| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Asthma Education Clinic (Dr. Balter), Mt. Sinai Hospital, University of Toronto, Toronto; the Department of Communication Sciences and Disorders (Dr. Adams), Elborn College, University of Western Ontario, London, Ontario; and the Asthma Center of The Toronto Hospital (Dr. Chapman), University of Toronto, Toronto, Canada.
Correspondence to: Meyer S. Balter, MD, FCCP, Mt. Sinai Hospital, 640–600 University Ave, Toronto, Ontario M5G 1X5; e-mail: mbalter{at}mtsinai.on.ca
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Objective: We undertook the following study to test the ability of an acoustic analysis technique to quantify changes in vocal function in steroid-naive patients with asthma who receive inhaled beclomethasone dipropionate (BDP), 1,000 µg/d for 4 months.
Methods: Patients self-administered one of four regimens of inhaled BDP. Group 1 patients received one 250-µg puff qid via metered-dose inhaler (MDI); group 2 patients received one 250-µg puff qid via MDI with a holding chamber; group 3 patients received two 250-µg puffs bid via MDI; and group 4 patients received two 250-µg puffs bid via MDI with a holding chamber. A smaller cohort of nonsmoking asthmatic patients was managed without steroid intervention for 4 months. At baseline and again at 8 weeks and 16 weeks after the initiation of BDP treatment, patients underwent spirometry and methacholine challenge. At baseline and again at 2, 4, 8, 12, and 16 weeks, patients underwent voice recording for analysis of voice parameters. The recorded vowels were low-pass filtered (10 KHz), digitized (22 KHz), and analyzed by software to obtain two acoustic measures: (1) jitter, the cycle-to-cycle variation in the time period of the voice signal; and (2) shimmer, the cycle-to-cycle variation in voice signal amplitude.
Results: We recruited 77 patients for randomization to inhaled steroid therapy and 10 patients who continued to receive only occasional inhaled bronchodilator therapy. In all active treatment groups, FEV1, FVC, and provocative concentration of methacholine causing a 20% fall in FEV1 improved significantly after BDP treatment. Mean jitter scores, a measurement of variation in voice pitch, were not significantly influenced by BDP treatment. However, mean shimmer scores, a reflection of perturbation in vocal amplitude, fell significantly (p < 0.05) in the active treatment groups. These reductions in shimmer scores were not significantly different in the active treatment groups. Shimmer scores in the bronchodilator-treated group were unchanged during the 16 weeks of follow-up.
Conclusions: Our data show that a simple and noninvasive acoustic analysis of voice is sensitive to subclinical changes associated with inhaled corticosteroid therapy. We have shown that 1,000 µg/d of inhaled BDP actually improves specific acoustic measures of voice in patients with inadequately controlled asthma. These improvements were uninfluenced by dosing schedule and whether a spacing chamber was used.
Key Words: asthma • inhaled steroids • voice
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the development of novel corticosteroids and their delivery devices, the demonstration of clinical efficacy is relatively straightforward. Depending on the severity of asthma under study, a variety of pulmonary function and clinical end points can be used to quantify the improvement in asthma control. By contrast, the evaluation of side effects is more difficult. Hoarseness, in particular, is a highly subjective symptom. Patients who rely on their voices professionally might be sensitive to and intolerant of a trivial change in voice quality that would go unnoticed by patients who do not use their voices in such a fashion. Although self-reporting by patients may lack the necessary detail and sensitivity, it is the usual means of monitoring for upper-airway side effects in clinical trials of inhaled corticosteroids. Visual inspection of the upper airway for Candida growth or quantitative cultures for Candida species are also occasional components of inhaled corticosteroid clinical trials; however, these strategies measure a phenomenon usually unrelated to changes in voice quality itself. Williams and colleagues6 attributed the dysphonia of inhaled corticosteroid use to steroid myopathy of the vocal cords. Using laryngoscopy, they studied a small group of patients prospectively and found subtle myopathic changes appearing in the vocal cords of patients even before changes in voice quality were clinically apparent. Laryngoscopy is an invasive procedure that is ill suited for routine clinical monitoring of voice abnormalities. During the past 20 years, several acoustic measures have been developed for the objective and noninvasive measurement of vocal function. In addition, advances in computer hardware and software have led to the development of many inexpensive and commercially available systems for obtaining these objective acoustic measures of voice.7 8 Two of the most commonly used acoustic measures of voice quality are vocal jitter and vocal shimmer. These measures have been reported in at least 30 published articles during the past 20 years.9
