Journal Logo


Hyperandrogenism May Explain Reproductive Dysfunction in Olympic Athletes


Author Information
Medicine & Science in Sports & Exercise: June 2009 - Volume 41 - Issue 6 - p 1241-1248
doi: 10.1249/MSS.0b013e318195a21a
  • Free


In many sport disciplines, including those involving endurance, esthetic factors, and weight classes, low body weight and fat mass are essential for optimal performance. It is well known that excessive efforts to lower body fat content are common among participants in such sports and may lead to eating disorders and chronic energy deficiency, with ensuing major medical problems such as disturbances of the menstrual cycle, decreased bone formation, and enhanced bone resorption (5,6,16).

In 1993, the concept of the female athlete triad was introduced as an association of eating disorders, amenorrhea, and osteoporosis (36). This triad is also associated with an increased incidence of musculoskeletal injuries and higher risk for stress fractures (13). In 2007, the American College of Sports Medicine (ACSM) redefined the concept of the triad as a set of physiological mechanisms relating energy availability to menstrual function and bone density (22).

The hormonal status characteristic of energy deficiency includes low circulating levels of insulin and insulin-like growth factor I (IGF-I), together with elevated levels of IGF-binding protein (IGFBP-1) and cortisol (15,17,28). The lowered activity of IGF-I attenuates stimulation of the hypothalamic-pituitary-gonadal (HPG) axis, whereas the enhanced secretion of cortisol inhibits hypothalamic secretion of gonadotropin-releasing hormone, effects which, in combination, might explain the suppression of the pulsatile secretion of circulating levels of luteinizing hormone (LH) in amenorrheic athletes (15,17,28).

It is necessary to remember, however, that not all female athletes who exhibit a disturbance in their menstrual cycle are in a catabolic state, suggesting the possible involvement of factors other than hypothalamic inhibition. We have recently demonstrated that essential hyperandrogenism, such as that associated with the polycystic ovary syndrome (PCOS), represents an alternative cause of menstrual disorders in athletes (27,28). PCOS, the most common hormonal aberration in women of fertile age, is characterized by anovulation, manifestations of hyperandrogenism, and polycystic ovarian morphology (8). In a subgroup of oligo-/amenorrheic endurance athletes, we observed enhanced diurnal secretion of testosterone and LH, low circulating levels of sex hormone-binding globulin (SHBG), and the highest frequency of polycystic ovaries (PCO). Hyperandrogenism in athletes seems to improve physical performance and to provide protection from the metabolic complications of estrogen deficiency (27).

Despite the importance of endocrine balance and optimal body composition to elite athletic performance, these parameters have not yet been explored fully in female Olympic athletes. The purpose of this study was to investigate the menstrual status, body composition, and biomarkers of energy availability in 90 female Olympic athletes participating in power, endurance, and technical sport disciplines. Furthermore, we aimed to clarify the endocrine mechanisms underlying menstrual disturbances in these women.



The recruitment of subjects was initiated at the Swedish pre-Olympic training camp in 2001, where all of the female athletes received written information concerning this study. Subsequently, several of the athletes of at least 18 yr of age and considered to be representative of the Swedish participation in the upcoming summer and Winter Olympic Games (11) were approached in person. Of those invited to participate, 22 (20%) declined, giving a busy schedule, together with living, training, and competing at distant locations as the primary reasons. In the end, 90 of Sweden's top elite female athletes were included in this study. The study protocol was approved by the local ethics committee of the Karolinska Hospital, and informed consent was given by all the subjects.

These athletes participated in a total of 27 different sport disciplines, which were divided into three groups (Table 1): Power (i.e., disciplines involving high mechanical load and short bursts of intense exertion), Endurance (involving prolonged periods of submaximal exertion and low mechanical load), and Technical (with an emphasis on technical skills and low demands on physical exertion and mechanical load).

Classification of the Olympic sport events in which our subjects participated into power, endurance, and technical disciplines.*

All of these athletes were subjected to urine tests for doping at several random times each year, in accordance with the initiative and decisions made by the WADA, IOC, International Sports Federation, and the organization Antidoping Sweden. None had tested positively for androgens or other unallowed substances.

Study design.

