- Email: [email protected]
Investigating the Effects of Glottal Stop Productions on Voice in Children With Cleft Palate Using Multidimensional Voice Assessment Methods
ARTICLE IN PRESS Investigating the Effects of Glottal Stop Productions on Voice in Children With Cleft Palate Using Multidimensional Voice Assessment Methods *Fatma Esen Aydınlı, *Esra Özcebe, *Mavis¸ E. Kulak Kayıkçı, †Taner Yılmaz, and ‡Fatma F. Özgür, *†‡Ankara, Turkey
Summary: Objectives. The aim was to investigate the effects of glottal stop productions (GS) on voice in children with cleft palate using multidimensional voice assessment methods. Study Design. This is a prospective case-control study. Methods. Children with repaired cleft palate (n = 34) who did not have any vocal fold lesions were separated into two groups based on the results of the articulation test. The glottal stop group (GSG) consisted of 17 children who had GS. The control group (CG) consisted of an equal number of age- and gender-matched children who did not have GS. The voice evaluation protocol included acoustic analysis, Pediatric Voice Handicap Index (pVHI), and perceptual analysis (Grade, Roughness, Breathiness, Asthenia, Strain method). The velopharyngeal statuses of the groups were compared using the nasopharyngoscopy and the nasometer. Results. The total pVHI score and the subscales of the pVHI were found to be significantly higher in the GSG. The F0, jitter, and shimmer were found to be numerically higher in the GSG with the difference being statistically significant in jitter (P < 0.05). Audioperceptual analysis revealed a difference in overall voice quality and roughness between the groups. Greater incidence of significant velopharyngeal insufficiency and higher nasalance scores were found in the GSG (P < 0.05). Conclusions. These results may indicate that the vocal quality characteristics of children with GS differ from children who do not have this type of production. It is suggested that children with cleft palate who have GS should receive a comprehensive speech and language pathology intervention including voice therapy techniques. Key Words: Cleft palate–Glottal stop–Velopharyngeal insufficiency–Laryngeal compensation–Dysphonia.
INTRODUCTION Cleft lip and palate is a congenital malformation seen in approximately 1 out of every 700 live births.1 Many researchers have found that children with cleft palate have a tendency to experience voice problems.2–7 Hoarseness, breathiness, reduced loudness, abnormal pitch, and vocal fry are frequently reported in the literature.3,5,6 In addition to these symptoms, vocalfold nodules, edema, inflammation, and hyperplasia are vocalfold pathologies that are also reported in children with cleft palate.5,6,8–10 Many research studies have concluded that there is a clear relationship between velopharyngeal function and dysphonia.5,6,8 The laryngeal compensation hypothesis has been accepted widely as the underlying responsible mechanism.3,5,6,11 According to Warren,12 when air pressure loss occurs due to velopharyngeal insufficiency (VPI), a compensation mechanism works to regulate pressure in the vocal tract by adducting the vocal folds abruptly.4,12,13 This phenomenon is known as glottal stop production. This type of production, which is one of the most common compensatory articulation errors, is a strategy used to ensure enough pressure in the vocal tract.14 When glottal stops are substituted for oral plosives, the manner remains the same Accepted for publication October 14, 2015. From the *Department of Speech and Language Therapy, Hacettepe University Faculty of Health Sciences, Ankara, Turkey; †Department of Otorhinolaryngology, Hacettepe University Medical Faculty, Ankara, Turkey; and the ‡Hacettepe University Medical Faculty, Ankara, Turkey. Address correspondence and reprint requests to Fatma Esen Aydınlı, Department of Speech Language Therapy, Hacettepe University Faculty of Health Science, Sıhhiye, 06100 Ankara, Turkey. E-mail address: [email protected] Journal of Voice, Vol. ■■, No. ■■, pp. ■■-■■ 0892-1997 © 2015 The Voice Foundation. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jvoice.2015.10.007
while the place of articulation changes.12,14,15 Glottal stops can replace all high-pressure consonants including oral fricatives and affricatives,14 and they can be coarticulated with oral placement.4 Even when the velopharyngeal structure is restored, these learned behaviors may still persist; therefore, speech therapy is required.16–18 Studies examining the vocal characteristics of children with VPI included acoustic and perceptual evaluation methods.6,8,10,19,20 Van Lierde et al8 compared the dysphonia severity index score between the healthy and cleft palate children, and they found higher scores in children with VPI. Villafuerte-Gonzales et al19 found vocal abnormalities in F0, jitter, and shimmer, wherein they included 14 children with VPI with healthy couples. Lewis et al6 found more perturbation in the acoustic signal in their study and suggested that results may relate to compensatory articulation patterns. Two retrospective studies10,20 searched for a relationship between VPI and hoarseness10,20 and did not find a relationship between the two. Hamming et al20 suggested that the laryngeal compensation mechanism may not create enough vocal trauma to cause an increase in vocal disturbances. Zajac and Linville7 found a relationship between hoarseness and shimmer. Prior and recent studies examined vocal abnormalities in children with cleft palate by considering only the VPI status.6,8,10,19–21 However, it is clear that every child with VPI may not have glottal stops, and in those who do, the glottal stops may be caused by other factors such as dental malocclusion or fistula.15 Furthermore, there are differences across the studies in terms of the accepted criteria of VPI and the categorization of the VPI status.6,8,10,20 Some of them rely only on perceptual evaluation,
ARTICLE IN PRESS 2 and classify VPI and hoarseness as present or not.10,20 As we know, none of the study searching for the vocal quality characteristics in children with cleft palate included participants who have glottal stop productions. The main purpose of this study is to determine the effects of glottal stop productions on voice. Our research question is: Are the vocal quality characteristics of children with glottal stop productions different from those who do not have that type of productions? METHODS In this study, participants were recruited from the Hacettepe University Hospital Cleft Lip and Palate Council where people applied for examination from November 2014 to February 2015. The effects of glottal stops on voice were prospectively researched using a case-controlled design methodology. All the evaluations were carried out in the Department of Ear, Nose, and Throat and the Audiology and Speech Pathology Unit at Hacettepe University Hospital. This study was approved by the Ethics Committee of Hacettepe University (Approval Number: GO 15/ 26-17). All of the children’s parents gave informed consent for participation, consistent with the Code of Ethics of the World Medical Association (Declaration of Helsinki). Participants Participants with all types of repaired cleft palate, ranging from 4 to 12 years of age, were included in the study. Children were excluded from the study if they had undergone secondary VPI surgery, had fistula, had accompanying syndromes/hearing loss, or had flu on the day of the evaluation. Hearing loss was determined as having more than mild hearing loss in the worse ear, based on the results of an audiometric test that had been conducted within the previous 3 months. Speech therapy history was questioned. The participants who received speech therapy were followed up in the same center. The speech therapy involved standard articulation therapy principles for establishing correct articulatory placement, and none of them included any voice therapy techniques. Speech therapy was received postoperatively once a week for a minimum of 2 months. A total of 60 eligible children with glottal stop (n = 20) and without glottal stop (n = 40) underwent video laryngoscopy evaluation. Participants who had vocal-fold lesions were excluded from the study. In total, glottal stop group (GSG) consisted of 17 children who had glottal stop productions, and control group (CG) consisted of an equal number of age- and sex-matched children who did not have glottal stop productions. The median age of the participants was 6 years and 3 months in the CG, and 5 years and 9 months in the GSG. The upper age limit in this study was determined based on the fact that puberty (after age 12) is a critical period for voice changes. Assessment protocol Articulation assessment This evaluation was conducted using a standardized Turkish Articulation and Phonology Test.22 The subunit of this test, the “Articulation Screening Sub Test,” was performed by a speech
Journal of Voice, Vol. ■■, No. ■■, 2015
language pathologist (SLP) in a silent room. Digital audio and video recordings were taken using Sony handycam HDRCX11E (Sony Corporation, Tokyo, Japan) with a built-in microphone. During the evaluations, the children’s utterances were recorded phonetically as much as possible. Evaluations were completed in approximately 15–30 minutes. Later, the recordings were listened to again, in a silent room to check for International Phonetic Alphabet coding and symbols used for cleft-related errors.23 This articulation test was done to determine whether the children had glottal stop errors. Compatible with this study’s purpose, the consonant errors were classified into three categories, as advised by Henningsson et al.24 These categories were compensatory articulation errors, developmental/ phonological errors, and other articulation errors. Because it is difficult to differentiate the substitution or coarticulation of glottal stops, both were accepted as examples of glottal stop.16 Reliability. Before every rating session, a training evaluation session was performed by listening to compensatory error production education samples (Misarticulation examples: excerpts from Trost-Cardamone [1987]) as found in the compact disc contained in Peterzone et al’s book.4 In addition to these samples, the investigators listened to recordings of two Turkish children with cleft palate. These recordings had been previously edited by two SLPs who agreed that one child had glottal stops and the other had normal articulation. Every recording was analyzed live and with the audio recording by the first author. Intrajudge agreement between the two conditions was determined to be between 92% and 100%. As defined in Henningsson et al’s study,24 error production had to occur more than once in the controlled sampling context for the error type to be considered a reliable error. In addition to the articulation test, conversational speech samples and flexible fiber-optic videoendoscopy images were performed to confirm glottal stops. Voice evaluations Subjective evaluation. Subjective evaluation of voice was conducted using Grade, Roughness, Breathiness, Asthenia, Strain (GRBAS) method25 and Pediatric Voice Handicap Index (pVHI).26 The Turkish pVHI, which is a valid and reliable instrument, includes 23 questions about the functional, physical, and emotional effects of a voice disorder.26 Parents were instructed to answer the questions by considering their child’s voice status during the previous month.26 Speech and voice samples were gathered audiovisually by the SLP in a sound-isolated room. Samples that consisted of recordings that were at least 2 minutes long included counting from 1 to 10, sustaining /a/ phonation, and connected speech.27 Recorded samples were evaluated in another session by two SLPs who were experienced in voice disorders. Recordings were listened to many times until the SLPs agreed on the values of G (grade), R (roughness), and B (breathiness) parameters.25 Grade represents the overall voice quality, roughness represents irregular fluctuations, and breathiness represents the turbulence of airflow in the voice sample.25 Those parameters were evaluated as they were thought to be more reliable in children.28,29
