Keywords
velopharyngeal sphincter - voice - cleft lip - cleft palate - physiology
Introduction
The velopharyngeal sphincter functions as a valve that closes like a sphincter. It
extends along the lateral and posterior pharyngeal walls and its anterior boundary
is close to the soft palate.[1]
[2]
[3]
[4]
[5] Velopharyngeal sphincter closure is accomplished by elevating and retracting the
soft palate at the same time as the nasopharyngeal walls are constricted.[3]
[6]
[7] The primary function of the velopharyngeal sphincter is to ensure the physiological
maintenance of this region.[1]
[8]
Speech may be affected in different ways when the velopharyngeal closure pattern is
disturbed. The most common symptoms of velopharyngeal inadequacy are hypernasality,
nasal air escape, and articulation problems. However, all these symptoms depend on
how much the soft palate has been affected.[9]
Malfunctioning of the velopharyngeal closure mechanism, which constitutes the velopharyngeal
dysfunction, may be associated with several underlying diseases such as neurological
disorders, sequelae from surgical interventions, structural changes, etc. Malformations
in the palate region lead to physiological abnormalities in the velopharyngeal sphincter,
such as characteristics of cleft lip and palate.[10]
The speech and voice inadequacy caused by velopharyngeal insufficiency is a major
stigma of patients with cleft palate.[11] Several abnormal characteristics that impair communication can be detected in the
speech of such individuals.[12] Primary and secondary speech disorders relate to velopharyngeal dysfunction. Hypernasality
and nasal air escape are primary disorders, whereas compensatory articulation and
its associated facial movement are secondary disorders.[13]
Velopharyngeal dysfunction is a term used to describe abnormalities in the velopharyngeal
mechanism in general; nevertheless, different terms are common to describe different
disorders. Velopharyngeal insufficiency has a structural cause; the soft palate movement
is normal, but it is too short to accomplish velopharyngeal closure. Another type
of velopharyngeal dysfunction is velopharyngeal incompetence, which is caused by a
neuromotor disorder; the soft palate has a normal structure, although it does not
move sufficiently to achieve velopharyngeal closure.[5]
[14]
[15]
[16] Failure of velopharyngeal closure may persist even after surgical repair of the
palate.[17]
Understanding the physiology of the craniofacial structures affected by this disease
is essential to choose the most appropriate therapeutic modality. The region of the
velopharyngeal sphincter is the most difficult for understanding the pathophysiology
of the functional changes present in this malformation.[8]
There are few studies in literature correlating velopharyngeal closure with regular,
weak, or strong voice intensity. Therefore, the objective of the present study was
to evaluate this association, comparing the three voice intensities produced by patients
without velopharyngeal dysfunction and patients with cleft lip and palate.
Methods
We conducted a cross-sectional, comparative, and contemporary study. The Scientific
Committee and Research Ethics Committee evaluated and approved the project (protocol
no. 13–0360).
Considering a 95% confidence level, an estimated standard deviation at 7% of the percentage
of velopharyngeal closure, and a margin of error of 5%, we included 27 individuals
in each group. This sample size calculation was based on the dissertation by Dornelles.
Nevertheless, we evaluated 32 individuals in the control group; however, 16 tests
were not properly recorded due to technical equipment failure. In the case group,
we could not achieve the size initially calculated for the sample because of the long
period of routine patient care. Therefore, after applying the inclusion and exclusion
criteria, we selected 16 patients for the control group and 16 patients for the case
group. The participants voluntarily agreed to participate in the study.
According to the inclusion criteria, we included in the control group female and male
participants without velopharyngeal dysfunction, whose age ranged from 18 to 50 years
old. The case group included female and male participants with velopharyngeal dysfunction
and cleft lip and palate, aged 7 to 51 years, who had undergone repair surgeries and
were receiving follow-up at the Outpatient Clinic of Otolaryngology and Cleft Lip
and Palate between March and July 2014.
