Effects of a Circumferentially Vented Mask on Breathing Patterns of

1 downloads 0 Views 115KB Size Report
Thus we sought to identify whether changes in breathing patterns occur with the use of a specific .... nursery rhyme before coming to the session, and before beginning the ... Model 1700 sound pressure meter) outside the pneumo- tachograph ...
472

JSLHR JSLHR,, Volume Volume 41, 41, 472–478, 472–478, June June 1998 1998

Effects of a Circumferentially Vented Mask on Breathing Patterns of Women as Measured by Respiratory Kinematic Techniques Jessica E. Huber Elaine T. Stathopoulos State University of New York at Buffalo

Lori A. Bormann The Speech Pathology Group Walnut Creek, CA

Kenneth Johnson State University of New York at Buffalo

Since pneumotachograph masks are commonly used in studies of speech breathing, the purpose of this study was to measure the differences in respiratory volumetric and frequency measures during speech under two conditions: with and without a circumferentially vented pneumotachograph mask coupled to the face. Thus we sought to identify whether changes in breathing patterns occur with the use of a specific face mask, because these patterns are accepted as representative of normal speech breathing. Subjects were 10 normal-speaking women, each of whom produced a syllable train and a connected speech task, both at comfortable intensity levels. Respiratory measures were made using linearized magnetometers during speech production. The measurements included lung volume, rib cage volume, and abdominal volume at utterance initiation and termination, volume excursions during the utterance, and the number of breath groups during the speech task. There were no significant differences between the mask-on and mask-off conditions in volumetric and frequency measures. A significant task difference for abdominal initiation was found. It was concluded that the use of a circumferentially vented pneumotachograph mask does not alter the reliability of respiratory volume and frequency measures for studies of voice. KEY WORDS: women, pneumotachograph mask, breathing, kinematics

A

pneumotachograph mask is commonly used to measure the airflow produced during speech. Previous research has indicated that a full face mask can restrict articulatory movements, especially those involving the jaw, as well as alter auditory feedback (Klatt, Stevens, & Mead, 1968). Studies have shown that the Rothenberg circumferentially vented pneumotachograph mask (Rothenberg, 1973) does not affect selected acoustic measures (e.g., Rothenberg, 1973; Till, Jafari, Crumley, & Law-Till, 1992). There is still concern on the part of researchers that a face mask may affect the normal breathing patterns of subjects (e.g., T. J. Hixon, personal communication, February 1995; L. Ramig, personal communication, September 1996; Winkworth & Davis, 1997). Winkworth and Davis state, “Use of a mask to measure [laryngeal and/or supralaryngeal data] might have constituted a barrier between the subject and her communication partner and possibly changed other speech and respiratory characteristics” (p. 167). Even though the concerns of these researchers have not been verified in any speech studies for a particular type of mask, their concerns

Journal of 472 Journal Speech, of Speech, Language, Language, and Hearing and Hearing ResearchResearch

1092-4388/98/4103-0472

©1998, American Speech-Language-Hearing Association

473

Huber et al.: Effect of Airflow Mask on Breathing Patterns

may be substantive according to the respiratory physiology literature. There are at least three possible reasons that a standard respiratory face mask could alter breathing patterns. These include (1) a buildup of carbon dioxide (CO2) in the mask, (2) a neural response due to the stimulation of sensory receptors in the face, and (3) a psychological or emotional response to the experimental situation and/or instrumentation (Askanazi et al., 1980; Comroe, 1950; Hirsch & Bishop, 1982; Winkworth & Davis, 1997). Because the circumferentially vented pneumotachograph face mask continues to be an integral part of speech research, it is important to determine whether wearing this face mask changes the normal breathing patterns of subjects.

Effects of Carbon Dioxide Buildup on Breathing The buildup of CO2 in the circumferentially vented pneumotachograph mask is a possible cause for change in speech breathing patterns (Comroe, 1950). There are two major ways a buildup in CO2 could alter breathing patterns. First, an increase in CO2 has an effect on volume (Hirsch & Bishop, 1982). If CO2 were to build up in the mask, the most likely response would be an increase in depth of breathing, thereby increasing the volume initiation for an utterance. Second, the frequency of breathing may increase, causing the subject to use more breath groups to accomplish the speaking task. Changes in the tidal volume with a mask on versus off may be due to the dead space in the mask, the space between the subject’s face and the end of the mask where CO2 can build up (Askanazi et al., 1980). It should be noted that a standard respiratory mask used by respiratory physiologists is different from the circumferentially vented pneumotachograph mask typically used in speech studies, which has a negligible dead space.

