Amenorrhea in Olympic marathon runners*t*

Transcription

1 FERTILITY AND STERILITY Copyright 1987 The American Fertility Society Printed in U.S.A. Amenorrhea in Olympic marathon runners*t* Allan R. Glass, M.D. II~** Patricia A. Deuster, Ph.D.tt Susan B. Kyle, M.A.tt Janet A. Yahiro, Ph.D.** Robert A. Vigersky, M.D. II~** Eric B. Schoomaker, M.D., Ph.D. II~** Uniformed Services University of Health Sciences, Bethesda, Maryland and Walter Reed Army Medical Center, Washington, District of Columbia Women who exercise heavily may develop secondary amenorrhea. Since the mechanism of so-called "runner's amenorrhea" has not been conclusively established, the authors examined the occurrence of amenorrhea in one of the most intensively exercising groups of female runners in the United States (average, 70 miles/week): those women participating in the marathon trials for the 1984 Olympics. Nineteen percent of these Olympic runners were amenorrheic. When compared with eumenorrheic marathon runners, these amenorrheic runners were significantly (P < 0.05) younger (24.8 ± 1.2 [standard error of the mean] versus 30.8 ± 0.8 years), lighter (108.4 ± 2.5 versus ± 1.7 lb), and leaner (11.2 ± 0.5 versus 12.5 ± 0.3% body fat). There were no differences between the two groups in weekly training mileage, proportion completing the marathon trial, finishing time, basal serum prolactin, or postmarathon serum prolactin. Although basal serum cortisol was slightly higher in the amenorrheic group (26.6 ± 0.8 versus 22.3 ± 0.7 /-Lg/dl; P < 0.05), postmarathon serum cortisol was similar in the two groups. This study supports the concept that training intensity above a certain threshold seems to have little effect on the development of runner's amenorrhea, and vigorously training national caliber marathon runners have a lower incidence of amenorrhea than previously predicted. Fertil Steril 48:740, 1987 Received April 21, 1987; revised and accepted July 17, * The opinions or assertions contained herein are the private views ofthe authors and are not to be construed as official or as reflecting the views of the Department of the Army or the Department of Defense. t Funding was provided by the Department of Clinical Investigation, Walter Reed Army Medical Center, and by grant C076AD from the Uniformed Services University of Health Sciences. :j: Reprint requests: Allan R. Glass, M.D., Endocrinology Service, 7D, Walter Reed Army Medical Center, Washington, DC Department of Medicine, Uniformed Services University of Health Sciences. II Department of Medicine, Walter Reed Army Medical Center. ~ Department of Clinical Investigation, Walter Reed Army Medical Center. ** Department of Physiology, Uniformed Services University of Health Sciences. tt Department of Military Medicine, Uniformed Services University of Health Sciences. 740 Glass et al. Amenorrhea in Olympic runners The recent explosion of interest in running as a form of exercise has led to the recognition that there is an increased incidence of secondary amenorrhea in female runners. Several factors have been suggested as leading to this amenorrhea, including weight loss, low percentage of body fat, energy drain, changes in diet, or exercise-induced hormonal changes. Despite numerous studies, the mechanism of runner's amenorrhea remains incompletely defined. To explore this question further, we chose as a model for study the most elite, intensely training group of female long-distance runners in the United States: those who qualified to participate in the women's Olympic trials marathon. We postulated that this heavily exercising group (running an average of 70 miles/week) would have a high incidence of amenorrhea. Furthermore, we theorized that differences between amenorrheic and eumenorrheic runners in this population would Fertility and Sterility

2 be magnified when compared with less intensively exercising groups and that examination of such differences would shed light on the mechanism of runner's amenorrhea. MATERIALS AND METHODS Women who had qualified to compete in the 1984 women's Olympic trials marathon by running a certified marathon (42.25 km) in 2:51:19 or less were contacted by mail and invited to participate in this study. Of the 202 eligible women, 80 agreed to take part in one or more aspects of this investigation. The distribution of marathon finishing times in this subgroup of 80 women was similar to that in the entire group of marathon participants, indicating comparable athletic performance. Participants in the study were asked to complete questionnaires detailing their menstrual history and their exercise history. Subjects' height and weight were determined, and skin-fold thickness was measured at seven sites for calculation of body fat percentage. Participants also were asked to provide a single fasting baseline blood sample between 6:00 and 10:00 A.M. (no exercise in previous 12 hours) on one of the 3 days prior to the race, and