- Email: [email protected]
The Effect of Hemodialysis on Voice: An Acoustic Analysis
The Effect of Hemodialysis on Voice: An Acoustic Analysis *Abdul-Latif Hamdan, †Walid Medawar, *Abbas Younes, *Hala Bikhazi, and *Nabil Fuleihan Beirut, Lebanon
Summary: Because respiration is part of the well-coordinated process necessary for phonation, this study was conducted with the purpose of analyzing the effect of chronic hemodialysis on voice characteristics of patients with chronic renal failure. A total of 57 patients were recruited for the study, including 31 males and 26 females ranging in age from 16 to 85 years. Patients underwent evaluation of their voice directly before and after hemodialysis using the Kay Elemetric VISI Pitch (Model 330; Kay Elemetric Corporation, Lincoln Park, New Jersey). The vocal acoustic parameters studied include habitual pitch, pitch range, relative average perturbation, shimmer, noise-to-harmonic ratio, voice turbulence index, maximum phonation time, and voice energy. The data were analyzed using the paired t-test for the total sample and the nonparametric test for the female and male subgroups. The total sample analysis showed a statistically significant increase in the habitual pitch after the hemodialyis (p ⬍ 0.05), with a borderline increase in the pitch range and maximum phonation time (p ⬍ 0.10). In the female group, there was a statistically significant increase in the habitual pitch and a borderline increase in the relative average perturbation. In the male group, there was a significant increase in the habitual pitch with a borderline increase in maximum phonation time. Discussion of the after-mentioned results is presented. Key Words: Chronic renal failure—Hemodialysis—Acoustic analysis.
INTRODUCTION Accepted for publication January 9, 2004. From the *Department of Otlaryngology—Head and Neck Surgery, American University of Beirut Medical Center, Beirut, Lebanon; †Department of Internal Medicine, American University of Beirut Medical Center, Beirut, Lebanon. Address correspondence and reprint requests to Abdul-Latif Hamdan, C/O Mohamad Hamdan, 5733 Cavender Drive, Plano, TX 75093. E-mail: [email protected] Journal of Voice, Vol. 19, No. 2, pp. 290–295 0892-1997/$30.00 쑕 2005 The Voice Foundation doi:10.1016/j.jvoice.2004.01.011
The phonatory system reflects a person’s overall well-being. The patient’s behavior and medical condition contribute to his or her vocal characteristics. As a product of well-coordinated processes, respiration, phonation, and resonation, the vocal sound reflects the delicate laryngeal muscular interplay with breathing.1 The sound produced is then enhanced in timbre intensity by the different air-filled cavities along its passage from the vocal folds to the outside air by virtue of resonance. Hence, vocal sounds can 290
HEMODIALYSIS ON VOICE be modified variably in relation to the characteristics of the resonators.2 Patients with chronic renal failure presenting for hemodialysis have been noticed to have generalized weakness, fatigue, and shortness of breath, at times affecting their voice rendering it weak, perceptually. We know that chronic renal failure can affect several body systems, among which is the respiratory system. Being the generator of the vocal signal, any respiratory system changes (eg, respiratory flow) can affect the vocal signal intensity and frequency, at least hypothetically. Chronic renal failure can result in a variety of conditions leading to muscle weakness. The muscle weakness can be caused by acid/base imbalance, electrolyte disorders, circulating ureic toxins, immune suppression, volume overload, and anemia.3 Hemodialysis is known to affect many of these factors by improving muscle weakness and endurance by virtue of its negative fluid imbalance.4,5 The clinical observation that patients with chronic renal failure improve in their overall condition shortly after dialysis has prompted us to study the effect of hemodialysis on voice. A review of the literature revealed one study by Nesic et al6 exploring the link between the psychosomatic state of patients on dialysis and the acoustic parameters of the vowel “e.” The vocal indicators of stress were considered to be the fundamental frequency, duration, and intensity. Changes in these vocal parameters, mainly an increase in fundamental frequency, were seen before the dialysis and were correlated to changes in the psychophysiological state of these patients. Because voice is the end result of aligned bodily functions rather than the consequences of psyche changes, we have undertaken the present acoustic study looking further at the effect of hemodialysis on voice in patients with renal failure, with the hypothesis that the intrinsic laryngeal muscles can be affected in a similar manner to the respiratory muscles.