Effects of Facial Stimulation on Breathing The second reason suggested for a change in breathing pattern has been that the contact pressure caused by a face mask stimulates the sensory receptors in the face (Hirsch & Bishop, 1982). The exact neural mechanisms linking the oral-facial system to breathing patterns are unknown; however, it is believed that stimulation of the trigeminal receptors in the oral-facial region exercise important control over respiratory output (Capra, 1995; Hirsch & Bishop, 1982). Hirsch and Bishop (1982) imply that a change in breathing pattern could occur with the use of a standard respiratory mask pressing on the face; however, data showing specific pattern changes were not presented. It is unclear whether a standard respiratory mask significantly changes breathing patterns, however.

Askanazi et al. (1980) concluded that “the use of a mask…induces an increase in both tidal volume and minute ventilation…[which] may be due…to sensory stimuli arising from the face, mouth, and nose” (p. 579).

Reaction to Experimental Situation/ Instrumentation A third reason for possible changes in respiratory patterns could be a reaction of the subject to the use of instrumentation in the experimental situation. Studies have been completed comparing breathing patterns with the use of a variety of facial apparatus (including standard respiratory face masks, mouthpieces, and noseclips) to breathing patterns without any apparatus (Askanazi et al., 1980; Gilbert, Auchincloss, Brodsky, & Boden, 1972; Hirsch & Bishop, 1982; Sackner, Nixon, Davis, Atkins, & Sackner, 1980). Askanazi et al. (1980) concluded that changes “may be due to a psychological load imposed by the mask” (p. 579). Again, Winkworth and Davis (1997) suggested that the face mask may create a barrier between the subject and his/her communication partner, thereby possibly changing the person’s breathing patterns. The present speech study was designed to investigate the question whether the use of a Rothenberg (1973) circumferentially vented pneumotachograph face mask alters volumetric and frequency measures of subjects during speech. It was believed that if any one of the three factors mentioned above changed breathing patterns, the change would be reflected in respiratory kinematic and volumetric measures. The specific hypothesis was that breathing patterns would change with the Rothenberg circumferentially vented pneumotachograph mask coupled to the face versus no face-mask-coupling.

Method Subjects Ten women participated in this study. The mean age of the women was 22 years 7 months (range: 21 years 9 months to 24 years 5 months). Men were not included, even though it has been established that men and women differ in volumetrics (Stathopoulos & Sapienza, 1993). The focus of the current study was how these patterns change with the addition of a pneumotachograph face mask and not whether differences exist. None of the previous investigations of the effects of a standard respiratory mask indicate that men are more susceptible to the effects of CO2 buildup, facial stimulation, or instrumentation than women. Additionally, we know from previous investigations that the O2 consumption and CO2 production is the same for men and women when the

Journal of Speech, Language, and Hearing Research

474 values are normalized for mass, indicating no expectation that men would differ from women in their breathing patterns with the use of a face mask (Bunn & Mead, 1971; Russell, Stathopoulos, & Cerny, 1998). Criteria for subject selection included (a) normal speech and voice; (b) general North American English dialect; (c) no major health problems, such as head or neck surgery, asthma, pneumonia, or history of smoking; (d) freedom from allergies, infections, and colds on the day of testing; and (e) no professional singing and/or voice training.

