Power Spectral Analysis of Heart Rate Variability in Healthy Young Women During the Normal Menstrual Cycle

Transcription

1 Power Spectral Analysis of Heart Rate Variability in Healthy Young Women During the Normal Menstrual Cycle NOZOMI SATO, MA, SHINJI MIYAKE, PHD, JUN'ICHI AKATSU MD, AND MASAHARU KUMASHIRO, PHD This study investigated the fluctuations of autonomic nervous activities during the menstrual cycle. Twenty college females were tested for cardiovascular reactivity to mental challenge during both follicular and luteal phases across two menstrual cycles. Power spectral analysis of heart rate variability (HRV) was used to examine the autonomic nervous activities. At baseline, although heart rate and blood pressure did not differ across the menstrual cycle, the low-frequency (LF) component in the HRV was higher and the high-frequency (HF) component in the HRV was lower during the luteal phase than during the follicular phase. The LF/HF ratio was also significantly greater in the luteal phase. These data suggest that sympathetic nervous activities are predominant in the luteal phase as compared with follicular phase. In addition, the power spectral analysis of HRV has more sensitivity than heart rate or blood pressure in assessing the slight fluctuations of autonomic nervous activities during the menstrual cycle. Key words: menstrual cycle, heart rate variability, mental workload. INTRODUCTION Many studies have focused on the physiological reactivity to mental challenge in relationship to the risk of cardiovascular disease (1). Some investigators have suggested that physiological functions fluctuate with the menstrual cycle (2, 3), and others have observed no menstrual phase effects (4, 5). One possible reason for these discrepant results is that the definitions of menstrual cycle and the measures to identify the menstrual phases vary by research (6). Some studies simply divided the menstrual cycle into several phases (7, 8). A few studies strictly monitored hormone levels and basal body temperature to identify the menstrual phases (9, 10). Experimental design, whether employing withinsubjects design (5, 9) or between-subjects design (3, 11), may also have led to inconsistent results. Another possible reason may be related to the sensitivity of physiological parameters. Traditionally, investigations of physiological reactivity have used heart rate, blood pressure, and skin conductance responses (9, 12, 13). As an alternative, power spectral analysis of heart rate variability (HRV) has From the Department of Ergonomics, Institute of Industrial Ecological Sciences, University of Occupational and Environmental Health, Japan Address reprint requests to: N. Sato, Department of Ergonomics, Institute of Industrial Ecological Sciences, University of Occupational and Environmental Health, 1-1 Iseigaoka, Yahatanishi-ku, Kitakyusyu 807, Japan. Received for publication April 18, 1994; revision received September 22, been used as a sensitive index of autonomic nervous activities (14-17). In humans, power spectral analysis of R-R interval variability has revealed that there are two major spectral components: the high frequency (HF) component at the respiratory frequency and the low frequency (LF) component at.03 to.15 Hz. The HF component corresponds to the respiratory sinus arrhythmia and is modulated solely by the parasympathetic nervous system (16, 18, 19), whereas the LF component corresponds to blood pressure oscillations occurring around.1 Hz, (i.e., the Mayer waves) and is jointly modulated by the sympathetic and parasympathetic nervous systems (20, 21). In addition, the LF/HF ratio is also a useful parameter that reflects the balance of autonomic nervous activities (16, 22, 23). The primary purpose of the study was to determine whether the menstrual cycle affected the fluctuations of autonomic nervous activities in the basal period through the use of power spectral analysis of HRV under strict identification of menstrual cycle. A second purpose of this study was to examine changes in cardiovascular reactivity to mental challenge during the menstrual cycle. We hypothesized that the sympathetic nervous activities might be predominant in the luteal phase and might exert influences on HRV parameters. METHODS Subjects Twenty female college students (age 19.2 ±.6 years) participated in the study. Informed consent was obtained from each subject. The subjects reported regular menstrual cycles between Psychosomatic Medicine 57: (1995) /95/ $03.00/0 Copyright 1995 by the American Psychosomatic Society

