Original Paper Folia Phoniatr Logop 2013;65:143–147 DOI: 10.1159/000355913
Published online: November 26, 2013
Vocal Characteristics during Child Development: Perceptual-Auditory and Acoustic Data Leonardo Wanderley Lopes Ivonaldo Leidson Barbosa Lima Elma Heitmann Mares Azevedo Maria Fabiana Bonfim de Lima-Silva Débora Pontes Cavalcante Larissa Nadjara Alves de Almeida Anna Alice Figueirêdo de Almeida Department of Speech and Language Pathology, Federal University of Paraíba, João Pessoa, Brazil
Key Words Voice quality · Dysphonia · Auditory perception · Acoustics · Child
Abstract Objective: To analyze perceptual-auditory and acoustic characteristics of children’s voices of different age ranges. Patients and Methods: Ninety-three 3- to 10-year-old children grouped from 3 to 5, 6 to 7, and 8 to 10 years served as participants. The severity of vocal deviation and the parameters of roughness, breathiness, strain, and instability were assessed using a visual analog scale. We calculated the mean and standard deviation of fundamental frequency (F0), jitter, shimmer, and glottal-to-noise excitation ratio for the sustained vowel, and the mean of F0 variability for connected speech. Results: The most affected voices were in the age range 8–10 years, and only the phonation tension level was reduced as a result of aging. There were significant differences between children aged 3–5 years and the other age ranges for F0 mean for sustained vowels and F0 variability. Conclusion: Children aged 8–10 years had the highest severity of vocal deviation. There was a significant reduction of phonation tension and measure of F0, jitter, and shimmer after the age of 5 years. © 2013 S. Karger AG, Basel
© 2013 S. Karger AG, Basel 1021–7762/13/0653–0143$38.00/0 E-Mail [email protected] www.karger.com/fpl
Introduction
Voice problems are the most common communication problems during the development of the individual. During childhood, the prevalence of dysphonia ranges from 0.5 to 23% [1–3]. Dysphonia may have a negative impact on the child’s overall health, effective communication, educational development, self-esteem, self-image, and even participation in social activities. Therefore, it is important to develop educational programs to promote vocal health that include parents and educators, who are not aware of their children’s or students’ vocal disorders, in addition to developing assessment tools for early detection of vocal changes in childhood; this may prevent negative effects on development. As voice is a multidimensional function, effective vocal assessment of children involves perceptual, acoustic, aerodynamic, and laryngological measures and self-assessment, which is influenced by anatomic-physiologic, emotional, behavioral, organic, and environmental aspects. Therefore, the assessment must map and correlate most vocal characteristics to provide an overall understanding of the real cause and impact of dysphonia [4]. Nevertheless, one of the main difficulties of vocal assessment is the definition of ‘abnormality’ for this age Leonardo Wanderley Lopes Departamento de Fonoaudiologia, Centro de Ciências da Saúde Universidade Federal da Paraíba, Cidade Universitária, Campus I Castelo Branco, João Pessoa, PB 58051-900 (Brazil) E-Mail lwlopes @ hotmail.com
range, given that children’s vocal production is unstable. For example, tension and breathiness are expected in children due to neuromuscular immaturity and the rudimentary structure of the larynx [5]. Although the importance of characterizing the pediatric voice with subjective and objective data is widely recognized (as this would lead to better identification of vocal deviation and monitoring of the progression of vocal therapy), there are few publications with normal-range data for voice in childhood [6]. This survey study is part of an epidemiological study that aims to create a database of normative acoustic and perceptual parameters throughout the life cycle, from infancy to senescence. The results of this survey may allow us to propose measures that may be used in screening procedures as well as diagnosis and monitoring of voice disorders at different ages through comparisons with normative data. Within this context, we analyzed the perceptual-auditory and acoustic characteristics of the voices of children aged 3–10 years with no previous diagnosis of vocal disorder. This research will provide a better understanding of vocal manifestations throughout child development.
Methods This study was quantitative, descriptive, and cross-sectional. The design and procedures were assessed and approved by the Research Ethics Committee of the University Hospital Lauro Wanderley and the Federal University of Paraiba in Brazil (HULW/ UFPB), protocol No. 775/10. Participants Ninety-three children aged 3–10 years participated in the study (48 girls and 45 boys). All were students in a school managed by the federal educational authorities. Seventy-one of these children had already participated in a previous, related study [7]. Children with cognitive disorders, those with an upper airway infection during data collection, or those who did not perform the requested activities were excluded from the study. Recordings and Procedures Before starting data collection, we visited the site where the analyses were to be conducted to find a silent area (environmental noise below 50 dB sound pressure level). On the same occasion, the school principal was informed about the procedures and objectives of the study. She agreed to the study protocol and gave her authorization. Later on, we submitted the informed consent form to the parents of potential participants. After the informed consent forms were signed and returned, we visited the school again to record the children’s voices at predefined times, as per our agreement with the school supervisor. The teachers sent the children individually to the recording room
144
Folia Phoniatr Logop 2013;65:143–147 DOI: 10.1159/000355913