ARTICLE IN PRESS Fatma E.S.E.N. Aydinli et al
Glottal Stop Productions in Children with Cleft Palate
Instrumental evaluation of voice. Acoustic analysis. In the acoustic analysis of voice, the Computerized Speech Lab (CSL) (Model 4300 B) of Kay Elemetrics (Lincoln Park, NJ, USA) was used in conjunction with the Multi-Dimensional Voice Program (MDVP) software (Lincoln Park,NJ, USA). Recordings were made by the same SLP in a sound-isolated room. Participants were asked to sustain /a/ phonation in their conversational speaking pitch and loudness while in an upright position. The SLP held the microphone 15 cm away from the participant’s mouth.30 Before the recordings, the SLP (Manufacturer is the same Key Elemetrics, Lincoln Park,NJ, USA) explained the procedure to the child and gathered trial phonation for a minimum of two times. While analyzing the recordings, the first and last parts were cut and the middle 3 seconds of the phonation interval was analyzed.31 Among the parameters, fundamental frequency (F0), percentage of jitter (%), and the percentage of shimmer (%) were taken into account. Video laryngoscopy. This evaluation was performed via either flexible 3.7-mm diameter steerable fiber-optic laryngoscopy (Optim, Sturbridge, MA, USA) or rigid video laryngostroboscopy using a Kay Pentax digital strobe (Kay Pentax, Lincoln Park, NJ, USA). Kay Pentax Rls 9100 B equipment (Key Elemetrics, Lincoln Park, NJ, USA) was used to capture the images and record the voices. The evaluation method was determined by considering the participant’s age and compatibility with the test method. Images were examined in a separate session by an experienced laryngologist who was not informed about the patient’s condition. Participants who had vocal-fold lesions were excluded from the study. Velopharyngeal function measurements Velopharyngeal function measurements were done to reveal the difference between the two groups with regard to VPI status. Instrumental evaluation methods; nasopharyngoscopy 32 and nasometer33 were used for this purpose. Nasopharyngoscopy (flexible fiber-optic videoendoscopy). Nasopharyngoscopy evaluations were performed by an experienced laryngologist and an SLP. The 3.7-mm diameter flexible fiber-optic nasopharyngoscope was used (Optim, Sturbridge, MA, USA). During the evaluation, the participants were asked to repeat speech samples for a minimum of five times. These samples consisted of nonsense syllables, including high pressure consonants /p/, /t/, /k/, /s/, combined with /ς /vowel, meaningful word composed of high pressure consonants [kɯɾksɛcɪz], meaningful word composed of low pressure consonants [lʌlɛ], and counting from 1 to 10 with high pressure consonants. Audio and video recordings were saved in the device’s hard drive for the purpose of rating in another session. The recordings were evaluated by an SLP who had 8 years of experience in this field. Velopharyngeal closure was rated according to the following three classifications: inefficient, borderline, and efficient.32 Nasometer evaluation. The Kay Pentax Nasometer II (Model 6450) device (Kay Pentax, Lincoln Park, NJ, USA) was
3
used for the evaluation. The test was performed according to the manufacturer’s instructions.34 Although the nasometer protocol includes longer passages in general, shorter samples can be used.34 The speech samples used in this study were the same with those used in a Turkish normative study.35 The children were asked to repeat nonsense syllables, combining high-pressure consonants /p/, /t/, /k/, /s/, with vowels /ʌ/ and /Ι/. Every utterance was repeated for a minimum of eight times. Afterward, participants were asked to count from 1 to 10. Mean nasalance scores were recorded for analysis.
Statistical analysis The IBM SPSS Statistics 21 program (IBM corporation, Armonk, NY, USA) was used in this study’s statistical analysis. Numeric variables were represented by average, standard deviation, median, minimum and maximum values, and interquartile ranges. Normality of the data was investigated using Shapiro Wilk test. Although some of the data (F0, pΙ, tΙ, kΙ, counting) were found to be distributed normally, and majority were not distributed normally, nonparametric tests were used in the analyses. Categorical variables were represented by frequency and percentage. The relation between categorical variables was analyzed using the Chi-square test. The numeric variable differences between the two groups were analyzed using the Mann-Whitney U test. The median and interquartile range values are shown in the tables. The significance value was accepted as 0.05.
RESULTS Demographics of the participant Comparison of the age of palatal surgery The median age at the time of the cleft palate surgery was 10 months in the CG, and 12 months in the GSG. The difference was analyzed using the Mann-Whitney U test. There was not a statistically significant difference detected between the two groups (P > 0.05). Comparison of the speech therapy ratio The difference in the ratio of receiving speech therapy was analyzed using Chi-square test. In both groups, 47% (8/17) of the total participants received speech therapy.
Articulation test results The consonant production errors for both groups were analyzed according to error types and frequency. In the GSG, the glottal stops constituted 127 (32.5%) of the 390 total errors. Other compensatory errors consisted of 36 nasal fricatives (9.2%), 34 posterior nasal fricatives (8.7%), 33 pharyngeal fricatives (8.5%), 28 nasal consonant for oral pressure consonant (7.1%), and 12 middorsum palatal stops (3.1%). Other articulation errors included 61 oral misarticulation errors (15.7%) and 59 developmental articulation/phonological errors (15.1%). In the GSG, 14 participants exhibited glottal stop errors in 3–6 consonants, and 3 participants exhibited glottal stop errors in 7–9 consonants.
ARTICLE IN PRESS 4
Journal of Voice, Vol. ■■, No. ■■, 2015
TABLE 1. Comparison of the Acoustic Vocal Parameters (Analyzed by CSL) Between the Groups Control Group
Glottal Stop Group
Acoustic Parameters
(Median ± IQR)
(Median ± IQR)
P Value*
F0 (Hz) Jitter (%) Shimmer (%)
301.55 ± 43.31 0.56 ± 0.5 2.98 ± 2.11
313.02 ± 63.64 1.06 ± 1.19 3.67 ± 2.89
NS 0.049* NS
Notes: Table 1 shows the acoustic analysis results. Numerical values of all parameters were found to be higher in the GSG. Jitter parameter was also found to be statistically significant. * Significant at P < 0.05. Abbreviations: IQR, interquartile range, NS, nonsignificant.
Voice evaluations Comparison of the acoustic analysis between the groups The difference between the groups was analyzed using the MannWhitney U test. In a Turkish study,36 in searching for normal values in 203 healthy children between 4 and 14 years old, difference was not found in F0 between gender under the age of 13, and F0 values were not inspected with respect to gender. Similarly, two studies36,37 showed no age and sex differences in jitter and shimmer values. Therefore, we inspected these variables independent of arbitrary age categories. (Table 1). The F0, jitter, and shimmer values were found to be numerically higher in the GSG. The difference was statistically significant for the jitter value (P < 0.05). Comparison of the pVHI scores The subunits and total scores of the pVHI were compared using the Mann-Whitney U test between the groups. All the subunit scores and the total score for the GSG were found to be significantly higher than those of the CG (P < 0.05). (Table 2). The total score for the GSG was found to be approximately three times higher than that of the CG. Comparison of the GRBAS results The differences in the G, R, and B parameters between the groups were evaluated using the Chi-square test, and the difference was detected for the G and R parameters (P < 0.05). Chi-square coefficients and P values were found for the G, R, and B parameters,
respectively, (χ2 = 13.592, P = 0.000), (χ2 = 7.798, P = 0.005), (χ2 = 1.417, P = 0.234). The voice quality was perceived to have a mild grade of hoarseness in eight (47%) participants and with a mild grade of roughness in five (29.4%) participants of the GSG. The G and R parameters were perceived to be normal in all of the participants in the CG. The distribution of the B parameter was almost identical in both of the groups. The maximum value was 1 for both groups. Velopharyngeal function evaluations Comparison of the nasometer results The Mann-Whitney U test was used to determine the differences between the two groups. In every speech utterance, the nasalance scores were found to be significantly higher (P < 0.05) in the glottal stop compared with the CG, except for the sample of [kΙ]. Table 3 displays the comparison of the nasometer scores between the two groups. Velopharyngeal closure evaluation One participant’s velopharyngeal closure evaluation was excluded in the CG due to nonreliability. The Chi-square test was used to determine the difference between the velopharyngeal closure (χ2 = 23,733 and P = 0 < 0.01). In the CG, 11 (64.7%) had efficient closure, 4 (23.5%) had borderline closure, and 1 (5.8%) had inefficient closure. In the GSG, 12 (70.5%) had borderline closure, and 5 (29.5%) had inefficient closure. Therefore, the GSG had greater incidence of significant VPI than the CG.
