general aerobic warm-up could set the stage for a less-demanding vocal-specific warm-up, especially for a high voice performing early in the morning. Key Words: Vocal–Voice–Vocal efficiency–Airflow–Subglottal pressure.
INTRODUCTION Classical singing is a complex and demanding act, yet must give the audience an impression of effortlessness. Although years of training are required to develop a healthy and efficient vocal technique, the preparation immediately before performance is equally important for optimal vocal production. Although performers differ markedly in how they use the time before taking the stage, preparation typically includes some version of a vocal warm-up. Although virtually all performers and voice teachers agree that vocal warm-up is essential before performing, there is little consensus regarding either method or purported impact of the exercise. Some have speculated that singing warm-ups may function to improve physiology, prepare the singer psychologically, or simply act as a placebo. Barr1 suggested that the physiological benefits of a warm-up may be crucial to less-trained voices, with the psychological benefits being greater for more accomplished performers. He argued that there may be a placebo effect when the warm-up has no physiological basis, but the singer nonetheless benefits, perhaps by feeling mentally ready to perform. Self-perceptions of warm-up may be deceiving, however. In one study, virtually all singers reported improvement in qualities such as ‘‘resonant voice sensations,’’
Accepted for publication August 14, 2015. Portions of this work were presented at the 44th Annual Symposium: Care of the Professional Voice. Philadelphia, June 2015. From the *Department of Speech-Language Pathology, School of Health Sciences and Practice, New York Medical College, Valhalla, New York; and the yMoores School of Music, University of Houston, Houston, Texas. Address correspondence and reprint requests to Monica McHenry, Department of SpeechLanguage Pathology, School of Health Sciences and Practice, New York Medical College, 30 Plaza West, Ste. 213, Valhalla, NY 10595. E-mail: [email protected] Journal of Voice, Vol. -, No. -, pp. 1-5 0892-1997/$36.00 Ó 2015 The Voice Foundation http://dx.doi.org/10.1016/j.jvoice.2015.08.011
with listeners noting significant change in some specific qualities, but not in the overall rating of ‘‘legit vocal quality.’’2 Contributing to the challenge of interpreting the benefit of warm-up is the considerable variability in warm-up methods, highlighted in a survey of nearly 120 singers at different levels of training.3 Only 53% of the participants reported always warming up before singing, although another 34% stated they usually did. It was noted that singers were more likely to warm up before singing solo than in an ensemble. The duration of the average warm-up was quite brief, typically 5–10 minutes. Most of the reported exercises would be considered part of a traditional vocal warm-up, focusing primarily on vocalizations, although 21% reported using general aerobic exercise as part of their warm-up regimen. The benefits of a traditional vocal warm-up have been documented in two studies2,4 focusing on acoustic changes. After a 25-minute warm-up, Moorcroft and Kenny2 found greater regularity and stability in vibrato. The most interesting finding, however, was the moderating of excessively fast and excessively slow mean vibrato rates. The authors speculated that warm-up may have a normalizing effect, increasing activation levels in some, while decreasing excess activation in others. In the second study of a traditional warm-up,4 participants spent an average of about 10 minutes performing relaxation, postural alignment, breathing exercises, and voice production at varying pitches and dynamic levels. The authors found post–warm-up improvement in measures of frequency and amplitude perturbation, as well as the noise-to-harmonics ratio. They argued against a purely vocal warm-up and concluded that the inclusion of more general exercises targeting breathing, posture, and relaxation contributed to the positive changes. The findings must be interpreted with caution, however, given the nonsystematic impact of vibrato on acoustic measures.5
2 As an alternative method of warm-up, interest in semioccluded vocal tract exercises continues to grow. These exercises are thought to facilitate ease of phonation because the back pressures keep the vocal folds optimally separated. Semioccluded exercises have been compared with a primarily articulatory exercise, saying fast rhymes, performed by professional speakers.6 Both warm-ups were performed relatively briefly before vocal tract observation with magnetic resonance imaging (MRI), as well as acoustic analysis. After warm-up, both speakers produced increased sound pressure level (SPL) and demonstrated a cluster of upper formants. MRI results differed between subjects who demonstrated laryngeal lowering, and wider pharyngeal area, respectively. In a second study of semioccluded vocal tract exercises, Duke et al7 compared traditional warm-up such as messa di voce, with phonation of the Star Spangled Banner through a 5-mm diameter straw for 6 minutes. The primary dependent variable was the singing power ratio (SPR), an indication of the relative amplitude of the highversus low-frequency resonance peaks. The presence of a singer’s formant would be indicated by a low SPR. Contrary to the proposed hypothesis, the investigators found no difference in SPR after traditional versus semioccluded warm-up in SPR. It is possible that aerodynamic, rather than acoustic, measures would have revealed differences, although the participants indicated no significant effect of either warm-up on perceived phonatory effort. The authors suggested that warmup may induce physiological differences that are not detectable with current acoustic measures. It is also possible that, for some people, traditional vocal warm-ups may not be adequate to bring about the physiological changes associated with more general physical warm-up for an activity. Because singing is a highly physical activity, it is reasonable to turn to exercise physiology literature when considering warm-up strategies. The most frequently cited effects of a physical warm-up are increased temperature and blood flow. The benefits of these two effects are extensive and include ‘‘faster muscle contraction and relaxation, greater economy of movement because of lowered viscous resistance within active muscles, improved oxygen delivery and use by muscles, facilitated nerve transmission and muscle metabolism, and increased blood flow through active tissues.’’8 Exercise researchers have also considered the various combinations of warm-up strategies. For example, in a stretching protocol in preparation for swimming sprints, times only improved when stretching was combined with a sport-specific warm-up, rather than with a general aerobic warm-up.9 Also investigating optimal warm-up strategies for swimming, other researchers highlighted the importance of individual differences in designing warm-ups and optimizing performance.10 The most interesting protocol was explored by Murphy et al,11 who varied the timing of aerobic activity combined with static stretching. They found that joint range of motion was optimized when stretching exercises were both preceded and followed with 5 minutes of running. Furthermore, they found that the beneficial effect lasted at least 30 minutes after the warm-up. They suggested that the second aerobic component following the static stretching may have maintained the
Journal of Voice, Vol. -, No. -, 2015
