Speech Articulation of Low-Dose Oral Contraceptive Users *Elis ea Maria Meurer, †Giana Valeria Fagundez Fontoura, ‡Helena von Eye Corleta, and §Edison Capp, *yzPorto Alegre, Brazil, and xHeidelberg, Germany Summary: Objectives. In the female life cycle, hormonal fluctuations may result in impaired verbal efficiency and vocal worsening during the premenstrual phase. Oral contraceptives may interfere with vocal range. Voice, resonance, and articulation variations clarify speech content. To investigate the phonoarticulatory sounds produced by oral contraceptive users aged between 20 and 30 years. Study Design. This is a cross-sectional study. Methods. Our study included four groups of women (n ¼ 66): two groups used low-dose oral contraceptives and two groups did not use any oral contraceptives. Questionnaires and sound records were used. Acoustic analysis was performed using the Computerized Speech Laboratory program, Model 4341 (Kay Elemetrics Corp, Lincoln Park, New Jersey). The statistical analysis of the SPPS database, version 13.0, was performed by means of generalized estimating equation. Results. In the groups that did not use oral contraceptives, sustained vowel tones were more acute in the two phases and cycles of women older than 25 years (w/oOC1, 175 ± 74 to 190 ± 55 Hz; w/oOC2, 194 ± 56 to 210 ± 32 Hz). At the midfollicular phase (Fph) and midluteal phase (Lph) of the two cycles, the speed of the speech was slower in this group (w/oOC1: Fph, 5.3 ± 1.6/s and Lph, 5.4 ± 1.4/s; w/oOC2: Fph, 4.5 ± 1.7/s and Lph, 4.8 ± 1.1/s). In both groups that used oral contraceptives, there was a higher modulation frequency in the sentences when compared with nonusers (OC1, 33 ± 10 Hz; w/oOC1, 28 ± 10 Hz; OC2, 34 ± 10 Hz; w/oOC2, 27 ± 10 Hz). Vocal intensity was closer between the OC1 (62 ± 4 dB), w/oOC1 (61 ± 3 dB), and OC2 (63 ± 4 dB) groups when compared with the w/oOC2 (67 ± 6 dB) group. Conclusions. We demonstrated hormonal influences on speech articulation of contraceptive users and nonusers. Key Words: Menacme–Menstrual cycles–Oral contraceptives–Speech articulation–Isolated emissions–Connected speech–Acoustic records. INTRODUCTION During menacme, the reproductive phase of the female life cycle, the dynamic interaction between the hypothalamicpituitary-ovarian axis and the genital tract establishes menstrual cycles at intervals of around 4 weeks.1 The duration of these cycles depends on the time required for follicular maturation (variable) and the functional duration of the corpus luteum (from 10 to 16 days), thus determining the cyclicity of menstruations. Ovulation occurs between these two phases and, if there is no fertilization, the endometrium lining is shed (menstruation).2 Reproductive capacity may be inhibited by the use of contraceptives containing estrogen and progesterone. These drugs are also used in the treatment of disorders such as premenstrual tension syndrome or premenstrual dysphoric disorder,3,4 as well as acne, hyperandrogenism, increased menstrual bleeding, primary dysmenorrhea, and so forth. Sex steroids have an influence on the motor processes of human verbal expression.5,6 Verbal expression may be a Accepted for publication December 3, 2014. From the *Graduate Program in Health Sciences, Department of Obstetrics and Gynecology, School of Medicine, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil; ySchool of Psychology, Universidade da Regi~ao da Campanha, URCAMP/Bage Campus, Porto Alegre, Brazil; zDepartment of Obstetrics and Gynecology, Hospital de Clı´nicas de Porto Alegre, School of Medicine, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil; and the xExcellence Initiative, Department of Gynecological Endocrinology and Reproductive Medicine, University Hospital Heidelberg, Heidelberg, Germany. Address correspondence and reprint requests to Edison Capp, Servic¸o de Ginecologia e Obstetrı´cia, Hospital de Clı´nicas de Porto Alegre, Rua Ramiro Barcelos, 2400/11 andar, 90035003 Porto Alegre, Rio Grande do Sul, Brazil. E-mail: [email protected] Journal of Voice, Vol. 29, No. 6, pp. 743-750 0892-1997/$36.00 Ó 2015 The Voice Foundation http://dx.doi.org/10.1016/j.jvoice.2014.12.002