In connection with a visit by each subject to the Women's Health Research Unit in the morning after an overnight fast, body weight and height were recorded, and a detailed medical and gynecologic history was obtained. Their menstrual status was categorized as amenorrhea (absence of menstruation during at least the previous 3 months), oligomenorrhea (five to nine periods during the past year occurring at intervals greater than 6 wk), or regular menstruation (RM; consistent periods at an interval of 22-34 d). Subjects in the first two categories were defined as exhibiting menstrual dysfunction (MD). In addition, all of the subjects estimated the degree of their physical training (on a subjective scale ranging from 2 = very, very light, 4 = very light, 6 = light, 8 = medium, 10 = heavy, 12 = very heavy, and 14 = very, very heavy) during the week preceding the visit.

Furthermore, a peripheral venous blood sample was collected from each woman while resting. In the case of women who were not using hormonal contraceptives, these samples were taken early in the follicular phase of the menstrual cycle (days 1-5) if they were menstruating or, otherwise, on an arbitrary day. Blood was drawn from women using hormonal contraceptives (HC, defined as any contraceptive regimen involving sex steroids, i.e., oral preparations of estradiol and gestagen or a gestagen alone, subdermal slow-release implants, or intrauterine gestagen-releasing devices) either during the last week of the active hormone phase of the OC treatment cycle if the HC was used cyclically or on an arbitrary day, if use was continuous. The corresponding sera, obtained by centrifugation, were stored at −20°C until being assayed.

In connection with the visit to our unit, a gynecologic examination (including ultrasonography with the Siemens Sonoline SI-250 (Siemens AG, Erlangen, Germany) equipped with a transvaginal transducer operating between 5.0 and 7.5 MHz) was performed. The ovarian parameters assessed were the maximum number of follicles present in a single plane and total volume (calculated as described previously 10). Women with at least one ovary demonstrating 12 or more follicles in one plane and/or a volume of greater than 10 mL were defined as exhibiting PCO (29).

Body composition.

Whole body composition in terms of fat mass, lean body mass (LBM), and bone mineral density (BMD, g·cm−2) was determined using dual-energy x-ray absorptiometry (DEXA) using the Lunar model Prodigy (GE Healthcare, Madison, WI). The software of the Lunar automatically calculates the amount of fat in the trunk and legs (the limit between these regions being defined by the line drawn from the upper margin of the iliac crest to the neck of the femur). The ratio of trunk-leg fat mass was taken as an estimate of the ratio of upper-lower body fat mass.

From the whole body DEXA, spinal BMD was also determined. The spinal subregion consisted of the lower portion of the cervical spine, together with the thoracic and most of the lumbar spine (approximately L1-L4). The t- and z-scores were estimated from the mean BMD, and their SD values were supplied by the manufacturer of the Lunar equipment. The whole body BMD was defined as osteopenia if the t-score was 1-2.5 SD lower than that of young adults (peak bone mass), as osteoporosis if the t-score was even lower (in accordance with the recommendations of the World Health Organization), or as exhibiting a z-score more than 2 SD lower than the mean for age-matched control individuals (in accordance with the guidelines of the International Society for Clinical Densitometry concerning low BMD in premenopausal women 35). The reproducibility in the determination of whole body BMD in this manner has been calculated to be less than 0.01 g·cm−2 or 0.1 SD (3).

Analytical procedures.

Serum levels of follicle-stimulating hormone (FSH), LH, SHBG, prolactin (PRL), thyroid-stimulating hormone (TSH), free thyroxine (fT4), and dehydroepiandrosterone sulfate (DHEAS) were determined by enzyme immunoassays involving direct chemiluminescence, as previously described (20,21). Serum levels of testosterone (T), androstenedione (A4), 17α-hydroxyprogesterone (17OHP), and estradiol-17β (E2) were determined by radioimmunoassay (RIA) (20,21).

Serum levels of insulin were measured by RIA as described elsewhere (12). After separation from IGFBP by acidic ethanol extraction and cryoprecipitation, IGF-I was determined by RIA, using des [1-3] IGF-I as the radioligand to minimize interference from any remaining IGFBP (2). The concentrations of IGFBP-1 in serum samples were determined by RIA according to Póvoa et al. (25). Serum cortisol was quantified by enzyme immunoassay involving direct chemiluminescence (21). Finally, glucose was determined using a glucose analyzer (YSI2300 Stat Plus; Yellow Springs Instruments, Yellow Springs, OH) on the basis of an enzymatic electrochemical procedure.

