This article was downloaded by: [Monash University Library] On: 04 February 2015, At: 10:33 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Health Care for Women International Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uhcw20
Body‐fat measurements and athletic menstrual irregularity a
b
Patricia J. Estok RN, PhD, FAAN , Ellen B. Rudy RN, PhD, FAAN & Jean A. Just RN, MSN a
School of Nursing , Kent State University ,
b
School of Nursing , Case Western Reserve University ,
c
c
Doctor's Hospital , Columbus, Ohio Published online: 14 Aug 2009.
To cite this article: Patricia J. Estok RN, PhD, FAAN , Ellen B. Rudy RN, PhD, FAAN & Jean A. Just RN, MSN (1991) Body‐fat measurements and athletic menstrual irregularity, Health Care for Women International, 12:2, 237-248, DOI: 10.1080/07399339109515944 To link to this article: http://dx.doi.org/10.1080/07399339109515944
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
BODY-FAT MEASUREMENTS AND ATHLETIC MENSTRUAL IRREGULARITY Patricia J. Estok, RN, PhD, FAAN School of Nursing, Kent State University
Ellen B. Rudy, RN, PhD, FAAN School of Nursing, Case Western Reserve University
Jean A. Just, RN, MSN Downloaded by [Monash University Library] at 10:33 04 February 2015
Doctor's Hospital, Columbus, Ohio
The purpose of this study was to compare four measurements used as estimates of body fat that have been used in previous studies to determine whether the association between body fat and athletic menstrual irregularity (AMI) is measurement dependent. In a sample of 112 marathon runners, 94 responded to questions regarding their menstrual cycle. Of these, 30 (32%) reported irregular or absent menses, and 64 (68%) reported regular menses. Of the 30 subjects reporting menstrual irregularities or amenorrhea, 13 (43%) reported having had menstrual irregularity or amenorrhea prior to taking up running. Estimates of body fat were based on Mellits and Cheek's (1970) equation for estimating percentage of body water, Lutter and Cushman's (1982) height and weight categories, actual gross body weight, and weight loss of 10 lb (4.5 kg) or more since taking up the sport. In this sample of marathon runners, none of the four methods used to estimate body fat supported a relationship between menstrual irregularity and low body fat. A significant (p < .001) relationship was found between prior menstrual irregularity and the development of AMI after starting to run.
Athletic menstrual irregularity (AMI) has received widespread attention. Many factors have been studied for their effect on AMI, but one of particular interest is the proposed relationship between body fat and AMI. A basis for this concern is Frisch and McArthur's (1974) claim that menarche and the maintenance of regular ovulating cycles in women depend on the maintenance of a critical level of body fat. A major problem in drawing conclusions about the relationship beHealth Care for Women International, 12:237-248, 1991 Copyright © 1991 by Hemisphere Publishing Corporation
237
238
P. J. Estok et al.
tween AMI and body fat found in early survey studies is the lack of consistency in measures used to estimate body fat. To study the effect of this inconsistency, we took measures used in four different studies and applied them to a single sample of women runners. Our purpose in conducting this study was to determine whether inconsistencies in previous findings were measurement dependent. This article is a report of our findings.
Downloaded by [Monash University Library] at 10:33 04 February 2015
RESEARCH QUESTIONS The four measurements used to represent body fat were total body water content (Mellits & Cheek, 1970; Sanborn, Albrecht, & Wagner, 1984), height-to-weight ratio (Lutter & Cushman, 1982), gross body weight (Galle, Freeman, Galle, Huggins, & Sondheimer, 1983; Speroff & Redwine, 1980), and weight loss (Bachman & Kemmann, 1982; Speroff & Redwine, 1980). There were five research questions: 1. Is there a relationship between percentage of body fat estimated from total body water content and AMI in female marathon runners? 2. Is there a relationship between reported height-to-weight ratio and AMI in female marathon runners? 3. Do female marathon runners who weigh 115 lb (52.3 kg) or less have a higher incidence of AMI than those weighing more than 115 1b? 4. Is there a relationship between loss of 10 lb (4.5 kg) or more since onset of running and AMI in female marathon runners? 5. What is the relationship among weight loss of 10 lb (4.5 kg) or more, total body weight of 115 lb (52.3 kg) or less, height-toweight ratio, and percentage of body fat estimated from total body water content in female marathon runners? Body Water and Fat Content A widely used method of determining body fat is the formula provided by Mellits and Cheek (1970) that predicts total body water based on the subject's height and weight. Because body water constitutes 72% of fat-free mass in the adult, body fat can be estimated by simple arithmetic. The equation has been used by many researchers as a convenient means to estimate fat when more complicated body-composition estimation techniques are not possible. The Mellits and Cheek estimation is based on regression equations and should be applied to samples that are
Downloaded by [Monash University Library] at 10:33 04 February 2015
Athletic Menstrual Irregularity
239
similar in age, sex, and race to the sample from which the equations were derived. Scott (1984) cautioned the use of this formula for bodycomposition estimates when subjects are extremely thin or extremely fat. In a cross sample of 105 subjects, he found the Mellits and Cheek formula to overestimate fat in leaner women by as much as 25% and to underestimate fat percentages in fatter women. This finding causes particular concern for the use of Mellits and Cheek's formula to estimate body fat for female athletes because these women are generally more physically fit and leaner than the general population, thus making overestimation of body fat a possibility. Height-to-Weight Ratio Frisch and McArthur (1974) suggested that menarche and the maintenance of regular ovulatory cycles are dependent on a minimum heightto-weight ratio that represents a critical fat storage. This hypothesis is based on the fact that aromatization of androgens to estrogen takes place in adipose tissue and that athletes who require high energy outputs have a reduction in body fat. This effect of high energy drain combined with low body fat on menstrual dysfunction has been found in ballet dancers (Frisch, Wyshak, & Vincent, 1980a, 1980b; Warren, 1980) and runners (Lutter & Cushman, 1982; Schwartz et al., 1981; Speroff & Redwine, 1980). In addition, dancers, runners, and swimmers who began their training at young ages prior to menarche have been shown to have a delayed aged of menarche (Frisch et al., 1981; Warren, 1980). There are studies that do not fully support the relationship between height-to-weight ratio and menstrual dysfunction in female athletes. In a study of a group of middle-distance runners, there was no difference in height-to-weight ratio between runners with regular menstrual cycles and those with amenorrhea (Feicht, Johnson, Martin, Sparks, & Wagner, 1978). Wakat, Sweeney, and Rogol (1982) also found that normally menstruating female athletes in their study were indistinguishable in height and weight from those athletes reporting oligomenorrhea or amenorrhea. In addition to contradictory findings relating menstrual function to height-to-weight ratios, several authors have criticized the accuracy of the Frisch and McArthur (1974) nomogram used in earlier studies to determine percentage of body fat (Lutter & Cushman, 1982; Reeves, 1979; Scott & Johnson, 1985). Frisch and McArthur's nomogram for estimating body fat on the basis of the subject's height and weight is itself based on the formula by Mellits and Cheek (1970) that estimates total body water. Such an extrapolation makes the assumption that the Mellits and Cheek formula is correct, and the lower variance in Frisch