MATERIALS AND METHODS A total of 57 patients with chronic renal failure undergoing hemodialysis at several medical centers in Beirut, Lebanon (ie, American University of Beirut Medical Center, Middle East Hospital, and
291
Beirut Hospital) were included in this study. Specifically, 31 males ranging in age from 20 to 86 years and 26 females aged 13 to 74 years participated. Patients underwent an evaluation of their voice immediately before and after hemodialysis using the Kay Elemetric VISI Pitch (Model 3300; Kay Elemetrics Corporation, Lincoln Park, New Jersey). With the patient seated in a quiet office, the patient’s vocal signal was recorded directly into the system (sampling rate ⫽ 50 kHz) using a condenser microphone at a distance of 15 cm from the mouth. Two VISI Pitch modules were used: the Vocal Quality Assessment (VQA) and the Pitch and Energy (PE). The variables analyzed included (1) the habitual pitch, (2) maximum phonation time, (3) pitch range, (4) voice energy, (5) relative average perturbation, (6) shimmer, (7) noise-to-harmonic ratio, and (8) voice turbulence index. The fundamental frequency in connected speech is assessed during the habitual pitch task. Specifically, during the task, the patient was asked to count to ten in a normal voice to establish habitual pitch. The maximum phonation time reveals information about the patient’s ability to sustain phonation over a long period. In this study, maximum phonation time (MPT) was calculated by asking the patient to take a deep breath and sustain phonation for as long as possible. The duration of phonation was noted. The pitch range was calculated by asking the patient to sustain the vowel/ i/ at a normal (or modal) pitch, at the lowest pitch, and then at the highest pitch (ie, from the low register to high register). Voiced energy, reported in decibels, is defined as vocal intensity, and it was measured by having the patient sustain a vowel using a vocal intensity similar to that employed in normal conversation. The relative average perturbation (RAP) measure evaluates the variability of the pitch period within a voice sample (jitter). Specifically, RAP refers to the ratio of the short-term average pitch period durations to the momentary pitch period duration, and it is presented here in percent. Shimmer (SHIM) is defined as the cycle-to-cycle amplitude variation (in dB SPL) within a voice sample. Shimmer is reported here in percent. Voice turbulence index (VTI) and noise-to-harmonic ratio (NHR) are measures unique to the VISI Pitch system and are defined as follows. VTI “is an average ratio of the spectral inharmonic high-frequency energy in the Journal of Voice, Vol. 19, No. 2, 2005
292
ABDUL-LATIF HAMDAN ET AL
range 2800-5800 Hz to the spectral harmonic energy in the range 70-4500 Hz in areas of the signal where the influence of the frequency and amplitude variations, voice breaks, sub-harmonic components are minimal.”7 NHR “is an average ratio of energy of the inharmonic components in the range 1500-4500 Hz to the harmonic components energy in the range 70-4500 Hz. It is a general evaluation of the noise presence in the analyzed signal.”7 All of these variables (ie, RAP, SHIM, VTI, and NHR) were obtained by asking the patient to sustain the vowel “ah” for 2 seconds, using the VQA module of the VISI Pitch system.
RESULTS The acoustic parameters for each subject were collected before and after hemodialysis and were analyzed using the appropriate corresponding statistical analysis. The analysis took into consideration the design (before and after) and modality of data collection (paired data). The paired t-test was conducted for the total sample (number 57) and the nonparametric test for the subgroups, the female subgroup of 26 patients, and the male subgroup of 31 patients. For the total sample analysis, there was a statistically significant increase (p ⬍ 0.05) in the habitual pitch (p ⬍ 0.000) after the hemodilaysis. Also there was an increase with borderline significance (p ⬍ 0.10) in the pitch range from 128.7814 to 154.0740 (p ⬍ 0.060) and in the maximum phonation time from 10.2730 seconds to 11.4023 seconds (p ⬍ 0.062). The rest of the acoustic parameters did not show any significant change after the hemodialysis (Table 1). The statistically significant increase in the habitual pitch noticed in the total sample group was consistently present in both the female and male subgroups. In the female subgroup, the relative average perturbation parameter also showed borderline significant increase after the hemodialysis from 0.8623 to 1.3415 (p ⬍ 0.089). The remaining acoustic variables did not show any significant changes. In the male subgroup, besides the statistically significant increase in the habitual pitch, there was also a borderline increase in the maximum phonation time from 11.9335 seconds to 13.5935 seconds Journal of Voice, Vol. 19, No. 2, 2005
(p ⬍ 0.060). The remaining acoustic variables did not show any significant changes.