during the first hours of class in the morning (from 8 a.m. to 10 a.m.) and the afternoon (from 1 p.m. to 3 p.m.). The recording sessions lasted a maximum of 5 min. They began with a sustained vowel /ε/ production on maximum phonation time, followed by connected speech counting from 1 to 10. Data collection was carried out by using an HP notebook computer (Hewlett-Packard Development Co., Palo Alto, Calif., USA) and a Logitech headset microphone (Logitech, Fremont, Calif., USA). We used PRAAT version 5.1.44 software (by P. Boersma and D. Weenink, University of Amsterdam, The Netherlands). The sampling rate was 44,100 Hz. The recordings were made from March to October 2012. Later, the voices were edited using Sound Forge software version 10.0 (Sony, Tokyo, Japan). The initial and final 2 s of the vocal emissions were eliminated due to large irregularities. Each production was at least 3 s. The sustained vowel and connected speech were normalized using the normalize control of Sound Forge at peak level mode. Thus, the audio output was standardized between –6 and 6 dB. Perceptual-Auditory Analysis For the perceptual-auditory voice analysis, a visual analog scale with 0- to 100-mm metrics was used. We assessed the severity (overall level) of vocal deviation (GG), roughness (RG), breathiness (BG), strain (SG), and instability (IG). The closer to zero, the smaller the vocal abnormality; the closer to 100, the greater the vocal deviation. The assessment was based on the consensus of 3 voice specialists (speech therapists) experienced in vocal perceptual-auditory assessment. The perceptual assessment session occurred in a silent room. Initially, the judges were told that the voices should be considered within the normal variability of vocal quality (NVVQ) when the issue was socially acceptable for a child. They were also advised that roughness corresponded to the presence of vibratory irregularities, breathiness was related to the exhaust air of the audible voice, strain corresponded to the perception of vocal effort, and instability would be identified by the presence of fluctuating voice quality and frequency over time. Each sustained vowel emission was presented 3 times over loudspeakers at a comfortable level, as reported by the examiners. Next, they identified the presence or absence of vocal deviation (NVVQ); if there was any deviation, they classified the voice according to its predominant type (rough, breathy, strained, or unstable) and severity of vocal deviation for each studied parameter (i.e., GG, RG, BG, SG, and IG). At the end of the perceptual analysis session, 10% of the samples of the vowel /ε/ were repeated randomly for a reliability analysis of the trial by a consensus of the evaluators through Cohen’s kappa coefficient. The kappa value was 0.80, indicating good interrater agreement. The examiners assessed correspondence between the visual analog scale and a numeric scale. Voice samples were categorized as grade 1 (0–35.5 mm), NVVQ or absence of deviation; grade 2 (35.6–50.5 mm), mild deviation; grade 3 (50.6–90.5 mm), or grade 4 (90.6–100 mm) [4, 8]. Acceptable variations that may represent vocal style, age, vocal use preference, professional characterization, or a mild vocal disorder were included in the NVVQ category [8].
Lopes et al.
Acoustic Analysis The acoustic analysis was carried out with the vocal analysis and vocal quality modules of VoxMetria software version 4.5h (CTS, Parana, Brazil). To that end, the following measures were extracted: the mean and standard deviation of fundamental frequency (F0), jitter, shimmer, and glottal-to-noise excitation ratio of a sustained vowel, as well as the F0 mean and variability of connected speech (i.e., number counting). We conducted a descriptive statistical analysis of the studied variables. Analysis of variance (ANOVA) was used to compare the perceptual-auditory and acoustic measures according to vocal deviation level and age ranges in the sample. We used Tukey’s Honestly Significant Difference test for the post hoc comparisons. The α level was 0.05. We used STATISTICA version 6.0 software (StatSoftInc, Tulsa, Okla., USA) for the statistical analyses.
Results
In the analysis of the perceptual parameters according to age group, there were significant between-group differences in the degree of vocal deviation (GG, p = 0.003) and vocal strain (SG, p = 0.001, table 1). There were significant differences in GG between the children aged 6–7 years and those aged 8–10 years (p = 0.003), and between 3–5 year-old children and both the 6- to 7- (p = 0.03) and the 8- to 10-year-olds (p = 0.001). Thus, there was a progressive reduction of the phonation strain level as a result of increased age. In the statistical analysis of the acoustic parameters according to age, between-group differences were detected in the mean F0 (p < 0.001), jitter (p = 0.018) and shimmer (p = 0.001) of the sustained vowel (p < 0.001) and the mean (p < 0.001) and variability (p = 0.003) of F0 in connected speech (table 2). Post hoc analyses revealed differences between the children aged 3–5 years and both the 6- to 7- (p = 0.04) and 8- to 10-year-old children (p = 0.001) in average F0 of the sustained vowel. There were differences between the 3- to 5-year-old children and the 6- to 7- (p = 0.001) and 8- to 10-year-old children (p = 0.001) in the average F0 in connected speech. With regard to the variability in connected speech F0, significant differences were observed between the 3- to 5-year-old group and the 6- to 7- (p = 0.007) and 8- to 10-year-old group (p = 0.009). Overall, there was a greater decrease in F0 after 5 years of age. Regarding jitter values, there was a significant difference between children aged 3–5 years and children aged 7–8 years (p = 0.02). The shimmer values were significantly different for children aged 3–5 years and children aged 8–10 years (p = 0.01). Vocal Characteristics in Children
Table 1. Visual analog scale values (mean ± SD) and statistical re-
sults of ANOVAs for each perceptual variable, according to age Variables
3 – 5 years
6 – 7 years
8 – 10 years
p value
GG RG BG SG IG
43.07 ± 4.39 41.19 ± 14.83 37.11 ± 22.27 48.03 ± 11.67 48.26 ± 9.59
40.31 ± 5.74 39.13 ± 14.96 30.27 ± 19.37 37.22 ± 15.96 44.13 ± 11.52
46.73 ± 9.03 44.47 ± 14.21 40.34 ± 17.44 34.41 ± 15.36 45.34 ± 11.77
0.003* 0.338 0.140 0.001* 0.403
Significant values: * p < 0.05 (ANOVA).