The ratio between total serum T and SHBG (the T/SHBG ratio or "free androgen index") is considered to be a useful indicator of T activity in women (29) and was, therefore, also calculated here. LH values of <2 U·L−1 were considered to indicate inhibition of the HPG axis. According to the Rotterdam consensus (29), PCOS is diagnosed when at least two of the three following phenomena are present: 1) oligo- or anovulation, 2) clinical and/or biochemical signs of hyperandrogenism, and/or 3) detection of PCO by ultrasound.

Statistical analyses.

The values obtained for variables that exhibit a normal distribution are expressed as means ± SD, whereas nonparametric variables are presented as medians (quartile range, Q25-Q75). The values for different groups were compared using the unpaired t-test or ANOVA followed by Fisher LSD post hoc analyses. Variables exhibiting significant skewness were transformed to logarithms before analysis. When comparing the IGF-I levels of different groups, differences in age were adjusted for using ANCOVA. Categorical variables were compared using the Fisher exact test, and correlations were assessed using Spearman's rank-order correlation. Forward stepwise multiple linear regression analysis was used to evaluate the extent to which variations in the BMD could be explained by different factors. Regression coefficients are expressed as standardized b (bs) and the total variance as total adjusted R2. P values < 0.05 were considered to be statistically significant. All analyses were performed with the Statistica™ 8.0 software (StatSoft®, Inc, Tulsa, OK).


General characteristics.

The mean ages of our subjects who participated in power and endurance sports were comparable (24.0 ± 3.6 and 24.1 ± 3.6 yr, respectively), whereas the average age of our technical athletes was older (30.3 ± 4.3 yr, P < 0.001, compared with either of the other two groups). The average age of menarche in the entire population was 13.3 ± 1.4 yr and did not differ among the three groups. Of the athletes, 47% were using HC at the time of our study. With the exception of one athlete who was using oral contraceptives for regulation of menstruation, all of the women were using HC for birth control.

Menstrual status.

Among the athletes who were not using HC, 27% had menstrual disturbances. A significantly larger proportion of the endurance athletes (6/9, 66.7%) than of those participating in power disciplines (4/30, 13.3%; P < 0.01) exhibited such disturbances, and the same tendency was seen upon comparing endurance and technical athletes (3/9, 33.3%; P = 0.06). The frequencies of regular menstruation (RM), menstrual dysfunction (MD), and usage of HC in the three groups are presented in Figure 1.

The proportion of athletes having regular menstruation (black), menstrual dysfunction (white), and using hormonal contraception (hatched) in power (n = 55), endurance (n = 20), and technical (n = 15) disciplines. Values are presented as percent of all athletes in each sport group. A significantly larger proportion of the endurance athletes (66.7%) than of those participating in power disciplines (13.3%; P < 0.01) exhibited menstrual disturbances.

The highest frequency of PCO detected by ultrasound in women not using HC was demonstrated by the power athletes (12/28, 42.9% vs 3/9, 33.3% in endurance and 2/9, 22.2% in technical athletes), although these differences were not statistically significant. Furthermore, this condition was more prevalent in women who were not using HC than in those who were (37.0% vs 12.5%; P < 0.01). There were no differences in circulating levels of sex hormones among the three sport groups (data not shown).

A total of 13 athletes exhibited MD, of which 5 were classified as amenorrhea and 8 as oligomenorrhea. Each underwent a thorough individual assessment of the disorder. The most common endocrine abnormality associated with MD was PCOS, which was demonstrated by one amenorrheic and five oligomenorrheic athletes. Besides MD, these women had PCO on ultrasound and clinical or biochemical signs of hyperandrogenism. In addition, a single case of hypothalamic inhibition and one of hyperprolactinemia were diagnosed among the amenorrheic athletes. No specific endocrine abnormalities were observed in the remaining five athletes with MD, which were therefore considered as idiopathic.

The serum levels of the various hormones analyzed here in the athletes with RM, PCOS, and non-PCOS MD are documented in Table 2. The mean levels of LH, LH/FSH ratio, T, T/SHBG ratio, DHEAS, and 17OHP were significantly higher, and SHBG was significantly lower in athletes with PCOS compared with regularly menstruating subjects, as expected. In contrast, athletes with non-PCOS MD had significantly lower FSH and fT4 than women with RM.

Serum levels of hormones in athletes with regular menstruation (RM) and menstrual dysfunction, with and without polycystic ovary syndrome (PCOS and non-PCOS MD).

