According to the new ACSM position stand, the female athlete triad is now defined as a condition involving three interrelated spectrums, namely, energy availability (defined as energy intake minus exercise energy expenditure), menstrual function, and bone mineral density (BMD) (23), with the athlete exhibiting the full triad presenting chronic low energy availability (LEA), functional hypothalamic amenorrhea, and osteopenia/osteoporosis (23). Although a female competitor participating in any sport may develop the conditions characteristic of the triad, previous studies have reported elevated prevalence estimates of menstrual irregularity (MI) and low BMD among collegiate and postcollegiate endurance runners (7,8,11,17). Laboratory studies have identified a direct effect of LEA on reducing estradiol, lowering bone deposition, and increasing bone resorption (15,19,33). Therefore, the elevated prevalence of MI and low BMD among young adult runners may be due, in part, to the high energy demands of running and the tendency of competitive runners to inadequately replenish calories expended after a training session or competitive event (18,28).
Adolescence is a period of rapid bone mineral accumulation where, under normal conditions, approximately 40-50% of adult bone mineral content (BMC) is accumulated and 90% of overall BMC is achieved by age 16.9 ± 1.3 yr (13,32). Because LEA has been reported to disrupt hormones that regulate bone metabolism and alter markers of bone turnover among young adult women (15), a chronic low-energy state occurring during adolescence may contribute to an inadequate bone mineral accumulation and a low peak bone mass. This effect may by irreversible (17) and may also increase the risk of an individual developing osteopenia or osteoporosis. Furthermore, adolescence marks the developmental stage where (except during pregnancy and lactation) energy needs are highest. Consequently, female adolescent endurance runners may represent a population at risk for developing LEA and may represent a group where LEA may have the most detrimental, long-term effects.
Most studies that have reported on the prevalence of individual or multiple components of the triad have examined collegiate and postcollegiate, recreational and elite athletes (7,8,11,14,21,29,30). Few reports exist that provide important information on the prevalence of low BMD or traits associated with low bone mass in adolescent runners. Identifying and understanding predictors of low bone mass among adolescent runners is important because these behaviors may be limiting bone mineral accumulation, which may lead to a reduced peak bone mass. Therefore, the purpose of our study was to determine the prevalence of and traits that may increase the risk of low BMD in a large sample of female adolescent endurance runners.
METHODS
Participants
A total of 106 female cross-country runners participating on interscholastic teams were recruited from six high schools in southern California during the 2004 cross-country season. Runners were included if they were 13 to 18 yr old and had initial onset of menarche, or, if not menstruating, were ages 15 to 18 yr (2,23). The study was approved by the University's Institutional Review Board. Written parental consent and subject assent were obtained from each participant before data collection. Data from athletes who reported taking any medications known to affect bone mass were excluded from analysis.
Data Collection
At the beginning of the cross-country season, the Eating Disorder Examination Questionnaire (EDE-Q), a menstrual history questionnaire, and a questionnaire assessing sports participation history were administered. We also measured each girl's height and weight without shoes (to the closest 0.5 inch and 0.5 lb, respectively). Body weight was measured using a digital scale (Health-O-Meter; Sunbeam Products, Inc, Bridgeview, IL).
Eating attitudes and behaviors.
The EDE-Q is composed of four subscales: weight concern, shape concern, eating concern, dietary restraint, and a global score (a composite mean score of the four subscales). Scores ranging from 0 to 6 on a Likert scale correspond to the number of days over the past 4 wk the respondent had experienced a specific attitude, feeling, or behavior. A mean cutoff of ≥3.0 was used to categorize runners as having an elevated value for each subscale. We chose this cutoff because it indicated that a specific attitude or behavior was reported on ≥14 of the previous 28 d and because it corresponded with the highest 6-18% of scores for each subscale. The EDE-Q has high internal consistency (20) and moderate to high concurrent and criterion validity (22). Furthermore, the EDE-Q has been validated in adolescent populations and has been reported to have high test-retest and interrater reliability in adolescent athletes (24,25). In addition to its psychometric properties, we chose the EDE-Q over other widely used questionnaires because it addressed a specific time frame and assessed the frequency of eating and/or pathologic behaviors.
Menstrual status.