Downloaded by [Monash University Library] at 10:33 04 February 2015
240
P. J. Estok et al.
and McArthur's analysis of body fat is merely a statistical artifact of the procedure used to calculate fat from height and weight (Scott & Johnson, 1985). If the Mellits and Cheek formula overestimates fat content for leaner women (Scott, 1984), the resulting nomogram by Frisch and McArthur for height-to-weight ratio may also be faulty. Lutter and Cushman (1982) calculated body fat by underwater weighing and found the Frisch and McArthur (1974) nomogram to be inaccurate in its linear correlation between the two categories. Because of this finding and their dissatisfaction with other methods of establishing ideal weight, Lutter and Cushman developed body-weight categories based on body height that assigned subjects into low, medium, and high categories. Although these height-weight categories do not follow the nomogram provided by Frisch and McArthur, they are nonetheless based on a height-weight relationship, and such relationships may or may not reflect actual body fat. In their study of 350 marathon and 10-km runners, Lutter and Cushman found that runners with amenorrhea weighed the least and trained the most, but that height-to-weight ratio alone did not explain differences in menstrual functioning. Body Weight Gross measurement of body weight has been shown to be related to development of AMI. Warren (1980), in a study of 15 ballet dancers, found that dancers weighed significantly less than control subjects at every measured interval and that their menarche was delayed. Of the 13 subjects who reached menarche, 11 (84%) later developed amenorrhea. Energy drain due to strenuous practice and in combination with weight below a critical level was proposed as a possible causative factor. Speroff and Redwine (1980) surveyed approximately 900 recreational runners and reported that body weight less than 115 1b (52.3 kg) was associated with an increased incidence of oligomenorrhea or amenorrhea. They did not find that intensity of running (distance run per week) influenced menstrual frequency; however, only 18.6% of the participants ran more than 20 mi (32.2 km) per week. They also did not report on menstrual irregularity prior to taking up running. Galle et al. (1983), in a study of 105 runners, did not find a significant difference in mean body weight or in height-to-weight ratio of runners with regular, irregular, or amenorrheic cycles, but they did find that a significant number of amenorrheic runners weighed 115 1b (52.3 kg) or less compared with runners with regular cycles. Gross body weight has been criticized because of the variability in skeletal size of individuals, making height-toweight ratio a more acceptable measurement.
Athletic Menstrual Irregularity
241
Downloaded by [Monash University Library] at 10:33 04 February 2015
Weight Loss Another gross estimation of body fat is reported weight loss (Bachman & Kemmann, 1982; Dale, Gerlach, & Wilhite, 1970; Speroff & Redwine, 1980; Wentz, 1980). This method of estimating body fat has been criticized primarily because the loss of body fat when replaced by lean body mass may not be reflected in a gross measurement of bodyweight loss. However, because these data are easily retrievable from survey questionnaires, the measure has been used in studies of AMI, usually without benefit of other body-fat estimates for comparisons. OTHER MEASUREMENTS Other methods that have been used to estimate body-fat content include skinfold measurements (Schwartz et al., 1981; Wakat et al., 1982) body-impedance analysis (BIA; Hutcheson, Latin, & Berg, 1988; Kushner & Schoeller, 1986), and hydrostatic weighing (Boyden, Pamenter, Stanforth, Rotkis, & Wilmore, 1983; Sanborn et al., 1984; Sinning, Little, Wilson, & Bowers, 1984). Skinfold measurements and BIA require contact with each subject and cannot be calculated from survey data. Skinfold measurements are thought to be investigator dependent, with room for many subjective decisions, and may measure only subcutaneous fat (Frisch, 1984). BIA is based on the concept that the electrical conductivity of fat-free mass greatly exceeds that of fat mass because of the larger electrolyte content of lean tissue (Pethig, 1979). Correlations of BIA with underwater hydrostatic weighing have not supported the accuracy of this method for estimating body fat, and body-hydration levels have also caused problems with this method. Hydrostatic weighing is judged to be an accurate assessment of body fat but requires actual underwater weighing of subjects, thus limiting its usefulness for large numbers of runners. SUMMARY The hypothesis that there is a need for a critical level of stored body fat for the maintenance of menstrual function has resulted in several methods to estimate body fat in order to determine the correlation with menstrual dysfunction. There is criticism of all methods presently used for large groups of subjects (survey data). No one has examined the relationship between multiple methods of estimating body fat and menstrual irregularity in order to assess the possibility of measurementdependent outcomes.
242
P. J. Estok et al.
METHOD
Downloaded by [Monash University Library] at 10:33 04 February 2015
Sample The sample for this study was a subsample of a study by Estok and Rudy (1986), which included both male and female marathon runners. There were 112 women in the original study and they ranged in age from 18 to 53 years. Most (89%) were between 60 and 67 in. (150167.5 cm) tall; the range was 60-71 in. (150-177.5 cm). Their mean weight was 54.68 kg (SD = 6.15 kg), and their weight ranged from 42.6 to 76.2 kg. Most (n — 64; 57%) had been running 5 years or longer. Sixty-two percent (n = 69) reported running between 30 to 50 mi (48.3 to 80.5 km) per week, and 64 percent (n •= 72) stated they ran 6 or 7 days per week. Most (n = 85; 76%) reported 4-8 mi (6.4-12.9 km) as an average run. Ninety-six percent of the 112 women responded to questions regarding their menstrual cycle, and 2 subjects were eliminated because they were 50 years of age or older, leaving a sample of 94 women for this study. The age range for this subgroup was 18-47 (M •= 32). Of these 94 subjects, 30 (32%) reported irregular or absent menses, and 64 (68%) reported regular menses. Of the 30 women reporting postrunning menstrual irregularities or amenorrhea, 13 (43%) reported having had menstrual irregularity or amenorrhea prior to taking up the sport. Procedure and Instruments A questionnaire designed to obtain demographic information, running history, and health survey data was used. The estimate of body fat was based on actual body weight, weight loss (subtracting current weight from prerunning weight), Lutter and Cushman's (1982) height and weight categories, and Mellits and Cheek's (1970) equation for estimating percentage of body water. The Mellits and Cheek (1970) equation for estimating body water and percentage of body fat is as follows: Total body water (TBW)
10.33 + [0.252 weight (kg)] + [0.154 height (cm)]
TRW 1ÖW
body weight
% body water
Athletic Menstrual Irregularity
Downloaded by [Monash University Library] at 10:33 04 February 2015
% body fat - 1 0 0 -
243
% body water 0.72
Lutter and Cushman (1982) used a method for determining heightweight categories of low, medium, and high body weights. For example, a woman 60 in. (150 cm) tall who weighed 95 lb (43 kg) or less was placed in the low weight category; if her weight was 98-115 lb (44.552.3 kg), she was in the medium category; and if over 115 lb (52.5 kg), she was in the high weight category. This method is admittedly somewhat arbitrary but is based on Lutter and Cushman's dissatisfaction with other methods of weight classification and inaccuracies found in body fat estimates based on the Frisch and McArthur (1974) nomogram. They based their categories on underwater weighing of 20 female runners. Scoring of the variables was as follows: Runners who reported a weight loss of 10 lb (4.5 kg) or greater received à score of 1; those with less weight loss were scored 0. Runners weighing 115 lb (52.3 kg) or less were scored 0; those weighing more than 115 lb (52.3 kg) were scored 1. Height-weight categories based on Lutter and Cushman's (1982) work were scored as follows: low «= 1, medium = 2, and high — 3. The percentage of body fat was calculated from total body water according to Mellits and Cheek's (1970) formula and was scored as calculated. ANALYSIS OF RESEARCH QUESTIONS Question 1: Is there a relationship between percentage of body fat estimated from total body water content and AMI in female marathon runners? The percentage of body fat for the participants ranged from 19.8 to 35.2 (M = 26.3, SD 3.2). Runners with menstrual alterations had a mean percentage of body fat of 26.2 (SD 3.97); those without menstrual irregularity had a mean of 26.3 (SD 2.8). There was no significant difference between the groups, r(92) = 0.13. Question 2: Is there a relationship between reported height-to-weight ratio and AMI in female marathon runners? The women were assigned to low, medium, and high body-weight categories on the basis of a height-weight chart developed by Lutter and Cushman (1982). Chi-square analysis to examine the relationship between regular and irregular or absent menstrual cycles and high, medium, and low weight categories revealed a significant association among these variables, x2 (2, N = 94) •= 6.00, p < .05. Seven of the