Body composition.

The body compositions and self-estimated training loads of our Olympic athletes are summarized in Table 3. Although there were no differences in the average weights of the three groups, endurance athletes had a significantly lower mean body mass index (BMI) than those in the power group and were, on average, taller than both the technical and power athletes. The proportion of body fat ranged between 7% and 38.1% (median value, 22.2%), with a significantly higher mean value in the case of technical than endurance or power athletes (Fig. 2A). The body fat contents of athletes with RM and MD were similar (23.6 ± 5.0% vs 21.1 ± 6.2; not significant), whereas the ratio of upper-lower body fat mass was highest in the power group (Table 3). Even after correction for the differences in height, the technical athletes demonstrated the lowest LBM of the three groups (Table 3; lean mass of the legs in Fig. 2B), whereas power and endurance athletes were similar in this respect.

The body compositions and self-estimated training loads of athletes participating in power, endurance, and technical sports.
A, Total body fat content, medians, Q25-Q75, and total range in athletes participating in power (n = 55), endurance (n = 20), and technical (n = 15) disciplines. B, Lean mass in legs, means ± SD, in athletes participating in power (n = 55), endurance (n = 20), and technical (n = 15) disciplines. C, Whole body z-score, means ± SD, and extreme values in athletes participating in power (n = 55), endurance (n = 20), and technical (n = 15) disciplines. Significance levels are indicated as follows: *P < 0.05, **P < 0.01, and ***P < 0.001.

In general, the BMD of our athletes was high (Table 3), and none demonstrated osteopenia or osteoporosis, as assessed on the basis of t-scores (Table 3) and z-scores (Fig. 2C). Power athletes exhibited significantly higher BMD values (both total body and for the spine) than the endurance and technical athletes, for whom these values were similar. All three groups reported, on the average, a training load of medium intensity.

Biomarkers of energy availability.

With the exception of serum levels of IGF-I, no differences in biomarkers of energy availability were observed among the three sport groups (Table 4). Although technical athletes demonstrated lower IGF-I levels than either power or endurance athletes, these differences were not statistically significant after adjustment for the differences in age. In all cases, the average values for these biomarkers were within the normal ranges (12,24).

Serum levels of biomarkers of energy availability in our power, endurance, and technical athletes.

Although treatment with oral contraceptives is well known to reduce circulating levels of IGF-I (1) and increase those of IGFBP-1 and cortisol, consideration of the use of HC still resulted in no significant differences in these parameters among the sport groups (not shown). Furthermore, the biomarkers of energy availability in the regularly menstruating athletes and those with PCOS and non-PCOS MD did not differ (data not shown). Five of the athletes who were not using HC exhibited IGFBP-1 levels that were higher than the upper clinical reference value of 45 μg·L−1. Of these, one endurance and one power athlete had oligomenorrhea. Otherwise, there were no other aberrations indicative of energy deficiency.

Multiple regression analysis.

A stepwise multiple regression analysis demonstrated that participation in power sports was associated with higher values of both total and spinal BMD (total BMD: bs = +0.54, P < 0.001; spinal BMD: bs = +0.32, P < 0.05; reference group: technical athletes). Furthermore, higher serum levels of cortisol were associated with reduced values for these same parameters (for total BMD: bs = −0.32, P < 0.05; for spinal BMD: bs = −0.42, P < 0.01). Moreover, higher serum levels of insulin were independently associated with higher spinal BMD values (bs = +0.29, P < 0.05). Menstrual dysfunction was not a significant factor predicting BMD. The variables discussed here accounted together for 36% and 30% of the variances in the total and spinal BMD values.


Female athletes competing on various levels for whom low body weight may be advantageous have been reported to be prone to develop chronic energy deficiency (5,16). This condition may be associated with suppression of the HPG axis, resulting in disturbances of the menstrual cycle and demineralization of the skeleton, i.e., the female athlete triad (22,36). In the present investigation on 90 female Olympic athletes participating in power, endurance, and technical sports, we found that almost all of the women had an anabolic body composition without any obvious indications of being in a catabolic state.

Overall, 27% of the Olympic athletes were found to have menstrual disturbances, which were particularly common in endurance athletes, as demonstrated previously (26). Almost half of the Olympic athletes were using HC for birth control, a rate which is comparable to the frequency of use by women of the same age in the general Swedish population (14). The most common endocrinological mechanism underlying menstrual disturbances was the hyperandrogenic condition PCOS, rather than the hypothalamic inhibition because of caloric deficiency. The group of athletes with PCOS demonstrated a hormonal pattern that is typical for this condition, such as an elevated LH/FSH ratio and higher androgen levels than those who menstruated regularly. Moreover, these subjects had higher circulating levels of 17OHP, which, however, were far below pathologic levels and could therefore not be due to a mild form of congenital adrenal hyperplasia. In contrast to what might be expected, only one of our athletes was diagnosed with hypothalamic inhibition. The occurrence of hyperprolactinemia in another Olympic athlete with MD further emphasizes the importance of considering various types of endocrine disorders in athletes.

The present important finding that PCOS is the predominant cause of MD in Olympic athletes is supported by our earlier observation of essential hyperandrogenism in a subgroup of endurance athletes competing at a less elite level than the athletes here (27,28). In previous reports (27,28), we demonstrated that athletic amenorrhea is associated with a hormonal profile consistent with a hypometabolic state, namely, attenuated diurnal secretion of LH, low serum levels of insulin, and enhanced diurnal secretion of GH and cortisol. In contrast, the oligomenorrheic athletes displayed elevated diurnal secretion of both LH and testosterone and a high incidence of PCO. Thus, the hormonal pattern of amenorrheic athletes was consistent with hypothalamic inhibition because of energy deficiency, whereas the oligomenorrhea in the athletes was a reflection of PCOS. The present investigation also reveals that oligomenorrhea can be a symptom of hyperandrogenism, that is, five of the eight Olympic athletes who exhibited oligomenorrhea were diagnosed with PCOS.

Polycystic ovaries are a well-recognized characteristic of a hyperandrogenic status, such as PCOS. The incidence of PCO observed with ultrasound in our athletes who were not using HC was high (37%) compared with the corresponding estimated incidence of approximately 20% in the general population (18). PCO was most common among the power athletes and least frequent among those participating in technical sports.

We demonstrated previously that athletes with PCOS performed better than normoandrogenic athletes in physical tests, that is, exhibiting higher maximal oxygen uptake (27). Androgens are well known to increase muscle strength in a dose-dependent manner and to enhance physical performance in both men and women (4,31). Therefore, we speculate that the presence of PCO/PCOS may reflect an anabolic state that could be advantageous for physical performance and may thereby play a key role in the achievement of a high competitive standard by female athletes.

The body mass index in our Olympic athletes ranged from 18.8 to 26.6 kg·m−2, which is very close to the normal range defined by the World Health Organization (i.e., 18.5-24.9 kg·m−2) (34). The ideal body fat content in young, sedentary, healthy women is considered to be 20-25% (9) and 39% of the Olympic athletes were within this range, 33% had values below it, and 28% were above this interval. In 1983, the average body fat content (as assessed by hydrostatic weighing) of 298 women participating in 15 Olympic events was reported to be 16.3% (7). Sparling et al. (30) observed that this value for the United States Olympic women's field hockey team in 1998 was 17.6% (as determined by DEXA). In addition, Meyer et al. (19) recorded an average body fat content of 17.5 ± 3.8% in 14 female athletes participating in the Winter Olympic Games in 2004. Although direct comparison is not possible, the average proportion of body fat observed here seems to be notably higher than in these investigations of American athletes.

As expected, endurance athletes exhibited the lowest BMI and less body fat than the women participating in the other two categories of sport. A relatively low fat mass has been associated with menstrual disturbances (7,9). However, in the present study, the Olympic athletes with such disturbances did not have a lower proportion of body fat than those menstruating regularly.

In general, bone mineral density was high among our subjects, especially in power athletes (whose mean z-score was greater than +2 SD). Frequent and intense mechanical loading are well known to exert a pronounced enhancing effect on BMD (19,33). Accordingly, the fact that power athletes should subject themselves to considerably higher impact forces and weight loads than endurance and technical athletes probably explains the significantly higher BMD of the former.

Several studies have documented negative effects of menstrual disturbances and energy deficiency on BMD in female athletes ([6], and references therein). Here, although more than a quarter of the study subjects had MD, none of the 90 female Olympic athletes exhibited osteopenia or osteoporosis. Furthermore, MD did not exert a negative effect on BMD (as calculated by multiple regression analysis). This latter observation might reflect the fact that these cases of MD were due primarily to PCOS, which should instead exert an anabolic impact on the BMD (23).