2 N. SATO et al. 24 and 34 days over the previous 6 months. They had not used oral contraceptives within the preceding 6 months, were not pregnant, did not have heart disease, and did not smoke. Identification of the Menstrual Cycle The subjects measured their basal body temperature (BBT) each morning and recorded the day of menses to verify a recent history of menstrual cycle regularity at least 3 months before entering the study and throughout the 5-month experimental period. On the basis of these data, the experimental days were determined; the luteal phase was scheduled as days 3 to 7 before the first day of bleeding, and the follicular phase was scheduled as days 7 to 10 after the first day of bleeding. We also measured blood levels of progesterone to confirm whether menstrual phases were pinpointed accurately. Task A complex tracing task was designed by the authors, because this kind of task elicits irritability and robust cardiovascular change (24, 25). A complex pathway was presented on a visual display unit. The subjects traced the pathway as accurately and as fast as they could using a mouse. The mouse horizontal and/or vertical axis control elements were reversed. Physiological Measures Electrocardiogram [ECG), respiration, and electroencephalogram (EEG) were monitored with a polygraph system (NEC-360 San-ei, Tokyo, Japan) and recorded on an FM tape recorder (TEAC XR7000, Tokyo, Japan) continuously. Heart rate (HR) was derived by calculating the number of R waves (CM 5 lead) of ECG. Respiration was monitored by a strain-gauge placed around the subject's abdomen halfway between the rib cage and navel. Systolic blood pressure (SBP) and diastolic blood pressure (DBF) were measured before the pretask resting and after the posttask resting periods by an automatic blood pressure monitoring system (Nihon Colin ABPM-630, Aichi, Japan) from a cuff placed on the left upper arm. Referential records of EEG were kept from C 3 to linked ear A 2 and were used to monitor the subject's arousal level. Questionnaire A Menstrual Distress Questionnaire (MDQ) (26) was translated into Japanese by the authors and used to assess subjective feelings associated with symptoms and mood during the menstrual cycle. A Japanese version of Spielberger's State Anxiety Inventory (STAI) (27) was also employed. We used the State Anxiety Scale to measure the fluctuations of anxiety with the menstrual cycle. As a subjective measurement of mental workload on the task, National Aeronautics and Space Administration Task Load Index (NASA-TLX) ratings (28) were used. The ratings consist of six component scales: mental demand, physical demand, temporal demand, performance, effort, and frustration level. The subjects rated their subjective workload on six 10-cm visual-analog scales. Average weighted workload score (WWL) was computed from six-component scores of NASA-TLX ratings. Procedure A within-subject design was used to eliminate the individual differences. Each subject individually visited the laboratory before entering the experiment, and detailed instructions were given on the experimental protocol. The subjects attended for 3 consecutive days during the follicular and luteal phases across two cycles, for a total of 12 sessions. The day of investigation was chosen with respect to the length of each subjects' menstrual cycle. Half of the subjects entered the experiment in the follicular phase, and the other half were tested first in the luteal phase. Each session was identical and was scheduled for the same time of day. All subjects were requested to refrain from eating and drinking for at least 3 hours before the experiment and to avoid heavy exercise and intake of alcohol-containing beverages after the night before the experiment. The subjects sat in a comfortable chair located in a small chamber (temperature 22 C, relative humidity 50%). After electrodes were attached, the subjects filled out the MDQ and STAI. Subjects were instructed to close their eyes, relax, and breathe at their own rate in a supine position for 20 minutes, then to rest seated quietly for 10 minutes. After completion of these periods, blood pressure was measured. Then, a pretask respiratory period followed. Subjects controlled their respiratory frequency during data readings because the HRV component is affected by the breathing pattern (29, 30). The subjects paced their breathing with an auditory tone that signaled inspiration and expiration for 5 minutes. The pace of respiration was determined individually by counting the last 5 minute-resting period in the supine position. The task period consisted of three complex tracing trials alternating with 30 second NASA-TLX ratings installed into the task program. Each task was approximately identical in difficulty. After the task period, the subjects continued to pace their breathing for 5 minutes. Finally, blood pressure was measured again. Blood sampling was performed on the second day of each experimental phase. The blood was centrifuged immediately, and the serum was withdrawn and stored until assay. Serum levels of progesterone were determined by radioimmunoassay (RIA) using commercial kits from BML (Tokyo, Japan). Data Analysis Off-line analysis was performed by a personal computer. The ECG data were digitized at a sampling frequency of 1 khz by a analog/digital converter (Canopus ADX-98E, Kobe, Japan). Each 10 minute-task period was divided into the first 5-minute and last 5-minute readings for subsequent analysis. Eight blocks of ECG data were analyzed; pretask respiratory