In the present study, we obtained vocal jitter and shimmer measures in an attempt to quantify changes in vocal function in patients who were receiving inhaled corticosteroids. In particular, we sought to determine the effect of inhaled beclomethasone dipropionate (BDP), 1,000 µg/d for 4 months, on acoustic measures of voice in steroid-naïve patients with asthma. In addition to testing the pretreatment and posttreatment voice measures of asthmatics receiving BDP, we compared voice measures in the steroid-treated group to voice measures in a population of patients with asthma who had not received inhaled steroids. Finally, we assessed the impact of a holding chamber and the impact of twice-daily vs four-times-daily dosing schedules on voice changes resulting from inhaled steroid use.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2.0 mg/mL. Potential subjects were excluded if they (1) had
smoked tobacco in the year before enrollment; (2) had received
treatment with systemic or inhaled steroids in the 2 months before
enrollment; (3) had been hospitalized for asthma management in the 2
months before enrollment; (4) had suffered an upper respiratory tract
infection in the month before enrollment; or (5) had a history of
laryngeal disease. Each patient signed an informed consent form. The
study was approved by the Human Subjects Review Committee of the
University of Toronto.
Study Design
In two university hospital asthma centers, we recruited
nonsmoking patients with asthma who were not receiving inhaled
corticosteroids, and we randomized them to one of four treatment
groups. For 4 months, patients self-administered one of four regimens
of inhaled BDP. Group 1 patients received one 250-µg puff qid via
metered-dose inhaler (MDI); group 2 patients received one 250-µg puff
qid via MDI with a holding chamber (Aerochamber; Trudell Medical;
London, Ontario); group 3 patients received two 250-µg puffs
bid via MDI; and group 4 patients received two 250-µg puffs bid via
MDI with a holding chamber. All patients were taught the proper use of
their delivery devices and were advised to rinse their mouths with
water, and gargle and spit after receiving inhaled corticosteroids. We
also recruited a smaller cohort of nonsmoking patients with asthma who
were not in need of inhaled steroid therapy or did not wish to use
them, and who were monitored without steroid intervention for 4 months.
At baseline and again at 2, 4, 8, 12, and 16 weeks after the initiation
of inhaled BDP therapy, patients underwent voice recording for later
blinded analysis of voice parameters. At baseline and again at 8 weeks
and 16 weeks after the initiation of inhaled BDP treatment, patients in
the active treatment groups underwent spirometry and methacholine
challenge. Patients in the control group were assessed by spirometry
but not by methacholine challenge.
Procedures
Spirometry was performed using a dry rolling seal spirometer,
following the recommendations of the American Thoracic Society.
Methacholine challenges were performed by a modification of the method
of Cockcroft et al.10
All voice recordings were made using a head-mounted condenser microphone (model HY3; TOA; Edmonton, Alberta), with a constant mouth-to-microphone distance of 7 cm, and an audiocassette recorder (model 112R; TASCAM; Mississauga, Ontario). After a demonstration, patients were asked to take a deep breath and to sustain the vowel sound "ah" steadily at a comfortable intensity and pitch for as long as possible. This process was repeated three times. Each subject’s second production of the vowel sound was used in all subsequent analyses.
At the completion of the study, recordings were supplied in random sequence and in blinded fashion to a speech pathologist who was unaware of the patient’s treatment allocation and whether the recording represented a pretreatment or posttreatment condition. The recorded vowels were low-pass filtered (10 KHz), digitized (22 KHz), and analyzed using a computer, an IBM Audio Capture and Playback Adapter board (IBM; Toronto, Ontario), and CSpeech Software (Madison, WI). A 1,000-ms segment starting 500 ms from the onset of the vowel was used to obtain two acoustic measures: jitter, the cycle-to-cycle variation in the time period of the voice signal; and shimmer, the cycle-to-cycle variation in voice signal amplitude.
Jitter and shimmer are indirect acoustic measures of variations in vocal fold vibration. Several previous studies11 12 13 14 15 16 17 have found that jitter and shimmer can be significantly correlated with the perception of abnormal voice qualities, including hoarseness, breathiness, and harshness. During the production of voice, the vocal folds rapidly alternate between adduction (the folds come together) and abduction (the folds separate). For most adults, this adductory/abductory cycle generally occurs at rates between 100 cycles per second and 300 cycles per second. Thus, the period between vocal cord vibratory cycles is between approximately 4 ms and 10 ms. Each of these adductory/abductory voice cycles is associated with the production of a transient pulse of sound. Using acoustic analysis techniques, it is possible to precisely measure the peak amplitude and timing of these sound pulses. Vocal shimmer is determined by measuring the amplitude of one vocal pulse and subtracting it from the amplitude of the next adjacent vocal pulse. This difference in amplitude is calculated for each set of adjacent pulses across several hundred pairs of pulses. The average of these amplitude difference values is calculated and is referred to as shimmer. Thus, shimmer is an estimate of the cycle-to-cycle variation in the amplitude of vocal pulses produced during voice production. Shimmer units are often expressed as a percentage of the average vocal pulse amplitude in decibels (dB%). The average shimmer value obtained for normal adults is approximately 1.6 dB% (SD, ± 0.5 dB%).18 19
Vocal jitter is determined by measuring the duration between adjacent vocal pulses. The difference in time is calculated for each pair of adjacent pulses across several hundred pairs of voice pulses. The average of these durational difference values is calculated and referred to as jitter. Thus, jitter is an estimate of the cycle-to-cycle variation in the time period between adjacent vocal pulses produced during phonation. Jitter is often expressed in millisecond units. The average jitter value for normal adults is approximately 0.02 ms (SD, ± 0.01 ms).18 19 More detailed descriptions of jitter, shimmer, and related acoustic vocal perturbation measures can be found elsewhere.20 21
Data Analysis
All statistical procedures were performed using software
(Minitab; State College, PA). Pretreatment and posttreatment
spirometric variables were compared by Student paired t
test. Differences among groups in spirometric variables were compared
by one-way repeated analysis of variance with specific pairwise
comparisons between groups when indicated by unpaired t
tests. Wilcoxon signed-rank test was used to compare pretreatment and
posttreatment voice parameter scores and PC20
within treatment groups. The Mann-Whitney U test was used to
examine differences among groups. Differences were considered
significant at the p < 0.05 level.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Respiratory Responses to Inhaled BDP
In all active treatment groups, FEV1, FVC,
and PC20 improved significantly after inhaled BDP
treatment. Mean FEV1 rose from 2.60 ± 0.52 to
2.92 ± 0.69 L (p < 0.0001), mean FVC rose from 3.86 ± 0.94 to
4.06 ± 1.26 L (p < 0.005), and mean PC20
fell from 4.60 ± 5.77 to 1.33 ± 2.52 mg/mL (p < 0.001).
Response of Voice Parameters to Inhaled BDP
Mean jitter scores were not significantly influenced by inhaled
BDP treatment. However, mean shimmer scores fell significantly in the
active treatment groups (p < 0.05). Changes were evident at week 2
and were essentially maximal by week 4 (Fig 1
). Subsequently, shimmer scores did not change significantly during the
remaining 12 weeks of active treatment. Overall, the combined active
treatment groups showed average vocal shimmer values that fell from
approximately 65% above normal to approximately 30% above normal
after treatment. These reductions in shimmer score were not
significantly different among the active treatment groups. Shimmer
values in the bronchodilator-treated group remained unchanged and
approximately equivalent to previously reported normal values during
the 16 weeks of follow-up.
