Portuguese Version

Year:  2003  Vol. 69   Ed. 6 - (10º)

Artigo Original

Pages: 791 to 799

PDF PT  

Auditory perceptual, acoustic, computerized and laryngological analysis of young smokers' and nonsmokers' voice

Author(s): Daniele C. de Figueiredo1,
Paula R. F. de Souza2,
Maria Inês R. Gonçalves3,
Noemi G. de Biase4

Keywords: voice, smoking, tobacco, voice quality, fundamental frequency

Abstract:
Aim: The goal of this study was to make the laryngological, auditory perceptual and acoustic computer analyses of young adults' (smokers and non-smokers) voices, without vocal complaint, compare them and verify the incidence of vocal alterations. Study design: Clinical comparative. Material and Method: The voices of 80 individuals with age range from 20 to 40 years were analyzed. These individuals were divided in four groups: 20 male smokers, 20 male non-smokers, 20 female smokers and 20 female non-smokers. This analysis involved laryngoscopy, which was performed and interpreted by an otolaryngologist, and cassette tape recordings of the sustained vowels /a/, /µ/, /i/ e /u/, number counting from 1 to 20, speech of the days of the week, months of the year and singing of "Happy Birthday". The cassette tape recordings were edited for further spectrographic analysis and auditory perceptual evaluation made by four speech therapists experienced in the major of voice. Results: After the analysis, a slight reduction of the fundamental frequency of the voice of smokers of both sexes was noticed, as well as higher incidence of hoarseness and laryngeal alterations among the smokers.

INTRODUCTION

Phonation is a result of vocal fold vibration during expiration, which promotes the displacement of the mucosa cover and the formation of a wave from bottom to top, up to the point that the free margins touch again. In order to have good vocal quality and richness of harmonics, it is necessary to have elasticity and flexibility of the mucosa. The mucous-undulation movement of the mucosa depends on transglottic airflow. In normal conditions, air slides through the walls of the larynx with reduced friction, whereas in smokers, the mucosa cover is modified and increases air friction. Therefore, there is air disturbance rather than air sliding, causing a consequent disarrangement of the vibration cycle, modifying the vocal quality and causing sudden vocal attacks, sensation of burning, tightening and throat clearing during speech 1.

It is known that the consumption of tobacco causes many affections to the whole body, but what are the specific damages caused to the vocal mechanism of subjects?

In smokers, during phonation seen at stroboscopic light, we can detect loss of harmonic movement of the mucosa cover, since mucous is more viscous 1. Moreover, the mucosa is irritated and, therefore, it is more edematous, interfering in the vocal fold vibration capacity. Before the lesions are properly in place, there are mucous affections and impairment of the mucous-undulation movement of the mucosa cover of the vocal folds 1.

In a study conducted by Sorensen & Horii (1982), they recorded voices of 80 normal adult speakers, 40 smokers and 40 nonsmokers, being 20 men and 20 women in each group, aged between 25 years and 1 month and 49 years and 11 months. All subjects were submitted to pure tone audiometry in frequencies of 500, 1000 and 2000Hz and those whose hearing acuity was not within the normal range were excluded from the sample, since hearing affections can impact vocal fundamental frequency characteristics. The voices were recorded during reading, spontaneous speech and production of sustained vowels /a/, /i/ and /u/; next, they were digitalized and analyzed by the computer. There was reduction of fundamental frequency in the group of smokers compared to the nonsmokers, and the difference was statistically significant in males but not in females, concluding that smoking causes irritation in the laryngeal epithelium, which can damage the laryngeal tissues and modify the vocal signal that is generated there.

According to Lee et al. (1999), fundamental frequency is normally slightly reduced in passive smokers, but at irrelevant levels, despite the changes in vocal fold cover layers. As to vocal quality, Medina et al. (1990) studied 100 hospitalized Chilean women aged 42 to 51 years among which 45 had never smoked, 36 were smokers and 19 were former smokers. They conducted an auditory-perceptive analysis and there was high incidence of hoarseness among smokers, observing higher incidence of permanent hoarse voice in smokers.

The purpose of the present study was to conduct a comparative analysis of laryngology assessment, auditory-perceptive and computed acoustic analyses of smoker and nonsmoker young adults with no vocal complaints.