elevated temperature for a longer period of time than either running alone, or running only before stretching. In an initial exploration of the potential impact of exercise on vocal warm-up, 20 actors performed either vocal warm-up in isolation, or vocal warm-up preceded by aerobic exercise.12 The 20-minute vocal warm-up consisted of relaxation with soft, gentle phonation that gradually increased in both pitch and loudness. In a second condition, the actors performed aerobic exercise before the vocal warm-up. They completed exercises such as stepping, marching, jogging, and jumping jacks to maintain their calculated aerobic target heart rate for 5 minutes. Results of this study differed by gender. Women demonstrated a significant reduction in phonatory threshold pressure in the vocal plus aerobic warm-up condition. Men, on the other hand, demonstrated improvement in acoustic measures in the vocal warm-up only condition. The authors suggested two possible reasons for this finding. First, men and women may differ in vocal fold composition, particularly regarding hyaluronic acid, a water-binding molecule. Second, the finding may be attributed to a quirk of the sample, in that the men’s responses to the NASA Physical Activity Questionnaire indicated they were in better physical condition. The differential impact cited previously highlights the importance of tailoring the warm-up to the individual. Variables that may impact the construction of a warm-up may include a performer’s vocal health and general health, as well as voice use and hydration on the day of performance. Furthermore, the demands and expectations of the performance must also be considered.4 It is clear that there are many variations on the theme of warming up to optimize performance. Although some benefit has been found for traditional vocal warm-up strategies, as well as semioccluded vocal tract exercises, and vocal exercises combined with aerobic exercise, warming up exclusively with an aerobic exercise has not been studied. The present study was designed to determine if a general, aerobic exercise would impact voice production in highly trained classical singers. It was hypothesized that after exercise, singers would increase SPL, prephonatory inspiration, estimated subglottal pressure, airflow, and vocal efficiency. METHODS The study was approved by the Committee for the Protection of Human Subjects at the University of Houston. All participants signed an informed consent before beginning the study. Participants Sixteen graduate and six undergraduate students in an academic vocal performance program were participated. There were 11 sopranos, one mezzo-soprano, five tenors, three baritones, and two bass/baritones. The participants reported no history of asthma or other respiratory problems, nor did they indicate any current respiratory condition such as a cold that might impact aerobic function. Age ranged from 19–38 years, with a mean of 24 (standard deviation, 4.4). All participants completed the Physical Activity Readiness Questionnaire
Monica McHenry and Joseph Evans
3
Impact of Aerobic Exercise
(PAR-Q) and indicated they were safe to exercise. NASA Physical Activity Ratings were obtained, where 0 indicates an avoidance of walking and exertion, and 10 indicates running over 25 miles/wk, or 12 hours in comparable activity. Participants’ activity levels ranged from 2–8, with an average of 6.8, indicating that the majority were in good physical condition. Each singer’s aerobic heart range was calculated as 60–80% of their maximum heart rate (calculated as 220 age). Aerobic exercise Singers wore a heart rate monitor while on the treadmill. They completed a treadmill workout, being monitored to ensure that 30 minutes were in their target aerobic heart range. Treadmill speed was adjusted as needed to increase or decrease heart rate. After the workout, the singers rested and drank a half liter of water until their heart rate returned to its pre-workout level. Aerodynamic data acquisition Data during singing were acquired before and after the treadmill workout. Data were collected with a Phonatory Aerodynamic System (PAS, PENTAX Medical, Montvale, New Jersey, USA), which was calibrated before each data collection session. Singers wore a face mask attached to a pneumotachograph to capture airflow, with a pressure-sensing tube placed behind the lips, not touching the tongue or teeth. In full voice, participants sang the first seven notes of the Star Spangled Banner on ‘‘pah,’’ repeating the seventh note seven times, at 1.5 syllables/s after an inhalation. Full voice was defined as voice production common to the Western operatic singing style, with a clear, resonating timbre, and full glottal closure. Singers were instructed to use a self-determined dynamic level of mezzo-forte. The key was determined by voice type, with the target note within the range of passaggio for men, and in head voice for women. For all 10 sopranos, the key was E-flat. The two mezzo-sopranos sang in the key of D. The keys for the tenors were varied, including two in Eflat, two in D, and one in C. Keys were also varied for the baritones and bass/baritones, with two in B-flat, one in A, and the remaining baritone in B-flat. Keys were determined by the
second author (J.E.), on the basis of his familiarity with the students’ voices, and the condition of their voice on the morning of data collection. Data analysis The airflow waveforms were inspected to ensure that no airflow was present during /p/ closure. The first and last syllables of the seven-syllable train were eliminated from analysis. In most cases, automated selection of the pressure peaks and airflow was performed. Occasionally, vibrato was too great for the automated system to accept the airflow as steady state. In these cases, airflow was marked manually. Automated selection generated values for mean peak pressure, mean airflow during voicing, and SPL, and calculated values of estimated laryngeal resistance, acoustic power, and vocal efficiency. To obtain measures of inspiration preceding the targeted syllable train, inspiration was marked manually. Reliability To determine intrajudge reliability, 15% of the sample, predata and postdata from three participants, was reanalyzed. Average differences between original and repeated measures for SPL, estimated subglottal pressure, airflow during voicing, and inspiration, respectively, were 0 dB, 1.7 cm H2O, 5 cc/s, and 25 cc. RESULTS Individual data are presented in Tables 1 and 2. In Table 1, it can be seen that five students produced SPLs more than 1 dB softer when singing after aerobic exercise. This was considered reflective of fatigue. The statistical analysis focused on the 17 singers who maintained or increased SPL after exercise, whose data are summarized in Table 2. Paired t tests were performed to assess differences before and after exercise. Mean SPL increased from 105.1 to 106.2, significant at P ¼ 0.01 (df ¼ 16; t ¼ 2.44). The SPL increases ranged from essentially the same to a 7.5-dB increase. As expected, estimated subglottal pressure also increased, from a mean of 26.5 to 29.4, an increase that approached significance (P ¼ 0.06; df ¼ 16; t ¼ 2.01). Mean airflow during voicing
TABLE 1. Sound Pressure Level (SPL), Estimated Subglottal Pressure (Psg), Airflow During Voicing (Flow), Prephonatory Inspiration (Insp), Aerodynamic Power (Power), Laryngeal Resistance (Resist), and Vocal Efficiency (Effic) for Participants Who Demonstrated Fatigue After Exercise Fatigued Pre-Insp Post Insp Subject Age Gender SPL Diff. Pre-Psg Post-Psg Pre-Flow Post-Flow Vol Vol Pre-Effic Post-Effic 1 2 3 4 5 Mean
26 19 18 25 32 24
M F F M M
6.5 3.8 2.7 1.9 1.5 3.3
33.6 23.2 21.8 31.6 15.2 25.1
29.3 23.1 20.3 36.6 25.0 26.9
270.0 330.0 280.0 560.0 150.0 318.0
230.0 310.0 230.0 610.0 160.0 308.0
1.3 0.8 0.9 1.3 0.3 0.9
1.4 0.6 0.9 1.5 0.7 1.02
Note: These five participants were not included in the statistical analysis because of their decreased SPL after warm-up. Abbreviations: M, male; F, female.
