Survey of Voice Acoustic Parameters in Iranian Female Teachers.

OBJECTIVES: Teachers are one of the professional voice users. Voice problems are common among them. Female teachers are known to have more voice problems than male ones. Furthermore, there are only few studies on the voice of Iranian female teachers. The present study investigated the acoustic parameters of voice in Iranian female teachers and compares them with nonteachers.
METHODS: In this cross-sectional study, 90 Iranian female elementary teachers, 30-50 years old, and 90 Iranian female nonteachers in the same age were assessed between May 2010 and October 2011. Data collection was carried out, using the Dr. Speech software (subprogram: vocal assessment Version 4.0 from Tiger Electronics) at the speech therapy clinic under a comfortable phonation. Normal voice in practitioners was judged by the perceptual evaluation by a voice therapist and indirect laryngoscopy examination by an otorhinolaryngologist. Voice characteristics were assessed with GRBAS scale. The speech sample was sustained /â/ using habitual and constant vocal for 10 seconds. Three tokens from each subject were obtained. Then, each subject was asked to read a standard passage in Farsi. Eventually, the difference measures of F0, jitter, shimmer, harmonic to noise ratio (HNR), and maximum of phonation time (MPT) between two groups were investigated by statistics software SPSS 19.0 (IBM corp.2010).
RESULTS: Results showed that the values of F0 were higher in teachers (210.03 Hz) than in nonteachers (194.11 Hz; P < 0.001). In addition, the values of perturbation measures were greater in teachers (jitter 0.32% and shimmer 4.63%) than those in the control group (jitter 0.22% and shimmer 3.15%; P < 0.001), but in HNR and MPT values, nonteachers showed higher levels (P < 0.001). The value of HNR in teachers was (18.84±1.56) but it was (21.3±1.73) in non-teachers and MPT value in teachers was (16.83±3.65) and in non-teachers was (22.5±5.2).
CONCLUSIONS: It can be concluded that vocal overuse, abuse, or misuse during teaching over a period of time result in achievement of inadequate phonatory pattern with excessive musculoskeletal tension, and the possible result is tissue changes in teacher's voice. In addition, acoustic analysis of voice parameters for teachers may significantly contribute to the objective voice examination of this group.

Introduction

The term “professional voice user” means those people who depend on a consistent, special, or appealing voice quality as a primary tool of trade, and those who are afflicted with dysphonia or aphonia would generally be discouraged in their jobs and seek alternative employment. Thus, teachers as professional voice users are thought to be at a higher risk of voice problems. Various studies have reported that voice problems are common among teachers.2, 3 According to many questionnaire studies, 50–80% of teachers experienced voice problems,4, 5 and teaching constitutes one of the 10 occupations that often require medical help for voice difficulties. Roy et al reported the prevalence of lifetime voice disorders to be significantly higher among teachers (57.7%) than in nonteachers (28.8%). In Spain, the prevalence of clinically diagnosed vocal disorders was 57%, and 79% of teachers reported symptoms of vocal alteration, and 20% reported a confirmed diagnosis of laryngeal injury. The incidence was 3.9 new cases per year per 1000 teachers. In Finland, laryngeal evaluation detected alterations in 51% of the teachers.

Voice use in teaching profession is highly demanding, and the hazardous factors are teaching often at high voice output level because of the presence of background noise, poor classroom acoustics, and poor working posture, long speaking distance, poor quality of air ventilation, stress, and nonavailability of or poor-quality aids. Noise is one important factor. Markides and Pekkarinen reported that background noise and reverberation time are higher than acceptable limits in many classrooms. In a noisy classroom, teacher does not notice that she has raised her voice but she has done it unconsciously. Loud speaking and increase of voice straining may lead to vocal fatigue and to vocal fold tissue damage.10, 11 Contributing cofactors are individual endurance, gender, living habits, vocal experiences, and so forth. Majority numbers of elementary schools teachers are women. Many studies reported that the voice disorders are twice in female teachers compared with their male peers.11, 12 Voice complaints apart from being a problem for teachers can also reduce their professional effectiveness.13, 14 In a study investigating the effect of the teacher's voice quality on the pupil's ability to process spoken language, it was observed that children performed better when recalling the words presented by a female teacher with a normal voice, as opposed to a female teacher with a dysphonic voice. Voice problems are therefore not only detrimental to the teacher concerned but also to their pupils and employers.

There are a few studies about voice changes induced by vocal loading. Fundamental frequency (F0), sound pressure level, jitter, shimmer, and long-time average spectra have been used for documenting vocal changes. Many studies reported common result that F0, jitter, and shimmer rise after loading.10, 15 In a field study conducted by Rajasudhakar and Savithri in five elementary school teachers, reported after 6 hours of teaching, fundamental frequency of phonation, jitter, and speaking fundamental frequency were increased compared with the preteaching condition. There is a lack of data on some acoustic measures such as noise-to-harmonics ratio in teachers, which may be one of the best acoustic predictors of perceptual qualities.16, 17 In one study, teachers who have more professional teaching showed lower level of HNR.