Speech Tasks Two types of speech tasks were performed: (1) a syllable train, and (2) a connected speech task. A syllable train was chosen because it is a very common speech task in studies that examine respiratory kinematics and aerodynamic measures. Such studies require use of an airflow mask. For the connected speech task, a nursery rhyme was chosen instead of reading, because it was a recitation task that could be standardized across subjects. Additionally, it is more like conversational speech than reading, because the subjects inhale when they choose without written punctuation affecting decisions to inhale. Winkworth et al. (1994) found that subjects inhaled at the same point when reading; they concluded that “the existence of individual breathing patterns during speech would be difficult to determine from reading tasks” (p. 551). For the syllable train, each subject produced 7 syllables of /pA/ at their comfortable sound pressure level. Six trials were produced for each condition of mask-on versus mask-off. The order of the mask condition (on versus off) was randomly selected for each subject. The syllable train was produced at a rate of 1.5 syllables per second on one breath. An electronic metronome (Metone model 23-F) was used to help the subjects produce the syllable task at the designated rate. The mask was on for about 1 minute for each of the syllable train tasks, and when two mask-on conditions were performed consecutively, the mask was not removed between trials. No instruction regarding depth of inspiration was given to the subject. When the subject reached a steady endexpiratory level, she was told to begin the syllable train when ready. For the connected speech task, each subject was required to recite six trials of the nursery rhyme Baa Baa Black Sheep. The subjects were asked to learn the nursery rhyme before coming to the session, and before beginning the speech task, the subjects were given an opportunity to practice the rhyme to ensure they were comfortable with the task. The first three trials were produced with the face mask off, and the last three trials were produced with the face mask on. Thus, the face

Journal of Speech, Language, and Hearing Research

JSLHR, Volume 41, 472–478, June 1998

mask was not removed between the three trials of the connected speech task, which lasted about 2 to 3 minutes. Each subject recited the nursery rhyme from memory in a natural, conversational manner. There was no instruction regarding depth of inspiration or frequency of breathing given to the subject. When the subject reached a steady end-expiratory level, she was told to begin the nursery rhyme when ready.

Equipment and Procedures A circumferentially vented pneumotachograph mask (Rothenberg, 1973), hereafter referred to as a pneumotachograph mask, was used for the mask-on conditions. The acoustic signal was obtained from a free field microphone (Quest condenser microphone with a Quest Model 1700 sound pressure meter) outside the pneumotachograph mask, with a mouth-to-microphone distance of 30 cm. Respiratory function was monitored with linearized magnetometers (GMG Scientific, Inc., Burlington, MA) using procedures developed in the Speech Production Laboratory at the State University of New York at Buffalo (Stathopoulos & Sapienza, 1993) and modeled after Hoit and Hixon (1987). Signals from the magnetometers were monitored on an x–y oscilloscope (Tektronix 5111A) during all speech tasks. The wide-band airflow signal was sensed and transduced with a circumferentially vented wire screen pneumotachograph face mask and transducer (PTW; Rothenberg, 1973). The data were digitized directly to CSpeech 3.0 (Milenkovic, 1989) using a computer (IBM P90) with an analog-to-digital conversion board (Data Translation 12-bit DT2821) at a sampling rate of 2 kHz per channel for respiratory kinematic measures. Calibration of distance-to-voltage output of the magnetometers was determined by using a ruler marked off in centimeter steps. One magnetometer was held stationary while the other was systematically moved across a 22.3-cm range. The output signal of the magnetometers was linear across this range. The air pressure transducer (Glottal Enterprises) and mask were calibrated with a known flow through a rotameter (Fischer-Porter Co. RS-5771) before collecting data on each subject. A lung volume waveform was generated by integrating the average airflow waveform sensed during tidal volume breathing. The integration procedure was verified by displacing 2 liters of air from a Collins Respirometer (9.0-liter model) through the mask/transduction system and confirming that the integrated flow signal on CSpeech represented the full 2 liters of air displaced (Stathopoulos & Sapienza, 1993). All volume measures from the integration procedure were found to be within ±5% of the volume measured from the respirometer.

475

Huber et al.: Effect of Airflow Mask on Breathing Patterns

Calibration respiratory maneuvers were performed in a seated position a minimum of three times by each subject. These included an isovolume maneuver, vitalcapacity/maximum-rib-cage-capacity maneuver, maximum-abdominal-capacity maneuver, and relaxation maneuver (Hoit & Hixon, 1987). Data in centimeters were converted into percent vital capacity (%VC), percent rib cage capacity (%RCC), and percent abdominal capacity (%ABC), all ranging from 0% to 100%, during the measurement analysis procedure.

Lung volume initiation (LVI) Lung volume termination (LVT) Lung volume excursion (LVE; LVI minus LVT) 2.

Rib cage volume initiation (RCVI) Rib cage volume termination (RCVT) Rib cage volume excursion (RCVE; RCVI minus RCVT) 3.