MATERIAL AND METHOD

We conducted ENT assessment and auditory-perceptive and computed acoustic analyses of the voices of 80 subjects, being 20 male smokers, 20 male nonsmokers, 20 female smokers and 20 female nonsmokers, aged 20 to 40 years.

Before starting the study, the project was previously analyzed and approved by the Research Ethics Committee at Federal University of Sao Paulo, Hospital Sao Paulo (CEP nº478/01).

To participate in the study, the subjects had to be within the defined age range and be either tobacco users or should have never smoked. Next, they answered a questionnaire whose model is shown in the Annex to learn, map and control vocal habits, and we selected those that did not present vocal complaints and had no other predisposing factors to the development of dysphonia, such as previous consumption of tobacco (former smokers) and professional vocal use (teachers, singers, actors, etc.).

The first procedure was the ENT examination, including laryngostroboscopy, conducted and analyzed by an experienced otorhinolaryngologist. We used the classical method: subjects were sitting down with open mouth and protracted tongue, involved in gauze and maintained in position by the physician's fingers. Topical anesthesia was made by spraying xylocaine and after it, we introduced the rigid laryngoscope up to the point of complete visualization of the larynx. During assessment, the subjects were instructed to breathe through the mouth, without stress, and when asked, they had to produce vowel /µ/ at sustained intensity and tone, as close as possible to habitual production.

At that moment of the study, many subjects were excluded from the sample because they could not undergo the test or presented organic-functional and organic pathologies. In such cases, we referred them to complete assessment and vocal therapy.

Laryngostroboscopy of subjects that remained in the study were analyzed and considered normal or abnormal, according to the observation of presence of edema, hyperemia, epithelial thickness and mucosa wave affections, in one or both vocal folds.

Next, the selected subjects were recorded in a acoustic booth, maintaining 15cm distance between the mouth and the microphone during the production of sustained vowels /a/, /µ/, /i/ and /u/, number counting from 1 to 20, emission of days of the week and months of the year, and singing "Happy Birthday ", in habitual conditions of frequency and intensity. We used unidirectional microphone Audiotechnica MB 4000C and voices were recorded in cassette tapes with professional tape recorder Marantz PMD 221. Based on this recording, we performed two analyses: computed analysis and auditory perceptive analysis.
In the computed acoustic analysis, the recording of vowel /µ/ for each subject was analyzed by the software SoundScope 2.0, by G.W. Instruments, in a Pentium PC. The analysis provided detailed and objective vocal data of each selected subject: fundamental frequency (f0) jitter (J); shimmer (S) and harmonic-noise ratio (HNR).

The last procedure was the auditory-perceptive analysis conducted by four speech and voice therapists experienced in voice. To prepare to the analysis, the original tape was edited being that the voice therapists did not know the identification of the subjects. Voices were classified concerning vocal quality, pitch and loudness, according to what was proposed by Behlau & Pontes (1995).

The data obtained were statistically compared between smokers and nonsmokers of each gender and among male and female groups.

Table 1. Comparison of values of jitter, shimmer, fundamental frequency and harmonic-noise ratio among the group of female nonsmokers (MNF) and female smokers (MF).



Table 2. Comparison of jitter, shimmer, fundamental frequency and harmonic-noise ratio between the groups of female nonsmokers (HNF) and male smokers (HF).



Table 3. Results of the auditory-perceptive analysis (type of voice, pitch and loudness) and laryngoscopy in the group of female nonsmokers.



Table 4. Results of the auditory-perceptive analysis (type of voice, pitch and loudness) and laryngoscopy in the group of male nonsmokers.



Table 5. Results of the auditory-perceptive analysis (type of voice, pitch and loudness) and laryngoscopy in the group of female smokers.



Table 6. Results of the auditory-perceptive analysis (type of voice, pitch and loudness) and laryngoscopy in the group of male smokers.

RESULTS AND DISCUSSION

After answering the questionnaire, as previously referred, subjects were submitted to the ENT assessment by the physician. The results are shown in Tables 3, 4, 5 and 6, always in the fourth column, and she considered only vocal fold abnormalities, such as presence of edema, hyperemia, epithelial thickness and mucosa wave affections.

