Results

Abstract

Knowledge about the changes of the vocal folds during puberty in children has been focused upon in many cases. Döllinger et al. write that children demonstrate greater cycle-to-cycle variability in oscillations compared to adults [1]. Based on videokymography and high-speed videos, Cavalli et al. suggest changes of treatment of pediatric voice disorders because there are important differences between the developmental approach and disorders for surgical and therapeutic management [2]. As late as 2012, an evaluation of pathology was made by Martins et al. of 304 children from 4 to 18 years where the findings related to normal pubertal development were not commented on [3].

FormalPara Core Messages

• High-speed video examples can on the vocal folds show marking of two maximal closures indicating four pubertal registers, two child registers, and two adult registers in pubertal boys. In girls, the differences between prepubertal, pubertal, and postpubertal vocal folds are not clearly seen for pubertal changes.

• Voice Range Profiles had bigger dynamics in the older pupils in both sexes, and in boys after the pubertal register shift with reminiscences of a child—and beginning adult registers. A falling of the lowest tones is seen in both sexes.

• The fundamental frequency in the reading of a standing text with a conventional voice falls in boys with an octave and in girls with one-third to one-fourth octave, related significantly to an increase of the tonal range in the continuous speech of four to five semitones on a chromatic scale, for both girls and boys.

• The statistically significant results are that the change of fundamental frequency in puberty is related to testosterone in boys and estrone sulfate (log E1 sulfate) in girls. In boys, it is predicted by the fall of sex hormone-binding globulin (SHBG), and in girls by a widening of the tonal range during speech (log E1 sulfate).

• Prospectively, testosterone values over 10 nmol/L suggest a boy in vocal puberty. A girl after menarche and with a tonal range in the continuous speech of five semitones is postpubertal.

4.1 High-Speed Videos

Knowledge about the changes of the vocal folds during puberty in children has been focused upon in many cases. Döllinger et al. write that children demonstrate greater cycle-to-cycle variability in oscillations compared to adults [1]. Based on videokymography and high-speed videos, Cavalli et al. suggest changes of treatment of pediatric voice disorders because there are important differences between the developmental approach and disorders for surgical and therapeutic management [2]. As late as 2012, an evaluation of pathology was made by Martins et al. of 304 children from 4 to 18 years where the findings related to normal pubertal development were not commented on [3].

Videos at best with high-speed setups of normal child development are necessary for defining the limitations of pathology and treatment hereof. Presented are pictures from high-speed films from the three traditional childhood/adolescence periods in pediatrics—the first period is traditional childhood after adrenarche: the prepubertal period ending at 12.9 years; the second period is the pubertal period from 13 to 15.9 years; and the third postpubertal period is the beginning of 16 years of age. Traditionally, no variations of vocal fold surface are attributed to puberty, but some differences have been found that can be considered related to puberty.

The material presented includes 18 examples from HSV with three films of girls and boys: at the beginning, middle, and end of each period. The examples correspond to the three pediatric periods in the literature, later described in our material (prepubertal, pubertal, and postpubertal groups, Figs. 4.30 and 4.31) [4]. The three films of variations inside the three periods are of interest, especially to boys.

The prepubertal girls and boys do in some cases have some thickness and slight irregularity of the vocal folds. Interestingly, in some boys in the pubertal group, two maxima of the edges of the vocal folds are seen, heard as cracks, and found with Voice Range Profiles as four registers as shown later (Fig. 4.6). In girls, in most cases, a rear insufficiency of the vocal fold is seen during intonation—no valid characteristic changes are found (4.1.1). Probably, a main finding in the postpubertal group is the thickness and irregular surfaces of the vocal folds, especially in boys. It can be concluded that high-speed video alone is not sufficient to define a pubertal voice neither in girls nor in boys. Kymographic video pictures from the middle of the vocal folds are added for documentation. It is noted that the kymography closures were different at other places of the glottis during vocal fold development.

Few descriptions of the surfaces of the vocal folds on HSV of pubertal children were found in the literature. Lee et al. described the glottal gaps found in some choir children at various frequencies, using videoendoscopies [5]. The descriptions of our results presented are made based on the high-speed videos with voiced intonation as in conversations of an /a/ using a stiff endoscope. The findings are described tentatively. It was decided to present the pubertal high-speed videos first in girls and boys, thereafter the postpubertal, and at last, the prepubertal ones of girls and boys to include all five Tanner stages of pediatric development [4].

4.1.1 Findings on HSV of the Vocal Folds in Pubertal Girl in the Beginning, Middle, and End

At the beginning of the pubertal stage in the girls’ examples on the HSV, the surface of the vocal folds is shiny and slightly irregular. The closure is irregular with a rear glottal gap. There is slight thickening especially at the rear, seen on the film, but no mucus (Fig. 4.1). In the middle of the pubertal stage, there are no clear signs of puberty. The surface of the vocal folds is matte partly in the rear, and partly in the front, but regular. There is a rear glottal gap and a slight thickening of the vocal folds (Fig. 4.2). and at the end of the pubertal stage, the surface of the vocal folds is shiny without irregularities, with a small rear glottal gap, slight thickening of the vocal folds, and no mucus (Fig. 4.3).

Fig. 4.1

Girl at the beginning of the pubertal stage, 13 years. Image: A002_0152–A002_0177. F0 = 311, dB(A) = 79, kymography, in the middle of the vocal folds. The surface of the vocal folds is shining and slightly irregular. The closure is irregular with a rear glottal gap. There is slight thickening especially at the rear, seen on the film, but no mucus

Fig. 4.2

Girl in the middle of the pubertal stage, 14 years, F0 = 272, loudness (dB(A)) = 83. Image: A001_0000–A001_0025, kymography, in the middle of the vocal folds, no clear signs of puberty. The surface of the vocal folds is partly matte, but regular. There is a rear glottal gap and a slight thickening of the vocal folds

Fig. 4.3

Girl at the end of the pubertal stage, 15 years, F0 = 340, loudness (dB(A)) = 83. Image: A004_0000–A004_0025, kymography in the middle of the vocal folds. The surface of the vocal folds is partly shiny, with no irregularities, and a small rear glottal gap

4.1.2 Findings on HSV of the Vocal Folds in Pubertal Boys in the Beginning, Middle, and End

At the beginning of the pubertal stage with the HSV of the boys, the surface of the vocal folds is in part shining, in part matte, and regular with a moderate glottal rear gap. A rather big thickening of the vocal folds is seen, but no mucus (Fig. 4.4). In the middle of the pubertal stage, the surface of the vocal folds is mostly matte, and irregular without a rear glottal gap, but with a slight marking of 2 maxima of the glottal closure (Fig. 4.5). At the end of the pubertal stage, the surface of the vocal folds is matte and irregular with 2 maxima of closure covered with a small amount of mucus. Slight thickening of the vocal folds is seen but with no rear glottal gap (Fig. 4.6).

Fig. 4.4

Boy at the beginning of the pubertal stage, 13 years, F0 = 300, loudness (dB(A)) = 79. Image: A002_016–A002_0190, kymography in the middle of the vocal folds. The surface of the vocal folds is in part shining, in part matte, and regular with a moderate glottal rear gap. A rather big thickening of the vocal folds is seen, but no mucus

Fig. 4.5

Boy in the middle of the pubertal stage, 15 years, F0 = 159, loudness (dB(A)) = 74. Image: A002_0000–A002_0025, kymography, in the middle of the vocal folds. The surface of the vocal folds most matte, and irregular without a rear glottal gap, but with slight marking of two maxima of the glottal closure

Fig. 4.6

Boy at the end of the pubertal stage. 15 years, F0 = 145, loudness (dB(A)) = 77. Image: A003_0004–A003_0029, kymography between the two maxima contact of the vocal folds. The surface of the vocal folds is matte and irregular with two maxima of closure covered with a small amount of mucus. Slight thickening of the vocal folds, but with no rear glottal gap

4.1.3 Findings on HSV of the Vocal Folds in Postpubertal Girls in the Beginning, Middle, and End

At the beginning of the postpubertal stage in the HSV of the girls’ examples, the surface of the vocal folds is partly shining, partly matte, and slightly thickened and irregular. The closure is slightly irregular with slight rear insufficiency. Mucus is seen in the front of the glottis (Fig. 4.7). In the middle of the postpubertal stage, the surface of the vocal folds is matte, nearly regular. There is a slight rear glottal gap and slight thickening of the vocal folds, but no mucus (Fig. 4.8). At the end of the postpubertal stage, the surface of the vocal folds is shiny and mostly regular with a small rear glottal gap, some thickening of the vocal folds, and a small amount of mucus on the vocal folds (Fig. 4.9).

Fig. 4.7

Girl at the beginning of the postpubertal stage, 16 years, F0 = 205, loudness (dB(A)) = 82. Image: A004_0000–A004_0025, kymography, in the middle of the vocal folds. The surface of the vocal folds is partly shining, partly matte, and slightly thickened and irregular. The closure is slightly irregular with slight rear insufficiency. Mucus in the front of the glottis

Fig. 4.8

Girl in the middle of the postpubertal stage, 17 years, F0 = 200, loudness (dB(A)) = 88. Image: A001_0329–A001_0354, kymography, in the middle of the vocal folds. The surface of the vocal folds is matte, nearly regular. There is a slight rear glottal gap and slight thickening of the vocal folds, but no mucus

Fig. 4.9

Girl at the end of the postpubertal stage, 19 years, F0 = 328, loudness (dB(A)) = 85. Image: A001_0002–A001_0027, kymography, in the middle of the vocal folds. The surface of the vocal folds is shiny and mostly regular with a small rear glottal gap, some thickening of the vocal folds, and a small amount of mucus on the right vocal fold

4.1.4 Findings on HSV of the Vocal Folds in Postpubertal Boys in the Beginning, Middle, and End

At the beginning of the postpubertal stage in the boys’ examples, the surface of the vocal folds is mostly shiny but irregular, with a slight marking of two maxima of contact between the vocal folds with mucus and a minimal rear glottal gap (Fig. 4.10). In the middle of the postpubertal stage, the surface of the vocal folds is mostly shiny and slightly irregular on the left side with a hint of two closure maxima. Minimal rear glottal gap, and no mucus (Fig. 4.11). At the end of the postpubertal stage, the surface of the vocal folds is mostly shiny and regular. There is no rear glottal gap and no mucus. Slightly increased thickness of the vocal folds is seen (Fig. 4.12).

Fig. 4.10

Boy at the beginning of the postpubertal stage, 16 years. Image: A001_0031–A001_0056, F0 = 202, dB(A) = 77, kymography, in the middle of the vocal folds. The surface of the vocal folds is mostly shiny but irregular, with a slight marking of two maxima of contact between the vocal folds, the mucus, and a minimal rear glottal gap

Fig. 4.11

Boy in the middle of the postpubertal stage, 17 years, F0 = 143, loudness (dB(A)) = 75. Image: A001_0016–A001_0041, kymography, in the middle of the vocal folds. The surface of the vocal folds is mostly shiny, slightly irregular on the left side with a hint of two closure maxima. Minimal rear glottal gap, and no mucus

Fig. 4.12

Boy at the end of the postpubertal stage, 18 years, F0 = 154, loudness (dB(A)) 70. Image: A002_0014–A002_0039, kymography, in the middle of the vocal folds. The surface of the vocal folds is mostly shiny and regular. There is no rear glottal gap and no mucus. Slightly increased thickness of the vocal folds

4.1.5 Findings on HSV of the Vocal Folds in Prepubertal Girls in the Beginning, Middle, and End

At the beginning of the prepubertal stage in the HSV of the girls’ examples, the surface of the vocal folds is matte but regular, and there is a moderate rear glottal gap but no mucus (Fig. 4.13). In the middle of the prepubertal stage, the surface of the vocal folds is partly shiny and partly matte. There is a slight irregularity of the glottis with mucus in the middle and a big rear glottal gap (Fig. 4.14). At the end of the prepubertal stage, the surface of the vocal folds is partly shiny, partly matte, and regular. There is a big rear glottal gap and no mucus (Fig. 4.15).

Fig. 4.13

Girl at the beginning of the prepubertal stage, 11 years. Image: A001_0000–A001_0017. F0 = 315, dB(A) = 70, kymography, in the middle of the vocal folds. The surface of the vocal folds is matte but regular; there is a moderate rear glottal gap but no mucus

Fig. 4.14

Girl in the middle of the prepubertal stage, 11 years. Image A001_0001–A001_0025, F0 = 304, loudness (dB(A)) = 72, kymography, in the middle of the vocal folds. The surface of the vocal folds is partly shiny and partly matte. There is a slight irregularity of the glottis with mucus in the middle and a big rear glottal gap

Fig. 4.15

Girl at the end of the prepubertal stage, 12 years, F0 = 274, loudness (dB(A)) = 83. Image: A002_0000–A002_0025, kymography, short recording in the middle of the vocal folds. The surface of the vocal folds is partly shiny, partly matte, and regular. There is a big rear glottal gap and no mucus

4.1.6 Findings on HSV of the Vocal Folds in Prepubertal Boys in the Beginning, Middle, and End

At the beginning of the prepubertal stage, on the HSV, the vocal folds are matte. There is a slight irregularity of the glottis, maybe at 2 points especially, and a moderate rear glottal gap. No mucus is seen (Fig. 4.16). In the middle of the prepubertal stage, the surface of the vocal folds is matte and slightly irregular; there is a small rear glottal gap and some thickening of the vocal folds. There is no mucus (Fig. 4.17). At the end of the prepubertal stage, the surface of the vocal folds is mostly shiny but slightly irregular. There is a small amount of mucus in the middle of the membranous part of the vocal folds. There is a big rear glottal gap (Fig. 4.18).

Fig. 4.16

Boy at the beginning of the prepubertal stage, 9 years, F0 = 327, loudness (dB(A)) = 79. Image: A001_0010–A001_0027, kymography in the middle of the vocal folds. The surface of the vocal folds is matte. There is a slight irregularity of the glottis, maybe at 2 points especially, and a rear glottal gap. No mucus is seen

Fig. 4.17

Boy in the middle of the prepubertal stage, 10 years, F0 = 313, loudness (dB(A)) = 99. Image: A002_0000–A002_0025, kymography, in the middle of the vocal folds. The surface of the vocal folds is matte and slightly irregular; there is a rear glottal gap and some thickening of the vocal folds. There is no mucus

Fig. 4.18

Boy at the end of the prepubertal stage, 11 years, F0 = 356, loudness (dB(A)) = 73. Image: A002_0000–A002_0025, kymography, in the middle of the vocal folds. The surface of the vocal folds is mostly shiny but slightly irregular. There is a small amount of mucus in the middle of the membranous part of the vocal folds. There is a big rear glottal gap

The high-speed film examples supplement the description of childhood voices including pubertal vocal fold studies, especially in boys where the cracks of voice could correspond to two maxima of contact between the vocal folds during puberty. In girls, a specific lack of contact between the vocal folds in the rear part of the vocal folds could be the reason for a lack of vocal intensity/loudness found during the pubertal period. Overall, there is a rather big variation in the appearance of the vocal folds—from normal shining to matte, with thickening and irregularity.

The high-speed video setups will most likely include more pixels in the future, and a new study might refine the results and understanding of the pubertal vocal fold changes. We have used boys and girls from an elementary school and high school with an amateur choir for the study—because a kind of standard reference for voice development is needed. The choir was evaluated as an amateur choir by guests from the professional Thomanerchor in Leipzig.

The puberty phenomena dominate the vocal fold appearance over the specific minimum musicality demanded. The results are therefore usable also without specific musical tests.

4.2 Voice Range Profiles During Voice Development

Voice Range Profiles are of interest to describe the developing voice. Till now, the change of fundamental frequency in speech has been used for comparison to pediatric and hormonal development. For the children, it is informative to understand normal voice development—as part of their identity. This is the case for both girls and boys for the deepest tone, as well as tone range and dynamics in speech and singing. Intensity ranges are included in Voice Range Profiles. The profile is in itself informative, e.g., before and after training and treatment. In our study, calculation of Voice Range Profile areas, using the diatonic scale, is included for the information on intensity variation and comparison with pediatric and hormonal development.

The risk for pathology is diminished when the youngsters can be informed about the normal changes. Some youngsters use registers for pop singing, and later also as a basis for adult upper register voice management, and have great fun as amateurs or professionals. With Voice Range Profiles, many voice nuances may be found, e.g., whether the voice has a higher intensity range for upper or lower tones.

The normal Voice Range Profiles of girls’ voices show measurable changes during childhood and puberty (Fig. 4.19). In childhood before the pediatric defined puberty (see Fig. 4.30), they demonstrate smaller Voice Range Profiles. At the beginning of puberty, there are modest changes. But then at the age of about 14.5 years, alterations with a change of registers take place, including a passing reduction of the intensity in the middle of the tone range [6].

Fig. 4.19

Girls’ Voice Range Profiles of different ages. (a) 8.9 years. (b) 11.7 years, typical child’s voice with dominating intensity in the upper part, change of register at 330–392 Hz. (c) 13.8 years, voice with slight register changes with greater dynamic breadth in the lower part. (d) 14.8 years, pubertal voice with passing reduced intensity in the middle

In this cross-sectional stratified study, Voice Range Profiles in girls did change in a more well-defined way than in high-speed videos. Voice Range Profiles in Fig. 4.20 show a difference in Voice Range Profiles in girls before and after the pediatric defined puberty: register change, lowest tones, maximal intensity variations, and Voice Range Profile areas. For girls, it is a piece of good information to know that a voice can normally be very light (soprano) or dark (alto). The differences are measured with different intensity areas, with a marking on the abscissa of the position of the “artistic” singing rage for prepubertal 1 and 2 soprano and alto—postpubertal 1 and 2 soprano and alto.

Fig. 4.20

Girls’ Voice Range Profile development during childhood. In the upper frequency range, there is a bigger intensity for the sopranos, and in the lower frequency range for the altos. Prepubertal first and second soprano and alto and postpubertal first and second soprano and alto are shown. The singing range of the “artistic” voice is given on the abscissa

Figure 4.21 shows the boys’ development of the Voice Range Profile in childhood, during puberty, and after the maximal pubertal change defined by pediatrics analysis [7]. In puberty, the Voice Range Profile is smaller; the lowest tone, the total tone range, and the registers are altered. The boys, in a way, have two times two registers, one with a reminiscence of a child’s voice and one with a kind of adult sound, which was illustrated in our high-speed video descriptions with two contact maxima of the vocal folds in some cases. After the pediatric defined puberty, the lowest tone lies deeper, and the Voice Range Profile areas of the intensity of the lower and the upper parts increase.

Fig. 4.21

Boys’ biological Voice Range Profile from childhood over puberty to past puberty representing child voices and beginning adult voice ranges. (The range of the “artistically” usable singing voice is marked on the abscissa. (a) 9-Year-old child at adrenarche. (b, c) Child voice (soprano) with higher intensity for upper tones. (d) Child voice (alto) with higher intensity for lower tones. (e) Voice in puberty. (f, g) Beginning adult voice (tenor). (h, i) Bass

The development of the voice can be described by average Voice Range Profiles with ranges every year from 8th–9th to 19th age. The number of children in each group is given in the pediatric section where the pediatric and hormone measures are combined with the measured mean fundamental frequency (F0), the total semitone range, the lowest tones, the semitone ranges in continuous speech on the chromatic scale, and the Voice Range Profile areas on the diatonic scale. The computer-assisted calculations of the measurement results have opened some new possibilities: it is possible to determine the “average Voice Range Profile” for each year from the Voice Range Profile of the individuals with the PG-200 software of our phonetograph earlier described and present the ranges of the Voice Range Profiles. The standard deviations of the lowest and highest semitones were calculated, usable for the lowest tones; for the highest tones, the spread is higher related to other factors such as talent and training (Figs. 4.22 and 4.23) [8].

Fig. 4.22

Girls’ average Voice Range Profiles with ranges, as a function of age. The abscissa is tones and frequency in Hz, and the ordinate is dB(A)

Fig. 4.23

Boys’ average and range of Voice Range Profiles. The abscissa is tones and frequency in Hz, and the ordinate is dB(A)

As earlier presented regarding individual girls in Fig. 4.19, a slight reduction of intensity in the upper part of the Voice Range Profile is seen in the 14- and 15-year-old girls, but thereafter the tone range is extended. In Fig. 4.22, the yearly changes in girls with ranges are shown. In boys, the yearly average range of Voice Range Profiles changes from childhood to adulthood with a reduction during the middle of the pubertal period. This is the case both when measuring the semitone range on the chromatic scale and when measuring the Voice Range Profile area on the diatonic scale. The standard deviations are shown for the lowest and highest semitones in the prepubertal, pubertal, and postpubertal groups. For the highest semitones, some of the standard deviations are rather high, not good for comparison with pediatric development. They were often not significantly related to pubertal biological development.

The development of the voice during puberty was investigated in the current work within the framework of a prospective stratified randomized study. Longitudinal prospective studies have in this context the advantage that intraindividual comparisons can be performed. For this reason, we have investigated three boys over the period of one school year (from the beginning to the end of the eighth grade). Measurements were carried out every 2 months. The six Voice Range Profiles for one of the boys are shown in Fig. 4.24 [9]. The average Voice Range Profiles and ranges for the three boys before and during the change of voice were also worked out with our Voice Range Profile software (PG-200, means, ranges, and standard deviations for the lowest and highest tones) (Fig. 4.25). The start of the change of voice happened for all three boys during their eighth school year. The vocal changes during this year were not dependent on age; the deepest tone in the Voice Range Profile, which our investigations had shown to be highly correlated with the fundamental frequency of the speaking voice (F0), was significantly dependent on the SHBG level. SHBG showed itself in this study to be the most sensitive parameter for predicting the lowering of the frequency of the voice as later discussed.

Fig. 4.24

6 Voice Range Profiles of one boy measured at intervals of 2 months (age 13.7–14.6 years) in the eighth school class. The third Voice Range Profile (December) has the biggest area and shows the smallest irregularities (c1 = C4 = 262 Hz)

Fig. 4.25

Three boys’ average Voice Range Profiles and standard deviations for the highest and lowest tones (I–III) were involved in the prospective longitudinal study of the eighth school year from August to June. The Voice Range Profiles before and after the change of voice were compared. For test boy I, only one Voice Range Profile was made that showed mutation, and for boy II, only one was before mutation. For test person III, three Voice Range Profiles were measured (in December) before and three during the pubertal change of the Voice Range Profile

Only one segment of the entire period of puberty was investigated in the eighth school class, and no significant relationships between the changes in the Voice Range Profile areas and testosterone level were found. With respect to the serum testosterone level, there is during this stage of puberty considerably more interindividual variation than in earlier or later stages. The Voice Range Profile areas likewise change markedly over a short period of time: the Voice Range Profile simply becomes more irregular, and the changes between registers appear more distinctly. Attempts to give a mathematical description of the irregularities (by a characteristic number for the Voice Range Profile, or a fractal dimension) have, however, so far not produced any satisfactory results; incorporation of these values in the statistical calculations was not meaningful [10, 11].

Voice Range Profiles are informative to show the normal development of voice. The results are usable in schools if parents, teachers, or normal pupils doubt their voices or just want to know more—even more so in cases of various kinds of normality and pathology. A routine service in schools can very well be established not only for pathology.

4.3 Fundamental Frequency with Electroglottography and Register Analysis

The fundamental frequency of children has been measured in many ways—more or less exact methods have been used. Probably measuring the mean fundamental semitone of spontaneous speech is sufficient in the daily routine. To compare F0 with especially the hormonal development, a well-defined measurement independent of outer phenomena, especially noise and also harmonic overtones, was chosen using electroglottography during continuous speech, defined as reading of a standard text with a conversational voice. The changes with age and during puberty can vary very much, and especially in girls, they can be small.

The change in the fundamental frequency in continuous speech for girls is—given in Hz—smaller and less pronounced than the change in the deepest tone in the Voice Range Profile. The clearest changes take place in the semitonal range of continuous speech (the postpubertal group of girls’ 5 semitones on a chromatic scale) as shown in Fig. 4.26 [6, 12]. A marking of the related breast stages was given, as discussed in the next chapter.

Fig. 4.26

Girls’ graphical representation of mean vocal parameters, Voice Range Profile areas, semitonal range of the voice in continuous speech, total semitone range, deepest semitones, mean fundamental frequency (F0) of the voice in conversational speech, as a function of age (abscissa): filled circle: breast development stage 1; open circle: breast development stages 2–4; open triangle: breast development stages 5–6

A commonly overlooked fact is that the physiological voice changes in Hz have larger physical effects for girls due to the position in one higher octave than in boys. The mean fundamental frequency in continuous speech for girls changes from 256 Hz in the prepubertal group to 241 Hz in the postpubertal group. The semitonal range in continuous speech (F0) increases from its prepubertal value of 3.7 semitones to a postpubertal value of 5.2 semitones; this change is significant to 99%.

The fundamental frequency in conversational, continuous speech (F0) is by itself a frequently investigated parameter used for describing the development of the voice. As later mentioned, there is for girls no significant correlation between the fundamental frequency during continuous speech (F0) and the Voice Range Profile areas; however, there are correlations between the Voice Range Profile areas, the deepest tone of the Voice Range Profile, and the tonal range in semitones in continuous speech.

For boys, the fundamental frequency in continuous speech (F0) gets deeper with age in a manner parallel to that of girls. The deepest tone of the Voice Range Profile falls at the same time as the semitonal range in continuous speech, and the Voice Range Profile areas expand, apart from the age of around 14.5 years where a reduction in the Voice Range Profile takes place (see Fig. 4.27) [7, 8].

Fig. 4.27

Boys’ graphical representation of the mean fundamental frequency in continuous speech (F0), the semitonal range of the mean fundamental frequency of voice in continuous speech with the chromatic scale with 12 semitones in an octave, the deepest tone, and the Voice Range Profile area with the diatonic scale with 7 tones in an octave. All as a function of age—filled circle: 8.7–12.9 years; open circle: 13–15.9 years; open triangle: 16–19.5 years

In Fig. 4.28 of boys’ voices, some age-related changes are presented in a drawing of 25 boys in an earlier study. In figure A, the lowering of the fundamental frequency in continuous speech (F0) is given, combined with the total semitone range in Hz and the semitonal range in continuous speech. An arrow marks the voice-related pubertal change. The change in the boys’ height is added. In Fig. 4.28b, the fundamental frequency in continuous speech (F0) is compared with the total serum testosterone and an arrow is made of the beginning and end of the voice-related pubertal period.

Fig. 4.28

(a) Boys’ mean fundamental frequency in continuous speech (F0), tonal range of voice in continuous speech in Hz, and total tone range in Hz compared to body height (ordinate) and age (abscissa) in 25 boys. The arrow indicates the end of the voice change. (b) Boys’ mean fundamental frequency during continuous speech (F0) in the 25 boys compared to total serum testosterone level. The abscissa shows the age in years. The arrows indicate the beginning and end of the pubertal voice change

The development of girls’ voices shows noticeable differences compared to boys. For girls, the average fundamental frequency in continuous speech (F0) changes independently of the Voice Range Profile areas (r = 0.29), whereas for boys, the dependency between these two parameters persists (r = 0.50). For the semitonal range of the fundamental frequency in continuous speech, there was no difference between the two sexes; both are related to the mean F0 (girls: r = 0.54; boys: r = 0.49) as discussed in 4.4 and 4.5. The changes in the Voice Range Profile areas depend on the stage of pubic hair development for girls (r = 0.51) and for boys (r = 0.65). For girls, there is also a connection to breast development. The statistics are elaborated on in Sect. 4.5.

Register shifts were analyzed specifically and averaged in boys. The relation between total serum testosterone and registers was measured. Three groups were calculated based on the boys in our referred study (see also 4.4.2) of serum testosterone of less than 1 nmol/L, 1–10 nmol/L, and more than 10 nmol/L (Fig. 4.28). The lowest and highest biological tones were measured in Hz. Hereafter, the tones usable in artistical singing are defined by self-evaluation, and the register changes of singing are defined. There was a clear change of register shift of the artistic singing tone range with group 3, where the total serum testosterone was more than 10 nmol/L. The standard deviations for the lowest tones in the groups were low; note the high standard deviations for the highest tones [13] (Fig. 4.29).

Fig. 4.29

Boys’ register changes in our study (calculations in Log Hz) grouped according to serum testosterone level. The tonal range of the voice is given here, first as the biological range and second as the classic artistically usable tonal range

In this study, detailed measurements of fundamental frequency in continuous speech (F0) are given for mean values and semitone ranges on a chromatic scale in girls and boys. In the pediatric and hormonal literature, there is a traditional division of groups into prepubertal, pubertal, and postpubertal groups. More details are presented usable in various kinds of normality and in pathologies including genetic disorders. The details can be compared separately to various defects and various kinds of genetic and social sexuality.

4.4 Puberty Stages and Hormonal Analysis

Pubertal stages are well defined in the hormonal and pediatric history based on standards [14, 15]. There are differences between the two sexes. But for the pubertal grouping, a comparison can be of interest relating voice development to the traditional pediatric defined two sexes. The hormonal regulation of not only the mean fundamental frequency in boys (F0) but also the lowest measurable, “biological” tone is of interest in the sexes. To use tones defined only by the quality of sound to be heard as artistical singing would be impossible in many cases of puberty. This is the reason for a choice of a measurable sound only. For their lowest tones, girls have a significant hormone-related change. The relation is weak for the mean fundamental frequency. The results give a background for further studies also related to the adrenarche, which is represented by the prepubertal period.

In the following, the hormonal and pediatric mean measurements are presented for girls and boys in the prepubertal, pubertal, and postpubertal groups as defined by Tanner [14]. In girls, E1 (estrone) is especially of interest in our study where a difference was found between the three groups changing from 57 pmol/L to 123 pmol/L. The semitone range in continuous speech changed from 3.7 to 5.2 semitones and the lowest tone changed from 166 to 154 Hz. The mean fundamental frequency in continuous speech changed only from 236 to 241 Hz (Fig. 4.30).