Discussion
The high prevalence of dysphonia in children requires special attention in the assessment and diagnosis of children’s voices. The use of objective measures may provide a better understanding of the severity of vocal deviations and their manifestations at different age ranges from 3 to 9 years [6, 7, 9, 10]. In our statistical analysis of perceptual parameters according to age range, there were more abnormal voices in the age range 8–10 years as compared to the other two age groups (p = 0.003, table 1). However, there was a progressive reduction of the phonation strain level as a result of aging (p = 0.001, table 1). The 8- to 10-year-old children exhibited an increase in the level of roughness and breathiness compared to the other age ranges, even though the comparisons were not statistically significant. The grade of phonation tension proved to be high in the early childhood years, compatible with the phonation effort required to project the voice. The voice quality changes are normally associated with an increase in vocal fold mass or abnormalities in glottic closure, the main determining factors for vocal disorders. In general, glottic chinks caused by glottic proportion deviations, such as in pediatric larynges, may overload the phonation mechanism in an attempt to increase vocal intensity, especially in group activities and in the presence of competitive background noise [11–16]. Moreover, the use of a higher F0 in the early childhood years also predisposes children to increased phonation strain. The repetitive use of a hyperfunctional pattern may even cause the development of vocal fold damage such as vocal nodules or mucosa thickness, laryngeal lesions commonly found in epidemiological studies [5]. With regard to the acoustic measures, there were significant differences in the mean F0 for sustained vowel Folia Phoniatr Logop 2013;65:143–147 DOI: 10.1159/000355913
145
Table 2. Mean values (±SD) for each acoustic measure and statistical results of ANOVAs, according to age
Variables
3 – 5 years
6 – 7 years
8 – 10 years
p value
F0 sustained vowel F0 SD F0 mean counting F0 Var counting Jitter Shimmer GNE
271.59 ± 37.87 5.61 ± 3.12 277.68 ± 30.6 205.88 ± 81.32 1.14 ± 1.91 8.02 ± 4.68 0.77 ± 0.21
242.53 ± 22.84 5.05 ± 2.72 242.43 ± 26.99 130.53 ± 69.48 0.27 ± 0.13 6.09 ± 2.13 0.83 ± 0.12
244.08 ± 28.09 4.29 ± 4.13 248.6 ± 34.39 143.99 ± 89.53 0.5 ± 0.67 5.74 ± 2.7 0.75 ± 0.19
0.000* 0.313 0.000* 0.003* 0.018* 0.001* 0.252
Significant values: * p < 0.05 (ANOVA). F0 Var = Fundamental frequency variability; GNE = glottal-to-noise excitation ratio.
(p < 0.001), mean frequency of connected speech (p < 0.001), and the variability of F0 in connected speech between early childhood (3–5 years) and the other age ranges; lower voice F0 and less variability was found after 5 years of age (table 2). F0 is closely related to length, mass, tension, vocal fold rigidity, and subglottic pressure [17]. Some studies [6, 17, 18] have shown a decrease in F0 standard throughout development, explained by an increase in vocal fold length and mass, expansion and development of the vocal tract, and the approximation of male and female voice-related adult models. A recent study [6] demonstrated that there is no typical F0 for each age range and that there is no correct age for changes to occur, but that there are periods of mild changes for both boys and girls, starting between 11 and 12 years of age. In the present study, there were significant changes after 5 years, corresponding to the end of early childhood, although we have not studied older children to determine whether there are also significant changes in early adolescence. The values for jitter (p = 0.018) and shimmer (p = 0.001) were significantly different for the 3- to 5-year-old children compared to the 6- to 7- and 8- to 10-year-old children (table 2). The normal range for jitter and shimmer for a sustained vowel /ε/ in the software used varied from 0 to 0.6% and 0 to 6.5%, respectively. Thus, the major difference between the early childhood years and other age ranges is that these values are the standard variation for the human voice in the age ranges 6–7 and 8–10 years. It is difficult to make comparisons between these data and the other standard ranges reported in the literature [6], which recommend jitter and shimmer values of 1.8 and 4–5%, respectively, for children; this is because, in most 146
Folia Phoniatr Logop 2013;65:143–147 DOI: 10.1159/000355913
of the studies, the vowel /a/ was used. The use of a different vowel may explain the difference in values. Jitter and shimmer may help the description and quantification of perceptual characteristics and vocal fold vibration patterns. They are widely used in clinical practice and research studies. They are descriptive and can be used for diagnostic purposes. Moreover, they provide data relevant to the assessment and screening process for vocal disorders and are measures used for documentation and monitoring of vocal treatment approaches [19–22]. Small irregularities of the sound signal are considered the standard variation associated with the physiological mechanism of vocal production. However, the increase in the values of acoustic signal perturbation may indicate an inappropriate operation of the glottic mechanism, which may be used to detect the presence of vocal abnormality. It is important to bear in mind that children’s vocal folds have not completed the process of differentiation of the intermediate and deep layers of the lamina propria [14]. Moreover, the children are still developing their neuromuscular control, which means that abnormal measures of perturbation/irregularity in children are to be expected, especially during early childhood. The normal operating range for jitter and shimmer in children should take into account the anatomic and physiologic characteristics of this important developmental stage [23]. The normal-range values used to classify a regular voice versus a deviant voice in adults might not be effective in discriminating voices in children. For this reason, it is important to carry out research studies that analyze acoustic measures at different age ranges during the childhood years associated with auditory perceptual analysis of the voices. Even though we found higher values for jitter, shimmer, F0, and phonation strain in early childhood (3–5 Lopes et al.
years), the most disturbed voices were detected in the age range of 8–10 years. This finding reflects the fact that there is greater vulnerability for children below 5 years to develop laryngeal disorders because of the recurrence of upper airway infections during this period and the immaturity of vocal tract structures. The presence of a vocal abnormality in school-aged children seems to be the result of chronic and excessive voice abuse in group activities, which can easily be noticed at about age 10 [3]. During childhood, especially by its end, the child has likely participated in group activities that involve excessive and abusive use of voice such as school and family, posing a risk for the development of vocal disorders caused by an overload of the phonation system and an imbalance in phonation physiology. If this pattern of behavior is maintained, it may lead to additional vocal fold lesions [24]. For this reason, it is necessary to create strategies to promote voice preservation in children. These strategies should be sensitive to the child’s environment and include educational measures that can improve and optimize oral
communication, especially during early childhood when phonation is made with additional effort. These strategies may prevent the development and fixation of inappropriate laryngeal adjustments in the preadolescent years, which could be maintained into adulthood.
