Body weight is reported to be an important determinant of cycling and running exercise capacity (22), and restricted eating behavior and menstrual dysfunction such as secondary functional hypothalamic amenorrhea (SFHA) are common among female athletes in weight-sensitive sports (2,19,25,28,29). Menstrual dysfunction is often ignored within the athletic community and regarded as a natural and sometimes more-or-less obligatory consequence of intense training despite lower bone mineral density (2,28), with increased risk of premature osteoporosis (28,33) and increased incidence of injuries (30) being linked to this condition. The etiology of SFHA involves persistent low energy and glucose availability (20), leading to several endocrine alterations such as reduced luteinizing hormone (LH) pulsatility, hypoestrogenism, suppressed triiodothyronine (T3), and increased levels of cortisol (11,19,20,28,36,39).
Estrogen is known to influence muscle contractile properties, muscle repair, regenerative processes, and postexercise muscle damage. Low levels of estrogen have also been associated with a higher incidence of injury and delayed recovery (7).
Elevated cortisol and lower T3, concurrent with hypoglycemia, are frequently reported in women with SFHA (11,19,20,28). High cortisol levels have been demonstrated to increase proteolysis (6,18). Lower levels of T3, although still within the normal range, have been shown to affect neuromuscular function assessed using electromyography (31) and to decrease muscle strength in nonathletic women (7,18,31). Furthermore, low T3 has been shown to decrease intracellular Ca2+ concentration, as well as the release and the uptake of Ca2+ by the sarcoplasmic reticulum, which may further limit neuromuscular performance (18) and contribute to fatigue and delayed recovery (3,8). The inability to maintain sufficient glucose supply to the contracting muscle has also been shown to decrease the intracellular Ca2+ concentration via a depressed release rate of Ca2+ from the sarcoplasmic reticulum, potentially contributing to fatigue and delayed recovery (3,8). Hence, alterations in glucose homeostasis, endogenous sex steroids, and stress hormone levels, as well as circulating thyroid hormones, may contribute to lowered neuromuscular performance in female athletes.
The aim of this study was therefore to investigate the link between reproductive function, metabolic and endocrine alterations, and neuromuscular performance in elite female endurance athletes.
MATERIALS AND METHODS
The study population was recruited through the Danish and Swedish sport federations in weight-bearing endurance sports and through local competitive endurance sports clubs in the Oresund region. Before inclusion, all participants were informed orally and in writing of all study procedures and gave their informed consent. Permission to undertake the study was provided by the data inspectorate and the regional ethical committees in both Sweden and Denmark (Nos. 2011/576 and H-4-2011-096) and endorsed by the Swedish and Danish Confederations of Sports and Team Denmark. The protocol was registered at www.clinicaltrials.gov.
Participants included were competitive athletes 18–39 yr of age and training a minimum of five times per week. Exclusion criteria included pregnancy, chronic illness, smoking, use of hormonal contraceptives including local intrauterine hormone–releasing systems 6 wk before investigation, injuries preventing the athlete from exercise training ≥2 wk, or menstrual dysfunctions other than SFHA. Forty-five athletes completed the study protocol; 15 were diagnosed with a menstrual dysfunction other than SFHA (primary FHA (n = 4), oligomenorrhea (n = 6), polycystic ovary syndrome (PCOS; n = 4), other menstrual dysfunction (n = 1)) and were therefore excluded from the present analysis. Sixteen participants were classified as being eumenorrheic (EUM) and 14 as having SFHA and included in the analysis.
Menstruating athletes were examined and performed all tests in the early follicular phase, on the third to the fifth day of menstruation, to eliminate hormonal fluctuations during the luteal phase and thereby limiting the variability between subjects. Participants performed the physical tests on two consecutive days followed by a 7-d registration period in the athletes’ normal environment. The first day comprised examinations involving bone health and reproductive function. The second day included examinations of energy metabolism, aerobic capacity, and neuromuscular performance. They were instructed not to exercise for more than 30 min at low or moderate intensity the day before and on the first day of testing and to arrive at the clinic in a fasted (from midnight the previous night) and rested state.