DISCUSSION Sound requires four essential components for its production: a vibrating object (oscillator), an actuator or a power source to stimulate vibration, a medium to transmit the vibrations, and a receiver. Breathing (respiration) is the most important component. It acts as the power source for the emission of sound, without which the vocal folds (oscillator) cannot transform the kinetic energy of airflow into audible sounds. Breathing has basically three stages: (1) inhalation, (2) exhalation, and (3) recovery period. The process of inhalation and exhalation requires several factors, with the most important of which being contraction and relaxation of muscles of inspiration and expiration, and adequate lung elasticity to allow adequate recoil of pulmonary tissue. Patients with chronic renal failure are known to have impairment of respiratory muscle strength and endurance, which can predispose them to respiratory fatigue.8 Weiner et al4 showed that, after hemodialysis, ten patients had a significant increase in both respiratory muscle strength and endurance. They attributed it to changes in the biochemical parameters that have affected muscle performance. The vocal signal is characterized by its pitch and intensity. Both of these characteristics are the end result of harmony between airflow and laryngeal muscle biomechanics. Factors known to affect the pitch are the vocal fold length, tension, and mass, in combination with the subglottic pressure. Modifications in glottic tension and mass mediated by the interplay of intrinsic laryngeal muscles result in pitch changes. An increase in pitch can be brought about by the antagonistic action of the two vocal tensors, the thyroarytenoid and the cricothyroid, and the laryngeal dilator muscles. Contraction of muscle bundles within the thyroarytenoid, with equal vocal cord length for a given pitch, will result in an increase in tension that is isometric.9,10 Similar to respiratory muscles, the improvement in muscle performance after dialysis may reflect better contractility of the intrinsic laryngeal muscles, more specifically, the pitch raising muscles. This may explain
HEMODIALYSIS ON VOICE
293
TABLE 1. Average Values of Voice Variables for the Total Sample Including Females (n ⫽ 26) and Males (n ⫽ 31) Before and After Hemodialysis Treatment. The Variables Analyzed were Relative Average Perturbation (RAP), Shimmer (SHIM), Noise to Harmonic Ratio (NHR), Voice Turbulence Index (VTI), Habitual Pitch, Pitch Range, Maximum Phonation Time (MPT), and Voice Energy Vocal Acoustic Parameters RAP(%) SHIM (%)
NHR
VTI
Habitual Pitch (Hz) Pitch Range (Hz) MPT (Sec) Voice Energy (db)
Before After P-value
0.8075 0.9960 0.223
4.7049 4.9060 0.644
0.1772 0.04147 0.2016 0.03819 0.507 0.430
Total (n ⫽ 57) 145.0314 168.3777 0.000**
128.7814 154.0740 0.060*
10.2730 11.4023 0.062*
62.2537 62.4658 0.764
Before After P-value
0.8623 1.3415 0.089*
4.7965 5.6654 0.341
0.1434 0.03450 0.2281 0.03962 0.863 0.257
Females (n ⫽ 26) 179.7458 206.0046 0.000**
142.0246 157.8450 0.412
8.2931 8.7896 0.424
60.4665 60.6412 0.820
Before After P-value
0.7616 0.7061 0.888
4.6281 4.2690 0.518
0.2055 0.04732 0.1794 0.03700 0.241 0.348
Males (n ⫽ 31) 115.9161 136.8197 0.000**
117.6742 150.9113 0.158
11.9335 13.5935 0.060*
63.7526 63.9961 0.412
**p-value ⬍ 0.05; *p-value ⬍ 0.10.
the increase in the habitual pitch of patients with chronic renal failure after dialysis. Another important factor is the negative fluid balance effect of hemodialysis. Computerized tomography (CT) densitometry studies have shown reduction of both intravascular and extravascular volumes of the lung fluids after dialysis.11 Extrapolating this further to the larynx, a parallel decrease in the fluid content within the lamina propria of the vocal cord can induce an increase in pitch as a result of a decrease in mass. A decrease in mass per unit length can result in an increase in the frequency of vibration and hence in the vocal pitch as there is a one-toone relationship between fundamental frequency and rate of vocal fold vibration.9 The enhancement of muscle performance in the thyroarytenoid and vocalis muscle as a result of the general improvement of muscle contraction and endurance, and the reduction in the water content within the lamina propria because of the overall loss of weight and fluid after dialysis can explain the increase in the habitual pitch and pitch range in the total sample analysis as well as the increase in the habitual pitch in both female and male subgroups. Another important factor to consider in the increase in pitch level is the change
in the subglottic pressure. An increase in subglottic pressure with laryngeal tension held constant could produce a negligible rise in pitch. This information cannot be extracted from the data because airflow measures were not recorded. The increase in maximum phonation time seen in the total group and male subgroup can be accounted for by the possible increase in the vital capacity, which ought to present the quantity of air that can be exhaled after a deep inhalation. Three anatomical factors usually can affect the vital capacity: (1) the position of the body, (2) the strength of the respiratory muscles, and (3) pulmonary compliance.9 In patients with chronic renal failure, pulmonary function tests have shown a restrictive pattern with a reduction in total lung capacity, functional residual capacity, and vital capacity. This has been attributed to several factors such as fluid overload, reduction in the size of the normally aerated area, recurrent infection, and respiratory muscle weakness. After hemodialysis, there was a significant increase in total lung capacity and functional residual capacity because of a reduction in the pulmonary hypervolemia and perialveolar edema.3,11 These Journal of Voice, Vol. 19, No. 2, 2005
294
ABDUL-LATIF HAMDAN ET AL