Conclusion
It was confirmed that children in the age range of 8–10 years had the highest severity of vocal deviation. There was a significant reduction of phonation strain and measure of F0 and jitter after 5 years. Shimmer was significantly lower after the age of 8.
Disclosure Statement The authors have no conflicts of interest to declare.
References 1 Tavares ELM, Labio RB, Martins RHG: 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. 2 Carding P, Roulstone S, Northstone K, ALSPAC Study team: The prevalence of childhood dysphonia: a cross-sectional study. J Voice 2006;20:623–630. 3 Sederholm E, McAllister A, Dalkvist J, Sundberg J: Aetiologic factors associated with hoarseness in ten-year-old children. Folia Phoniatr Logop 1995;47:262–278. 4 Dejonckere PH, Bradley P, Clemente P, Cornut G, Crevier-Buchman L, Friedrich G, Van De Heyning P, Remacle M, Woisard V: A basic protocol for functional assessment of voice pathology, especially for investigating the efficacy of (phonosurgical) treatments and evaluating new assessment techniques. Eur Arch Otorhinolaryngol 2001;258:77–82. 5 Tavares ELM, Brasolotto A, Santana MF, Padovan CA, Martins RHG: Epidemiological study of dysphonia in 4–12 year-old children. Braz J Otorhinolaryngol 2011; 77: 736– 746. 6 Maturo S, Hill C, Bunting G, Ballif C, Maurer R, Hartnick C: Establishment of a normative pediatric acoustic database. Arch Otolaryngol Head Neck Surg 2012;138:956–961. 7 Lopes LW, Barbosa Lima IL, Alves Almeida LN, Cavalcante DP, de Almeida AA: Severity of voice disorders in children: correlations between perceptual and acoustic data. J Voice 2012;26:819.e7–819.e12.
Vocal Characteristics in Children
8 Yamasaki R, Leo SHS, Madazio G, Padovani M, Azevedo R, Behlau M: Correspondence between visual analog scale and the numerical rating scale perceptual-hearing voices. 16th Brazilian Congress of Speech, Campos de Jordao. San Paulo, Brazilian Society of Speech, 2008, pp 1080–1084. 9 Niedzielska G: Acoustic analysis in the diagnosis of voice disorders in children. Int J Pediatr Otorhinolaryngol 2001;57:189–193. 10 Martins RHG, Ribeiro CBH, Mello BMZF, Branco A, Tavares EL: Dysphonia in children. J Voice 2012;26:674.e17–674.e20. 11 Uloza V, Verikas A, Bacauskiene M, Gelzinis A, Pribuisiene R, Kaseta M, Saferis V: Categorizing normal and pathological voices: automated and perceptual categorization. J Voice 2011;25:700–708. 12 Trani M, Ghidini A, Bergamini G, Presuti L: Voice therapy in pediatric functional dysphonia: a prospective study. Int J Pediatr Otorhinolaryngol 2007;71:379–384. 13 Connor NP, Cohen SB, Theis SM, Thibeault SL, Heatley DG, Bless DM: Attitudes of children with dysphonia. J Voice 2008; 22: 197– 209. 14 McAllister AM, Granqvist S, Sjölander P, Sundberg J: Child voice and noise: a pilot study of noise in day cares and the effects on 10 children’s voice quality according to perceptual evaluation. J Voice 2009; 23: 587– 593. 15 Roy N, Holt KI, Redmond S, Muntz H: Behavioral characteristics of children with vocal fold nodules. J Voice 2007;21:157–168.
16 Van Houte E, Van Lierde K, Claeys S: Pathophysiology and treatment of muscle tension dysphonia: a review of the current knowledge. J Voice 2011;25:202–207. 17 Campisi P, Tewfik TL, Pelland-Blais E, Husein M, Sadeghi N: MultiDimensional Voice Program analysis in children with vocal cord nodules. J Otolaryngol 2000;29:302–308. 18 Nicollas R, Garrel R, Ouaknine M, Giovani A, Nazarian B, Triglia JM: Normal voice in children between 6 and 12 years of age: database and nonlinear analysis. J Voice 2008;22:671–675. 19 Ma EPM, Yu EML: Multiparametric evaluation of dysphonic severity. J Voice 2006; 20: 380–390. 20 Araújo SA, Grellet M, Pereira JC: Standardization of acoustic measures of the normal voice. Br J Otorhinolaryngol 2002;68:540–544. 21 Sader RCM, Hanayama EM: Theoretical considerations on the acoustic approach of the infant voice. Rev CEFAC 2004;6:312–318. 22 Brockmann M, Drinnan MJ, Storck C, Carding P: Reliable jitter and shimmer measurements in voice clinics: the relevance of vowel, gender, vocal intensity, and fundamental frequency effects in a typical clinical task. J Voice 2011;25:44–53. 23 McAllister A, Sederholm E, Ternström S, Sundberg J: Perturbation and hoarseness: a pilot study of six children’s voices. J Voice 1996;10:252–261. 24 Takeshita TK, Ricz LA, Isaac ML, Ricz H, Lima WA: Comportamento vocal de crianças em idade pré-escolar. Arq Int Otorrinolaringol/ Int Arch Otorhinolaryngol 2009;13:252–258.
Folia Phoniatr Logop 2013;65:143–147 DOI: 10.1159/000355913
147
Copyright: S. Karger AG, Basel 2014. Reproduced with the permission of S. Karger AG, Basel. Further reproduction or distribution (electronic or otherwise) is prohibited without permission from the copyright holder.
