Visual evoked potential and electroencephalogram of healthy females during the menstrual cycle Yasuhiro Kaneda, Takumi Ikuta, Hiroshi Nakayama, Kouichi Kagawa and Noriko Furuta Department of Neuropsychiatry, The University of Tokushima School of Medicine, Tokushima, Japan Abstract:Flash visual evoked potential (VEP) and electroencephalogram (EEG) changes during the menstrual cycle were studied using healthy females having regular menstruation, with21at the follicular phase (FP) and23at the luteal phase (LP). The following results were obtained. (1)The waveforms of Group Mean VEPs of both groups had approximately similar triphasic contours, consisting of16components of P1-N8up to500msec of latency. (2)Latencies tended to be longer in LP. (3)Interpeak amplitudes tended to be larger in LP, and one VEP interpeak amplitude (P5-N7) of long latency component was significantly larger at LP after eliminating the effect of body height by ANCOVA for2CH. (4)Quantitative analysis of EEGs between FP and LP resulted in a tendency for increased α, and decreased β power % at LP. Since estrogen increases the VEP amplitude, and decreases the VEP latency and the α activity of EEGs, the large VEP amplitude, the tendency for prolonged VEP latency, and the tendency for increased α power % at LP observed in this study indicate that the VEP amplitude at LP reflects the effect of estrogen, and that the VEP latency and EEGs at LP reflect the effect of progesterone. J. Med. Invest. 44:41-46, 1997 Keywords:visual evoked potential, electroencephalogram, menstrual cycle, estrogen, progesterone INTRODUCTION Visual evoked potentials (VEPs) provide information regarding the central nervous system (CNS) including the visual system and brain excitability (10), and are often used in clinical neurophysiology (13, 23, 30). Variabilities in VEP amplitudes and latencies are the main criteria of pathology. VEP latencies are shorter and amplitudes are larger in females than in males in the normative data (1, 4, 15, 27, 33). In addition, in our previous study using100females and100males, the same result was verified not only in the shorter or middle latency components (up to240msec), but also in the longer latency components (up to500msec) (15). In females, the mood (11, 34) or psychometric perfor-mance (6) changes in normal women has been reported during normal menstruation, and some neurophysiologic studies have demonstrated electroencephalogram (EEG) changes with the menstrual cycle (6, 35). However, there have been few reports of VEP changes during the menstrual cycle (22, 29), and they related only to the shorter or middle latency components (up to240msec). Therefore, in the present study, the statistical differences in VEPs of females, including those of later components (up to500msec) were studied between FP and LP, using44subjects out of100having regular menstrual cycles. MATERIALS AND METHODS 1. Materials The subjects in the present study were 44 female students (paid volunteers) who had regular menstrual cycles. Of them, twenty-one were at the follicular phase (FP) (generally day1-13of a28-day cycle, when estrogen alone is raised) at the time of recording, and the remaining23were at the luteal phase (LP) (generally day15-28, when estrogen and progesterone are raised). Average body height was significantly greater at FP than at LP, but average body weight or age was not (t-test, p<0.05) (Table1). All subjects had no neuropsychiatric history, visual disorders nor ongoing medication. EEGs of all subjects were within the normal range. One out of21females at FP and2out of23females at LP were left-handed. 