fundamental resource of communication in interpersonal relationships and professional activities, such as teaching, radio and television journalism, sales, and entertaining activities (singing and performing arts). Vocal characteristics help to identify people and also provide tips on their health, lifestyle, and emotional state.7,8 Our best vocal performance occurs between 25 and 45 years.9 As a consequence of the synchrony between motor movements and vocal result, we can perceive changes in the voice (human sound waves), resonance (amplification and damping of the vocal sound waves), and articulation (variations of opening and closing of the vocal tract used to produce vowel and consonant sounds). The effects of these variations create suprasegmental features, such as intonation, vocal intensity, and verbal fluency (speed, rhythm, relationship between sound duration and intervals during emissions).10,11 These processes are essential to clarify the content of the speakers’ messages, being dependent on the integration of central and peripheral motor processes.5,12–14 Likewise, specific interconnections between the cortical hubs enable the production of different types of speech such as spontaneous speech, repetitive speech, or verbal reading. Gender and phases of menstrual cycles may also cause variations in the neural processes.15 During the reproductive phase of the female life cycle, hormonal variations may result in reduced verbal efficiency and temporary vocal worsening in the premenstrual or luteal phase.7,16–19 Women with symptoms of premenstrual syndrome showed increased jitter during this phase.20 Oral contraceptives may interfere with the vocal range.3,21,22
744 Emissions of prolonged vowel sounds and sound sequences differ from connected speech in sentences because of the influence of suprasegmental aspects and speakers’ specific characteristics.23,24 With the purpose of addressing these peculiar aspects, the objective of the present study was to investigate acoustic variations of voice and speech in oral contraceptive users aged between 20 and 30 years old during isolated and contextualized emission tasks.
Journal of Voice, Vol. 29, No. 6, 2015
- Women aged between 20 and 24 years, 11 months, and 29 days: B Using low-dose oral contraceptive (containing estrogen 0.03, 0.02, or 0.015 mg): OC1 group; B No oral contraceptive use: w/oOC1; - Women aged between 25 and 30 years: B Using low-dose oral contraceptive (containing estrogen 0.03, 0.02, or 0.015 mg): OC2 group; B No oral contraceptive use: w/oOC2;
vocal intensity (DDKi); (4) five repetitions of the vowel combination /iu/ to study the speaker’s ability to reinforce sound groups derived from the fundamental voice frequency in the nasopharyngeal-oral route (F2max and F2min); and (5) the sentence ‘‘Irei a Gramado nas ferias de inverno.’’ [I’m going to go Gramado over winter break] in six variations, which consisted of neutral, exclamatory, and interrogative (prosodic) intonations and expressing sadness, happiness, and anger (emotional intonations) to record the speakers’ abilities regarding frequency variations (Nf0, Ef0, If0, Sf0, Hf0, Af0, and Ff0) and vocal intonation (Nstd, Estd, Istd, Sstd, Hstd, Astd, and Fstd). This sentence refers to a gaucho tourist spot. Patients were instructed in producing the desired intonational intended as in previous studies.25,26 A demo model of intonation was only provided when demanded by the patient. The emission of this sentence using neutral intonation demonstrated the mean time of production of syllables (Fspeed) and rhythm (Frhythm). Recordings were conducted during the follicular phase (days 5–8) and the premenstrual phase (days 18–23) for women with regular menstrual cycles. These recording sessions were repeated in two menstrual cycles. Oral contraceptive users recorded their vocal production after the third day they started a new pill pack. We used a Sony MZR70-S1 MiniDisc Recorder (Sony Corporation, Tokyo, Japan), a Shure 16A microphone (SHURE Incorporated, Evanston, Illinois), and a Sony Recordable Mini Disc–74 Minutes (Sony Corporation). This cardioid, unidirectional, polar-pattern microphone had frequency response between 50 and 15 000 Hz. In the range between 500 and 8000 Hz, the variation was less than 4 dB with a peak between 6000 and 7000 Hz for high-fidelity recording. The microphone was placed 10 cm from the mouth of the participants, who remained standing during the recording to facilitate their verbal emission processes. The emissions were trained whenever training was requested by the participants to avoid interference of individual differences such as, for example, professional activity. The records were repeated until analyzable patterns were achieved according to the recorder sensor. Continuous analysis of the results was performed using the Motor Speech Profile program, model 4341, coupled to the Computerized Speech Laboratory software (Kay Elemetrics Corp, Lincoln Park, New Jersey). Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) 13.0 for Windows. The final analysis was performed using generalized estimating equations (GEE) considering differences of 30% between the variables for a beta error of 0.10, effect size of 1, and significance level of 0.05.
