We are thrilled to share the news that the manuscript “Harmonic Amplitude Differences Before and After Voice Treatment for Parkinson’s Disease and Their Relationship to Voice Quality and Speech Intelligibility” is now ‘in press’ in the Journal of Voice! The outcomes of this study uniquely document a pre post treatment effect and an impact on intelligibility. I hope you enjoy this important paper! – Lori Ramig
We asked one of the authors, Dr. Michael Cannito, to summarize the article by sharing his perspectives. Plus, all of the authors share their thoughts through indvidual quotes.
Why study Harmonic Amplitude Differences?
These data can provide insight into vocal fold closure, voice qualities (breathiness or strain). They also can reflect increasing vocal effort. For example, a large difference between the first and second harmonic indicates breathy voice quality, while a large negative difference (second harmonic greater than the first harmonic) indicates strained voice quality. It also correlates with increased vocal effort. Therefore, as voice quality improves after voice treatment, we should also see changes in the harmonic amplitude differences.
Michael Cannito, PhD, CCC-SLP
Both as an SLP trained in LSVT, and as an applied speech scientist, I am so very excited about these data. They demonstrate that the treatment improves not only overall loudness but also has a dramatic affect on the harmonic structure of the voice. This explains post treatment improvement in intelligibility often observed by clinicians and reported in the research literature.
How do you define Harmonic Amplitude Differences?
The human voice is comprised of many frequencies. Fundamental frequency (F0) is measured by the number of vocal fold vibrations per second and is reflected in the pitch of the voice measured in Hertz (cycles per second). This can be seen on the left side of the Figure of Spectra. If the vocal folds vibrate faster, there are more cycles per second and the pitch goes up. If the vocal folds vibrate slower, there are fewer cycles per second and the pitch goes down. However, the F0 is not the only frequency comprising the voice. Importantly, the voice also contains a sequence of harmonics or “overtones” that give the voice good healthy voice its rich and melodious sound. The fundamental frequency is considered the first harmonic (H1). If it was 100Hz, the second harmonic would be 200Hz, the third harmonic, 300Hz and the fourth harmonic 400Hz and so on. When we subtract the intensity of a higher harmonics from the first harmonic, the resulting value is the Harmonic Amplitude Difference.
Angela Halpern, MS, CCC-SLP
This study highlights the clinical impact of LSVT LOUD treatment by bridging research outcomes of acoustic change in harmonic amplitude differences to perceptual and intelligibility improvements in speech for people with Parkinson’s disease.
What did prior research show about Harmonic Amplitude Differences?
Researchers have reported relationships between specific harmonics and vocal fold activity. For example, Hanson et al. (2001) reported that the difference between the first harmonic and the second harmonic indicates the degree of vocal fold closure and the time the vocal folds are open during each cycle of voicing. Large differences between harmonic one and harmonic two indicates greater vocal fold opening and correlates with perceived breathiness, while small or negative differences (harmonic two greater than harmonic one) indicates greater vocal fold closure and correlates with perceived vocal strain. We also know that as overall harmonic intensity increases near the fundamental frequency (harmonics 1, 2, and 3) after LSVT, (Alharbi, et al., 2019).
Jennifer Spielman, MA, CCC-SLP
I really enjoyed the collaborative work on this important article. I learned a great deal about Harmonic Amplitude Differences and their value in documenting pre post LSVT LOUD changes as well as the impact on intelligibility.
What did you study and why?
In order determine how LSVT affects Harmonic Amplitude Differences in Parkinson’s Disease (PD), our research consisted of two studies. In Study 1 we measured harmonic 1 minus harmonic 2 as well as the difference between harmonics at six-month follow up. We used speech recordings from a previously published clinical trial (Halpern et al, 2012). We found that both harmonic 1 and 2 and decreased, indicating greater degree of glottal closure and shorter open time during each glottal cycle. In addition, greater harmonic intensity in the higher frequency range was associated with the vowel formants as represented by formant 3. These findings are consistent with earlier studies demonstrating increased vocal fold closure and reduced breathiness post LSVT.
Our second study utilized speech recordings from another prior study in which a carefully controlled listening experiment had been conducted to examine intelligibility of sentences produced by individuals with PD before and after LSVT (Cannito et al, 2012). That study demonstrated improved intelligibility of PD speech, presented to normal hearing listeners in the sound field, in the presence of noise, following LSVT. Importantly, the pre-and-post LSVT sentences were presented to listeners at the same loudness level to rule out overall loudness as a cause for intelligibility change. We measured the differences between harmonic 1 and the harmonic at the third formant, as well as differences between harmonic 1 and the harmonics at the first and second formants, all of which are important for speech intelligibility. We also examined the relationship between the Sentence Intelligibility Test scores and the harmonic amplitude differences.
Lorraine Ramig, PhD, CCC-SLP
I especially appreciated the long-term collaborative relationship between the Colorado and Louisiana teams in this study. The blending of the science and clinical expertise was excellent. I personally learned much about harmonic amplitude differences (HADs) and the value they bring to documenting both pre post LSVT changes and improved intelligibility. The conclusion that ‘increased glottal closure following LSVT strengthened harmonic energy at the second and third formants, which are critical for speech perception, made them more perceptible for listeners, thereby improving intelligibility.’ This is a significant conclusion.
What did you find in the new study?
In Study 1 we found that after LSVT, harmonic amplitude differences between the first and second harmonic decreased toward a normal level which correlated with improved voice quality. We also found the difference between the first harmonic and the harmonic at the third formant decreased toward a normal level due to the strengthening to the intensity of vocal tract resonance. These changes were maintained at six-month follow-up. This reinforced our view that changes intelligibility should be related to change in harmonic amplitude differences.
In Study 2, we found that the differences between the first and second harmonics and between the first harmonic and the harmonics associated with the first, second and third formants all decreased significantly from pre-to-post LSVT. This indicated greater harmonic energy occurring across the frequency range after treatment. Importantly, we also found that the changes in difference between the first harmonic and the harmonic at the second formant as well as the difference between the first harmonic and the harmonic located at the third formant were strongly correlated with the pre-to-post LSVT improvement in listeners’ sentence intelligibility from the original study. We interpreted this to indicate that increased glottal closure, which is known to occur following LSVT, strengthened harmonic energy generally and the strengthening of harmonic energy at the second and third formants. These elements are critical for speech perception and made them more perceptible for listeners, thereby improving intelligibility.
Qiang Li, PhD, CCC-SLP
Acoustic spectra of vowels is an important perspective to assess voice and speech changes. The relative relationship between harmonics and formants as revealed in the current study demonstrates its potential clinical applications to indicate changes in speech intelligibility.
References
Alharbi GG, Cannito MP, Buder EH, Awan SN. Spectral/cepstral analyses of phonation in Parkinson’s disease before and after voice treatment: A preliminary study. Folia Phoniatrica et Logopaedica. 2019;71(5-6):275-85.
Cannito MP, Suiter DM, Beverly D, Chorna L, Wolf T, Pfeiffer RM. Sentence intelligibility before and after voice treatment in speakers with idiopathic Parkinson’s disease. Journal of Voice. 2012;26:214-219.
Halpern AE, Ramig LO, Matos CEC, et al. Innovative technology for the assisted delivery of intensive voice treatment (LSVT®LOUD) for Parkinson disease. Am. J. Speech-Lang. Pathol. 2012;21:354-367.
Hanson HM, Stevens KN, Hong-Kwang JK, Chen MY, Slifka J. Towards models of phonation. Journal of Phonetics.2001;29:451-480.