2. Recording methods of VEPs and EEGs Recording electrodes were placed on the scalp according to the international10-20system, in preference to linked ear lobes. Electrode impedance was kept below5kΩ. While recording, the subjects laid, with eyes closed, in a reclining chair at70degrees in a dark shielded room, at 24-25°C. The consciousness level of subjects was monitored by EEG. Flash stimuli from the XENON tube of acoustically shielded Retinogragh MSP-2R (Nihon Kohden) were presented from a distance of30cm once every5sec. The intensity of the flash was0.6J. EEGs containing VEPs were derived from the two derivations, 2CH (O1→A1+2) and 5CH (O1→CZ) (28), amplified by Preamplifier AB-622M with0.1sec time constant, with100Hz high range filter and without using a hum eliminating filter. Data were recorded by Data Recorder RX-50L (TEAC) onto magnetic tapes with trigger pulses synchronized to the flash stimuli. Reproducing EEGs and trigger pulses from the tape by Data Recorder RX-50L, VEPs derived through the two derivations of each subject were recorded by averaging100single responses by ATAC-210 (1024address×220 bit) for1024msec of analysis time. 3. Methods of data processing 3·1 Data processing of VEPs VEPs were processed by subsequent computation. Each VEP was adjusted by least squares so that sums of squares of instantaneous values from the baseline were minimum. 3·1·1 Group Mean VEPs The group Mean VEPs for each group were obtained for2CH and5CH, respectively. The components (P1-P8, N1-N8) in Group Mean VEPs were identified on a CRT monitor. The differences in the waveforms, latencies and interpeak amplitudes of Group Mean VEPs between FP and LP were compared for each derivation. 3·1·2 Individual VEPs Referring to the components in Group Mean VEPs, those in Individual VEPs were identified on a CRT monitor, respectively for each derivation. The differences in the latencies and interpeak amplitudes of Individual VEPs between FP and LP were compared and tested by t-test, and by analysis of covariance (ANCOVA) with reference to body height for each derivation. 3·2 Data processing of EEGs Absolute power values of EEGs were calculated with the program (QP-130B "RHYTHM") by quantitative frequency analysis. Eight epochs (32sec) with 128Hz sampling rate and 512points were analyzed every 0.25Hz by fast Fourier transformation. Through division into 6 frequency bands, δ (2.0-3.75Hz) and θ (4.0-7.75Hz), α1(8.0-9.75Hz), α2(10.00-12.75Hz), β1(13.00-19.75Hz), β2 (20.00-30.00Hz), absolute amplitudes and power % of each band were calculated. Then, the average absolute amplitudes and average power % were compared and tested between FP and LP by t-test for each derivation. RESULTS 1. Waveforms of Group Mean VEPs of each group The waveforms of Group Mean VEPs of each group were approximately triphasic up to500msec in latency for both2CH and5CH, and consisted of16components, P1,N1, P2, N2, P3, N3, P4, N4, P5, N5, P6, N6, P7, N7,P8and N8 (Fig.1). The highest positive peak of each Group Mean VEP was P5in both FP and LP for both derivations. The lowest negative peaks were N3in both FP and LP for2CH, N8at FP and N7at LP for5CH. The maximal interpeak amplitudes of Group Mean VEPs (2CH:N3-P5, 5CH:P5-N7) were larger at LP than in FP. 2. Component analysis for Individual VEPs 2·1 Waveforms of Individual VEPs The waveforms of Individual VEPs were similar to those of Group Mean VEPs derived from the same derivations. The appearance rates of each component varied from67 to 100 % (Table2). Five peaks (N3, P5, N7-N8) at FP, and eight peaks (P2, N3, P5, N6-N8) at LP for2CH, six peaks (P2-N3, P5and N8) at FP, and eight peaks (N2-N3, P5, P7-N8) at LP for 5CH were consistently identified. The prominent interpeak amplitudes of VEPs at LP for2CH (N3-P5) and for 5CH (P5-N7) were7 % and 9 % larger, respectively, than those at FP (N3-P5for 2CH, P4-N8 for5CH). 2·2 Differences in Individual VEP components between FP and LP 2·2·1 Differences in Individual VEP latencies between FP and LP The VEP latencies of9components were shorter at LP than at FP for 2CH (Table2). The latency of P3 was significantly shorter at LP than at FP, and that of P2was significantly longer at LP for2CH by t-test (p<0.05). After eliminating the effect of body height, these differences were not significant by ANCOVA. Although eleven latencies were longer at LP than at FP for5CH, the differences were not significant by t-test nor ANCOVA (Table2). 