244
P. J. Estok et al.
Downloaded by [Monash University Library] at 10:33 04 February 2015
14 women in the low weight category and 7 of the 14 women in the high weight category reported AMI. Many (3 of 4) of the amenorrheic women were in the low weight category. Post hoc analysis indicated that the low and high weight categories each contributed 35% to the chisquare values and that the medium category contributed 30%. This suggests that both high and low weight categories were associated with AMI. Question 3: Do female marathon runners who weigh 115 lb (52.3 kg) or less have a higher incidence of AMI than those weighing more than 115 Ib (52.3 kg)? Thirty-eight (40%) of the women reported weighing 115 lb (52.3 kg) or less; 56 (60%) were above that weight. There was no significant difference in AMI on the basis of the weight category of the subjects, x2 (\,N = 94) - 0.16. Question 4: Is there a relationship between loss of 10 lb (4.5 kg) or more since onset of running and AMI in female marathon runners? Of the 92 participants providing information about weight loss and menstrual cycle functions, 52 (56%) reported they had lost 10 or more pounds (4.5 kg) since beginning to run. The chi-square statistic indicated that the occurrence or nonoccurrence of a weight loss of 10 Ib or more was not significantly associated with menstrual cycle changes, x (\,N — 92) = 0.02. Although this association was not significant when amenorrhea and menstrual irregularities were considered together, 4 of the 5 amenorrheic women reported having lost 10 or more pounds (4.5 kg) since starting to run. Question 5: What is the relationship among weight loss of 10 Ib (4.5 kg) or more, total body weight of 115 lb (52.3 kg) or less, height-to-weight ratio, and percentage of body fat estimated from total body water content in female marathon runners? Pearson's product-moment and biserial correlations were used to examine these relationships. See Table 1 for the correlation matrix. Loss of 10 lb (4.5 kg) since .beginning to run reported by the female marathon runners explained little of the variation in the other measurements. A strong correlation was found between height-to-weight ratio and percentage body fat estimated from total body water content (r = .86), indicating that variations in one of these variables explained 74% of the variation in the other. There was also a moderate relationship between gross body weight and percentage of body fat estimated from total body water content (r — .59).
Athletic Menstrual Irregularity
245
Table 1. Correlations Between Weight Loss, Height-to-Weight Ratio, and Total Body Water Content in Female Marathon Runners Variable
Downloaded by [Monash University Library] at 10:33 04 February 2015
1. 2. 3. 4.
Weight loss 101b Weight: 115 lb Height-to-weight ratio Percentage of body fat
1
2
3
4
1.00
.12
-.06
-.07
.12
1.00
.43
-.06 -.07
.43 .59
1.00
.59 .86
.86
1.00
Note: N - 92 (2 women did not provide weight-loss data).
ADDITIONAL FINDING In the total sample of 94 women runners, 30 reported irregular or absent menses, with 13 of the 30 reporting menstrual irregularity prior to taking up running. One woman did not respond to the question on prior menstrual regularity. By a chi-square measurement of association, this relationship between prior menstrual irregularity was significantly related to present menstrual irregularity, x2 (1, N •= 93) = 20.87, p < .001. DISCUSSION The incidence of AMI, including amenorrhea for this group of female marathon runners, was 32%. This is in contrast to a 5% incidence of menstrual irregularities in the general population (Shangold, 1981) and to a 24% incidence of menstrual irregularities in female marathon runners reported by Shangold and Levine (1982). The 32% incidence is consistent with a 30% incidence of menstrual irregularities in female athletes reported by Galle et al. (1983). More than 40% of the women with AMI in our sample, however, reported menstrual irregularity prior to beginning the sport. In the sample of marathon runners, none of the four methods used to estimate body fat supported a clear relationship between menstrual irregularity and low body fat. This suggests that inconsistencies in earlier studies were not measurement dependent; furthermore, the low correlations between weight loss of 10 lb (4.5 kg) and the other measurements make weight loss of questionable value in estimating body fat. Perhaps highly conditioned marathon runners, who are likely to have developed muscle tissue that weighs more than fat, are particularly poor candidates for estimating body fat by loss of weight since beginning to run. Findings also support that Lutter and Cushman's (1982) formulas for height-