TABLE 2. Comparison of the pVHI Scores: Subunits’ and Their Total Scores Between the Two Groups Control Group
Glottal Stop Group
Subunits of pVHI
(Median + IQR)
(Median + IQR)
P Value
Functional score Physical score Emotional score Total score
2 ± 5.0 2 ± 6.00 1 ± 3.0 5 ± 14.00
6 ± 4.50 5 ± 9.00 5 ± 8.50 16 ± 25.00
0.018* 0.005* 0.005† 0.014*
Notes: Table 2 displays the median and interquartile range values for pVHI subunits and their total score. In all parameters, GSG had higher values, and difference was statistically important. * Significant at P < 0.05. † Significant at P < 0.01. Abbreviations: IQR, interquartile range; M, median.
ARTICLE IN PRESS Fatma E.S.E.N. Aydinli et al
5
Glottal Stop Productions in Children with Cleft Palate
TABLE 3. Comparison of the Speech Utterances’ Nasometer Values Between the Two Groups
Speech Utterance Pʌ PΙ Tʌ TΙ Kʌ KΙ Sʌ SΙ Counting
Control Group
Glottal Stop Group
(Median + IQR)
(Median + IQR)
P Value
15.00 ± 27.00 26.00 ± 20.00 10.00 ± 30.00 30.00 ± 36.00 10.00 ± 19.00 31.00 ± 35.00 12.00 ± 40.00 38.00 ± 27.00 26.00 ± 25.00
35.00 ± 29.00 58.00 ± 46.00 41.00 ± 29.50 56.00 ± 29.00 33.00 ± 35.50 58.00 ± 43.50 40.00 ± 26.50 59.00 ± 29.00 56.00 ± 26.50
0.016* 0.006† 0.010† 0.04* 0.010† NS 0.011* 0.018* 0.001†
Notes: Table 3 displays the median and interquartile range of speech utterances’ nasometer values. In all utterances, GSG had higher values and difference was statistically important except for the sample of KΙ. * Significant at P < 0.05. † Significant at P < 0.01. Abbreviations: IQR, interquartile range; M, median; NS, nonsignificant.
DISCUSSION Many studies have been carried out to search for vocal-quality characteristics in children with VPI.6–8,19Among these studies, Zajac and Linville’s7 and Van Lierde et al’s8 studies included the multiparameter approach, whereas the studies of Lewis et al6 and Villafuerte-Gonzalez et al 19 included only acoustic analysis. It is recommended that voice evaluation include auditoryperceptual methods, self-assessment tools, and objective measures.38 Perceptual ratings are often used to validate objective measures.39 The present study included not only acoustic analysis, but also audio perceptual analysis and self-assessment tool. According to the pVHI scores, not only the total score, but also the functional, physical, and emotional subscale scores were found to be higher in the GSG. The numerical value for the total score was higher than 8.6 in the GSG, which was the limit determined in healthy subjects in the original validity and reliability study of this index.26 These results may indicate that parents of children in GSG perceive their children’s voice quality to be somewhat influenced. In the literature, only one study investigated voice performance by a self-assessment tool (pediatric voice outcomes survey in patients with VPI). Boseley and Hartnick found that scores were improved after surgery.40 Care should be taken when interpreting the pVHI results because there is a probability that the parents’ responses to the questions were based not only on their impression of their child’s voice, but also on the articulation and resonance characteristics of the child’s voice. This may arise from the fact that the pVHI was not designed to differentiate resonance disorders from articulation disorders or glottal dysphonias.40 The jitter and shimmer values found in the present study were not higher than the normative values determined in a Turkish normative study36 in both groups. However, higher values of F0, jitter, and shimmer in the GSG may indicate that the glottal stop articulations can be associated with vocal perturbations. This may result from the struggle to regulate laryngeal and respiratory aerodynamics.7 On the other hand, roughness value was found
to differ in GSG from CG, which is compatible with the significantly higher jitter value detected in GSG. Similarly, Zajac and Linville7 found significantly higher value in jitter, wherein they included 10 children with VPI and 5 healthy couples. Villafuerte-Gonzales et al19 concluded that vocal abnormalities expressed by higher values in F0, jitter, and shimmer in their study where they included 14 children with healthy couples. Yang et al41 found lower cepstral peak prominence values when comparing pre–post VPI surgery results for vowels. They suggested that the articulatory movements that have an effect on maintaining adequate intraoral pressure might be the underlying cause. Van Lierde et al8 supported the idea of decreased vocal quality results from the attempts to compensate VPI as a result of their acoustic analysis. The voice quality was perceived to have a mild grade of hoarseness and roughness in some participants of the GSG. This result may imply that irregularity in frequency was perceptually evident in some of the participants with glottal stop. However, it should be noted that the range of the values were very narrow (0–1) for all parameters. In the literature, similarly, Van Lierde et al8 found mild degree of disturbance in the G and R parameters. Zajac and Linville7 measured hoarseness rate at 1.90 points in a 7-point equal-interval scale. Similarly, D’Antonio et al42 rated voices audio-perceptually in an equal-interval scale in children with VPI. They reported hoarseness as the most frequent symptom, and abnormal pitch and reduced loudness as the most frequent tertiary symptoms in children with VPI. In this present study, listeners concluded that perceptual analysis was more difficult when there was an accompanying VPI. This can be a result of both hypernasality and nasal emission secondary to VPI, which can mask audible effects of breathiness. Breathiness was rated abnormal in a few participants in both groups. As Imatomi43 discussed, breathiness may conflict with hypernasality. Imatomi43 noted that when there is mild hypernasality, breathiness causes it to be perceived as more severe. Conversely, when there is severe hypernasality, breathiness causes it to be perceived as milder. Another theory suggested by Tavares et al44 posits that the
ARTICLE IN PRESS 6 posterior triangular slit glottic configuration may make it difficult to identify breathiness in childhood. In this study, velopharyngeal function evaluations revealed that there was a clear relationship between velopharyngeal function and the searched glottal stop production as generally accepted in the literature.12,14 Leder and Lerman45 showed through a spectrogram that severe hypernasality was associated with glottal stops. In healthy voice production, it is known that the respiration, phonation, resonance, and articulation systems must be in good balance. Titze and Laukkanen46 suggested that supraglottic vocal tract characteristics have a direct effect on vocal-fold oscillations. Hence, compensatory articulation productions may lead to conflict among the respiration, phonation, resonance, and articulation systems. Although glottal stops constitute the higher percentage of all compensatory errors, many of the participants in the present study exhibited two or more types of compensatory errors. Although it seems reasonable that glottal stops are the primary compensatory errors to have potential effect on voice because they occur at the laryngeal level, other errors may contribute to vocal quality by attempting to regulate pressure in the vocal tract. This factor makes reaching definitive conclusions difficult in this study. In future studies, it is reasonable to classify participants according to compensatory articulation error types and explore their aerodynamic profiles by measuring the subglottal pressure and laryngeal airflow rate.47 Besides, it may be more advantageous to design a long-term follow-up study. This is because the severity of glottal stops can change from time to time. It may be a better option to evaluate the participants’ voice performance periodically in relation to glottal stops. Furthermore, increasing the number of participants could contribute to the increase of the validity and generalizability of the study’s results. CONCLUSIONS These results may indicate that the vocal quality characteristics of children with glottal stops differ from children who do not have this type of production. The difference was most pronounced in jitter parameter of acoustic analysis. GRBAS showed difference in overall voice quality and roughness. pVHI showed that parents perceived their children’s voice quality to be somewhat influenced. Although these results support that children with glottal stops may be at more risk to develop vocal pathologies, other types of compensatory articulations should be taken into account in future studies with a higher number of participants. In conclusion, it is suggested that children with cleft palate who have glottal stop productions should receive a comprehensive speech and language pathology intervention including voice therapy techniques. REFERENCES 1. Dixon MJ, Marazita ML, Beaty TH, et al. Cleft lip and palate: synthesizing genetic and environmental influences. Nat Rev Gen. 2011;12:167–178. 2. Peterson-Falzone SJ, Hardin-Jones M, Karnell MP. Cleft Palate Speech. 3rd ed. St. Louis, MO: Mosby-Elsevier; 2001. 3. Kummer AW. Cleft Palate and Craniofacial Anomalies Effects on Speech and Resonance. 2nd ed. Albany, NY: Thomson Delmar Learning; 2008.
