Percephtaland Motor Skills, 1990, 70, 271-278. O Perceptual and Motor Skills 1990
AUDITORY SENSITIVITY A N D TONE-SEQUENCE REPRODUCTION I N ORAL CONTRACEPTIVE USERS A N D NONUSERS DAHNA M. GRIEZE-JURGELEVICIUS, TAMARA N. CHERNOS, AND J. TIMOTHY PETERSIK'
Ripon College Summary.-In two stages spannlng a single academic year, 10 women using oral contraceptives and 11 not using oral contraceptives were tested on their auditory threshold sensitivity to six frequencies (250, 750, 1500, 3000, 6000, and 8000 Hz) as a function of phases of the menstrual cycle. 11 women (5 users and 6 nonusers) were also tested on their ability to reproduce successively longer tone sequences. Results showed that there was little variation due to phase of the menstrual cycle, with the exception that oral contraceptive users showed lower relative thresholds at low frequencies during the postmenstrual phase. Over-all, oral contraceptive users had hgher absolute and relative thresholds than nonusers, except at 3000 Hz. Although not statistically significant, results on the tone-reproduction test showed an enhanced ability to reproduce tones during the premenstrual phase for women using oral contraceptives. Results are discussed in relation to previous findings.
A growing concern with the psychological effects of the human menstrual cycle has been seen in recent years. For example, significant differences across the menstrual cycle have been found for galvanic skin response, heart rate, and blood pressure (Rosenberg, 1980); for subjective estimation of time intervals (Kopell, Lunde, Clayton, & Moos, 1969); and for EEG (Becker, Creutzfeldt, Schwibbe, & Wuttke, 1982). Given these and other physiological changes associated with the menstrual cycle, it is reasonable to expect sensory and perceptual variations as well. Parlee (1983) summarized much of the earlier research that showed sensitivity on a number of sensory dimensions to be lowest around ovulation. Of interest to the present study is the fact that a number of reports (but not all; see below) have shown cyclic variations in auditory performance (Mehta, Mazumbar, Pathak, & Skandhan, 1977; Parlee, 1983; Tobias, 1965), with the largest departures from the mean occurring near menstruation and ovulation. Relatively little work has been done on the relationships among phases of the menstrual cycle, hormone therapy (i.e., oral contraceptive use), and basic perceptual abilities. I n their study of EEG and physiological changes over the menstrual cycle, Becker, et al. (1982) found that typical changes in mean alpha frequencies were not present in women using oral contraceptives (OCs). Wong and Tong (1974) found that, although both O C users and nonusers showed fluctuations in temporal discrimination over the menstrual cycle, over-all sensitivity was greater among the former. Swanson and 'Address correspondence to J. Timothy Petersik, Department of Psychology, a p o n , WI 54971.
272
D. M. GRIEZE-JURGELEVICIUS, ETAL
Dengerink (1988) compared auditory thresholds in women using and not using OCs and found that users had elevated thresholds during the ovulation and luted phases of the menstrual cycle. However, their testing was limited to tones of 4000 and 6000 Hz, so the data lack generality. The first two of the above studies suggest that O C use minimizes changes due to hormonal fluctuations during the menstrual cycle whereas the third suggests a decrement in auditory detectability. The present experiments concerned auditory sensitivity and the abihty to reproduce successively longer tone sequences as functions of three phases of the menstrual cycle and O C use. I n the case of auditory sensitivity, we were struck by the fact that, although numerous studies have shown sensory abilities vary with the menstrual cycle (Parlee, 1983), at least two showed no effect on auditory sensitivity (Haggard & Gaston, 1978; Schubert, Meyer, & Washer, 1975). No previous research on memory for tones (or tone sequencing) and menstrual-cycle phases was found. Given the above-cited research, however, it seems reasonable to predict that cycle-related variations in memory for tone sequences exist, since memory could vary with arousal and temporal sensitivity. S i d a r l y , if memory for tones is affected by the menstrual cycle, it seems likely that OC users and nonusers would show differences. GENERAL METHOD The experiment testing auditory sensitivity was run in two stages The first stage was conducted during the spring semester of an academic year. At this Lime, auditory sensitivities were measured over one complete menstrual cycle for each of 10 women. To assess the possible effects of seasonal or academic influences, the second stage of the experiment was run during the fall semester of the subsequent academic year. A group of 11 women were tested on auditory sensitivity; a group of men were also tested to control for the possible influence of envitonmental (e.g., seasonal or academic) events that might have occurred during the time of testing. Each man was "yoked" to a woman, i.e., he was tested on the same days as the corresponding woman. And, each woman was tested on the limits of her tone-sequence memory.' Stibjects
AU subjects were between the ages of 19 and 22 yr. None participated i n both stages of the experiment. Sfage 1.-Ten female college students participated, having been solicited through a campus survey seeking women who had regular 28-day menstrual cycles. Five of the subjects were not using OCs and had not for at least three months prior to the onset of the experiment. Five subjects were currently using OCs,' and had been for at least three months prior to the onset of 'Henceforth, we use the terms "tone memory" and "tonal memory" as shorthand terms referring to the tone-sequence reproduction task described below. In our o inion th are reasonable hypothetical constructs regarding the psychological processes require% to reproyuce our successively longer tone sequences (i.e., some storage and recall capacities). Nonetheless, we wish to avoid theoretical polemics regarding the task and its interpretation. The majority of women in this study used Ortho Novum 71717, which contains a primarily progestational compound, norethindrone, in a phase-adjusted dose along with an estrogenic compound, ethinyl esrradiol, in a constant dose, administered for 21 days. One subject used Norinyl 1 + 35 and one used Ovcon 35, each of which contain both norethindrone and ethinyl