Published online: November 26, 2013
Vocal Characteristics during Child Development: Perceptual-Auditory and Acoustic Data Leonardo Wanderley Lopes Ivonaldo Leidson Barbosa Lima Elma Heitmann Mares Azevedo Maria Fabiana Bonfim de Lima-Silva Débora Pontes Cavalcante Larissa Nadjara Alves de Almeida Anna Alice Figueirêdo de Almeida Department of Speech and Language Pathology, Federal University of Paraíba, João Pessoa, Brazil
Key Words Voice quality · Dysphonia · Auditory perception · Acoustics · Child
Abstract Objective: To analyze perceptual-auditory and acoustic characteristics of children’s voices of different age ranges. Patients and Methods: Ninety-three 3- to 10-year-old children grouped from 3 to 5, 6 to 7, and 8 to 10 years served as participants. The severity of vocal deviation and the parameters of roughness, breathiness, strain, and instability were assessed using a visual analog scale. We calculated the mean and standard deviation of fundamental frequency (F0), jitter, shimmer, and glottal-to-noise excitation ratio for the sustained vowel, and the mean of F0 variability for connected speech. Results: The most affected voices were in the age range 8–10 years, and only the phonation tension level was reduced as a result of aging. There were significant differences between children aged 3–5 years and the other age ranges for F0 mean for sustained vowels and F0 variability. Conclusion: Children aged 8–10 years had the highest severity of vocal deviation. There was a significant reduction of phonation tension and measure of F0, jitter, and shimmer after the age of 5 years. © 2013 S. Karger AG, Basel
© 2013 S. Karger AG, Basel 1021–7762/13/0653–0143$38.00/0 E-Mail [email protected] www.karger.com/fpl
Introduction
Voice problems are the most common communication problems during the development of the individual. During childhood, the prevalence of dysphonia ranges from 0.5 to 23% [1–3]. Dysphonia may have a negative impact on the child’s overall health, effective communication, educational development, self-esteem, self-image, and even participation in social activities. Therefore, it is important to develop educational programs to promote vocal health that include parents and educators, who are not aware of their children’s or students’ vocal disorders, in addition to developing assessment tools for early detection of vocal changes in childhood; this may prevent negative effects on development. As voice is a multidimensional function, effective vocal assessment of children involves perceptual, acoustic, aerodynamic, and laryngological measures and self-assessment, which is influenced by anatomic-physiologic, emotional, behavioral, organic, and environmental aspects. Therefore, the assessment must map and correlate most vocal characteristics to provide an overall understanding of the real cause and impact of dysphonia [4]. Nevertheless, one of the main difficulties of vocal assessment is the definition of ‘abnormality’ for this age Leonardo Wanderley Lopes Departamento de Fonoaudiologia, Centro de Ciências da Saúde Universidade Federal da Paraíba, Cidade Universitária, Campus I Castelo Branco, João Pessoa, PB 58051-900 (Brazil) E-Mail lwlopes @ hotmail.com
range, given that children’s vocal production is unstable. For example, tension and breathiness are expected in children due to neuromuscular immaturity and the rudimentary structure of the larynx [5]. Although the importance of characterizing the pediatric voice with subjective and objective data is widely recognized (as this would lead to better identification of vocal deviation and monitoring of the progression of vocal therapy), there are few publications with normal-range data for voice in childhood [6]. This survey study is part of an epidemiological study that aims to create a database of normative acoustic and perceptual parameters throughout the life cycle, from infancy to senescence. The results of this survey may allow us to propose measures that may be used in screening procedures as well as diagnosis and monitoring of voice disorders at different ages through comparisons with normative data. Within this context, we analyzed the perceptual-auditory and acoustic characteristics of the voices of children aged 3–10 years with no previous diagnosis of vocal disorder. This research will provide a better understanding of vocal manifestations throughout child development.
Methods This study was quantitative, descriptive, and cross-sectional. The design and procedures were assessed and approved by the Research Ethics Committee of the University Hospital Lauro Wanderley and the Federal University of Paraiba in Brazil (HULW/ UFPB), protocol No. 775/10. Participants Ninety-three children aged 3–10 years participated in the study (48 girls and 45 boys). All were students in a school managed by the federal educational authorities. Seventy-one of these children had already participated in a previous, related study [7]. Children with cognitive disorders, those with an upper airway infection during data collection, or those who did not perform the requested activities were excluded from the study. Recordings and Procedures Before starting data collection, we visited the site where the analyses were to be conducted to find a silent area (environmental noise below 50 dB sound pressure level). On the same occasion, the school principal was informed about the procedures and objectives of the study. She agreed to the study protocol and gave her authorization. Later on, we submitted the informed consent form to the parents of potential participants. After the informed consent forms were signed and returned, we visited the school again to record the children’s voices at predefined times, as per our agreement with the school supervisor. The teachers sent the children individually to the recording room
144
Folia Phoniatr Logop 2013;65:143–147 DOI: 10.1159/000355913