Journal of Voice, Vol. ■■, No. ■■, 2015 4. Peterson-Falzone SJ, Trost-Cardamone JE, Karnell MP, et al. Effect of cleft and non-cleft VPI on speech in older children. In: Falk K, ed. The Clinician’s Guide to Treating Cleft Palate Speech. St. Louis, MO: Mosby Elsevier; 2006:5–16. 5. McWilliams BJ, Bluestone CD, Musgrave RH. Diagnostic implications of vocal cord nodules in children with cleft palate. Laryngoscope. 1969;79:2072–2080. 6. Lewis JR, Andreassen ML, Leeper HA, et al. Vocal characteristics of children with cleft lip/palate and associated velopharyngeal incompetence. J Otolaryngol. 1993;22:113–117. 7. Zajac DJ, Linville RN. Voice perturbations of children with perceived nasality and hoarseness. Cleft Palate J. 1989;26:226–231. 8. Van Lierde KM, Claeys S, De Bodt M, et al. Vocal quality characteristics in children with cleft palate: a multiparameter approach. J Voice. 2004;18:354–362. 9. Hocevar-Boltezar I, Jarc A, Kozelj V. Ear, nose and voice problems in children with orofacial clefts. J Laryngol Otol. 2006;120:276–281. 10. Robison JG, Otteson TD. Prevalence of hoarseness in the cleft palate population. Arch Otolaryngol Head Neck Surg. 2011;137:74–77. 11. McWilliams BJ, Lavorato AS, Bluestone CD. Vocal cord abnormalities in children with velopharyngeal valving problems. Laryngoscope. 1973;83:1745–1753. 12. Warren DW. Compensatory speech behaviours in individuals with cleft palate: a regulation/control phenomenon? Cleft Palate J 1986;23:251–260. 13. Trindade IE, Paciello RZ, Trindade Junior AS, et al. Consonantal intraoral air pressure characteristics in Brazilian postoperative cleft palate speakers with velopharyngeal disorders. Braz J Med Biol Res. 1989;22:667–674. 14. Golding-Kushner KJ. Therapy Techniques for Cleft Palate Speech and Related Disorders. San Diego, CA: Singular Thomson Learning; 2001. 15. Kummer AW. Disorders of resonance and airflow secondary to cleft palate and/or velopharyngeal dysfunction. Semin Speech Lang. 2011;32:141– 149. 16. Gooch JL, Hardin-Jones M, Chapman KL, et al. Reliability of listener transcriptions of compensatory articulations. Cleft Palate J 2001;38:59–67. 17. Kummer AW. Speech therapy for errors secondary to cleft palate and velopharyngeal dysfunction. Semin Speech Lang. 2011;32:191–198. 18. Bispo NH, Whitaker ME, Aferri HC, et al. Speech therapy for compensatory articulations and velopharyngeal function: a case report. J Appl Oral Sci. 2011;19:679–684. 19. Villafuerte-Gonzalez R, Valadez-Jimenez VM, Hernandez-Lopez X, et al. Acoustic analysis of voice in children with cleft palate and velopharyngeal insufficiency. Int J Pediatr Otorhinolaryngol. 2015;79:1073–1076. 20. Hamming KK, Finkelstein M, Sidman JD. Hoarseness in children with cleft palate. Otolaryngol Head Neck Surg. 2009;140:902–906. 21. Van Lierde KM, Bonte K, Baudonck N, et al. Speech outcome regarding overall intelligibility, articulation, resonance and voice in Flemish children a year after pharyngeal flap surgery, a pilot study. Folia Phoniatr Logop. 2008;60:223–232. 22. Topbas¸ S. Türkçe Sesletim–Sesbilgisi Testi: Geçerlik-Güvenirlik ve Standardizasyon Çalıs¸ması. Türk Psikoloji Dergisi. 2006;21:39–56. 23. Trost JE. Articulatory additions to the classical description of the speech of persons with cleft palate. Cleft Palate J. 1981;18:193–203. 24. Henningsson G, Kuehn DP, Sell D, et al. Universal parameters for reporting speech outcomes in individuals with cleft palate. Cleft Palate J 2008;45: 1–17. 25. De Bodt MS, Wuyts FL, Van de Heyning PH, et al. Test-retest study of the GRBAS scale: influence of experience and professional background on perceptual rating of voice quality. J Voice. 1997;11:74–80. 26. Tadıhan Özkan E, Tüzüner A, Demirhan E, et al. Reliability and validity of the Turkish Pediatric Voice Handicap Index. Int J Pediatr Otorhinolaryngol. 2015; 79:680–684. 27. de Labio RB, Tavares EL, Alvarado RC, et al. Consequences of chronic nasal obstruction on the laryngeal mucosa and voice quality of 4- to 12-year-old children. J Voice. 2012;26:488–492. 28. Tavares EL, Brasolotto A, Santana MF, et al. Epidemiological study of dysphonia in 4–12 year-old children. Braz J Otorhinolaryngol. 2011;77:736– 746.
ARTICLE IN PRESS Fatma E.S.E.N. Aydinli et al
Glottal Stop Productions in Children with Cleft Palate
29. Schindler A, Tiddia C, Ghidelli C, et al. Adaptation and validation of the Italian Pediatric Voice Handicap Index. Folia Phoniatr Logop. 2011;63:9–14. 30. Souza LS, Crespo AN, Medeiros JLA. Laryngeal vocal and endoscopic alterations after thyroidectomy under local anesthesia and hypnosedation. Braz J Otorhinolaryngol. 2009;75:511–516. 31. Kent RD, Vorperian HK, Duffy JR. Reliability of the multi-dimensional voice program for the analysis of voice samples of subjects with dysarthria. Am J Speech Lang Pathol. 1999;8:129–136. 32. Antonio D, MuntzHR LL, Marsh JL, et al. Practical application of flexible fiberoptic nasopharyngoscopy for evaluating velopharyngeal function. Plast Reconst Surg. 1988;82:611–618. 33. Watterson T, Lewis KE, Foley-Homan N. Effect of stimulus length on nasalance scores. Cleft Palate J 1999;36:243–247. 34. Kay Pentax. Instruction Manual Nasometer II Model 6450, Issue C. Lincoln Park, NJ: KayPentax; 2010. 35. Sarac ET, Kayikci ME, Ozkan S. Nasality evaluation of Turkish phonemes in vowel-consonant combinations. Int J Pediatr Otorhinolaryngol. 2011;75:894–898. 36. Akmese PP, Kulak Kayikci ME, Atas A. Acoustic characteristics of Turkish speaking children ages between 4 and 14 years old. Int Adv Otol. 2012;8:399–406. 37. Maturo S, Hill C, Bunting G, et al. Establishment of a normative pediatric acoustic database. Arch Otolaryngol Head Neck Surg. 2012;138:956–961. 38. Ziwei Y, Zheng P, Pin D. Multiparameter voice assessment for voice disorder patients: a correlation analysis between objective and subjective parameters. J Voice. 2014;28:770–774.
7
39. Hillebrand J, Cleveland RA, Erickson RL. Acoustic correlates of breathy vocal quality: dysphonic voices and continuous speech. J Speech Hear Res. 1994;39:311–321. 40. Boseley ME, Hartnick CJ. Assessing the outcome of surgery to correct velopharyngeal insufficiency with the pediatric voice outcomes survey. Int J Pediatr Otorhinolaryngol. 2004;68:1429–1433. 41. Yang Z, Fan J, Tian J, et al. Cepstral analysis of voice in children with velopharyngeal insufficiency after cleft palate surgery. J Voice. 2014;28:789– 792. 42. D’Antonio LL, Muntz HR, Province MA, et al. Laryngeal/voice findings in patients with velopharyngeal dysfunction. Laryngoscope. 1988;98:432– 438. 43. Imatomi S. Effects of breathy voice source on ratings of hypernasality. Cleft Palate J 2005;42:641–648. 44. Tavares EL, Labio RB, Martins RH. Normative study of vocal acoustic parameters from children from 4 to 12 years of age without vocal symptoms: a pilot study. Braz J Otorhinolaryngol. 2010;76:485–490. 45. Leder SB, Lerman JW. Some acoustic evidence for vocal abuse in adult speakers with repaired cleft palate. Laryngoscope. 1985;95:837–840. 46. Titze IR, Laukkanen AM. Can vocal economy in phonation be increased with an artificially lengthened vocal tract? A computer modeling study. Logoped Phoniatr Vocol. 2007;32:147–156. 47. Gillespie AI, Gardner-Schmidt J, Rubinstein EN, et al. Aerodynamic profiles of women with muscle tension dysphonia/aphonia. J Speech Lang Hear Res. 2013;56:481–488.
Summary: Objectives. The aim was to investigate the effects of glottal stop productions (GS) on voice in children with cleft palate using multidimensional voice assessment methods. Study Design. This is a prospective case-control study. Methods. Children with repaired cleft palate (n = 34) who did not have any vocal fold lesions were separated into two groups based on the results of the articulation test. The glottal stop group (GSG) consisted of 17 children who had GS. The control group (CG) consisted of an equal number of age- and gender-matched children who did not have GS. The voice evaluation protocol included acoustic analysis, Pediatric Voice Handicap Index (pVHI), and perceptual analysis (Grade, Roughness, Breathiness, Asthenia, Strain method). The velopharyngeal statuses of the groups were compared using the nasopharyngoscopy and the nasometer. Results. The total pVHI score and the subscales of the pVHI were found to be significantly higher in the GSG. The F0, jitter, and shimmer were found to be numerically higher in the GSG with the difference being statistically significant in jitter (P < 0.05). Audioperceptual analysis revealed a difference in overall voice quality and roughness between the groups. Greater incidence of significant velopharyngeal insufficiency and higher nasalance scores were found in the GSG (P < 0.05). Conclusions. These results may indicate that the vocal quality characteristics of children with GS differ from children who do not have this type of production. It is suggested that children with cleft palate who have GS should receive a comprehensive speech and language pathology intervention including voice therapy techniques. Key Words: Cleft palate–Glottal stop–Velopharyngeal insufficiency–Laryngeal compensation–Dysphonia.
INTRODUCTION Cleft lip and palate is a congenital malformation seen in approximately 1 out of every 700 live births.1 Many researchers have found that children with cleft palate have a tendency to experience voice problems.2–7 Hoarseness, breathiness, reduced loudness, abnormal pitch, and vocal fry are frequently reported in the literature.3,5,6 In addition to these symptoms, vocalfold nodules, edema, inflammation, and hyperplasia are vocalfold pathologies that are also reported in children with cleft palate.5,6,8–10 Many research studies have concluded that there is a clear relationship between velopharyngeal function and dysphonia.5,6,8 The laryngeal compensation hypothesis has been accepted widely as the underlying responsible mechanism.3,5,6,11 According to Warren,12 when air pressure loss occurs due to velopharyngeal insufficiency (VPI), a compensation mechanism works to regulate pressure in the vocal tract by adducting the vocal folds abruptly.4,12,13 This phenomenon is known as glottal stop production. This type of production, which is one of the most common compensatory articulation errors, is a strategy used to ensure enough pressure in the vocal tract.14 When glottal stops are substituted for oral plosives, the manner remains the same Accepted for publication October 14, 2015. From the *Department of Speech and Language Therapy, Hacettepe University Faculty of Health Sciences, Ankara, Turkey; †Department of Otorhinolaryngology, Hacettepe University Medical Faculty, Ankara, Turkey; and the ‡Hacettepe University Medical Faculty, Ankara, Turkey. Address correspondence and reprint requests to Fatma Esen Aydınlı, Department of Speech Language Therapy, Hacettepe University Faculty of Health Science, Sıhhiye, 06100 Ankara, Turkey. E-mail address: [email protected] Journal of Voice, Vol. ■■, No. ■■, pp. ■■-■■ 0892-1997 © 2015 The Voice Foundation. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jvoice.2015.10.007
while the place of articulation changes.12,14,15 Glottal stops can replace all high-pressure consonants including oral fricatives and affricatives,14 and they can be coarticulated with oral placement.4 Even when the velopharyngeal structure is restored, these learned behaviors may still persist; therefore, speech therapy is required.16–18 Studies examining the vocal characteristics of children with VPI included acoustic and perceptual evaluation methods.6,8,10,19,20 Van Lierde et al8 compared the dysphonia severity index score between the healthy and cleft palate children, and they found higher scores in children with VPI. Villafuerte-Gonzales et al19 found vocal abnormalities in F0, jitter, and shimmer, wherein they included 14 children with VPI with healthy couples. Lewis et al6 found more perturbation in the acoustic signal in their study and suggested that results may relate to compensatory articulation patterns. Two retrospective studies10,20 searched for a relationship between VPI and hoarseness10,20 and did not find a relationship between the two. Hamming et al20 suggested that the laryngeal compensation mechanism may not create enough vocal trauma to cause an increase in vocal disturbances. Zajac and Linville7 found a relationship between hoarseness and shimmer. Prior and recent studies examined vocal abnormalities in children with cleft palate by considering only the VPI status.6,8,10,19–21 However, it is clear that every child with VPI may not have glottal stops, and in those who do, the glottal stops may be caused by other factors such as dental malocclusion or fistula.15 Furthermore, there are differences across the studies in terms of the accepted criteria of VPI and the categorization of the VPI status.6,8,10,20 Some of them rely only on perceptual evaluation,
ARTICLE IN PRESS 2 and classify VPI and hoarseness as present or not.10,20 As we know, none of the study searching for the vocal quality characteristics in children with cleft palate included participants who have glottal stop productions. The main purpose of this study is to determine the effects of glottal stop productions on voice. Our research question is: Are the vocal quality characteristics of children with glottal stop productions different from those who do not have that type of productions? METHODS In this study, participants were recruited from the Hacettepe University Hospital Cleft Lip and Palate Council where people applied for examination from November 2014 to February 2015. The effects of glottal stops on voice were prospectively researched using a case-controlled design methodology. All the evaluations were carried out in the Department of Ear, Nose, and Throat and the Audiology and Speech Pathology Unit at Hacettepe University Hospital. This study was approved by the Ethics Committee of Hacettepe University (Approval Number: GO 15/ 26-17). All of the children’s parents gave informed consent for participation, consistent with the Code of Ethics of the World Medical Association (Declaration of Helsinki). Participants Participants with all types of repaired cleft palate, ranging from 4 to 12 years of age, were included in the study. Children were excluded from the study if they had undergone secondary VPI surgery, had fistula, had accompanying syndromes/hearing loss, or had flu on the day of the evaluation. Hearing loss was determined as having more than mild hearing loss in the worse ear, based on the results of an audiometric test that had been conducted within the previous 3 months. Speech therapy history was questioned. The participants who received speech therapy were followed up in the same center. The speech therapy involved standard articulation therapy principles for establishing correct articulatory placement, and none of them included any voice therapy techniques. Speech therapy was received postoperatively once a week for a minimum of 2 months. A total of 60 eligible children with glottal stop (n = 20) and without glottal stop (n = 40) underwent video laryngoscopy evaluation. Participants who had vocal-fold lesions were excluded from the study. In total, glottal stop group (GSG) consisted of 17 children who had glottal stop productions, and control group (CG) consisted of an equal number of age- and sex-matched children who did not have glottal stop productions. The median age of the participants was 6 years and 3 months in the CG, and 5 years and 9 months in the GSG. The upper age limit in this study was determined based on the fact that puberty (after age 12) is a critical period for voice changes. Assessment protocol Articulation assessment This evaluation was conducted using a standardized Turkish Articulation and Phonology Test.22 The subunit of this test, the “Articulation Screening Sub Test,” was performed by a speech
Journal of Voice, Vol. ■■, No. ■■, 2015
language pathologist (SLP) in a silent room. Digital audio and video recordings were taken using Sony handycam HDRCX11E (Sony Corporation, Tokyo, Japan) with a built-in microphone. During the evaluations, the children’s utterances were recorded phonetically as much as possible. Evaluations were completed in approximately 15–30 minutes. Later, the recordings were listened to again, in a silent room to check for International Phonetic Alphabet coding and symbols used for cleft-related errors.23 This articulation test was done to determine whether the children had glottal stop errors. Compatible with this study’s purpose, the consonant errors were classified into three categories, as advised by Henningsson et al.24 These categories were compensatory articulation errors, developmental/ phonological errors, and other articulation errors. Because it is difficult to differentiate the substitution or coarticulation of glottal stops, both were accepted as examples of glottal stop.16 Reliability. Before every rating session, a training evaluation session was performed by listening to compensatory error production education samples (Misarticulation examples: excerpts from Trost-Cardamone [1987]) as found in the compact disc contained in Peterzone et al’s book.4 In addition to these samples, the investigators listened to recordings of two Turkish children with cleft palate. These recordings had been previously edited by two SLPs who agreed that one child had glottal stops and the other had normal articulation. Every recording was analyzed live and with the audio recording by the first author. Intrajudge agreement between the two conditions was determined to be between 92% and 100%. As defined in Henningsson et al’s study,24 error production had to occur more than once in the controlled sampling context for the error type to be considered a reliable error. In addition to the articulation test, conversational speech samples and flexible fiber-optic videoendoscopy images were performed to confirm glottal stops. Voice evaluations Subjective evaluation. Subjective evaluation of voice was conducted using Grade, Roughness, Breathiness, Asthenia, Strain (GRBAS) method25 and Pediatric Voice Handicap Index (pVHI).26 The Turkish pVHI, which is a valid and reliable instrument, includes 23 questions about the functional, physical, and emotional effects of a voice disorder.26 Parents were instructed to answer the questions by considering their child’s voice status during the previous month.26 Speech and voice samples were gathered audiovisually by the SLP in a sound-isolated room. Samples that consisted of recordings that were at least 2 minutes long included counting from 1 to 10, sustaining /a/ phonation, and connected speech.27 Recorded samples were evaluated in another session by two SLPs who were experienced in voice disorders. Recordings were listened to many times until the SLPs agreed on the values of G (grade), R (roughness), and B (breathiness) parameters.25 Grade represents the overall voice quality, roughness represents irregular fluctuations, and breathiness represents the turbulence of airflow in the voice sample.25 Those parameters were evaluated as they were thought to be more reliable in children.28,29