period (PRE), first task period (Tl-1, Tl-2), second task period (T2-1, T2-2), third task period (T3-1, T3-2), and posttask respiratory period (POST). ECG data were interpolated by the cubic-spline method and were resampled at 2 Hz for spectral analysis. To remove trends, these equidistant R-R interval data were high-pass filtered with a.05 Hz cut-off frequency by digital filter. Next, the autoregressive (AR) power spectrum from trendgram was calculated by the fixed loth-order AR model. The order was determined empirically. The power spectral density was divided into a sum of spectral components by means of a spectral decomposition method and center frequency, and the power of every spectral component was obtained. Only components greater than 5% of the total power were considered to be significant. A spectral component with a center frequency between.05 and.15 Hz was considered as the LF component. A spectral component with the same cycle as the 332 Psychosomatic Medicine 57: (1995)

3 HEART RATE VARIABILITY DURING THE MENSTRUAL CYCLE respiratory component and a center frequency of around.25 Hz was considered as the HF component. Each spectral component was normalized (%LF, %HF) by dividing it by total power less the first order component, if present (21). The LF/HF ratio was also calculated to assess the sympathovagal balance. Task performance was calculated by dividing the total traced length on the pathway by the number of deviations from the pathway. A 2 X 8 MANOVA (2 phases X eight measurement blocks) and post hoc t tests were performed on the HR, %LF, %HF, and LF/HF. Base-line levels across the menstrual cycle were assessed by using Student's paired t test. Data are expressed as means ± SE. <.01). The mean HF components were lower during the luteal phase (F(l,19) = 26.28, p <.001). Regarding the LF/HF ratios analysis, a highly significant main effect for phase was obtained. As shown in Figure 1, LF/HF ratios increased during the luteal phase (F(l,19) = 12.76, p <.01). Except for the T3-1 period, LF/HF ratios did not increase significantly, relative to the base-line levels (p <.01). As for SBP and DBP, no significant differences were obtained between the follicular and luteal phases. The task elicited slight increases after the posttask respiratory period on both SBP and DBP compared to the pretask respiratory period (p <.01). Phase X Block interactions were not significant for any cardiovascular parameters. RESULTS Hormone Assays Mean serum levels of progesterone were.35 ±.03 ng/ml (range.20 to.65) for the follicular phase and 7.23 ±.85 ng/ml (range 3.40 to 16.40) for the luteal phase, respectively. Therefore, subjects were tested in the target phases correctly. Task Performance and NASA-TLX Table 1 illustrates mean values for task performance and NASA-TLX WWL scores across the two menstrual phases. For task performance, significant decreases were observed over task periods (F(2,38) = 19.38, p <.001). NASA-TLX WWL scores increased gradually as a function of task periods (F(2,38) = 3.61, p <.05). Physiological Measures First, values of pretask respiratory period (defined as base-line levels) were compared for the physiological measures between the follicular and luteal phases. There were no significant differences for HR and BP across the menstrual cycle. As for the power spectral analysis of HRV, there were significant phase differences (p <.05). Mean base-line levels during the follicular and luteal phases were ± 2.68% and ± 2.91% for the LF components, ± 2.68% and ± 2.91% for the HF components, and 1.07 ±.13 and 1.68 ±.23 for the LF/HF ratios, respectively. Results from the MANOVA on HR revealed no significant phase effect, although values were marginally higher during the luteal phase (f(l,19) = 3.15, p <.10). The LF components were significantly higher during the luteal phase (F(l,19) = 28.19, p <.001). In comparison to base-line levels, LF components only increased during the T3-1 period (p Questionnaire Mean self-ratings for MDQ were 1.45 ±.07 for the follicular phase and 1.50 ±.09 for the luteal phase. No significant phase differences were obtained for ratings of MDQ. The results of STAI also showed no phase-related differences (39.90 ± 1.37 for the follicular and ± 1.22 for the luteal phase, respectively). DISCUSSION The results of this study indicate that autonomic nervous activities fluctuate during the menstrual cycle. We observed significant physiological differences between the follicular and luteal phases by using the power spectral analysis of HRV. PRE T1-1 T1-2 T2-1 T2-2 T POST R I O 0 K Fig. 1. LF to HF changes (mean ± SE) during experimental session; **p <.01. Psychosomatic Medicine 57: (1995) 333