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our findings are important in several ways. First, the data are reassuring for patients with asthma who fear impaired voice performance if inhaled corticosteroids are prescribed. Second, our findings provide a potentially useful clinical and, more important, clinical trial tool for assessment and quantification of vocal function during long periods of time. All patients received corticosteroid therapy for frequent episodes of cough, wheeze, or both; at baseline, the average FEV1 was 20% below the predicted normal value. Our data are most consistent with impairment in voice acoustics arising from troublesome cough in patients with untreated or undertreated asthma. We cannot rule out the possibility that coexisting upper-airway disease, particularly allergic rhinoconjunctivitis, contributed to the elevated baseline shimmer scores in the treated groups. However, our therapy was not directed toward the upper airway, and we doubt that significant changes in nasal symptoms or postnasal drip occurred as a consequence of the active treatment. It might be argued that the use of increased shimmer scores as an index of voice impairment could be the consequence of frequent ß2-agonist use. ß2-Agonists do produce skeletal muscle tremor, an effect most notable in the outstretched extremities. We cannot rule out such an effect on vocal cord motion arising from ß2-agonist use. One could postulate that the inhaled corticosteroids reduced ß2-agonist use and thus resulted in less perturbation of normal vocal cord motion by the adrenergic drug. We think that this explanation of our findings is unlikely. We found no evidence of impaired voice acoustics in patients treated with ß2-agonists alone.
Some limitations to our findings should be noted. Our study was limited to 4 months. We cannot rule out the possibility that continued inhaled corticosteroid use would have resulted in subsequent deterioration in voice acoustics. One explanation for dysphonia after inhaled corticosteroid use is vocal cord muscle atrophy. Nevertheless, the present trend toward controlling asthma is to obtain control at a reasonable initial dose of inhaled corticosteroids and then to titrate the dosage downwards to the least amount needed to maintain control. Our findings would reassure that the voice will not likely be impaired by doses of BDP as high as 1,000 µg/d during at least a 4-month period, a period sufficient to obtain control in the majority of asthma sufferers. Our findings did not compare various doses of inhaled corticosteroid, and we do not know whether the changes in voice acoustics are dose related. We suspect that improvement in voice quality is the consequence of improved asthma control and as such is unlikely to have any simple dose-response relationship. Our study did not examine the influence of alternative delivery systems on vocal function. The development of alternatives to chlorofluorocarbon-containing MDIs will include alternative propellant pressurized inhalers, dry-powder inhalers, and other technologies. However, our study does provide information on how the impact of these technologies on voice acoustics could be evaluated. Our study did not incorporate a placebo-treated group. Therefore, we cannot rule out changes in voice occurring spontaneously or as "regression toward the mean" in a group of patients who presented with symptomatic asthma and associated voice abnormalities. However, we believed that the incorporation of a placebo-treated group would be unethical for the 4-month treatment of patients with poorly controlled disease. We also think that the time course of improvement in shimmer scores is more consistent with active treatment response than seasonal variability in asthma.
The results of this vocal acoustics study provide significant evidence of a change in vocal function in asthmatic patients after treatment with inhaled BDP. These results also suggest that this treatment may be associated with improvements in perceived voice quality. Additional studies using listener judgments and perceptual analysis procedures could be useful. Despite studies23 24 25 that have questioned the sensitivity and reliability of perceptual voice quality measures, we believe that both acoustic and perceptual measures can provide important information about vocal function in patients receiving treatment for asthma.
Our study removes one of the smaller barriers that may have discouraged early administration of inhaled corticosteroids to patients with poorly controlled asthma. The unexpected finding of improvement in voice acoustics is most readily explained by a reduction in cough with a consequent reduction in upper-airway stress. The measurement and analytic techniques described are simple enough for clinical use but are most likely to find their application in clinical research.
|
Footnotes |
|---|
Supported in part by a grant of the PSI Foundation of Ontario, Canada (Dr. Balter).
Received for publication August 3, 1999. Accepted for publication June 13, 2001.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Derendorf, R. Nave, A. Drollmann, F. Cerasoli, and W. Wurst Relevance of pharmacokinetics and pharmacodynamics of inhaled corticosteroids to asthma. Eur. Respir. J., November 1, 2006; 28(5): 1042 - 1050. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