There are only few studies on the Iranian voice. In one study that performed on the Iranian voice, Mohseni et al compared the F0 and intensity between 200 normal adult Iranian men and women. Their result showed that the value of F0 and intensity in Iranian female were significantly higher than men. In another study, Dehghan and Scherer studied 15 male and 15 female teachers and found that female Iranian teachers seem to be more susceptible to voice stability change than the male ones. Furthermore, the incidence of vocal disorders is twice in female teachers compared with their male peers. Thus, we planned to investigate the acoustic parameters of voice only in Iranian female teachers in much more samples (90 Iranian female teachers and 90 Iranian female nonteachers) and in a different area.

Methods and materials

Participants

Ninety Iranian female teachers of elementary school in the age range of 30–50 years with 8–23 years of teaching experience volunteered to participate in this study. The control group consisted of 90 Iranian female nonteachers in the age range of 30–50 years. All of them were housewife and did not have heavy vocal loading. The number of classes taken by teachers per day was six and duration of each class was about 1 hour. The average number of students in each class was 20–28. The exclusion criteria were previously treated dysphonia, smoking or alcohol habits, neurologic or endocrine disease, psychiatric disturbances, acid reflux, multiple medical complaints, vocal fold lesion as polyps, paralysis, papillomas, or severe dysphonia requiring urgent intervention, allergies, asthma, and other recurrent upper respiratory tract diseases, singing training and speech, language, and hearing or voice problems. None of them were at menstrual period.

Normal voice in practitioner was judged by the perceptual evaluation by a voice therapist and indirect laryngoscopy examination by an otorhinolaryngologist. All the subjects had normal larynx. Both two groups were assessed for voice characteristics with the GRBAS scale, which stands for grade, roughness, breathiness, asthenicity, and strain. Voice sampling was on spontaneous speech for 1 minute. Those with ratings higher than 0, even if it was on only one measure, were excluded from the study. All subjects spoke Farsi and were from the same dialectal region. The voice samples were performed in a sound-treated room with the subjects in a seated position.

Instrumentation

Data collection was performed, using the Dr. Speech software (Tiger Electronics Inc) (subprogram: vocal assessment Version 4.0 from Tiger Electronics) at the speech therapy clinic. The voice was recorded by a microphone (type: ECM-717 electret condenser microphone; Sony Corporation, Japan) placed on a stand at 10 cm from the mouth. We used the Real Analysis program of Dr. Speech software for determining the mean of fundamental frequency (F0), jitter (%), shimmer (%), and the HNR (dB).

Voice sample

Data collection took place between May 2010 and October 2011. Before recording the samples' voices, the experimental and control groups were informed about the aims of the survey. The voice sample consisted of (1) sustained vowel for assessing jitter (%), shimmer (%), HNR (dB), and MPT (seconds) and (2) reading a standard passage in Farsi for assessing F0. The procedure of data collection was the same for both groups. The participants were asked to produce vowel /â/ in a comfortable and habitual way, for 10 seconds and three tokens from each subject were obtained. A mid-5-second segment of each vowel prolongation was subjected to the acoustic analyses. Another part of data collection was reading from a standard passage in Farsi. The passage contained 138 commonly used words. All 23 consonants and six vowels in Farsi were involved in the passage. Before testing, all subjects were asked to practice the reading passage to induce the fluency of the voice. All subjects were asked to read the Farsi passage in a comfortable and habitual way.

Statistical analysis

The average of the three tokens produced by each subject was determined. Then, we used Student t test to examine the mean values of variables between two groups. Data were analyzed with the statistics software SPSS 19 for windows. P < 0.001 was considered statistically significant.

Results

For each measure, the summary data for Iranian female teachers and nonteachers are summarized in Table 1 and Figure 1, Figure 2, Figure 3, Figure 4, Figure 5.

Table 1

Results of F0, Jitter, Shimmer, HNR, and MPT for Female Teachers and Nonteachers

Participant	F0 (Hz)	Parameters
Jitter (%)	Shimmer (%)	HNR (dB)	MPT (s)
Teachers	210.03 (±25.65)	0.32 (±0.307)	4.63 (±1.86)	18.84 (±1.56)	16.83 (±3.65)
Nonteachers	194.11 (±20.81)	0.22 (±0.1)	3.15 (±0.85)	21.3 (±1.73)	22.5 (±5.2)

Notes: Student t test; P < 0.001.

Figure 1

F0 differences between female teachers and nonteachers.

Figure 2

Jitter differences between female teachers and nonteachers.

Figure 3

Shimmer differences between female teachers and nonteachers.

Figure 4

HNR differences between female teachers and nonteachers.

Figure 5

MPT differences between female teachers and nonteachers.

As indicated in the table and figures, the t test showed significant differences between the teachers and control group in all parameters (P < 0.001). As the data are shown in Table 1 and Figure 1, the measure of F0 in Iranian female teachers had significantly higher frequency (210.03 Hz) than in the control group (194.11 Hz).