We excluded individuals who (or whose guardians) did not agree to participate in the
study by not signing the informed consent form, as well as patients with cognitive
and/or behavioral disorders with associated syndromes, individuals with dysphonia,
use of nasoenteric tube that could prevent or hinder the performance of videonasoendoscopy,
and patients with anatomical and functional abnormalities that prevented the performance
of routine examinations and clinical follow-up.
We searched the speech-screening database for the patients' medical history. We also
conducted brief interviews with the participants and/or their guardians. For those
who agreed to participate in the study, we performed videonasoendoscopy examination
with flexible fiber to the area of the velopharyngeal sphincter to capture images
of the studied structures. The survey had a standardized dynamic, better access to
nostril without anesthetic, with the patient sitting in front of the medical examiner.
The examiner recorded a speech sample consisting of two sentences at three different
intensities: regular, weak, and strong. The sentences contained plosive sounds (Papai pediu pipoca, Dad asked for popcorn) and fricative sounds (Juju saiu cedo, Juju left early). Actual intensities were computed in accordance with what was requested
and measured by the DL decibel meter, Model 4020 (ICEL, Manaus). The microphone was
positioned 5 cm away from the subject́s mouth, in his lapel. Before starting the exam,
the examiner measured environmental noise. The patient was asked to perform the steps
in the assigned protocol, under constant monitoring by the researcher in charge. For
the videonasoendoscopy procedure, we used equipment belonging to the Otorhinolaryngology
Department. We edited the images to include the two sentences at the three different
intensities, recorded them on DVD, and showed them to three professionals experienced
in the assessment and treatment of patients with cleft lip and palate. The audio was
deleted from the video records and the assessors were supposed to rate each sentence
according to the appropriate closure (similar closure pattern, large gap, small gap,
and moderate gap). Next, we analyzed images using a computer program that is under
improvement, used only in research aiming to analyze their actual contributions in
the same. We used a computational model for the analysis of the wall motion of the
velopharyngeal sphincter.
The quantitative variables were expressed as mean and standard deviation, whereas
the qualitative variables were expressed as absolute and relative frequencies. We
used student's t-test and Fisher's exact test to compare the means between the groups. To calculate
the clinical estimate of the velopharyngeal sphincter closure, we reached agreement
between the analyses of the three assessors considering the whole sample. As there
was no agreement using the Kappa test, we used the analysis of the assessors with
greater scientific and technical knowledge in the field as a reference for data analysis.
We also evaluated the agreement produced by the computer program using the Kappa test.
We performed the interpretation of Kappa coefficients as proposed by Landis and Koch.
Kappa values may range from 1 (perfect agreement) to < 0 (no agreement). The authors
proposed a six-level scale so that the values < 0 indicated no agreement, the values
from 0.00 to 0.20 showed very poor agreement, from 0.21 to 0.40 showed poor agreement,
from 0.41 to 0.60 showed moderate agreement, from 0.61 to 0.80 showed good agreement,
and from 0.81 to 1.00 showed perfect or almost perfect agreement. We evaluated the
association between the intensity of the sentences of velopharyngeal sphincter closure
between the assessors and the computer program using Spearman's correlation coefficient.
We compared the proportions between groups using Pearson's chi-square test and Fisher's
exact test. In both analyses, we used a significance level of 5%, and the SPSS version
18.0 for the analyses.
Results
Thirty-two individuals participated in the study: 16 in the control group and 16 in
the case group. In the control group, 12 (75%) participants were female and four (25%)
were male; their mean age was 27.6 years (±9.5), ranging between 18 and 51 years old.
In the case group, eight (50%) participants were female and eight (50%) were male;
their mean age was 15.6 years (±11.5), ranging between 7 and 50 years old.
