G Model

PEDOT-7560; No. of Pages 4

International Journal of Pediatric Otorhinolaryngology xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Pediatric Otorhinolaryngology

journal homepage: www.elsevier.com/locate/ijporl

Acoustic analysis of voice in children with cleft palate and

velopharyngeal insufficiency
Rocio Villafuerte-Gonzalez a, Victor M. Valadez-Jimenez a,
Xochiquetzal Hernandez-Lopez a, Pablo Antonio Ysunza b,*
a
Department of Speech & Language Pathology, Instituto Nacional de Rehabilitacion, Mexico City, Mexico
b
Ian Jackson Craniofacial and Cleft Palate Clinic, Neuroscience Program, Beaumont Health, Royal Oak, MI, United States

A R T I C L E I N F O A B S T R A C T

Article history: Background: Acoustic analysis of voice can provide instrumental data concerning vocal abnormalities.
Received 2 March 2015 These findings can be used for monitoring clinical course in cases of voice disorders. Cleft palate severely
Received in revised form 17 April 2015 affects the structure of the vocal tract. Hence, voice quality can also be also affected.
Accepted 18 April 2015
Objective: To study whether the main acoustic parameters of voice, including fundamental frequency,
Available online xxx
shimmer and jitter are significantly different in patients with a repaired cleft palate, as compared with
normal children without speech, language and voice disorders.
Keywords:
Materials and methods: Fourteen patients with repaired unilateral cleft lip and palate and persistent or
Cleft palate
Speech
residual velopharyngeal insufficiency (VPI) were studied. A control group was assembled with healthy
Voice volunteer subjects matched by age and gender. Hypernasality and nasal emission were perceptually
Acoustics assessed in patients with VPI. Size of the gap as assessed by videonasopharyngoscopy was classified in
patients with VPI. Acoustic analysis of voice including Fundamental frequency (F0), shimmer and jitter
were compared between patients with VPI and control subjects.
Results: F0 was significantly higher in male patients as compared with male controls. Shimmer was
significantly higher in patients with VPI regardless of gender. Moreover, patients with moderate VPI
showed a significantly higher shimmer perturbation, regardless of gender.
Conclusion: Although future research regarding voice disorders in patients with VPI is needed, at the
present time it seems reasonable to include strategies for voice therapy in the speech and language
pathology intervention plan for patients with VPI.
ß 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Acoustic analysis includes measurement of several parameters.

Perceptual assessment of voice is voice is fundamental for the
evaluation of patients with voice disorders. Digital acoustic
analysis of voice provides instrumental data complementing
perceptual assessment. Digital acoustic analysis increases the
knowledge concerning specific voice characteristics and it is
helpful for understanding the mechanisms of voice production.
Moreover, acoustic data are useful for following the patient’s
clinical course, documenting clinical improvement and providing
visual biofeedback which can be used during the voice intervention
[1].
* Corresponding author. Tel.: +1 248 551 2100; fax: +1 248 551 4692.
E-mail address: antonio.ysunza@beaumont.edu (P.A. Ysunza).
The most important for clinical purposes are Fundamental
Frequency (F0), Jitter and Shimmer. FO is determined by the
frequency at which the vocal cords vibrate during voice production
and it is measured in Hz. Normal values of F0 for children have
been previously reported [2]. F0 in normal female children ranges
from 268 to 295 Hz. A mean F0 of 247 Hz in children under 12 years
of age has also been reported, without significant differences when
boys and girls were compared [3].
During consistent vocal cord vibration, F0 demonstrates a slight
variation and also a variation of cycle-to-cycle amplitude, these
phenomena are known as frequency perturbation or jitter and
amplitude perturbation or shimmer. These variations or changes
are a consequence of the mass and tension differences which
characterize the vocal cords. Jitter and shimmer values can be
expressed in a percentage. There are normative data for these
percentages. Jitter percentage should be below 1% whereas

http://dx.doi.org/10.1016/j.ijporl.2015.04.030
0165-5876/ß 2015 Elsevier Ireland Ltd. All rights reserved.

