Soccer is the most popular sport in the world with 265 million subjects across all ages, sexes, and skill levels competing in the game (14). Female soccer players represent a growing portion of this population; however, research related to the physiological changes that occur in women because of soccer-specific training demands is lacking. A recent review by Datson et al. (7) evaluated the current literature on the demands that high-level female soccer players experience in games and found on average, these athletes cover about 10 km, perform 76 skill involvements (passing, dribbling, headers, and shooting), and experience 1,350–1,650 changes in activity during competitive play. In addition, these athletes maintain an average body fat (BF) percentage of 14.6–20.1% and a VO2max of 49.4–57.6 ml·kg−1·min−1 (7). Thus, to maintain this high level of play, coaches and athletes must optimize training to elicit the greatest performance benefits and maximize physiological attributes. To do this, systematic athlete monitoring has become increasingly common.
Athlete tracking and monitoring approaches have made recent technological advancements to encompass internal physiological markers (heart rate, heart rate variability, and biomarkers) (17). Heart rate monitoring is a commonly used technique to monitor on-field training load (TL), which represents internal “effort” to complete a physical task (4). This effort is quantified as TL via algorithms based on heart rate response specific to each athlete or as exercise energy expenditure (EEE) (4,6). Unfortunately, many TL-monitoring techniques (including heart rate monitoring) only account for what is happening on the field and are unable to capture off-field stressors. Implementation of additional monitoring tools, such as blood biomarkers, can give insight regarding athlete health, performance, and recovery status by encompassing both on and off the field stressors. Blood biomarkers can provide a comprehensive analysis of the physiological and biochemical response to TL that would otherwise be undetected through the more traditional monitoring techniques (15). Markers such as cortisol, testosterone, creatine kinase (CK), sex hormones, cytokines, hematological panels, and nutritional markers have been used to assess athletes' response to TL (11,12,15,18,26). In research, however, the use of biomarkers has been far more prevalent in male athletes with far less emphasis on female athletes despite known sex differences that could affect performance and recovery. This lack of diversity in the research is primarily driven by an unwillingness to work with female athletes because of hormonal variations associated with the menstrual cycle and use of oral contraceptives, although it seems counterproductive to exclude a large portion of the athletic population because of these factors.
Use of monitoring techniques allows for the athlete management to optimize performance and to potentially prevent injury and long-term decrements to accumulated TL and stress, which may manifest as nonfunctional overreaching (NFOR) or overtraining syndrome (OTS). Nonfunctional overreaching is defined as an accumulation of stress, physical workload, and psychological strain, resulting in short-term performance decrements without physiological and psychological maladaptation, whereas OTS includes both physiological and psychological maladaptations (17). Not only are NFOR/OTS detrimental to athletic performance and overall health, but full recovery may take months to years (17,25,33).
Applying these methods to team sports presents a unique challenge of assessing the team as a group and as individuals, in addition to considering the external stressors athletes are facing (10). Accounting for the individualized response to TL and accumulated stress provides coaches, trainers, and sport scientists the ability to tailor the athletes' workload and required recovery individually. Adequate monitoring is especially vital in collegiate athletes who experience increased physical and psychological stress because of the combination of a condensed season, TL, travel, academic requirements, changes in the environment, diet, and sleep patterns, which may all interact to inhibit athletic performance and recovery (16). Therefore, the purpose of this observational study was to evaluate the cumulative effects of season-long TL in conjunction with changes in performance and blood-based biomarkers associated with health, performance, and recovery in high-level Division I female soccer players. It was hypothesized there would be alterations in blood-based biomarkers, performance, and body composition over the course of the full season.
Experimental Approach to the Problem
This observational study sought to evaluate the season-long effects of training on various biomarkers in a real-world setting using high-level female athletes. Performance testing was preformed before the start of the season and 4–6 days after the final match to observe any performance changes. Training load variables were monitored throughout the duration of the competitive season including the preseason play, regular season play, and tournament play in National Collegiate Athletic Association (NCAA) Division I female soccer players. Biomarkers were analyzed before the start of preseason (T1) and every 4 weeks, approximately 18–36 hours after a game thereafter to evaluate the effects of accumulated stress of training on biomarkers representing performance, recovery, and general athlete health.
Twenty-five Division I female soccer players (± SD; Mage = 20 ± 1.1 years) were included. Descriptive and baseline performance data are presented in Table 1. All subjects performed testing as part of regular team activity associated with their sport science program. Subjects were asked not to change their diet over this period. All subjects received clearance by the Rutgers University sports medicine staff before testing and at the start of the season. This research was approved, and written consent was waived by the Rutgers University Institutional Review Board for the Protection of Human Subjects. All procedures performed were in accordance with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standard.