Twelve of the 30 participants included used hormonal contraceptives (5 of the 14 SFHA and 7 of the 16 EUM athletes) and were requested to stop for a minimum of 6 wk before the gynecological examination to secure a sufficient washout period for exogenous estrogen and progesterone (17). Six weeks was confirmed to be a sufficient washout period, because sex hormone–binding globulin (SHBG) was within normal range (20–150 nmol·L−1 for women 20–40 yr old) for all 12 hormonal contraceptive using participants (mean, 64.9 nmol·L−1; range, 40–116 nmol·L−1). A pregnancy test was performed on the arrival at the clinic, followed by an examination of menstrual function assessed by an experienced gynecologist using a transvaginal ultrasound examination (Ultrasound Scanner, Class 1 type B, B-K Medical REF TYPE 2202). The maximum number of ovarian follicles present in a single plane was counted, and total volume was assessed. Ovarian volumes were calculated using the prolate ellipsoid approximation formula [abc × (π/6)] (14). In addition to sex hormone status, an anamnestic assessment (i.e., age of menarche, previous menstrual irregularities, use of hormonal contraceptives, and number of menstrual cycles during the last year) was performed using the Low Energy Availability (EA) in Females Questionnaire (LEAF-Q) (25). After the gynecological examination, participants were asked to keep a menstrual bleeding calendar and were contacted monthly by the research team during a follow-up period of a minimum of 3 months. Participants were then classified as follows: 1) EUM (menstrual cycles of 28 ± 7 d and normal sex hormones), 2) oligomenorrhea (menstrual cycles >35 d), FHA (either primary (no menarche after 15 yr of age) or secondary (absence of ≥3 consecutive menstrual cycles)), 3) other menstrual dysfunction (anatomic defects, hyperprolactinemia, or other dysfunctional ovarian conditions), or 4) PCOS, defined as two or more of the following: 1) enlarged ovaries with a volume >10 mL and/or one or more ovaries demonstrating ≥12 follicles in one plane, 2) oligomenorrhea/amenorrhea, and 3) clinical symptoms of excess androgenic hormones (excess hair growth, acne, and obesity with android fat deposition).
Dual-energy x-ray absorptiometry (DXA; Hologic Model Discovery A (QDR series), software: body composition and subregional analysis; Hologic Inc., Bedford, MA) was used to determine whole-body fat-free (FFM) and fat (FM) mass as well as FFM in the tested right leg (FFMleg) All measurements and scans were performed between 7:30 AM and 9:00 AM in the fasted state and after voiding, assessed by the same technician, and performed on the same scanner. Calibration of the scanner was performed weekly using a phantom provided by the manufacturer.
Participants were transported by a car to the clinic on the morning of the second test day to minimize physical activity before assessment of resting metabolic rate (RMR). RMR was assessed between 7:00 and 8:30 AM after an overnight fast using a ventilated open hood system (Oxycon Pro 4, Jaeger, Hoechberg, Germany). The system was automatically calibrated before each test, and again weekly, using an alcohol-burning test. After voiding, participants laid supine for 15 min; after which measurements of oxygen uptake (V˙O2) and carbon dioxide output (V˙CO2) were recorded for 35 min. After RMR assessment, a venous blood sample was taken to ensure hormone concentrations to be measured in the fasted resting state. A test of work efficiency was then performed, with the participants remaining in the fasted state. The work efficiency test started with the participant sitting on the bicycle ergometer (Monark 939E; Monark Exercise AB, Vansbro, Sweden) for 6 min, followed by cycling for 6 min at 0, 50, and 100 W, respectively. An airtight mask covering the mouth and nose was used to measure respiratory gas exchange breath-by-breath (Oxycon Pro 4, Jaeger). Work efficiency was calculated by dividing the increase in external work on the cycle ergometer (from 0 and 50 W to 100 W) by the corresponding increase in total energy expenditure for the same period and expressing this as a percentage. Capillary blood glucose from the fingertip was measured at rest and after cycling for 5 min at 0, 50, and 100 W. An average of the fasting capillary blood glucose at rest and during light physical activity (taken directly after measuring RMR and after cycling for 5 min at 0, 50, and 100 W during the work efficiency test) was used in the statistical analysis (Fig. 1).