In the case group, the participants had cleft lip and palate. Based on the classification
suggested by Brazil Cleft, we found higher prevalence of unilateral cleft lip and
palate (7; 43.75%) and lower frequency of unilateral cleft lip (0; 0%), followed by
full cleft palate (4; 25%), bilateral cleft lip and palate (3; 18.75%), and bilateral
cleft lip and incomplete cleft palate (1; 6.25%). All participants had undergone previous
surgery. Eleven (68.75%) underwent nose and lip repair surgery, 15 (93.75%) underwent
palate repair surgery, and one (6.25%) underwent pharyngeal flap surgery.
[Table 1] shows the characteristics of the sample. Controls were significantly older than
cases. There was no difference between the groups in terms of gender.
Table 1
Characteristics of the sample
|
Variables
|
Cases
|
Controls
|
p
|
|
Age (years) - Mean ± SD
|
15.6 ± 11.5
|
27.6 ± 9.5
|
0.003[*]
|
|
Gender – n (%)
|
−
|
−
|
0.273[**]
|
|
Male
|
8 (50.0)
|
4 (25.0)
|
−
|
|
Female
|
8 (50.0)
|
12 (75.0)
|
−
|
* Student's t-test;
** Fisher's exact test.
[Table 2] shows that patients had higher voice intensity (which was measured using a decibel
meter) in the regular and weak fricative sentences when compared with the control
group, showing a statistically significant relationship.
Table 2
Comparison of voice intensity between cases and controls according to the decibel
meter
|
Variables
|
Cases
Mean ± SD
|
Controls
Mean ± SD
|
p[*]
|
|
Plosive sentence
|
|
|
|
|
Regular
|
73.8 ± 3.2
|
70.9 ± 4.7
|
0.053
|
|
Weak
|
69.6 ± 3.5
|
70.6 ± 6.7
|
0.602
|
|
Strong
|
79.9 ± 6.3
|
82.4 ± 7.8
|
0.325
|
|
Fricative sentence
|
|
|
|
|
Regular
|
74.2 ± 3.9
|
70.3 ± 2.9
|
0.003
|
|
Weak
|
70.6 ± 4.1
|
67.1 ± 2.2
|
0.005
|
|
Strong
|
79.4 ± 6.1
|
77.9 ± 6.6
|
0.529
|
* Student's t-test.
[Table 3] shows the agreement between the assessors in terms of intensity and velopharyngeal
sphincter closure. There was significant agreement between assessors 2 and 3 considering
the plosive sentence at regular and strong intensity. However, these agreements were
weak according to Landis and Koch. There was also significant agreement between assessors
1 and 3 regarding all intensities of the plosive sentence and in terms of the regular
intensity of the fricative sentence. Of the four variables showing agreement, one
(25%) was very weak, two (50%) were weak, and one (25%) was moderate (plosive sentence
at strong intensity).
Table 3
Agreement between assessors
|
Comparisons
|
SCP/LG/SG/MG %
|
Agreement (%)
|
Kappa
|
p
|
|
Assessor 1 versus Assessor 2
|
|
Plosive sentence
|
|
|
|
|
|
Regular intensity
|
65.6/0/21.9/12.5 versus 56.3/12.5/21.9/9.4
|
15/32 = 46.8%
|
0.13
|
0.377
|
|
Weak intensity
|
68.8/12.5/6.3/12.5 versus 46.9/15.6/18.8/18.8
|
13/32 = 40.6%
|
0.05
|
0.647
|
|
Strong intensity
|
53.1/28.1/18.8/0 versus 40.6/34.4/15.6/9.4
|
16/32 = 50.0%
|
0.29
|
0.033
|
|
Fricative sentence
|
|
|
|
|
|
Regular intensity
|
68.8/6.3/12.5/12.5 versus 40.6/18.8/18.8/21.9