The next phase consisted of recording the following items: (1) noise level of the recording room for 15 seconds to exclude possible interferences from environmental noise on verbal records; (2) sustained utterance of the vowel ‘‘a’’ to investigate the fundamental voice frequency (/a/F0) and the ability to maintain the synchrony of vocal fold vibration (/a/std); (3) five repetitions of verbal diadochokinesia (DDK) /pataka/ to investigate articulatory agility (DDK speed and DDK rhythm) and
RESULTS All participants completed a questionnaire with the purpose of collecting data on their vocal history and their perception of vocal changes (Table 1). In all groups, the habit of talking a lot was reported (43% SCO1, 68% CO1, 43% CO2, and 82% SCO2). In three groups, the habit of talking fast was reported (43% SCO1, 64% CO1, 48% and CO2).
MATERIAL AND METHODS Study design We conducted a cross-sectional study. Population and sample The present study was approved by the Research Ethics Committee of the Research and Graduate Studies Group of Hospital de Clı´nicas de Porto Alegre (#03–230). This study is ethically and methodologically adequate according to the Guidelines and Regulatory Standards of Research Involving Human Subjects (Resolution 196/96 of the Brazilian National Health Council). The sample consisted of women seen at the gynecology outpatient clinic of the Hospital de Clı´nicas de Porto Alegre. We included nonsmoking women who did not have vocal training in speech or singing and were native speakers of Brazilian Portuguese. None of the participants reported previous hormone therapy or other organic, neurologic, cognitive, or emotional limitations. Sixty-six women of reproductive age were invited to participate in the study. Of these, 21 had never been pregnant before, had regular menstrual cycles, and had not been using hormonal drugs and contraceptives for more than 3 months. The other 45 women had been using oral contraceptives for 3 months or longer. After agreeing to participate and signing the written consent form, participants completed a questionnaire with the purpose of collecting data on their vocal and gynecologic history. Participants were divided into four groups:
a Maria Meurer, et al Elise
745
Speech Articulation of Low-Dose Oral Contraceptive Users
Of the 66 volunteers who participated in our study, 36 were aged between 20 years and 24 years, 11 months, and 29 days and 30 women were aged between 25 years and 29 years, 11 months, and 29 days. The mean age of the groups were as follows: w/oOC1 (n ¼ 14), 22 ± 1.3; OC1 (n ¼ 22), 22 ± 0.8; w/oOC2 (n ¼ 7), 27 ± 1.3; and OC2 (n ¼ 23), 27 ± 1.1 years. Most participants were students (w/oOC1, 93%; OC1, 100%; w/oOC2, 71%; and OC2, 74%) and were enrolled in college programs or had already graduated from college (w/oOC1, 85%; OC1, 77%; w/oOC2, 100%; and OC2, 74%). The number of years of education was similar between the groups (w/oOC1, 15 ± 2; OC1, 16 ± 2; w/oOC2, 17 ± 2; and OC2, 18 ± 2 years). The age of menarche was between 12 and 13 years for all women. The mean duration of contraceptive use was 4 ± 2 years in the OC1 group and 7 ± 3.3 years in the OC2 group. The contraceptives used contained ethinyl estradiol 0.02 mg and gestodene 0.075 mg (OC1, 46% and OC2, 35%) or desogestrel 0.15 mg (OC1, 9% and OC2, 22%). Other contraceptive contained only ethinyl estradiol 0.03 mg (OC1, 4%) or combined with gestodene 0.075 mg (OC2, 17%), desogestrel 0.15 mg (OC1, 4%), levonorgestrel 0.15 mg (OC2, 4%), or drospirenone 3 mg (OC1, 32% and OC2, 9%). In addition, some participants also used contraceptives containing ethinyl estradiol 0.015 mg and gestodene 0.060 mg (OC1, 8% and OC2, 9%). The statistical analyses of acoustic results were performed in three phases. In the first phase, statistically significant interactions in the w/oOC groups between age groups (AG), groups and cycles (GC), groups and phases (GP), and groups, cycles, and phases (GCP) were investigated. During the second phase, the mean value of the acoustic variables was calculated in the follicular and luteal phases because there were no major discrepancies between these phases in the cycles. Thus, we investigated the interactions between groups (G) and GP. In the third phase of analysis, the means between the follicular and luteal phases