AUDITORY SENSITIVITY/ORAL CONTRACEPTIVES
273
the experiment. None of the subjects were using any other medication at the time the experiment was conducted. Stage 2.-Eleven female and 11 male college students participated. Solicitation and selection procedures for the women were the same as in Stage 1. Five of the women were using OCs, and six were not. The men were volunteers from psychology classes. For both stages of the experiment women were given code numbers at the time they volunteered for the experiment. An experimenter recorded whether the subject was or was not using OCs and set up a schedule so that the subject would be tested during the middle of her premenstrual, menstrual, and postmenstrual phases. Subsequent testing was conducted by a diffeent experimenter who had no knowledge of the subject's O C use or phase of the menstrual cycle. Data were analyzed so confidentiality of the women was maintained. Stimuli and Apparatus In borh stages of the experiment, the auditory-threshold test was conducted with a Beltone Portable Audiometer, 1000 series, Model No. 110. The subject wore a set of headphones, and broadband-masking noise was presented to the left ear continuously at 25 dB. Threshold testing was conducted using the right ear only. Testing was conducted in a small cubicle that admitted a minimum of environmental noise. A partition separated the subject from the experimenter. I n Stage 2 of the experiment, tonal memory was tested using an APPLE I1 + microcomputer. A program generated sequences of broad-spectrum tones (low, medium, and high, cenrered on approximately 700 Hz, 1400 Hz, and 2100 Hz, respectively, easily discriminable by all subjects) in a random order. O n the first sequence, one tone was presented; on the second two tones, on the third three tones, etc. Each tone was presented for .5 sec., with .5 sec. between successive tones. Each tone could be reproduced by pressing one of three keys on the computer's keyboard. On each sequence, the subject's cask was to reproduce the original set of tones in the appropriate order. There was no time limit for doing so. Procedure For all tests in both stages of the experiment, all women were tested in the middle of the premenstrual, menstrual, and postmenstrual phases of a single menstrual cycle.' In Stage 2 of the experiment, each of the 11 women was randomly assigned a male subject, who was subsequently tested on the same days. The time of day of testing was held constant as closely as possible. The auditory threshold test was conducted in the following way: six different frequencies (250, 750, 1500, 3000, 6000, and 8000 Hz) were tested. The subject held a response button and was instructed to press it only after hearing rhree consecutive beeps of a tone. Testing was conducted in two consecutive blocks of trials. Each block consisted of a random order of the six test frequencies. On any individual trial, the experimenter initiated a stimulus at 0-dB sound
estradiol in constant doses for 21 days. There was no apparent difference in the data of these subjects in comparison to the rest. We included subjects using different contraceptive preparations and doses in hopes of reaching greater generality regarding the influence of oral contraceptives on auditory sensitivity. 'For testing purposes, the phases of the menstrual cycle were defined in the following ways: premenstrual phase as five days preceding onset of menstruation, menstruation as five days in middle of menstruation, and postmenstrual phase as five days following cessation of menstruation. 'In Stage 2 of the experiment, women were tested on auditory sensitivity a fourth time corresponding to the date of first testing plus 28 days (and so testing was actually conducted over 1 113 cycles) as a check to assess whether thresholds had drifted over t+,e and whether the riming of the menstrual cycle had been accurately assessed. We found these additional" thresholds to be comparable to the first set that had been collected, and so the data were not incorporated into the present analysis.
274
D. M. GRIEZE-JURGELEVICIUS, ET AL.
pressure level and increased the tone's intensity by 5 dB approximately every 5 sec. until the subject responded. The intensity was then lowered by 2 dB for one additional presentation. Threshold was recorded as the last intensity in the sequence to which the subject responded. At least one practice trial (at 1000 Hz) was conducted prior to testing i n each experimental session. In Stage 2 of the experiment, the procedure for testing tonal memory was as follows: before testing, each subject was allowed four or more practice trials to establish the association between [he tone pitches and corresponding keys. A trial consisted of the presentation of a single tone, followed by the subject's depressing the appropriate key on the keyboard. Next, two tones (the original plus a new, second tone) were presented in sequence, again followed by the subject's depressing the corresponding keys in the proper order, The length of the tone sequence was successively increased in this way until the subject made an error. The length (i.e., number of cones) of the last sequence prior to an error was recorded as the subject's score on that trial. Two trials were conducted in each session, and their average was the subject's datum.