Food intake and exercise training were recorded by the participants for seven consecutive days to calculate current EA (19). Energy intake was calculated from prospective weighed food records. Participants were instructed to maintain and follow their normal dietary and exercise training regimen. Heart rate monitors (PolarRS400; Polar, Oulu, Finland) and exercise training logs were used to estimate exercise energy expenditure (EEE) during each exercise training session using individual prediction equations on the basis of measured heart rate and corresponding energy expenditure obtained during the incremental maximal exercise test in the laboratory (37). To calculate daily total energy expenditure, heart rate monitors (Polar RS400®) were used to assess energy expenditure during bicycle transportation, whereas actigraphy (ActiGraph GT3X; Pensacola, FL, USA) and the data analysis software ActiLife 5 (ActiGraph) were used for assessment of nonexercise activity thermogenesis (NEAT). Subjects were instructed to wear an accelerometer on the wrist during sleep and on the hip from rising in the morning until bedtime, and only to remove it during showering, swimming, bicycle transportation, and training. EA was calculated using the energy intake and EEE determined within the same 7-d period and FFM determined by DXA. EEE represented only the energy attributable to the training on the basis of training log and heart rate measurement for each training session. Finally, the estimate of the energy expended for RMR and NEAT throughout the duration of training was subtracted from the estimate of EEE (25).
Two hours after a standardized breakfast, an incremental exercise test on the bicycle ergometer was performed, initiated by cycling for 6 min at 50 W, followed by an increase in workload of 12–14 W·min−1 until exhaustion. An airtight mask covering the mouth and nose was worn to measure breath-by-breath V˙O2 and V˙CO2, for determination of V˙O2peak (Oxycon Pro 4; Jaeger). Heart rate was recorded continuously throughout the test.
After 10 min of rest, the reaction time (RT) test was performed on a portable computer connected to a linear encoder (MuscleLab 4010; Ergotest Innovation, Langensund, Norway). RT was assessed by measurement of time taken to press the space bar on the keyboard when the screen changed color while the finger rested on the space bar. After three trials for familiarization, the test procedure began. Each participant performed five reaction trials. The shortest RT of the five trials was used in the statistical analysis.
Knee muscular strength and Knee muscular endurance
Directly after the RT test, knee muscular strength (KMS) and endurance (KME) measurements were performed using isokinetic dynamometry, and each participant performed strength and endurance tests using the right leg in a consistent order. The isokinetic tests were performed on a computerized dynamometer (Biodex Multi-Joint System IV; Biodex Medical Systems, Inc., Shirley, NY), and the dynamometer calibration was verified on the morning of the test day. Participants were seated in a comfortable position with the backrest angled at 100° to the seat. Velcro straps were placed across the thigh, pelvis, and chest to minimize body movements and to isolate the movement of the knee joint. The arms were kept crossed over the chest to limit movement during the test. The mechanical axis of the dynamometer was aligned with the transverse axis of the knee joint, with the lateral femoral epicondyle used as the bony landmark. The weight of the right leg was recorded and a gravity adjustment made using the computer software. The range of movement was from 10° to 100° of knee flexion. Both isokinetic and concentric knee extension and flexion muscle strength, as well as endurance, were assessed using two test protocols. To analyze KMS, five maximal concentric extensor/flexor cycles at 60°·s−1, with no rest between repetitions, were performed. After a 60-s rest, participants performed 30 consecutive maximal extensor/flexor cycles at 180°·s−1 to analyze KME. Before initiating the KMS and KME test protocol, the participants performed two submaximal extensor/flexor cycles at the preselected angular velocity for familiarization. During the test, each participant was instructed to exert maximal force by performing each contraction as hard and as fast as possible. Peak torque was used in the calculation of KMS and defined as the sum of the peak torques in extension and flexion. Total work was used in the calculation of KME and was defined as the total work done in extension and flexion over the total 30 s.
