ACOUSTIC AND RESPIRATORY MEASURES AS A FUNCTION OF AGE IN THE MALE VOICE

The purpose of this study was to extend understanding of the effects of aging on the male voice by obtaining and analyzing both acoustic and respiratory measures across the aging continuum. Aerodynamic measurements such as vital capacity (VC), maximum phonation time (MPT) and the acoustic measurement Speaking Fundamental Frequency (SFF) are used commonly in Speech-language Pathology to aid in the assessment and treatment outcomes of vocal dysfunction. However, current research lacks analysis of the interaction across these parameters within males and younger ages. This information may be important in understanding the normal changes in the speech mechanism with age and provide important direction for voice assessment and therapy outcomes. This study examined the changes of these parameters and interactions in males across various age groups. Acoustic measures of SFF, MPT, and VC were obtained in age groups of 20-29, 30-39, 40-49, 50-59, and 60-69, N=35. A statistically significant decrease in SFF with age was observed. No statistically significant interaction was observed between MPT and VC. Additionally, no statistically significant interaction was observed between MPT and age, or between VC and age.

changes, but in terms of changes in societal roles and capabilities (Glascock & Feinman, 1980). One of the major hallmarks of aging is changes in capabilities as evidenced by change in physical characteristics. Physical changes are the most predominant measure of aging. An outstanding physical characteristic of aging is in the area of vocal change.
After decades of relative vocal stability, noticeable changes in the voice occur as a function of the aging process. For example, as the body ages, there is loss in muscle mass, thinning of mucosal membranes, as well as in coordination. These changes not surprisingly, are reflected in laryngeal function that leads to changes in the voice.
Common age attributed characteristics of the elderly voice are hoarseness, breathiness, roughness, instability, reduced acoustic volume, changes in pitch and vocal tremor (Gorham-Rowan & Laures-Gore, 2006;Hartman, 1979;Verdonck-de Leeuw & Mahieu, 2004). The quality of voice resulting from air loss, laryngeal tension, tremor, and altered fundamental frequency associated with age may allow listeners to easily distinguish elderly voices from younger voices (Lundy, Silva, Casiano, Lu, & Xue, 1998). Elderly individuals experience voice disorders and dysphonia at a higher rate than younger individuals. The prevalence of voice disorders in the elderly has been estimated to be between 12% to 47% (Golub, Chen, Otto, Hapner, & Johns, 2006;Roy, Stemple, Merrill, & Thomas, 2007;Turley & Cohen, 2009). The most commonly reported age related voice complaint is reduction in vocal quality (Verdonck-de Leeuw & Mahieu, 2004).
Decreased voice quality secondary to aging has been shown to negatively impact quality of life. Many elderly individuals report an inability to speak in a noisy situation, insufficient air, reduced ability to practice one's profession, and social isolation ( Anatomical and physiological changes in the speech mechanism due to the aging process have been reported by several researchers. Some established age related changes in the speech mechanism are due to structural changes in the thoracic skeleton and chest cavity, decreased lung capacity, poor laryngeal valving, ossification and calcification of laryngeal cartilages, atrophy of laryngeal muscles, changes in blood supply, and significant changes in the vocal folds (Kahane, 1990;S. Xue & Hao, 2003). These collective changes are often referred to as presbylaryngis. Age related changes in both the structure and physiology of the speech mechanism are believed to impact voice production and acoustic qualities. These changes are not uniform in men and women.
Some degenerative changes occur earlier in the life and to a greater extent in men than in women (Kahane, 1990). Therefore, one would expect greater changes in acoustic and respiratory measures to begin earlier and occur faster in the lifespan of males.
Additionally, there are different structural and physiological changes in the speech mechanism in men and women (Kahane, 1981). However, structural changes in the speech mechanism may not clearly produce functional changes in speech production. It has been suggested that men and women may differentially adjust their speech to accommodate these changes (Linville & Rens, 2002). The nature and extent of these anatomical, physical, and acoustic changes are still being investigated.