RESULTS
Auditory Sensitivity Because the corresponding auditory-sensitivity overlapped considerably for the two stages of the experiment (ruling out effects from season or academic schedule), data from the two stages were combined for analysis. To reduce variability associated with individual differences in auditory sensitivity, data were normalized around the point of lowest threshold in each protocol. I n all cases, this turned out to be 3000 H z . Normalization consisted of subtracting the threshold at 3000 H z from the thresholds at all other frequencies. For all subjects in all conditions, the relative threshold at 3000 Hz was 0 dB. As expected, men showed no significant effects for variables other than frequency (i.e., the cycle phase and condition of OC use of their corresponding-yoked-women subjects) so their data are not considered further. The two panels of Fig. 1 show the resulting relative thresholds for women in each condition of the experiment. As was expected, there was a significant main effect of stimulus frequency (F,,9,= 54.36, p < ,001). As can be seen in the comparison of the two panels of Fig. 1, there was also a significant main effect of whether the subjects used or did not use OCs (F,,,,= 10.57, p < ,005). At all frequencies other than 3000 Hz, relative thresholds are higher for users than for nonusers. Interestingly, when the original threshold data were examined, users had lower absolute thresholds than nonusers at 3000 H z during the premenstrual and menstrual phases of the menstrual cycle by an average of 2 dB. Although small, this effect was consistent and nearly significant (p .05). Examination of Fig. 1
AUDITORY SENSITIVITY/ORAL CONTRACEPTIVES
275
shows clearly that phase of the menstrual cycle had no effect on the sensitivity of nonusers. However, for users there is a suggestion that thresholds are higher in the premenstrual and menstrual phases than in the postmenstrual phase, especially at low frequencies. Although this effect did not reach significance in the over-all data analysis, there was a significant interaction for oral contraceptive use x cycle phase x frequency in the data obtained in Stage 2 of the experiment. Further analysis showed the interaction to be the result of significantly lower thresholds at 250, 750, and 1500 Hz during the postmenstrual phase for users only.
C
a,
.-a>
-
25-
Nonusers
0
CL 15
5
-5
-
: %.,
100
1000
loo00
Frequency ( H z )
.
Fig. 1. Normalized auditory thresholds for oral contraceptive users (upper panel) and nonusers (lower panel). Separate curves show results obtained during three phases of the menstrual cycle: ( * premenstrual cycle, 0 menstrual, and postmenstrual). Error bars show the group standard error during the menstrual phase.
The relative lack of effect of the phase of the menstrual cycle was demonstrated in two other ways. First, we examined the magnitude of the absolute threshold differences between users and nonusers during each phase of the menstrual cycle. The results are shown in the upper panel of Fig. 2 . Two features of the data stand out. First, it is apparent that threshold elevation is not consistently related to cycle phase in any way (although there is some evidence for a premenstrual/postmenstrual difference at low frequencies). Second, it is clear that the change in threshold with oral contraceptive use increases progressively on either side of 3000 Hz.
276
D. M. GRIEZE-JURGELEVICIUS, ETAL.
Our second attempt to find differences due to cycle phase was to examine the standard errors of each group at each frequency and cycle phase. Once again, there was no consistent relationship, although for users there was some indication that variability was greater during the menstrual cycle. The average standard deviations of the two groups, irrespective of cycle phase, are shown in the lower panel of Fig. 2. Two major findings arise from this experiment. First, our data agree with others' that have shown no over-all effect of cycle phase on auditory sensitivity. Second, it appears as though use of oral contraceptives reduces, rather than enhances, auditory detectability, especially for frequencies progressively higher and lower than 3000 Hz.
Frequency
(Hz)
FIG. 2. Upper panel: absolute threshold difference between oral contraceptive usen and nonusers as a function of frequency. Curve parameter is cle phase: (e premenstrual, 0 menstrual, and postmenstrual). Lower panel: group standard yeviations on auditory-sensitivity test users, as a function of frequency. Lower curve parameter condition of contraceptive use: 0 nonusers. Data are averaged over cycle phase.
Memory for Tonex The average lengths of tone sequence correctly repeated for users over cycle phase were: premenstrual 7.9, menstrual 5.0, postmenstrual 6.4. The corresponding means for the nonusers were: premenstrual 5.3, menstrual 7.0, postmenstrual 7.0. There was no average difference between users and non= .15, p > .05). Although the data suggest that users had better users (F,,,
AUDITORY SENSITIVITY/ORAL CONTRACEPTIVES
277
tonal memory than nonusers during the premenstrual phase and that differences between the premenstrual and menstrual phases are reversed in the two groups, the interaction between condition of oral contraceptive use and phase only approached traditional significance (F,, = 2.92, p = . l 3 ) . Further testing with a larger number of subjects could provide a more powerful test. Finally, there was no over-all main effect of cycle phase (F,,, = .27, p > .05). DISCUSSION Our data show that, relative to their lowest threshold (at 3000 H z in this study), there is little variation in threshold for tones spanning the 150to 8000-Hz range as a function of cycle phase in nonusers. Oral contraceptive users show only a modest relative threshold elevation during the premenstrual and menstrual phases at lower auditory frequencies (250 and 750 Hz). The first of these findings is consistent with others that have shown no effect of cycle phase on auditory threshold (Haggard & Gaston, 1978; Schubert, et al., 1975). Nonetheless, some studies have indicated cycle-related differences (Mehta, Mazumbar, Pathak, & Skandhan, 1977; Parlee, 1983; Tobias, 1965). Our interpretation is unique in examining relative rather than absolute threshold. Our finding that there was little relative-threshold change as a function of cycle phase in users is consistent with the work of Swanson and Dengerink (1988), who tested subjects at 4000 and 6000 H z . However, we also find a small decrement in relative threshold during the postmenstrual phase. I n comparing relative threshold elevations of users and nonusers, we find a greater elevation in users, particularly at lower frequencies. I n comparing the absolute thresholds of users and nonusers, we have found that users show elevated thresholds during all phases and at all frequencies except 3000 H z . Our data also give some evidence that the difference between the thresholds of users and nonusers is greatest during the menstrual phase at frequencies below 3000 Hz. Swanson and Dengerink have ~ublishedthe relative hormone levels of oral contraceptive users (Ortho Novum 71717) and nonusers over the phases of the menstrual cycle. A comparison of their information with our data at the lower frequencies (where differences were most apparent) suggests that the smallest differences between users and nonusers occur during the postmenstrual phase when both estorgen and progesterone levels are beginning to rise in both groups. O n the other hand, differences were largest during the menstrual phase during which time both estrogen and progesterone levels were virtually nil among users and low among nonusers. This interpretation is consistent with the suggestion of Miller and Gould (1967) that progesterone raises hearing thresholds and also suggests a role for estrogen. Our data also permit the suggestion that during the premenstrual phase, oral contraceptive users show enhanced ability to reproduce tone sequences
278
D. M. GRIEZE-NRGELEVICIUS, E T A L .
relative to both nonusers and the other two phases of their cycles. However, the modest difference we observed may have simply been due to greater concentration exerted by women who knew they were in the premenstrual phase of their cycles. REFERENCES BECKER,D., CREUTZFELDT,0. D., SCHWBE, M., & W U ~ K EW., (1982) Changes in physiological, EEG, and psychological parameters in women during the spontaneous menstrual cycle and FolLowing oral contraceptives. Psychoneuroendocrinology,7, 75-90. HAGGARD, M., & GASTON, J . B. (1978) Changes in auditory perception in the menstrual cycle. British Iournal of Audiology, 12, 105-118. KOPELL,B. S., LUNDE,D. T., CLAYTON, R. B., & MOOS, R. H. (1969) Variations in some measures of arousal during the menstrual cycle. The Journal of Nervous and Mental Disease, 148, 180-187. MEHTA,Y. B., MAZUMBAR, B. N., P A ~ I A K J. ,D., & SKANDHAN, K. P. (1977) Auditory reaction time during different phases of menstrual cycle. Indian Journal of Medical Sciences, 31, 107-110. PARLEE,M. B. (1983) Menstrual rhythms in sensory processes: a review of fluctuations in vision, olfaction, audicion, taste, and touch. Psychological Bulletin, 93, 539-548. ROSENBERG, E. J . (1980) The relationship between hormonal changes and psychophysiological measures in women. Biological Psychology Bulletin, 6, 46-56. SCHUBERT,G . W., MEYER,R. C., & WASHER,S. H. (1975) Responses to short-duration signals, pre- and postmenses, in subjects using oral contraceptives and subjects not using oral contraceptives. Journal of the American Audiological Sociely, 1, 112-118. SWANSON, S. J., & DENGERINK, H. A. (1988) Changes in pure-tone chresholds and temporary threshold shifts as a Eunction of menstrual cycle and oral contraceptives. Journal of Speech and Hearing Research, 31, 569-574. TOBIAS,J. V. (1965) Consistency of sex differences in binaural-beat perception. International Audiology, 4 , 179-182. WONG,S., 8L TONG, J . E. (1974) Menstrual cycle and contraceptive hormonal effects on temporal discrimination. Perceptual and Motor Skills, 39, 103-108.