The menstrual history questionnaire was derived from an athletic preparticipation medical history form (31). The criteria for classifying athletes with MI were as follows: primary amenorrhea (no onset of menarche by age 15 yr), secondary amenorrhea (absence of three consecutive menstrual cycles in the past year), or oligomenorrhea (interval between menses <21 d and >35 d in the past year) (1,2,23). Because the descriptive traits for athletes with oligomenorrhea and amenorrhea were similar, runners with MI were combined into a single (oligo/amenorrheic) group and compared to runners with normal menses (eumenorrheic).
At the time of data collection, three runners, aged <15 yr, had recently began menstruating (within ≤8 months) and reported that the interval between one and more of their cycles was >35 d. Because a menstrual cycle length of >35 d is typical during the first year of a female's menstrual cycle (2), we did not classify these girls with MI.
Bone mineral density (BMD).
Two to four weeks after the administration of the questionnaires, the runners received a dual-energy x-ray absorptiometry (DXA) BMD scan. Areal BMD (g·cm−2) at the spine (L1-L4), proximal femur, and total body and body composition were assessed by DXA using a Lunar DPX-NT densitometer (Lunar/GE Corp, Madison, WI). Quality assurance tests were performed each morning of testing. The coefficient of variation in BMD in our laboratory is 0.6% for the total hip, 1.2% for the spine (L1-L4), and 0.99% for total body. Runners were classified as having low bone mass for their age if their values at the spine or total body were 1 or 2 SD or more below the age-matched, gender-specific reference data from the GE/Lunar pediatric database (z-score of ≤−1 or ≤−2, respectively) (23). At the time of data collection, z-scores for the hip were not available for children.
Statistical Analyses
Mean and SE were used to describe characteristics of the overall study sample. Mean differences of selected physical and performance traits by BMD status were determined by ANOVA. The ANOVA was also used to evaluate the mean differences of selected physical characteristics, BMD and BMD z-scores, and running performance traits by menstrual function status in the past year. ANCOVA were performed to determine mean differences of BMD and total body and lumbar spine BMD z-scores, adjusting for age, body mass index (BMI), and lean tissue mass (total body, total hip, and lumbar spine BMD) and BMI and lean tissue mass (total body and lumbar spine z-scores). Bonferroni pairwise comparisons were used for ANOVA and ANCOVA analyses. Pearson's correlations were conducted between the number of menstrual cycles in the past year, summer mileage, lifetime seasons, and BMD z-scores. Multivariate logistic regression was then fit to determine the final adjusted risk models for predictors of low BMD at z-scores of −1 SD and −2 SD levels.
RESULTS
Descriptive characteristics.
Of the 106 runners assessed, complete data were obtained from 93 girls (87.7%). Descriptive characteristics of the entire sample are presented in Table 1. On average, although runners exhibited a normal age at menarche, they reported having only 10 menstrual cycles in the past year. The runners' mean BMI and percent body fat levels were also within normal ranges.
;)
TABLE 1: Descriptive characteristics among 93 adolescent runners (mean ± SE).
Bone mineral density.
Among our sample of adolescent runners, low BMD was determined using −1 and −2 z-score criteria. The prevalence of −1 and −2 z-score criteria for low BMD was 39.8% (n = 37) and 11.8% (n = 11), respectively. Figure 1 depicts the prevalence of low BMD among our sample of adolescent runners compared with prevalence estimates of low BMD in young collegiate and postcollegiate runners and athletes participating in nonrunning sports. Our sample of adolescent runners (N = 93) demonstrated a 1.8- and a 4.1-times higher prevalence of low BMD using the −1 z-score cutoff and a 2.9- and a 6.6-times higher prevalence using the −2 z-score cut-points compared with the adolescent high school (24) and US collegiate athletes (7) participating in various lean-build and non-lean-build sports (Fig. 1).
FIGURE 1: Prevalence of low
bone mass (determined by a low total body or lumbar spine
z-score) of current study sample of female adolescent endurance runners (
N = 93) compared with the prevalence estimates reported in other competitive female running and sport populations using the
z-score and
t-score cut-points of −1, −2, and −1, −2.5, respectively (
7,8,24,29).
1Prevalence of low BMD in the current sample.
Table 2 depicts runners' physical characteristics and performance traits by their BMD status. Runners with normal BMD exhibited significantly higher weight, BMI, lean tissue mass, and percent body fat than runners in the two low BMD groups (Table 2). Runners in the low (≤−2) BMD group had significantly fewer menses in the past year than runners with their lowest BMD z-score between −1 and −2 and normal BMD (Table 2). In addition, runners in the low BMD (≤−2) group ran more miles over the three summer months before data collection and displayed a trend toward participating in a higher number of lifetime competitive endurance running seasons (P = 0.06) than runners with normal BMD.