Downloaded by [Monash University Library] at 10:33 04 February 2015
246
P. J. Estok et al.
to-weight ratio and percentage of body fat based on Mellits and Cheek's (1970) formulas for total body water content are strongly correlated (r = .86), decreasing the likelihood that the use of these two measurements of body fat would produce inconsistent results. Lutter and Cushman's (1982) measurements indicated a significant relationship between weight and menstrual functioning in the 96 female marathon runners in the current sample; however, this difference was confounded by AMI's association with.both high and low weight categories. That is, runners in the high body-weight category were as likely to report AMI as those in the low weight categories; 50% of each reported menstrual irregularities. Amenorrhea, however, was more frequent in women in the low weight category. As with the Lutter and Cushman's study, height-to-weight ratio alone could not explain differences in menstrual functioning in this sample of women. It is noteworthy that so many (40%) women with AMI in this study reported having menstrual irregularity prior to running; specifically, they reported that their menstrual periods were occasionally skipped, irregular, or absent. Menstrual irregularity prior to running has been reported in other studies of female runners (Lutter & Cushman, 1982; Shangold & Levine, 1982); however, the majority of survey-type reports on female athletes have not reported pretraining menstrual patterns. Prior history of menstrual irregularity appears to be an important variable in relationship to menstrual irregularity in female runners. Such a finding makes it less likely that body fat alone or exercise alone is the causative factor in AMI. Three categories characterize women with AMI: those who develop menstrual irregularity after beginning regular exercise, those who have a lifelong pattern of menstrual irregularity, and those who have a prior history of menstrual irregularity and develop irregularities again after beginning to run. Failure to take this pattern into account may result in an inflated reporting of AMI. Our findings support those of authors who suggest that AMI incidence based on survey data may under- or overestimate a menstrual response to running. In the first instance, covert changes such as a shortened luteal phase or anovulation may not be reported (Shangold, 1981). In the second instance, lack of prerunning menstrual history data can imply an AMI effect of exercise that does not exist because menstrual irregularity may reflect a préexistent endocrine fragility and a lifelong history of menstrual irregularity. Third, self-reporting of menstrual irregularity is difficult to interpret because the basis for the assessment is not provided. The literature is replete with studies examining menstrual irregularity in the female athlete. Because of the hypothesis that body fat is related to the maintenance of menstrual functioning, authors have tried through
Downloaded by [Monash University Library] at 10:33 04 February 2015
Athletic Menstrual Irregularity
247
various indirect methods to assess the relationship between body fat and the incidence of menstrual irregularity in female athletes. In this study, regardless of the measurement used to reflect body fat, little relationship was found between menstrual irregularity and low body fat. The fact that the measurements, except for loss of 10 lb (4.5 kg) or more, were moderately to highly correlated indicates that measurements were not responsible for the inconsistent findings in the literature; that is, findings were not measurement dependent. Unlike other studies that have attempted to examine the relationship between body fat, intensity of exercise, and AMI, in our study the women were all at a high exercise intensity (marathon runners), making level of exercise less variable. Therefore, the belief that low body fat directly causes amenorrhea is simplistic and thus open to conflicting and ambiguous results. High levels of exercise, such as occur in ballet dancers and runners, may indeed result in low body fat but may also result in psychological stress and hormonal imbalance, each of which may contribute to menstrual cycle irregularity in these athletes. Furthermore, a prior history of menstrual irregularities appears to be an important factor that needs further exploration. When seeking explanation for menstrual irregularity in female athletes, a multifactional approach appears to be appropriate. REFERENCES Bachman, G. A., & Kemmann, E. (1982). Prevalence of oligomenorrhea and amenorrhea in a college population. American Journal of Obstetrics and Gynecology, 144, 98-102. Boyden, T. W., Pamenter, R. W., Stanforth, P., Rotkis, T., & Wilmore, J. H. (1983). Sex steroids and endurance running in women. Fertility and Sterility, 39, 629-632. Dale, E., Gerlach, D. H., & Wilhite, A. L. (1979). Menstrual dysfunction in distance runners. Obstetrics and Gynecology, 54, 47-53. Estok, P., & Rudy, E. (1986). Marathon running: Comparison of physical and psychological risks for men and women. Research in Nursing and Health, 10, 79-85. Feicht, C. B., Johnson, T. S., Martin, B. J., Sparks, K. E., & Wagner, W. W. (1978). Secondary amenorrhea in athletes. Lancet, 2, 1145-1146. Frisch, R. E. (1984). Body fat, puberty and fertility. Biological Review, 59, 161-188. Frisch, R. E., von Gotz-Welbergen, A., McArthur, J. W., Albright, T., Witschi, J., Bullen, B., Brinholz, J., Reed, R. B., & Hermann, H. (1981). Delayed menarche and amenorrhea of college athletes in relation to age of onset of training. Journal of the American Medical Association, 246, 1559-1563. Frisch, R. E., & McArthur, J. M. (1974). Menstrual cycles: Fatness as a determinant of minimum weight for height necessary for their maintenance or onset. Science, 185, 949-951.
Downloaded by [Monash University Library] at 10:33 04 February 2015
248
P. J. Estok et al.
Frisch, R. E., Wyshak, G., & Vincent, L. (1980a). Delayed menarche and amenorrhea of ballet dancers. New England Journal of Medicine, 303, 17-19. Frisch, R. E., Wyshak, G., & Vincent, L. (1980b). More on fatness and reproduction. New England Journal of Medicine, 303, 1125-1126. Galle, P. C , Freeman, E. W., Galle, M. G., Huggins, G. R., & Sondheimer, S. J. (1983). Physiologie and psychologie profiles in a survey of women runners. Fertility and Sterility, 39, 633-639. Hutcheson, L., Latin, R. W., & Berg, K. E. (1988). Body impedance analysis and body water loss. Research Quarterly for Exercise and Sport, 59(4), 359-362. Kushner, R. F., & Schoeller, D. A. (1986). Estimation of total body water by bioelectrical impedance analysis. American Journal of Clinical Nutrition, 44, 417-424. Lutter, J. M., & Cushman, S. (1982). Menstrual patterns in female runners. The Physician and Sportsmedicine, 10, 61-72. Mellits, E. D., & Cheek, D. B. (1970). The assessment of body water and fatness from infancy to adulthood. Monographs of Social Research on Child Development, 35,
12-26. Pethig, R. (1979). Dielectrical and electronic properties of biological materials. Chichester, England: Wiley. Reeves, J. (1979). Estimating fatness. Science, 204, 881. Sanborn, C. F., Albrecht, B. H., & Wagner, W. W. (1984). Athletic amenorrhea: The role of body fat. Medicine and Science in Sports and Exercise, 16, 118 (abstract). Schwartz, B., Cummings, D. C., Riordan, E., Selye, H., Yen, S. S. C , & Rebar, R. W. (1981). Exercise-associated amenorrhea: A distant entity. American Journal of Obstetrics and Gynecology, 141, 662-670. Scott, E. C. (1984). Estimation of total water and fatness from weight and height: Inaccurate for lean women. American Journal of Physical Anthropology, 64, 83-87. Scott, E. C , & Johnson, F. E. (1985). Science, nutrition, fat and policy: Test of the critical fat hypothesis. Current Anthropology, 26(4), 463-473. Shangold, M. (1981, June). Sports gynecology. The Runner, pp. 35-38. Shangold, M. M., & Levine, H. S. (1982). The effect of marathon training upon menstrual function. American Journal of Obstetrics and Gynecology, 143, 862-869. Sinning, W. E., Little, K. D., Wilson, J. R., & Bowers, B. M. (1984). Body composition and menstrual function in women athletes. Unpublished manuscript, Kent State University. Speroff, L., & Redwine, D. B. (1980). Exercise and menstrual function. The Physician and Sportsmedicine, 8, 42-52. Wakat, K. D., Sweeney, K. A., & Rogol, A. D. (1982). Reproductive system function in women cross-country runners. Medicine and Science in Sports and Exercise, 14, 263-269. Warren, M. P. (1980). The effects of exercise on pubertal progression and reproductive function in girls. Journal of Clinical Endocrinology and Metabolism, 51, 1150-1157. Wentz, A. C. (1980). Body weight and amenorrhea. Obstetrics and Gynecology, 143,862-869.