during the first hours of class in the morning (from 8 a.m. to 10 a.m.) and the afternoon (from 1 p.m. to 3 p.m.). The recording sessions lasted a maximum of 5 min. They began with a sustained vowel /ε/ production on maximum phonation time, followed by connected speech counting from 1 to 10. Data collection was carried out by using an HP notebook computer (Hewlett-Packard Development Co., Palo Alto, Calif., USA) and a Logitech headset microphone (Logitech, Fremont, Calif., USA). We used PRAAT version 5.1.44 software (by P. Boersma and D. Weenink, University of Amsterdam, The Netherlands). The sampling rate was 44,100 Hz. The recordings were made from March to October 2012. Later, the voices were edited using Sound Forge software version 10.0 (Sony, Tokyo, Japan). The initial and final 2 s of the vocal emissions were eliminated due to large irregularities. Each production was at least 3 s. The sustained vowel and connected speech were normalized using the normalize control of Sound Forge at peak level mode. Thus, the audio output was standardized between –6 and 6 dB. Perceptual-Auditory Analysis For the perceptual-auditory voice analysis, a visual analog scale with 0- to 100-mm metrics was used. We assessed the severity (overall level) of vocal deviation (GG), roughness (RG), breathiness (BG), strain (SG), and instability (IG). The closer to zero, the smaller the vocal abnormality; the closer to 100, the greater the vocal deviation. The assessment was based on the consensus of 3 voice specialists (speech therapists) experienced in vocal perceptual-auditory assessment. The perceptual assessment session occurred in a silent room. Initially, the judges were told that the voices should be considered within the normal variability of vocal quality (NVVQ) when the issue was socially acceptable for a child. They were also advised that roughness corresponded to the presence of vibratory irregularities, breathiness was related to the exhaust air of the audible voice, strain corresponded to the perception of vocal effort, and instability would be identified by the presence of fluctuating voice quality and frequency over time. Each sustained vowel emission was presented 3 times over loudspeakers at a comfortable level, as reported by the examiners. Next, they identified the presence or absence of vocal deviation (NVVQ); if there was any deviation, they classified the voice according to its predominant type (rough, breathy, strained, or unstable) and severity of vocal deviation for each studied parameter (i.e., GG, RG, BG, SG, and IG). At the end of the perceptual analysis session, 10% of the samples of the vowel /ε/ were repeated randomly for a reliability analysis of the trial by a consensus of the evaluators through Cohen’s kappa coefficient. The kappa value was 0.80, indicating good interrater agreement. The examiners assessed correspondence between the visual analog scale and a numeric scale. Voice samples were categorized as grade 1 (0–35.5 mm), NVVQ or absence of deviation; grade 2 (35.6–50.5 mm), mild deviation; grade 3 (50.6–90.5 mm), or grade 4 (90.6–100 mm) [4, 8]. Acceptable variations that may represent vocal style, age, vocal use preference, professional characterization, or a mild vocal disorder were included in the NVVQ category [8].
Lopes et al.
Acoustic Analysis The acoustic analysis was carried out with the vocal analysis and vocal quality modules of VoxMetria software version 4.5h (CTS, Parana, Brazil). To that end, the following measures were extracted: the mean and standard deviation of fundamental frequency (F0), jitter, shimmer, and glottal-to-noise excitation ratio of a sustained vowel, as well as the F0 mean and variability of connected speech (i.e., number counting). We conducted a descriptive statistical analysis of the studied variables. Analysis of variance (ANOVA) was used to compare the perceptual-auditory and acoustic measures according to vocal deviation level and age ranges in the sample. We used Tukey’s Honestly Significant Difference test for the post hoc comparisons. The α level was 0.05. We used STATISTICA version 6.0 software (StatSoftInc, Tulsa, Okla., USA) for the statistical analyses.
Results
In the analysis of the perceptual parameters according to age group, there were significant between-group differences in the degree of vocal deviation (GG, p = 0.003) and vocal strain (SG, p = 0.001, table 1). There were significant differences in GG between the children aged 6–7 years and those aged 8–10 years (p = 0.003), and between 3–5 year-old children and both the 6- to 7- (p = 0.03) and the 8- to 10-year-olds (p = 0.001). Thus, there was a progressive reduction of the phonation strain level as a result of increased age. In the statistical analysis of the acoustic parameters according to age, between-group differences were detected in the mean F0 (p < 0.001), jitter (p = 0.018) and shimmer (p = 0.001) of the sustained vowel (p < 0.001) and the mean (p < 0.001) and variability (p = 0.003) of F0 in connected speech (table 2). Post hoc analyses revealed differences between the children aged 3–5 years and both the 6- to 7- (p = 0.04) and 8- to 10-year-old children (p = 0.001) in average F0 of the sustained vowel. There were differences between the 3- to 5-year-old children and the 6- to 7- (p = 0.001) and 8- to 10-year-old children (p = 0.001) in the average F0 in connected speech. With regard to the variability in connected speech F0, significant differences were observed between the 3- to 5-year-old group and the 6- to 7- (p = 0.007) and 8- to 10-year-old group (p = 0.009). Overall, there was a greater decrease in F0 after 5 years of age. Regarding jitter values, there was a significant difference between children aged 3–5 years and children aged 7–8 years (p = 0.02). The shimmer values were significantly different for children aged 3–5 years and children aged 8–10 years (p = 0.01). Vocal Characteristics in Children
Table 1. Visual analog scale values (mean ± SD) and statistical re-
sults of ANOVAs for each perceptual variable, according to age Variables
3 – 5 years
6 – 7 years
8 – 10 years
p value
GG RG BG SG IG
43.07 ± 4.39 41.19 ± 14.83 37.11 ± 22.27 48.03 ± 11.67 48.26 ± 9.59
40.31 ± 5.74 39.13 ± 14.96 30.27 ± 19.37 37.22 ± 15.96 44.13 ± 11.52
46.73 ± 9.03 44.47 ± 14.21 40.34 ± 17.44 34.41 ± 15.36 45.34 ± 11.77
0.003* 0.338 0.140 0.001* 0.403
Significant values: * p < 0.05 (ANOVA).