Maximum Phonation Time
There are many acoustic and respiratory measures available to researchers to ascertain information about the speech and respiratory mechanisms. Some of these measures are MPT, VC, and SFF. MPT is an accepted standard clinical task in speechlanguage pathology for the assessment of respiratory and phonatory function. (Kent, Kent, & Rosenbek, 1987;Pearl Solomon, Garlitz, & Milbrath, 2000).
Maximum phonation time is defined as the longest period during which an individual can sustain phonation of a vowel sound, typically /a/. Usually a timer and audio recorder, with or without audio analysis, are the only instruments needed to measure MPT. MPT is used as a quick, noninvasive, low-cost diagnostic tool to assess vocal function. It measures laryngeal function in different pathological circumstances such as dysphonia and Parkinson's disease. It is also used to measure improvement after voice therapy (Maslan, Leng, Rees, Blalock, & Butler, 2011;Speyer, 2008).
Several researchers have reported norms for MPT, however, there have been inconstancies among these findings Maslan et al. (2011) MPTs seem to be longer in males than females presumably due to an average larger VC (Kent et al., 1987;Maslan et al., 2011). Maslan et al. (2011) found that MPTs were longer for individuals over 65 than previously reported; however, times were shorter among younger individuals. Still, it is not clear to what extent MPT is influenced by age. Kent, Kent, & Rosenbek (1987) sought to gain data concerning several common clinical tests, one of which was such as maximum phonation time. Kent, Kent, & Rosenbek (1987) noted that a reduced MPT may be attributable to an inadequate volume of air used during phonation or to excessive wasting of air during phonation as a result of poor laryngeal valving. Kent et al. (1987) concluded that MPT alone is not a useful determinate of respiratory inefficiency.
Pearl Solomon, Garlitz, & Milbrath (2000) reported a correlation between MPT duration and the presence of organic and functional voice disorders. They also found an inverse relationship between MPT and the severity of a voice disorder. MPT durations have been used in pre-post therapeutic measurements to assess treatment outcomes (Stemple, Weinrich, & Brehm, 2008). Selected studies reporting MPT for typical males are presented in Table I.  Kent (1987) Older adults (1) 65-75 14.6 5.9 Ptacek & Sanders (1966) Older adults (2)

Vital Capacity
In order to produce sound, air must be expelled from the lungs into the vocal tract.
Vital capacity (VC) is defined as the greatest volume of air that can be expelled from the lungs after taking the deepest possible breath. VC is related to the quantity of air available for phonation. Age related reduction in the vital capacity directly reduces the amount of air available to be expelled for phonation. This reduction in air available for phonation contributes to various age related changes in speech production (Kahane, 1990).
Total lung Capacity (TLC) is the total volume of air in the lungs after a maximal inhalation. Tidal Volume (TV) is the amount of air that is inspired and expired from the lungs during a cycle of quiet respiration. Inspiratory reserve volume (IRV) is the volume that can be inhaled after a tidal inspiration, while, Expiratory Reserve Volume (ERV) is the amount of air that can be expired following a tidal expiration. Even after a maximal exhalation there is air left within the lungs, this volume is referred to as Residual Volume (RV). Total lung capacity (TLC) and residual volume remain relatively the same across the adult lifespan, while vital capacity (VC), inspiratory capacity (IC), and expiratory reserve volume (ERV) diminish with age (Hoit & Hixon, 1987). Rochet (1991) reported that changes in pulmonary function due to aging become measurable at around age 40.
While there are declines in the respiratory system of both men and women, these changes are greater in women than men (Gorham-Rowan & Laures-Gore, 2006).
VC is known to be influenced by an individual's age, sex, and height (Kent, Kent, & Rosenbek, 1987). There are numerous studies which provide normative data for both men and women (Kent, Kent, & Rosenbek, 1987;Yiu, Yuen, Whitehill, & Winkworth, 2004;Zraick, Smith-Olinde, & Shotts, 2012). One recent study utilizing the Phonatory Aerodynamic System (PAS) found that the mean expiratory volume for males was 4.14 for ages 18-38; 4.19 for ages 40-59, and 3.09 for ages 60-89 ( Zraick, Smith-Olinde, Shotts, 2012). The results of this study are presented in Table II. A speaker's fundamental frequency is not constant; rather, there is variability of the frequencies produced. This variability of frequency is measured in one of two ways, standard deviation of F o or in semitones called pitch sigma. (Baken & Orlikoff, 2000) define pitch sigma is a "measure of the average distance of values from the mean." This measure is the standard deviation (SD) of the frequencies included in a speech sample, which is the "square root of the sum of the squares of the deviations from the mean" (Baken & Orlikoff, 2000). The average standard deviation of fundamental frequency For women, F 0 continues to decrease with age, or it stays constant until menopause after which time it decreases anywhere from 10 to 35 Hz (Hollien & Shipp, 1972;McGlone & Hollien, 1963;Sataloff, Rosen, Hawkshaw, & Spiegel, 1997).
Men tend to exhibit an increase in F 0 due to vocal fold atrophy, in contrast, females tend to exhibit a decrease in F 0 post-menopause as a result of reductions in vocal fold mass.
There have been several investigations into changes of fundamental frequency across age groups that show this trend (Awan, 2006;Linville, 1987;Ramig & Ringel, 1983