ARTICLE IN PRESS Fatma E.S.E.N. Aydinli et al
Glottal Stop Productions in Children with Cleft Palate
Instrumental evaluation of voice. Acoustic analysis. In the acoustic analysis of voice, the Computerized Speech Lab (CSL) (Model 4300 B) of Kay Elemetrics (Lincoln Park, NJ, USA) was used in conjunction with the Multi-Dimensional Voice Program (MDVP) software (Lincoln Park,NJ, USA). Recordings were made by the same SLP in a sound-isolated room. Participants were asked to sustain /a/ phonation in their conversational speaking pitch and loudness while in an upright position. The SLP held the microphone 15 cm away from the participant’s mouth.30 Before the recordings, the SLP (Manufacturer is the same Key Elemetrics, Lincoln Park,NJ, USA) explained the procedure to the child and gathered trial phonation for a minimum of two times. While analyzing the recordings, the first and last parts were cut and the middle 3 seconds of the phonation interval was analyzed.31 Among the parameters, fundamental frequency (F0), percentage of jitter (%), and the percentage of shimmer (%) were taken into account. Video laryngoscopy. This evaluation was performed via either flexible 3.7-mm diameter steerable fiber-optic laryngoscopy (Optim, Sturbridge, MA, USA) or rigid video laryngostroboscopy using a Kay Pentax digital strobe (Kay Pentax, Lincoln Park, NJ, USA). Kay Pentax Rls 9100 B equipment (Key Elemetrics, Lincoln Park, NJ, USA) was used to capture the images and record the voices. The evaluation method was determined by considering the participant’s age and compatibility with the test method. Images were examined in a separate session by an experienced laryngologist who was not informed about the patient’s condition. Participants who had vocal-fold lesions were excluded from the study. Velopharyngeal function measurements Velopharyngeal function measurements were done to reveal the difference between the two groups with regard to VPI status. Instrumental evaluation methods; nasopharyngoscopy 32 and nasometer33 were used for this purpose. Nasopharyngoscopy (flexible fiber-optic videoendoscopy). Nasopharyngoscopy evaluations were performed by an experienced laryngologist and an SLP. The 3.7-mm diameter flexible fiber-optic nasopharyngoscope was used (Optim, Sturbridge, MA, USA). During the evaluation, the participants were asked to repeat speech samples for a minimum of five times. These samples consisted of nonsense syllables, including high pressure consonants /p/, /t/, /k/, /s/, combined with /ς /vowel, meaningful word composed of high pressure consonants [kɯɾksɛcɪz], meaningful word composed of low pressure consonants [lʌlɛ], and counting from 1 to 10 with high pressure consonants. Audio and video recordings were saved in the device’s hard drive for the purpose of rating in another session. The recordings were evaluated by an SLP who had 8 years of experience in this field. Velopharyngeal closure was rated according to the following three classifications: inefficient, borderline, and efficient.32 Nasometer evaluation. The Kay Pentax Nasometer II (Model 6450) device (Kay Pentax, Lincoln Park, NJ, USA) was
3
used for the evaluation. The test was performed according to the manufacturer’s instructions.34 Although the nasometer protocol includes longer passages in general, shorter samples can be used.34 The speech samples used in this study were the same with those used in a Turkish normative study.35 The children were asked to repeat nonsense syllables, combining high-pressure consonants /p/, /t/, /k/, /s/, with vowels /ʌ/ and /Ι/. Every utterance was repeated for a minimum of eight times. Afterward, participants were asked to count from 1 to 10. Mean nasalance scores were recorded for analysis.
Statistical analysis The IBM SPSS Statistics 21 program (IBM corporation, Armonk, NY, USA) was used in this study’s statistical analysis. Numeric variables were represented by average, standard deviation, median, minimum and maximum values, and interquartile ranges. Normality of the data was investigated using Shapiro Wilk test. Although some of the data (F0, pΙ, tΙ, kΙ, counting) were found to be distributed normally, and majority were not distributed normally, nonparametric tests were used in the analyses. Categorical variables were represented by frequency and percentage. The relation between categorical variables was analyzed using the Chi-square test. The numeric variable differences between the two groups were analyzed using the Mann-Whitney U test. The median and interquartile range values are shown in the tables. The significance value was accepted as 0.05.
RESULTS Demographics of the participant Comparison of the age of palatal surgery The median age at the time of the cleft palate surgery was 10 months in the CG, and 12 months in the GSG. The difference was analyzed using the Mann-Whitney U test. There was not a statistically significant difference detected between the two groups (P > 0.05). Comparison of the speech therapy ratio The difference in the ratio of receiving speech therapy was analyzed using Chi-square test. In both groups, 47% (8/17) of the total participants received speech therapy.
Articulation test results The consonant production errors for both groups were analyzed according to error types and frequency. In the GSG, the glottal stops constituted 127 (32.5%) of the 390 total errors. Other compensatory errors consisted of 36 nasal fricatives (9.2%), 34 posterior nasal fricatives (8.7%), 33 pharyngeal fricatives (8.5%), 28 nasal consonant for oral pressure consonant (7.1%), and 12 middorsum palatal stops (3.1%). Other articulation errors included 61 oral misarticulation errors (15.7%) and 59 developmental articulation/phonological errors (15.1%). In the GSG, 14 participants exhibited glottal stop errors in 3–6 consonants, and 3 participants exhibited glottal stop errors in 7–9 consonants.
ARTICLE IN PRESS 4
Journal of Voice, Vol. ■■, No. ■■, 2015
TABLE 1. Comparison of the Acoustic Vocal Parameters (Analyzed by CSL) Between the Groups Control Group
Glottal Stop Group
Acoustic Parameters
(Median ± IQR)
(Median ± IQR)
P Value*
F0 (Hz) Jitter (%) Shimmer (%)
301.55 ± 43.31 0.56 ± 0.5 2.98 ± 2.11
313.02 ± 63.64 1.06 ± 1.19 3.67 ± 2.89
NS 0.049* NS
Notes: Table 1 shows the acoustic analysis results. Numerical values of all parameters were found to be higher in the GSG. Jitter parameter was also found to be statistically significant. * Significant at P < 0.05. Abbreviations: IQR, interquartile range, NS, nonsignificant.
Voice evaluations Comparison of the acoustic analysis between the groups The difference between the groups was analyzed using the MannWhitney U test. In a Turkish study,36 in searching for normal values in 203 healthy children between 4 and 14 years old, difference was not found in F0 between gender under the age of 13, and F0 values were not inspected with respect to gender. Similarly, two studies36,37 showed no age and sex differences in jitter and shimmer values. Therefore, we inspected these variables independent of arbitrary age categories. (Table 1). The F0, jitter, and shimmer values were found to be numerically higher in the GSG. The difference was statistically significant for the jitter value (P < 0.05). Comparison of the pVHI scores The subunits and total scores of the pVHI were compared using the Mann-Whitney U test between the groups. All the subunit scores and the total score for the GSG were found to be significantly higher than those of the CG (P < 0.05). (Table 2). The total score for the GSG was found to be approximately three times higher than that of the CG. Comparison of the GRBAS results The differences in the G, R, and B parameters between the groups were evaluated using the Chi-square test, and the difference was detected for the G and R parameters (P < 0.05). Chi-square coefficients and P values were found for the G, R, and B parameters,
respectively, (χ2 = 13.592, P = 0.000), (χ2 = 7.798, P = 0.005), (χ2 = 1.417, P = 0.234). The voice quality was perceived to have a mild grade of hoarseness in eight (47%) participants and with a mild grade of roughness in five (29.4%) participants of the GSG. The G and R parameters were perceived to be normal in all of the participants in the CG. The distribution of the B parameter was almost identical in both of the groups. The maximum value was 1 for both groups. Velopharyngeal function evaluations Comparison of the nasometer results The Mann-Whitney U test was used to determine the differences between the two groups. In every speech utterance, the nasalance scores were found to be significantly higher (P < 0.05) in the glottal stop compared with the CG, except for the sample of [kΙ]. Table 3 displays the comparison of the nasometer scores between the two groups. Velopharyngeal closure evaluation One participant’s velopharyngeal closure evaluation was excluded in the CG due to nonreliability. The Chi-square test was used to determine the difference between the velopharyngeal closure (χ2 = 23,733 and P = 0 < 0.01). In the CG, 11 (64.7%) had efficient closure, 4 (23.5%) had borderline closure, and 1 (5.8%) had inefficient closure. In the GSG, 12 (70.5%) had borderline closure, and 5 (29.5%) had inefficient closure. Therefore, the GSG had greater incidence of significant VPI than the CG.