Discussion
The high prevalence of dysphonia in children requires special attention in the assessment and diagnosis of children’s voices. The use of objective measures may provide a better understanding of the severity of vocal deviations and their manifestations at different age ranges from 3 to 9 years [6, 7, 9, 10]. In our statistical analysis of perceptual parameters according to age range, there were more abnormal voices in the age range 8–10 years as compared to the other two age groups (p = 0.003, table 1). However, there was a progressive reduction of the phonation strain level as a result of aging (p = 0.001, table 1). The 8- to 10-year-old children exhibited an increase in the level of roughness and breathiness compared to the other age ranges, even though the comparisons were not statistically significant. The grade of phonation tension proved to be high in the early childhood years, compatible with the phonation effort required to project the voice. The voice quality changes are normally associated with an increase in vocal fold mass or abnormalities in glottic closure, the main determining factors for vocal disorders. In general, glottic chinks caused by glottic proportion deviations, such as in pediatric larynges, may overload the phonation mechanism in an attempt to increase vocal intensity, especially in group activities and in the presence of competitive background noise [11–16]. Moreover, the use of a higher F0 in the early childhood years also predisposes children to increased phonation strain. The repetitive use of a hyperfunctional pattern may even cause the development of vocal fold damage such as vocal nodules or mucosa thickness, laryngeal lesions commonly found in epidemiological studies [5]. With regard to the acoustic measures, there were significant differences in the mean F0 for sustained vowel Folia Phoniatr Logop 2013;65:143–147 DOI: 10.1159/000355913
145
Table 2. Mean values (±SD) for each acoustic measure and statistical results of ANOVAs, according to age
Variables
3 – 5 years
6 – 7 years
8 – 10 years
p value
F0 sustained vowel F0 SD F0 mean counting F0 Var counting Jitter Shimmer GNE
271.59 ± 37.87 5.61 ± 3.12 277.68 ± 30.6 205.88 ± 81.32 1.14 ± 1.91 8.02 ± 4.68 0.77 ± 0.21
242.53 ± 22.84 5.05 ± 2.72 242.43 ± 26.99 130.53 ± 69.48 0.27 ± 0.13 6.09 ± 2.13 0.83 ± 0.12
244.08 ± 28.09 4.29 ± 4.13 248.6 ± 34.39 143.99 ± 89.53 0.5 ± 0.67 5.74 ± 2.7 0.75 ± 0.19
0.000* 0.313 0.000* 0.003* 0.018* 0.001* 0.252
Significant values: * p < 0.05 (ANOVA). F0 Var = Fundamental frequency variability; GNE = glottal-to-noise excitation ratio.
(p < 0.001), mean frequency of connected speech (p < 0.001), and the variability of F0 in connected speech between early childhood (3–5 years) and the other age ranges; lower voice F0 and less variability was found after 5 years of age (table 2). F0 is closely related to length, mass, tension, vocal fold rigidity, and subglottic pressure [17]. Some studies [6, 17, 18] have shown a decrease in F0 standard throughout development, explained by an increase in vocal fold length and mass, expansion and development of the vocal tract, and the approximation of male and female voice-related adult models. A recent study [6] demonstrated that there is no typical F0 for each age range and that there is no correct age for changes to occur, but that there are periods of mild changes for both boys and girls, starting between 11 and 12 years of age. In the present study, there were significant changes after 5 years, corresponding to the end of early childhood, although we have not studied older children to determine whether there are also significant changes in early adolescence. The values for jitter (p = 0.018) and shimmer (p = 0.001) were significantly different for the 3- to 5-year-old children compared to the 6- to 7- and 8- to 10-year-old children (table 2). The normal range for jitter and shimmer for a sustained vowel /ε/ in the software used varied from 0 to 0.6% and 0 to 6.5%, respectively. Thus, the major difference between the early childhood years and other age ranges is that these values are the standard variation for the human voice in the age ranges 6–7 and 8–10 years. It is difficult to make comparisons between these data and the other standard ranges reported in the literature [6], which recommend jitter and shimmer values of 1.8 and 4–5%, respectively, for children; this is because, in most 146
Folia Phoniatr Logop 2013;65:143–147 DOI: 10.1159/000355913
of the studies, the vowel /a/ was used. The use of a different vowel may explain the difference in values. Jitter and shimmer may help the description and quantification of perceptual characteristics and vocal fold vibration patterns. They are widely used in clinical practice and research studies. They are descriptive and can be used for diagnostic purposes. Moreover, they provide data relevant to the assessment and screening process for vocal disorders and are measures used for documentation and monitoring of vocal treatment approaches [19–22]. Small irregularities of the sound signal are considered the standard variation associated with the physiological mechanism of vocal production. However, the increase in the values of acoustic signal perturbation may indicate an inappropriate operation of the glottic mechanism, which may be used to detect the presence of vocal abnormality. It is important to bear in mind that children’s vocal folds have not completed the process of differentiation of the intermediate and deep layers of the lamina propria [14]. Moreover, the children are still developing their neuromuscular control, which means that abnormal measures of perturbation/irregularity in children are to be expected, especially during early childhood. The normal operating range for jitter and shimmer in children should take into account the anatomic and physiologic characteristics of this important developmental stage [23]. The normal-range values used to classify a regular voice versus a deviant voice in adults might not be effective in discriminating voices in children. For this reason, it is important to carry out research studies that analyze acoustic measures at different age ranges during the childhood years associated with auditory perceptual analysis of the voices. Even though we found higher values for jitter, shimmer, F0, and phonation strain in early childhood (3–5 Lopes et al.
years), the most disturbed voices were detected in the age range of 8–10 years. This finding reflects the fact that there is greater vulnerability for children below 5 years to develop laryngeal disorders because of the recurrence of upper airway infections during this period and the immaturity of vocal tract structures. The presence of a vocal abnormality in school-aged children seems to be the result of chronic and excessive voice abuse in group activities, which can easily be noticed at about age 10 [3]. During childhood, especially by its end, the child has likely participated in group activities that involve excessive and abusive use of voice such as school and family, posing a risk for the development of vocal disorders caused by an overload of the phonation system and an imbalance in phonation physiology. If this pattern of behavior is maintained, it may lead to additional vocal fold lesions [24]. For this reason, it is necessary to create strategies to promote voice preservation in children. These strategies should be sensitive to the child’s environment and include educational measures that can improve and optimize oral
communication, especially during early childhood when phonation is made with additional effort. These strategies may prevent the development and fixation of inappropriate laryngeal adjustments in the preadolescent years, which could be maintained into adulthood.