Please cite this article in press as: R. Villafuerte-Gonzalez, et al., Acoustic analysis of voice in children with cleft palate and
velopharyngeal insufficiency, Int. J. Pediatr. Otorhinolaryngol. (2015), http://dx.doi.org/10.1016/j.ijporl.2015.04.030
G Model
PEDOT-7560; No. of Pages 4
2 R. Villafuerte-Gonzalez et al. / International Journal of Pediatric Otorhinolaryngology xxx (2015) xxx–xxx
shimmer percentage should be below 10%. It should be considered
that shimmer and jitter are dependent of F0. Thus, in cases of
abnormal voices with aperiodicity which precludes F0 measure-
ment, shimmer and jitter are not useful [1].
Systematic changes of F0 can occur as a consequence of stress or
intonation, which are inherent to speech. These changes affect the
acoustic assessment of F0. Thus, the clinical evaluation of F0,
shimmer and jitter is usually performed during the production of
a sustained vowel sound. The most acoustically consistent vowel
sound is the vowel/a/ [4,5].
Vocal qualities are the result of individual anatomical and
physiological characteristics which are strongly influenced by age,
gender and physical structure [6]. Therefore, for an adequate vocal
production it is necessary that all the components of the vocal
tract, from the glottis to the lips and including the resonance
chambers of the nasal cavities and sinuses be structurally and
functionally intact.
The most common craniofacial anomaly affecting resonance
and speech is a palatal cleft. According to Kernahan and Stark [7],
the most common cleft is the total cleft of the primary and
secondary palate. Patients with a cleft of the secondary palate
show abnormal nasal resonance and nasal emission, which are a
consequence of the defective or incomplete seal of the velophar-
yngeal sphincter during speech. This anomaly is known as
velopharyngeal insufficiency (VPI) and significantly affects speech
intelligibility [8]. Besides hypernasality and nasal emission, some
patients with cleft palate also develop compensatory behaviors
known as compensatory articulation errors. These articulation
disorders affect intelligibility even more severely than VPI [9].
Some papers have focused on the acoustic characteristics of
voice in patients with cleft palate. Frequency instability, high
intensity, abnormal formant distribution and abnormal spectro-
grams have been reported [10–12]. In 1989, Zajac and Linvile [11]
studied voice perturbations in patients with VPI. They performed
acoustic analyses and found a significant relationship between
abnormally high jitter values and hypernasality.
The purpose of this paper is to study acoustic parameters of
voice including F0, shimmer and jitter in children with persistent
or residual VPI after surgical repair of a cleft palate. The acoustic
parameters of these children were compared with a group of
normal controls matched by age and gender, without any voice or
speech disorder.
2. Materials and methods
A comparative, cross-sectional study was carried out at the
Department of Speech and Language Pathology of the National
Institute of Rehabilitation in Mexico City from January 2012 to
December 2013. The protocol was approved by the Internal Review
Board of the Institute.
2.1. Patients
Patients who met the following criteria were recruited for the
study: (a) surgically corrected non-syndromic unilateral total cleft
lip and palate (UCLP) with persistent or residual VPI as
demonstrated by perceptual analysis of voice and videonasophar-
yngoscopy. The patients should present with UCLP as an isolated
malformation, not associated with a sequence or syndrome; (b)
normal hearing as demonstrated by a behavioral pure – tone
audiometry; (c) 6–10 years of age at the time to be recruited for
the study. This age group was selected in order to assure
appropriate compliance during the behavioral audiometry and
the acoustic assessment, as well as to prevent voice changes as
a consequence of puberty. Patients with palatal fistula or with a
history of laryngeal pathology or hoarseness were excluded.
Please cite this article in press as: R. Villafuerte-Gonzalez, et al., Acoustic analysis of voice in children with cleft palate and
velopharyngeal insufficiency, Int. J. Pediatr. Otorhinolaryngol. (2015), http://dx.doi.org/10.1016/j.ijporl.2015.04.030
Patients present with compensatory articulation errors including
glottal stops were excluded from the study group. Patients with
neurological disorders or intellectual – cognitive deficits were
excluded. Patients with severe nasal emission causing nasal
rustle which was being picked up during the recordings procedures
were also excluded.
During the recruiting period, a total of 14 patients met the
criteria for being included in the study. Eleven patients were males
and 3 patients were females. The age of the patients ranged from
7 to 9 years of age with a median of 8 years. The study protocol
was carefully explained to the parents and/or legal guardians and
all of them signed an informed consent form.
2.2. Controls
A control group was assembled, including 14 subjects matched
by age and gender. These subjects were recruited at a nearby
elementary school. All these subjects voluntarily participated in
the study. The study protocol was carefully explained to the
parents and/or legal guardians and all of them signed an informed
consent form. All of these patients were examined in order to verify
that there were no palatal anomalies (including submucous cleft
palate), no voice disorders, no compensatory articulation errors
and no perceptual signs of VPI (normal nasal resonance during
speech).
2.3. Procedure
In patients with repaired cleft palate and persistent VPI,
hypernasality and nasal emission were perceptually classified as
previously reported [13,14]. Two experienced physicians certified
by the Board of Speech and Language Pathology of Mexico (first
and second authors of this paper) were in charge of conducting the
examination. The examinations were performed independently.
A Cohen’s Kappa concordance index of 85 was found between both
examiners. Regarding nasal emission a Cohen’s Kappa index of
92 was found. Whenever there was disagreement, each case was
discussed until a consensus had been reached. Hypernasality was
detected in all patients. Only 4 patients presented with nasal
emission across plosive and fricative phonemes. It should be
pointed out that none of these patients showed nasal rustle that
could be picked up by the microphone during the recording
procedures. Eight patients presented with mild hypernasality. Six
patients presented with moderate hypernasality.
All patients with repaired cleft palate and persistent VPI
underwent a flexible nasoendoscopy following a protocol which
has been previously reported [15]. Velopharyngeal movements
during speech including velar and lateral pharyngeal motion, as
well as presence of Passavant’s ridge were classified as ratios,
considering the position at rest as 0 and full contact or closure as
1.0. The size of the gap during speech was estimated as a ratio
considering total velopharyngeal space at rest as 1.0. Eight patients
demonstrated only bubbling during speech without a discernible
gap. These cases were conventionally considered as a 10% gap and
a mild VPI. Six patients demonstrated gaps with a size of 20%, that
is, a moderate VPI. It should be pointed out that size of the gap as
assessed by videonasopharyngoscopy is merely an estimate and
not hard data. In other words, inter-observer reliability is not
statistically significant. This is why the gaps are conventionally
assessed in 10-point intervals (10–20–30%, and so on). It should be
pointed out that the endoscope was passed further into the vocal
tract in order to assess the vocal folds. All patients showed normal
mobility of both vocal folds and no masses were identified in any
of the cases. No stroboscopy was performed.
Acoustic assessments were performed according to the protocol
that was previously reported [16,17]. The patients and subjects
G Model
PEDOT-7560; No. of Pages 4

R. Villafuerte-Gonzalez et al. / International Journal of Pediatric Otorhinolaryngology xxx (2015) xxx–xxx 3

Table 1
Distribution of acoustic parameters.

Variable Patients Control subjects P; 95% CI Patients Control subjects P; 95% CI

Males Females

F0 270.4 Hz SD  7.01 257.9 Hz SD  6.57 0.000 259.1 Hz SD  35.3 240.2 Hz SD  2.1 0.407
Shimmer 5.2 SD  0.86 0.78 SD  0.21 0.000 7.7 SE  3.5 0.6 SD  0.04 0.024
Jitter 1.24 SD  0.26 0.23 SD  0.09 0.000 1.2 SD  1.8 1.01 SD  0.07 0.826

Table displays mean values and standard deviations (SD) of F0, shimmer and jitter of the groups of patients and control subjects. Patients and subjects were divided by gender
for comparison purposes. Probability values (P) are also included. A 95% confidence interval was selected for considering P values as statistically significant.

were seated in a straight position on a clinical chair. They were 4. Discussion

asked to produce a sustained vowel sound (/a/) for at least 3 s at an
intensity that was considered as comfortable by the patients and
subjects. A Perfect Choice microphone PC-110279 LO2311027 was
used at a distance of 10 cm from the lips and an angle of 30 degrees
toward the patient. F0, jitter and shimmer were determined.
2.4. Statistical analysis
Conventional central tendency measures, including mean and
standard deviation were used for analyzing F0, shimmer and jitter.
Acoustic parameters were compared between groups (active –
patients and control – subjects) through a non-paired two tailed
student t test. The SPSS Statistics 20 software was used for
performing the statistical tests.
3. Results
F0 was significantly higher in males from the group of patients
as compared with male controls. In contrast, female patients and
subjects did not demonstrate a significant difference (see Table 1).
Shimmer was significantly higher in the group of patients as
compared with controls, regardless of gender (see Table 1). These
data suggest that patients with persistent VPI seem to show higher
variability of peak-to-peak amplitude between consecutive
periods.
Jitter was significantly higher in males from the group of
patients as compared with male controls. Female patients and
subjects did not demonstrate a significant difference (see Table 1).
These data suggest that male patients with persistent VPI seem to
show higher variability in F0 signal periodicity.
When patients present with VPI with nasal emission were
compared with patients with VPI without nasal emission, non-
significant differences in F0, shimmer or jitter were found.
When the patients were stratified according to the severity of
VPI (mild to moderate), a significant difference was demonstrated
between shimmer values of patients with mild VPI (10% gaps)
as compared with patients with moderate VPI (20% gaps). Patients
with moderate VPI showed a higher shimmer perturbation. This
difference was independent of gender (see Table 2).
Table 2
Comparison by severity of VPI.
Variable VPI
Mild Moderate
F0 263.2 SD  11.9 274.4 SD  19.3
Shimmer 4.7 SD  0.38 7.1 SD  2.27
Jitter 0.20 SD  0.09 0.79 SD  1.2
Table displays mean values and standard deviations (SD) of F0, shimmer and jitter
in the group of patients with velopharyngeal insufficiency (VPI). VPI was classified
according to size of the gap and hypernasality (mild to moderate). Probability (P)
values are also included. A 95% confidence interval was selected for considering P
values as statistically significant.
From the results of this study it seems that patients present
with VPI show vocal abnormalities which were expressed by
differences in F0 and perturbations of voice frequency (jitter) and
amplitude (shimmer).
It should be considered that although the group of patients was
homogeneous because of the relatively strict inclusion and
exclusion criteria, the number of cases studied herein is reduced,
which prevents drawing definite conclusions.
The acoustic abnormalities found in patients present with VPI
could be explained by the presence of VPI during speech
development. VPI occurs most commonly in individuals with a
history of cleft palate. Although surgeons attempt to achieve as
much velar lengthening and motion at early ages, through different
surgical techniques, 20–40% of patients with cleft palate will
present with persistent or residual VPI after surgical palatal repair
[18,19]. Thus, if there is a significant VPI during speech
development it seems plausible that several patients will show
vocal compensatory behaviors for the lack of intraoral pressure.
It should be pointed out that the patients studied for this paper
did not present with compensatory articulation errors. Therefore,
the vocal compensatory behaviors mentioned herein seem to be
independent of compensatory articulation errors such as glottal
stops and pharyngeal fricatives.
Patients with VPI showed acoustic abnormalities which may be
associated with vocal fold adduction and sudden release. Patients
with vocal fold paresis show similar although more severe acoustic
findings, but the acoustic findings in patients with VPI may be
indicating a low concentration of acoustic energy at the F0 and
relative to the overall energy of the signal.
Patients present with VPI may be attempting to regulate active
vocal tract resistances as compensation. Moreover, increased
glottal resistances during vowel production would decrease air
flow and hence facilitate the regulation of subglottal pressure
which is essential for continuous phonation [20].
A possible explanation for some of the acoustic abnormalities
described herein, especially the apparent relationship between
severity of VPI and shimmer perturbation, may be that the
individual is attempting to regulate subglottal pressure and vocal
tract aerodynamics in a condition in which there is insufficient
coupling between the resonant nasal cavities and the rest of the
vocal tract [21], potentially increasing vocal noise.
Another interesting finding of this study was gender dependent
F0 difference. Males showed higher F0 as compared with male
P controls. Moreover, jitter was also higher in male patients as
compared with male controls. These data suggest that male
patients with persistent VPI seem to show higher variability in F0
0.204 signal periodicity.
0.011
Van Lierde et al. [22] described vocal quality characteristics in
0.207
children with cleft palate through a multiparameter approach.
They reported that the results of their study supported the
hypothesis that vocal quality disorders characterized by a more
negative DSI-value occur in male cleft palate children. These
authors also described that decreased vocal quality expressed as

Please cite this article in press as: R. Villafuerte-Gonzalez, et al., Acoustic analysis of voice in children with cleft palate and
velopharyngeal insufficiency, Int. J. Pediatr. Otorhinolaryngol. (2015), http://dx.doi.org/10.1016/j.ijporl.2015.04.030
G Model
PEDOT-7560; No. of Pages 4
4 R. Villafuerte-Gonzalez et al. / International Journal of Pediatric Otorhinolaryngology xxx (2015) xxx–xxx
acoustic abnormalities in male children with cleft palate may be
caused by a direct laryngeal pathology or may be an indirect
response to VPI, which supports the hypothesis that vocal
hyperfunction and abnormal vocal quality in children with VPI
represent the children’s attempts to compensate for the VPI.
Although age does not affect mean F0, shimmer and jitter are
age related. Thus, the rather broad age range of the patients studied
in this paper could have affected the acoustic assessments.
Moreover, in this study, nasal resonance and nasal emission were
assessed perceptually using a previously published classification
which has not been entirely validated concerning its reliability. It
will be necessary to perform future studies using instrumental
assessments including nasometry and aerodynamics measure-
ments for assessing nasal resonance and emission.
5. Conclusion
The results of this study suggest that acoustic parameters of
voice are significantly affected in patients with VPI. Jitter seems
to be more affected in male patients and there seems to be a
relationship between severity of VPI and perturbation of shimmer.
The vocal abnormalities found in this study seem to be associated
with compensatory behaviors related with VPI and they are
independent of compensatory articulation errors such as glottal
stops. Patients present with VPI may be at risk of developing vocal
pathology. Future research regarding vocal disorders in patients
with VPI is needed, including collection of vocal tract resonance
measures (spectrograms and cepstral peak prominence measure-
ments). However, at the present time, it seems reasonable to
include strategies for voice therapy in the speech & language
pathology intervention plan for patients with VPI.
References
[1] K.C. Toran, B.K. Lal, Objective analysis of voice in normal young adults, Kathmandu
Univ. Med. J. 7 (2009) 374–377.
[2] L. Wanderley Lopes, S.L. Do Nascimento Cunha Costa, W.C. De Almeida Costa, S.E.
Nobrega Correia, V.J. Dias Vieira, Acoustic assessment of the voices of children
using nonlinear analysis: proposal for assessment and vocal monitoring, J. Voice
28 (2014) 565–573.
[3] M. Guzman, D. Munoz, M. Vivero, N. Marin, M. Ramirez, M.T. Rivera, et al., Acoustic
markers to differentiate gender in prepubescent children’s speaking and singing
voice, Int. J. Pediatr. Otorhinolaryngol. 78 (2014) 1592–1598.
[4] E.T. Doherty, Evaluation of selected acoustic parameters for use in speaker
identification, J. Acoust. Soc. Am. 58 (S107) (1975) 1121.
Please cite this article in press as: R. Villafuerte-Gonzalez, et al., Acoustic analysis of voice in children with cleft palate and
velopharyngeal insufficiency, Int. J. Pediatr. Otorhinolaryngol. (2015), http://dx.doi.org/10.1016/j.ijporl.2015.04.030
[5] M. Grellet, J.C. Pereira, M. Oliveira Rosa, Standardization of acoustic measures of
the normal voice, Braz. J. Otorhinolaryngol. 68 (2002) 540–544.
[6] S. Maturo, C. Hill, G. Burting, C. Balif, R. Maurer, C. Hartnick, Establishment of
a normative pediatric acoustic data base, Otolaryngol. Head Neck Surg. 138 (2012)
956–961.
[7] D.A. Kernahan, R.B. Stark, A new classification for cleft lip and cleft palate, Plast.
Reconstr. Surg. 22 (1958) 435–441.
[8] A. Ysunza, M. Pamplona, F. Molina, M. Drucker, J. Felemovicius, E. Ramirez, et al.,
Surgery for speech, Int. J. Pediatr. Otorhinolaryngol. 68 (2004) 1499–1505.
[9] M.C. Pamplona, A. Ysunza, S. Morales, Strategies for treating compensatory
articulation in patients with cleft palate, Int. J. Biomed. Sci. 10 (2014) 43–51.
[10] R. Kataoka, D.W. Warren, D.J. Zajac, R. Mayo, R.W. Lutz, The relationship between
spectral characteristics and perceived hypernasality in children, J. Acoust. Soc.
Am. 16 (2001) 205–210.
[11] D.J. Zajac, R.N. Linville, Voice perturbations of children with perceived nasality
and hoarseness, Cleft Palate J. 26 (1989) 226–231.
[12] J. Chenghui, T.L. Whitehill, B. McPherson, M. Lawrence, Spectral features and
perceptual judgment of place of affricate in Putonghua-speaking pre-adolescents
with normal and cleft palate, Int. J. Pediatr. Otorhinolaryngol. 79 (2015) 179–185.
[13] G. Henningsson, D.P. Kuehn, D. Sell, T. Sweeney, J. Trost-Cardamone, T. Whitehill,
et al., Universal parameters for reporting speech outcomes in individuals with
cleft palate, Cleft Palate Craniofac. J. 45 (2008) 1–17.
[14] A. Ysunza, A. Pamplona, S. Morales, Velopharyngeal valving during speech, in
patients with velocardiofacial syndrome and patients with non-syndromic clefts
after surgical and speech pathology management, Int. J. Pediatr. Otorhinolaryn-
gol. 75 (2011) 1255–1259.
[15] A. Ysunza, R. Merson, Videonasopharyngoscopy of the velopharyngeal sphincter
during speech and swallowing, in: D.M. Grange (Ed.), Endoscopy and Endoscopic
Procedures: Management, Technologies and Methods of Improvement, Nova,
New York, 2014, , ISBN: 978-1-63117-0805.
[16] V. Valadez, A. Ysunza, E. Ocharan-Hernandez, N. Garrido-Bustamante, A. Sanchez-
Valerio, M.C. Pamplona, Voice parameters and videonasolaryngoscopy in children
with vocal nodules. A longitudinal study before and after voice therapy, Int. J.
Pediatr. Otorhinolaryngol. 76 (2012) 1361–1365.
[17] A. Garcı́a Pérez, X. Hernández-López, V.M. Valadez, A. Minor-Martı́nez, A. Ysunza,
Synchronous electrical stimulation of laryngeal muscles: an alternative for en-
hancing recovery of unilateral recurrent laryngeal nerve paralysis, J. Voice 28
(2014) 524e1–524e7.
[18] A. Ysunza, K. Shaheen, D.J. Aughton, M.A. Micale, R. Merson, K. Rutkowski,
Velopharyngeal insufficiency, submucous cleft palate and a phonological disorder
as the associated clinical features leading to the diagnosis of Jacobsen syndrome.
Case report and review of the literature, Int. J. Pediatr. Otorhinolaryngol. 77
(2013) 1601–1605.
[19] J. Nyberg, P. Peterson, A. Lohmander, Speech outcomes at age 5 and 10 years in
unilateral cleft lip and palate after one-stage palatal repair with minimal incision
technique – a longitudinal perspective, Int. J. Pediatr. Otorhinolaryngol. 78 (2014)
1662–1670.
[20] Z. Yang, J. Fan, J. Tian, L. Liu, C. Gan, W. Chen, et al., Cepstral analysis in children
with velopharyngeal insufficiency after cleft palate surgery, J. Voice 28 (2014)
789–792.
[21] S. Ha, H. Sim, M. Zhi, D.P. Kuehn, An acoustic study of the temporal characteristics
of nasalization in children with and without cleft palate, Cleft Palate Craniofac. J.
41 (2004) 535–543.
[22] K.M. Van Lierde, S. Claeys, M. De Bodt, V. Van Cauwenberge, Vocal quality
characteristics in children with cleft palate: a multiparameter approach, J. Voice
(2004) 354–362.