In summary, 90 female Olympic athletes participating in different events were shown here to display a generally anabolic body composition and exhibit biomarkers of energy availability that were within the normal ranges, that is, they had no obvious indications of chronic energy deficiency. Furthermore, most of the cases of MD observed were caused by PCOS rather than by hypothalamic inhibition. These findings challenge the contemporary concept of the female athlete triad, which postulates that reproductive dysfunction in athletic women is typically a consequence of chronic energy deficiency resulting from eating disorders. Indeed, Torstveit and Sundgot-Borgen (32) recently demonstrated that the prevalence of the triad among elite female Norwegian athletes was low (4.3%) and similar to that in a nonathletic control group (3.4%). Although the female athlete triad was not specifically investigated here, not a single athlete had osteopenia or osteoporosis; hence, no one fulfilled the criteria for such a state.

Our observations may suggest that because of the intense focus on their health issues, the training and nutritional habits of female athletes have improved in recent years. An alternative interpretation that cannot be ruled out is that Scandinavian athletes have different such habits compared with American elite sportswomen. Our results might also partially be explained by a selection of athletes with anabolic characteristics to top elite performance at the Olympic level. Certainly, the present findings are encouraging in their indication that female athletes can perform exceedingly well at an elite level without endangering their health.

This study was financed by grants from the Swedish Research Council (04224, 20324), the Swedish Olympic Committee, the Centre for Athletic Research, and the Karolinska Institutet in Stockholm. The authors also thank Berit Legerstam and Veronica Marklund for their logistical support. The results of the present study do not constitute endorsement by ACSM.

Conflicts of interest: None to our knowledge.