findings in the literature may explain the increase in the maximum phonation time reported in our study. The increase in the relative average perturbation in the female subgroup after dialysis can be attributed to a decrease in the phonatory control. Recall that relative average perturbation gives an evaluation of the variability of the pitch period within the analyzed voice sample. It is an acoustic correlate of erratic vibratory patterns and hence reflects the stability of the phonatory system.12,13 The acute effects of hemodialysis on muscle performance may affect the phonatory control by enhancing the inherent sloppiness in the sustained contraction of the intrinsic laryngeal muscles.14,15 The negative fluid balance of hemodialysis may increase the relative average frequency perturbation by altering the aerodynamic influences and shift of the laryngeal mucus, which have been referred to in the understanding of the source of jitter.16,17 The voice turbulence index and noise/harmonic ratio reflects laryngeal efficiency. An incomplete glottic closure or a prolonged glottic opening results in excessive airflow that becomes turbulent and is perceived as aperiodic noise. This turbulence of noise has no harmonics, and hence, its energy is spread out across all frequencies. With no apparent cause for changes in the glottic opening or competence with hemodialysis, both the voice turbulence index and noise/harmonic ratios did not vary before and after the dialysis as expected. Stroboscopic examination of the larynx, if available, would have been of importance in the documentation of the absence of change in the glottic opening after dialysis. Unfortunately, stroboscopy was not conducted for patients in this study. Both vocal intensity and amplitude perturbation did not change after hemodialysis for males. The major determinants of loudness are the respiratory flow and subglottic pressure. With no apparent reason for changes in any of these determinants, the vocal intensity will remain the same. Because we lack proper airflow measures, we are unable to explain the consistency in vocal intensity before and after dialysis. Our results are in partial accordance with the study conducted by Nesic et al.6 These researchers reported no change in vocal intensity as in our study; however, they had higher fundamental frequency and duration preceding the dialysis that Journal of Voice, Vol. 19, No. 2, 2005
has been attributed to the anticipatory stress. This is in contrast to our study where the habitual pitch increased after hemodialysis based on the aforementioned rationale mainly changes in vocal strength. Probably a questionnaire to objectively evaluate the extent of stress in patients undergoing hemodialysis would have been helpful in delineating clearly the presence and role of stress in this group of individuals. CONCLUSION Vocal acoustic parameters may reflect early signs of various medical disorders. Patients with chronic renal failure and a weak voice may improve after hemodialysis. The effect of hemodialysis on voice characteristics showed a consistent increase in the habitual pitch seen in both male and female groups. These acoustic changes may not be perceptually noticeable to the patient or others in view of their subtle existence. These findings may not carry diagnostic potentials but do substentiate further the strong interplay between the different bodily systems and laryngeal muscles. Follow-up studies that include stroboscopy and airflow measures may help further in understanding the present findings.
REFERENCES 1. McKinney JC. The diagnosis & correction of vocal faults. Nashville, TN: Genevox Music Group; 1994. 2. Brown LB. Discover your Voice. New York: Singular Publishing Group; 1996. 3. Prezant DJ. Effect of uremia and its treatment on pulmonary function. Lung. 1990;168:1–14. 4. Weiner P, Zidan F, Zonder HB. Hemodialysis treatment may improve inspiratory muscle strength and endurance. Isr J Med Sci. 1997;33:134–138. 5. Chen CW, Lee CH, Chang HY, Hsiue TR, Sung JM, Huang JJ. Respiratory mechanics before and after hemodialysis in mechanically ventilated patients. J Formos Med Assoc. 1998;97:271–277. 6. Nesic M, Veljkovic S, Obrenovic J, Cekic S, Velickovic D, Radenkovic M. Voice frequencies in patients treated with chronic hemodialysis. Srp Arh Celok Lek. 1996;124(Suppl 1): 99–101. 7. Kay Elemetrics Corporation. Visi-Pitch II Model 3300 Instruction Manual. Lincoln Park, New Jersey, 1996:120. 8. Bark H, Heimer D, Chaimovitz C, Mostoslovski M. Effect of chronic renal failure on respiratory muscle strength. Respiration. 1988;54:153–161.
HEMODIALYSIS ON VOICE 9. Zemlin W. Speech and Hearing Science. Englewood Cliffs, NJ: Prentice Hall; 1981. 10. Greene M, Mathieson L. The Voice and its Disorders. New York: Whurr Publishers Limited; 1989. 11. Metry G, Wegenius G, Hedenstrom H, Wikstrom B, Danielson BG. Computed tomographic measurement of lung density changes in lung water with hemodialysis. Nephron. 1997;75:394–401. 12. Beckett RL. Pitch perturbation as a function of subjective vocal constriction. Folia Phonialis. 1969;21:416–425. 13. Sorensen D, Horii Y, Leonard R. Effects of laryngeal topical anesthesia on voice fundamental frequency perturbation. J Speech Hear Res. 1980;23:274–283.
295
14. Baer T. Effect of single motor unit firings on fundamental frequency of phonation. J Acoust Soc Am. 1978;64:S90. 15. Baer T. Vocal jitter: a neuromuscular explanation. In: Lawrence V, Veinberg B, eds. Transcripts of the Eighths Symposium: Care of the Professional Voice. New York: Voice Foundation; 1987:19–24. 16. Broad D. The new theories of vocal fold vibration. In: Lass NJ ed. Speech and Language: Advances in Basic Research and Practice. New York: Academic Press; 1979: 203–256. 17. Isshiki N, Tanabe M, Ishizaka K, Broad D. Clinical significance of asymmetrical vocal cord tension. Ann Otol Rhinol Laryngol. 1977;86:58–66.