Health Care for Women International Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uhcw20
Body‐fat measurements and athletic menstrual irregularity a
b
Patricia J. Estok RN, PhD, FAAN , Ellen B. Rudy RN, PhD, FAAN & Jean A. Just RN, MSN a
School of Nursing , Kent State University ,
b
School of Nursing , Case Western Reserve University ,
c
c
Doctor's Hospital , Columbus, Ohio Published online: 14 Aug 2009.
To cite this article: Patricia J. Estok RN, PhD, FAAN , Ellen B. Rudy RN, PhD, FAAN & Jean A. Just RN, MSN (1991) Body‐fat measurements and athletic menstrual irregularity, Health Care for Women International, 12:2, 237-248, DOI: 10.1080/07399339109515944 To link to this article: http://dx.doi.org/10.1080/07399339109515944
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
BODY-FAT MEASUREMENTS AND ATHLETIC MENSTRUAL IRREGULARITY Patricia J. Estok, RN, PhD, FAAN School of Nursing, Kent State University
Ellen B. Rudy, RN, PhD, FAAN School of Nursing, Case Western Reserve University
Jean A. Just, RN, MSN Downloaded by [Monash University Library] at 10:33 04 February 2015
Doctor's Hospital, Columbus, Ohio
The purpose of this study was to compare four measurements used as estimates of body fat that have been used in previous studies to determine whether the association between body fat and athletic menstrual irregularity (AMI) is measurement dependent. In a sample of 112 marathon runners, 94 responded to questions regarding their menstrual cycle. Of these, 30 (32%) reported irregular or absent menses, and 64 (68%) reported regular menses. Of the 30 subjects reporting menstrual irregularities or amenorrhea, 13 (43%) reported having had menstrual irregularity or amenorrhea prior to taking up running. Estimates of body fat were based on Mellits and Cheek's (1970) equation for estimating percentage of body water, Lutter and Cushman's (1982) height and weight categories, actual gross body weight, and weight loss of 10 lb (4.5 kg) or more since taking up the sport. In this sample of marathon runners, none of the four methods used to estimate body fat supported a relationship between menstrual irregularity and low body fat. A significant (p < .001) relationship was found between prior menstrual irregularity and the development of AMI after starting to run.
Athletic menstrual irregularity (AMI) has received widespread attention. Many factors have been studied for their effect on AMI, but one of particular interest is the proposed relationship between body fat and AMI. A basis for this concern is Frisch and McArthur's (1974) claim that menarche and the maintenance of regular ovulating cycles in women depend on the maintenance of a critical level of body fat. A major problem in drawing conclusions about the relationship beHealth Care for Women International, 12:237-248, 1991 Copyright © 1991 by Hemisphere Publishing Corporation
237
238
P. J. Estok et al.
tween AMI and body fat found in early survey studies is the lack of consistency in measures used to estimate body fat. To study the effect of this inconsistency, we took measures used in four different studies and applied them to a single sample of women runners. Our purpose in conducting this study was to determine whether inconsistencies in previous findings were measurement dependent. This article is a report of our findings.
Downloaded by [Monash University Library] at 10:33 04 February 2015
RESEARCH QUESTIONS The four measurements used to represent body fat were total body water content (Mellits & Cheek, 1970; Sanborn, Albrecht, & Wagner, 1984), height-to-weight ratio (Lutter & Cushman, 1982), gross body weight (Galle, Freeman, Galle, Huggins, & Sondheimer, 1983; Speroff & Redwine, 1980), and weight loss (Bachman & Kemmann, 1982; Speroff & Redwine, 1980). There were five research questions: 1. Is there a relationship between percentage of body fat estimated from total body water content and AMI in female marathon runners? 2. Is there a relationship between reported height-to-weight ratio and AMI in female marathon runners? 3. Do female marathon runners who weigh 115 lb (52.3 kg) or less have a higher incidence of AMI than those weighing more than 115 1b? 4. Is there a relationship between loss of 10 lb (4.5 kg) or more since onset of running and AMI in female marathon runners? 5. What is the relationship among weight loss of 10 lb (4.5 kg) or more, total body weight of 115 lb (52.3 kg) or less, height-toweight ratio, and percentage of body fat estimated from total body water content in female marathon runners? Body Water and Fat Content A widely used method of determining body fat is the formula provided by Mellits and Cheek (1970) that predicts total body water based on the subject's height and weight. Because body water constitutes 72% of fat-free mass in the adult, body fat can be estimated by simple arithmetic. The equation has been used by many researchers as a convenient means to estimate fat when more complicated body-composition estimation techniques are not possible. The Mellits and Cheek estimation is based on regression equations and should be applied to samples that are
Downloaded by [Monash University Library] at 10:33 04 February 2015
Athletic Menstrual Irregularity
239
similar in age, sex, and race to the sample from which the equations were derived. Scott (1984) cautioned the use of this formula for bodycomposition estimates when subjects are extremely thin or extremely fat. In a cross sample of 105 subjects, he found the Mellits and Cheek formula to overestimate fat in leaner women by as much as 25% and to underestimate fat percentages in fatter women. This finding causes particular concern for the use of Mellits and Cheek's formula to estimate body fat for female athletes because these women are generally more physically fit and leaner than the general population, thus making overestimation of body fat a possibility. Height-to-Weight Ratio Frisch and McArthur (1974) suggested that menarche and the maintenance of regular ovulatory cycles are dependent on a minimum heightto-weight ratio that represents a critical fat storage. This hypothesis is based on the fact that aromatization of androgens to estrogen takes place in adipose tissue and that athletes who require high energy outputs have a reduction in body fat. This effect of high energy drain combined with low body fat on menstrual dysfunction has been found in ballet dancers (Frisch, Wyshak, & Vincent, 1980a, 1980b; Warren, 1980) and runners (Lutter & Cushman, 1982; Schwartz et al., 1981; Speroff & Redwine, 1980). In addition, dancers, runners, and swimmers who began their training at young ages prior to menarche have been shown to have a delayed aged of menarche (Frisch et al., 1981; Warren, 1980). There are studies that do not fully support the relationship between height-to-weight ratio and menstrual dysfunction in female athletes. In a study of a group of middle-distance runners, there was no difference in height-to-weight ratio between runners with regular menstrual cycles and those with amenorrhea (Feicht, Johnson, Martin, Sparks, & Wagner, 1978). Wakat, Sweeney, and Rogol (1982) also found that normally menstruating female athletes in their study were indistinguishable in height and weight from those athletes reporting oligomenorrhea or amenorrhea. In addition to contradictory findings relating menstrual function to height-to-weight ratios, several authors have criticized the accuracy of the Frisch and McArthur (1974) nomogram used in earlier studies to determine percentage of body fat (Lutter & Cushman, 1982; Reeves, 1979; Scott & Johnson, 1985). Frisch and McArthur's nomogram for estimating body fat on the basis of the subject's height and weight is itself based on the formula by Mellits and Cheek (1970) that estimates total body water. Such an extrapolation makes the assumption that the Mellits and Cheek formula is correct, and the lower variance in Frisch