TABLE 2: Mean ± SE physical characteristics and performance traits among runners by bone mass status (N = 93).a,b
Menstrual function and BMD.
In our sample, 25.8% of the runners reported MI. Of 24 runners with MI, 3.2% (n = 3) were classified with primary amenorrhea, 17.2% (n = 16) with secondary amenorrhea, and 5.4% (n = 5) with oligomenorrhea. Table 3 presents the difference in physical characteristics and bone mass values by menstrual status. Runners with MI exhibited significantly lower weight, BMI, lean tissue mass, percent body fat, gynecological age, and fewer menses during the past year. In addition, BMD and BMD z-score values were significantly lower in runners with MI at each bone site. However, after adjusting for age, BMI, and lean tissue mass (for BMD values) and BMI and lean tissue mass (for BMD z-score values), only total hip BMD and lumbar spine BMD and BMD z-score group differences remained significant. No significant differences were identified for lifetime seasons run or summer mileage between menstrual groups.
TABLE 3: Mean ± SE physical characteristics, bone mass values, and performance traits according to runners' menstrual status (N = 93).a
In our sample of adolescent runners, total body and lumbar spine BMD z-score exhibited a direct relationship with number of menstrual cycles in the past year (r = 0.33, P < 0.005 and r = 0.37, P < 0.001 for total body and lumbar spine, respectively). Runners reporting zero menstrual cycles in the past year exhibited a lower lumbar spine and total body BMD z-score than runners reporting 10 or more cycles in the past year (Fig. 2). Furthermore, runners reporting one to three cycles in the past year displayed a trend (P = 0.08) toward a lower lumbar spine BMD z-score than girls reporting 10 or more cycles (Fig. 2).
FIGURE 2: Total body and lumbar spine BMD z-scores by number of menstrual cycles in the past year. ANOVA were used to compare mean differences in BMD z-score. a,bSimilar letters indicate significant differences. Lumbar spine and total body BMD z-scores were significantly lower in runners with 0 cycles in the past year compared with those reporting 10 or more cycles in the past year (P < 0.05). N = 90, three girls who recently began menstruating (≤8 months, n = 3) were excluded.
Running and BMD.
Figure 3 depicts the inverse relationship between lifetime seasons of endurance running and total body and lumbar spine BMD z-score after adjusting for number of menses in the past year, BMI, and lean tissue mass (r = −0.23, P < 0.05 and r = −0.31, P < 0.005, respectively). Mean total body and lumbar spine BMD z-scores were significantly different between girls that participated in less than three and five or morelifetime seasons of endurance running (Fig. 3). Furthermore, the number of miles run over the three summer months before data collection was inversely associated with lumbar spine BMD z-score (r = −0.24, P< 0.05).
FIGURE 3: Adjusted total body and lumbar spine BMD z-score values according to the number of seasons of lifetime competitive endurance running. Z-score values are adjusted for BMI, lean tissue mass, and number of menstrual cycles in the past year. Runners that participated in five or more compared with less than three seasons of endurance running had significantly lower total body and lumbar spine BMD z-score values. Similar letters indicate significant differences, a P < 0.01, b P < 0.05, ANCOVA.
Predictors of low BMD.
After adjusting for age, ethnicity, lean tissue mass, BMI, gynecological age, and elevated dietary restraint (a score ≥3.0 on the EDE-Q dietary restraint subscale), our adjusted model indicated that MI and lifetime participation in five or more (compared with less than five) seasons of endurance running significantly increased the likelihood of low BMD at the −1 z-score criterion (Table 4). In addition, BMI and lean tissue mass were found to be inversely associated with low BMD at the −1 z-score criterion. Similarly, after controlling for age, ethnicity, lean tissue mass, BMI, and gynecological age, our adjusted model indicated that only MI and lifetime participation in five or more seasons of endurance running were associated with low BMD at the −2 z-score criterion (Table 4).