Journal of Voice, Vol. 19, No. 2, 2005
Summary: Because respiration is part of the well-coordinated process necessary for phonation, this study was conducted with the purpose of analyzing the effect of chronic hemodialysis on voice characteristics of patients with chronic renal failure. A total of 57 patients were recruited for the study, including 31 males and 26 females ranging in age from 16 to 85 years. Patients underwent evaluation of their voice directly before and after hemodialysis using the Kay Elemetric VISI Pitch (Model 330; Kay Elemetric Corporation, Lincoln Park, New Jersey). The vocal acoustic parameters studied include habitual pitch, pitch range, relative average perturbation, shimmer, noise-to-harmonic ratio, voice turbulence index, maximum phonation time, and voice energy. The data were analyzed using the paired t-test for the total sample and the nonparametric test for the female and male subgroups. The total sample analysis showed a statistically significant increase in the habitual pitch after the hemodialyis (p ⬍ 0.05), with a borderline increase in the pitch range and maximum phonation time (p ⬍ 0.10). In the female group, there was a statistically significant increase in the habitual pitch and a borderline increase in the relative average perturbation. In the male group, there was a significant increase in the habitual pitch with a borderline increase in maximum phonation time. Discussion of the after-mentioned results is presented. Key Words: Chronic renal failure—Hemodialysis—Acoustic analysis.
INTRODUCTION Accepted for publication January 9, 2004. From the *Department of Otlaryngology—Head and Neck Surgery, American University of Beirut Medical Center, Beirut, Lebanon; †Department of Internal Medicine, American University of Beirut Medical Center, Beirut, Lebanon. Address correspondence and reprint requests to Abdul-Latif Hamdan, C/O Mohamad Hamdan, 5733 Cavender Drive, Plano, TX 75093. E-mail: [email protected] Journal of Voice, Vol. 19, No. 2, pp. 290–295 0892-1997/$30.00 쑕 2005 The Voice Foundation doi:10.1016/j.jvoice.2004.01.011
The phonatory system reflects a person’s overall well-being. The patient’s behavior and medical condition contribute to his or her vocal characteristics. As a product of well-coordinated processes, respiration, phonation, and resonation, the vocal sound reflects the delicate laryngeal muscular interplay with breathing.1 The sound produced is then enhanced in timbre intensity by the different air-filled cavities along its passage from the vocal folds to the outside air by virtue of resonance. Hence, vocal sounds can 290
HEMODIALYSIS ON VOICE be modified variably in relation to the characteristics of the resonators.2 Patients with chronic renal failure presenting for hemodialysis have been noticed to have generalized weakness, fatigue, and shortness of breath, at times affecting their voice rendering it weak, perceptually. We know that chronic renal failure can affect several body systems, among which is the respiratory system. Being the generator of the vocal signal, any respiratory system changes (eg, respiratory flow) can affect the vocal signal intensity and frequency, at least hypothetically. Chronic renal failure can result in a variety of conditions leading to muscle weakness. The muscle weakness can be caused by acid/base imbalance, electrolyte disorders, circulating ureic toxins, immune suppression, volume overload, and anemia.3 Hemodialysis is known to affect many of these factors by improving muscle weakness and endurance by virtue of its negative fluid imbalance.4,5 The clinical observation that patients with chronic renal failure improve in their overall condition shortly after dialysis has prompted us to study the effect of hemodialysis on voice. A review of the literature revealed one study by Nesic et al6 exploring the link between the psychosomatic state of patients on dialysis and the acoustic parameters of the vowel “e.” The vocal indicators of stress were considered to be the fundamental frequency, duration, and intensity. Changes in these vocal parameters, mainly an increase in fundamental frequency, were seen before the dialysis and were correlated to changes in the psychophysiological state of these patients. Because voice is the end result of aligned bodily functions rather than the consequences of psyche changes, we have undertaken the present acoustic study looking further at the effect of hemodialysis on voice in patients with renal failure, with the hypothesis that the intrinsic laryngeal muscles can be affected in a similar manner to the respiratory muscles.