Research Questions and Hypothesis
The purpose of this investigation is to provide preliminary data that would address the following questions: 1) Are MPT and VC related in males?
2) Does MPT decline across age groups in males?
3) Does SFF change across age groups in males?
It is hypothesized that: 1) MPT and VC will be related in males 2) MPT and VC will decline across age groups in males 3) SFF will increase across age groups of males Concerning the first question, it was hypothesized that MPT and VC will be related. Kent, Kent, & Rosenbek (1987) concluded that since MPT requires the voluntary expulsion of air, a reduction in available air from a reduced VC would naturally reduce the duration of MPT. Conversely, an increase in VC would provide more air for phonation and lengthen the duration of MPT. To the author's knowledge, Awan's 2006 study was the only study that directly compared VC and MPT; however, Awan (2006) only used female participants. Since there are established differences in vital capacities and respiratory aging patterns in females, it is necessary to test this correlation in males.
Addressing the second question, it is hypothesized that both MPT and VC will reduce as a function of age. VC is known to decrease with age in both males and females (Spector, 1956, p. 267). However, there is less agreement about the nature of changes in MPT with age. Kent, Kent, & Rosenbek (1987)  The third hypothesis is that SFF will increase with age in males. This is consistent with previous reports (Higgins & Saxman, 1991;Hollien & Shipp, 1972;McGlone & Hollien, 1963;Mysak, 1959;Sataloff, Rosen, Hawkshaw, & Spiege, 1997).

METHODOLOGY
The investigation, materials, and procedures were approved by the Institutional University and members from the greater community. A total of thirty five individuals participated in this study. There were no financial incentives provided for participation.

Consent Form
The investigator recruited, screened and collected data for all participants. All data were collected in the voice laboratory of the Speech and Hearing Clinic at Cleveland State University. All participants agreed and signed the consent form after discussing with the examiner. The participants were also offered a reference copy.

Screening
Based on self-report, prospective participants were screened for laryngeal pathologies and other health conditions that could affect the voice. Exclusionary conditions that participants were asked to self-report were asthma, sinus problems, acid reflux, use antihistamines, vocal fold pathology, emphysema or neuromotor impairment that may impact the voice. Prospective participants were also asked to self-report if they currently had a respiratory infection. They were also asked if they have used tobacco consistently in the past five years. The questionnaire used to screen participants can be found in Appendix B.

Data Collection
Two instruments were used to record the acoustic and respiratory measures. The Each participant received a disposable plastic mouthpiece which was discarded after use.
Each data elicitation task was preceded by the investigator reading an explanatory script explaining what the participant needed to do in order to perform a task.

Vital Capacity
To obtain vital capacity measures participants were asked to breathe in as deeply as possible and exhale maximally into a handheld spirometer in order to measure vital capacity. The investigator offered an example of a maximal inhalation and exhalation.
The highest of two trials was taken as a measure of vital capacity. A copy of this script can be found in Appendix C.

Maximum Phonation Time Procedure
Participants were asked to take the largest breath possible and sustain the vowel /a/ for the longest possible time. The experimenter provided a verbal description of the maximum phonation time task using a script, then offered a demonstration. A copy of this script can be found in Appendix C. The longest of two trials was used. This measure was calculated using the Visi-Pitch-IV Maximum Phonation Time protocol.

Average Fundamental Frequency
Participants spoke into a hand-held microphone placed four to six inches away from their mouths. Participants were asked to read an excerpt from the "Rainbow Passage" at a comfortable volume (Fairbanks, 1969)

Participant Data
Descriptive statistics for the participants and all measures were calculated. Raw data for all participants can be found in Appendix F. The number of participants in each age group, mean age within group, and standard deviation are presented in Table III. Group means and standard deviations for SFF, VC, and MPT are presented in Table IV of SFF, VC, and MPT. The highest SFF was within the 30-39 year old group.
The highest mean MPT and VC were in the 40-49 year old group.