The testing timeline is presented in Figure 1. Athletes reported to the Rutgers University Center for Health and Human Performance (CHHP) before the start of preseason (P1) and 4–6 days after the final competitive match (P2) to complete a battery of 3 fitness tests in one session. Subjects were instructed to arrive euhydrated, at least two-hours fasted, and without having trained 24 hours before testing.
Body composition was assessed by air displacement plethysmography via the BodPod (BOD POD; COSMED, Concord, CA, USA) (8) to determine percent BF (%BF), fat-free mass (FFM), and fat mass. After a general systemic warm up, subjects were given 3 attempts for maximal single countermovement vertical jump (VJ) with arm swing using the Just Jump system (Probotics, Huntsville, AL, USA), with the highest jump recorded. After this, a maximal graded exercise test was used to measure maximal aerobic capacity (VO2max) and ventilatory threshold (VT) via direct gas exchange measured by a TrueOne 2,400 Metabolic Measurement System using a modified Bruce protocol (Parvo Medics, Sandy, UT, USA). Subjects continued the test with encouragement from the lab staff until volitional fatigue. At least 3 of the following criteria were met for attainment of VO2max: a leveling off or plateauing of VO2 with an increase in exercise intensity, attainment of age-predicted heart rate max, a respiratory exchange ratio greater than 1.10, and/or an RPE ≥18. Heart rate was continuously monitored using a Polar S610 heart rate monitor to accurately obtain maximal heart rate (HRmax) (Polar Electro Co., Woodbury, NY, USA). Ventilatory threshold was calculated after the completion of each test as the point where ventilation begins to increase nonlinearly with VO2, which is expressed as a percentage of VO2max (9).
Season Training Monitoring
All practices and games were monitored using the Polar Team2 system (Polar Electro Co.). Resistance training, although minimal throughout the season, was not monitored and details were not consistently provided to the researchers. Full-season TL can be found in Figure 2. The Team2 system monitored each player's individual workload, energy expenditure, and time spent at percentages of HRmax (55–65, 66–75, 76–85, 86–95, and 96–100). The quantification of an individual player's workload was estimated by total kcal expenditure (EEE) and TL, the latter being calculated via an algorithm developed by Polar based on physiological attributes of the player obtained from laboratory testing, which was entered for each player (height, weight, HRmax, VO2max, and VT), and physical workload measured (6).
Sample Collection and Analysis
The players reported to the CHHP for blood draw samples during 5 time points throughout the season. Preseason samples were drawn one day before the first day of practice, with players having refrained from training for 36 hours (T1); subsequent blood draws were conducted every 4 weeks approximately 18–36 hours after a game until the last competitive match (T2–T5). Athletes arrived at least 8 hours fasted and euhydrated between 07:00 and 09:00 hours. Blood samples were centrifuged for 10 minutes at 4,750 rpm (Allegra x-15R Centrifuge; Beckman Coulter, Brea, CA, USA) and were shipped to Quest Diagnostics for analysis by liquid chromatography-mass spectrometry/mass spectrometry-based assays. Biomarkers analyzed include free and total cortisol (CORTF, CORTT), prolactin (PRL), T3, IL-6, CK, sex-hormone binding globulin (SHBG), omega-3 (n-3FA), vitamin-D (Vit-D), iron (Fe), hematocrit (HcT), ferritin (Fer), percent saturation (%Sat), and total iron-binding capacity (TIBC).
Biomarker, performance, and body composition testing data were analyzed using repeated measures (RM) multivariate analysis of variances with RM analysis of variance univariate follow-ups (IBM SPSS v23). Planned simple contrasts were conducted using the baseline values as the comparison term. Pairwise contrasts were included in the case of significant univariate findings using the least significant difference method. The null hypothesis was rejected when p ≤ 0.05. Cohen's d was used to calculate effect sizes (ESs).
Performance and Training Load
Body composition and performance values can be found in Table 1. Body mass and %BF decreased from P1 to P2 (p < 0.05, ES = −0.17; p < 0.05, ES = −0.39) with no significant difference in FFM. VO2max and VJ decreased from P1 to P2 (p < 0.05, ES = −0.93; p < 0.05, ES = −0.28, respectively) with no significant difference in VT.
Training load was evaluated as the total sum during the 4-week training block between time points and can be found in Figure 2. T1–T2 had 18 practices (6 double sessions and 2 exhibition matches) and 4 games, T2–T3 had 15 practices and 6 games, T3–T4 had 13 practices and 7 games, T4–T5 had 11 practices and 7 games (including the first 3 rounds of the NCAA tournament). All subsequent training blocks were significantly lower (p < 0.05) than the initial preseason training block (T1–T2) (Figure 3). After the preseason, there was a substantial decrease in TL in the second training block (T2–T3) (ES = −1.35) followed by a further reduction in the third training block (T3–T4) (ES = −0.94) before normalizing through the last training block through the NCAA tournament.