TABLE 4: Adjusted odds ratios for potential risk factors of low BMD in adolescent runners (N = 93).a - c
DISCUSSION
The prevalence estimates of low BMD, using −1 and −2 z-score criterion points (39.8% and 11.8%, respectively), found in our sample of endurance runners are considerably higher than estimates previously reported for adolescent, collegiate, and postcollegiate athletes participating in other sports. The percentage of our runners who met the −1 criterion for low BMD is approximately twice that reported in a previous study by our research group among a sample of high school athletes representing multiple sports (24). Our findings are also higher than those reported by Beals and Hill and by Torstveit and Sundgot-Borgen who used a BMD z-score (7) or t-score (29) criterion of −1, and both reported an approximately 10% prevalence of low BMD in athletes who competed in various lean-build and non-lean-build sports. The reduced prevalence of low BMD in the sports considered "nonrunning" may be partially due to their less energy-demanding nature or that athletes who participate in non-lean-build sports may put less emphasis on obtaining or maintaining a lean or thin body frame. In addition, sports such as volleyball, soccer, and tennis involve high-impact and high-variable strain on bone, which are more effective in promoting bone mass gains than lower-impact repetitive loading activities (10,29).
Therefore, our findings suggest that female adolescent runners may represent a population at risk for developing, potentially irreversible low BMD. Given the cross-sectional nature of our study, we are unclear whether the low BMD in these runners was a result of bone loss or inadequate bone mineral accrual. However, because our sample of runners are still within their critical window of bone mineral accrual (13,32), they may have the potential to increase their BMD to normal levels if they begin practicing behaviors that promote bone mass gains. Future studies are necessary to evaluate this possibility.
Approximately 26% of the high school runners reported MI. Because a long interval between cycles occurs most often within the first 3 yr postmenarche and because a normal cycle length typically occurs near the sixth gynecologic year (2), it is unclear whether our sample exhibited an elevated prevalence of MI. Although according to the American Academy of Pediatrics, the absence of menses for more than 3 months or persistent oligomenorrhea is considered abnormal (1). Our finding is similar to estimates of MI reported in adolescent competitors (24%) (24), Norwegian elite athletes participating in leanness sports (28%) (30), and those reported in US collegiate and postcollegiate elite cross-country runners (36%) (8) but was less than those reported among elite women runners (66%) (11). Since approximately 98% of girls reach menarche by age 15 (2), our runners' 12.9% occurrence of late age at menarche (at ≥15 yr) is higher than would be expected in a group of normal age-matched peers. It is important to note that only clinical MI was measured and that we did not assess subclinical menstrual disturbances, such as anovulation or shortened luteal phase length. Thus, our 26% estimate of MI may underestimate the actual occurrence of menstrual dysfunction in these runners.
Although our study is consistent with numerous studies that have reported a negative association between MI and bone mass (3,6,8,11,16,24,29), our findings also indicate a negative relationship between participation in competitive endurance running and bone mass, independent of menstrual function. Our finding is consistent with Petit et al. (26), who observed a significant inverse relationship between kilometers run per week and change in BMD during a 1-yr period among premenopausal runners with normal menstrual cycle lengths. Interestingly, Petit et al. (26) found this relationship in women with shortened luteal phase lengths but not among normally ovulating women. Therefore, among our sample of adolescent runners, subclinical menstrual disturbances (which have previously been associated with bone loss) may have been associated with the inverse relationship between lifetime seasons of endurance running and bone mass.
The negative relationship between endurance running and bone mass may be, at least, partially related to the high amount of energy expended during endurance running. Thus, if a female endurance runner's diet is not adequate to support her high energy needs, she may enter a state of LEA. If this pattern is continued over time, the chronic energy deficit may affect bone mass. This is supported by studies of young adult women, by Loucks et al. (15,19), that have identified LEA (≤30 kcal·kg−1 FFM) inducing changes in reproductive and growth hormones and markers of bone turnover in ways that promote bone loss. Therefore, among adolescent runners, LEA may also promote bone loss or inadequate bone mineral accumulation. However, the direct effects of LEA on markers of bone turnover or bone mass have not yet been assessed in adolescents.
The negative relationship between seasons of endurance running and BMD identified in our study is opposite those reported in previous studies that suggested a positive relationship between years and/or hours of participation and bone mass (4,5). Although we observed this negative relationship, we do not believe that endurance running has direct negative effects on bone health. Previous studies have reported higher BMD values at bone sites exposed to high impact, namely, the trochanter, the femoral neck, and the total lower body, in young adult distance runners when compared with sedentary controls (9). Increases in cancellous bone of the lumbar spine have been reported after 1 yr of running among normally ovulatory women (26). In addition, a higher distal tibial bone strength index has been reported in adolescent middle-distance runners compared with nonrunners (12). Therefore, we do not interpret our findings to suggest that endurance running itself exerts detrimental bone health effects; rather, other factors associated with this sport, such as those promoting LEA, may contribute to the relationship between increased participation and low bone mass.