of the w/oOC groups were compared with the means of the OC groups for analysis of interactions between AG, contraceptive use (OC), groups and contraceptive use (GOC). For date readings, we grouped isolated characteristics of vocal emissions (/a/F0, /a/std, /iu/F2min and /iu/F2max, DDK speed, DDK rhythm, DDKi) and connected speech (NF0, IF0, EF0, SF0, AF0, HF0; Nstd, Istd, Estd, Sstd, Astd, Hstd, FF0, Fstd, Fspeed, Frhythm). To present the acoustic measures in tables, we used the criterion of increasing it by one when the first digit after the comma was greater than 4, except for speed parameters where the measures were segments per second, and there were standard deviations less than 1. After completing the first model, we found statistically significant interactions in all isolated emissions and in most data of connected speech. The interactions between AG were higher pitch tones in sentences uttered using exclamatory intonation (EF0) and sadness (SF0) by the w/oOC2 group. In the interactions considering only GC, we found less regularity of sustained utterance (/a/std) in the first cycle of the w/oOC1 group. In the w/oOC2 group, there was lower speed (DDK speed) and slower rhythm (DDK rhythm) in diadochokinesia, and we found greater vocal modulation in the means of the sentences (Fstd) in the first cycle. The only statistically significant interaction between GP suggested greater regularity between the phases of the w/oOC1 group regarding the sentence uttered using normal modulation (NF0). In the variables showing combined interactions between GC and between GP, we found a greater variety of minimumsecond formants (F2min) in the luteal phases of the w/oOC2 group, and the same event was detected between the phases of the first cycle in the w/oOC1 group. During the follicular phase of the second cycle of the w/oOC2 groups, we found the lowest pitch values of the maximum-second formants (F2max) in this group. Furthermore, there was greater vocal
TABLE 1. Vocal Profile Aged 25 Years
SCO
CO
SCO
CO
0 (6) 43% (6) 43% (3) 21% (1) 7% (2) 14% (3) 21%
(1) 4% (15) 65% (15) 65% (7) 30% (1) 4% (2) 9% (3) 13%
0 (3) 43% 0 (2) 29% (3) 43% (1) 14%
(1) 5% (19) 86% (10) 45% (8) 36% (4) 18% (2) 9% (6) 27%
(2) 14%
(1) 4% (2) 9%
0 0
(1) 5% 0
0 (2) 14% 0 0
0 (7) 30% 0 0
0 (2) 29% 0 0
0 0 0 0
746
6.3 ± 1 163 ± 25 67 ± 6 5.4 ± 0.5 187 ± 18 67 ± 8 5.2 ± 1.2 210 ± 85 65 ± 6 5.7 ± 0.6 177 ± 18 68 ± 5 6.4 ± 1.7 180 ± 108 62 ± 3 5.8 ± 1.7 200 ± 103 61 ± 3 5.4 ± 1.3 204 ± 91 61 ± 3 6.1 ± 1.1 168 ± 34 62 ± 3 /pataka/DDK speed (/s) /pataka/DDK pattern (ms) /pataka/DDK intensity pattern (dB)
2545 ± 282 2342 ± 366 2486 ± 304 2533 ± 181 2495 ± 187 2500 ± 172 2447 ± 203 /iu/F2 max (Hz)
2390 ± 299
9 ± 15 788 ± 111 /a/variation (Hz) /iu/F2 min (Hz)
6±7 906 ± 186
6±4 863 ± 179
8±9 878 ± 133
8±8 831 ± 117
6±4 901 ± 201
8±8 828 ± 117
7±9 728 ± 136
GP 0.024 GCP 0.043 GC 0.03 GC 0.010 GP 0.001 GC 0.000 GP 0.029 GC 0.012 GC 0.026 G 0.018 GC 0.001 GCP 0.11 194 ± 56 202 ± 64 207 ± 36 210 ± 32 175 ± 74 190 ± 51 185 ± 45 Isolated emissions /a/F0 (Hz)
190 ± 55
Significance Luteal Phase Follicular Phase Luteal Phase Follicular Phase Luteal Phase Follicular Phase Luteal Phase Follicular Phase
Cycle 1
Variable
Cycle 2 Cycle 2
Cycle 1
Aged >25 Years Aged
fundamental resource of communication in interpersonal relationships and professional activities, such as teaching, radio and television journalism, sales, and entertaining activities (singing and performing arts). Vocal characteristics help to identify people and also provide tips on their health, lifestyle, and emotional state.7,8 Our best vocal performance occurs between 25 and 45 years.9 As a consequence of the synchrony between motor movements and vocal result, we can perceive changes in the voice (human sound waves), resonance (amplification and damping of the vocal sound waves), and articulation (variations of opening and closing of the vocal tract used to produce vowel and consonant sounds). The effects of these variations create suprasegmental features, such as intonation, vocal intensity, and verbal fluency (speed, rhythm, relationship between sound duration and intervals during emissions).10,11 These processes are essential to clarify the content of the speakers’ messages, being dependent on the integration of central and peripheral motor processes.5,12–14 Likewise, specific interconnections between the cortical hubs enable the production of different types of speech such as spontaneous speech, repetitive speech, or verbal reading. Gender and phases of menstrual cycles may also cause variations in the neural processes.15 During the reproductive phase of the female life cycle, hormonal variations may result in reduced verbal efficiency and temporary vocal worsening in the premenstrual or luteal phase.7,16–19 Women with symptoms of premenstrual syndrome showed increased jitter during this phase.20 Oral contraceptives may interfere with the vocal range.3,21,22
744 Emissions of prolonged vowel sounds and sound sequences differ from connected speech in sentences because of the influence of suprasegmental aspects and speakers’ specific characteristics.23,24 With the purpose of addressing these peculiar aspects, the objective of the present study was to investigate acoustic variations of voice and speech in oral contraceptive users aged between 20 and 30 years old during isolated and contextualized emission tasks.
Journal of Voice, Vol. 29, No. 6, 2015
- Women aged between 20 and 24 years, 11 months, and 29 days: B Using low-dose oral contraceptive (containing estrogen 0.03, 0.02, or 0.015 mg): OC1 group; B No oral contraceptive use: w/oOC1; - Women aged between 25 and 30 years: B Using low-dose oral contraceptive (containing estrogen 0.03, 0.02, or 0.015 mg): OC2 group; B No oral contraceptive use: w/oOC2;
vocal intensity (DDKi); (4) five repetitions of the vowel combination /iu/ to study the speaker’s ability to reinforce sound groups derived from the fundamental voice frequency in the nasopharyngeal-oral route (F2max and F2min); and (5) the sentence ‘‘Irei a Gramado nas ferias de inverno.’’ [I’m going to go Gramado over winter break] in six variations, which consisted of neutral, exclamatory, and interrogative (prosodic) intonations and expressing sadness, happiness, and anger (emotional intonations) to record the speakers’ abilities regarding frequency variations (Nf0, Ef0, If0, Sf0, Hf0, Af0, and Ff0) and vocal intonation (Nstd, Estd, Istd, Sstd, Hstd, Astd, and Fstd). This sentence refers to a gaucho tourist spot. Patients were instructed in producing the desired intonational intended as in previous studies.25,26 A demo model of intonation was only provided when demanded by the patient. The emission of this sentence using neutral intonation demonstrated the mean time of production of syllables (Fspeed) and rhythm (Frhythm). Recordings were conducted during the follicular phase (days 5–8) and the premenstrual phase (days 18–23) for women with regular menstrual cycles. These recording sessions were repeated in two menstrual cycles. Oral contraceptive users recorded their vocal production after the third day they started a new pill pack. We used a Sony MZR70-S1 MiniDisc Recorder (Sony Corporation, Tokyo, Japan), a Shure 16A microphone (SHURE Incorporated, Evanston, Illinois), and a Sony Recordable Mini Disc–74 Minutes (Sony Corporation). This cardioid, unidirectional, polar-pattern microphone had frequency response between 50 and 15 000 Hz. In the range between 500 and 8000 Hz, the variation was less than 4 dB with a peak between 6000 and 7000 Hz for high-fidelity recording. The microphone was placed 10 cm from the mouth of the participants, who remained standing during the recording to facilitate their verbal emission processes. The emissions were trained whenever training was requested by the participants to avoid interference of individual differences such as, for example, professional activity. The records were repeated until analyzable patterns were achieved according to the recorder sensor. Continuous analysis of the results was performed using the Motor Speech Profile program, model 4341, coupled to the Computerized Speech Laboratory software (Kay Elemetrics Corp, Lincoln Park, New Jersey). Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) 13.0 for Windows. The final analysis was performed using generalized estimating equations (GEE) considering differences of 30% between the variables for a beta error of 0.10, effect size of 1, and significance level of 0.05.