TABLE 2. Comparison of the pVHI Scores: Subunits’ and Their Total Scores Between the Two Groups Control Group
Glottal Stop Group
Subunits of pVHI
(Median + IQR)
(Median + IQR)
P Value
Functional score Physical score Emotional score Total score
2 ± 5.0 2 ± 6.00 1 ± 3.0 5 ± 14.00
6 ± 4.50 5 ± 9.00 5 ± 8.50 16 ± 25.00
0.018* 0.005* 0.005† 0.014*
Notes: Table 2 displays the median and interquartile range values for pVHI subunits and their total score. In all parameters, GSG had higher values, and difference was statistically important. * Significant at P < 0.05. † Significant at P < 0.01. Abbreviations: IQR, interquartile range; M, median.
ARTICLE IN PRESS Fatma E.S.E.N. Aydinli et al
5
Glottal Stop Productions in Children with Cleft Palate
TABLE 3. Comparison of the Speech Utterances’ Nasometer Values Between the Two Groups
Speech Utterance Pʌ PΙ Tʌ TΙ Kʌ KΙ Sʌ SΙ Counting
Control Group
Glottal Stop Group
(Median + IQR)
(Median + IQR)
P Value
15.00 ± 27.00 26.00 ± 20.00 10.00 ± 30.00 30.00 ± 36.00 10.00 ± 19.00 31.00 ± 35.00 12.00 ± 40.00 38.00 ± 27.00 26.00 ± 25.00
35.00 ± 29.00 58.00 ± 46.00 41.00 ± 29.50 56.00 ± 29.00 33.00 ± 35.50 58.00 ± 43.50 40.00 ± 26.50 59.00 ± 29.00 56.00 ± 26.50
0.016* 0.006† 0.010† 0.04* 0.010† NS 0.011* 0.018* 0.001†
Notes: Table 3 displays the median and interquartile range of speech utterances’ nasometer values. In all utterances, GSG had higher values and difference was statistically important except for the sample of KΙ. * Significant at P < 0.05. † Significant at P < 0.01. Abbreviations: IQR, interquartile range; M, median; NS, nonsignificant.
DISCUSSION Many studies have been carried out to search for vocal-quality characteristics in children with VPI.6–8,19Among these studies, Zajac and Linville’s7 and Van Lierde et al’s8 studies included the multiparameter approach, whereas the studies of Lewis et al6 and Villafuerte-Gonzalez et al 19 included only acoustic analysis. It is recommended that voice evaluation include auditoryperceptual methods, self-assessment tools, and objective measures.38 Perceptual ratings are often used to validate objective measures.39 The present study included not only acoustic analysis, but also audio perceptual analysis and self-assessment tool. According to the pVHI scores, not only the total score, but also the functional, physical, and emotional subscale scores were found to be higher in the GSG. The numerical value for the total score was higher than 8.6 in the GSG, which was the limit determined in healthy subjects in the original validity and reliability study of this index.26 These results may indicate that parents of children in GSG perceive their children’s voice quality to be somewhat influenced. In the literature, only one study investigated voice performance by a self-assessment tool (pediatric voice outcomes survey in patients with VPI). Boseley and Hartnick found that scores were improved after surgery.40 Care should be taken when interpreting the pVHI results because there is a probability that the parents’ responses to the questions were based not only on their impression of their child’s voice, but also on the articulation and resonance characteristics of the child’s voice. This may arise from the fact that the pVHI was not designed to differentiate resonance disorders from articulation disorders or glottal dysphonias.40 The jitter and shimmer values found in the present study were not higher than the normative values determined in a Turkish normative study36 in both groups. However, higher values of F0, jitter, and shimmer in the GSG may indicate that the glottal stop articulations can be associated with vocal perturbations. This may result from the struggle to regulate laryngeal and respiratory aerodynamics.7 On the other hand, roughness value was found
to differ in GSG from CG, which is compatible with the significantly higher jitter value detected in GSG. Similarly, Zajac and Linville7 found significantly higher value in jitter, wherein they included 10 children with VPI and 5 healthy couples. Villafuerte-Gonzales et al19 concluded that vocal abnormalities expressed by higher values in F0, jitter, and shimmer in their study where they included 14 children with healthy couples. Yang et al41 found lower cepstral peak prominence values when comparing pre–post VPI surgery results for vowels. They suggested that the articulatory movements that have an effect on maintaining adequate intraoral pressure might be the underlying cause. Van Lierde et al8 supported the idea of decreased vocal quality results from the attempts to compensate VPI as a result of their acoustic analysis. The voice quality was perceived to have a mild grade of hoarseness and roughness in some participants of the GSG. This result may imply that irregularity in frequency was perceptually evident in some of the participants with glottal stop. However, it should be noted that the range of the values were very narrow (0–1) for all parameters. In the literature, similarly, Van Lierde et al8 found mild degree of disturbance in the G and R parameters. Zajac and Linville7 measured hoarseness rate at 1.90 points in a 7-point equal-interval scale. Similarly, D’Antonio et al42 rated voices audio-perceptually in an equal-interval scale in children with VPI. They reported hoarseness as the most frequent symptom, and abnormal pitch and reduced loudness as the most frequent tertiary symptoms in children with VPI. In this present study, listeners concluded that perceptual analysis was more difficult when there was an accompanying VPI. This can be a result of both hypernasality and nasal emission secondary to VPI, which can mask audible effects of breathiness. Breathiness was rated abnormal in a few participants in both groups. As Imatomi43 discussed, breathiness may conflict with hypernasality. Imatomi43 noted that when there is mild hypernasality, breathiness causes it to be perceived as more severe. Conversely, when there is severe hypernasality, breathiness causes it to be perceived as milder. Another theory suggested by Tavares et al44 posits that the
ARTICLE IN PRESS 6 posterior triangular slit glottic configuration may make it difficult to identify breathiness in childhood. In this study, velopharyngeal function evaluations revealed that there was a clear relationship between velopharyngeal function and the searched glottal stop production as generally accepted in the literature.12,14 Leder and Lerman45 showed through a spectrogram that severe hypernasality was associated with glottal stops. In healthy voice production, it is known that the respiration, phonation, resonance, and articulation systems must be in good balance. Titze and Laukkanen46 suggested that supraglottic vocal tract characteristics have a direct effect on vocal-fold oscillations. Hence, compensatory articulation productions may lead to conflict among the respiration, phonation, resonance, and articulation systems. Although glottal stops constitute the higher percentage of all compensatory errors, many of the participants in the present study exhibited two or more types of compensatory errors. Although it seems reasonable that glottal stops are the primary compensatory errors to have potential effect on voice because they occur at the laryngeal level, other errors may contribute to vocal quality by attempting to regulate pressure in the vocal tract. This factor makes reaching definitive conclusions difficult in this study. In future studies, it is reasonable to classify participants according to compensatory articulation error types and explore their aerodynamic profiles by measuring the subglottal pressure and laryngeal airflow rate.47 Besides, it may be more advantageous to design a long-term follow-up study. This is because the severity of glottal stops can change from time to time. It may be a better option to evaluate the participants’ voice performance periodically in relation to glottal stops. Furthermore, increasing the number of participants could contribute to the increase of the validity and generalizability of the study’s results. CONCLUSIONS These results may indicate that the vocal quality characteristics of children with glottal stops differ from children who do not have this type of production. The difference was most pronounced in jitter parameter of acoustic analysis. GRBAS showed difference in overall voice quality and roughness. pVHI showed that parents perceived their children’s voice quality to be somewhat influenced. Although these results support that children with glottal stops may be at more risk to develop vocal pathologies, other types of compensatory articulations should be taken into account in future studies with a higher number of participants. In conclusion, it is suggested that children with cleft palate who have glottal stop productions should receive a comprehensive speech and language pathology intervention including voice therapy techniques. REFERENCES 1. Dixon MJ, Marazita ML, Beaty TH, et al. Cleft lip and palate: synthesizing genetic and environmental influences. Nat Rev Gen. 2011;12:167–178. 2. Peterson-Falzone SJ, Hardin-Jones M, Karnell MP. Cleft Palate Speech. 3rd ed. St. Louis, MO: Mosby-Elsevier; 2001. 3. Kummer AW. Cleft Palate and Craniofacial Anomalies Effects on Speech and Resonance. 2nd ed. Albany, NY: Thomson Delmar Learning; 2008.