Estradiol, androstenedione, and total testosterone were measured using tandem mass spectrometry, progesterone using ADVIA Centaur Immunoassay Systems (Siemens Healthcare Diagnostics Products GmbH, Marburg, Germany), serum SHBG using a chemiluminescent assay (Architect I 2000 system; Abbott, Europe), serum T3 using the ARCHITECT system essay (Abbott Laboratories, Longford, Ireland), cortisol and thyroid-stimulating hormone (TSH) levels using Roche Electro Chemiluminescence Immunoassay (Roche Diagnostic, Bromma, Sweden), insulin using an Access Ultrasensitive Insulin assay (Beckman Coulter, Bromma, Sweden), and capillary blood glucose using Biosen C Line (EKF Diagnostic, Barleben, Germany). For a more detailed description, the readers are referred to Melin et al. (25).
The data set was checked for missing data, and nonnormality before statistical tests were performed. Participants were divided by their reproductive function: EUM or SFHA. When comparing differences between these two groups, a Student’s unpaired t-test was used for comparison of means in normally distributed data, and the Mann–Whitney U test was used for skewed data. ANOVA was used when adjusting for covariates. To test for associations between continuous variables, Pearson coefficient of correlation (r) and Spearman rank correlation coefficient (rho) were used for normally distributed and skewed data, respectively. Multiple linear regression analyses, both backward and forward, were performed. In the backward regression, independent variables were excluded due to colinearity during multiple regression analyses. When no colinarites were found between independent variables, a forward regression analysis was performed. Dependent variables were RT, KMS, and KME. Independent variables were FFMleg, blood glucose, cortisol, T3, and estrogen. The models are reported together with the adjusted R 2 values as assessments of the best fit. Statistical significance was defined as P < 0.05. Potential differences in KMS and KME expressed as ratios of FFM in the tested leg were also tested using a one-way ANOVA. All statistical analyses were performed using SPSS for Windows (version 22.0; IBM, Chicago, IL).
Participant (n = 30) characteristics are shown in Table 1. Antral follicle count was normal in all EUM and SFHA athletes, and there were no ultrasound indications of PCOS from the ovarian morphology. The SFHA athletes weighed less and had lower body mass index (BMI) compared with EUM athletes. The body composition also differed, with lower absolute and relative whole-body FM and FFMleg in SFHA compared with EUM athletes.
There were no differences in energy intake, total energy expenditure, EA, or the amount of endurance training per week between the groups. However, SFHA athletes performed more resistance exercise training compared with EUM athletes (Table 1). Furthermore, SFHA athletes were characterized by lower RMR, the ratio between measured and predicted RMR using the Cunningham equation (RMRratio), T3, and blood glucose, but higher cortisol and work efficiency compared with EUM athletes (Table 2). We found positive associations between relative FM and T3 (r = 0.40, P < 0.05), RMR (r = 0.45, P < 0.05), and RMRratio (r = 0.42, P < 0.05), as well as between T3 and the RMR (r = 0.43, P < 0.05) and the RMRratio (r = 0.37, P < 0.05). A subanalysis including only SFHA subjects showed a strong negative association between T3 and total FFM (r = − 0.72, P < 0.01). Twelve of the 14 athletes (86%) with SFHA had hypoglycemia (blood glucose levels <4.0 mmol·L−1) in the fasted state both at rest and after cycling for 5 min at 0, 50, and 100 W respectively, whereas only 2 of the 16 EUM athletes (13%) were hypoglycemic during the same period (Fig. 1). Estrogen levels were lower and androgen levels were higher in SFHA compared with EUM athletes, although all levels were within normal range. The calculated cortisol-to-insulin ratio was 39% higher, and the estrogen-to-cortisol ratio was 50% lower in SFHA athletes compared with EUM athletes (Table 2).
