This Article Abstract Full Text (PDF) Submit a related Letter to the Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Moore, C. Articles by Baumann, G. Search for Related Content PubMed PubMed Citation Articles by Moore, C. Articles by Baumann, G. Related Collections Neuroendocrinology and Pituitary

Voice Abnormality in Adults with Congenital and Adult-Acquired Growth Hormone Deficiency

North Western Medical Physics (C.M., K.M.) and Department of Endocrinology (S.S.), Christie Hospital, M20 4BX Manchester, United Kingdom; North Western Medical Physics (T.W.), Withington Hospital, M20 2LR Manchester, United Kingdom; and Division of Endocrinology, Metabolism, and Molecular Medicine (H.M., G.B.), Northwestern University; Feinberg School of Medicine, Chicago, Illinois 60611

Address all correspondence and requests for reprints to: Gerhard Baumann, M.D., Northwestern University Feinberg School of Medicine, 303 East Chicago Avenue, Chicago, Illinois 60611. E-mail: gbaumann{at}northwestern.edu; or Professor C. J. Moore, North Western Medical Physics, Christie Hospital National Health Service Trust, Wilmslow Road, Withington, Manchester M20 4BX, United Kingdom. E-mail: chris.moore{at}physics.cr.man.ac.uk.

	Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Context: Adult males with congenital, untreated, severe GH deficiency (GHD) due to genetic GHRH receptor deficiency exhibit distinctive, high-pitched, and raspy voice characteristics.

Objective/Design: To determine the physical underpinning of this phenomenon, we performed voice recordings, translarynx impedance measurements, spectral analysis, and estimates of spectral complexity [approximate entropy (ApEn)] in four affected men. Results were compared with those obtained in four men with untreated adult-onset GHD and a normal male population.

Results: Congenital GHD subjects had a high-pitched voice with a fundamental frequency typical of normal females (174–266 Hz). Their frequency spectra were characterized by abnormal harmonics, with reversal/interruption of the normal amplitude decay among higher-order harmonics, findings consistent with a creaky quality of the voice. Patients with adult-onset GHD, acquired at ages 31, 38, and 40 yr, had a normal male pitch (fundamental frequency, 117–154 Hz) but pathologically low ApEn values, corresponding to a breathy quality of the voice and suggesting abnormal vocal fold function. A fourth patient who acquired GHD at age 22 yr had a pitch intermediate between male and female, high ApEn, and a spectral pattern similar to the congenital GHD patients.

Conclusions: This study demonstrates an effect of GH on laryngeal size and vocal fold compliance that results in a high pitch and disordered spectral quality. The time of onset of GHD determines which type of abnormality predominates.

	Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
VOICE ANOMALIES ARE well recognized as part of the clinical spectrum in GH disorders, such as GH deficiency (GHD) and acromegaly. Thus far, this impression has been based primarily on subjective observations. During recent studies of the phenotype of congenital, severe, untreated GHD in patients with an inactivating mutation in the GHRH receptor gene (1, 2), we noted that the voices of adult males were high pitched and raspy, with a screeching quality (1). This voice abnormality is most striking, second in prominence only to extreme short stature. To evaluate the physical basis of this phenomenon, we conducted objective voice studies in four affected young adult males during a visit from their native Pakistan to Manchester, United Kingdom (2). For comparison, we similarly studied four men who acquired GHD in adult life. Both groups were compared with normal adults using a single-parameter measure of voice normality, which has recently been developed and validated (3, 4, 5).

Theory in brief

The vibration of the vocal folds can be measured by exploiting the highly correlated translarynx electrical impedance variation (6, 7). A time series of such impedance measurements is termed the electroglottogram (EGG), which can be objectively quantified using spectral analysis (5, 8) and approximate entropy (ApEn), a measure of spectral complexity (9, 10).

The fundamental frequency, f₀, and the harmonics, defined as exact multiples of f₀, correspond to the subjective perceptions of pitch and timbre, respectively. They are augmented by the potential existence of a continuum of frequencies between these distinct harmonic frequencies. These extra frequencies are said to be anharmonic; subjectively they contribute to the perception of "noise".

Typically, a spectral power series is arranged in order of ascending frequency to form an orchestrated power spectrum, with the distribution of power among the spectral components termed power spectral density (PSD). To facilitate interpretation, the raw PSD is refined by fundamental harmonic normalization (FHN), which normalizes the EGG spectrum relative to the power and frequency of f₀ (8). In acoustic terms, amplitude is related to power, f₀ to pitch, harmonic frequencies to timbre, and ApEn to voice clarity.

	Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Voice recordings were obtained in four adult males with genetic GHRH receptor deficiency and congenital, severe, untreated GHD due to a homozygous nonsense mutation (E72X) in the GHRH receptor gene (age range, 23–30 yr; Fig. 1; patients 5, 35, 23, and 4 in Ref. 1). These patients have isolated GHD; the other pituitary hormonal axes are intact (1). Key elements of the phenotype are short stature, somatotrope/pituitary hypoplasia, and relative microcephaly (1, 11).

View larger version (30K): [in this window] [in a new window]   FIG. 1. Voice spectral characteristics in two men representative of normal group 1 and group 2 (top two rows) and congenital GHD patients shown in row order, left to right: vowel "I" frame averaged PSD, FHN of the PSD, and truncated log10FHN. Note that all spectra, especially patient Rs, exhibit nonexponential falls in their spectral envelope because of the multiple occurrence of unexpectedly strong harmonics. The arrow denotes the weak fundamental frequency f0 in patient Rs (see text).

View larger version (34K): [in this window] [in a new window]   FIG. 2. Voice spectral characteristics in adult-onset, acquired GHD patients, shown analogous to Fig. 1. Only patient R shows well-maintained, distinct harmonic peaks extending to the higher harmonics in the FHN spectrum. The effect is clearly seen in the truncated log10FHN representation, which is flat only for patient R.

Using external throat sensors, translarynx EGG signals for the vowel "I" were sampled at 20 kHz for up to 4 sec and analyzed using spectral analysis and ApEn as described in detail previously (4).

All computations were performed using software written in Research Systems International IDL version 5.5. Statistical analysis was performed using t test and for the normal population, a Gaussian mixtures model fitting based on maximum likelihood. Results are expressed as mean ± SD, and P <0.05 was considered significant.

	Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Normal population

The salient aspects of the EGG voice spectra in the normal population have been previously reported (4). Two groups with distinct voice characteristics were identified: the larger (n = 55) group, group 1, had strong spectral features extending across the harmonic range with slowly decaying harmonics amplitudes, subjectively perceived as a clear voice. The smaller (n = 30) group, group 2, had weak spectral features that decayed rapidly toward the higher harmonics, which was consistent with a perceived breathy voice. Group 1 had high complexity with an ApEn of 0.338 ± 0.036 (mean ± SD), and group 2 had significantly lower complexity with an ApEn of 0.175 ± 0.049 (P < 0.001). In contrast, pitch analysis showed no difference between the two groups, with both exhibiting typical male f₀ distributions of 124 ± 29 Hz.

Congenital GHD subjects

The spectra for the congenital GHD patients (and for two representative normal subjects) are shown in Fig. 1. These patients differ from normal men in two important respects. First, the f₀ values are 174–266 Hz, well above the distribution for normal males (P < 0.01) and within the range seen in adult females (8, 12). Second, the progressive decay of the spectral envelope is interrupted and can undergo multiple reversals rather than the single reversal that is characteristic of the bright normal voice seen in group 1. ApEn values differ significantly from group 1 (P 0.025) but not from group 2 normals.

Patient Rs deserves particular mention. His spectrum is apparently dominated by a 460-Hz peak and another prominent peak at 690 Hz. This frequency ratio indicates that the true f₀ is 230 Hz, corresponding to a very small, broad peak that has less than one tenth the power of the 460-Hz harmonic. Effectively, patient Rs exhibits an extremely large, dual-peak reversal of spectral envelope decay, beyond which the spectrum is essentially featureless. A pathologically low complexity value of 0.08 is the result. This is consistent with the notably creaky characteristic of his voice.

Adult-onset, acquired GHD patients

Figure 2 shows the spectra for the four patients with adult-onset GHD. Patient R has an f₀ of 172 Hz, intermediate between normal male and female f₀. Based on a survey of normal f₀ values (12), an f₀ of 160 Hz has been suggested as a perceptual boundary for transsexuals, below which subjects are considered to be male and above which they are considered female (13). Patient R exhibits a strong spectral pattern, which is well maintained to high harmonic levels and similar to spectra of congenital GHD patients Ra, La, and Gh. The spectral envelope is erratic with amplitude decay reversed twice. The creaky quality of the voice in this patient is pronounced.