Data Collection and Measurements End-expiratory level (EEL) was determined by allowing the subjects to stabilize their rest breathing before each speech task. End-expiratory level was taken as the average of three consecutive minimum values before each utterance from the lung, rib cage, and abdominal volume waveforms (see Figure 1, horizontal line C). Thereafter, all volumes were measured relative to each subject’s EEL. Any displacement values above EEL were designated as positive, and values below EEL were negative. The following measurements (Hoit & Hixon, 1987; Stathopoulos & Sapienza, 1993) were obtained under both the mask-on and mask-off conditions: 1.

Lung volume measures, all relative to EEL and expressed as %VC:

Rib cage measures, all relative to EEL and expressed as %RCC:

Abdominal measures, all relative to EEL and expressed as %ABC: Abdominal volume initiation (ABVI) Abdominal volume termination (ABVT) Abdominal volume excursion (ABVE; ABVI minus ABVT)

Also, the relative volume contribution (RVC) of the rib cage versus the abdomen towards lung volume change was measured. Frequency of breathing was defined as the number of breath groups during the connected speech task. The volumetric and frequency measures were chosen because they are the typical measures used to track lung, rib cage, and abdominal volumes during speech (e.g., Hoit & Hixon, 1987; Hoit, Hixon, Watson, & Morgan, 1990; Stathopoulos & Sapienza, 1997). Lung volume measures were estimated from the sum of the rib cage and abdominal displacements based

Figure 1. Audio, rib cage (RC), and abdomen (AB) waveforms. Point A = utterance initiation; point B = utterance termination; horizontal line C = end-expiratory level (EEL).

Journal of Speech, Language, and Hearing Research

476

JSLHR, Volume 41, 472–478, June 1998

on the theory that lung volume change is the sum of the changes in the anteroposterior dimension of the rib cage and abdomen (Konno & Mead, 1967). The rib cage and abdominal contribution coefficients were estimated for the expiratory volumes by using a least squares analysis (Chadha et al., 1982). For the syllable train, each utterance initiation was defined from the acoustic waveform as the point when voicing began for the first /A/ of the syllable train (see Figure 1, point A). The utterance termination was defined by the end of voicing for the last vowel of the syllable train (see Figure 1, point B). For the connected speech task, each utterance initiation was defined on the acoustic waveform as the point when voicing began for the first phoneme of the breath group, whereas the utterance termination was defined by the end of voicing for the last phoneme of the breath group. Respiratory kinematic measurements were taken from the rib cage and abdominal displacements at utterance initiation and termination of each respective waveform, with the excursion measures calculated as initiation amplitude minus termination amplitude (see Figure 1, points A and B).

Statistical Analysis Descriptive and inferential statistics were used to analyze the respiratory volumetric and breathing frequency data. The measures listed above were obtained from each trial of the syllable train (6 trials × 2 mask conditions × 10 subjects = 120 observations). The values were then averaged to calculate group means and standard deviations. For the connected speech task, measures were obtained from each breath group (3 trials × 2 mask conditions × 10 subjects with a varied number of breath groups). The breath group measures for each

subject were then averaged to calculate group means and standard deviations. The group means were analyzed using a repeated-measures analysis of variance (ANOVA) (JMP, version 3.1.5). The repeated measures were the speech task (syllable versus connected) and mask condition (mask on versus mask off). The alpha level was set conservatively at a probability level of .01 because of the increased number of conditions. The mask-on versus mask-off comparison for breathing frequency was determined by a paired t test (JMP, version 3.1.5), with an alpha level of 0.05.

Results The means and standard deviations for the dependent variables are shown in Table 1. Examination of the means shows that there was not a consistent direction of effect and that many times the differences were quite small. During the syllable train, 6 of 10 measures were larger for the mask-on condition than the mask-off condition. For the connected speech task, the mean differences were generally larger for the mask-off condition than the mask-on condition, but the differences were quite small. The statistical analyses showed that there was not a significant main effect for the mask condition (off versus on) for any of the measures. The main effect for speech task (syllable versus connected) was significant for ABVI (p = 0.01). There were no significant interaction effects between the mask and task conditions. The paired t test used to test differences in frequency of breathing during the connected speech task showed there were no significant differences between the face mask on and the face mask off in the number of breath groups in the connected speech task, p(9) = 0.6783.