EUM athletes had 7% shorter RT (Table 3), and a faster RT was found to be associated with higher blood glucose, T3, and estrogen levels as well as with lower cortisol (Table 4). EUM athletes were also found to have greater KMS (11%) and KME (20%) compared with SFHA athletes (Table 3). Although differences in FFMleg explained individual differences in KMS, both reproductive function and FFMleg independently explained the variability in KME (Table 3). However, when KME was divided by FFM in the tested leg, the difference between the groups did not reach the predefined level of significance (Table 3). KMS and KME were positively associated with FFMleg and T3 but negatively with cortisol (Table 4). T3 explained 14% of the variance in RT and 16% of the variance in KMS, with an additional 15% explained when FFMleg was included in the model. Cortisol accounted for 35% of the variance in KME, whereas FFMleg accounted for another 30% of the variance (Table 5). There were, however, no relationships between testosterone or androgen levels and performance characteristics. No differences in absolute exercise capacity or relative to body weight between the SFHA and EUM athletes were found.
This study is the first to demonstrate a lower neuromuscular performance in SFHA athletes compared with EUM athletes. SFHA athletes had lower BMI, FM, FFMleg, RMR, blood glucose, T3, and estrogen, as well as higher work efficiency and cortisol levels when compared with EUM athletes. Although SFHA had a lower body weight, we found no improved exercise capacity compared with EUM subjects. Reduced neuromuscular performance was associated with lower blood glucose, T3, estrogen, and FFMleg and with high cortisol levels—metabolic alterations that are frequently reported in female athletes with SFHA (11,20,28).
Although performing more resistance training and having higher levels of androgen and testosterone, which are normally associated with a greater muscle mass (5,15,35), participants with SFHA weighed less, had lower FM and FFMleg, and had a tendency toward a lower total FFM compared with EUM athletes (Table 1). The differences in body composition in this study could be a result of increased gluconeogenesis among SFHA athletes due to elevated cortisol levels in relation to insulin, because insulin is important not only in the regulation of glucose metabolism but also in the prevention of muscular proteolysis (6). Glucose infusions have been shown to suppress cortisol (23), linking low blood glucose levels to higher proteolysis, because a higher cortisol-to-insulin ratio has been shown to accelerate proteolysis (27). Estrogen has also been shown to have a protective effect on muscle protein (7,21,34). In the present study, the estrogen-to-cortisol ratio was 50% lower in the SFHA athletes compared with EUM athletes, whereas no difference was found in the testosterone-to-cortisol ratio (Table 1). Maintaining estrogen levels (21,34) will most likely be of greater importance than testosterone for women compared with men when considering skeletal muscle protein synthesis and breakdown (5). Adipose tissue is an estrogen-preserving and -producing tissue. In both relative and absolute terms, SFHA athletes in this study had lower FM compared with EUM athletes. Hence, the low levels of blood glucose, T3, and estrogen as well as the higher cortisol-to-insulin ratio may be important contributing factors explaining the altered body composition among the SFHA athletes compared with the EUM athletes.
Low glucose availability has been shown to suppress the hypothalamic–pituitary axis leading to reduced circulating hormones such as T3 (20,39), and several studies have reported a lower RMR in SFHA athletes, in line with the results in the present study (16,27,33,38). We also found a better work efficiency in the SFHA athletes, suggesting that the metabolic adaptation may not be limited to the resting condition. Rosenbaum et al. (32) reported a 35% increase in muscle efficiency in weight-reduced subjects resulting in lower energy expenditure during low-intensity activities. A lowered circulating thyroid hormone concentrations during energy deficiency has been suggested as a potential mechanism because it leads to changes in the recruitment of specific types of muscle fiber, favoring greater use of slower-twitch (primarily oxidative) fibers. Therefore, the increased work efficiency found in SFHA athletes in the present study could be due to the lower T3.