Analysis of Acoustic and Respiratory Measures
A series of Pearson product-moment correlations between all variables: age, MPT, VC, and SFF were calculated. Additionally, scatter plot diagrams are provided with the measures that were compared using the Pearson product-moment with the line of best fit. SigmaPlot 11.0 was used to calculate the results and scatter plot figures.
correlation between age and MPT. A Pearson product-moment correlation was calculated to determine the relationship between age and MPT. There was no significant correlation between age and MPT (r=-1.46, p>0.05). A scatterplot representing the relationship between age and MPT is presented in Figure 2. Age vs MPT Regression correlation between age and VC. A Pearson product-moment correlation was calculated to determine the relationship between age and Vital Capacity. There was no significant correlation between age and MPT (r=-3.08, p>0.05). A scatterplot representing the relationship between age and Vital Capacity is presented in Figure 3. Age vs VC Regression correlation between age and SFF. A Pearson product-moment correlation was calculated to determine the relationship between age and Speaking Fundamental Frequency. There was a significant correlation between age and SFF (r=-3.06, p<0.05). A scatterplot representing the relationship between age and SFF is presented in Figure 4. Age vs SFF Regression correlation between VC and MPT. A Pearson product-moment correlation was calculated to determine the relationship between VC and MPT. There was no significant correlation between VC and MPT (r=.323, p>0.05). A scatterplot representing the relationship between VC and MPT is presented in Figure 5. correlation between MPT and SFF. A Pearson product-moment correlation was calculated to determine the relationship between MPT and SFF. There was no significant correlation between MPT and SFF (r=.089, p>0.05). A scatterplot representing the relationship between MPT and SFF is presented in Figure 6. The data from all age groups were analyzed with respect to whether there was a correlation between VC and SFF. A Pearson product-moment correlation was calculated to determine the relationship between VC and SFF There was no significant correlation between VC and SFF (r=.185, p>0.05). A scatterplot representing the relationship between VC and SFF is presented in Figure 7. ANOVA revealed no statistical significance between MPT and VC (F=1.202,df=27,p=.43,p<0.05). The results of the ANOVA are presented in table VI.

Research Questions
The research questions proposed in this study are as follows: 1) MPT and VC will be related in males 2) MPT and VC will decline across age groups in males 3) SFF will increase across age groups of males Regarding the first question, visual inspection of the scatterplot in Figure 5 shows that an increased VC is consistent with an increased MPT. However, the Pearson product-moment found no statistical correlation between MPT and VC,(r=.323,p>.05).
Additionally the ANOVA that was calculated comparing MPT and VC showed no correlation,(F=1.202,df=27,p=.43,p<.05). The lack of correlation could be attributable to several factors. Firstly, this study was comprised of only thirty five participants. The scatterplot shown in Figure 5 visually suggests a correlation but statistical tests failed to show a correlation. Perhaps with more participants a statistical significance would have been reached. Alternatively, participants with lower vital capacities may be able to effectively compensate while producing MPTs.
Regarding the second hypothesis, MPT and age were not found to be statistically significant, (r=-1.46, p>.05). VC and age were also not found to be statistically significant (r=-3.08, p>.05). However, a visual inspection of Figure 2 comparing MPT with age and Figure 3 comparing VC with age visually show a negative correlation of both measures with age. The failure to reach statistical significance could be attributable to a small number of participants. Additionally, chronological age may not be a strong predictor of these measures; rather, other factors including physiological age, height, weight, and activity levels may be better predictive factors.
Regarding the third hypothesis, SFF will increase with age; the results of this study revealed a statistically significant negative correlation of age and SFF in males by calculating a Pearson product-moment (r=-3.06, p<.05). Figure 4 shows a decrease in SFF with age. These results were inconsistent with previous reports of fundamental frequency increasing with age in males. It has been suggested that the possibility for elderly men to attempt to compensate for high-pitched unstable voice and for elderly woman to attempt to avoid a deep voice (Pontes, Brasolotto, & Behlau, 2005). Perhaps, men compensated for a natural increase in fundamental frequency while reading the passage.