2·2·2 Differences in Individual VEP interpeak amplitudes between FP and LP Thirteen out of 23 VEP interpeak amplitudes including the prominent interpeak amplitude were larger at LP than at FP for2CH (Table3). The interpeak amplitude of P5-N7was significantly larger at LP than at FP for2CH by t-test (p<0.01), even after eliminating the effect of body height by ANCOVA (p<0.05). Although eighteen interpeak amplitudes including the prominent interpeak amplitude were larger at LP for5CH, the difference was not significant by t-test nor ANCOVA (Table3). 3. Differences in EEGs by quantitative frequency analysis between FP and LP 3·1 EEG absolute amplitudes and their differences between FP and LP There was no significant difference in EEG absolute amplitudes between FP and LP in either derivation by Mann Whitney's U-test (Table4). 3·2 EEG power % and their differences between FP and LP Although there was a tendency for increased α, and decreased β power % in LP, there was no significant difference in EEG power % between FP and LP in either derivation by Mann Whitney's U-test (Table4). DISCUSSION In general, shorter people tends to have a smaller brain(31), and to have shorter VEP latency (1, 15). However, latency tended to be longer at LP in the present study, though there were no significant latency differences after eliminating the effect of body height by ANCOVA. This result is consistent with the previous report by Simpson, et al. (29). Physical conditions such as relaxed state (8), sleep (5), and neuroendocrinological factors such as estrogen (7, 25) and progesterone (25) are supposed to affect VEP latency both at FP and LP. The consciousness level of the subjects was monitored by EEG to be awake during recording. FP is a period when estrogen alone is raised, and LP is a period when both estrogen and progesterone are raised. The effects of estrogen on the CNS are likely to be antagonized by progesterone and its metabolites (17, 25). In addition, estrogen has been shown to shorten the latency of VEPs in animals (7, 25), whereas progesterone prolongs latency (25). Therefore, the tendency for the prolonged VEP latency at LP is thought to reflect the effect of progesterone more than estrogen. In the present study, VEP amplitudes of long components (up to500msec), tended to be larger at LP. Moreover, there was a significantly larger amplitude (P5-N7 for 2CH) at LP than at FP after eliminating the effects of body height by ANCOVA (p<0.05). This result is consistent with other studies of VEPs (22, 29) or ERPs(19). Physical conditions such as relaxed state (8), attention (21), sleep (5) or body temperature (2), and neuroendocrinological factors such as estrogen (7, 25) and progesterone (25) are supposed to affect VEP amplitude both at FP and LP. Increased body temperature reduces the amplitude of the VEP (2), and lower temperature increases compound action potential amplitude (3). However, averaged body temperature is reported to be higher at LP than at FP (9). Therefore, the difference in body temperature is unlikely to account for the observed amplitude differences. Lehtonen et al. (22) suggested the correlation between increased VEP amplitude at LP (22) and increased photic driving in EEGs (35), which is thought to reflect central adrenergic processes (36). However, increased VEP amplitude at LP does not correlate with increased photic driving in EEGs. In animals, VEP amplitude has been shown to be increased by estrogen (7, 25) directly and/or indirectly through L-type voltage-dependent calcium channels (12), acetylcholine (16), monoa-mines (18, 20), γ-aminobutyric acid (24, 26, 32) or glutamate (32), and to be inhibited by progesterone (25) directly and/or indirectly through γ-aminobutyric acid or glutamate (32). In humans, photic driving has been shown to be reduced by estrogen (18, 35) and to be enhanced by progesterone (35). Therefore, larger VEP amplitude at LP observed in the present and previous studies (22, 29) indicates that VEP amplitude at LP reflects the effect of estrogen more than progesterone, while increased photic driving in EEGs at LP (35) indicates that EEGs at LP reflect the effect of progesterone more than estrogen. In the present study, although no significant differences were seen in EEGs by