Fig. 4.30

Girls’ geometrical averages of hormonal, pubertal, and vocal parameters in three groups by age. The relative standard deviation lay between 11 and 140%. (Significance of the differences between the groups: **p < 0,01; *p < 0,05)

In boys, the changes in the pediatric and hormonal values are clearly related to the development of the mean fundamental frequency (F0) and to the lowest tone as well. The total serum testosterone changes from 0.54 nmol/L to 18.9 nmol/L. The fundamental frequency (F0) changes from 273 Hz to 125 Hz. The tone range in semitones of the mean fundamental frequency (F0) is of the same range in the two sexes—even if the difference in frequency (Hz) is of the half range in boys. The many measured hormonal parameters give further information usable for the prediction of voice change and for the pathology of voice.

For boys, the average annual changes were evaluated, for example, the fundamental frequency of the speaking voice (F0) (11%) and the Voice Range Profile areas (9.2%) (Fig. 4.31). Another significantly changing parameter is the deepest tone of the Voice Range Profile, which falls 16% to a similar extent to the fundamental frequency of the speaking voice (F0) (12%). The androgen level rises, and the level of SHBG falls [16].

Fig. 4.31

Boys’ geometrical average of hormonal, pubertal, and vocal parameters (grouped according to age) and the annual change in these parameters in %

For the girls, there were significant differences between the groups of all prepubertal against postpubertal voice categories with respect to the Voice Range Profile areas and the semitonal range in continuous speech. The pubertal group was not defined. For the mean fundamental frequency in continuous speech (F0), however, no significant differences between the groups could be seen. For E1 and E1 sulfate, we found significant differences between the pre- and postpubertal girls (p < 0.001); this was also the case for androstenedione and DHEAS. No significant differences were found inside the pre- and postpubertal groups. As referred, E1 (serum estrone level) rises from 57 to 123 pmol/L (pico means nano/1000). Body weight was on average for the youngest group of girls 37.8 kg and for the oldest group 64.4 kg. In the age group of 8.6–12.9-year-old girls, 4 out of 18 had already reached menarche; in the age group of 16–19.5-year-olds, all the girls had reached menarche (Figs. 4.30 and 4.32) [6, 12].

Fig. 4.32

Girls’ geometrical averages of vocal and hormonal parameters for different voice categories in the group (8–19 years). (1): Child’s voice first soprano. (2): Child’s voice second soprano. (3): Child’s voice alto. (4): Mutating voice (no values shown). (5): Adult voice first soprano. (6): Adult voice second soprano. (7): Adult voice alto. SD—standard deviation of the mean values. Significance calculated using t-test: prepubertal groups 1–3 versus postpubertal groups 5–7. No significant differences were found inside the prepubertal and the postpubertal groups for the various voice categories. (NS = no significance)

There was a linear correlation between the SHBG level and the arrival of menarche for girls (r = 0.93); this correlation could however not be confirmed if the statistical calculation was based on logarithmically transformed values [17].

For our stratified study, we also divided the boys into three groups: prepubertal, pubertal, and postpubertal voices, including the voice categories inside the groups. The Voice Range Profiles of the groups differed significantly (p < 0.01) with respect to the areas, the lowest tone, and the total tonal range in semitones between groups. The same is true for the fundamental frequency in continuous speech (F0), for serum testosterone, and for SHBG (Fig. 4.33) [8]. No significant difference was found inside the pre- and postpubertal groups for the voice categories.

Fig. 4.33

Boys’ mean values compared between voice categories in the groups. (1): Non-differentiated beginners, (2): first soprano, (3): second soprano, (4): alto, (5): puberty, (6): first tenor, (7): second tenor, (8): first bass, (9): second bass with respect to Voice Range Profile area, the fundamental frequency of the speaking voice (mean F0), total tone range of the voice in semitones, SHBG, stage of pubic hair development and free testosterone. Voice categories were measured, but no differences were found inside the prepubertal and postpubertal groups. Mutual SD within groups (f = 39) in percent of mean: Free testosterone nmol/L = 315, SHBG nmol/L = 61, VRP area semitones * dB(A) = 1184, F0 in Hz = 17, tone range semitones = 17

Statistical methods also appear to be useful for advanced descriptions of voice during childhood and adolescence when discussing the placement of voices in choirs. It is also of interest to predict voice change in puberty for the mean fundamental frequency during continuous speech (mean F0). There are many aspects hereof for transsexualism and pathology, especially genetic voice disorders. That is why we calculated all measured parameters with a view of prediction. For girls, we used the logarithmic results and showed that estrone—Log(E1SO4)—and raising semitonal range during continuous speech were significantly predicting the fall of the fundamental frequency. Since menarche is a dominating phenomenon in female puberty, a differentiation between pre- and post-menarche phenomena was made, and the pre-menarche results showed an even more significant relation to E1, estrone—Log(E1SO4)—and height change was also a significant predicting factor. After menarche, again the increasing semitone range in continuous speech but also time after menarche, as well as age, had an influence on the prediction.

For girls for predicting factors as referred to above, there was a significant correlation to an increasing semitone range in continuous speech and an increasing level of estrone sulfate (E1SO4) (p < 0.05), independent of age. Before menarche, there exists a correlation between the level of E1SO4, body height, and stage of development of pubic hair. After menarche, a highly significant dependency (p < 0.001) appeared of the semitonal range in continuous speech, and also with regard to age and the period of time which had passed since menarche: the larger the semitonal range of the speaking voice, the lower the mean fundamental frequency in continuous speech (F0) for the speaking voice of girls in puberty (Fig. 4.34) [6, 12].