In the group of male nonsmokers, 50% of the subjects (n=10) presented vocal fold abnormalities. It is believed that despite technology development and medical experience, some subjects could have had minimal structural abnormalities, which were not diagnosed during the tests since we emphasized the search for affections such as edema, hyperemia, epithelial thickness and mucosa wave affections. According to Gonçalves (1994), 30% of laryngeal minimal alterations are diagnosed only during surgical procedure.

According to Garnett (2001), smoking is closely related to laryngeal affections, which can cause edemas (such as Reinke's edema), which are inflammatory processes, up to leukoplasia, hyperplasia and cancer, requiring early intervention, which can be surgical and/or speech therapy, but it requires cessation of smoking. Subjects that presented evident organic lesion, resultant or not from smoking, were excluded from the study, since our objective was to identify initial affections, which did not characterize a disease yet. This fact can explain why there were no statistically significant differences between smoking and nonsmoking groups concerning abnormal vocal folds findings.

The first speech and voice analysis was the acoustic computed analysis. It provided data concerning fundamental frequency, jitter, shimmer and harmonic-noise ratio (Tables 1 and 2).

Fundamental frequency is related to frequency of vocal fold vibration. In men, it ranges from 80 to 150Hz; in women, from 150 to 250Hz, and in children it is above 250Hz 4. It is determined by size, tension and mass of the vocal fold cover combined with subglottic pressure. The vibration increases when tension increases and mass reduces 2. Taking into account that in smokers this mass tends to be larger, vibration would be reduced. We expected, therefore, to find reduction of fundamental frequency in the group of smokers compared to the group of nonsmokers, which indeed happened. However, it was not a statistically significant difference even though in the group of women, the value of significance was almost achieved (p=0.07). In the study conducted by Sorensen & Horii (1982), the reduction of fundamental frequency was more evident when they compared smokers and nonsmokers in advanced age than in our study. This fact suggests that the effects of cigarette smoking are more marked the longer the exposure. We believe, therefore, that if the age range of the volunteers in the present study had been older and the number of subjects in the study had been higher, the reduction in fundamental frequency would have been statistically significant.

Jitter indicates the variability of fundamental frequency in short term measured by glottic cycles and it expresses the quantitative difference between the period and its previous or next one, ignoring voluntary changes in frequency 2. In normal subjects, jitter value is low; however, the presence of vocal fold lesions causes more aperiodicity, which is expressed by higher values. Jitter is expressed in percentages and the normal limit value is 0.5%. In the group of female nonsmokers, only 5% of the subjects (n=1) presented jitter above this limit; in the group of female smokers, all jitter values were within the normal ranges; in male nonsmokers, 20% (n=4) presented jitter affection and in male smokers, 10% (n=2). We can see that many cases of jitter affections were present in the group that had greater amount of vocal fold abnormalities.

Shimmer indicates variability of sound wave amplitude in the short term and represents the irregular affections of glottic cycle amplitude, from one cycle to the other 3. Similarly to jitter, it presents high values in cases of laryngeal abnormality. Shimmer is also related to noise in vocal production; thus, it tends to be higher in cases of breathiness, for example. Its limit for normal range is 3%. In the group of female nonsmokers, only 5% of the subjects (n=1) presented values greater than those; in female smokers, 10% (n=2) had the same manifestation; in the male nonsmokers, 15% (n=3), and in male smokers, 10% (n=2). Once again, the highest number of abnormalities was found in the group that had more laryngeal affections.

The term harmonic-to-noise ratio (HNR) is related to the harmonic component as opposed to the noise component of the acoustic wave 3. Its measure is obtained in dB and it is lower in men and higher in women, which was also detected in the present study. There was statistically significant difference between the groups of male and female nonsmokers, but not among smokers.

The last analysis we conducted was auditory-perceptive analysis, which provided data about vocal quality (physical perception of laryngeal tone modified by resonance cavities), pitch (psychophysical sensation of fundamental frequency) and loudness (psychophysical sensation of vocal intensity) of subjects and whose data are exposed in Tables 3, 4, 5, and 6. This classical assessment of vocal quality is essential, even though subjective, since what takes patients to vocal therapist is the auditory perception that he/she and the others have of the voice 3.