4065.0 887.1 6064.5 2243.3 7577.0 4167.4
1433.1 447.9 4098.1 1118.9 4573.4 2334.3
4
Journal of Voice, Vol. -, No. -, 2015
increased significantly from a mean of 261 cc/s to 302 cc/s (P ¼ 0.02; df ¼ 16; t ¼ 2.63). Mean airflow was essentially unchanged in the participants who decreased SPL. Vocal efficiency is based on airflow during voicing, estimated subglottal pressure, and SPL. There was no significant difference in vocal efficiency after the aerobic exercise (P ¼ 0.85; df ¼ 16; t ¼ 0.19), primarily due to large standard deviations within the pre- and post-exercise conditions. The magnitude of prephonatory inspiration on average, increased only slightly, and not significantly after warm-up (P ¼ 3.87, df ¼ 15, with one singer who did not inhale before syllable train production; t ¼ 2.13). Because meaningful results can be overlooked with averaged data, we would like to highlight the extremes: three individuals who appeared to fatigue with exercise, and three who appeared to benefit the most. Table 3 summarizes these participants. For the individuals who demonstrated fatigue, all aerodynamic measures were negatively impacted. With an average loss of about 4 dB, we see a reduction in estimated subglottal pressure and airflow, with a resultant halving of vocal efficiency. The decline did not appear to be age or gender dependent; there were two women and one man, and the age ranged from 18 to 26 years. By contrast, the individuals who markedly improved with aerobic exercise demonstrated a 6 dB increase. There was a concomitant increase in estimated subglottal pressure by more than a third. They averaged over 100 cc/s increase in airflow, and vocal efficiency doubled for all three. The individuals who demonstrated most improvement were not the youngest in the study; their ages were typical of graduate students. The
oldest singer improved, but not to the extent of the three who benefitted the most. This suggests that singers may benefit from this warm-up, regardless of age or experience.
DISCUSSION This study was designed to determine if individuals completing an aerobic workout would demonstrate changes in various features of voice production. It was hypothesized that prephonatory inspiration, SPL, estimated subglottal pressure, airflow, and vocal efficiency would improve. After eliminating five participants who appeared to fatigue with exercise, all hypothesized trends were seen, with the exception of prephonatory inspiration. It is critical to note that these changes were accomplished without incorporating any vocal warm-up exercises at all. Within the general findings, there were individual differences in the response to the aerobic warm-up. Clearly, if a person is not in good physical condition, and is fatigued by aerobic exercise, this warm-up is not appropriate; however, given the apparent benefits, we would encourage singers to become conditioned enough to incorporate an aerobic warm-up in their performance preparation. A second consideration is a clear conception of an aerobic, versus an anaerobic exercise. An aerobic training level is functionally defined as ‘‘conversational exercise.’’8 The inability to converse is caused by lactate accumulation and shortness of breath, associated with a more vigorous, and possibly anaerobic workout. Thus, in the absence of heart rate monitoring to maintain a target rate between 55%
TABLE 2. Sound Pressure Level (SPL), Estimated Subglottal Pressure (Psg), Airflow During Voicing (Flow), Prephonatory Inspiration (Insp), Aerodynamic Power (Power), Laryngeal Resistance (Resist), and Vocal Efficiency (Effic) for Participants Who Demonstrated Improvement After Exercise Improved Pre-Insp Post-Insp Subject Age Gender SPL Diff. Pre-Psg Post-Psg Pre-Flow Post-Flow Vol Vol Pre-Effic Post-Effic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Mean
22 21 22 23 22 24 25 24 22 28 23 29 21 23 23 26 24
F M F M F M F F F F M M F M F F M
0.9 0.9 0.8 0.7 0.6 0.4 0.2 0.4 0.9 1.4 1.6 1.7 2.4 3.4 4.9 6.1 7.5 1.5
22.7 29.2 20.6 41.5 27.9 26.9 27.1 27.8 26.7 36.7 15.9 36.9 17.9 32.7 19.8 19.3 17.8 26.3
Abbreviations: F, female; M, male; NA, not applicable.