We did not find a difference in current EA between SFHA and EUM athletes, and our results indicate that the threshold for insufficient EA maintaining SFHA may be higher (9) than the 30 kcal·(kg FFM)−1·d−1 shown to induce a reduction in LH pulsatility in EUM women (20). Alternative explanations may be that our ability to assess EA in free-living subjects is not sensitive enough, or that the duration for assessing EA (1 wk) is too short to sufficiently reflect the habitual EA (26).
In the present study, the findings of lower blood glucose, T3, FFM, and higher cortisol as well as cortisol-to-insulin levels in the SFHA group compared with the EUM group may all, directly or indirectly, contribute to the observed reduced neuromuscular performance. Eighty-six percent of the SFHA athletes were hypoglycemic, not only in the fasted and rested state but also after low-intensity exercise (Fig. 1). This indicates a general inability to maintain glucose homeostasis that is also present during exercise. Blood glucose levels have been shown to influence athletic performance (1,4,10), and KME is dependent on energy production via glycolysis (1) as well as on muscle mass in the tested leg for force production. Energy supply is partly maintained by increased proteolysis, mainly from skeletal muscles, to produce free amino acids for the increased gluconeogenesis in the liver (6). In the present study, FFMleg was positively associated with muscle performance during both KMS and KME tests, suggesting that FFMleg might be involved in mediating the effects on KMS and KME. However, when statistically adjusted for FFMleg, the difference in endurance, but not strength, was maintained between SFHA and EUM athletes (Table 3), supporting an effect of menstrual function on KME independent of FFMleg.
The lower T3 (38) and higher cortisol levels found among SFHA athletes, just as with the lower FFM, may contribute to the observed differences in KMS and KME. Because of the duration of the KME test, substrate requirements by the working muscle will predominantly come not only from intramyocellular energy sources such as adenosine triphosphate (ATP) and phosphorylated creatine (PCr) (3,13) but also from glycogen. T3 has an effect on the mitochondria’s ability to produce ATP from glycogen (40), affecting the muscle’s ability to regenerate PCr (12). The finding that high levels of cortisol are related to low KME (Table 4) could be a consequence of low availability of glucose in the working muscle, explaining the contribution from cortisol to the results of KME. However, because of the short duration of the KMS test, the substrate requirements of the working muscle will be unaffected by glucose availability. Furthermore, although body weight is reported to be an important factor for exercise capacity (22), we found similar exercise capacity (mL O2·kg−1·min−1) between groups despite a lower body weight in SFHA subjects.
RT depends partly on the speed of the neural signal reaching the contracting muscle of the index finger (24). T3 is known to affect neural function through the mitochondria’s ability in the neuron to produce ATP for the Na+/K+-ATPase pump (40). This may contribute to the longer RT among the SFHA athletes. The excitability of the neurons innervating the contracting muscles is also important for producing high KMS, also highlighting the potential importance of T3 levels for the differences in KMS found in this study (31).
Anovulation and luteal-phase deficiency are commonly reported in female athletes (3,28). It is therefore a limitation to this study that EUM subjects were not assessed in this regard. However, excluding EUM subjects with subclinical menstrual dysfunction would potentially have increased the differences in metabolic and endocrine markers as well as in neuromuscular performance test between SFHA and EUM subjects even further.
In summary, we report neuromuscular performance to be lower in SFHA as compared with EUM endurance athletes. Athletes with SFHA were also characterized by lower blood glucose, T3, and estrogen levels but higher cortisol, as well as showing signs of energy-sparing adaptations. We additionally found a number of associations between metabolic markers as well as muscle mass (FFMleg) and performance.
Although this observational study is unable to provide evidence as to the specific effects these markers may have on mediating performance, the interplay between these markers and their potential link to performance are biologically plausible. We therefore hypothesize that habitually low blood glucose level may lead to increased levels of cortisol and reduced levels of T3, as well as a lower muscle mass in the long term, all of which have been implicated in impaired neuromuscular performance.
Consequently, low blood glucose levels may, both directly and indirectly, be an important factor leading to the lower neuromuscular performance found in endurance athletes with SFHA. This needs to be examined in future mechanistic studies.