Vocal quality is defined as perception of physical complexity of laryngeal tone modified by resonance cavities 1 and it provides information about physical characteristics and even educational background of subjects 3. One of the characteristics of vocal quality is type of voice, defined as the basic pattern of emission of the subject and it is related to selection of motor adjustments made, both at the level of the vocal folds and resonance 3.

According to Aronson (1990), it is very difficult to define what normal voice is, since variety of voices is unlimited and the vocal appropriateness parameters are very wide. Regardless, we can consider a normal voice the one that has pleasant quality and pitch, loudness and flexibility within the normal ranges. In this study, in the group of female nonsmokers, 45% of the subjects (n=9) had normal voice; in the group of female smokers, the number was down to 15% (n=3); in male nonsmokers, it was 50% (n=10), and in male smokers it was only 15% (n=3). All the other subjects had vocal types considered abnormal in some aspect. The most frequent affections were hoarseness, breathiness and the association between both.

In the group of female nonsmokers there was pure hoarseness in only 5% of the cases (n=1), breathiness in 20 (4 cases) and both associated in 30% of the cases (N= 6). In the group of female smokers, the incidence of breathiness increased to 35% of the cases (n=7), as well as hoarseness associated with breathiness; the cases of pure breathiness were less common (15% of the cases - n=3). In male nonsmokers, there was 15% hoarseness (n=3), as well as breathiness, which curiously were detected only in Asian descendents; the association of both characteristics was present in 20% of the cases (n=4). In the group of male smokers, there was high incidence of hoarseness (50% of the cases - n=10), contrasting with the absence of pure breathiness; 25% of the subjects (n=5) presented hoarse-breathy voice; 5% (n=1) had hoarse-cracking voice, and other 5% (n=1) had strained voice. There was, therefore, a high incidence of hoarseness in both groups of smokers.

It is important to be very careful concerning episodes of hoarseness. The study conducted by Lieberman (1963) showed that hoarseness was unanimously perceived by hearing people only in severe and advanced laryngeal cases, such as non-initial cancer. Therefore, it is fundamental to conduct laryngeal assessment of any subjects that present manifestation for a long time.

If we conceptualize dysphonia, as described by Behlau (2001), as a disorder of oral communication in which voice can not fulfill its basic role of transmitting the verbal and emotional message of a subject, representing differences in vocal emission that prevent natural voice, we can say that in this study, even though subjects had no vocal complaint, many of them presented mild dysphonia.

According to Garnett (2001), speech and voice therapy are essential to eliminate compensation speech techniques (which can be the case among smokers); however, if the subject does not cease smoking, little is achieved. Such compensation mechanisms were also described by Koufman and Blalock (1991) in smokers that had triggered glottic affections and functional dysphonia.

Thus, we believe that for subjects who present abnormalities, speech therapy is specific for each type of affection found and it can promote favorable results in vocal quality improvement, provided that the first measure taken is smoking cessation. Fortunately, the study by Spiegel, Sataloff and Emerich (1997), showed that tobacco intake has started to reduce among young adults.

CONCLUSION

Based on the analyses conducted, we concluded that:

1. Vocal fold abnormalities were more common in male subjects;
2. Vocal quality was significantly worse among smokers, being the presence of hoarseness its most frequent characteristic;
3. There was reduction of fundamental frequency in the groups of smokers compared to nonsmokers, but it was not statistically significant.

ACKNOWLEDGMENTS

We would like to thank all the people how had somewhat contributed to the completion of the present study.

To our tutor, Prof Maria Inês Rebelo Gonçalves, we would like to thank her dedication, theoretical support, availability and friendship and for having calmed us down even during the most critical periods, which encouraged our scientific interest.

To Prof. Noemi Grigoletto de Biase, we would like to thank her dedication and availability to conduct the tests and analysis of all laryngology examinations.

To Speech Therapists and Audiologist Vanessa Pedrosa Vieira, we would like to thank her support in the computer analysis and her availability as a volunteer for the auditory perceptive analysis.
To Speech Therapist and Audiologist Carla Bogossian, we would like to thank for her availability in conducting the auditory perceptive analysis.