Fig. 4.34

Girls’ prediction of the fundamental frequency fall of the speaking voice (F0), evaluated for all test persons and divided into two groups (before and after menarche). A linear correlation coefficient of SHBG with menarche: r = 0.93. Significance: *p < 0.05; **p < 0.01; ***p < 0.001, p = p-value of t-test

The prediction in boys for the change of fundamental frequency in continuous speech (F0) based on coefficients of a multiple regression analysis of all parameters for voice analysis and hormonal and pediatric analysis showed a different result. The pubic Tanner stages 2–4 with mean values of fundamental frequency (F0) of 219 Hz and age of 13.5 years were predicted by a falling Log SHBG (Fig. 4.35). SHBG binds testosterone, and the fall is regulated centrally in the brain. We do not know till now what stimulates the fall of SHBG, as it was discussed earlier. Interestingly, there is also a linear fall of SHBG in girls related to menarche.

Fig. 4.35

Boys’ coefficients between the fundamental frequency of the speaking voice (mean F0) and age, hormonal parameters, and stage of puberty evaluated within the framework of a multiple regression analysis. Independent parameters are not included. Change of mean F0 is predicted by the fall of SHBG in the pubertal group and age. Mean values of the remaining parameters according to grouping. The coefficient is significantly different from zero (*p < 0.05)

With our material, we have—with respect to the mean fundamental frequency in continuous speech (F0)—performed a one-way multivariate analysis, and this has enabled us to predict the timing of the change of boys’ voices in relation to the hormonal and bodily changes in the individual case (Fig. 4.35) [18, 19]. As referred, for boys in the group at stages 2–4 of pubic hair development and an average age of 13.5 years, a correlation between the lowering of the average fundamental frequency in continuous speech (F0) and the falling SHBG level was found. This means that a drop in the fundamental frequency (F0) can be expected when the SHBG level falls under 91 nmol/L in this pubescent stage.

4.5 Further Results from the Statistical Analysis

The tight connection between voice development in childhood, pediatric, and hormonal development was further analyzed statistically, for relations between the three groups of parameters studied: voice parameters, pediatric puberty stages, and hormones. This is of interest when some of the parameters of development are deviant, among others in genetic disorders. In the future, also a focus on the best quality voices can be of interest. The voice parameters were related to hormonal and pediatric development—there is an indication that even the best quality voices are dependent hereon, as shown for example the Voice Range Profile area in girls in Fig. 4.36. Also, in boys, the comparison showed the dependency of the measured voice parameters on hormones and pediatric values in Fig. 4.37.

Fig. 4.36

Girls’ correlation coefficients of different voice and hormonal parameters in relation to age and Voice Range Profile area (age/Voice Range Profile area: r = 0.65). Significance: *p < 0.05 (r ≥ 0.30). Significance: **p < 0.01 (r ≥ 0.39). Significance: ***p < 0.001 (r ≥ 0.49)

Fig. 4.37

Boys’ logarithmic correlation coefficient for vocal and hormonal parameters in relation to age and Voice Range Profile area, respectively. Age/Voice Range Profile area: r = 0.66. All p < 0.01

An increase in body weight is recognized as a normal phenomenon in puberty. The correlation between the development of the Voice Range Profile areas and somatic changes during puberty is significant for both sexes in the case of the stage of pubic hair development, of body weight, and for girls also of mamma (breast) development. Concerning hormonal parameters, androgens play a significant role, both for girls and for boys. For girls, a significant correlation could also be found between E1 and E1 sulfate and the development of the voice. For the height of girls, there is no significant age dependency, while for all other parameters including voice parameters, change is related to age (Fig. 4.36) [6, 12].

The Voice Range Profile areas for boys changed depending on the volume of the testicles, corresponding to the serum testosterone level. There was no significant relation to the voice category as measured with Voice Range Profiles in prepubertal and postpubertal boys. The changes in the Voice Range Profile areas during puberty are however a very complex matter, where age-related development plays a decisive role (Fig. 4.37) [7].

Figure 4.38 shows the mean fundamental frequency in continuous speech (F0) for boys as abscissa compared to the stage of pubic hair development, testicle volume, serum testosterone, and SHBG [7, 8]. In the earlier pilot study of 25 boys, we were able to demonstrate that the fundamental frequency of the speaking voice is high until the age of 13 and that for the age group of 13–15-year-olds the fundamental frequency is also still above 195 Hz, while the serum testosterone level has already risen up to 10 nmol/L (see Fig. 4.32) [16]. Not until they reach the age group of 15 years does the fundamental frequency in continuous speech (F0) fall to below 150 Hz, while the serum testosterone level of this age group is at least 10 nmol/L. The high serum testosterone level shows a correlation with the changing semitonal range, the high tones, and the changes in registers. All young men of 17–18 years had an adult tone of voice. The mean fundamental frequency in continuous speech (F0) was 8–12 semitones above the deepest tone in the Voice Range Profile (Fig. 4.28).

Fig. 4.38

Boys’ graphical representation of the stages of pubic hair development, testicle volume, serum testosterone (total), and SHBG as a function of the fundamental frequency in continuous speech (F0) (abscissa): filled circle: 8.7–12.9 years; open circle: 13–15.9 years; open triangle: 16–19.5 years

A corresponding table was made for girls for the semitone range of the fundamental frequency, lowest measured tone, and age in relation to the mean fundamental frequency in continuous speech (F0). There is a significant relation between E1 and the lowest tone and the mean F0 in continuous speech. Semitone range in continuous speech, lowest tone, and age with multivariate analysis were also related to other parameters where DHEAS is of interest. Among the pediatric puberty phenomena, the pubic hair stage and weight had high significance (Fig. 4.39).

Fig. 4.39

Girls’ correlation coefficients between the fundamental frequency in continuous speech (F0) and semitone range in continuous speech (F0), total tone range, lowest tone, and age compared to the female sex hormones, androgens, stage of pubic hair development, stage of breast development (significance: **p < 0.01; *p < 0.05)

Other calculations and comparisons were also made for girls and boys. The relations to the Voice Range Profiles as abscissa are shown.

Figure 4.40 gives a graphical representation of the changes in the voice category in girls. Here, the voice changes are presented as related to androstenedione, estrone, body weight, and stage of pubic hair development [6, 12]. Figure 4.41 gives a graphical representation of how the changes in the voice category for boys are related to the falling level of SHBG and the rising level of serum testosterone. There are also correlations between testicle volume, stage of pubic hair development, and Voice Range Profile areas [17,18,19].