In addition, as presented in Table 1 and Figure 2, Figure 3, for the perturbation acoustic parameters (jitter % and shimmer %), the female teacher group had significantly higher values than the control group.

Similarly, the HNR measures for the Iranian female teachers were significantly lower than for their corresponding control group (Figure 4).

Discussion

The results of present study indicated that there were significantly different values on all acoustic parameters of voice between two groups. Teachers often speak loudly for long duration in presence of high-level background noise and not in a conductive environment. Because of this, most teachers suffer from vocal fatigue at the end of the workday. The accompanying feature along with vocal fatigue would be an increase in the level of muscular tension of vocal folds, thus leading to a higher F0 value.24, 25, 26, 27, 28 In the present study, the measure of F0 in teachers was higher than nonteachers (210.03 Hz in teachers and194.11 Hz in nonteachers; P < 0.001). The findings of present study were in agreement with the findings of Rantala et al. They reported that the F0 increased in teachers. The literature offered some explanation for F0 rise. According to Stemple et al, increase of F0 is a due to weakness of the thyroarytenoid muscle. When the muscular layer of the thyroarytenoid muscle slackens, the cover and transition layers of the vocal folds stiffen. This leads to increasing the rate of vibrations in the vocal folds and a rise of the F0. Vilkman et al29, 30 have suggested another explanation. The rise of F0 was caused by the speaker's compensatory reaction to alteration in their voice. When compensating for the physiological changes, which could be alterations in the mucosa, the speaker increases the frequency of vocal fold vibration and the glottal adductory forces. This increased constriction influences the F0 indirectly.

Relative to the perturbation acoustic parameters, jitter refers to the vibration of fundamental frequency present in all speakers to some degree and detected when the subject is attempting to produce steady, sustained vowels. Jitter is the result of instability of the vocal fold vibration. As such, jitter reflects the biomechanical characteristics of the vocal folds, as well as variations of neuromuscular control. We found that jitter value in the female teacher group was significantly higher than that in the control group (0.32 in teachers and 0.22 in nonteachers; P < 0.001). These findings were in agreement with the findings of previous studies.25, 26, 27, 29 Jitter increasing correlates to fatigue of the throat in the healthy. This reduction might result from lowered muscle tonus and impaired neuromotor control of the larynx because of fatigue. Sorenson proposed that the voice of female teachers is related to high values in jitter measures.

Shimmer is a cycle-to-cycle, short-term perturbation in the amplitude of the voice. Our data indicated that the shimmer in female teachers had significantly higher value compared with nonteachers (4.63 in teachers and 3.15 in nonteachers; P < 0.001). Rantala et al reported similar findings that perturbation values increased in teachers. It has been hypothesized in a research that increasing of jitter and shimmer correlated with fatigue of the throat in the healthy. This reduction might result from lowered muscle tonus and impaired neuromotor control of the larynx because of fatigue.

The HNR measures for female teachers were significantly lower than those for nonteachers group (18.84 in teachers and 21.3 in nonteachers; P < 0.001). HNR quantifies the relative amount of additive noise in the voice signal. Inadequate closure of the vocal folds, or aperiodic vocal fold vibration, allows excessive airflow through the glottis, giving rise to turbulence. The resulting friction noise results in a higher noise level in the spectrum, reflected in a lower HNR. De Krom proposed that HNR is one of the parameters that can be used to relate physiological aspects of voice production to a perceptual impression of the voice because the degree of spectral noise is related to the quality of the vocal output. It seems that vocal loading in female teachers is important to decrease of HNR than in nonteachers. Verstraete et al reported decrease of HNR in female teachers after a day work. The MPT indicates the efficiency of the coordination between the levels of breathing and phonation because, for maximum sustained phonation, the individual uses the maximum of her/her vital capacity (VC) to maintain a phoneme for as long as possible, reflecting the neuromuscular control and aerodynamic vocal production.37, 38, 39, 40 Our results indicated that the value of MPT in teachers was significantly lower in teachers than nonteachers (16.83 in teachers and 22.5 in nonteachers; P < 0.001). Through the MPT relationship, it is possible to check the occurrence of glottic hypercontraction or air leak to phonation because the MPT is a prediction about the time that the subject should sustain the emission on the basis of the values of VC, and MPT is time that it was effectively performed by the subject.41, 42 Decreased MPT values are suggestive of transglottic air escape during phonation and increased values suggest increased glottal closure and increased muscle tension or phonation.39, 41, 42, 43, 44, 45 MPT results suggest pneumophonoarticulatory incoordination. This may be a risk factor for the development of voice problems with increasing vocal demand for teachers. It seems that female teachers, because of high vocal demand, have air leakage during phonation. The end result is low value of MPT.

Conclusions

In this study of Iranian female teachers of elementary school, the female teachers produced significantly higher F0 values and greater perturbation values for the sustained vowel /â/, but in HNR and MPT values, nonteachers showed higher levels. It can be concluded that vocal overuse, abuse, or misuse during teaching over a period of time result in achievement of inadequate phonatory pattern with excessive musculoskeletal tension, and the possible result is tissue changes in teacher's voice.