4 N. SATO et al. TABLE 1. Mean Values of Task Performance and NASA-TLX WWL Scores in the Luteal and Follicular Phases of the Menstrual Cycle" Task period Task 1 Task 2 Task 3 Task performance Luteal Follicular ± ±4.80 NASA-TLX WWL scores Luteal ± ± Follicular ± * Values are means ± SE. On the basis of preliminary studies, the HF component in the HRV has been considered as a useful index of vagal activities because it decreases with tilting or atropine administration and increases with propranolol administration or respiratory control (15, 16, 18, 19). On the other hand, an acute intravenous administration of propranolol blunts the increase of the LF component in the HRV induced by tilting (15). Therefore, the LF component has been felt to be mainly a measure of sympathetic activities. Accordingly, Lombardi et al. (22) reported that, in patients 2 weeks after acute myocardial infarction, the LF/HF ratio was significantly greater than in healthy subjects. Therefore, it could be a convenient index of sympathovagal interaction (16, 23). This study provides findings that the luteal phase of the menstrual cycle was associated with a greater increase in the LF component and a greater decrease in the HF component, resulting in a higher LF/HF ratio. This suggests sympathetic nervous activities are predominant in the luteal phase during the menstrual cycle. Little et al. (31) showed higher heart and respiration rates during the luteal phase compared with other phases. Kirsch et al. (13), Asso et al. (32), and Vila et al. (33) reported that women are more autonomically aroused during the luteal phase than the follicular and/or ovulatory phases through the use of galvanic skin response. Our results are in keeping with these conclusions. Although speculative, high concentration levels of progesterone, estrogen, and/or other hormones might influence autonomic nervous activities during the luteal phase. Therefore, consideration should be given to examinations evaluating physiological reactivity to mental challenge in women. Currently, the most reliable method to verify the menstrual phase is measuring the levels of reproductive hormones in blood (4, 10). In our study, we measured progesterone levels as a reliable marker of menstrual phase and observed a significant rise in this hormone during the luteal phase. For heart rate and blood pressure, no significant differences were observed between the luteal and follicular phases. These may partly depend on the sensitivity of each parameter. Kamada et al. (34) found the differences in autonomic balance by LF/HF ratio in HRV, which could not be observed in heart rate and blood pressure. Similar trends are shown in this study. This suggests that power spectral analysis of HRV has more sensitivity than heart rate and/or blood pressure to assess the autonomic nervous activities. HRV has been also used as a useful index of mental workload (35). However, in the present study, significant changes were not observed in HRV parameters during task periods. One possible explanation is that the task did not act as a stressor to induce robust reactivity. Subjects might have habituated the task over the course of the total of 12 sessions. In addition, it has been reported that the effects of tasks on physiological reactivity vary by task characteristics (36). Therefore, various laboratory stressors should be administered to examine the difference in reactivity in future research. Subjects' performance of the task did not differ because of the menstrual cycle. Contradictory results may be related to the fact that the various tasks may measure aspects of information-processing functions that are differentially responsive or insensitive to slight hormonal fluctuations (37). Again, it would seem important for future investigations to specify what kind of tasks are influenced by the menstrual cycle. Symptoms and moods during the menstrual cycle did not fluctuate in the present study. This could be partly the cause of large individual differences in rating scores. As Olasov and her colleague suggested (38), daily mood associated with the menstrual cycle may depend on individual awareness of the menstrual phases, at least for non-premenstrual syndrome (PMS) women. In summary, this study clarifies that for normallycycling women, sympathetic nervous activities are more predominant in the luteal phase at rest and during the task periods. Future investigations focusing on physiological reactivity to mental challenge should consider the differences in physiological base-line values during the menstrual cycle for women. In addition, some researchers have suggested that stress-response patterns differ between PMS-women and non-pms women (12, 13). Further studies examining the autonomic nervous system in PMS-women using power spectral analysis of HRV are desirable for the clinical application. 334 Psychosomatic Medicine 57: (1995)