Table 1. Means (and standard deviations) of the two speech tasks (syllable and connected speech) under the two conditions (face mask on and face mask off). Syllable train Measure LVI LVT LVE RCVI RCVT RCVE ABVI ABVT ABVE RVC No. Breath groups

Connected speech

Face mask off

Face mask on

15.08 2.58 12.51 14.51 –2.69 17.19 9.27 –7.11 16.38 73.10

17.38 3.50 13.88 14.52 –0.82 15.34 14.64 –6.18 20.82 69.23

(6.5) (8.5) (7.8) (7.6) (10.3) (7.6) (7.4) (12.6) (10.7) (35.3)

(10.6) (10.3) (4.7) (7.5) (10.9) (8.1) (11.6) (12.9) (10.8) (22.8)

Face mask off 16.30 (8.6) 1.84 (8.1) 14.47 (8.0) 12.67 (6.7) –1.83 (6.8) 14.51 (7.9) 5.96 (9.9) –10.19 (14.7) 16.15 (18.2) 73.19 (23.6) 2.90 (1.1)

Note. All initiations and terminations are in relation to end-expiratory level.

Journal of Speech, Language, and Hearing Research

Face mask on 16.27 2.80 13.47 12.07 –0.38 12.45 7.46 –7.38 14.84 68.36 3.00

(9.1) (7.2) (6.1) (6.8) (5.8) (6.9) (11.3) (14.9) (12.3) (18.8) (1.2)

477

Huber et al.: Effect of Airflow Mask on Breathing Patterns

Discussion In the present study, there were no differences in speech breathing measures whether or not subjects wore the pneumotachograph mask. This indicated that none of the possible effects, CO2 buildup in the mask, facial stimulation, or reaction to the instrumentation, had an effect on the kinematic or volumetric measures of the subjects. No breathing pattern predictions were made relative to sensory stimulation or reaction to equipment, because no previous data had ever substantiated these effects. The predicted response to CO2 buildup was higher volume initiations and more breath groups while the mask was coupled to the face. The data from the current study were not consistent with these expected volumetric and frequency responses in a situation where CO2 would build up. The reason for the apparently negligible buildup of CO2 in the pneumotachograph mask is the reduction of ventilatory dead space because of the inclusion of multiple screens that allow free gas exchange. Furthermore, when small volume changes occurred at initiations, the LVI and RCVI with the face mask on were lower than with the face mask off. It would be logical to expect that the breathing patterns during the connected speech task would show a larger difference than in the syllable task, because the mask was left on for a longer time. This was not observed. The abdominal initiations were more different in the syllable task (5% higher for mask on than mask off) than in the connected speech task (1.5% higher for mask on than mask off). A buildup of CO2 in the mask could cause an increase in the frequency of breathing, causing the subjects to use more breath groups to complete the connected speech task. However, there were no significant differences in the number of breath groups with the mask on versus the mask off. In the present study, ABVI was higher in the syllable than in the connected speech task; this does not alter the fact the mask did not change breathing patterns for either the syllable or connected speech task. Studies by Higgins and Saxman (1993) and Sapienza and Stathopoulos (1995) have shown task effects for acoustic and aerodynamic measures. In both studies, some measures changed as a function of task whereas some did not. The present study along with previous studies indicate that task effects are quite variable. It is suggested that further research investigating task effects be pursued. The results of this study indicate that there is no effect on lung, rib cage, and abdominal volumes and frequency of breathing while wearing a circumferentially vented pneumotachograph face mask. It is concluded that researchers can use the circumferentially vented

pneumotachograph face mask during studies of respiratory kinematics and have confidence that their results are generalizable to nonmask conditions.

Acknowledgments This work was supported by National Institute on Deafness and Other Communication Disorders (NIDCD) grant R1DC02261A.