We thank Dr. Mubeena Aziz and Dr. Katrine Haahr at Herlev Hospital for performing gynecological examinations, Ulla Kjærulff-Hansen at Herlev Hospital for assisting with blood sampling, Hanne Udengaard at Herlev Hospital for logistics assistance, and Sara Siig Møller and Janni Borg Hoffmeyer for analyzing dietary records. Finally, we highly appreciate the extraordinary cooperation of the athletes participating in this study and the support of the Swedish and Danish national sports federations and Team Danmark.
This study was funded by research grants from the Faculty of Science, University of Copenhagen; World Village of Women Sports Foundation; and Arla Foods Ingredients.
The results of the present study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation, and do not constitute endorsement by the American College of Sports Medicine. None of the authors have any conflicts of interest.
Å. B. T. and A. M. contributed to study planning, data collection and analysis, and manuscript writing. F. M. K. and A. J. contributed to data collection and revision of the manuscript. S. S. and J. F. contributed to study planning, data collection and analysis, and revision of the manuscript. A. S. contributed to study planning, data analysis, and manuscript writing. All authors approved the final manuscript.
1. Bangsbo J, Graham TE, Kiens B, Saltin B. Elevated muscle glycogen and anaerobic energy production during exhaustive exercise in man. J Physiol
2. Cialdella-Kam L, Guebels CP, Maddalozzo GF, Manore MM. Dietary intervention restored menses in female athletes with exercise-associated menstrual dysfunction with limited impact on bone and muscle health. Nutrients
3. Connett RJ, Sahlin K. Handbook of Physiology
. New York (NY): Oxford University Press; 1996. pp. 870–911.
4. Costill D, Gollnick PD, Jansson ED, Saltin B, Stein EM. Glycogen depletion pattern in human muscle fibres during distance running. Acta Physiol Scand
5. Crewther BT, Cook C, Cardinale M, Weatherby RP, Lowe T. Two emerging concepts for elite athletes: the short-term effects of testosterone and cortisol
on the neuromuscular system and the dose-response training role of these endogenous hormones. Sports Med
6. Darmaun D, Matthews DE, Bier DM. Physiological hypercortisolemia increases proteolysis, glutamine, and alanine production. Am J Physiol
7. Enns DL, Tiidus PM. The influence of estrogen on skeletal muscle: sex matters. Sports Med
8. Gejl KD, Hvid LG, Frandsen U, Jensen K, Sahlin K, Ortenblad N. Muscle glycogen content modifies SR Ca2+
release rate in elite endurance athletes. Med Sci Sports Exerc
9. Gibbs JC, Williams NI, Mallinson RJ, Reed JL, Rickard AD, De Souza MJ. Effect of high dietary restraint on energy availability
and menstrual status. Med Sci Sports Exerc
10. Gollnick PD, Piehl K, Saltin B. Selective glycogen depletion pattern in human muscle fibres after exercise of varying intensity and at varying pedalling rates. J Physiol
11. Gordon CM. Functional hypothalamic amenorrhea
. N Engl J Med
12. Haber VJ, Petersen SR, Chilibeck PD. Thyroid hormone concentrations and muscle metabolism in amenorrheic and eumenorrheic athletes. Can J Appl Physiol
13. Katz A, Sahlin K. Effect of decreased oxygen availability on NADH and lactate contents in human skeletal muscle during exercise. Acta Physiol Scand
14. Kelsey TW, Dodwell SK, Wilkinson AG, et al. Ovarian volume throughout life: a validated normative model. PLoS One
15. Knuttgen HG. Strength training and aerobic exercise: comparison and contrast. J Strength Cond Res
16. Lebenstedt M, Platte P, Pirke KM. Reduced resting metabolic rate in athletes with menstrual disorders. Med Sci Sports Exerc
17. Lindstedt RJ, Becker C, Axel O, Kågedal B. Clinical Chemistry in Praxis
. Student literature, Lund, Sweden; 2012. pp. 329–336.