In contrast, the remaining three patients with adult-onset GHD have f₀ values (117–154 Hz) within the normal male range. They exhibit the rapidly decaying spectral envelope characteristic of group 2 normal men. Progressing from patient H to W to S, there is obliteration of spectral features by noise, as evidenced by the upward shift in base line power and the filling between low power harmonics. Subjectively, vowel phonation in patients H and W has a mild breathy quality. The corresponding ApEn levels are low and comparable with group 2 normal males. For patient S, the breathy quality of phonation is pronounced and clearly perceived as abnormal. This impression is consistent with the paucity of spectral features and a pathologically low ApEn value of 0.09. The ApEn values for the whole group differ from those in normal group 1 (P < 0.01) but not group 2.

	Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

 
The present study identifies the fundamental physics underlying the peculiar voice of our male patients with extreme short stature due to congenital, severe, untreated GHD. The principal findings are a high f₀ and an abnormally interrupted/partially reversed harmonic spectrum. f₀ values this high are typical of the normal female population; the distorted harmonics account for the creaky or raspy voice. Based on ApEn (complexity) data, three of four congenital GHD patients exhibit good vocal fold functionality. Patient Rs is the exception, with a pathologically low ApEn that is consistent with an observed difficulty in attempting to phonate the vowel "I".

The patients with adult-onset GHD acquired in midadulthood (patients H, W, and S), whose larynx development was complete before the onset of GHD, had a normal male f₀ but exhibited low ApEn, indicating disordered vocal fold functionality. Similar features are also seen in larynx cancer patients; this suggests that there may be a pathological basis for the findings in adult GHD patients.

The fourth adult-onset GHD patient (patient R) combines features of both congenital and adult-acquired GHD. His onset of GHD was in relative youth (21–22 yr), occurring after achievement of final height but before full adult body mass accretion. Voice spectral structure in patient R is comparable with that for the patients with congenital GHD, with a dual reversal of the spectral envelope decay. He has an intermediate f₀ typically associated with transsexuals and a high ApEn similar to group 1 normal males. Therefore, patient R may represent an intermediate between the congenital GHD patients and the patients whose larynx was fully developed before the onset of GHD.

The strikingly abnormal voice in our patients with congenital GHD partially results from the small size of their larynx (it should be noted that they were fully masculinized and had completed normal, if somewhat delayed, puberty). In particular, pitch (f₀) is known to be correlated with laryngeal size and vocal chord length (14). GHD also appears to play a role in larynx function independent of size, as evidenced by the abnormal voice recordings in our patients with adult-onset GHD. In addition to laryngeal factors, small body size and cranial anomalies probably contribute to distorted tonal characteristics because of altered resonance within underdeveloped nasopharyngeal cavities and perhaps subnormally pneumatized craniofacial bones. Based on our findings, it appears that laryngeal size determines the f₀ but that the distorted harmonics are a function of other aspects of GHD.

A high-pitched voice has long been recognized as a feature of GHD, with improvement of the abnormality after GH treatment (15). Similarly, genetic GH insensitivity is typically associated with a high-pitched voice in both genders (16, 17). Generally there is no mention of a creaky quality of the voice in the literature; this aspect may be particularly evident in our congenital GHD patients because they are adults with longstanding, untreated GHD. To our knowledge, no comprehensive EGG studies have been published in GHD or GH insensitivity patients.

In conclusion, spectral analysis of the distinctive voice distortion in adult men with congenital, untreated GHD (genetic GHRH receptor deficiency) yields objective evidence for disordered phonation. EGG analysis of vocal fold functionality identifies a high f₀ and abnormal reversals in harmonic spectral decay. This indicates perturbed larynx development and vocal fold function and explains the characteristic high pitched, raspy, and screechy voice. GHD acquired in early adulthood, with onset between late puberty and mature adulthood, may also affect phonation by disrupting normal larynx development. When GHD is acquired in midadulthood, well after completion of puberty, laryngeal size and vocal pitch are normal, but vocal fold functionality may be reduced as a consequence of GHD.

	Acknowledgments

 
The authors thank Susan Jones and her staff (Department of Speech and Language Therapy, Wythenshawe Hospital, Manchester, UK) for their expert help in gathering the phonation data for this study.

	Footnotes

 
This work was supported in part by Engineering and Physical Sciences Research Council Grant GR/R047 13/01; a merit review grant from the Department of Veterans Affairs; and a travel grant from the Pharmacia Corp.

First Published Online May 3, 2005

Abbreviations: ApEn, Approximate entropy; EGG, electroglottogram; f₀, fundamental frequency; FHN, fundamental harmonic normalization; GHD, GH deficiency; PSD, power spectral density.

Received December 28, 2004.

Accepted April 26, 2005.