5 HEART RATE VARIABILITY DURING THE MENSTRUAL CYCLE The authors thank Ms. Mihoko Matsunami for her help with this study. REFERENCES 1. Krantz DS, Manuck SB: Acute psychophysiologic reactivity and risk of cardiovascular disease: A review and methodological critique. Psychol Bull 96(3): , Damas-Mora J, Davies L, Taylor W, Jenner FA: Menstrual respiratory changes and symptoms. Br J Psychiatry 136: , Johnston VS, Wang XT: The relationship between menstrual phase and the P3 component of ERPs. Psychophysiology 28(4): , Stoney CM, Owens JF, Matthews KA: Influences of the normal menstrual cycle on physiologic functioning during behavioral stress. Psychophysiology 27(2): , Weidner G, Helmig L: Cardiovascular stress reactivity and mood during the menstrual cycle. Women Health 16(3):5-21, Doty RL: A procedure for combining menstrual cycle data. J Clin Endocrinol Metab 48(6): , Diamond M, Diamond L, Mast M: Visual sensitivity and sexual arousal levels during the menstrual cycle. J Nerv Ment Dis 155: , Zimmerman E, Parlee MB: Behavioral changes associated with the menstrual cycle: An experimental investigation. J Appl Psychol 3: , Collins A, Eneroth P, Landgren BM: Psychoneuroendocrine stress responses and mood as related to the menstrual cycle. Psychosom Med 47(6): , Laessle RG, Tuschl RJ, Schweiger U, Pirke KM: Mood changes and physical complaints during the normal menstrual cycle in healthy young women. Psychoneuroendocrinology 15(2): , Hunter S, Schraer R, Landers DM, Buskirk ER, Harris DV: The effects of total oestrogen concentration and menstrual-cycle phase on reaction time performance. Ergonomics 22(3): , Girdler SS, Pedersen CA, Stern RA, Light KC: Menstrual cycle and premenstrual syndrome: Modifiers of cardiovascular reactivity in women. Health Psychol 12(3): , Kirsch JR, Geer JH: Skin conductance and heart rate in women with premenstrual syndrome. Psychosom Med 50: , Sayers B McA: Analysis of heart rate variability. Ergonomics 16(l):17-32, Pomeranz B, Macaulay RJB, Caudill MMA, et al: Assessment of autonomic function in humans by heart rate spectral analysis. Am J Physiol 248:H151-H153, Pagani M, Lombardi F, Guzzetti S, et al: Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res 59: , Malliani A, Pagani M, Lombardi F, Cerutti S: Cardiovascular neural regulation explored in the frequency domain. Circulation 84(2): , Eckberg DL: Human sinus arrhythmia as an index of vagal cardiac outflow. J Appl Physiol 54(4): , Akselrod S, Gordon D, Madwed JB, Snidman NC, Shannon DC, Cohen RJ: Hemodynamic regulation: Investigation by spectral analysis. Am J Physiol 249:H867-H875, Baselli G, Cerutti S, Civardi S, et al: Spectral and crossspectral analysis of heart rate and arterial blood pressure variability signals. Comp Biomed Res 19: , Baselli G, Cerutti S, Civardi S, Malliani A, Pagani M: Cardiovascular variability signals: Towards the identification of a closed-loop model of the neural control mechanisms. IEEE Trans Biomed Eng 35(12): , Lombardi F, Sandrone G, Pernpruner S, et al: Heart rate variability as an index of sympathovagal interaction after acute myocardial infarction. Am J Cardiol 60: , Rimoldi O, Pierini S, Ferrari A, Cerutti S, Pagani M, Malliani A: Analysis of short-term oscillations of R-R and arterial pressure in conscious dogs. Am J Physiol 258:H967- H Saab PG, Matthews KA, Stoney CM, McDonald RH: Premenopausal and postmenopausal women differ in their cardiovascular and neuroendocrine responses to behavioral stressors. Psychophysiology 26(3): , Owens JF, Stoney CM, Matthews KA: Menopausal status influences ambulatory blood pressure levels and blood pressure changes during mental stress. Circulation 88: , Moos RH: The development of a menstrual distress questionnaire. Psychosom Med 30: , Spielberger CD, Gorsuch RL, Lushene RE: STAI manual. Palo Alto, Consulting Psychologist Press, Hart SG, Staveland LE: Development of NASA-TLX(Task Load Index): Results of empirical and theoretical research. In Hancock PA, Meshkati (eds.), Human Mental Workload. Amsterdam, Elsevier Science Publishers B.V, 1988, Hirsch JA, Bishop B: Respiratory sinus arrhythmia in humans: How breathing pattern modulates heart rate. Am J Physiol 241:H620-H629, Grossman P, Karemaker J, Wieling W: Prediction of tonic parasympathetic cardiac control using respiratory sinus arrhythmia: The need for respiratory control. Psychophysiology 28(2): , Little BC, Zahn TP: Changes in mood and autonomic functioning during the menstrual cycle. Psychophysiology 11(5): , Asso D, Beech HR: Susceptibility to acquisition of a conditioned response in relation to the menstrual cycle. J Psychosom Res 19: , Vila J, Beech HR: Vulnerability and conditioning in relation to the menstrual cycle. Br J Soc Clin Psychol 16:69-75, Kamada T, Miyake S, Kumashiro M, Monou H, lnoue K: Power spectral analysis of heart rate variability in Type As and Type Bs during mental workload. Psychosom Med 54: , Cerutti S, Fortis G, Liberati D, Baselli G, Civardi S: Power spectrum analysis of heart rate variability during a mental arithmetic task. J Ambulatory Monitoring l(3): , Kasprowicz AL, Manuck SB, Malkoff SB, Krantz DS: Individual differences in behaviorally evoked cardiovascular response: Temporal stability and hemodynamic patterning. Psychophysiology 27(6): , Phillips SM, Sherwin BB: Variations in memory function and sex steroid hormones across the menstrual cycle. Psychoneuroendocrinology 17(5): , Olasov B, Jackson J: Effects of expectancies on women's reports of moods during the menstrual cycle. Psychosom Med 49(l):65-78, 1987 Psychosomatic Medicine 57: (1995) 335