Downloaded by [Monash University Library] at 10:33 04 February 2015
240
P. J. Estok et al.
and McArthur's analysis of body fat is merely a statistical artifact of the procedure used to calculate fat from height and weight (Scott & Johnson, 1985). If the Mellits and Cheek formula overestimates fat content for leaner women (Scott, 1984), the resulting nomogram by Frisch and McArthur for height-to-weight ratio may also be faulty. Lutter and Cushman (1982) calculated body fat by underwater weighing and found the Frisch and McArthur (1974) nomogram to be inaccurate in its linear correlation between the two categories. Because of this finding and their dissatisfaction with other methods of establishing ideal weight, Lutter and Cushman developed body-weight categories based on body height that assigned subjects into low, medium, and high categories. Although these height-weight categories do not follow the nomogram provided by Frisch and McArthur, they are nonetheless based on a height-weight relationship, and such relationships may or may not reflect actual body fat. In their study of 350 marathon and 10-km runners, Lutter and Cushman found that runners with amenorrhea weighed the least and trained the most, but that height-to-weight ratio alone did not explain differences in menstrual functioning. Body Weight Gross measurement of body weight has been shown to be related to development of AMI. Warren (1980), in a study of 15 ballet dancers, found that dancers weighed significantly less than control subjects at every measured interval and that their menarche was delayed. Of the 13 subjects who reached menarche, 11 (84%) later developed amenorrhea. Energy drain due to strenuous practice and in combination with weight below a critical level was proposed as a possible causative factor. Speroff and Redwine (1980) surveyed approximately 900 recreational runners and reported that body weight less than 115 1b (52.3 kg) was associated with an increased incidence of oligomenorrhea or amenorrhea. They did not find that intensity of running (distance run per week) influenced menstrual frequency; however, only 18.6% of the participants ran more than 20 mi (32.2 km) per week. They also did not report on menstrual irregularity prior to taking up running. Galle et al. (1983), in a study of 105 runners, did not find a significant difference in mean body weight or in height-to-weight ratio of runners with regular, irregular, or amenorrheic cycles, but they did find that a significant number of amenorrheic runners weighed 115 1b (52.3 kg) or less compared with runners with regular cycles. Gross body weight has been criticized because of the variability in skeletal size of individuals, making height-toweight ratio a more acceptable measurement.
Athletic Menstrual Irregularity
241
Downloaded by [Monash University Library] at 10:33 04 February 2015
Weight Loss Another gross estimation of body fat is reported weight loss (Bachman & Kemmann, 1982; Dale, Gerlach, & Wilhite, 1970; Speroff & Redwine, 1980; Wentz, 1980). This method of estimating body fat has been criticized primarily because the loss of body fat when replaced by lean body mass may not be reflected in a gross measurement of bodyweight loss. However, because these data are easily retrievable from survey questionnaires, the measure has been used in studies of AMI, usually without benefit of other body-fat estimates for comparisons. OTHER MEASUREMENTS Other methods that have been used to estimate body-fat content include skinfold measurements (Schwartz et al., 1981; Wakat et al., 1982) body-impedance analysis (BIA; Hutcheson, Latin, & Berg, 1988; Kushner & Schoeller, 1986), and hydrostatic weighing (Boyden, Pamenter, Stanforth, Rotkis, & Wilmore, 1983; Sanborn et al., 1984; Sinning, Little, Wilson, & Bowers, 1984). Skinfold measurements and BIA require contact with each subject and cannot be calculated from survey data. Skinfold measurements are thought to be investigator dependent, with room for many subjective decisions, and may measure only subcutaneous fat (Frisch, 1984). BIA is based on the concept that the electrical conductivity of fat-free mass greatly exceeds that of fat mass because of the larger electrolyte content of lean tissue (Pethig, 1979). Correlations of BIA with underwater hydrostatic weighing have not supported the accuracy of this method for estimating body fat, and body-hydration levels have also caused problems with this method. Hydrostatic weighing is judged to be an accurate assessment of body fat but requires actual underwater weighing of subjects, thus limiting its usefulness for large numbers of runners. SUMMARY The hypothesis that there is a need for a critical level of stored body fat for the maintenance of menstrual function has resulted in several methods to estimate body fat in order to determine the correlation with menstrual dysfunction. There is criticism of all methods presently used for large groups of subjects (survey data). No one has examined the relationship between multiple methods of estimating body fat and menstrual irregularity in order to assess the possibility of measurementdependent outcomes.
242
P. J. Estok et al.
METHOD
Downloaded by [Monash University Library] at 10:33 04 February 2015
Sample The sample for this study was a subsample of a study by Estok and Rudy (1986), which included both male and female marathon runners. There were 112 women in the original study and they ranged in age from 18 to 53 years. Most (89%) were between 60 and 67 in. (150167.5 cm) tall; the range was 60-71 in. (150-177.5 cm). Their mean weight was 54.68 kg (SD = 6.15 kg), and their weight ranged from 42.6 to 76.2 kg. Most (n — 64; 57%) had been running 5 years or longer. Sixty-two percent (n = 69) reported running between 30 to 50 mi (48.3 to 80.5 km) per week, and 64 percent (n •= 72) stated they ran 6 or 7 days per week. Most (n = 85; 76%) reported 4-8 mi (6.4-12.9 km) as an average run. Ninety-six percent of the 112 women responded to questions regarding their menstrual cycle, and 2 subjects were eliminated because they were 50 years of age or older, leaving a sample of 94 women for this study. The age range for this subgroup was 18-47 (M •= 32). Of these 94 subjects, 30 (32%) reported irregular or absent menses, and 64 (68%) reported regular menses. Of the 30 women reporting postrunning menstrual irregularities or amenorrhea, 13 (43%) reported having had menstrual irregularity or amenorrhea prior to taking up the sport. Procedure and Instruments A questionnaire designed to obtain demographic information, running history, and health survey data was used. The estimate of body fat was based on actual body weight, weight loss (subtracting current weight from prerunning weight), Lutter and Cushman's (1982) height and weight categories, and Mellits and Cheek's (1970) equation for estimating percentage of body water. The Mellits and Cheek (1970) equation for estimating body water and percentage of body fat is as follows: Total body water (TBW)
10.33 + [0.252 weight (kg)] + [0.154 height (cm)]
TRW 1ÖW
body weight
% body water
Athletic Menstrual Irregularity
Downloaded by [Monash University Library] at 10:33 04 February 2015
% body fat - 1 0 0 -
243
% body water 0.72
Lutter and Cushman (1982) used a method for determining heightweight categories of low, medium, and high body weights. For example, a woman 60 in. (150 cm) tall who weighed 95 lb (43 kg) or less was placed in the low weight category; if her weight was 98-115 lb (44.552.3 kg), she was in the medium category; and if over 115 lb (52.5 kg), she was in the high weight category. This method is admittedly somewhat arbitrary but is based on Lutter and Cushman's dissatisfaction with other methods of weight classification and inaccuracies found in body fat estimates based on the Frisch and McArthur (1974) nomogram. They based their categories on underwater weighing of 20 female runners. Scoring of the variables was as follows: Runners who reported a weight loss of 10 lb (4.5 kg) or greater received à score of 1; those with less weight loss were scored 0. Runners weighing 115 lb (52.3 kg) or less were scored 0; those weighing more than 115 lb (52.3 kg) were scored 1. Height-weight categories based on Lutter and Cushman's (1982) work were scored as follows: low «= 1, medium = 2, and high — 3. The percentage of body fat was calculated from total body water according to Mellits and Cheek's (1970) formula and was scored as calculated. ANALYSIS OF RESEARCH QUESTIONS Question 1: Is there a relationship between percentage of body fat estimated from total body water content and AMI in female marathon runners? The percentage of body fat for the participants ranged from 19.8 to 35.2 (M = 26.3, SD 3.2). Runners with menstrual alterations had a mean percentage of body fat of 26.2 (SD 3.97); those without menstrual irregularity had a mean of 26.3 (SD 2.8). There was no significant difference between the groups, r(92) = 0.13. Question 2: Is there a relationship between reported height-to-weight ratio and AMI in female marathon runners? The women were assigned to low, medium, and high body-weight categories on the basis of a height-weight chart developed by Lutter and Cushman (1982). Chi-square analysis to examine the relationship between regular and irregular or absent menstrual cycles and high, medium, and low weight categories revealed a significant association among these variables, x2 (2, N = 94) •= 6.00, p < .05. Seven of the
244
P. J. Estok et al.
Downloaded by [Monash University Library] at 10:33 04 February 2015
14 women in the low weight category and 7 of the 14 women in the high weight category reported AMI. Many (3 of 4) of the amenorrheic women were in the low weight category. Post hoc analysis indicated that the low and high weight categories each contributed 35% to the chisquare values and that the medium category contributed 30%. This suggests that both high and low weight categories were associated with AMI. Question 3: Do female marathon runners who weigh 115 lb (52.3 kg) or less have a higher incidence of AMI than those weighing more than 115 Ib (52.3 kg)? Thirty-eight (40%) of the women reported weighing 115 lb (52.3 kg) or less; 56 (60%) were above that weight. There was no significant difference in AMI on the basis of the weight category of the subjects, x2 (\,N = 94) - 0.16. Question 4: Is there a relationship between loss of 10 lb (4.5 kg) or more since onset of running and AMI in female marathon runners? Of the 92 participants providing information about weight loss and menstrual cycle functions, 52 (56%) reported they had lost 10 or more pounds (4.5 kg) since beginning to run. The chi-square statistic indicated that the occurrence or nonoccurrence of a weight loss of 10 Ib or more was not significantly associated with menstrual cycle changes, x (\,N — 92) = 0.02. Although this association was not significant when amenorrhea and menstrual irregularities were considered together, 4 of the 5 amenorrheic women reported having lost 10 or more pounds (4.5 kg) since starting to run. Question 5: What is the relationship among weight loss of 10 Ib (4.5 kg) or more, total body weight of 115 lb (52.3 kg) or less, height-to-weight ratio, and percentage of body fat estimated from total body water content in female marathon runners? Pearson's product-moment and biserial correlations were used to examine these relationships. See Table 1 for the correlation matrix. Loss of 10 lb (4.5 kg) since .beginning to run reported by the female marathon runners explained little of the variation in the other measurements. A strong correlation was found between height-to-weight ratio and percentage body fat estimated from total body water content (r = .86), indicating that variations in one of these variables explained 74% of the variation in the other. There was also a moderate relationship between gross body weight and percentage of body fat estimated from total body water content (r — .59).
Athletic Menstrual Irregularity
245
Table 1. Correlations Between Weight Loss, Height-to-Weight Ratio, and Total Body Water Content in Female Marathon Runners Variable
Downloaded by [Monash University Library] at 10:33 04 February 2015
1. 2. 3. 4.
Weight loss 101b Weight: 115 lb Height-to-weight ratio Percentage of body fat
1
2
3
4
1.00
.12
-.06
-.07
.12
1.00
.43
-.06 -.07
.43 .59
1.00
.59 .86
.86
1.00
Note: N - 92 (2 women did not provide weight-loss data).
ADDITIONAL FINDING In the total sample of 94 women runners, 30 reported irregular or absent menses, with 13 of the 30 reporting menstrual irregularity prior to taking up running. One woman did not respond to the question on prior menstrual regularity. By a chi-square measurement of association, this relationship between prior menstrual irregularity was significantly related to present menstrual irregularity, x2 (1, N •= 93) = 20.87, p < .001. DISCUSSION The incidence of AMI, including amenorrhea for this group of female marathon runners, was 32%. This is in contrast to a 5% incidence of menstrual irregularities in the general population (Shangold, 1981) and to a 24% incidence of menstrual irregularities in female marathon runners reported by Shangold and Levine (1982). The 32% incidence is consistent with a 30% incidence of menstrual irregularities in female athletes reported by Galle et al. (1983). More than 40% of the women with AMI in our sample, however, reported menstrual irregularity prior to beginning the sport. In the sample of marathon runners, none of the four methods used to estimate body fat supported a clear relationship between menstrual irregularity and low body fat. This suggests that inconsistencies in earlier studies were not measurement dependent; furthermore, the low correlations between weight loss of 10 lb (4.5 kg) and the other measurements make weight loss of questionable value in estimating body fat. Perhaps highly conditioned marathon runners, who are likely to have developed muscle tissue that weighs more than fat, are particularly poor candidates for estimating body fat by loss of weight since beginning to run. Findings also support that Lutter and Cushman's (1982) formulas for height-