some subjects also agreed to provide another blood sample within 60 minutes after completing the Olympic trials marathon. Blood samples were allowed to clot at room temperature and the serum subsequently was kept frozen until analyzed. Baseline serum samples were assayed for luteinizing hormone (LH), follicle-stimulating hormone (FSH), estradiol (E 2), cortisol, and prolactin (PRL) by radioimmunoassay. Postmarathon blood samples were assayed for cortisol and PRL. Menstrual status was determined from the responsesto the questionnaire. Women were defined as amenorrheic if they reported two or fewer menstrual cycles within the previous year and no menstrual cycles within the previous three months. Women were classified as eumenorrheic if they reported ten or more menstrual cycles in the previous year. Women not fitting into either of these categories were defined as oligomenorrheic. Of the 80 participants, 8 were excluded from analysis because they were taking hormonal medications, thus precluding assignment of menstrual status. Also excluded were other subjects in whom menstrual status could not be assigned (1 hysterectomy, 1 pregnant, 1 lactating, 2 insufficient data), yielding 67 classifiable subjects. The number of subjects taking part in each section of the study differed, and the appropriate numbers are indicated in the Results section. Assignment of code numbers to each participant ensured confidentiality by rendering it impossible for those analyzing the data to identify any individual. Statistical comparisons between amenorrheic and eumenorrheic groups were carried out by Student's t-test (serum E2 analyzed after logarithmic transformation) or Wilcoxon rank sum test, as appropriate, with a nominal significance level of This study was approved by the Human Use Review Committee of the Uniformed Services University of Health Sciences and the Athletic Congress of the United States Olympic Committee. RESULTS Of the 67 women who could be classified according to menstrual status, 13 were amenorrheic and 44 were eumenorrheic. The remaining 10 women were classified as oligomenorrheic (mean of 5.9 menstrual periods during the previous year) and were not included in subsequent analyses. The amenorrheic runners were significantly younger than the eumenorrheic runners, and they were also slightly, but significantly, lighter and leaner than the eumenorrheic runners (Table 1). There was no difference between the two groups in weight changes over the previous 5 years. The intensity of training was similar in the two groups, as was the mean finishing time for those completing the Olympic trial marathon (Table 1). The age at menarche was not significantly different in the two groups (Table 1), but none of the amenorrheic group, as compared with 18 of the eumenorrheic group, reported a previous pregnancy. Table 1 Characteristics of Olympic-Caliber Runners Age (yr) Height (in) Weight (lb) Weight change in past 5 years (lb) Body fat (%) Age at menarche (yr) Training intensity (miles/wk) Marathon time (min)" Amenorrheic 24.8 ± 1.2" 64.7 ± ± ± ± ± ± ± 3.4 Eumenorrheic 30.8 ± 0.8 b 64.2 ± ± 1.7 b -3.9 ± ± 0.3 b 12.8 ± ± ± 1.3 " Mean ± standard error of the mean (n = 11 to 13 for amenorrheic; n = 36 to 44 for eumenorrheic). b p < 0.05 versus amenorrheic. C During Olympic trials. Nonfinishers are excluded (2/13 amenorrheic, 4144 eumenorrheic). Glass et al. Amenorrhea in Olympic runners 741

3 I Hormone concentrations in the amenorrheic and eumenorrheic runners are shown in Table 2, with values for serum LH, FSH, and E2 in the eumenorrheic group limited to those subjects in the follicular phase of the menstrual cycle (days 2 to 9) at the time the blood sample was obtained. Serum LH and FSH were not significantly different in the two groups, but serum E2 was significantly lower in the amenorrheic group. Serum PRL, both basally and after completing the marathon, was the same in both groups, and both groups showed a significant increase in serum PRL after the marathon (Table 2). Baseline serum cortisol was slightly, but significantly higher in the amenorrheic group, but postmarathon serum cortisol levels were similar in the two groups (Table 2). DISCUSSION Intensive endurance exercise training in women can lead to secondary amenorrhea. Vigorous exercise also is associated with various physiologic changes, many of which may not have etiologic significance for the development of amenorrhea. However, parameters that show differences between amenorrheic runners and eumenorrheic runners may be important in leading to loss of menses. For example, one of the first factors proposed as leading to amenorrhea in runners was the intensity of training. Feicht et al. 1 showed a correlation between the incidence of amenorrhea in runners and the number of training miles per week. By extrapolation, they predicted that national-cali- Table 2 Runners Hormone Concentrations in Olympic Caliber Serum LH (miu/m!) Serum