MATERIALS AND METHODS A total of 57 patients with chronic renal failure undergoing hemodialysis at several medical centers in Beirut, Lebanon (ie, American University of Beirut Medical Center, Middle East Hospital, and
291
Beirut Hospital) were included in this study. Specifically, 31 males ranging in age from 20 to 86 years and 26 females aged 13 to 74 years participated. Patients underwent an evaluation of their voice immediately before and after hemodialysis using the Kay Elemetric VISI Pitch (Model 3300; Kay Elemetrics Corporation, Lincoln Park, New Jersey). With the patient seated in a quiet office, the patient’s vocal signal was recorded directly into the system (sampling rate ⫽ 50 kHz) using a condenser microphone at a distance of 15 cm from the mouth. Two VISI Pitch modules were used: the Vocal Quality Assessment (VQA) and the Pitch and Energy (PE). The variables analyzed included (1) the habitual pitch, (2) maximum phonation time, (3) pitch range, (4) voice energy, (5) relative average perturbation, (6) shimmer, (7) noise-to-harmonic ratio, and (8) voice turbulence index. The fundamental frequency in connected speech is assessed during the habitual pitch task. Specifically, during the task, the patient was asked to count to ten in a normal voice to establish habitual pitch. The maximum phonation time reveals information about the patient’s ability to sustain phonation over a long period. In this study, maximum phonation time (MPT) was calculated by asking the patient to take a deep breath and sustain phonation for as long as possible. The duration of phonation was noted. The pitch range was calculated by asking the patient to sustain the vowel/ i/ at a normal (or modal) pitch, at the lowest pitch, and then at the highest pitch (ie, from the low register to high register). Voiced energy, reported in decibels, is defined as vocal intensity, and it was measured by having the patient sustain a vowel using a vocal intensity similar to that employed in normal conversation. The relative average perturbation (RAP) measure evaluates the variability of the pitch period within a voice sample (jitter). Specifically, RAP refers to the ratio of the short-term average pitch period durations to the momentary pitch period duration, and it is presented here in percent. Shimmer (SHIM) is defined as the cycle-to-cycle amplitude variation (in dB SPL) within a voice sample. Shimmer is reported here in percent. Voice turbulence index (VTI) and noise-to-harmonic ratio (NHR) are measures unique to the VISI Pitch system and are defined as follows. VTI “is an average ratio of the spectral inharmonic high-frequency energy in the Journal of Voice, Vol. 19, No. 2, 2005
292
ABDUL-LATIF HAMDAN ET AL
range 2800-5800 Hz to the spectral harmonic energy in the range 70-4500 Hz in areas of the signal where the influence of the frequency and amplitude variations, voice breaks, sub-harmonic components are minimal.”7 NHR “is an average ratio of energy of the inharmonic components in the range 1500-4500 Hz to the harmonic components energy in the range 70-4500 Hz. It is a general evaluation of the noise presence in the analyzed signal.”7 All of these variables (ie, RAP, SHIM, VTI, and NHR) were obtained by asking the patient to sustain the vowel “ah” for 2 seconds, using the VQA module of the VISI Pitch system.
RESULTS The acoustic parameters for each subject were collected before and after hemodialysis and were analyzed using the appropriate corresponding statistical analysis. The analysis took into consideration the design (before and after) and modality of data collection (paired data). The paired t-test was conducted for the total sample (number 57) and the nonparametric test for the subgroups, the female subgroup of 26 patients, and the male subgroup of 31 patients. For the total sample analysis, there was a statistically significant increase (p ⬍ 0.05) in the habitual pitch (p ⬍ 0.000) after the hemodilaysis. Also there was an increase with borderline significance (p ⬍ 0.10) in the pitch range from 128.7814 to 154.0740 (p ⬍ 0.060) and in the maximum phonation time from 10.2730 seconds to 11.4023 seconds (p ⬍ 0.062). The rest of the acoustic parameters did not show any significant change after the hemodialysis (Table 1). The statistically significant increase in the habitual pitch noticed in the total sample group was consistently present in both the female and male subgroups. In the female subgroup, the relative average perturbation parameter also showed borderline significant increase after the hemodialysis from 0.8623 to 1.3415 (p ⬍ 0.089). The remaining acoustic variables did not show any significant changes. In the male subgroup, besides the statistically significant increase in the habitual pitch, there was also a borderline increase in the maximum phonation time from 11.9335 seconds to 13.5935 seconds Journal of Voice, Vol. 19, No. 2, 2005
(p ⬍ 0.060). The remaining acoustic variables did not show any significant changes.