|
17/32 = 53.1%
|
0.29
|
0.003
|
|
Weak intensity
|
71.9/12.5/15.6/0 versus 53.1/9.4/31.3/6.3
|
19/32 = 59.3%
|
0.18
|
0.161
|
|
Strong intensity
|
75/12.5/3.1/9.4 versus 40.6/40.6/6.3/12.5
|
15/32 = 46.8%
|
0.16
|
0.109
|
|
Assessor 2 versus Assessor 3
|
|
Plosive sentence
|
|
|
|
|
|
Regular intensity
|
56.3/12.5/21.9/9.4 versus 56.3/12.5/25/6.3
|
17/32 = 53.1%
|
0.23
|
0.050
|
|
Weak intensity
|
46.9/15.6/18.8/18.8 versus 53.1/12.5/12.5/21.9
|
15/32 = 46.8%
|
0.20
|
0.059
|
|
Strong intensity
|
40.6/34.4/15.6/9.4 versus 46.9/34.4/6.3/12.5
|
17/32 = 53.1%
|
0.30
|
0.008
|
|
Fricative sentence
|
|
|
|
|
|
Regular intensity
|
40.6/18.8/18.8/21.9 versus 25/6.3/37.5/31.3
|
7/32 = 21.8%
|
−0.04
|
0.649
|
|
Weak intensity
|
53.1/9.4/31.3/6.3 versus 28.1/15.6/28.1/28.1
|
6/32 = 18.7%
|
−0.11
|
0.244
|
|
Strong intensity
|
40.6/40.6/6.3/12.5 versus 25/53.1/9.4/12.5
|
12/32 = 37.5%
|
0.06
|
0.630
|
|
Assessor 1 versus Assessor 3
|
|
Plosive sentence
|
|
|
|
|
|
Regular intensity
|
65.6/0/21.9/12.5 versus 56.3/12.5/25/6.3
|
18/32 = 56.2%
|
0.32
|
0.027
|
|
Weak intensity
|
68.8/12.5/6.3/12.5 versus 53.1/12.5/12.5/21.9
|
18/32 = 56.2%
|
0.25
|
0.021
|
|
Strong intensity
|
53.1/28.1/18.8/0 versus 46.9/34.4/6.3/12.5
|
19/32 = 59.3%
|
0.47
|
0.001
|
|
Fricative sentence
|
|
|
|
|
|
Regular intensity
|
68.8/6.3/12.5/12.5 versus 25/6.3/37.5/31.3
|
13/32 = 40.6%
|
0.20
|
0.022
|
|
Weak intensity
|
71.9/12.5/15.6/0 versus 28.1/15.6/28.1/28.1
|
12/32 = 37.5%
|
0.22
|
0.067
|
|
Strong intensity
|
75/12.5/3.1/9.4 versus 25/53.1/9.4/12.5
|
11/32 = 34.3%
|
0.10
|
0.193
|
*Abbreviations: LG, Large gap; MG, Moderate gap; SCP, Similar closure pattern; SG,
Small gap.
[Table 4] shows the association between the intensity of the sentences and the closure according
to assessor 1 and the computer program. There was no significant association between
the intensity of the sentences and closure both according to the assessors and the
computer program; that is, regardless of the intensity, closure remained similar.
Table 4
Association between sentence intensity and closure according to the assessor and the
computer program using Spearman's correlation coefficient
|
Intensity of the sentences
|
Case group
|
Control group
|
|
Assessor
|
Computer program
|
Assessor
|
Computer program
|
|
Plosive sentence
|
|
|
|
|
|
Regular
|
0.120
|
−0.196
|
−0.320
|
*
|
|
Weak
|
0.029
|
0.214
|
0.181
|
*
|
|
Strong
|
0.296
|
−0.169
|
0.416
|
*
|
|
Fricative sentence
|
|
|
|
|
|
Regular
|
−0.267
|
0.206
|
0.197
|
*
|
|
Weak
|
−0.049
|
−0.235
|
−0.072
|
*
|
|
Strong
|
−0.085
|
0.185
|
−0.366
|
*
|
* We could not perform statistical test because all controls showed similar closure
pattern according to the computer program.
[Table 5] shows data on the agreement between the assessors and the computer program in terms
of velopharyngeal sphincter closure. There was no significant agreement between the
assessors and the computer program.