20.3 27.3 24.1 43.1 23.9 18.7 24.4 32.7 18.9 39.3 47.87 38.9 20.6 29.6 24.3 24.9 32.0 28.9
160.0 180.0 180.0 130.0 380.0 200.0 400.0 240.0 400.0 320.0 200.0 340.0 370.0 320.0 270.0 220.0 130.0 261.2
210.0 170.0 190.0 160.0 410.0 170.0 410.0 300.0 300.0 340.0 220.0 450.0 480.0 370.0 320.0 320.0 320.0 302.4
0.7 0.9 1.0 NA 1.1 0.6 1.0 1.0 1.0 0.8 1.0 1.3 1.0 1.4 0.7 0.5 0.9 0.9
0.5 0.6 0.8 NA 1.4 1.1 1.0 1.2 1.0 0.8 1.6 1.1 1.0 1.2 0.8 0.8 1.0 1.0
21321.9 19676.3 14695.9 11001.2 5806.3 1406.5 10072.3 18874.4 5017.9 9998.1 13976.9 4466.5 5326.6 4613.7 10328.4 4508.5 1151.9 9543.7
23762.2 18137.6 8853.9 7348.2 5907.9 2172.3 11847.4 14917.2 7626.6 12053.5 5168.4 4686.2 6165.8 6638.0 18718.53 7884 2643.0 9678.3
Monica McHenry and Joseph Evans
5
Impact of Aerobic Exercise
TABLE 3. Extremes of Changes in SPL and Related Aerodynamic Measures Subject
Age Gender Pre-SPL Post-SPL Difference Pre-Psg Post-Psg Pre-Flow Post-Flow Pre-Effic Post-Effic
3 dB or more decrease 1 26 M 2 19 F 3 18 F Mean 5 dB or more increase 15 23 F 16 26 F 17 24 M Mean
103.8 96.7 103.9 101.5
97.3 92.9 101.2 97.1
6.5 3.8 2.7 4.3
33.6 23.2 21.8 26.2
29.3 23.1 20.3 24.2
270.0 330.0 280.0 293.3
230.0 310.0 230.0 256.7
4065.0 887.1 6064.5 3672.2
1433.1 447.9 4098.1 1993.1
105.2 100.3 92.2 99.2
110.0 106.4 99.7 105.4
4.9 6.1 7.5 6.2
19.8 19.3 17.8 18.9
24.3 24.9 32.0 27.1
270.0 220.0 130.0 206.7
320.0 320.0 320.0 320.0
10328.4 4508.5 1151.9 5329.6
18718.5 7884.0 2643.0 9748.5
Abbreviations: SPL, sound pressure level in dB; Psg, estimated subglottal pressure; Effic, vocal efficiency; M, male; F, female.
and 70% of maximum, the ability to speak during exercise would likely keep one in the optimal aerobic range. It should be noted that this is a cardio workout, as opposed to a weighttraining regimen. We do not advocate weight training as a strategy to warm up the voice. In fact, the laryngeal valving associated with many weight-training activities could negatively impact the voice. Singers experience some situations in which a vocal warmup is particularly necessary, and potentially difficult to accomplish. These include Sitzprobe, morning rehearsals, Young Artist Programs, and church jobs. Sitzprobes (literally, ‘‘seat tests,’’ or a seated rehearsals with the orchestra), particularly in major opera houses, are routinely conducted around 10 AM to accommodate the additional professional demands of orchestra musicians. Because the purpose of the Sitzprobe is to allow the conductor to determine orchestral and singer balance, performance level projection is required. Young Artist Programs in opera companies around the country often require participants to perform in outreach programs that are frequently scheduled at 8 AM in elementary, junior high, and high schools. Although these are typically with piano accompaniment, the vocal demands of the operatic repertoire, and performance without amplification, are expected. Finally, many young professionals rely on Sunday morning church jobs to supplement income. Leading a choir section and performing solos can lead to vocal fatigue, particularly after a Saturday evening vocal performance.13 These situations, and others encountered by working singers, highlight the need for adequate vocal preparation before performing. Despite the potential promise of aerobic exercise to warm up the voice, many details remain unanswered. For example, the optimal timing of an aerobic workout in relation to performance is unknown. Exercise data suggest that it should be close enough that the body is still experiencing the increase in temperature, which seems to benefit muscle function. A second consideration is the willingness of singers to complete an aerobic workout before performing. It is possible that in certain situations, a general aerobic warm-up could set the stage for a less-demanding vocal-specific warm-up, especially for a high voice performing early in the morning. Muscle temperature
would increase, circulation would improve, fluid accumulation in the superficial layer would likely dissipate, and vocal production may be easier. As with all warm-up protocols, customizing the warm-up to the individual is critical. Each singer will need to determine the optimal combination of general aerobic warm-up, with vocalspecific exercises. It is likely that the proportion of each may change with the upcoming vocal demands and time of day, as well as the vocal health of the singer. REFERENCES 1. Barr S. Singing warm-ups: physiology, psychology, or placebo? Logoped Phoniatr Vocol. 2009;34:142–144. 2. Moorcroft L, Kenny DT. Singer and listener perception of vocal warm-up. J Voice. 2013;27:258.e1–258.e13. 3. Gish A, Kuncuk M, Sims L, McWhorter AJ. Vocal warm-up practices and perceptions in vocalists: a pilot survey. J Voice. 2012;26:e1–e10. 4. Amir O, Amir N, Michaeli O. Evaluating the influence of warmup on singing voice quality using acoustic measures. J Voice. 2005;19:252–260. 5. Sundberg J, La FMB, Gill BP. Automatic Measurements of Formant Frequency and Aperiodicity of Vibrato Vowels. 2015. Presented at the 44th Annual Symposium: Care of the Professional Voice, Philadelphia, PA. 6. Laukkanen A-M, Horacek J, Havlı´k R. Case-study magnetic resonance imaging and acoustic investigation of the effects of vocal warm-up on two voice professionals. Logoped Phoniatr Vocol. 2012;37:75–82. 7. Duke E, Plexico LW, Sandage MJ, Hoch M. The effect of traditional singing warm-up versus semioccluded vocal tract exercises on the acoustic parameters of singing voice. J Voice. 2015. http://dx.doi.org/10.1016/j.jvoice. 2014.12.009 [Epub ahead of print]. 8. McArdle WD, Katch FL, Katch VL. Exercise Physiology. 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001:575. 479. 9. Samson M, Button DC, Chaouachi A, Behm DG. Effects of dynamic and static stretching within general and activity specific warm-up protocols. J Sports Sci Med. 2012;11:279–285. 10. Al-Nawaiseh A, Albiero A, Bishop P. Impact of different warmup procedures on a 50-yard swimming spring. Intl J Acad Res. 2013;5:44–48. 11. Murphy JR, Di Santo MC, Alkanani T, Behm DG. Aerobic activity before and following short-duration static stretching improves range of motion and performance vs. a traditional warm-up. Appl Physiol Nutr Metab. 2010;35: 679–690. 12. McHenry M, Johnson J, Foshea B. The effect of specific versus combined warm-up strategies on the voice. J Voice. 2009;25:572–576. 13. McHenry M, Evans J, Powitzky E. Vocal assessment before, after, and the day after opera performance. J Voice. 2015. http://dx.doi.org/10.1016/j. jvoice.2015.02.009 [Epub ahead of print].