Fig. 4.40

Girls’ graphical representation of the parameters with the highest correlation with Voice Range Profile area. Filled circle: Breast development stage 1; open circle: breast (mamma) development stages 2–4; open triangle: breast development stages 5–6; Voice Range Profile area in tones * dB(A) in the diatonic scale. 1–2–3 prepubertal voice category, 5–6–7 postpubertal voice category

Fig. 4.41

Boys’ graphical representation of the stage of pubic hair development, testicle volume, serum testosterone level (total), SHBG, and voice category. (a): beginner, (b and c): soprano, (d): alto, (e): voice in puberty, (f and g): tenor, (h and i): bass as a function of Voice Range Profile area (abscissa). Filled circle: 8.7–12.9 years; open circle: 13–15.9 years; open triangle: 16–19.5 years. Voice Range Profile area in the diatonic scale

Further statistical analysis has underlined the primary results.

To assess the possible influences of local peculiarities in the Copenhagen school system or voice categories, either on voice parameters or on hormonal values, we performed an investigation of members of the professional Leipzig Thomanerchor school: a prepubertal group: boy sopranos, before the change of voice, and a pubertal group: those whose voices had just felt broken. Four subjects were investigated in each group (Fig. 4.42).

Fig. 4.42

Boys’ average Voice Range Profiles with standard deviation for the cohort of four sopranos and of pubertal change groups (mutants) from the Leipzig Thomanerchor school. The hormonal parameters were similar to those of the boys in the Danish school system

The prepubertal soprano group of the Thomanerchor school was comparable to the soprano group of the boys in Copenhagen with respect to the deepest tone and the mean Voice Range Profile areas, and the mean semitone range (Figs. 4.43 and 4.44). In the Voice Range Profiles of the sopranos of the Thomaner groups, there is a smaller range for all high tones in the Voice Range Profiles which could be a quality symbol. The difference is possibly due to a stricter selection of talented boys or a better technical mastery of the voice. As we have shown, voice categories seem not to be related directly to hormonal and pediatric development.

Fig. 4.43

Girls’ average Voice Range Profile and ranges with standard deviation for the lowest and highest tones from a Danish ordinary and high school with choirs, as a function of voice category. The abscissa is divided up into tones, and the frequency in Hz is indicated. The scale of the ordinate is dB(A). One group could not be securely defined during puberty

Fig. 4.44

Boys’ average Voice Range Profiles with standard deviation in a Danish ordinary school and high school, as a function of voice category. The abscissa is divided up into tones, and the frequency in Hz is indicated. The scale of the ordinate is dB(A)

We also performed a pilot study of the Thomaner school boys on an electroglottographic determination of the changes of register. The boys first sang a rising chromatic scale as softly as possible and then as loudly as possible; during this process, the electroglottograms were drawn, and register changes were comparable. With respect to hormonal values, there was no difference between the Leipzig and the Copenhagen subjects [20, 21].

Before the use of Voice Range Profiles was established as a method for simultaneous registration of the tonal and dynamic range of voice, the development of the voice was mostly described by the F0 and total tonal range. Already at an early stage in the history of phoniatrics, investigations of the tonal range for normal school children were carried out [22]. A summary of the results of research into children’s voices was proposed at the Conference of Logopedics and Phoniatrics in 1936 and subsequently performed by Weiss [23]. This summary covers a period of 4000 years and shows that people concerned themselves almost exclusively with boys and eunuchs’ voices. The average age for the change of voice was 14.5 years; the fundamental frequency in continuous speech (F0) for boys dropped by about an octave and for girls by about 1/3 octave. Frank and Sparber and Wendler et al. arrived at comparable results [24, 25].

Blatt discussed the topic of voice training during puberty [26]. Komiyama et al. performed an analysis of Voice Range Profiles during puberty [27]. They did not, however, make any comparisons with other pubertal phenomena and fixed the lower measurement limit for intensity at 60 dB(A). In our investigations, the intensity of the voice during soft singing was significantly lower, and thus the measurements are not comparable.

Meuser and Nieschlag showed that the type of voice for adult men (tenor, baritone, bass) is related to the serum testosterone level [28]. Large and Iwata found differences between the formants, which depended on the voice type of adults [29]. We also believed that a distinction between the types of voice should be made if an exact appraisal of the development of the voice during the time of puberty is to be achieved. But we did not find hormonal-related voice categories in childhood in this study. This could in the future possibly be considered in investigations of the pathology of the voice. Pedersen et al. made a follow-up on voice disorders [30].

Klingholz et al. carried out Voice Range Profiles on members of the Tölzer boys choir; in addition, Konzelmann et al. investigated the Voice Range Profiles of choirboys [31, 32]. A summary of the literature can be found in the thesis of Bühring [33]. Behrendt followed the development of the falsetto register of the boys of the Thomanerchor school until the age of adulthood but did not relate the phenomena to other parameters [34]. Hacki used the shouting voice measurements in Voice Range Profiles and electroglottography [35, 36].

Voice Range Profiles help in the schoolwork also of music teachers and performers. With this method, it is possible to check the results of instruction on the regulation of dynamics (especially during soft singing) and the changes of register more precisely [37,38,39,40]. The voice development in children however cannot be assessed independently of other aspects in pathology [41, 42].

The results are based on a stratified population of girls and boys randomly chosen from the 3rd to 12th school classes. The voice measures of both sexes usable in the school system are given in Fig. 4.45. The hormonal and pubertal pediatric results correspond to the literature, voice being related hereto. The various voice parameters are in detail compared with the measured pubertal development. There are no specific results related to musicality or musical training as shown in the detailed tables of voice categories. But a relation between voice categories in childhood and hormonal and pediatric pubertal values and voice cannot be excluded.

Fig. 4.45

Boys’ and girls’ age-related comparison of the semitone range during continuous speech on the chromatic scale and the lowest tone, the Voice Range Profile area (with tones * dB(A) in the diatonic scale) and the mean fundamental frequency in continuous speech (F0): filled circle: girls; open circle: boys