	References

Top Abstract Introduction Subjects and Methods Results Discussion References

 

1. Maheshwari HG, Silverman BL, Dupuis J, Baumann G 1998 Phenotype and genetic analysis of a syndrome caused by an inactivating mutation in the growth hormone releasing hormone receptor: dwarfism of Sindh. J Clin Endocrinol Metab 83:4065–4074[Abstract/Free Full Text]
2. Maheshwari HG, Rahim A, Shalet SM, Baumann G 1999 Selective lack of growth hormone (GH) response to the GH-releasing peptide hexarelin in patients with GH-releasing hormone receptor deficiency. J Clin Endocrinol Metab 84:956–959[Abstract/Free Full Text]
3. Manickam K, Moore CJ, Willard T, Jones S, Slevin N 2002 Approximate entropy in the analysis and monitoring of voice quality changes in larynx cancer patients following radiotherapy. London: Proceedings IEE Medical Applications of Signal Processing, ISSN 0963-3308, Ref No 2002/110
4. Moore C, Manickam K, Willard T, Jones S, Slevin N, Shalet S 2004 Spectral pattern complexity analysis and the quantification of voice normality in healthy and radiotherapy patient groups. Med Eng Phys 26:291–301[Medline]
5. Moore CJ, Slevin N, Winstanley S, Woods H, Allan E, Birzgalis AR, Farrington WT 1999 Computerised quantification and 3D-visualisation of voice quality changes following radiotherapy for carcinoma of the larynx. Harrogate, UK: Brit Comp Soc Proceedings: Current Perspectives in Healthcare Computing HC99; 137–145, ISBN 0 9535 427 0 X
6. Titze IR 1994 Principles of voice production. Englewood Cliffs, NJ: Prentice-Hall
7. Fourcin AJ 1986 Electrolaryngographic assessment of vocal fold function. J Phon 14:435–442
8. Moore CJ, Slevin N, Winstanley S 1999 Characterising vowel phonation by fundamental spectral normalisation of Lx-waveforms. Proc International Workshop Models, Analysis of Vocal Emissions for Biomedical Applications. Florence, Italy: Firenze University Press; 1–6
9. Pincus SM 1991 Approximate entropy as a measure of system complexity. Proc Natl Acad Sci USA 88:2297–2301[Abstract/Free Full Text]
10. Pincus SM, Gevers EF, Robinson ICAF, Van den Berg G, Roelfsema F, Hartman ML, Veldhuis JD 1996 Females secrete growth hormone with more process irregularity than males in both human and rat. Am J Physiol 270:E107–E115
11. Baumann G 2002 GHRH receptor defects. J Pediatr Endocrinol Metab 15(Suppl 5):1431–1433
12. Traunmüller H, Eriksson A 1993 The frequency range of the voice fundamental in the speech of male and female adults. Stockholm: Institutionen för Lingvistik, Stockholms Universitet (http://www.ling.su.se/staff/hartmut/f0_m&f.pdf)
13. Clark S, To sound like a woman. Proc Fifth International Gender Dysphonia Conference GENDYS98, Manchester, UK, 1998 (http://www.gender.org.uk/conf/1998/clark.htm)
14. Williams RG, Eccles R 1990 A new clinical measure of external laryngeal size which predicts the fundamental frequency of the larynx. Acta Otolaryngol (Stockh) 110:141–148
15. Hoffman WH, Supal C, Tosi O 1984 Computer analysis of acoustical parameters in hypopituitary children before and after growth hormone treatment. Intl J Pediatr Otolaryngol 7:1–9[Medline]
16. Rosenbloom AL, Guevara Aguirre J, Rosenfeld RG, Fielder PJ 1990 The little women of Loja—growth hormone-receptor deficiency in an inbred population of southern Ecuador. N Engl J Med 323:1367–1374[Abstract]
17. Burren CP, Woods KA, Burren CP, Woods KA, Rose SJ, Tauber M, Price DA, Heinrich U, Gilli G, Razzaghy-Azar M, Al-Ashwal A, Crock PA, Rochiccioli P, Yordam N, Ranke MB, Chatelain PG, Preece MA, Rosenfeld RG, Savage M 2001 Clinical and endocrine characteristics in atypical and classical growth hormone insensitivity syndrome. Horm Res 55:125–130[CrossRef][Medline]

This Article Abstract Full Text (PDF) Submit a related Letter to the Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Moore, C. Articles by Baumann, G. Search for Related Content PubMed PubMed Citation Articles by Moore, C. Articles by Baumann, G. Related Collections Neuroendocrinology and Pituitary