INTRODUCTION Classical singing is a complex and demanding act, yet must give the audience an impression of effortlessness. Although years of training are required to develop a healthy and efficient vocal technique, the preparation immediately before performance is equally important for optimal vocal production. Although performers differ markedly in how they use the time before taking the stage, preparation typically includes some version of a vocal warm-up. Although virtually all performers and voice teachers agree that vocal warm-up is essential before performing, there is little consensus regarding either method or purported impact of the exercise. Some have speculated that singing warm-ups may function to improve physiology, prepare the singer psychologically, or simply act as a placebo. Barr1 suggested that the physiological benefits of a warm-up may be crucial to less-trained voices, with the psychological benefits being greater for more accomplished performers. He argued that there may be a placebo effect when the warm-up has no physiological basis, but the singer nonetheless benefits, perhaps by feeling mentally ready to perform. Self-perceptions of warm-up may be deceiving, however. In one study, virtually all singers reported improvement in qualities such as ‘‘resonant voice sensations,’’
Accepted for publication August 14, 2015. Portions of this work were presented at the 44th Annual Symposium: Care of the Professional Voice. Philadelphia, June 2015. From the *Department of Speech-Language Pathology, School of Health Sciences and Practice, New York Medical College, Valhalla, New York; and the yMoores School of Music, University of Houston, Houston, Texas. Address correspondence and reprint requests to Monica McHenry, Department of SpeechLanguage Pathology, School of Health Sciences and Practice, New York Medical College, 30 Plaza West, Ste. 213, Valhalla, NY 10595. E-mail: [email protected] Journal of Voice, Vol. -, No. -, pp. 1-5 0892-1997/$36.00 Ó 2015 The Voice Foundation http://dx.doi.org/10.1016/j.jvoice.2015.08.011
with listeners noting significant change in some specific qualities, but not in the overall rating of ‘‘legit vocal quality.’’2 Contributing to the challenge of interpreting the benefit of warm-up is the considerable variability in warm-up methods, highlighted in a survey of nearly 120 singers at different levels of training.3 Only 53% of the participants reported always warming up before singing, although another 34% stated they usually did. It was noted that singers were more likely to warm up before singing solo than in an ensemble. The duration of the average warm-up was quite brief, typically 5–10 minutes. Most of the reported exercises would be considered part of a traditional vocal warm-up, focusing primarily on vocalizations, although 21% reported using general aerobic exercise as part of their warm-up regimen. The benefits of a traditional vocal warm-up have been documented in two studies2,4 focusing on acoustic changes. After a 25-minute warm-up, Moorcroft and Kenny2 found greater regularity and stability in vibrato. The most interesting finding, however, was the moderating of excessively fast and excessively slow mean vibrato rates. The authors speculated that warm-up may have a normalizing effect, increasing activation levels in some, while decreasing excess activation in others. In the second study of a traditional warm-up,4 participants spent an average of about 10 minutes performing relaxation, postural alignment, breathing exercises, and voice production at varying pitches and dynamic levels. The authors found post–warm-up improvement in measures of frequency and amplitude perturbation, as well as the noise-to-harmonics ratio. They argued against a purely vocal warm-up and concluded that the inclusion of more general exercises targeting breathing, posture, and relaxation contributed to the positive changes. The findings must be interpreted with caution, however, given the nonsystematic impact of vibrato on acoustic measures.5
2 As an alternative method of warm-up, interest in semioccluded vocal tract exercises continues to grow. These exercises are thought to facilitate ease of phonation because the back pressures keep the vocal folds optimally separated. Semioccluded exercises have been compared with a primarily articulatory exercise, saying fast rhymes, performed by professional speakers.6 Both warm-ups were performed relatively briefly before vocal tract observation with magnetic resonance imaging (MRI), as well as acoustic analysis. After warm-up, both speakers produced increased sound pressure level (SPL) and demonstrated a cluster of upper formants. MRI results differed between subjects who demonstrated laryngeal lowering, and wider pharyngeal area, respectively. In a second study of semioccluded vocal tract exercises, Duke et al7 compared traditional warm-up such as messa di voce, with phonation of the Star Spangled Banner through a 5-mm diameter straw for 6 minutes. The primary dependent variable was the singing power ratio (SPR), an indication of the relative amplitude of the highversus low-frequency resonance peaks. The presence of a singer’s formant would be indicated by a low SPR. Contrary to the proposed hypothesis, the investigators found no difference in SPR after traditional versus semioccluded warm-up in SPR. It is possible that aerodynamic, rather than acoustic, measures would have revealed differences, although the participants indicated no significant effect of either warm-up on perceived phonatory effort. The authors suggested that warmup may induce physiological differences that are not detectable with current acoustic measures. It is also possible that, for some people, traditional vocal warm-ups may not be adequate to bring about the physiological changes associated with more general physical warm-up for an activity. Because singing is a highly physical activity, it is reasonable to turn to exercise physiology literature when considering warm-up strategies. The most frequently cited effects of a physical warm-up are increased temperature and blood flow. The benefits of these two effects are extensive and include ‘‘faster muscle contraction and relaxation, greater economy of movement because of lowered viscous resistance within active muscles, improved oxygen delivery and use by muscles, facilitated nerve transmission and muscle metabolism, and increased blood flow through active tissues.’’8 Exercise researchers have also considered the various combinations of warm-up strategies. For example, in a stretching protocol in preparation for swimming sprints, times only improved when stretching was combined with a sport-specific warm-up, rather than with a general aerobic warm-up.9 Also investigating optimal warm-up strategies for swimming, other researchers highlighted the importance of individual differences in designing warm-ups and optimizing performance.10 The most interesting protocol was explored by Murphy et al,11 who varied the timing of aerobic activity combined with static stretching. They found that joint range of motion was optimized when stretching exercises were both preceded and followed with 5 minutes of running. Furthermore, they found that the beneficial effect lasted at least 30 minutes after the warm-up. They suggested that the second aerobic component following the static stretching may have maintained the
Journal of Voice, Vol. -, No. -, 2015