Limitations.
To increase the likelihood of obtaining accurate responses, we incorporated protocols to ensure that the athletes understood the confidentiality of their responses, and clear definitions of potentially confusing terminology, such as binge eating, and so on, were provided before the questionnaires were administered. Despite these precautions, our subjects may have misreported their responses due to the sensitive nature of the questions. We recognize that our cross-sectional study design did not allow us to measure any direct causal relationships. Yet, we indirectly tried to use measures indicative of usual patterns of health (e.g., number of menstrual cycles in the past year, training volume from the previous 3 months, etc.) in determining factors associated with low BMD. Furthermore, the lack of an age-matched nonathlete control group limited our ability to compare the prevalence of low BMD among the runners to nonathletes. However, z-scores in our sample were calculated from the GE/Lunar pediatric database, and in a normal population distribution, the expected prevalence estimates of a z-score ≤−1 and ≤−2 were 15.9% and 2.3%, respectively. These estimates are considerably lower than the low BMD estimates observed our sample of young runners.
CONCLUSION
Our findings recognize that female adolescent endurance runners may represent a population with an increased risk of low bone mass. The factors significantly associated with low BMD in our sample of adolescent runners included MI, long-term participation in competitive endurance running, BMI, and lean tissue mass. We recommend that adolescent runners be periodically screened and monitored during competition in their sport to detect disturbances in their menstrual patterns. Because a history of long distance running might indicate a history of a chronic energy deficit, the nutritional and training regimens of these runners should be explored to determine whether appropriate eating and training practices are being followed. Because the research among adolescents is limited, future studies are needed to identify better the factors contributing to low bone mass in this population and to identify behavioral strategies that promote optimal bone mineral accumulation among competitive endurance runners during the adolescent years.
The authors thank the National Athletic Trainers Association and the San Diego State University, Department of Exercise and Nutritional Sciences, and Fred Kasch Endowment for funding this research study. The authors also acknowledge Mandra Lawson, Shoshi Barkai, and Kylie Edwards for their assistance with data collection and data entry. None of the authors had personal or financial conflicts of interest.
REFERENCES
1. American Academy of Pediatrics. Committee on Sports Medicine and Fitness. Medical concerns in the female athlete.
Pediatrics. 2000;106:610-3.
2. American Academy of Pediatrics. Committee on Adolescence, American College of Obstetricians and Gynecologists and Committee on Adolescent Health Care.
Pediatrics. 2006;118:2245-50.
3. Baer JT, Taper LJ, Gwazdauskas FG, et al. Diet, hormonal, and metabolic factors affecting bone mineral density in adolescent amenorrheic and eumenorrheic female runners.
J Sports Med Phys Fitness. 1992;32:51-8.
4. Bailey DA, McKay HA, Mirwald RL, Crocker PR, Faulkner RA. A six year longitudinal study of the relationship of physical activity to bone mineral accrual in growing children: the University of Saskatchewan Bone Mineral Accrual Study.
J Bone Miner Res. 1999;14:1672-9.
5. Barkai HS, Nichols JF, Rauh MJ, Barrack MT, Lawson MJ, et al. Influence of sports participation and menarche on bone mineral density of female
high school athletes.
J Sci Med Sport. 2007;10:170-9.
6. Barr SI, Prior JC, Vigna YM. Restrained eating and ovulatory disturbances: possible implications for bone health.
Am J Clin Nutr. 1994;59:92-7.
7. Beals KA, Hill AK. The prevalence of disordered eating, menstrual dysfunction, and low bone mineral density among US collegiate athletes.
Int J Sport Nutr Exerc Metab. 2006;16:1-23.
8. Cobb KL, Bachrach LK, Greendale G, et al. Disordered eating, menstrual irregularity, and bone mineral density in female runners.
Med Sci Sports Exerc. 2003;35(5):711-9.
9. Egan E, Reilly T, Giacomoni M, Redmond L, Turner C. Bone mineral density among female sports participants.
Bone. 2006;38:227-33.
10. Fehling PC, Alekel L, Clasey J, Rector A, Stillman RJ. A comparison of bone mineral densities among female athletes in impact loading and active loading sports.
Bone. 1995;17:205-10.
11. Gibson JH, Mitchell A, Harries MG, Reeve J. Nutritional and exercise-related determinants of bone density in elite female runners.