References Askanazi, J., Silverberg, P. A., Foster, R. J., Hyman, A. I., Milic-Emili, J., & Kinney, J. M. (1980). Effects of respiratory apparatus on breathing pattern. Journal of Applied Physiology: Respiration Environment Exercise Physiology, 48, 577–580. Bunn, J. C., & Mead, J. (1971). Control of ventilation during speech. Journal of Applied Physiology, 31, 870–872. Capra, N. F. (1995). Mechanisms of oral sensation. Dysphagia, 10, 235–247. Chadha, T. S., Watson, H., Birch, S., Jenouri, G. A., Schneider, A. W., Cohn, M. A., & Sacker, M. A. (1982). Validation of respiratory inductive plethysmography using different calibration procedures. American Review of Respiratory Disease, 125, 644–649. Comroe, J. H., Jr. (Ed.). (1950). Methods in Medical Research, 2. Chicago: The Year Book Publishers, Inc. Gilbert, R., Auchincloss, J. H. Jr., Brodsky, J., & Boden, W. (1972). Changes in tidal volume, frequency, and ventilation induced by their measurement. Journal of Applied Physiology, 33, 252–254. Higgins, M. B., & Saxman, J. H. (1993). Inverse filtered air flow and EGG measures for sustained vowels and syllables. Journal of Voice, 7, 47–53. Hirsch, J. A., & Bishop, B. (1982). Human breathing patterns on mouthpiece or face mask during air, CO2, or low O2. Journal of Applied Physiology: Respiration Environment Exercise Physiology, 53, 1281–1290. Hoit, J. D., & Hixon, T. J. (1987). Age and speech breathing. Journal of Speech and Hearing Research, 30, 351–366. Hoit, J., Hixon, T., Watson, P., & Morgan, W. (1990). Speech breathing in children and adolescents. Journal of Speech and Hearing Research, 33, 51–69. JMP 3.1.5 [Computer software]. (1996). Cary, NC: SAS Institute. Klatt, D. H., Stevens, K. N., & Mead, J. (1968). Studies of articulatory activity and airflow during speech. Annals of New York Academy of Sciences, 155, 42–54. Konno, K., & Mead, J. (1967). Measurement of the separate volume changes of rib cage and abdomen during breathing. Journal of Applied Physiology, 22, 407–422. Milenkovic, P. (1989). Cspeech (Version 3.0; Speech Analysis Software Program) [Computer Software]. Madison, WI: University of Wisconsin. Rothenberg, M. (1973). A new-inverse filtering technique for deriving the glottal airflow waveform during voicing. Journal of the Acoustical Society of America, 53, 1632–1645. Journal of Speech, Language, and Hearing Research

478 Russell, B. A., Stathopoulos, E. T., & Cerny, F. (1998). Effects of varied vocal intensity on ventilation and energy expenditure in women and men. Journal of Speech, Language, and Hearing Research, 41, 239–248. Sackner, J. D., Nixon, A. J., Davis, B., Atkins, N., & Sackner, M. A. (1980). Effects of breathing through external dead space on ventilation at rest and during exercise. II. American Review of Respiratory Disease, 122, 933–940. Sapienza, C. M., & Stathopoulos, E. T. (1995). Speech task effects on acoustic and aerodynamic measures of women with vocal nodules. Journal of Voice, 9, 413–418. Stathopoulos, E. T., & Sapienza, C. (1993). Respiratory and laryngeal measures of children during vocal intensity variation. Journal of the Acoustical Society of America, 94, 2531–2543. Stathopoulos, E. T., & Sapienza, C. M. (1997). Developmental changes in laryngeal and respiratory function with variations in sound pressure level. Journal of Speech, Language, and Hearing Research, 40, 595–614.

Journal of Speech, Language, and Hearing Research

JSLHR, Volume 41, 472–478, June 1998

Till, J. A., Jafari, M., Crumley, R. L., Law-Till, C. B. (1992). Effects of initial consonant, pneumotachograph mask, and oral pressure tube on vocal perturbation, harmonics-to-noise, and intensity measurements. Journal of Voice, 6, 217–223. Winkworth, A. L., & Davis, P. J. (1997). Speech breathing and the Lombard Effect. Journal of Speech, Language, and Hearing Research, 40, 159–169. Winkworth, A. L., Davis, P. J., Ellis, E., & Adams, R. D. (1994). Variability and consistency in speech breathing during reading: Lung volumes, speech intensity, and linguistic factors. Journal of Speech and Hearing Research, 37, 535–556. Received February 28, 1997 Accepted December 9, 1997 Contact author: Jessica Huber, SUNY at Buffalo, 109 Park Hall, North Campus, Buffalo, NY 14260. Email: [email protected]