elevated temperature for a longer period of time than either running alone, or running only before stretching. In an initial exploration of the potential impact of exercise on vocal warm-up, 20 actors performed either vocal warm-up in isolation, or vocal warm-up preceded by aerobic exercise.12 The 20-minute vocal warm-up consisted of relaxation with soft, gentle phonation that gradually increased in both pitch and loudness. In a second condition, the actors performed aerobic exercise before the vocal warm-up. They completed exercises such as stepping, marching, jogging, and jumping jacks to maintain their calculated aerobic target heart rate for 5 minutes. Results of this study differed by gender. Women demonstrated a significant reduction in phonatory threshold pressure in the vocal plus aerobic warm-up condition. Men, on the other hand, demonstrated improvement in acoustic measures in the vocal warm-up only condition. The authors suggested two possible reasons for this finding. First, men and women may differ in vocal fold composition, particularly regarding hyaluronic acid, a water-binding molecule. Second, the finding may be attributed to a quirk of the sample, in that the men’s responses to the NASA Physical Activity Questionnaire indicated they were in better physical condition. The differential impact cited previously highlights the importance of tailoring the warm-up to the individual. Variables that may impact the construction of a warm-up may include a performer’s vocal health and general health, as well as voice use and hydration on the day of performance. Furthermore, the demands and expectations of the performance must also be considered.4 It is clear that there are many variations on the theme of warming up to optimize performance. Although some benefit has been found for traditional vocal warm-up strategies, as well as semioccluded vocal tract exercises, and vocal exercises combined with aerobic exercise, warming up exclusively with an aerobic exercise has not been studied. The present study was designed to determine if a general, aerobic exercise would impact voice production in highly trained classical singers. It was hypothesized that after exercise, singers would increase SPL, prephonatory inspiration, estimated subglottal pressure, airflow, and vocal efficiency. METHODS The study was approved by the Committee for the Protection of Human Subjects at the University of Houston. All participants signed an informed consent before beginning the study. Participants Sixteen graduate and six undergraduate students in an academic vocal performance program were participated. There were 11 sopranos, one mezzo-soprano, five tenors, three baritones, and two bass/baritones. The participants reported no history of asthma or other respiratory problems, nor did they indicate any current respiratory condition such as a cold that might impact aerobic function. Age ranged from 19–38 years, with a mean of 24 (standard deviation, 4.4). All participants completed the Physical Activity Readiness Questionnaire
Monica McHenry and Joseph Evans
3
Impact of Aerobic Exercise
(PAR-Q) and indicated they were safe to exercise. NASA Physical Activity Ratings were obtained, where 0 indicates an avoidance of walking and exertion, and 10 indicates running over 25 miles/wk, or 12 hours in comparable activity. Participants’ activity levels ranged from 2–8, with an average of 6.8, indicating that the majority were in good physical condition. Each singer’s aerobic heart range was calculated as 60–80% of their maximum heart rate (calculated as 220 age). Aerobic exercise Singers wore a heart rate monitor while on the treadmill. They completed a treadmill workout, being monitored to ensure that 30 minutes were in their target aerobic heart range. Treadmill speed was adjusted as needed to increase or decrease heart rate. After the workout, the singers rested and drank a half liter of water until their heart rate returned to its pre-workout level. Aerodynamic data acquisition Data during singing were acquired before and after the treadmill workout. Data were collected with a Phonatory Aerodynamic System (PAS, PENTAX Medical, Montvale, New Jersey, USA), which was calibrated before each data collection session. Singers wore a face mask attached to a pneumotachograph to capture airflow, with a pressure-sensing tube placed behind the lips, not touching the tongue or teeth. In full voice, participants sang the first seven notes of the Star Spangled Banner on ‘‘pah,’’ repeating the seventh note seven times, at 1.5 syllables/s after an inhalation. Full voice was defined as voice production common to the Western operatic singing style, with a clear, resonating timbre, and full glottal closure. Singers were instructed to use a self-determined dynamic level of mezzo-forte. The key was determined by voice type, with the target note within the range of passaggio for men, and in head voice for women. For all 10 sopranos, the key was E-flat. The two mezzo-sopranos sang in the key of D. The keys for the tenors were varied, including two in Eflat, two in D, and one in C. Keys were also varied for the baritones and bass/baritones, with two in B-flat, one in A, and the remaining baritone in B-flat. Keys were determined by the
second author (J.E.), on the basis of his familiarity with the students’ voices, and the condition of their voice on the morning of data collection. Data analysis The airflow waveforms were inspected to ensure that no airflow was present during /p/ closure. The first and last syllables of the seven-syllable train were eliminated from analysis. In most cases, automated selection of the pressure peaks and airflow was performed. Occasionally, vibrato was too great for the automated system to accept the airflow as steady state. In these cases, airflow was marked manually. Automated selection generated values for mean peak pressure, mean airflow during voicing, and SPL, and calculated values of estimated laryngeal resistance, acoustic power, and vocal efficiency. To obtain measures of inspiration preceding the targeted syllable train, inspiration was marked manually. Reliability To determine intrajudge reliability, 15% of the sample, predata and postdata from three participants, was reanalyzed. Average differences between original and repeated measures for SPL, estimated subglottal pressure, airflow during voicing, and inspiration, respectively, were 0 dB, 1.7 cm H2O, 5 cc/s, and 25 cc. RESULTS Individual data are presented in Tables 1 and 2. In Table 1, it can be seen that five students produced SPLs more than 1 dB softer when singing after aerobic exercise. This was considered reflective of fatigue. The statistical analysis focused on the 17 singers who maintained or increased SPL after exercise, whose data are summarized in Table 2. Paired t tests were performed to assess differences before and after exercise. Mean SPL increased from 105.1 to 106.2, significant at P ¼ 0.01 (df ¼ 16; t ¼ 2.44). The SPL increases ranged from essentially the same to a 7.5-dB increase. As expected, estimated subglottal pressure also increased, from a mean of 26.5 to 29.4, an increase that approached significance (P ¼ 0.06; df ¼ 16; t ¼ 2.01). Mean airflow during voicing
TABLE 1. Sound Pressure Level (SPL), Estimated Subglottal Pressure (Psg), Airflow During Voicing (Flow), Prephonatory Inspiration (Insp), Aerodynamic Power (Power), Laryngeal Resistance (Resist), and Vocal Efficiency (Effic) for Participants Who Demonstrated Fatigue After Exercise Fatigued Pre-Insp Post Insp Subject Age Gender SPL Diff. Pre-Psg Post-Psg Pre-Flow Post-Flow Vol Vol Pre-Effic Post-Effic 1 2 3 4 5 Mean
26 19 18 25 32 24
M F F M M
6.5 3.8 2.7 1.9 1.5 3.3
33.6 23.2 21.8 31.6 15.2 25.1
29.3 23.1 20.3 36.6 25.0 26.9
270.0 330.0 280.0 560.0 150.0 318.0
230.0 310.0 230.0 610.0 160.0 308.0
1.3 0.8 0.9 1.3 0.3 0.9
1.4 0.6 0.9 1.5 0.7 1.02
Note: These five participants were not included in the statistical analysis because of their decreased SPL after warm-up. Abbreviations: M, male; F, female.