Conclusion
It was confirmed that children in the age range of 8–10 years had the highest severity of vocal deviation. There was a significant reduction of phonation strain and measure of F0 and jitter after 5 years. Shimmer was significantly lower after the age of 8.
Disclosure Statement The authors have no conflicts of interest to declare.
References 1 Tavares ELM, Labio RB, Martins RHG: 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. 2 Carding P, Roulstone S, Northstone K, ALSPAC Study team: The prevalence of childhood dysphonia: a cross-sectional study. J Voice 2006;20:623–630. 3 Sederholm E, McAllister A, Dalkvist J, Sundberg J: Aetiologic factors associated with hoarseness in ten-year-old children. Folia Phoniatr Logop 1995;47:262–278. 4 Dejonckere PH, Bradley P, Clemente P, Cornut G, Crevier-Buchman L, Friedrich G, Van De Heyning P, Remacle M, Woisard V: A basic protocol for functional assessment of voice pathology, especially for investigating the efficacy of (phonosurgical) treatments and evaluating new assessment techniques. Eur Arch Otorhinolaryngol 2001;258:77–82. 5 Tavares ELM, Brasolotto A, Santana MF, Padovan CA, Martins RHG: Epidemiological study of dysphonia in 4–12 year-old children. Braz J Otorhinolaryngol 2011; 77: 736– 746. 6 Maturo S, Hill C, Bunting G, Ballif C, Maurer R, Hartnick C: Establishment of a normative pediatric acoustic database. Arch Otolaryngol Head Neck Surg 2012;138:956–961. 7 Lopes LW, Barbosa Lima IL, Alves Almeida LN, Cavalcante DP, de Almeida AA: Severity of voice disorders in children: correlations between perceptual and acoustic data. J Voice 2012;26:819.e7–819.e12.
Vocal Characteristics in Children
8 Yamasaki R, Leo SHS, Madazio G, Padovani M, Azevedo R, Behlau M: Correspondence between visual analog scale and the numerical rating scale perceptual-hearing voices. 16th Brazilian Congress of Speech, Campos de Jordao. San Paulo, Brazilian Society of Speech, 2008, pp 1080–1084. 9 Niedzielska G: Acoustic analysis in the diagnosis of voice disorders in children. Int J Pediatr Otorhinolaryngol 2001;57:189–193. 10 Martins RHG, Ribeiro CBH, Mello BMZF, Branco A, Tavares EL: Dysphonia in children. J Voice 2012;26:674.e17–674.e20. 11 Uloza V, Verikas A, Bacauskiene M, Gelzinis A, Pribuisiene R, Kaseta M, Saferis V: Categorizing normal and pathological voices: automated and perceptual categorization. J Voice 2011;25:700–708. 12 Trani M, Ghidini A, Bergamini G, Presuti L: Voice therapy in pediatric functional dysphonia: a prospective study. Int J Pediatr Otorhinolaryngol 2007;71:379–384. 13 Connor NP, Cohen SB, Theis SM, Thibeault SL, Heatley DG, Bless DM: Attitudes of children with dysphonia. J Voice 2008; 22: 197– 209. 14 McAllister AM, Granqvist S, Sjölander P, Sundberg J: Child voice and noise: a pilot study of noise in day cares and the effects on 10 children’s voice quality according to perceptual evaluation. J Voice 2009; 23: 587– 593. 15 Roy N, Holt KI, Redmond S, Muntz H: Behavioral characteristics of children with vocal fold nodules. J Voice 2007;21:157–168.
16 Van Houte E, Van Lierde K, Claeys S: Pathophysiology and treatment of muscle tension dysphonia: a review of the current knowledge. J Voice 2011;25:202–207. 17 Campisi P, Tewfik TL, Pelland-Blais E, Husein M, Sadeghi N: MultiDimensional Voice Program analysis in children with vocal cord nodules. J Otolaryngol 2000;29:302–308. 18 Nicollas R, Garrel R, Ouaknine M, Giovani A, Nazarian B, Triglia JM: Normal voice in children between 6 and 12 years of age: database and nonlinear analysis. J Voice 2008;22:671–675. 19 Ma EPM, Yu EML: Multiparametric evaluation of dysphonic severity. J Voice 2006; 20: 380–390. 20 Araújo SA, Grellet M, Pereira JC: Standardization of acoustic measures of the normal voice. Br J Otorhinolaryngol 2002;68:540–544. 21 Sader RCM, Hanayama EM: Theoretical considerations on the acoustic approach of the infant voice. Rev CEFAC 2004;6:312–318. 22 Brockmann M, Drinnan MJ, Storck C, Carding P: Reliable jitter and shimmer measurements in voice clinics: the relevance of vowel, gender, vocal intensity, and fundamental frequency effects in a typical clinical task. J Voice 2011;25:44–53. 23 McAllister A, Sederholm E, Ternström S, Sundberg J: Perturbation and hoarseness: a pilot study of six children’s voices. J Voice 1996;10:252–261. 24 Takeshita TK, Ricz LA, Isaac ML, Ricz H, Lima WA: Comportamento vocal de crianças em idade pré-escolar. Arq Int Otorrinolaringol/ Int Arch Otorhinolaryngol 2009;13:252–258.
Folia Phoniatr Logop 2013;65:143–147 DOI: 10.1159/000355913
147
Copyright: S. Karger AG, Basel 2014. Reproduced with the permission of S. Karger AG, Basel. Further reproduction or distribution (electronic or otherwise) is prohibited without permission from the copyright holder.