Exercise energy expenditure was also evaluated as the total sum during the 4-week training block between time points (Figure 4). All subsequent training blocks were significantly lower (p < 0.05) than the initial preseason training block (T1–T2). After preseason, there was a substantial decrease in EEE in the second training block (T2–T3) (ES = −5.44), followed by a further reduction in the third training block (T3–T4) (ES = −3.79) before normalizing through the last training block.
All biomarker data can be found in Table 2. Compared with T1, CORTT was significantly higher at T3–T5 (p < 0.05). There was an initial increase at T3 (ES = 0.61) followed by a second larger increase at T5 (ES = 0.91). Similarly, CORTF was significantly higher at T3–T5 compared with T1 (p < 0.05). There was a similar pattern showing an initial increase at T3 (ES = 1.36), followed by a second increase at T5 (ES = 1.0). Compared with T1, PRL significantly decreased at T2 (p < 0.05, ES = −0.38) before significantly rising at T3 (p < 0.05, ES = 1.63) and remained elevated through T4 and T5 (p < 0.05). Creatine kinase significantly increased at T2 (p < 0.05, ES = 0.85) before returning to baseline values at T3 and T4. Creatine kinase increased again and reached its highest value at T5 (p < 0.05, ES = 1.08). IL6 remained at baseline values before significantly increasing at T5 (p < 0.05, ES = 5.73). T3 significantly increased at T2 (p < 0.05, ES = 1.15) and remained significant through T3 before returning to baseline values at T4 and T5. N-3FA was significantly lower at all time points (T2–T5) compared with T1 (p < 0.05). Compared with baseline, Vit-D significantly decreased at T4 (p < 0.05, ES = −0.44) and remained significantly lower through T5 (p < 0.05). There were no significant changes seen in SHBG.
All hematological values can be found in Table 3. Compared with T1, Fe was significantly lower at T2–T5 (p < 0.05). There was an initial decrease at T2 (ES = −0.83) which remained stable before a second decline happened at T5 (ES = −0.75). Similarly, Fer was significantly lower at T2–T5 compared with T1 (p < 0.05) with an initial decrease at T2 (ES = −0.76), which continued through T5 with little deviation. Compared with baseline, TIBC was significantly higher at T3–T5 (p < 0.05), whereas Hct was significantly lower at all time points (T2–T5) (p < 0.05).
The results of this study provide a comprehensive evaluation of the cumulative stress of a season incorporating TL, EEE, performance, and biomarkers of health, performance, and recovery in collegiate female athletes. This study exhibited the highest TL and EEE during preseason which corresponded with several physiological perturbations, that persisted throughout the season. The authors observed a first hormonal disruption occurring at week-8 (T3) with a second hormonal disruption at week-16 (T5). To our knowledge, this is the first study to track TL, EEE, and biomarkers in female team sport athletes through the preseason, competitive season, and tournament play.
The TL and EEE found in this study show the high metabolic demand these athletes encounter throughout a college soccer season, which is notable for its congested match fixture and short preseason. Both TL and EEE were significantly higher during preseason training compared with the rest of the season. The TL during the first block (T1–T2) corresponded to notable physiological changes, primarily seen in dietary and hematological markers. These changes may be a byproduct of NCAA restrictions on athlete monitoring during the summer months. This restriction shapes the nature of the collegiate preseason itself. It is a very short, intense, 2-week period which uses multiple practices per day in conjunction with the stress of competition along with team, academic, and administrative meetings, thus compromising optimal recovery. These results indicate the potential negative impact of a short preseason on the athletes if not adequately managed.
Although TL exhibited a steady decline as the team progressed through the season and entered tournament play, the decrease in weight, %BF, VO2max, and VJ observed at P2 depict the cumulative stress of the collegiate soccer season. Along with season-long TL, student-athletes experience the challenges and stressors of doing both athletics and academics creating increased overall stress that is often not accounted for (16). The changes in fitness found in this study have also been observed in other studies in female collegiate soccer players where significant decreases in VO2max were observed over the course of a season (32). However, the changes in body composition were unique regarding the decrease in total body weight and %BF without any change in FFM. This could also be due to the minimal resistance training performed throughout the season, although this cannot be confirmed because of the inability to monitor athletes during these activities. It is notable that, despite the maintenance of FFM, aerobic capacity and power still declined, and biomarker perturbation was still evident. These results show that during the most competitive phase of the season (tournament play), athletes are at their lowest fitness levels, possibly because of the underlying physiological response to the accumulated TL, EEE, or insufficient recovery.