FSH (miu/m!) Serum estradiol (pg/m!) Serum prolactin (ng/ml) Baseline Postmarathon Serum cortisol (llg/dl) Baseline Postmarathon Amenorrheic 10.8 ± 2.8 a 13.1 ± ± ± ± ± ± 3.6 Eumenorrheic 12.1 ± 1.6 b 13.1 ± 0.9 b 36.1 ± 6.3 b c 11.4 ± ± ± 0.7 c 42.0 ± 1.5 a Mean ± standard error of the mean (amenorrheics: n = 11 to 13 for baseline values, n = 8 to 10 postmarathon; eumenorrheics: n = 41 for baseline [n = 11 for follicular phase], n = 28 postmarathon). b Women in follicular phase of menstrual cycle (days 2 to 9). c P < 0.05 versus amenorrheic. ber marathon runners (such as those in the current study) would have an incidence of amenorrhea greater than 50%.1 Some studies supported this concept by showing that amenorrheic runners ran more miles per week than did eumenorrheic runners. 2,3 In contrast, in the current study, there were no differences between the amenorrheic and eumenorrheic runners in weekly training mileage, percentage completing the Olympic trials marathon, or finishing times for those completing the race. Moreover, the incidence of amenorrhea in this group of elite runners was 19%, much less than the 50% predicted by Feicht et al.1 and similar to the incidence of amenorrhea in women whose weekly training mileage was much 10wer.4-6 Although we have no reason to suspect that our study, which was portrayed to potential participants as more extensive than simply a menstrual survey, showed a selection bias in favor of either amenorrheic or eumenorrheic runners, especially since the study group constituted a large proportion (40%) of the entire eligible population and was representative of it in terms of marathon finishing times, we cannot exclude the possibility that our results may be colored by undetected causes of selection bias. Several previous studies also have failed to find a correlation between the incidence of amenorrhea in runners and the weekly training mileage,4,6,7 with one study actually finding amenorrhea more common in those whose weekly training mileage was lower.s Thus, the preponderance of evidence does not seem to support training intensity as a major factor in leading to runner's amenorrhea. Another factor proposed as leading to amenorrhea in exercising women is loss of body weight or body fatness. Since low body weight or body fatness is associated with amenorrhea in other situations, such as anorexia nervosa, it seemed logical to propose that the loss of body weight and/or body fatness typically seen in women engaged in intensive exercise training might similarly lead to amenorrhea. Consequently, one might expect to find lower body weight or body fatness in amenorrheic runners than in eumenorrheic runners. In fact, the current study of Olympic-caliber runners did show that amenorrheic athletes had slightly, but significantly, lower body weight and body fatness than did eumenorrheic runners. Several previous studies also have shown lower body weight and/or body fatness in amenorrheic runners. 2,5,6,9 Furthermore, when normal women begin an intensive exercise 742 Glass et a1. Amenorrhea in Olympic runners Fertility and Sterility

4 program, the incidence of subsequent menstrual abnormalities can be decreased if weight loss is averted.lo Thus, although some studies have shown that amenorrheic and eumenorrheic runners are similar in body weight and/or body fatness/,8 substantial evidence suggests that there may be some connection between these parameters and development of runner's amenorrhea. The exact nature of this connection, however, has not been defined. There may be a direct causal relationship between levels of body weight or body fatness and the persistence of menstrual cycles. Conversely, the development of amenorrhea coincident with a greater fall in body weight or fatness in some runners may be independent effects reflecting an underlying difference in metabolism or food intake. These theories should be interpreted with caution since, in our study (as in some others), the difference in body weight or body fatness between eumenorrheic and amenorrheic runners is quite small and of uncertain physiologic significance. Moreover, since some runners with very low body weight or body fatness maintain normal menses, these factors, although contributory, cannot be the exclusive cause of runner's amenorrhea. Recently, attention has been focused on the possibility that women who develop amenorrhea during exercise training may have had abnormal menstrual function prior to beginning training. Several studies have noted that amenorrheic runners have delayed menarche/,8 increased incidence of menstrual dysfunction prior to running,5,7 and decreased incidence of previous pregnancy,4,5,8 all of which might suggest that such women have abnormal underlying reproductive function independent of whatever additional stress running imposes. In contrast, our study, like several others,5,6 found normal age of menarche in amenorrheic runners. Some of the discrepancies on this point may reflect the age at which training began, since Frisch et aly have shown that menarche is delayed in athletes only if training begins before menarche. This latter finding suggests a specific exercise effect in leading to amenorrhea rather than an underlying abnormality in pituitary-ovarian function. Further supporting a specific role of exercise is the high incidence of menstrual abnormalities that develop when normal women are subjected to exercise training. 