We would also like to thank Liliana da Silva and Speech Therapist and Audiologist Camila Mercatelli for having helped us finding volunteers and for filling out the endless forms; Josefina Grandino (D. Josie), for her dedication and attention in the preparation of the references; Prof. Paulo Pontes, for having participated, even if indirectly, in the guidance for the conduction of the present study, and finally, we would like to thank all the volunteers that enabled the performance of this study.

REFERENCES

1. Behlau M, Pontes P. Avaliação e tratamento das disfonias. São Paulo: Lovise; 1995.
2. Greene MCL, Mathieson L. The voice and its disorders. 5th ed. San Diego: Singular Publishing; 1989.
3. Behlau M. Voz: O livro do especialista - Volume 1. São Paulo: Revinter; 2001.
4. Aronson AE. Clinical voice disorders. 3rd ed. New York: Thieme; 1990.
5. Fortes H. Dicionário médico inglês - português. 5ª ed. Rio de Janeiro: Fábio M. de Mello, 1968.
6. Garnett JD. Tobacco and laryngeal pathology. J Medical 2001; 97(1):13-16.
7. Gonçalves MIR. Alterações Estruturais Mínimas da Cobertura de Pregas Vocais - Fatores que interferem no diagnóstico. [tese] São Paulo: Universidade Federal de São Paulo; 1994.
8. Koufman J, Blalock D. Functional voice disorders. Otolaryngol Clin North Am 1991; 24(5):1059-73.
9. Lee L, Stemple JC, Geiger D, Goldwasser R. Effects of environmental tobacco smoke on objective measures of vocal production. Laryngoscope 1999;109(9):1531-4.
10. Lieberman P. Some acoustic measures of the fundamental periodicity of normal and pathologic larynges. J Acoust Soc Am 1963; 35 (3):344-53.
11. Medina E, Arteaga P, Pizarro L, Ahumada M. Algunos efectos del tabaquismo en la mujer. Rev Méd Chile 1999; 118:253-8.
12. Sorensen D, Horii Y. Cigarette smoking and voice fundamental frequency. J Commun Dis 1982; 15:135-44.
13. Spence AP. Anatomia humana básica. São Paulo: Manole; 1991. p. 519-21.
14. Spiegel JR, Sataloff RT, Emerich KA. The young adult voice. J Voice 1997; 11(2):138-43.
15. Rother ET, Braga MER. Como elaborar sua tese: Estruturas e referências. São Paulo: Copyright; 2001.

---

1 Speech Therapist and Audiologist, Federal University of Sao Paulo - Escola Paulista de Medicina. Specialization studies on
Oral Motricity in Oncology under course, Centro de Tratamento e Pesquisa Hospital do Câncer A. C. Camargo.
2 Speech Therapist and Audiologist, Federal University of Sao Paulo - Escola Paulista de Medicina. Specialization studies on Voice, Centro de Estudos da Voz - CEV.
3 Post-Doctorate, University of California, Davis, USA. Affiliated Professor, Department of Otorhinolaryngology and Human Communication Disorders, Federal University of Sao Paulo - Escola Paulista de Medicina. Joint Professor, Master Course in Human Communication Disorders, Universidade Tuiuti do Paraná.
4 Ph.D. in Otorhinolaryngology, Federal University of Sao Paulo - Escola Paulista de Medicina. Collaborating physician, Laryngology and Voice, Federal University of Sao Paulo - Escola Paulista de Medicina. Assistant Professor, School of Speech Therapy and Audiology, Pontifícia Universidade Católica de São Paulo.

Address correspondence to: Daniele Cannataro de Figueiredo - Rua Pedro Morcilla Filho, 256 V. Guilhermina São Paulo 03546-010
Tel/fax (55 11) 6958-9623 / 9646-8160 - E-mail: danyfig@yahoo.com.br
Scientific Study, Federal University of Sao Paulo - Escola Paulista de Medicina.
Project presented at X Congresso de Iniciação Científica na Universidade Federal de São Paulo - Escola Paulista de Medicina,
on November 10, 2002, financially supported by CNPq.
Article submitted on July 02, 2003. Article accepted on July 10, 2003.

Print: