Psychoneuroendocrinology, Vol. 17, No. 5, pp. 497-506,1992
0306-4530/92 $5.00+0.00 @1992 Pcxgtmon Pt~t Ltd.
Printed in Great Britain
VARIATIONS IN MEMORY FUNCTION AND SEX STEROID HORMONES ACROSS THE MENSTRUAL CYCLE SUSANA M. PHILLIPS1 and BARBARAg. SHERWIN2 1Department of Psychology,McCrillUniversity,and 2Department of Obstetrics and Gynecology,Sir Mortimer B. Davis-Jewish General Hospital Montreal, Qudbec, Canada (Received 24 June 1991; in final form 9 March 1992)
SUMMARY Memory,mood, and hormone levels were measured in 25 women during the menstrual and luteal phases of their cycles. Significantlylower visual memory (delayed recall) scores were found during the menstrual phase compared to the luteal phase. No phase differences were found on mood measures or on other memory measures including digit span, paired-associate learning, immediate recall of visual material, and immediate or delayed paragraph recall. The visual memory decrease was most prominent in approximately one-half of the sample and was significantlycorrelated with plasma progesterone in the luteal phase. For all subjects, paragraph recall scores were negatively correlated with free testosterone levels, whereas paired-associate learning was positively correlated with estradiol levels in the luteal phase. These results suggest that changes in memory test performance may be associated with sex steroidlevels,particularlyin some subgroups of women.
INTRODUCTION SEX STEROIDHORMONESinfluence several aspects of central nervous system function, such as neuronal enzyme activity and neurotransmitter uptake and turnover, which in turn may result in behavioral changes (McEwen & Parsons, 1982). Studies examining whether variations in cognitive or perceptual-motor abilities occur coincident with physiological fluctuations in endogenous gonadal hormones across the menstrual cycle have produced conflicting results (Sommer, 1982). For example, among investigators using tests purported to measure various aspects of perceptual-spatial abilities, some reported enhanced performance on a line orientation task (Chiarello et al., 1989) and on a mental rotation task (He et al., 1986) during the preovulatory/ late follicular phase, when estradiol (E2) levels are high, compared to the menstrual phase, when female sex hormone levels are low. In contrast, other investigators have noted significantly higher scores during the menstrual phase compared to the preovulatory phase on other visuospatial tasks, such as a spatial ability composite measure consisting of the Rod-andFrame, Hidden Figures, and Space Relations test scores (Hampson, 1990a). Some of these discrepancies may be related to the fact that the various tasks may measure aspects of cognitive or perceptual-motor function that are differentially responsive or even Address correspondenceand reprint requests to: Dr. Barbara B. Sherwin, Department of Psychology,McGill University, 1205 Dr. Pen_fieldAvenue,Monm~al,Qudbe~, CANADA,H3A 1B1. 497
498
S. M. PHILLIPS& B. B. Srt~WIN
insensitive to hormonal variations. For example, variations across the menstrual cycle generally have not been found on simple reaction time measures (Pierson & Loekhart, 1963; Kopell et al., 1969; Zimmerman & Parlee, 1973; Hunter et al., 1979), whereas decrements during the premenstrual-menstrual phase have emerged in some studies when choice reaction time tasks were administered (Landauer, 1974; Gamberale et al., 1975; Hunter et al., 1975), although not consistently (Zimmerman & Parlee, 1973; Hutt et al., 1980; Slade & Jenner, 1980). Conflicting findings from studies that used the same task may have occurred, in part, secondary to basic methodological shortcomings. Only a few studies have used radioimmunoassay techniques to measure plasma E 2 (Hunter et al., 1979; Hampson, 1990a) and progesterone (P) levels (Wuttke et al., 1976; Komnenich et al., 1978). Instead, some have used less accurate measures of hormone metabolites in the urine (Graham, 1980; Graham & Glasser, 1985), while others have relied on basal body temperature records to estimate hormone levels (Broverman et al., 1981) and/or ovulation (Zimmerman & Parlee, 1973; Hartley et al., 1987), a technique known to be unreliable (Bauman, 1981). In most cases, cycle phase was inferred by counting forward or backward from the day of menstrual onset, without subsequent endocrinological confirmation (e.g., Pierson & Lockhart, 1963; Kopell et al., 1969; Landauer, 1974; Gamberale et al., 1975; Dor-Shav, 1976; Golub, 1976; Hutt et al., 1980; Slade & Jenner, 1980; Hughes, 1983; Ho et al., 1986; Hampson & Kimura, 1988; Chiarello et al., 1989; Heister et al., 1989; Black & Koulis-Chitwood, 1990; Hampson, 1990b; Robinson & Kertzman, 1990). In addition, considerable differences exist between studies in the cycle days used to define a phase. For example, Graham (1980) defined the luteal phase as days 20-25, Hartley et al. (1987) interchangeably referred to testing on the 24 ta day as representing the premenstrual and luteal phase, Heister et al. (1989) defined the luteal phase as days 15-22 and premenstrual phase as days 23-28, and Hughes (1983) simply grouped subjects into one of "three stages": days 1-10, 11-17, or 18-28. As Whalen (1975) pointed out, spurious findings may result if cycle normality is simply assumed without using the currently available technology to accurately measure plasma sex steroids. Even when menstrual history or cycle lengths seem normal, serum radioimmunoassays can reveal abnormal hormonal variations (Abraham et al., 1972; Sherman & Korenman, 1974). In particular, studies of younger women should screen for anovulatory cycles, since there is evidence that only 62% of women aged 20-24 yr actually ovulate during each cycle (Metcalf & Mackenzie, 1980). Because anovulatory cycles are associated with an absence of the luteal phase rise in P levels, failure to detect and eliminate data from such cycles would decrease the probability of detecting changes in test performance across the menstrual cycle (Broverman et al., 1981). Therefore, some studies of small samples (e.g., 13 or less) of younger women might have reported negative findings not only due to reduced statistical power, but also possibly because some subjects may have had anomalous hormone levels. If that were the case, a hormone-behavior relationship that was indeed present might not have emerged. Among studies in which hormone levels were measured, higher test scores have been found during cycle phases in which plasma estrogen or both estrogen and P levels were elevated, including mental arithmetic (Wuttke et al., 1976; Graham, 1980) and articulatory tasks (e.g., speeded color naming) (Graham, 1980; Hampson, 1990a), "automatization" abilities (i.e., simple repetitive tasks) (Komnenich et al., 1978), and manual speed skills (e.g., finger tapping) (Hampson, 1990a). In addition, a positive relationship has been reported between E 2 and scores on "automatization" tasks (Komnenich et al., 1978) and between a P metabolite (pregnanediol) and "perceptual restructuring" tasks (i.e., tasks requiring some inhibition; e.g., Rodand-Frame, Embedded Figures, and mental subtraction tests) (Graham & Glasser, 1985).
MEMORYFIJNCTIONACRO~ THEMENSTRUALCYCLE
499
There is evidence in older posmaenopausal women that memory abilities, in particular, may be influenced by changes in E 2 levels (e.g., Caldwell & Watson, 1952; Sherwin, 1988). Golub (1976) and Hartley et al. (1987) examined whether memory performance varies across the menstrual cycle, but both failed to measure sex steroid hormones, leaving the interpretation of their results unclear. The rationale for the present study was based on the finding that supraphysiologieal E 2 levels were associated with an increase in verbal memory scores, as assessed by immediate paragraph recall scores, in surgically menopausal women treated with estrogen (Phillips & Sherwin, 1992). In that investigation, women who had randomly received placebo postoperatively showed significant declines in paired-associate learning scores in association with their drastically reduced E 2 levels from pre- to postoperatively. Thus, the goal of the present study was to examine whether specific aspects of memory function covary with fluctuating physiological levels of sex steroids, by administering the same memory test battery to young, naturally cycling women during the menstrual and luteal phases of their cycles. It was hypothesized that these same tests of verbal memory would covary with fluctuating E 2 levels across the menstrual cycle, and that higher verbal memory test scores would be obtained during the luteal compared to the menstrual phase. The association between specific areas of memory function and the other sex steroids, P and testosterone (T), also was examined. Plasma hormone levels, measured by radioimmunoassay, were used both to confirm cycle phases and to eliminate subjects with anovulatory or otherwise abnormal cycles. METHODS
Subjects Thirty-five female university students, ages 18 to 35 yr, were recruited through campus newspaper ads and posters. It was required that none of the subjects had taken oral contraceptives for the previous six months and that they were in good general health. Subjects' menstrual cycle lengths were regular, with a mean (+SD) of 28.4 + 1.3 days. Data from four subjects were not entered into subsequent analyses after plasma hormone assays indicated that they had had an anovulatory cycle (i.e., no luteal phase rise in P was detected). Another subject was dropped due to her unusually high E2 level during the menstrual phase. Five subjects did not return for the second test session either because they had scheduling conflicts or they had decided to withdraw from the study. The mean (:I:SD)age of the remaining 25 subjects who provided a complete set of data was 24:t: 5.7 yr.
Procedure First, the women signed a consent form that had been approved by a university ethics committee. They were told that the goal of the study was to examine cognitive functioning during two different phases of the menstrual cycle. Cycle phases were determined by counting forward from the first day of their last period and were subsequently confirmed by plasma hormone assays. The menstrual-phase test session was held on day 3 5 of the cycle; the first 2 days of flow were avoided, due to the possibly distracting effect of physical symptoms. Luteal-phase testing occurred between days 19 and 24, or at least 5 days before the start of the next period, the exact date of which was subsequently obtained by telephone contact. Test sessions were counterbalanced such that half the subjects were tested initially in the menstrual phase, and the other half were tested first in the luteal phase, Subjects were tested at approximately the same time of day in order to control for possible variations in alertness during the day. Within each hour-long test session, a blood sample was drawn, followed by an immediate-recall phase in which the subject was asked to recall each individual memory item immediately after presentation. After all the memory tests were presented, a 20- to 30-rain delay period ensued, during which the subject completed a series of questionnaires, including those listed below. During the delayed-recall phase that followed, the subject was asked to recall test items from each of the first three subtests listed below.
500
S. M. PHILLIPS~r B. B. SHERWIN
Test~ Wechsler Memory_ Scale (WMS) (Wechsler, 1945; Stone et al., 1946; with modification of Russell, 1975): The two forms of the WMS were counterbalanced across the two test sessions such that half the subjects rea~eivedForm I first and half received Form II. The following four subtests were administered: a) Logical Memory/Paragraph Recall: The subject listened to a short paragraph of about 4 - 5 sentences and then was asked to verbally recall what she remembered from the passage. Two paragraphs were presented once. b) Visual Reproduction: The subject was shown a design for 10 sec and then asked to draw what she remembered seeing on the card. Three design cards were presented once each. The subjects were given a separate sheet of paper for each drawing to minimize visual exposure to designs presented first. c) Associate Learning: The subject listened to a list of 10 word-pairs (paired-associates) presented in a random order, six of which were clearly related (easy-associate; e.g., fruit-apple) and four of which had no apparent relationship (hard-associate; e.g., necktie-cracker). After hearing the list, the subject was asked to recall the word that was paired (e.g., apple) with each cue word (e.g., fruit). Three trials of presentation and immediate recall were conducted. After the delay period, one final recall trial was administered. Two points were given for a correctly recalled hard-associate and one point for an easy-associate. d) Digit Span: The subject listened to a series of numbers and was instructed to verbally repeat them in the given order immediately afterwards. In order to ensure approximately identical test conditions across subjects, a manual of exact procedures was prepared, verbal test items were presented to subjects with a tape recorder, and verbal responses were recorded on tape and later transcribed. The protocols were scored independently by three coders who were blind to group assignment. Wechsler's (1945) original scoring criteria for the Logical Memory and Visual Reproduction subtests have been criticized as being poorly defined, which may lead to significant scoring discrepancies among coders (e.g., Mitchell, 1987). An improved scoring system therefore was developed for this study with well-defined criteria and high inter-rater reliability (r = 0.98 for Logical Memory; r = 0.96 for Visual Reproduction). Alternate-form analyses indicated the two forms were equivalent under these scoring criteria (Phillips et al., unpublished data). General Health and Menstrual Cycle Ouestionnalres: The subjects completed a general questionnaire whose items covered their menstrual history (e.g., past use of oral contraceptives, characteristics of their menstrual cycle, etc.), their health, and drugs used recently. These data were collected during the 30-rain delay between immediate and delayed memory testing. Mulfinle Affect Adjective Cheek List (MAA(:L) (Zuckerman & Lubin, 1965): In this brief, self-administered inventory, the subject indicated which of 132 adjectives best reflected her present mood state during each test session. This instrument yields scores for current levels of anxiety, depression, and hostility and is valid and reliable. It has been used previously in psychoendoerine studies to measure the relationship between affect and concurrent hormone levels (Sherwin & Gelfand, 1985). Plasma Hormone Assays During each test session, 10 ml blood were collected by venipuncture into heparinized Vacutainer tubes. The samples were immediately centrifuged and the plasma stored at -20 ° C until completion of data collection. Plasma steroid hormone levels were determined in duplicate by radioimmunoassay (RIA). P was measured with the Coat-A-Count progesterone kit and free T by the Coat-A-Count testosterone kit (Diagnostic Products Corporation, Los Angeles, CA, USA), both of which are solid phase 1251RIAs. E2 was measured by a double antibody technique with the Gamma-B Direct Oestradiol RIA kit (IDS Limited, Boldon, UK). The antisera used in these assays were highly specific for their respective hormones. Statistical Analyses The data were analyzed with the SPSS-X multivariate analysis of variance (MANOVA) statistical package. One-way repeated measures MANOVAs were carried out with cycle phase as the independent factor. If a significant multivariate F statistic (p 0.05). Mean plasma levels of all sex steroids fell within the cycle phase norms for the radioimmunoassay procedures used. Memory Measures Table II shows that scores on the delayed recall of the visual reproduction subtest were significantly lower in the menstrual phase compared to the luteal phase F1~4=6.9, p
0306-4530/92 $5.00+0.00 @1992 Pcxgtmon Pt~t Ltd.
Printed in Great Britain
VARIATIONS IN MEMORY FUNCTION AND SEX STEROID HORMONES ACROSS THE MENSTRUAL CYCLE SUSANA M. PHILLIPS1 and BARBARAg. SHERWIN2 1Department of Psychology,McCrillUniversity,and 2Department of Obstetrics and Gynecology,Sir Mortimer B. Davis-Jewish General Hospital Montreal, Qudbec, Canada (Received 24 June 1991; in final form 9 March 1992)
SUMMARY Memory,mood, and hormone levels were measured in 25 women during the menstrual and luteal phases of their cycles. Significantlylower visual memory (delayed recall) scores were found during the menstrual phase compared to the luteal phase. No phase differences were found on mood measures or on other memory measures including digit span, paired-associate learning, immediate recall of visual material, and immediate or delayed paragraph recall. The visual memory decrease was most prominent in approximately one-half of the sample and was significantlycorrelated with plasma progesterone in the luteal phase. For all subjects, paragraph recall scores were negatively correlated with free testosterone levels, whereas paired-associate learning was positively correlated with estradiol levels in the luteal phase. These results suggest that changes in memory test performance may be associated with sex steroidlevels,particularlyin some subgroups of women.
INTRODUCTION SEX STEROIDHORMONESinfluence several aspects of central nervous system function, such as neuronal enzyme activity and neurotransmitter uptake and turnover, which in turn may result in behavioral changes (McEwen & Parsons, 1982). Studies examining whether variations in cognitive or perceptual-motor abilities occur coincident with physiological fluctuations in endogenous gonadal hormones across the menstrual cycle have produced conflicting results (Sommer, 1982). For example, among investigators using tests purported to measure various aspects of perceptual-spatial abilities, some reported enhanced performance on a line orientation task (Chiarello et al., 1989) and on a mental rotation task (He et al., 1986) during the preovulatory/ late follicular phase, when estradiol (E2) levels are high, compared to the menstrual phase, when female sex hormone levels are low. In contrast, other investigators have noted significantly higher scores during the menstrual phase compared to the preovulatory phase on other visuospatial tasks, such as a spatial ability composite measure consisting of the Rod-andFrame, Hidden Figures, and Space Relations test scores (Hampson, 1990a). Some of these discrepancies may be related to the fact that the various tasks may measure aspects of cognitive or perceptual-motor function that are differentially responsive or even Address correspondenceand reprint requests to: Dr. Barbara B. Sherwin, Department of Psychology,McGill University, 1205 Dr. Pen_fieldAvenue,Monm~al,Qudbe~, CANADA,H3A 1B1. 497
498
S. M. PHILLIPS& B. B. Srt~WIN
insensitive to hormonal variations. For example, variations across the menstrual cycle generally have not been found on simple reaction time measures (Pierson & Loekhart, 1963; Kopell et al., 1969; Zimmerman & Parlee, 1973; Hunter et al., 1979), whereas decrements during the premenstrual-menstrual phase have emerged in some studies when choice reaction time tasks were administered (Landauer, 1974; Gamberale et al., 1975; Hunter et al., 1975), although not consistently (Zimmerman & Parlee, 1973; Hutt et al., 1980; Slade & Jenner, 1980). Conflicting findings from studies that used the same task may have occurred, in part, secondary to basic methodological shortcomings. Only a few studies have used radioimmunoassay techniques to measure plasma E 2 (Hunter et al., 1979; Hampson, 1990a) and progesterone (P) levels (Wuttke et al., 1976; Komnenich et al., 1978). Instead, some have used less accurate measures of hormone metabolites in the urine (Graham, 1980; Graham & Glasser, 1985), while others have relied on basal body temperature records to estimate hormone levels (Broverman et al., 1981) and/or ovulation (Zimmerman & Parlee, 1973; Hartley et al., 1987), a technique known to be unreliable (Bauman, 1981). In most cases, cycle phase was inferred by counting forward or backward from the day of menstrual onset, without subsequent endocrinological confirmation (e.g., Pierson & Lockhart, 1963; Kopell et al., 1969; Landauer, 1974; Gamberale et al., 1975; Dor-Shav, 1976; Golub, 1976; Hutt et al., 1980; Slade & Jenner, 1980; Hughes, 1983; Ho et al., 1986; Hampson & Kimura, 1988; Chiarello et al., 1989; Heister et al., 1989; Black & Koulis-Chitwood, 1990; Hampson, 1990b; Robinson & Kertzman, 1990). In addition, considerable differences exist between studies in the cycle days used to define a phase. For example, Graham (1980) defined the luteal phase as days 20-25, Hartley et al. (1987) interchangeably referred to testing on the 24 ta day as representing the premenstrual and luteal phase, Heister et al. (1989) defined the luteal phase as days 15-22 and premenstrual phase as days 23-28, and Hughes (1983) simply grouped subjects into one of "three stages": days 1-10, 11-17, or 18-28. As Whalen (1975) pointed out, spurious findings may result if cycle normality is simply assumed without using the currently available technology to accurately measure plasma sex steroids. Even when menstrual history or cycle lengths seem normal, serum radioimmunoassays can reveal abnormal hormonal variations (Abraham et al., 1972; Sherman & Korenman, 1974). In particular, studies of younger women should screen for anovulatory cycles, since there is evidence that only 62% of women aged 20-24 yr actually ovulate during each cycle (Metcalf & Mackenzie, 1980). Because anovulatory cycles are associated with an absence of the luteal phase rise in P levels, failure to detect and eliminate data from such cycles would decrease the probability of detecting changes in test performance across the menstrual cycle (Broverman et al., 1981). Therefore, some studies of small samples (e.g., 13 or less) of younger women might have reported negative findings not only due to reduced statistical power, but also possibly because some subjects may have had anomalous hormone levels. If that were the case, a hormone-behavior relationship that was indeed present might not have emerged. Among studies in which hormone levels were measured, higher test scores have been found during cycle phases in which plasma estrogen or both estrogen and P levels were elevated, including mental arithmetic (Wuttke et al., 1976; Graham, 1980) and articulatory tasks (e.g., speeded color naming) (Graham, 1980; Hampson, 1990a), "automatization" abilities (i.e., simple repetitive tasks) (Komnenich et al., 1978), and manual speed skills (e.g., finger tapping) (Hampson, 1990a). In addition, a positive relationship has been reported between E 2 and scores on "automatization" tasks (Komnenich et al., 1978) and between a P metabolite (pregnanediol) and "perceptual restructuring" tasks (i.e., tasks requiring some inhibition; e.g., Rodand-Frame, Embedded Figures, and mental subtraction tests) (Graham & Glasser, 1985).
MEMORYFIJNCTIONACRO~ THEMENSTRUALCYCLE
499
There is evidence in older posmaenopausal women that memory abilities, in particular, may be influenced by changes in E 2 levels (e.g., Caldwell & Watson, 1952; Sherwin, 1988). Golub (1976) and Hartley et al. (1987) examined whether memory performance varies across the menstrual cycle, but both failed to measure sex steroid hormones, leaving the interpretation of their results unclear. The rationale for the present study was based on the finding that supraphysiologieal E 2 levels were associated with an increase in verbal memory scores, as assessed by immediate paragraph recall scores, in surgically menopausal women treated with estrogen (Phillips & Sherwin, 1992). In that investigation, women who had randomly received placebo postoperatively showed significant declines in paired-associate learning scores in association with their drastically reduced E 2 levels from pre- to postoperatively. Thus, the goal of the present study was to examine whether specific aspects of memory function covary with fluctuating physiological levels of sex steroids, by administering the same memory test battery to young, naturally cycling women during the menstrual and luteal phases of their cycles. It was hypothesized that these same tests of verbal memory would covary with fluctuating E 2 levels across the menstrual cycle, and that higher verbal memory test scores would be obtained during the luteal compared to the menstrual phase. The association between specific areas of memory function and the other sex steroids, P and testosterone (T), also was examined. Plasma hormone levels, measured by radioimmunoassay, were used both to confirm cycle phases and to eliminate subjects with anovulatory or otherwise abnormal cycles. METHODS
Subjects Thirty-five female university students, ages 18 to 35 yr, were recruited through campus newspaper ads and posters. It was required that none of the subjects had taken oral contraceptives for the previous six months and that they were in good general health. Subjects' menstrual cycle lengths were regular, with a mean (+SD) of 28.4 + 1.3 days. Data from four subjects were not entered into subsequent analyses after plasma hormone assays indicated that they had had an anovulatory cycle (i.e., no luteal phase rise in P was detected). Another subject was dropped due to her unusually high E2 level during the menstrual phase. Five subjects did not return for the second test session either because they had scheduling conflicts or they had decided to withdraw from the study. The mean (:I:SD)age of the remaining 25 subjects who provided a complete set of data was 24:t: 5.7 yr.
Procedure First, the women signed a consent form that had been approved by a university ethics committee. They were told that the goal of the study was to examine cognitive functioning during two different phases of the menstrual cycle. Cycle phases were determined by counting forward from the first day of their last period and were subsequently confirmed by plasma hormone assays. The menstrual-phase test session was held on day 3 5 of the cycle; the first 2 days of flow were avoided, due to the possibly distracting effect of physical symptoms. Luteal-phase testing occurred between days 19 and 24, or at least 5 days before the start of the next period, the exact date of which was subsequently obtained by telephone contact. Test sessions were counterbalanced such that half the subjects were tested initially in the menstrual phase, and the other half were tested first in the luteal phase, Subjects were tested at approximately the same time of day in order to control for possible variations in alertness during the day. Within each hour-long test session, a blood sample was drawn, followed by an immediate-recall phase in which the subject was asked to recall each individual memory item immediately after presentation. After all the memory tests were presented, a 20- to 30-rain delay period ensued, during which the subject completed a series of questionnaires, including those listed below. During the delayed-recall phase that followed, the subject was asked to recall test items from each of the first three subtests listed below.
500
S. M. PHILLIPS~r B. B. SHERWIN
Test~ Wechsler Memory_ Scale (WMS) (Wechsler, 1945; Stone et al., 1946; with modification of Russell, 1975): The two forms of the WMS were counterbalanced across the two test sessions such that half the subjects rea~eivedForm I first and half received Form II. The following four subtests were administered: a) Logical Memory/Paragraph Recall: The subject listened to a short paragraph of about 4 - 5 sentences and then was asked to verbally recall what she remembered from the passage. Two paragraphs were presented once. b) Visual Reproduction: The subject was shown a design for 10 sec and then asked to draw what she remembered seeing on the card. Three design cards were presented once each. The subjects were given a separate sheet of paper for each drawing to minimize visual exposure to designs presented first. c) Associate Learning: The subject listened to a list of 10 word-pairs (paired-associates) presented in a random order, six of which were clearly related (easy-associate; e.g., fruit-apple) and four of which had no apparent relationship (hard-associate; e.g., necktie-cracker). After hearing the list, the subject was asked to recall the word that was paired (e.g., apple) with each cue word (e.g., fruit). Three trials of presentation and immediate recall were conducted. After the delay period, one final recall trial was administered. Two points were given for a correctly recalled hard-associate and one point for an easy-associate. d) Digit Span: The subject listened to a series of numbers and was instructed to verbally repeat them in the given order immediately afterwards. In order to ensure approximately identical test conditions across subjects, a manual of exact procedures was prepared, verbal test items were presented to subjects with a tape recorder, and verbal responses were recorded on tape and later transcribed. The protocols were scored independently by three coders who were blind to group assignment. Wechsler's (1945) original scoring criteria for the Logical Memory and Visual Reproduction subtests have been criticized as being poorly defined, which may lead to significant scoring discrepancies among coders (e.g., Mitchell, 1987). An improved scoring system therefore was developed for this study with well-defined criteria and high inter-rater reliability (r = 0.98 for Logical Memory; r = 0.96 for Visual Reproduction). Alternate-form analyses indicated the two forms were equivalent under these scoring criteria (Phillips et al., unpublished data). General Health and Menstrual Cycle Ouestionnalres: The subjects completed a general questionnaire whose items covered their menstrual history (e.g., past use of oral contraceptives, characteristics of their menstrual cycle, etc.), their health, and drugs used recently. These data were collected during the 30-rain delay between immediate and delayed memory testing. Mulfinle Affect Adjective Cheek List (MAA(:L) (Zuckerman & Lubin, 1965): In this brief, self-administered inventory, the subject indicated which of 132 adjectives best reflected her present mood state during each test session. This instrument yields scores for current levels of anxiety, depression, and hostility and is valid and reliable. It has been used previously in psychoendoerine studies to measure the relationship between affect and concurrent hormone levels (Sherwin & Gelfand, 1985). Plasma Hormone Assays During each test session, 10 ml blood were collected by venipuncture into heparinized Vacutainer tubes. The samples were immediately centrifuged and the plasma stored at -20 ° C until completion of data collection. Plasma steroid hormone levels were determined in duplicate by radioimmunoassay (RIA). P was measured with the Coat-A-Count progesterone kit and free T by the Coat-A-Count testosterone kit (Diagnostic Products Corporation, Los Angeles, CA, USA), both of which are solid phase 1251RIAs. E2 was measured by a double antibody technique with the Gamma-B Direct Oestradiol RIA kit (IDS Limited, Boldon, UK). The antisera used in these assays were highly specific for their respective hormones. Statistical Analyses The data were analyzed with the SPSS-X multivariate analysis of variance (MANOVA) statistical package. One-way repeated measures MANOVAs were carried out with cycle phase as the independent factor. If a significant multivariate F statistic (p 0.05). Mean plasma levels of all sex steroids fell within the cycle phase norms for the radioimmunoassay procedures used. Memory Measures Table II shows that scores on the delayed recall of the visual reproduction subtest were significantly lower in the menstrual phase compared to the luteal phase F1~4=6.9, p