Osteoporos Int. 2004;15:611-8.
12. Greene DA, Naughton GA, Briody JN, Kemp A, Woodhead H. Assessment of bone strength at differentially-loaded skeletal regions in adolescent middle-distance runners.
J Sci Med Sport. 2006;9:221-30.
13. Heaney RP, Abrams S, Dawson-Hughes B, et al. Peak
bone mass.
Osteoporos Int. 2000;11:985-1009.
14. Hulley AJ, Hill AJ. Eating disorders and health in elite women distance runners.
Int J Eat Disord. 2001;30:312-7.
15. Ihle R, Loucks AB. Dose-response relationships between energy availability and bone turnover in young exercising women.
J Bone Miner Res. 2004;19:1231-40.
16. Kaga M, Takahashi K, Ishihara T, et al. Bone assessment of female long-distance runners.
J Bone Miner Metab. 2004;22:509-13.
17. Keen AD. Irreversible bone loss in former amenorrheic athletes.
Osteoporos Int. 1997;7:311-5.
18. Kopp-Woodroffe SA, Manore MM, Dueck CA, Skinner JS, Matt KS. Energy and nutrient status of amenorrheic athletes participating in a diet and exercise training intervention program.
Int J Sport Nutr. 1999;9:70-88.
19. Loucks AB, Thuma JR. Luteinizing hormone pulsatility is disrupted at a threshold of energy availability in regularly menstruating women.
J Clin Endocrinol Metab. 2003;88:297-311.
20. Luce KH, Crowther JH. The reliability of the eating disorder examination-self-report questionnaire version.
Int J Eat Disord. 1999;25:349-51.
21. McLean JA, Barr SI, Prior JC. Dietary restraint, exercise, and bone density in young women: are they related?
Med. Sci. Sports Exerc. 2001;33(8):1292-6.
22. Mond JM, Hay PJ, Rodgers B, Owen C, Beumont PJ. Validity of the Eating Disorder Examination Questionnaire in screening for eating disorders in community samples.
Behav Res Ther. 2004;42:551-67.
23. Nattiv A, Loucks AB, Manore MM, Sanborn CF, Sundgot-Borgen J, et al. American College of Sports Medicine position stand. The
female athlete triad.
Med Sci Sports Exerc. 2007;39(10):1867-82.
24. Nichols JF, Rauh MJ, Lawson MJ, Ji M, Barkai HS. Prevalence of the
Female Athlete Triad syndrome among
high school athletes.
Arch Pediatr Adolesc Med. 2006;160:137-42.
25. Passi VA, Bryson SW, Lock J. Assessment of eating disorders in adolescents with anorexia nervosa.
Int J Eat Disord. 2003;33:45-54.
26. Petit MA, Prior JC, Barr SI. Running and ovulation positively change cancellous bone in premenopausal women.
Med Sci Sports Exerc. 1999;31(6):780-7.
27. Taaffe DR, Robinson TL, Snow CM, Marcus R. High-impact exercise promotes bone gain in well-trained female athletes.
J Bone Miner Res. 1997;12:255-60.
28. Tomten SE, Hostmart AT. Energy balance in weight stable athletes with and without menstrual disorders.
Scand J Med Sci Sports. 2006;16:127-33.
29. Torstveit MK, Sundgot-Borgen J. Low bone mineral density is two to three times more prevalent times more prevalent in non-athletic premenopausal women than in elite athletes: a comprehensive controlled study.
Br J Sports Med. 2005;39:282-7.
30. Torstveit MK, Sundgot-Borgen J. Participation in leanness sports but not training volume is associated with menstrual dysfunction: a national survey or 1276 elite athletes and controls.
Br J Sports Med. 2005;39:141-7.
31. Van de Loo DA, Johnson MD. The young female athlete.
Clin Sports Med. 1995;14:687-707.
32. Weaver CM. Adolescence: the period of dramatic bone growth.
Endocrine. 2002;17:43-8.
33. Williams NI, Helmreich DL, Parfitt DB, Caston-Balderrama A, Cameron JL. Evidence for a causal role of low energy availability in the induction of menstrual cycle disturbances during strenuous exercise training.
J Clin Endocrinol Metab. 2001;86:5184-93.
34. Yeager KK, Agostini R, Nattiv A, Drinkwater B. The
female athlete triad: disordered eating, amenorrhea, osteoporosis.
Med Sci Sports Exerc. 1993;25(7):775-7.