Table 5
Agreement between assessors and computer program
|
Comparisons
|
SCP/LG/
SG/MG %
|
Agreement (%)
|
Kappa
|
p
|
|
Assessors versus Computer program
|
|
Plosive sentence
|
|
|
|
|
|
Regular intensity
|
65.6/0/21.9/12.5 versus 93.8/3.1/3.1/0
|
20/32 = 62.5%
|
−0.07
|
0.586
|
|
Weak intensity
|
68.8/12.5/6.3/12.5 versus 93.8/0/3.1/3.1
|
22/32 = 68.7%
|
0.18
|
0.131
|
|
Strong intensity
|
53.1/28.1/18.8/0 versus 93.8/6.3/0/0
|
17/32 = 53.1%
|
0.06
|
0.634
|
|
Fricative sentence
|
|
|
|
|
|
Regular intensity
|
68.8/6.3/12.5/12.5 versus 93.8/0/3.1/3.1
|
21/32 = 65.6%
|
0.02
|
0.838
|
|
Weak intensity
|
71.9/12.5/15.6/0 versus 93.8/3.1/0/3.1
|
22/32 = 68.7%
|
−0.06
|
0.713
|
|
Strong intensity
|
75/12.5/3.1/9.4 versus 93.8/6.3/0/0
|
23/32 = 71.8%
|
−0.06
|
0.678
|
*Abbreviations: LG, Large gap; MG, Moderate gap; SCP, Similar closure pattern; SG,
Small gap.
Discussion
The velopharyngeal closure is the result of the action of a set of muscles. That is,
it consists of a mechanism that works in a coordinated and synergic manner to alternately
bring together or separate the oropharyngeal and nasopharyngeal cavities.[18]
[19] This mechanism is essential for the production of vowels and consonants; therefore,
it has a profound impact on speech intelligibility. Several clinical populations,
such as children with a history of cleft lip and palate or individuals with dysarthria,
have velopharyngeal dysfunctions that cause speech production difficulties.[20]
The main objective of the present study was to compare voice intensity and velopharyngeal
sphincter closure. Our sample consisted of individuals with and without velopharyngeal
dysfunction and patients with cleft lip and palate. The most frequent dysfunction
was unilateral cleft lip and palate. Given that we used a random sample and there
was no case of unilateral cleft lip, we regrouped the types of cleft. The most prevalent
cases were cleft lip and palate. Such data are consistent with the literature.[21]
[22]
[23] The second most common type was cleft lip, and the least prevalent type was cleft
palate.
Our findings demonstrated that controlpatients were significantly older than patients.
The control grouṕs mean age was 27.6 years, whereas the case grouṕs mean age was
15.6 years. The velopharyngeal sphincter is critical for successful feeding and communication.
Motor activities, such as speech, sucking, swallowing, gag reflex, and breathing,
need a point of maximum closure during the movement against the walls.[2]
[24]
[25]
[26] We could not find many studies describing aspects of voice, swallowing, and hearing
including anatomical and functional signs and functional decline in the literature.
However, previous studies have not found deterioration of velopharyngeal functions
with aging.[27]
[28]
We assessed all patients using videonasoendoscopy. We used a decibel meter to measure
the voice intensity of the fricative and plosive sentences. We found that cases had
higher intensities in regular and weak fricative sentences when compared with controls.
Fricative and plosive phonemes are included in the protocol because they require greater
intraoral pressure; therefore, they evidence the articulation difficulties of patients
with cleft.[29]
[30]
The authors,[31] in their study on the variation of voice intensity, found mean voice intensity of
63.4 dB at regular emission and 72.5 dB at high emission. Such findings are not in
agreement with the present study. In our study, the mean intensities were higher.
The participants of the case group had mean voice intensities of 73.8 and 79.9 dB
for the plosive sentence, and 74.2 and 79.4 dB for the fricative sentence; whereas,
the participants of the control group showed mean intensities of 70.9 and 82.4 dB
for the plosive sentence and 70.3 and 77.9 dB for the fricative sentence.
With the purpose of achieving the objectives of the study, we calculated the agreement
between the assessors for the velopharyngeal sphincter closure patterns using the
Kappa coefficient of agreement by pairing the findings of an assessor with each of
the other assessors, thus resulting in three pairs (1 × 2, 2 × 3, 1 × 3). We only
found six variables showing agreement; and of these, one had very weak intensity,
one had moderate intensity, and four had weak intensity. Assessments of auditory perception
are known to be unreliable[32] and are not necessarily correlated with the functioning of the velopharyngeal sphincter.[33]
[34]
[35] Therefore, the decisions of the auditory perception regarding the velopharyngeal
function are often complemented by instrumental evaluation.[20]
As demonstrated in the present study, there was no statistically significant relationship
between the intensity of the sentences and velopharyngeal sphincter closure, regardless
of the fact that the intensity of the closure pattern remained similar. In another
study, the authors have reported that the orifice of the velopharyngeal sphincter
does not become smaller when there is increased intensity.[36] Therefore, this may indicate that the speaker is already using the physiological
mechanisms to achieve maximum closure. These authors also stated that these results
demonstrate the use of high voice intensity as a strategic behavior for individuals
with poor performance of the velopharyngeal sphincter and/or hypernasality.[36] The authors[26] performed a study of 21 assessments aimed at investigating whether the nasal emission
test showed compatibility with the videonasoendoscopy findings in the evaluation of
the velopharyngeal mechanism. They found that the participants did not show articulatory
compensation and had large gaps in most productions of phonemes, tending to maintain
the same gap size in both plosive and fricative sentences. This is an interesting
find because it shows that it is possible to produce sounds without compensation,
even when there is velopharyngeal incompetence or insufficiency.
Although videonasoendoscopy allows us to view the velopharyngeal sphincter during
speech, the test has some limitations. It is an invasive method and has the disadvantage
of being subjective because it does not provide quantitative data, with arbitrary
inference of velopharyngeal gap size.[37] In the present study, there was not agreement between the assessors and the computer
program. The velopharyngeal sphincter closure assessed by the computer program was
similar in both groups, with no difference between the proportions found. However,
it is worth noting that the only two patients who had no such closure pattern were
in the case group. Therefore, it is important to standardize the assessment, so that
more objective data can be collected regarding the evaluation of the motion of the
velopharyngeal sphincter.
Many researchers have concluded that the velopharyngeal mechanism is highly complex
and, thus, it can only be understood if several images are combined due to the difficulties
in getting a general view of the area.[38]
[39] The need for accurate information about the closure patterns is considered vital
for planning surgical intervention[40]
[41] and enabling the assessment of the advances of surgical methods.[42]
With respect to the sample size, we performed calculations. However, we could not
include it due to the routine care front upon which the project depended and the deadline
for completion of the work. Nevertheless, we believe that it has not generated a false
negative, based on clinical experience coupled with years of observation. We believe
that our study may help to understand the velopharyngeal function by providing more
reliable clinical evaluation tools.
Conclusions
When the three voice intensities (regular, strong, and weak) related to velopharyngeal
sphincter closure, regardless of whether the patient had velopharyngeal dysfunction,
the computer program showed similar closure pattern. Thus, it proved to be a useful
tool in clinical practice to assess the functioning of the velopharyngeal sphincter.
There was no statistically significant correlation between voice intensity of speech
and degree of velopharyngeal sphincter closure in both groups. Based on our results,
there was no agreement between the three assessors regarding the velopharyngeal sphincter
closure pattern. Because perceptual analysis are often not reliable, there is need
for standardization of a protocol or a tool to assist in this assessment.
We could not establish a correlation between the groups with and without velopharyngeal
dysfunction in terms of closure mechanism and voice intensity. It was not possible
correlate the findings in both groups with clinical practice. Therefore, we suggest
using a larger sample size to check our findings. In addition, further studies should
be conducted on this topic to contribute to increase scientific knowledge.