4065.0 887.1 6064.5 2243.3 7577.0 4167.4
1433.1 447.9 4098.1 1118.9 4573.4 2334.3
4
Journal of Voice, Vol. -, No. -, 2015
increased significantly from a mean of 261 cc/s to 302 cc/s (P ¼ 0.02; df ¼ 16; t ¼ 2.63). Mean airflow was essentially unchanged in the participants who decreased SPL. Vocal efficiency is based on airflow during voicing, estimated subglottal pressure, and SPL. There was no significant difference in vocal efficiency after the aerobic exercise (P ¼ 0.85; df ¼ 16; t ¼ 0.19), primarily due to large standard deviations within the pre- and post-exercise conditions. The magnitude of prephonatory inspiration on average, increased only slightly, and not significantly after warm-up (P ¼ 3.87, df ¼ 15, with one singer who did not inhale before syllable train production; t ¼ 2.13). Because meaningful results can be overlooked with averaged data, we would like to highlight the extremes: three individuals who appeared to fatigue with exercise, and three who appeared to benefit the most. Table 3 summarizes these participants. For the individuals who demonstrated fatigue, all aerodynamic measures were negatively impacted. With an average loss of about 4 dB, we see a reduction in estimated subglottal pressure and airflow, with a resultant halving of vocal efficiency. The decline did not appear to be age or gender dependent; there were two women and one man, and the age ranged from 18 to 26 years. By contrast, the individuals who markedly improved with aerobic exercise demonstrated a 6 dB increase. There was a concomitant increase in estimated subglottal pressure by more than a third. They averaged over 100 cc/s increase in airflow, and vocal efficiency doubled for all three. The individuals who demonstrated most improvement were not the youngest in the study; their ages were typical of graduate students. The
oldest singer improved, but not to the extent of the three who benefitted the most. This suggests that singers may benefit from this warm-up, regardless of age or experience.
DISCUSSION This study was designed to determine if individuals completing an aerobic workout would demonstrate changes in various features of voice production. It was hypothesized that prephonatory inspiration, SPL, estimated subglottal pressure, airflow, and vocal efficiency would improve. After eliminating five participants who appeared to fatigue with exercise, all hypothesized trends were seen, with the exception of prephonatory inspiration. It is critical to note that these changes were accomplished without incorporating any vocal warm-up exercises at all. Within the general findings, there were individual differences in the response to the aerobic warm-up. Clearly, if a person is not in good physical condition, and is fatigued by aerobic exercise, this warm-up is not appropriate; however, given the apparent benefits, we would encourage singers to become conditioned enough to incorporate an aerobic warm-up in their performance preparation. A second consideration is a clear conception of an aerobic, versus an anaerobic exercise. An aerobic training level is functionally defined as ‘‘conversational exercise.’’8 The inability to converse is caused by lactate accumulation and shortness of breath, associated with a more vigorous, and possibly anaerobic workout. Thus, in the absence of heart rate monitoring to maintain a target rate between 55%
TABLE 2. Sound Pressure Level (SPL), Estimated Subglottal Pressure (Psg), Airflow During Voicing (Flow), Prephonatory Inspiration (Insp), Aerodynamic Power (Power), Laryngeal Resistance (Resist), and Vocal Efficiency (Effic) for Participants Who Demonstrated Improvement After Exercise Improved Pre-Insp Post-Insp Subject Age Gender SPL Diff. Pre-Psg Post-Psg Pre-Flow Post-Flow Vol Vol Pre-Effic Post-Effic 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Mean
22 21 22 23 22 24 25 24 22 28 23 29 21 23 23 26 24
F M F M F M F F F F M M F M F F M
0.9 0.9 0.8 0.7 0.6 0.4 0.2 0.4 0.9 1.4 1.6 1.7 2.4 3.4 4.9 6.1 7.5 1.5
22.7 29.2 20.6 41.5 27.9 26.9 27.1 27.8 26.7 36.7 15.9 36.9 17.9 32.7 19.8 19.3 17.8 26.3
Abbreviations: F, female; M, male; NA, not applicable.