10,12 Another difference between amenorrheic and eumenorrheic runners in this study was age: amenorrheic runners were considerably younger than eumenorrheic runners. This difference also has been noted in at least two prior studies,9,13 although it has not been stressed as a possible factor related to development of runner's amenorrhea. This age differential has not been apparent in many previous studies because of deliberate age-matching of amenorrheic and eumenorrheic runners or because subject selection was age-limited (e.g., college-age runners). Our study provides no information on why amenorrheic runners might be younger than eumenorrheic runners. In one previous study, this age difference reflected to some extent the finding that amenorrheic runners began exercise training at a younger age.13 This point raises the possibility that menstrual cycles might be more vulnerable to disruption by exercise in the early postmenarchal period than at later ages, as has been suggested.9,1l Alternatively, there may be some adaptation that occurs with aging, such that older subjects are less likely to develop amenorrhea following exercise than younger subjects. The hormonal mechanism of runner's amenorrhea has been studied extensively. Substantial evidence now indicates that runner's amenorrhea reflects an impairment in the pulsatile release of luteinizing hormone-releasing hormone (LH-RH) from the hypothalamus, which consequently leads to secondary deficiency of gonadotropins and estrogen along with anovulation.14 The data from the current study (low serum E2 levels without elevations in serum LH or FSH) are consistent with this hypothesis. However, the possibility remains that the loss of pulsatile LH-RH release in amenorrheic runners is itself secondary to other exercise-related hormonal changes. In particular, because exercise can lead to transient increases in serum PRL and/ or cortisol, and overproduction of cortisol or PRL can lead to amenorrhea under other circumstances, one might theorize that the intermittent hyperprolactinemia or hypercortisolemia that occurs during running disrupts the hypothalamic-pituitary-ovarian axis and leads to runner's amenorrhea. The current study of Olympic-caliber runners disclosed no difference in basal PRL or postmarathon serum PRL levels between amenorrheic and eumenorrheic runners. Likewise, most previous studies have shown normal basal serum PRL levels in amenorrheic runners.3,8,15 Considerable disagreement exists, however, about the magnitude of the PRL change after exercise in amenorrheic runners, with previous studies indicating that amenorrheic runners have greater/5 lesser,3,16 or the same17 PRL Glass et a1. Amenorrhea in Olympic runners 743

5 increase after acute exercise as controls (eumenorrheic runners or sedentary women). Thus, in the absence of more definitive data, one cannot conclude that runner's amenorrhea occurs only in those women with greater serum PRL increases during running. Some lines of evidence suggest that hypercortisolemia may be important in initiating runner's amenorrhea. Recent studies indicate that runners have higher serum and urine cortisol levels than normals.16,18 However, these studies failed to find a difference in cortisol levels between amenorrheic and eumenorrheic runners/8 and several other investigators have reported normal basal serum cortisol levels and normal serum cortisol responses to stimulation by corticotropin (ACTH) in amenorrheic runners.3,8,15,19 In the current study, although basal serum cortisol levels were significantly higher in amenorrheic runners than in eumenorrheic runners, the difference was small. Furthermore, serum cortisol levels in both groups were quite high, probably reflecting the stress of the upcoming marathon. There is also considerable controversy about cortisol responses to exercise in runners. Even though normal subjects starting an exercise training program will show increasing cortisol responses to acute exercise as the training intensity is increased,20 the serum cortisol increases after acute exercise in amenorrheic runners have been reported to be greater than,15 less than, 3 or the same16 as the responses in eumenorrheic runners. Our study also found similar postmarathon cortisol levels in amenorrheic and eumenorrheic runners. Thus, as with PRL, convincing evidence is lacking to conclude that a greater degree of intermittent hypercortisolemia, due to enhanced hormonal response to