DISCUSSION Sound requires four essential components for its production: a vibrating object (oscillator), an actuator or a power source to stimulate vibration, a medium to transmit the vibrations, and a receiver. Breathing (respiration) is the most important component. It acts as the power source for the emission of sound, without which the vocal folds (oscillator) cannot transform the kinetic energy of airflow into audible sounds. Breathing has basically three stages: (1) inhalation, (2) exhalation, and (3) recovery period. The process of inhalation and exhalation requires several factors, with the most important of which being contraction and relaxation of muscles of inspiration and expiration, and adequate lung elasticity to allow adequate recoil of pulmonary tissue. Patients with chronic renal failure are known to have impairment of respiratory muscle strength and endurance, which can predispose them to respiratory fatigue.8 Weiner et al4 showed that, after hemodialysis, ten patients had a significant increase in both respiratory muscle strength and endurance. They attributed it to changes in the biochemical parameters that have affected muscle performance. The vocal signal is characterized by its pitch and intensity. Both of these characteristics are the end result of harmony between airflow and laryngeal muscle biomechanics. Factors known to affect the pitch are the vocal fold length, tension, and mass, in combination with the subglottic pressure. Modifications in glottic tension and mass mediated by the interplay of intrinsic laryngeal muscles result in pitch changes. An increase in pitch can be brought about by the antagonistic action of the two vocal tensors, the thyroarytenoid and the cricothyroid, and the laryngeal dilator muscles. Contraction of muscle bundles within the thyroarytenoid, with equal vocal cord length for a given pitch, will result in an increase in tension that is isometric.9,10 Similar to respiratory muscles, the improvement in muscle performance after dialysis may reflect better contractility of the intrinsic laryngeal muscles, more specifically, the pitch raising muscles. This may explain
HEMODIALYSIS ON VOICE
293
TABLE 1. Average Values of Voice Variables for the Total Sample Including Females (n ⫽ 26) and Males (n ⫽ 31) Before and After Hemodialysis Treatment. The Variables Analyzed were Relative Average Perturbation (RAP), Shimmer (SHIM), Noise to Harmonic Ratio (NHR), Voice Turbulence Index (VTI), Habitual Pitch, Pitch Range, Maximum Phonation Time (MPT), and Voice Energy Vocal Acoustic Parameters RAP(%) SHIM (%)
NHR
VTI
Habitual Pitch (Hz) Pitch Range (Hz) MPT (Sec) Voice Energy (db)
Before After P-value
0.8075 0.9960 0.223
4.7049 4.9060 0.644
0.1772 0.04147 0.2016 0.03819 0.507 0.430
Total (n ⫽ 57) 145.0314 168.3777 0.000**
128.7814 154.0740 0.060*
10.2730 11.4023 0.062*
62.2537 62.4658 0.764
Before After P-value
0.8623 1.3415 0.089*
4.7965 5.6654 0.341
0.1434 0.03450 0.2281 0.03962 0.863 0.257
Females (n ⫽ 26) 179.7458 206.0046 0.000**
142.0246 157.8450 0.412
8.2931 8.7896 0.424
60.4665 60.6412 0.820
Before After P-value
0.7616 0.7061 0.888
4.6281 4.2690 0.518
0.2055 0.04732 0.1794 0.03700 0.241 0.348
Males (n ⫽ 31) 115.9161 136.8197 0.000**
117.6742 150.9113 0.158
11.9335 13.5935 0.060*
63.7526 63.9961 0.412
**p-value ⬍ 0.05; *p-value ⬍ 0.10.
the increase in the habitual pitch of patients with chronic renal failure after dialysis. Another important factor is the negative fluid balance effect of hemodialysis. Computerized tomography (CT) densitometry studies have shown reduction of both intravascular and extravascular volumes of the lung fluids after dialysis.11 Extrapolating this further to the larynx, a parallel decrease in the fluid content within the lamina propria of the vocal cord can induce an increase in pitch as a result of a decrease in mass. A decrease in mass per unit length can result in an increase in the frequency of vibration and hence in the vocal pitch as there is a one-toone relationship between fundamental frequency and rate of vocal fold vibration.9 The enhancement of muscle performance in the thyroarytenoid and vocalis muscle as a result of the general improvement of muscle contraction and endurance, and the reduction in the water content within the lamina propria because of the overall loss of weight and fluid after dialysis can explain the increase in the habitual pitch and pitch range in the total sample analysis as well as the increase in the habitual pitch in both female and male subgroups. Another important factor to consider in the increase in pitch level is the change
in the subglottic pressure. An increase in subglottic pressure with laryngeal tension held constant could produce a negligible rise in pitch. This information cannot be extracted from the data because airflow measures were not recorded. The increase in maximum phonation time seen in the total group and male subgroup can be accounted for by the possible increase in the vital capacity, which ought to present the quantity of air that can be exhaled after a deep inhalation. Three anatomical factors usually can affect the vital capacity: (1) the position of the body, (2) the strength of the respiratory muscles, and (3) pulmonary compliance.9 In patients with chronic renal failure, pulmonary function tests have shown a restrictive pattern with a reduction in total lung capacity, functional residual capacity, and vital capacity. This has been attributed to several factors such as fluid overload, reduction in the size of the normally aerated area, recurrent infection, and respiratory muscle weakness. After hemodialysis, there was a significant increase in total lung capacity and functional residual capacity because of a reduction in the pulmonary hypervolemia and perialveolar edema.3,11 These Journal of Voice, Vol. 19, No. 2, 2005
294
ABDUL-LATIF HAMDAN ET AL
findings in the literature may explain the increase in the maximum phonation time reported in our study. The increase in the relative average perturbation in the female subgroup after dialysis can be attributed to a decrease in the phonatory control. Recall that relative average perturbation gives an evaluation of the variability of the pitch period within the analyzed voice sample. It is an acoustic correlate of erratic vibratory patterns and hence reflects the stability of the phonatory system.12,13 The acute effects of hemodialysis on muscle performance may affect the phonatory control by enhancing the inherent sloppiness in the sustained contraction of the intrinsic laryngeal muscles.14,15 The negative fluid balance of hemodialysis may increase the relative average frequency perturbation by altering the aerodynamic influences and shift of the laryngeal mucus, which have been referred to in the understanding of the source of jitter.16,17 The voice turbulence index and noise/harmonic ratio reflects laryngeal efficiency. An incomplete glottic closure or a prolonged glottic opening results in excessive airflow that becomes turbulent and is perceived as aperiodic noise. This turbulence of noise has no harmonics, and hence, its energy is spread out across all frequencies. With no apparent cause for changes in the glottic opening or competence with hemodialysis, both the voice turbulence index and noise/harmonic ratios did not vary before and after the dialysis as expected. Stroboscopic examination of the larynx, if available, would have been of importance in the documentation of the absence of change in the glottic opening after dialysis. Unfortunately, stroboscopy was not conducted for patients in this study. Both vocal intensity and amplitude perturbation did not change after hemodialysis for males. The major determinants of loudness are the respiratory flow and subglottic pressure. With no apparent reason for changes in any of these determinants, the vocal intensity will remain the same. Because we lack proper airflow measures, we are unable to explain the consistency in vocal intensity before and after dialysis. Our results are in partial accordance with the study conducted by Nesic et al.6 These researchers reported no change in vocal intensity as in our study; however, they had higher fundamental frequency and duration preceding the dialysis that Journal of Voice, Vol. 19, No. 2, 2005
has been attributed to the anticipatory stress. This is in contrast to our study where the habitual pitch increased after hemodialysis based on the aforementioned rationale mainly changes in vocal strength. Probably a questionnaire to objectively evaluate the extent of stress in patients undergoing hemodialysis would have been helpful in delineating clearly the presence and role of stress in this group of individuals. CONCLUSION Vocal acoustic parameters may reflect early signs of various medical disorders. Patients with chronic renal failure and a weak voice may improve after hemodialysis. The effect of hemodialysis on voice characteristics showed a consistent increase in the habitual pitch seen in both male and female groups. These acoustic changes may not be perceptually noticeable to the patient or others in view of their subtle existence. These findings may not carry diagnostic potentials but do substentiate further the strong interplay between the different bodily systems and laryngeal muscles. Follow-up studies that include stroboscopy and airflow measures may help further in understanding the present findings.
REFERENCES 1. McKinney JC. The diagnosis & correction of vocal faults. Nashville, TN: Genevox Music Group; 1994. 2. Brown LB. Discover your Voice. New York: Singular Publishing Group; 1996. 3. Prezant DJ. Effect of uremia and its treatment on pulmonary function. Lung. 1990;168:1–14. 4. Weiner P, Zidan F, Zonder HB. Hemodialysis treatment may improve inspiratory muscle strength and endurance. Isr J Med Sci. 1997;33:134–138. 5. Chen CW, Lee CH, Chang HY, Hsiue TR, Sung JM, Huang JJ. Respiratory mechanics before and after hemodialysis in mechanically ventilated patients. J Formos Med Assoc. 1998;97:271–277. 6. Nesic M, Veljkovic S, Obrenovic J, Cekic S, Velickovic D, Radenkovic M. Voice frequencies in patients treated with chronic hemodialysis. Srp Arh Celok Lek. 1996;124(Suppl 1): 99–101. 7. Kay Elemetrics Corporation. Visi-Pitch II Model 3300 Instruction Manual. Lincoln Park, New Jersey, 1996:120. 8. Bark H, Heimer D, Chaimovitz C, Mostoslovski M. Effect of chronic renal failure on respiratory muscle strength. Respiration. 1988;54:153–161.
HEMODIALYSIS ON VOICE 9. Zemlin W. Speech and Hearing Science. Englewood Cliffs, NJ: Prentice Hall; 1981. 10. Greene M, Mathieson L. The Voice and its Disorders. New York: Whurr Publishers Limited; 1989. 11. Metry G, Wegenius G, Hedenstrom H, Wikstrom B, Danielson BG. Computed tomographic measurement of lung density changes in lung water with hemodialysis. Nephron. 1997;75:394–401. 12. Beckett RL. Pitch perturbation as a function of subjective vocal constriction. Folia Phonialis. 1969;21:416–425. 13. Sorensen D, Horii Y, Leonard R. Effects of laryngeal topical anesthesia on voice fundamental frequency perturbation. J Speech Hear Res. 1980;23:274–283.
295
14. Baer T. Effect of single motor unit firings on fundamental frequency of phonation. J Acoust Soc Am. 1978;64:S90. 15. Baer T. Vocal jitter: a neuromuscular explanation. In: Lawrence V, Veinberg B, eds. Transcripts of the Eighths Symposium: Care of the Professional Voice. New York: Voice Foundation; 1987:19–24. 16. Broad D. The new theories of vocal fold vibration. In: Lass NJ ed. Speech and Language: Advances in Basic Research and Practice. New York: Academic Press; 1979: 203–256. 17. Isshiki N, Tanabe M, Ishizaka K, Broad D. Clinical significance of asymmetrical vocal cord tension. Ann Otol Rhinol Laryngol. 1977;86:58–66.
Journal of Voice, Vol. 19, No. 2, 2005