18. Loucks AB. Thyriod hormone, or triiodothyronine. In: ACSM
’s Advanced Exercise Physiology
. Baltimore: Lippincott Williams & Wilkins; 2006. pp. 467–8.
19. Loucks AB. Energy availability
and infertility. Curr Opin Endocrinol Diabetes Obes
20. Loucks AB, Thuma JR. Luteinizing hormone pulsatility is disrupted at a threshold of energy availability
in regularly menstruating women. J Clin Endocrinol Metab
21. Lowe DA, Baltgalvis KA, Greising SM. Mechanisms behind estrogen’s beneficial effect on muscle strength in females. Exerc Sport Sci Rev
22. Lundby C, Montero D, Gehrig S, et al. Physiological, biochemical, anthropometric, and biomechanical influences on exercise economy in humans. Scand J Med Sci Sports
. 2017. doi: 10.1111/sms.12849.
23. MacLaren DP, Reilly T, Campbell IT, Hopkin C. Hormonal and metabolic responses to maintained hyperglycemia during prolonged exercise. J Appl Physiol (1985)
24. McIsaac G, Kiernan JA. Complete staining of neuromuscular innervation with bromoindigo and silver. J Anat
25. Melin A, Tornberg ÅB, Skouby S, et al. The LEAF questionnaire: a screening tool for the identification of female athletes at risk for the female athlete triad. Br J Sports Med
26. Melin A, Tornberg ÅB, Skouby S, et al. Energy availability
and the female athlete triad in elite endurance athletes. Scand J Med Sci Sports
27. Myerson M, Gutin B, Warren MP, et al. Resting metabolic rate and energy balance in amenorrheic and eumenorrheic runners. Med Sci Sports Exerc
28. Nattiv A, Loucks AB, Manore MM, et al. American College of Sports Medicine position stand. The female athlete triad. Med Sci Sports Exerc
29. Pollock N, Grogan C, Perry M, et al. Bone-mineral density and other features of the female athlete triad in elite endurance runners: a longitudinal and cross-sectional observational study. Int J Sport Nutr Exerc Metab
30. Rauh MJ, Macera CA, Trone DW, Shaffer RA, Brondine SK. Epidemiology of stress fracture and lower-extremity overuse injury in female recruits. Med Sci Sports Exerc
31. Reuters VS, Almeida Cde P, Teixeira Pde F, et al. Effects of subclinical hypothyroidism treatment on psychiatric symptoms, muscular complaints, and quality of life. Arq Bras Endocrinol Metabol
32. Rosenbaum M, Vandenborne K, Goldsmith R, et al. Effects of experimental weight perturbation on skeletal muscle work efficiency in human subjects. Am J Physiol Regul Integr Comp Physiol
33. Scheid JL, Williams NI, West SL, VanHeest JL, De Souza MJ. Elevated PYY is associated with energy deficiency and indices of subclinical disordered eating in exercising women with hypothalamic amenorrhea
34. Spangenburg EE, Geiger PC, Leinwand LA, Lowe DA. Regulation of physiological and metabolic function of muscle by female sex steroids. Med Sci Sports Exerc
35. Staron RS, Leonardi MJ, Karapondo DL, et al. Strength and skeletal muscle adaptations in heavy-resistance trained women after detraining and retraining. Pflugers Arch
36. Thong FS, McLean C, Graham TE. Plasma leptin in female athletes: relationship with body fat, reproductive, nutritional, and endocrine factors. J Appl Physiol (1985)
37. Tomten SE, Høstmark AT. Energy balance in weight stable athletes with and without menstrual disorders. Scand J Med Sci Sports
38. Vanheest JL, Rodgers CD, Mahoney CE, De Souza MJ. Ovarian suppression impairs sport performance in junior elite female swimmers. Med Sci Sports Exerc
39. Warren MP. Endocrine manifestations of eating disorders. J Clin Endocrinol Metab
40. Wrutniak-Cabello C, Casa F, Cabello G. Thyroid hormone action on mitochondria. J Mol End