exercise, is a causative factor in those women who develop loss of menses. It should be emphasized, however, that even if amenorrheic runners do not have enhanced cortisol or PRL responses to daily exercise when compared with their eumenorrheic counterparts, we cannot exclude the possibility that amenorrhea results from enhanced sensitivity in such individuals to these exercise-induced hormonal changes. In summary, the current study of the most intensively exercising group of American women Olympic-caliber marathon runners-indicates that amenorrhea is less common than would be expected if training intensity were a key factor in the loss of menses. As in previous survey studies of this type, we could not evaluate the incidence of repro- ductive abnormalities less severe than amenorrhea, such as anovulation or luteal phase defect, which might occur despite persistence of regular vaginal bleeding. Amenorrheic runners were somewhat lighter and leaner than their eumenorrheic counterparts, indicating a possible relationship between energy balance and the development of amenorrhea. The amenorrheic runners also were considerably younger than the normally menstruating runners, a fact whose significance remains undetermined. The development of amenorrhea in heavily exercising, elite runners is not associated with improved athletic performance. Eumenorrheic and amenorrheic runners did not differ in postmarathon serum levels of PRL or cortisol, and the exact mechanism by which exercise training might lead to loss of pulsatile LH -RH secretion and sub-. sequent amenorrhea remains undefined. Acknowledgment. We thank Mrs. Estelle Coleman for expert editorial assistance. REFERENCES 1. Feicht CB, Johnson TS, Martin BJ, Sparkes KE, Wagner WW: Secondary amenorrhea in athletes. Lancet 2:1145, Carlberg KA, Buckman MT, Peake GT, Riedesel ML: A survey of menstrual function in athletes. Eur J Appl Physiol 51:211, Loucks AB, Horvath SM: Exercise-induced stress responses of amenorrheic and eumenorrheic runners. J Clin Endocrinol Metab 59:1109, Dale E, Gerlach DH, Wilhite AL: Menstrual dysfunction in distance runners. Am J Obstet Gynecol 54:47, Schwartz B, Cumming DC, Riordan E, Selye M, Yen SSC, Rebar RW: Exercise-associated amenorrhea: a distinct entity? Am J Obstet Gynecol 141:662, Galle PC, Freeman EW, Galle MG, Huggins GR, Sondheimer SJ: Physiologic and psychologic profiles in a survey of women runners. Fertil Steril 39:633, Shangold MM: Exercise and amenorrhea. Semin Reprod Endocrinol 3:35, Baker ER, Mathur RS, Kirk RF, Williamson HO: Female runners and secondary amenorrhea: correlation with age, parity, mileage, and plasma hormonal and sex-hormonebinding globulin concentrations. Fertil Steril 36:183, Marcus R, Cann C, Madvig P, Minkoff J, Goddard M, Bayer M, Martin M, Gaudiani L, Haskell W, Genant H: Menstrual function and bone mass in elite women distance runners: endocrine and metabolic features. Ann Intern Med 102:158, Bullen BA, Skrinar GS, Beitins IZ, von Mering G, Burnbull BA, McArthur JW: Induction of menstrual disorders by strenuous exercise in untrained women. N Engl J Med 312:1349, Glass et ai. Amenorrhea in Olympic runners Fertility and Sterility

6 11. Frisch RE, Gotz-Welbergen AV, McArthur JW, Albright T, Witschi J, Bullen B, Birnholz J, Reed RB, Hermann H: Delayed menarche and amenorrhea of college athletes in relation to age of onset of training. JAMA 246:1559, Boyden TW, Pamenter RW, Stanforth P, Rotkis T, Wilmore JH: Sex steroids and endurance running in women. Fertil Steril 39:629, Fisher EC, Nelson ME, Frontera WR, Turksoy RN, Evans WJ: Bone mineral content and levels of gonadotropins and estrogens in amenorrheic running women. J Clin Endocrinol Metab 62:1232, Veldhuis JD, Evans WS, Demers LM, Thorner MO, Wakat D, Rogol AD: Altered neuroendocrine regulation of gonadotropin secretion in women distance runners. J Clin Endocrinol Metab 61:557, Miescher E, Stager JM, Niswender GD, Robertshaw D: Athletic amenorrhea and the serum cortisol and prolactin response to stress. Fed Proc 44:1371, Cumming DC, Rebar RW: Exercise and reproductive function in women. Am J Indust Med 4:113, Chang FE, Dodds WG, Sullivan M, Kim MH, Malarkey WB: The acute effects of exercise on prolactin and growth hormone secretion: comparison between sedentary women and women runners with normal and abnormal menstrual cycles. J Clin Endocrinol Metab 62:551, Villanueva AL, Schlosser C, Hopper B, Liu JH, Hoffman DI, Rebar RW: Increased cortisol production in women runners. J Clin Endocrinol Metab 63:133, Ronkainen HRA, Pakarinen AJ, Kauppila AJI: Adrenocortical function of female endurance runners and joggers. Med Sci Sports Exerc 18:385, Carr DB, Bullen BA, Skrinar GS, Arnold MA, Rosenblatt M, Beitins IZ, Martin JB, McArthur JW: Physical conditioning facilitates the exercise-induced secretion of betaendorphin and beta lipotropin in women. N Engl J Med 305:560, 1981 Glass et ai. Amenorrhea in Olympic runners 745