Downloaded by [Monash University Library] at 10:33 04 February 2015
246
P. J. Estok et al.
to-weight ratio and percentage of body fat based on Mellits and Cheek's (1970) formulas for total body water content are strongly correlated (r = .86), decreasing the likelihood that the use of these two measurements of body fat would produce inconsistent results. Lutter and Cushman's (1982) measurements indicated a significant relationship between weight and menstrual functioning in the 96 female marathon runners in the current sample; however, this difference was confounded by AMI's association with.both high and low weight categories. That is, runners in the high body-weight category were as likely to report AMI as those in the low weight categories; 50% of each reported menstrual irregularities. Amenorrhea, however, was more frequent in women in the low weight category. As with the Lutter and Cushman's study, height-to-weight ratio alone could not explain differences in menstrual functioning in this sample of women. It is noteworthy that so many (40%) women with AMI in this study reported having menstrual irregularity prior to running; specifically, they reported that their menstrual periods were occasionally skipped, irregular, or absent. Menstrual irregularity prior to running has been reported in other studies of female runners (Lutter & Cushman, 1982; Shangold & Levine, 1982); however, the majority of survey-type reports on female athletes have not reported pretraining menstrual patterns. Prior history of menstrual irregularity appears to be an important variable in relationship to menstrual irregularity in female runners. Such a finding makes it less likely that body fat alone or exercise alone is the causative factor in AMI. Three categories characterize women with AMI: those who develop menstrual irregularity after beginning regular exercise, those who have a lifelong pattern of menstrual irregularity, and those who have a prior history of menstrual irregularity and develop irregularities again after beginning to run. Failure to take this pattern into account may result in an inflated reporting of AMI. Our findings support those of authors who suggest that AMI incidence based on survey data may under- or overestimate a menstrual response to running. In the first instance, covert changes such as a shortened luteal phase or anovulation may not be reported (Shangold, 1981). In the second instance, lack of prerunning menstrual history data can imply an AMI effect of exercise that does not exist because menstrual irregularity may reflect a préexistent endocrine fragility and a lifelong history of menstrual irregularity. Third, self-reporting of menstrual irregularity is difficult to interpret because the basis for the assessment is not provided. The literature is replete with studies examining menstrual irregularity in the female athlete. Because of the hypothesis that body fat is related to the maintenance of menstrual functioning, authors have tried through
Downloaded by [Monash University Library] at 10:33 04 February 2015
Athletic Menstrual Irregularity
247
various indirect methods to assess the relationship between body fat and the incidence of menstrual irregularity in female athletes. In this study, regardless of the measurement used to reflect body fat, little relationship was found between menstrual irregularity and low body fat. The fact that the measurements, except for loss of 10 lb (4.5 kg) or more, were moderately to highly correlated indicates that measurements were not responsible for the inconsistent findings in the literature; that is, findings were not measurement dependent. Unlike other studies that have attempted to examine the relationship between body fat, intensity of exercise, and AMI, in our study the women were all at a high exercise intensity (marathon runners), making level of exercise less variable. Therefore, the belief that low body fat directly causes amenorrhea is simplistic and thus open to conflicting and ambiguous results. High levels of exercise, such as occur in ballet dancers and runners, may indeed result in low body fat but may also result in psychological stress and hormonal imbalance, each of which may contribute to menstrual cycle irregularity in these athletes. Furthermore, a prior history of menstrual irregularities appears to be an important factor that needs further exploration. When seeking explanation for menstrual irregularity in female athletes, a multifactional approach appears to be appropriate. REFERENCES Bachman, G. A., & Kemmann, E. (1982). Prevalence of oligomenorrhea and amenorrhea in a college population. American Journal of Obstetrics and Gynecology, 144, 98-102. Boyden, T. W., Pamenter, R. W., Stanforth, P., Rotkis, T., & Wilmore, J. H. (1983). Sex steroids and endurance running in women. Fertility and Sterility, 39, 629-632. Dale, E., Gerlach, D. H., & Wilhite, A. L. (1979). Menstrual dysfunction in distance runners. Obstetrics and Gynecology, 54, 47-53. Estok, P., & Rudy, E. (1986). Marathon running: Comparison of physical and psychological risks for men and women. Research in Nursing and Health, 10, 79-85. Feicht, C. B., Johnson, T. S., Martin, B. J., Sparks, K. E., & Wagner, W. W. (1978). Secondary amenorrhea in athletes. Lancet, 2, 1145-1146. Frisch, R. E. (1984). Body fat, puberty and fertility. Biological Review, 59, 161-188. Frisch, R. E., von Gotz-Welbergen, A., McArthur, J. W., Albright, T., Witschi, J., Bullen, B., Brinholz, J., Reed, R. B., & Hermann, H. (1981). Delayed menarche and amenorrhea of college athletes in relation to age of onset of training. Journal of the American Medical Association, 246, 1559-1563. Frisch, R. E., & McArthur, J. M. (1974). Menstrual cycles: Fatness as a determinant of minimum weight for height necessary for their maintenance or onset. Science, 185, 949-951.
Downloaded by [Monash University Library] at 10:33 04 February 2015
248
P. J. Estok et al.
Frisch, R. E., Wyshak, G., & Vincent, L. (1980a). Delayed menarche and amenorrhea of ballet dancers. New England Journal of Medicine, 303, 17-19. Frisch, R. E., Wyshak, G., & Vincent, L. (1980b). More on fatness and reproduction. New England Journal of Medicine, 303, 1125-1126. Galle, P. C , Freeman, E. W., Galle, M. G., Huggins, G. R., & Sondheimer, S. J. (1983). Physiologie and psychologie profiles in a survey of women runners. Fertility and Sterility, 39, 633-639. Hutcheson, L., Latin, R. W., & Berg, K. E. (1988). Body impedance analysis and body water loss. Research Quarterly for Exercise and Sport, 59(4), 359-362. Kushner, R. F., & Schoeller, D. A. (1986). Estimation of total body water by bioelectrical impedance analysis. American Journal of Clinical Nutrition, 44, 417-424. Lutter, J. M., & Cushman, S. (1982). Menstrual patterns in female runners. The Physician and Sportsmedicine, 10, 61-72. Mellits, E. D., & Cheek, D. B. (1970). The assessment of body water and fatness from infancy to adulthood. Monographs of Social Research on Child Development, 35,
12-26. Pethig, R. (1979). Dielectrical and electronic properties of biological materials. Chichester, England: Wiley. Reeves, J. (1979). Estimating fatness. Science, 204, 881. Sanborn, C. F., Albrecht, B. H., & Wagner, W. W. (1984). Athletic amenorrhea: The role of body fat. Medicine and Science in Sports and Exercise, 16, 118 (abstract). Schwartz, B., Cummings, D. C., Riordan, E., Selye, H., Yen, S. S. C , & Rebar, R. W. (1981). Exercise-associated amenorrhea: A distant entity. American Journal of Obstetrics and Gynecology, 141, 662-670. Scott, E. C. (1984). Estimation of total water and fatness from weight and height: Inaccurate for lean women. American Journal of Physical Anthropology, 64, 83-87. Scott, E. C , & Johnson, F. E. (1985). Science, nutrition, fat and policy: Test of the critical fat hypothesis. Current Anthropology, 26(4), 463-473. Shangold, M. (1981, June). Sports gynecology. The Runner, pp. 35-38. Shangold, M. M., & Levine, H. S. (1982). The effect of marathon training upon menstrual function. American Journal of Obstetrics and Gynecology, 143, 862-869. Sinning, W. E., Little, K. D., Wilson, J. R., & Bowers, B. M. (1984). Body composition and menstrual function in women athletes. Unpublished manuscript, Kent State University. Speroff, L., & Redwine, D. B. (1980). Exercise and menstrual function. The Physician and Sportsmedicine, 8, 42-52. Wakat, K. D., Sweeney, K. A., & Rogol, A. D. (1982). Reproductive system function in women cross-country runners. Medicine and Science in Sports and Exercise, 14, 263-269. Warren, M. P. (1980). The effects of exercise on pubertal progression and reproductive function in girls. Journal of Clinical Endocrinology and Metabolism, 51, 1150-1157. Wentz, A. C. (1980). Body weight and amenorrhea. Obstetrics and Gynecology, 143,862-869.