Accepted January 29, 1990.
AUDITORY SENSITIVITY A N D TONE-SEQUENCE REPRODUCTION I N ORAL CONTRACEPTIVE USERS A N D NONUSERS DAHNA M. GRIEZE-JURGELEVICIUS, TAMARA N. CHERNOS, AND J. TIMOTHY PETERSIK'
Ripon College Summary.-In two stages spannlng a single academic year, 10 women using oral contraceptives and 11 not using oral contraceptives were tested on their auditory threshold sensitivity to six frequencies (250, 750, 1500, 3000, 6000, and 8000 Hz) as a function of phases of the menstrual cycle. 11 women (5 users and 6 nonusers) were also tested on their ability to reproduce successively longer tone sequences. Results showed that there was little variation due to phase of the menstrual cycle, with the exception that oral contraceptive users showed lower relative thresholds at low frequencies during the postmenstrual phase. Over-all, oral contraceptive users had hgher absolute and relative thresholds than nonusers, except at 3000 Hz. Although not statistically significant, results on the tone-reproduction test showed an enhanced ability to reproduce tones during the premenstrual phase for women using oral contraceptives. Results are discussed in relation to previous findings.
A growing concern with the psychological effects of the human menstrual cycle has been seen in recent years. For example, significant differences across the menstrual cycle have been found for galvanic skin response, heart rate, and blood pressure (Rosenberg, 1980); for subjective estimation of time intervals (Kopell, Lunde, Clayton, & Moos, 1969); and for EEG (Becker, Creutzfeldt, Schwibbe, & Wuttke, 1982). Given these and other physiological changes associated with the menstrual cycle, it is reasonable to expect sensory and perceptual variations as well. Parlee (1983) summarized much of the earlier research that showed sensitivity on a number of sensory dimensions to be lowest around ovulation. Of interest to the present study is the fact that a number of reports (but not all; see below) have shown cyclic variations in auditory performance (Mehta, Mazumbar, Pathak, & Skandhan, 1977; Parlee, 1983; Tobias, 1965), with the largest departures from the mean occurring near menstruation and ovulation. Relatively little work has been done on the relationships among phases of the menstrual cycle, hormone therapy (i.e., oral contraceptive use), and basic perceptual abilities. I n their study of EEG and physiological changes over the menstrual cycle, Becker, et al. (1982) found that typical changes in mean alpha frequencies were not present in women using oral contraceptives (OCs). Wong and Tong (1974) found that, although both O C users and nonusers showed fluctuations in temporal discrimination over the menstrual cycle, over-all sensitivity was greater among the former. Swanson and 'Address correspondence to J. Timothy Petersik, Department of Psychology, a p o n , WI 54971.
272
D. M. GRIEZE-JURGELEVICIUS, ETAL
Dengerink (1988) compared auditory thresholds in women using and not using OCs and found that users had elevated thresholds during the ovulation and luted phases of the menstrual cycle. However, their testing was limited to tones of 4000 and 6000 Hz, so the data lack generality. The first two of the above studies suggest that O C use minimizes changes due to hormonal fluctuations during the menstrual cycle whereas the third suggests a decrement in auditory detectability. The present experiments concerned auditory sensitivity and the abihty to reproduce successively longer tone sequences as functions of three phases of the menstrual cycle and O C use. I n the case of auditory sensitivity, we were struck by the fact that, although numerous studies have shown sensory abilities vary with the menstrual cycle (Parlee, 1983), at least two showed no effect on auditory sensitivity (Haggard & Gaston, 1978; Schubert, Meyer, & Washer, 1975). No previous research on memory for tones (or tone sequencing) and menstrual-cycle phases was found. Given the above-cited research, however, it seems reasonable to predict that cycle-related variations in memory for tone sequences exist, since memory could vary with arousal and temporal sensitivity. S i d a r l y , if memory for tones is affected by the menstrual cycle, it seems likely that OC users and nonusers would show differences. GENERAL METHOD The experiment testing auditory sensitivity was run in two stages The first stage was conducted during the spring semester of an academic year. At this Lime, auditory sensitivities were measured over one complete menstrual cycle for each of 10 women. To assess the possible effects of seasonal or academic influences, the second stage of the experiment was run during the fall semester of the subsequent academic year. A group of 11 women were tested on auditory sensitivity; a group of men were also tested to control for the possible influence of envitonmental (e.g., seasonal or academic) events that might have occurred during the time of testing. Each man was "yoked" to a woman, i.e., he was tested on the same days as the corresponding woman. And, each woman was tested on the limits of her tone-sequence memory.' Stibjects
AU subjects were between the ages of 19 and 22 yr. None participated i n both stages of the experiment. Sfage 1.-Ten female college students participated, having been solicited through a campus survey seeking women who had regular 28-day menstrual cycles. Five of the subjects were not using OCs and had not for at least three months prior to the onset of the experiment. Five subjects were currently using OCs,' and had been for at least three months prior to the onset of 'Henceforth, we use the terms "tone memory" and "tonal memory" as shorthand terms referring to the tone-sequence reproduction task described below. In our o inion th are reasonable hypothetical constructs regarding the psychological processes require% to reproyuce our successively longer tone sequences (i.e., some storage and recall capacities). Nonetheless, we wish to avoid theoretical polemics regarding the task and its interpretation. The majority of women in this study used Ortho Novum 71717, which contains a primarily progestational compound, norethindrone, in a phase-adjusted dose along with an estrogenic compound, ethinyl esrradiol, in a constant dose, administered for 21 days. One subject used Norinyl 1 + 35 and one used Ovcon 35, each of which contain both norethindrone and ethinyl
AUDITORY SENSITIVITY/ORAL CONTRACEPTIVES
273
the experiment. None of the subjects were using any other medication at the time the experiment was conducted. Stage 2.-Eleven female and 11 male college students participated. Solicitation and selection procedures for the women were the same as in Stage 1. Five of the women were using OCs, and six were not. The men were volunteers from psychology classes. For both stages of the experiment women were given code numbers at the time they volunteered for the experiment. An experimenter recorded whether the subject was or was not using OCs and set up a schedule so that the subject would be tested during the middle of her premenstrual, menstrual, and postmenstrual phases. Subsequent testing was conducted by a diffeent experimenter who had no knowledge of the subject's O C use or phase of the menstrual cycle. Data were analyzed so confidentiality of the women was maintained. Stimuli and Apparatus In borh stages of the experiment, the auditory-threshold test was conducted with a Beltone Portable Audiometer, 1000 series, Model No. 110. The subject wore a set of headphones, and broadband-masking noise was presented to the left ear continuously at 25 dB. Threshold testing was conducted using the right ear only. Testing was conducted in a small cubicle that admitted a minimum of environmental noise. A partition separated the subject from the experimenter. I n Stage 2 of the experiment, tonal memory was tested using an APPLE I1 + microcomputer. A program generated sequences of broad-spectrum tones (low, medium, and high, cenrered on approximately 700 Hz, 1400 Hz, and 2100 Hz, respectively, easily discriminable by all subjects) in a random order. O n the first sequence, one tone was presented; on the second two tones, on the third three tones, etc. Each tone was presented for .5 sec., with .5 sec. between successive tones. Each tone could be reproduced by pressing one of three keys on the computer's keyboard. On each sequence, the subject's cask was to reproduce the original set of tones in the appropriate order. There was no time limit for doing so. Procedure For all tests in both stages of the experiment, all women were tested in the middle of the premenstrual, menstrual, and postmenstrual phases of a single menstrual cycle.' In Stage 2 of the experiment, each of the 11 women was randomly assigned a male subject, who was subsequently tested on the same days. The time of day of testing was held constant as closely as possible. The auditory threshold test was conducted in the following way: six different frequencies (250, 750, 1500, 3000, 6000, and 8000 Hz) were tested. The subject held a response button and was instructed to press it only after hearing rhree consecutive beeps of a tone. Testing was conducted in two consecutive blocks of trials. Each block consisted of a random order of the six test frequencies. On any individual trial, the experimenter initiated a stimulus at 0-dB sound
estradiol in constant doses for 21 days. There was no apparent difference in the data of these subjects in comparison to the rest. We included subjects using different contraceptive preparations and doses in hopes of reaching greater generality regarding the influence of oral contraceptives on auditory sensitivity. 'For testing purposes, the phases of the menstrual cycle were defined in the following ways: premenstrual phase as five days preceding onset of menstruation, menstruation as five days in middle of menstruation, and postmenstrual phase as five days following cessation of menstruation. 'In Stage 2 of the experiment, women were tested on auditory sensitivity a fourth time corresponding to the date of first testing plus 28 days (and so testing was actually conducted over 1 113 cycles) as a check to assess whether thresholds had drifted over t+,e and whether the riming of the menstrual cycle had been accurately assessed. We found these additional" thresholds to be comparable to the first set that had been collected, and so the data were not incorporated into the present analysis.
274
D. M. GRIEZE-JURGELEVICIUS, ET AL.
pressure level and increased the tone's intensity by 5 dB approximately every 5 sec. until the subject responded. The intensity was then lowered by 2 dB for one additional presentation. Threshold was recorded as the last intensity in the sequence to which the subject responded. At least one practice trial (at 1000 Hz) was conducted prior to testing i n each experimental session. In Stage 2 of the experiment, the procedure for testing tonal memory was as follows: before testing, each subject was allowed four or more practice trials to establish the association between [he tone pitches and corresponding keys. A trial consisted of the presentation of a single tone, followed by the subject's depressing the appropriate key on the keyboard. Next, two tones (the original plus a new, second tone) were presented in sequence, again followed by the subject's depressing the corresponding keys in the proper order, The length of the tone sequence was successively increased in this way until the subject made an error. The length (i.e., number of cones) of the last sequence prior to an error was recorded as the subject's score on that trial. Two trials were conducted in each session, and their average was the subject's datum.
RESULTS
Auditory Sensitivity Because the corresponding auditory-sensitivity overlapped considerably for the two stages of the experiment (ruling out effects from season or academic schedule), data from the two stages were combined for analysis. To reduce variability associated with individual differences in auditory sensitivity, data were normalized around the point of lowest threshold in each protocol. I n all cases, this turned out to be 3000 H z . Normalization consisted of subtracting the threshold at 3000 H z from the thresholds at all other frequencies. For all subjects in all conditions, the relative threshold at 3000 Hz was 0 dB. As expected, men showed no significant effects for variables other than frequency (i.e., the cycle phase and condition of OC use of their corresponding-yoked-women subjects) so their data are not considered further. The two panels of Fig. 1 show the resulting relative thresholds for women in each condition of the experiment. As was expected, there was a significant main effect of stimulus frequency (F,,9,= 54.36, p < ,001). As can be seen in the comparison of the two panels of Fig. 1, there was also a significant main effect of whether the subjects used or did not use OCs (F,,,,= 10.57, p < ,005). At all frequencies other than 3000 Hz, relative thresholds are higher for users than for nonusers. Interestingly, when the original threshold data were examined, users had lower absolute thresholds than nonusers at 3000 H z during the premenstrual and menstrual phases of the menstrual cycle by an average of 2 dB. Although small, this effect was consistent and nearly significant (p .05). Examination of Fig. 1
AUDITORY SENSITIVITY/ORAL CONTRACEPTIVES
275
shows clearly that phase of the menstrual cycle had no effect on the sensitivity of nonusers. However, for users there is a suggestion that thresholds are higher in the premenstrual and menstrual phases than in the postmenstrual phase, especially at low frequencies. Although this effect did not reach significance in the over-all data analysis, there was a significant interaction for oral contraceptive use x cycle phase x frequency in the data obtained in Stage 2 of the experiment. Further analysis showed the interaction to be the result of significantly lower thresholds at 250, 750, and 1500 Hz during the postmenstrual phase for users only.
C
a,
.-a>
-
25-
Nonusers
0
CL 15
5
-5
-
: %.,
100
1000
loo00
Frequency ( H z )
.
Fig. 1. Normalized auditory thresholds for oral contraceptive users (upper panel) and nonusers (lower panel). Separate curves show results obtained during three phases of the menstrual cycle: ( * premenstrual cycle, 0 menstrual, and postmenstrual). Error bars show the group standard error during the menstrual phase.
The relative lack of effect of the phase of the menstrual cycle was demonstrated in two other ways. First, we examined the magnitude of the absolute threshold differences between users and nonusers during each phase of the menstrual cycle. The results are shown in the upper panel of Fig. 2 . Two features of the data stand out. First, it is apparent that threshold elevation is not consistently related to cycle phase in any way (although there is some evidence for a premenstrual/postmenstrual difference at low frequencies). Second, it is clear that the change in threshold with oral contraceptive use increases progressively on either side of 3000 Hz.
276
D. M. GRIEZE-JURGELEVICIUS, ETAL.
Our second attempt to find differences due to cycle phase was to examine the standard errors of each group at each frequency and cycle phase. Once again, there was no consistent relationship, although for users there was some indication that variability was greater during the menstrual cycle. The average standard deviations of the two groups, irrespective of cycle phase, are shown in the lower panel of Fig. 2. Two major findings arise from this experiment. First, our data agree with others' that have shown no over-all effect of cycle phase on auditory sensitivity. Second, it appears as though use of oral contraceptives reduces, rather than enhances, auditory detectability, especially for frequencies progressively higher and lower than 3000 Hz.
Frequency
(Hz)
FIG. 2. Upper panel: absolute threshold difference between oral contraceptive usen and nonusers as a function of frequency. Curve parameter is cle phase: (e premenstrual, 0 menstrual, and postmenstrual). Lower panel: group standard yeviations on auditory-sensitivity test users, as a function of frequency. Lower curve parameter condition of contraceptive use: 0 nonusers. Data are averaged over cycle phase.
Memory for Tonex The average lengths of tone sequence correctly repeated for users over cycle phase were: premenstrual 7.9, menstrual 5.0, postmenstrual 6.4. The corresponding means for the nonusers were: premenstrual 5.3, menstrual 7.0, postmenstrual 7.0. There was no average difference between users and non= .15, p > .05). Although the data suggest that users had better users (F,,,
AUDITORY SENSITIVITY/ORAL CONTRACEPTIVES
277
tonal memory than nonusers during the premenstrual phase and that differences between the premenstrual and menstrual phases are reversed in the two groups, the interaction between condition of oral contraceptive use and phase only approached traditional significance (F,, = 2.92, p = . l 3 ) . Further testing with a larger number of subjects could provide a more powerful test. Finally, there was no over-all main effect of cycle phase (F,,, = .27, p > .05). DISCUSSION Our data show that, relative to their lowest threshold (at 3000 H z in this study), there is little variation in threshold for tones spanning the 150to 8000-Hz range as a function of cycle phase in nonusers. Oral contraceptive users show only a modest relative threshold elevation during the premenstrual and menstrual phases at lower auditory frequencies (250 and 750 Hz). The first of these findings is consistent with others that have shown no effect of cycle phase on auditory threshold (Haggard & Gaston, 1978; Schubert, et al., 1975). Nonetheless, some studies have indicated cycle-related differences (Mehta, Mazumbar, Pathak, & Skandhan, 1977; Parlee, 1983; Tobias, 1965). Our interpretation is unique in examining relative rather than absolute threshold. Our finding that there was little relative-threshold change as a function of cycle phase in users is consistent with the work of Swanson and Dengerink (1988), who tested subjects at 4000 and 6000 H z . However, we also find a small decrement in relative threshold during the postmenstrual phase. I n comparing relative threshold elevations of users and nonusers, we find a greater elevation in users, particularly at lower frequencies. I n comparing the absolute thresholds of users and nonusers, we have found that users show elevated thresholds during all phases and at all frequencies except 3000 H z . Our data also give some evidence that the difference between the thresholds of users and nonusers is greatest during the menstrual phase at frequencies below 3000 Hz. Swanson and Dengerink have ~ublishedthe relative hormone levels of oral contraceptive users (Ortho Novum 71717) and nonusers over the phases of the menstrual cycle. A comparison of their information with our data at the lower frequencies (where differences were most apparent) suggests that the smallest differences between users and nonusers occur during the postmenstrual phase when both estorgen and progesterone levels are beginning to rise in both groups. O n the other hand, differences were largest during the menstrual phase during which time both estrogen and progesterone levels were virtually nil among users and low among nonusers. This interpretation is consistent with the suggestion of Miller and Gould (1967) that progesterone raises hearing thresholds and also suggests a role for estrogen. Our data also permit the suggestion that during the premenstrual phase, oral contraceptive users show enhanced ability to reproduce tone sequences
278
D. M. GRIEZE-NRGELEVICIUS, E T A L .
relative to both nonusers and the other two phases of their cycles. However, the modest difference we observed may have simply been due to greater concentration exerted by women who knew they were in the premenstrual phase of their cycles. REFERENCES BECKER,D., CREUTZFELDT,0. D., SCHWBE, M., & W U ~ K EW., (1982) Changes in physiological, EEG, and psychological parameters in women during the spontaneous menstrual cycle and FolLowing oral contraceptives. Psychoneuroendocrinology,7, 75-90. HAGGARD, M., & GASTON, J . B. (1978) Changes in auditory perception in the menstrual cycle. British Iournal of Audiology, 12, 105-118. KOPELL,B. S., LUNDE,D. T., CLAYTON, R. B., & MOOS, R. H. (1969) Variations in some measures of arousal during the menstrual cycle. The Journal of Nervous and Mental Disease, 148, 180-187. MEHTA,Y. B., MAZUMBAR, B. N., P A ~ I A K J. ,D., & SKANDHAN, K. P. (1977) Auditory reaction time during different phases of menstrual cycle. Indian Journal of Medical Sciences, 31, 107-110. PARLEE,M. B. (1983) Menstrual rhythms in sensory processes: a review of fluctuations in vision, olfaction, audicion, taste, and touch. Psychological Bulletin, 93, 539-548. ROSENBERG, E. J . (1980) The relationship between hormonal changes and psychophysiological measures in women. Biological Psychology Bulletin, 6, 46-56. SCHUBERT,G . W., MEYER,R. C., & WASHER,S. H. (1975) Responses to short-duration signals, pre- and postmenses, in subjects using oral contraceptives and subjects not using oral contraceptives. Journal of the American Audiological Sociely, 1, 112-118. SWANSON, S. J., & DENGERINK, H. A. (1988) Changes in pure-tone chresholds and temporary threshold shifts as a Eunction of menstrual cycle and oral contraceptives. Journal of Speech and Hearing Research, 31, 569-574. TOBIAS,J. V. (1965) Consistency of sex differences in binaural-beat perception. International Audiology, 4 , 179-182. WONG,S., 8L TONG, J . E. (1974) Menstrual cycle and contraceptive hormonal effects on temporal discrimination. Perceptual and Motor Skills, 39, 103-108.
Accepted January 29, 1990.