Journal of Voice, Vol. ■■, No. ■■, 2015 4. Peterson-Falzone SJ, Trost-Cardamone JE, Karnell MP, et al. Effect of cleft and non-cleft VPI on speech in older children. In: Falk K, ed. The Clinician’s Guide to Treating Cleft Palate Speech. St. Louis, MO: Mosby Elsevier; 2006:5–16. 5. McWilliams BJ, Bluestone CD, Musgrave RH. Diagnostic implications of vocal cord nodules in children with cleft palate. Laryngoscope. 1969;79:2072–2080. 6. Lewis JR, Andreassen ML, Leeper HA, et al. Vocal characteristics of children with cleft lip/palate and associated velopharyngeal incompetence. J Otolaryngol. 1993;22:113–117. 7. Zajac DJ, Linville RN. Voice perturbations of children with perceived nasality and hoarseness. Cleft Palate J. 1989;26:226–231. 8. Van Lierde KM, Claeys S, De Bodt M, et al. Vocal quality characteristics in children with cleft palate: a multiparameter approach. J Voice. 2004;18:354–362. 9. Hocevar-Boltezar I, Jarc A, Kozelj V. Ear, nose and voice problems in children with orofacial clefts. J Laryngol Otol. 2006;120:276–281. 10. Robison JG, Otteson TD. Prevalence of hoarseness in the cleft palate population. Arch Otolaryngol Head Neck Surg. 2011;137:74–77. 11. McWilliams BJ, Lavorato AS, Bluestone CD. Vocal cord abnormalities in children with velopharyngeal valving problems. Laryngoscope. 1973;83:1745–1753. 12. Warren DW. Compensatory speech behaviours in individuals with cleft palate: a regulation/control phenomenon? Cleft Palate J 1986;23:251–260. 13. Trindade IE, Paciello RZ, Trindade Junior AS, et al. Consonantal intraoral air pressure characteristics in Brazilian postoperative cleft palate speakers with velopharyngeal disorders. Braz J Med Biol Res. 1989;22:667–674. 14. Golding-Kushner KJ. Therapy Techniques for Cleft Palate Speech and Related Disorders. San Diego, CA: Singular Thomson Learning; 2001. 15. Kummer AW. Disorders of resonance and airflow secondary to cleft palate and/or velopharyngeal dysfunction. Semin Speech Lang. 2011;32:141– 149. 16. Gooch JL, Hardin-Jones M, Chapman KL, et al. Reliability of listener transcriptions of compensatory articulations. Cleft Palate J 2001;38:59–67. 17. Kummer AW. Speech therapy for errors secondary to cleft palate and velopharyngeal dysfunction. Semin Speech Lang. 2011;32:191–198. 18. Bispo NH, Whitaker ME, Aferri HC, et al. Speech therapy for compensatory articulations and velopharyngeal function: a case report. J Appl Oral Sci. 2011;19:679–684. 19. Villafuerte-Gonzalez R, Valadez-Jimenez VM, Hernandez-Lopez X, et al. Acoustic analysis of voice in children with cleft palate and velopharyngeal insufficiency. Int J Pediatr Otorhinolaryngol. 2015;79:1073–1076. 20. Hamming KK, Finkelstein M, Sidman JD. Hoarseness in children with cleft palate. Otolaryngol Head Neck Surg. 2009;140:902–906. 21. Van Lierde KM, Bonte K, Baudonck N, et al. Speech outcome regarding overall intelligibility, articulation, resonance and voice in Flemish children a year after pharyngeal flap surgery, a pilot study. Folia Phoniatr Logop. 2008;60:223–232. 22. Topbas¸ S. Türkçe Sesletim–Sesbilgisi Testi: Geçerlik-Güvenirlik ve Standardizasyon Çalıs¸ması. Türk Psikoloji Dergisi. 2006;21:39–56. 23. Trost JE. Articulatory additions to the classical description of the speech of persons with cleft palate. Cleft Palate J. 1981;18:193–203. 24. Henningsson G, Kuehn DP, Sell D, et al. Universal parameters for reporting speech outcomes in individuals with cleft palate. Cleft Palate J 2008;45: 1–17. 25. De Bodt MS, Wuyts FL, Van de Heyning PH, et al. Test-retest study of the GRBAS scale: influence of experience and professional background on perceptual rating of voice quality. J Voice. 1997;11:74–80. 26. Tadıhan Özkan E, Tüzüner A, Demirhan E, et al. Reliability and validity of the Turkish Pediatric Voice Handicap Index. Int J Pediatr Otorhinolaryngol. 2015; 79:680–684. 27. de Labio RB, Tavares EL, Alvarado RC, et al. Consequences of chronic nasal obstruction on the laryngeal mucosa and voice quality of 4- to 12-year-old children. J Voice. 2012;26:488–492. 28. Tavares EL, Brasolotto A, Santana MF, et al. Epidemiological study of dysphonia in 4–12 year-old children. Braz J Otorhinolaryngol. 2011;77:736– 746.
ARTICLE IN PRESS Fatma E.S.E.N. Aydinli et al
Glottal Stop Productions in Children with Cleft Palate
29. Schindler A, Tiddia C, Ghidelli C, et al. Adaptation and validation of the Italian Pediatric Voice Handicap Index. Folia Phoniatr Logop. 2011;63:9–14. 30. Souza LS, Crespo AN, Medeiros JLA. Laryngeal vocal and endoscopic alterations after thyroidectomy under local anesthesia and hypnosedation. Braz J Otorhinolaryngol. 2009;75:511–516. 31. Kent RD, Vorperian HK, Duffy JR. Reliability of the multi-dimensional voice program for the analysis of voice samples of subjects with dysarthria. Am J Speech Lang Pathol. 1999;8:129–136. 32. Antonio D, MuntzHR LL, Marsh JL, et al. Practical application of flexible fiberoptic nasopharyngoscopy for evaluating velopharyngeal function. Plast Reconst Surg. 1988;82:611–618. 33. Watterson T, Lewis KE, Foley-Homan N. Effect of stimulus length on nasalance scores. Cleft Palate J 1999;36:243–247. 34. Kay Pentax. Instruction Manual Nasometer II Model 6450, Issue C. Lincoln Park, NJ: KayPentax; 2010. 35. Sarac ET, Kayikci ME, Ozkan S. Nasality evaluation of Turkish phonemes in vowel-consonant combinations. Int J Pediatr Otorhinolaryngol. 2011;75:894–898. 36. Akmese PP, Kulak Kayikci ME, Atas A. Acoustic characteristics of Turkish speaking children ages between 4 and 14 years old. Int Adv Otol. 2012;8:399–406. 37. Maturo S, Hill C, Bunting G, et al. Establishment of a normative pediatric acoustic database. Arch Otolaryngol Head Neck Surg. 2012;138:956–961. 38. Ziwei Y, Zheng P, Pin D. Multiparameter voice assessment for voice disorder patients: a correlation analysis between objective and subjective parameters. J Voice. 2014;28:770–774.
7
39. Hillebrand J, Cleveland RA, Erickson RL. Acoustic correlates of breathy vocal quality: dysphonic voices and continuous speech. J Speech Hear Res. 1994;39:311–321. 40. Boseley ME, Hartnick CJ. Assessing the outcome of surgery to correct velopharyngeal insufficiency with the pediatric voice outcomes survey. Int J Pediatr Otorhinolaryngol. 2004;68:1429–1433. 41. Yang Z, Fan J, Tian J, et al. Cepstral analysis of voice in children with velopharyngeal insufficiency after cleft palate surgery. J Voice. 2014;28:789– 792. 42. D’Antonio LL, Muntz HR, Province MA, et al. Laryngeal/voice findings in patients with velopharyngeal dysfunction. Laryngoscope. 1988;98:432– 438. 43. Imatomi S. Effects of breathy voice source on ratings of hypernasality. Cleft Palate J 2005;42:641–648. 44. Tavares EL, Labio RB, Martins RH. Normative study of vocal acoustic parameters from children from 4 to 12 years of age without vocal symptoms: a pilot study. Braz J Otorhinolaryngol. 2010;76:485–490. 45. Leder SB, Lerman JW. Some acoustic evidence for vocal abuse in adult speakers with repaired cleft palate. Laryngoscope. 1985;95:837–840. 46. Titze IR, Laukkanen AM. Can vocal economy in phonation be increased with an artificially lengthened vocal tract? A computer modeling study. Logoped Phoniatr Vocol. 2007;32:147–156. 47. Gillespie AI, Gardner-Schmidt J, Rubinstein EN, et al. Aerodynamic profiles of women with muscle tension dysphonia/aphonia. J Speech Lang Hear Res. 2013;56:481–488.