Limitations of the Study
There were several limitations to this study and the results represent preliminary data of a pilot study. The most significant limitation was the number of participants,

N=35. The number of participants in groups
This study was also limited by type of instrumentation used. Despite being a medical grade hand-held spirometer it was only accurate to .05 liters. Additionally, spirometer was analog, therefore requiring the experimenter to read the volume.
There were several known influencers of VC, MPT, and SFF that were not included in the study. There were no controls for the height or weight of participants which are both known influencers of both VC and F 0 (Boone, McFarlane, Von Berg, & Zraick, 2013). Additionally, the study did not control for the activity levels of participants. The activity level of a participant would be assumed to directly affect VC.
Race and ethnic background were not considered in the participation criteria or organization of the data and groups. Race and voice interactions have been suggested by Boone et al. (2011) and Richard (2013).

Overview
The purpose of this study was to examine the effects of age on acoustic and respiratory measures and the interaction between VC and MPT. Data analysis demonstrated that there was no statistically significant interaction between MPT and VC, and similarly no interaction between MPT and VC with age. There was however, a statistically significant interaction between age and speaking fundamental frequency.
There was a decrease in SFF as a function of age. The knowledge gained from this study is relevant for diagnosis and assessing treatment outcomes using acoustic and aerodynamic measurements.

Clinical Implications
VC and MPT are measured routinely used by speech-language pathologists during informal assessments and are incorporated within assessment protocols such as the Dysphonia Severity Index (Wuyts, Bodt, Molenberghs, Remacle, Heylen, Millet, Lierde, Raes, & Heyning, 2000). These measures are used by clinicians to make judgments about breath support for speech production. Breathe support has been defined as the reservoir of available air for speech production along with the efficiency of valving and air control at the level of the vocal folds. VC is a measure of the total amount of air available for phonation, while MPT provides a measure regarding an individual's functional ability to use available air. A large VC paired with a low MPT could indicate that an individual is unable to functionally use all of the air available during a sustained vowel task.
Conversely, a low VC paired with a high MPT could indicate a high level of efficiency at the glottal level. This study did not yield a statistical correlation between these two measures, nevertheless, a reduced VC or MPT both reflect an underlying deficit of the speech and respiratory system. Given the results of this study, a clinician cannot measure either variable and reliably assume about the other. It must be noted that any correlations or lack of correlation between MPT and VC in healthy speakers may not be true of those with vocal disorders or other disease processes. When analyzing functional use of voice, MPT is a test of maximum performance; therefore, it might not accurately reflect an individual's ability to use their voice during everyday speech tasks.
Interest in VC and MPT extends beyond the identification and treatment of vocal pathologies. VC and MPT are of clinical interest for individuals experiencing difficulty with adequate respiratory volume or ability to produce speech for sustained durations during activities such as lecturing, singing, and continued talking. Improvement of VC may result in an individual possessing a larger volume of air to use for phonation. An increase in VC could result in an increased ability to functionally use one's voice for longer durations and at increased volume. This study's findings that VC and MPT are not statistically correlated suggest that an individual interested in improving the length of ability to speak should not be primarily concerned with increasing VC.
The lack of correlation between both MPT and VC with age is important for clinicians making judgments about an individual's speech and respiratory system. The clinician should consider factors such as physiological age, height, weight, and activity levels in addition to chronological age.
Previous research has shown an increase in fundamental frequency with age in males; however this study showed a decrease in speaking fundamental frequency with age. Fundamental frequency is often obtained using a single vowel sound while SFF is obtaining using connected speech. Co-articulatory factors across a speech sample could influence the average fundamental frequency resulting in a lower frequency than when measured in isolated vowel sounds. The result of this study show an age related reduction in SFF similar to that which was displayed in females in Awan (2006). This demonstrates that a reduction in SFF within males is typical with age.
It has also been suggested that perhaps elderly men to attempt to compensate for high-pitched unstable voice (Pontes, Brasolotto, & Behlau, 2005). If indeed men's fundamental frequency increases with age and the results of this study reflect a compensatory behavior in men. If this is so, clinicians need to consider the results of data collected on measures like SFF considering the possibility of compensatory behaviors, not just physiologic changes.

FUTURE RESEARCH
Despite the clinical reliance on MPT, VC, and SFF during voice evaluations these measures have not been used extensively in the same research design and directly compared to each other. A future study could examine the interaction of these measures in both males and females in the same research design. A future study could group participants based on height, weight, physiologic age, and activity levels. Other parameters of respiratory function could also be included such as phonation quotient and peak expiratory flow. Future studies could incorporate newer digital instruments such as the Phonatory Aerodynamic System and the different respiratory parameters it provides.