The next phase consisted of recording the following items: (1) noise level of the recording room for 15 seconds to exclude possible interferences from environmental noise on verbal records; (2) sustained utterance of the vowel ‘‘a’’ to investigate the fundamental voice frequency (/a/F0) and the ability to maintain the synchrony of vocal fold vibration (/a/std); (3) five repetitions of verbal diadochokinesia (DDK) /pataka/ to investigate articulatory agility (DDK speed and DDK rhythm) and
RESULTS All participants completed a questionnaire with the purpose of collecting data on their vocal history and their perception of vocal changes (Table 1). In all groups, the habit of talking a lot was reported (43% SCO1, 68% CO1, 43% CO2, and 82% SCO2). In three groups, the habit of talking fast was reported (43% SCO1, 64% CO1, 48% and CO2).
MATERIAL AND METHODS Study design We conducted a cross-sectional study. Population and sample The present study was approved by the Research Ethics Committee of the Research and Graduate Studies Group of Hospital de Clı´nicas de Porto Alegre (#03–230). This study is ethically and methodologically adequate according to the Guidelines and Regulatory Standards of Research Involving Human Subjects (Resolution 196/96 of the Brazilian National Health Council). The sample consisted of women seen at the gynecology outpatient clinic of the Hospital de Clı´nicas de Porto Alegre. We included nonsmoking women who did not have vocal training in speech or singing and were native speakers of Brazilian Portuguese. None of the participants reported previous hormone therapy or other organic, neurologic, cognitive, or emotional limitations. Sixty-six women of reproductive age were invited to participate in the study. Of these, 21 had never been pregnant before, had regular menstrual cycles, and had not been using hormonal drugs and contraceptives for more than 3 months. The other 45 women had been using oral contraceptives for 3 months or longer. After agreeing to participate and signing the written consent form, participants completed a questionnaire with the purpose of collecting data on their vocal and gynecologic history. Participants were divided into four groups:
a Maria Meurer, et al Elise
745
Speech Articulation of Low-Dose Oral Contraceptive Users
Of the 66 volunteers who participated in our study, 36 were aged between 20 years and 24 years, 11 months, and 29 days and 30 women were aged between 25 years and 29 years, 11 months, and 29 days. The mean age of the groups were as follows: w/oOC1 (n ¼ 14), 22 ± 1.3; OC1 (n ¼ 22), 22 ± 0.8; w/oOC2 (n ¼ 7), 27 ± 1.3; and OC2 (n ¼ 23), 27 ± 1.1 years. Most participants were students (w/oOC1, 93%; OC1, 100%; w/oOC2, 71%; and OC2, 74%) and were enrolled in college programs or had already graduated from college (w/oOC1, 85%; OC1, 77%; w/oOC2, 100%; and OC2, 74%). The number of years of education was similar between the groups (w/oOC1, 15 ± 2; OC1, 16 ± 2; w/oOC2, 17 ± 2; and OC2, 18 ± 2 years). The age of menarche was between 12 and 13 years for all women. The mean duration of contraceptive use was 4 ± 2 years in the OC1 group and 7 ± 3.3 years in the OC2 group. The contraceptives used contained ethinyl estradiol 0.02 mg and gestodene 0.075 mg (OC1, 46% and OC2, 35%) or desogestrel 0.15 mg (OC1, 9% and OC2, 22%). Other contraceptive contained only ethinyl estradiol 0.03 mg (OC1, 4%) or combined with gestodene 0.075 mg (OC2, 17%), desogestrel 0.15 mg (OC1, 4%), levonorgestrel 0.15 mg (OC2, 4%), or drospirenone 3 mg (OC1, 32% and OC2, 9%). In addition, some participants also used contraceptives containing ethinyl estradiol 0.015 mg and gestodene 0.060 mg (OC1, 8% and OC2, 9%). The statistical analyses of acoustic results were performed in three phases. In the first phase, statistically significant interactions in the w/oOC groups between age groups (AG), groups and cycles (GC), groups and phases (GP), and groups, cycles, and phases (GCP) were investigated. During the second phase, the mean value of the acoustic variables was calculated in the follicular and luteal phases because there were no major discrepancies between these phases in the cycles. Thus, we investigated the interactions between groups (G) and GP. In the third phase of analysis, the means between the follicular and luteal phases
of the w/oOC groups were compared with the means of the OC groups for analysis of interactions between AG, contraceptive use (OC), groups and contraceptive use (GOC). For date readings, we grouped isolated characteristics of vocal emissions (/a/F0, /a/std, /iu/F2min and /iu/F2max, DDK speed, DDK rhythm, DDKi) and connected speech (NF0, IF0, EF0, SF0, AF0, HF0; Nstd, Istd, Estd, Sstd, Astd, Hstd, FF0, Fstd, Fspeed, Frhythm). To present the acoustic measures in tables, we used the criterion of increasing it by one when the first digit after the comma was greater than 4, except for speed parameters where the measures were segments per second, and there were standard deviations less than 1. After completing the first model, we found statistically significant interactions in all isolated emissions and in most data of connected speech. The interactions between AG were higher pitch tones in sentences uttered using exclamatory intonation (EF0) and sadness (SF0) by the w/oOC2 group. In the interactions considering only GC, we found less regularity of sustained utterance (/a/std) in the first cycle of the w/oOC1 group. In the w/oOC2 group, there was lower speed (DDK speed) and slower rhythm (DDK rhythm) in diadochokinesia, and we found greater vocal modulation in the means of the sentences (Fstd) in the first cycle. The only statistically significant interaction between GP suggested greater regularity between the phases of the w/oOC1 group regarding the sentence uttered using normal modulation (NF0). In the variables showing combined interactions between GC and between GP, we found a greater variety of minimumsecond formants (F2min) in the luteal phases of the w/oOC2 group, and the same event was detected between the phases of the first cycle in the w/oOC1 group. During the follicular phase of the second cycle of the w/oOC2 groups, we found the lowest pitch values of the maximum-second formants (F2max) in this group. Furthermore, there was greater vocal
TABLE 1. Vocal Profile Aged 25 Years
SCO
CO
SCO
CO
0 (6) 43% (6) 43% (3) 21% (1) 7% (2) 14% (3) 21%
(1) 4% (15) 65% (15) 65% (7) 30% (1) 4% (2) 9% (3) 13%
0 (3) 43% 0 (2) 29% (3) 43% (1) 14%
(1) 5% (19) 86% (10) 45% (8) 36% (4) 18% (2) 9% (6) 27%
(2) 14%
(1) 4% (2) 9%
0 0
(1) 5% 0
0 (2) 14% 0 0
0 (7) 30% 0 0
0 (2) 29% 0 0
0 0 0 0
746
6.3 ± 1 163 ± 25 67 ± 6 5.4 ± 0.5 187 ± 18 67 ± 8 5.2 ± 1.2 210 ± 85 65 ± 6 5.7 ± 0.6 177 ± 18 68 ± 5 6.4 ± 1.7 180 ± 108 62 ± 3 5.8 ± 1.7 200 ± 103 61 ± 3 5.4 ± 1.3 204 ± 91 61 ± 3 6.1 ± 1.1 168 ± 34 62 ± 3 /pataka/DDK speed (/s) /pataka/DDK pattern (ms) /pataka/DDK intensity pattern (dB)
2545 ± 282 2342 ± 366 2486 ± 304 2533 ± 181 2495 ± 187 2500 ± 172 2447 ± 203 /iu/F2 max (Hz)
2390 ± 299
9 ± 15 788 ± 111 /a/variation (Hz) /iu/F2 min (Hz)
6±7 906 ± 186
6±4 863 ± 179
8±9 878 ± 133
8±8 831 ± 117
6±4 901 ± 201
8±8 828 ± 117
7±9 728 ± 136
GP 0.024 GCP 0.043 GC 0.03 GC 0.010 GP 0.001 GC 0.000 GP 0.029 GC 0.012 GC 0.026 G 0.018 GC 0.001 GCP 0.11 194 ± 56 202 ± 64 207 ± 36 210 ± 32 175 ± 74 190 ± 51 185 ± 45 Isolated emissions /a/F0 (Hz)
190 ± 55
Significance Luteal Phase Follicular Phase Luteal Phase Follicular Phase Luteal Phase Follicular Phase Luteal Phase Follicular Phase
Cycle 1
Variable
Cycle 2 Cycle 2
Cycle 1
Aged >25 Years Aged