1. Balogh A, Kauf E, Vollanth R, Graser G, Klinger G, Oettel M. Effects of two oral contraceptives on plasma levels of insulin-like growth factor I (IGF-I) and growth hormone (hGH). Contraception. 2000;62(5):259-69.
2. Bang P, Eriksson U, Sara V, Wivall IL, Hall K. Comparison of acid ethanol extraction and acid gel filtration prior to IGF-I and IGF-II radioimmunoassays: improvement of determinations in acid ethanol extracts by the use of truncated IGF-I as radioligand. Acta Endocrinol (Copenh). 1991;124(6):620-9.
3. Brismar TB, Ringertz J. Effect of bone density of the head on total body DEXA measurements in 100 healthy Swedish women. Acta Radiol. 1996;37(1):101-6.
4. Cardinale M, Stone MH. Is testosterone influencing explosive performance? J Strength Cond Res. 2006;20(1):103-7.
5. De Souza MJ, Lee DK, Van Heest JL, Scheid JL, West SL, Williams NI. Severity of energy-related menstrual disturbances increases in proportion to indices of energy conservation in exercising women. Fertil Steril. 2007;88(4):971-5.
6. De Souza MJ, Williams NI. Beyond hypoestrogenism in amenorrheic athletes: energy deficiency as a contributing factor for bone loss. Curr Sports Med Rep. 2005;4(1):38-44.
7. Fleck SJ. Body composition of elite American athletes. Am J Sports Med. 1983;11(6):398-403.
8. Franks S. Controversy in clinical endocrinology: diagnosis of polycystic ovarian syndrome: in defense of the Rotterdam criteria. J Clin Endocrinol Metab. 2006;91(3):786-9.
9. Frisch RE. Body fat, menarche, fitness and fertility. Hum Reprod. 1987;2(6):521-33.
10. Goswamy RK, Campbell S, Whitehead MI. Screening for ovarian cancer. Clin Obstet Gynaecol. 1983;10(3):621-43.
11. Hagmar M, Hirschberg AL, Berglund L, Berglund B. Special attention to the weight-control strategies employed by Olympic athletes striving for leanness is required. Clin J Sport Med. 2008;18(1):5-9.
12. Hilding A, Brismar K, Degerblad M, Thoren M, Hall K. Altered relation between circulating levels of insulin-like growth factor-binding protein-1 and insulin in growth hormone-deficient patients and insulin-dependent diabetic patients compared to that in healthy subjects. J Clin Endocrinol Metab. 1995;80(9):2646-52.
13. Joy EA, Campbell D. Stress fractures in the female athlete. Curr Sports Med Rep. 2005;4(6):323-8.
14. Larsson M, Aneblom G, Odlind V, Tyden T. Reasons for pregnancy termination, contraceptive habits and contraceptive failure among Swedish women requesting an early pregnancy termination. Acta Obstet Gynecol Scand. 2002;81(1):64-71.
15. Laughlin GA, Yen SS. Nutritional and endocrine-metabolic aberrations in amenorrheic athletes. J Clin Endocrinol Metab. 1996;81(12):4301-9.
16. Loucks AB. Energy balance and body composition in sports and exercise. J Sports Sci. 2004;22(1):1-14.
17. 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(1):297-311.
18. Lowe P, Kovacs G, Howlett D. Incidence of polycystic ovaries and polycystic ovary syndrome amongst women in Melbourne, Australia. Aust N Z J Obstet Gynaecol. 2005;45(1):17-9.
19. Meyer NL, Shaw JM, Manore MM, et al. Bone mineral density of Olympic-level female winter sport athletes. Med Sci Sports Exerc. 2004;36(9):1594-601.
20. Naessen S, Carlstrom K, Garoff L, Glant R, Hirschberg AL. Polycystic ovary syndrome in bulimic women-an evaluation based on the new diagnostic criteria. Gynecol Endocrinol. 2006;22(7):388-94.
21. Naessen S, Carlstrom K, Glant R, Jacobsson H, Hirschberg AL. Bone mineral density in bulimic women-influence of endocrine factors and previous anorexia. Eur J Endocrinol. 2006;155(2):245-51.
22. Nattiv A, Loucks AB, Manore MM, Sanborn CF, Sundgot-Borgen J, Warren MP. American College of Sports Medicine. Position Stand. The female athlete triad. Med Sci Sports Exerc. 2007;39(10):1867-82.
23. Notelovitz M. Androgen effects on bone and muscle. Fertil Steril. 2002;77(Suppl 4):S34-41.
24. Nystrom FH, Ohman PK, Ekman BA, Osterlund MK, Karlberg BE, Arnqvist HJ. Population-based reference values for IGF-I and IGF-binding protein-1: relations with metabolic and anthropometric variables. Eur J Endocrinol. 1997;136(2):165-72.
25. Pòvoa G, Roovete A, Hall K. Cross-reaction of serum somatomedin-binding protein in a radioimmunoassay developed for somatomedin-binding protein isolated from human amniotic fluid. Acta Endocrinol (Copenh). 1984;107(4):563-70.
26. Redman LM, Loucks AB. Menstrual disorders in athletes. Sports Med. 2005;35(9):747-55.
27. Rickenlund A, Carlstrom K, Ekblom B, Brismar TB, von Schoultz B, Hirschberg AL. Hyperandrogenicity is an alternative mechanism underlying oligomenorrhea or amenorrhea in female athletes and may improve physical performance. Fertil Steril. 2003;79(4):947-55.
28. Rickenlund A, Thoren M, Carlstrom K, von Schoultz B, Hirschberg AL. Diurnal profiles of testosterone and pituitary hormones suggest different mechanisms for menstrual disturbances in endurance athletes. J Clin Endocrinol Metab. 2004;89(2):702-7.
29. Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril. 2004;81(1):19-25.
30. Sparling PB, Snow TK, Rosskopf LB, O'Donnell EM, Freedson PS, Byrnes WC. Bone mineral density and body composition of the United States Olympic women's field hockey team. Br J Sports Med. 1998;32(4):315-8.
31. Storer TW, Magliano L, Woodhouse L, et al. Testosterone dose-dependently increases maximal voluntary strength and leg power, but does not affect fatigability or specific tension. J Clin Endocrinol Metab. 2003;88(4):1478-85.
32. Torstveit MK, Sundgot-Borgen J. The female athlete triad exists in both elite athletes and controls. Med Sci Sports Exerc. 2005;37(9):1449-59.
33. Turner CH. Three rules for bone adaptation to mechanical stimuli. Bone. 1998;23(5):399-407.
34. World Health Organisation. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser. 2000;894:i-xii, 1-253.
35. Writing Group for the ISCD Position Development Conference. Technical standardization for dual-energy x-ray absorptiometry. J Clin Densitom. 2004;7(1):27-36.
36. 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.


©2009The American College of Sports Medicine