20.3 27.3 24.1 43.1 23.9 18.7 24.4 32.7 18.9 39.3 47.87 38.9 20.6 29.6 24.3 24.9 32.0 28.9
160.0 180.0 180.0 130.0 380.0 200.0 400.0 240.0 400.0 320.0 200.0 340.0 370.0 320.0 270.0 220.0 130.0 261.2
210.0 170.0 190.0 160.0 410.0 170.0 410.0 300.0 300.0 340.0 220.0 450.0 480.0 370.0 320.0 320.0 320.0 302.4
0.7 0.9 1.0 NA 1.1 0.6 1.0 1.0 1.0 0.8 1.0 1.3 1.0 1.4 0.7 0.5 0.9 0.9
0.5 0.6 0.8 NA 1.4 1.1 1.0 1.2 1.0 0.8 1.6 1.1 1.0 1.2 0.8 0.8 1.0 1.0
21321.9 19676.3 14695.9 11001.2 5806.3 1406.5 10072.3 18874.4 5017.9 9998.1 13976.9 4466.5 5326.6 4613.7 10328.4 4508.5 1151.9 9543.7
23762.2 18137.6 8853.9 7348.2 5907.9 2172.3 11847.4 14917.2 7626.6 12053.5 5168.4 4686.2 6165.8 6638.0 18718.53 7884 2643.0 9678.3
Monica McHenry and Joseph Evans
5
Impact of Aerobic Exercise
TABLE 3. Extremes of Changes in SPL and Related Aerodynamic Measures Subject
Age Gender Pre-SPL Post-SPL Difference Pre-Psg Post-Psg Pre-Flow Post-Flow Pre-Effic Post-Effic
3 dB or more decrease 1 26 M 2 19 F 3 18 F Mean 5 dB or more increase 15 23 F 16 26 F 17 24 M Mean
103.8 96.7 103.9 101.5
97.3 92.9 101.2 97.1
6.5 3.8 2.7 4.3
33.6 23.2 21.8 26.2
29.3 23.1 20.3 24.2
270.0 330.0 280.0 293.3
230.0 310.0 230.0 256.7
4065.0 887.1 6064.5 3672.2
1433.1 447.9 4098.1 1993.1
105.2 100.3 92.2 99.2
110.0 106.4 99.7 105.4
4.9 6.1 7.5 6.2
19.8 19.3 17.8 18.9
24.3 24.9 32.0 27.1
270.0 220.0 130.0 206.7
320.0 320.0 320.0 320.0
10328.4 4508.5 1151.9 5329.6
18718.5 7884.0 2643.0 9748.5
Abbreviations: SPL, sound pressure level in dB; Psg, estimated subglottal pressure; Effic, vocal efficiency; M, male; F, female.
and 70% of maximum, the ability to speak during exercise would likely keep one in the optimal aerobic range. It should be noted that this is a cardio workout, as opposed to a weighttraining regimen. We do not advocate weight training as a strategy to warm up the voice. In fact, the laryngeal valving associated with many weight-training activities could negatively impact the voice. Singers experience some situations in which a vocal warmup is particularly necessary, and potentially difficult to accomplish. These include Sitzprobe, morning rehearsals, Young Artist Programs, and church jobs. Sitzprobes (literally, ‘‘seat tests,’’ or a seated rehearsals with the orchestra), particularly in major opera houses, are routinely conducted around 10 AM to accommodate the additional professional demands of orchestra musicians. Because the purpose of the Sitzprobe is to allow the conductor to determine orchestral and singer balance, performance level projection is required. Young Artist Programs in opera companies around the country often require participants to perform in outreach programs that are frequently scheduled at 8 AM in elementary, junior high, and high schools. Although these are typically with piano accompaniment, the vocal demands of the operatic repertoire, and performance without amplification, are expected. Finally, many young professionals rely on Sunday morning church jobs to supplement income. Leading a choir section and performing solos can lead to vocal fatigue, particularly after a Saturday evening vocal performance.13 These situations, and others encountered by working singers, highlight the need for adequate vocal preparation before performing. Despite the potential promise of aerobic exercise to warm up the voice, many details remain unanswered. For example, the optimal timing of an aerobic workout in relation to performance is unknown. Exercise data suggest that it should be close enough that the body is still experiencing the increase in temperature, which seems to benefit muscle function. A second consideration is the willingness of singers to complete an aerobic workout before performing. It is possible that in certain situations, a general aerobic warm-up could set the stage for a less-demanding vocal-specific warm-up, especially for a high voice performing early in the morning. Muscle temperature
would increase, circulation would improve, fluid accumulation in the superficial layer would likely dissipate, and vocal production may be easier. As with all warm-up protocols, customizing the warm-up to the individual is critical. Each singer will need to determine the optimal combination of general aerobic warm-up, with vocalspecific exercises. It is likely that the proportion of each may change with the upcoming vocal demands and time of day, as well as the vocal health of the singer. REFERENCES 1. Barr S. Singing warm-ups: physiology, psychology, or placebo? Logoped Phoniatr Vocol. 2009;34:142–144. 2. Moorcroft L, Kenny DT. Singer and listener perception of vocal warm-up. J Voice. 2013;27:258.e1–258.e13. 3. Gish A, Kuncuk M, Sims L, McWhorter AJ. Vocal warm-up practices and perceptions in vocalists: a pilot survey. J Voice. 2012;26:e1–e10. 4. Amir O, Amir N, Michaeli O. Evaluating the influence of warmup on singing voice quality using acoustic measures. J Voice. 2005;19:252–260. 5. Sundberg J, La FMB, Gill BP. Automatic Measurements of Formant Frequency and Aperiodicity of Vibrato Vowels. 2015. Presented at the 44th Annual Symposium: Care of the Professional Voice, Philadelphia, PA. 6. Laukkanen A-M, Horacek J, Havlı´k R. Case-study magnetic resonance imaging and acoustic investigation of the effects of vocal warm-up on two voice professionals. Logoped Phoniatr Vocol. 2012;37:75–82. 7. Duke E, Plexico LW, Sandage MJ, Hoch M. The effect of traditional singing warm-up versus semioccluded vocal tract exercises on the acoustic parameters of singing voice. J Voice. 2015. http://dx.doi.org/10.1016/j.jvoice. 2014.12.009 [Epub ahead of print]. 8. McArdle WD, Katch FL, Katch VL. Exercise Physiology. 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001:575. 479. 9. Samson M, Button DC, Chaouachi A, Behm DG. Effects of dynamic and static stretching within general and activity specific warm-up protocols. J Sports Sci Med. 2012;11:279–285. 10. Al-Nawaiseh A, Albiero A, Bishop P. Impact of different warmup procedures on a 50-yard swimming spring. Intl J Acad Res. 2013;5:44–48. 11. Murphy JR, Di Santo MC, Alkanani T, Behm DG. Aerobic activity before and following short-duration static stretching improves range of motion and performance vs. a traditional warm-up. Appl Physiol Nutr Metab. 2010;35: 679–690. 12. McHenry M, Johnson J, Foshea B. The effect of specific versus combined warm-up strategies on the voice. J Voice. 2009;25:572–576. 13. McHenry M, Evans J, Powitzky E. Vocal assessment before, after, and the day after opera performance. J Voice. 2015. http://dx.doi.org/10.1016/j. jvoice.2015.02.009 [Epub ahead of print].