quantitative frequency analysis between FP and LP in both their absolute amplitudes and power % of each band, there was a tendency for increased α power % at LP. This result was consistent with other studies (6). In addition, Creutzfeldt, et al. (6) reported the slight decrease of θ power at LP, and our data shows a tendency for decreased β power % at LP. In a study of estradiol 17-n-valerate, the valerate-acid ester of the endogenous female estrogen, according to the power spectral analysis of EEG, an increase of slower as well as faster waves and a decrease of the activities9to12cps were demonstrated (14). Therefore, as Creutzfeldt, et al. (6) founded, EEG findings in this study also indicate that EEGs at LP reflect the effect of progesterone more than estrogen. In conclusion, the large VEP amplitude, and the tendency for increased α power % and at LP verified in this study were considered to indicate that the VEP amplitude at LP reflects the effect of estrogen more than progesterone, and that the VEP latency and EEG changes at LP reflect the effect of progesterone more than estrogen. We believe that VEP analysis is a useful tool for the study of the actions of gonadal hormones on CNS, not only in animals but also in humans. ACKNOWLEDGMENTS We appreciate the cooperation of our colleagues and students in the college. This study was presented at the 35th annual meeting of The Tokushima Society of Psychiatry and Neurology, Tokushima, June, 1997. REFERENCES 1. Allison T, Wood CC, Goff WR:Brain stem auditory, pattern-reversal visual, and short-latency somatosensory evoked potentials:latencies in relation to age, sex, and brain and body size. Electroenceph Clin Neurophysiol55:619-636, 1983 2. Bajada S, Mastaglia FL, Black JL, Collins DWK: Effects of induced hyperthermia on visual evoked potentials and saccade parameters in normal subjects and multiple sclerosis patients. J Neurol Neurosurg Psychiatry43:849-852, 1980 3. Bolton CF, Sawa GM, Carter K:Temperature effects on the size of human sensory compound action potentials. Can J Neuro Sci6:378, 1979 4. Buchsbaum MS, Henkin RI, Christiansen RL:Age and sex differences in averaged evoked responses in a normal population, with observations on patients with gonadal dysgenesis. Electroenceph Clin Neurophysiol37:137-144, 1974 5. Ciganek L:The EEG response (evoked potential) to light stimulus in man. Electroenceph Clin Neurophysiol13:165-172, 1961 6. Creutzfeldt O, Arnold P, Becker D, Langenstein S, Tirsch W, Wilhelm H, Wuttke W:EEG change during spontaneous and controlled menstrual cycles and their correlation with psychological performance. Electroenceph Clin Neurophysiol40:113-131, 1976 7. Curry JJ, Timiras PS:Development of evoked potentials in specific brain systems after neonatal administration of estradiol. Exp Neurol34:129-139, 1972 8. Danilova NN:Human evoked potentials and functional states. Electroenceph Clin Neurophysiol43:456,1977 9. Dehan CP, Jerger J:Analysis of gender differences in the auditory brainstem response. Laryngoscope100:18-24, 1990 10. Devos JE:The visual evoked response as a measure of brain excitability. Electroenceph Clin Neurophysiol43:544, 1977 11. Dickstein LJ:Menstrual disorders and stress in university students. Psychiat Ann14:436-441, 1984 12. Drouva SV, Rerat E, Bihoreau C, Laplante E, Rasolonjanahary R, Clauser H, Kordon C:Dihydropyridine-sensitive calcium channel activity related to prolactin, growth hormone, and luteinizing hormone release from anterior pituitary cells in culture:interactions with somatostatin, dopamine, and estrogens. Endocrinology123:2762-2773, 1988 13. Harding GFA:The visual evoked potentials in neuro-ophthalmic disorders. In:Desmedt JE, ed. Visual Evoked Potentials. Elsevier, Amsterdam, 1990, pp. 147-167 14. Itil TM, Herrmann WM:Effects of hormones on computer-analyzed human electroencephalogram. In:Lipton MA, DiMascio A, Killam KF, eds. Psychopharmacology:A Generation of Progress. Raven Press, New York, 1978, pp. 729-743 15. Kaneda Y, Nakayama H, Kagawa K, Furuta N, Ikuta T:Sex differences in visual evoked potential and electroencephalogram of healthy adults. Tokushima J Exp Med43:143-157, 1996 16. Kaufman H, Vadasz C, Lajtha A:Effects of estradiol and dexamethasone on choline acetyltransferase activity in various rat brain regions. Brain Res453:389-392, 1988 17. Kawakami M, Sawyer CH:Effects of sex hormones and antifertility steroids in brain thresholds in the rabbit. Endocrinology80:857-871, 1967 18. Klaiber EL, Broverman DM, Vogel W, Kobayashi Y, Moriarty D:Effects of estrogen therapy on plasma MAO activity and EEG driving responses of depressed women. Am J Psychiatry128:1492-1498, 1972 19. Kluck N, O'Connor S, Hesselbrock V, Tasman A, Maier D, Bauer L:Variation in evoked potential measures over the menstrual cycle:a pilot study. Prog Neuropsychopharmacol Biol Psychiatry16:901-911, 1992 20. Kobayashi T, Kobayashi T, Kato J, Menaguchi H: Cholinergic and adrenergic mechanisms in the female rat hypothalamus with special reference to feedback of ovarian steroid hormones. In:Pincus G, Nakao T, Tait J, eds. Steroid Dynamics. Academic Press, New York, 1966, pp. 303-339 21. Kopell BS, Wittner WK, Warrick GL:The effects of stimulus differences, light intensity, and selective attention on the amplitude of the visual averaged evoked potential in man. Electroenceph Clin Neurophysiol 26:619-622, 1969 22. Lehtonen J, Hyyppa MT, Kaihola HL, Kangasniemi P, Lang AH:Visual evoked potentials in menstrual migraine. Headache19:63-70, 1979 23. Maurer K, Bruner M, Hopf HC, Lowitzsch K:Visual pattern evoked responses (VER), acoustically evoked responses (AER) and electrically evoked blink reflexes in assessment of neurofibromatosis. Electroenceph Clin Neurophysiol43:524, 1977 24. Nicoletti F, Patti F, Ferrara N, Canonico PL, Giammona G, Condorelli DF, Scapagnini U:Comparative effects of estrogens and prolactin on nigral and striatal GAD activity. Brain Res232:238-241, 1982 25. Oshima K, Gorbman A: Effects of sex hormones on photically evoked potentials in frog brain. Gen Comp Endocr12:397-404, 1969 26. Parducz A, Perez J, Garcia-Segura LM:Estradiol induces plasticity of GABAergic synapses in the hypothalamus. Neuroscience53:395-401, 1993 27. Rodin EA, Grisell JL, Gudobba RD, Zachary G: Relationship of EEG background rhythms and photic evoked responses. Electroenceph Clin Neurophysiol19:301-304, 1965 28. Saito K, Enai T:The electrode settings based on the10-20electrode system for the selective recording of the VEP (Visual evoked potential). Shikoku Acta Medica (in Japanese) 41:263-277, 1985 29. Simpson D, Erwin CW, Linnoila M:Ethanol and menstrual cycle interactions in the visual evoked response. Electroenceph Clin Neurophysiol52:28-35, 1981 30. Sklar FH, Ehle AL, Clark WK:Visual evoked potentials:a noninvasive technique to monitor patients with shunted hydrocephalus. Neurosurgery4:529-534, 1979 31. Skullerud K:Variations in the size of the human brain. Influence of age, sex, body length, body mass index, alcoholism, Alzheimer changes, and cerebral atherosclerosis. Acta Neurol Scand (Suppl)102:1-94, 1985 32. Smith SS, Waterhouse BD, Woodward DJ:Sex steroid effects on extrahypothalamic CNS. II. Progesterone, alone and in combination with estrogen, modulates cerebellar responses to amino acid neurotransmitters. Brain Res422:52-62, 1987 33. Stockard JJ, Hughes JF, Sharbruogh FW:Visual evoked potentials to electronic pattern reversal: latency variations with gender, age and technical factors. Am J EEG Technol19:171-205, 1979 34. Van den Akker O, Steptoe A:The pattern and prevalence of symptoms during the menstrual cycle. Br J Psychiatry147:164-169, 1985 35. Vogel W, Broverman DM, Kaiku EL:EEG responses in regulatory menstruating women and in amenorrheic women treated with ovarian hormones. Science172:388-391, 1971 36. Vogel W, Broverman DM, Klaiber EL, Kun KJ:EEG response to photic stimulation as a function of cognitive style. Electroenceph clin Neurophysiol27:186-190, 1969 Received for publication March 5, 1997;accepted May 27, 1997. 1 Address correspondence and reprint requests to Yasuhiro Kaneda, M.D., Ph.D., Department of Neuropsychiatry, The University of Tokushima School of Medicine, 3-18-15, Kuramoto-cho, Tokushima, Japan