Along with changes in anthropometric and performance variables, biomarkers can provide further insight into the physiological changes that athletes experience during the season. One of the more common biomarkers used is cortisol, a primary hormone released in response to stress. During times of increased TL and less than optimal recovery, an elevation in resting cortisol can be seen due to disruption in homeostasis and the subsequent stress response (2,17). If this inappropriate TL continues, the stress response can become desensitized, resulting in a decreased cortisol response (17). It is important to note that is has been suggested that resting cortisol is not a useful measure because of the lack of change seen in male endurance athletes (17). However, a recent review indicated these notions may be overstated because of a contradictory literature base with varying results, thus highlighting the need for further investigation (2). Furthermore, these recommendations may not apply to female power endurance athletes who experience a more physical (contact-oriented) TL rather than the more aerobic-based male endurance athletes that are often studied (17). The nature of the college soccer season with congested match fixtures may further exacerbate this. The current study revealed significant elevations in both CORTT and CORTF at T3 with the second inflection at T5, indicating an accumulated stress response throughout the season. It is important to note that at all 5 time points, CORTT values were above the clinical range. The authors believe these chronically high cortisol levels could be due to the lack of sex-specific ranges or due to the effects of menstrual disruption and oral contraceptives on the hypothalamic pituitary adrenal (HPA) axis (21).
Somewhat surprisingly, there were no changes in cortisol immediately following the highest TL (T1–T2), which may indicate a potential delayed disruption of the HPA-axis or other compensatory adjustments (17). However, the increased cortisol response evident at T3 and T5, suggests the total cumulative stress may begin to alter HPA-axis activation. These results are contrary to the lack of change or even decreased cortisol response that has been seen in both collegiate and professional male soccer players, providing additional support for the need for more female-specific ranges and data (26).
Testosterone is commonly measured in conjunction with cortisol in males to show the anabolic:catabolic balance of the athlete. Although testosterone was not evaluated because of the low resting values in females, secondary indicators such as SHBG and PRL were used as exploratory alternatives. Sex-hormone binding globulin acts as a transport vessel for sex hormones and has been shown to increase in both males and females with exercise (15). Prolactin, which has been shown to increase in response to stress, hypoglycemia, and physical exercise, has relevance in female athletes because of its suppressive effects on estrogen when elevated (31). Similar to other observed hormones, PRL increased in the middle of the season (T3) and remained elevated throughout the rest of the season, whereas SHBG showed no change. The similar responses of PRL and CORT provide additional support for the likelihood of the cumulative effects of TL and stress on endocrine perturbation. Furthermore, an HPA/hypothalamic pituitary gonadal (HPG) axis disruption occurs with inappropriate TL (13), potentially affecting PRL which may play a role in reproductive cycle dysfunction and contribute to the female athlete triad (21). PRL appears to be an important biomarker in female athletes, although more research is needed along with the addition of markers such as estrogen. A lab system error at baseline resulted in the lack of completion of the estrogen analysis that was beyond the control of the researchers.
Along with the changes in the HPA/HPG responses, there were also changes in markers of muscle damage as reflected by CK (18). The initial increase in CK was associated with the high TL in the preseason. It is important to note that all markers remained in the athlete- and sex-specific ranges (female athlete reference range: 47–513 U·L−1) (20). Similar results have been seen in both collegiate and professional soccer players, with an increase in CK after preseason training, although within the normal athlete ranges (11,18,26). The results of high CK may indicate strenuous periods within the season where additional recovery strategies should be implemented.
IL6 was evaluated as a marker of inflammation, as it responds to decreased muscle glycogen and muscle contraction, along with muscle damage and injury to facilitate an immune response (22). With systemic inflammation resulting from chronic intense exercise, cytokines, such as IL6, activate the HPA-axis (29). Not surprisingly then, IL6 followed a similar trend as other hormones linked to the HPA-axis such as CORTT. There was a notable elevation of IL6 at T3 and again at T5, representing the highest state of physiological disruption. These results, specifically the elevation at T5, are consistent with the cytokine hypothesis of overtraining which posits that the increased systemic inflammation can be a major disruptor of the HPA-axis (29). Along with the effects mentioned above, IL6 also stimulates the expression of hepcidin in the liver, which has been shown to decrease iron absorption, further exacerbating the changes in iron observed in this study (27). The physiological alterations observed at T3 and T5 may result from a multi-faceted hormonal response to the cumulative stress of the condensed collegiate season.
Thyroid hormones are influenced by energy balance and contribute to performance and recovery by regulating metabolism (30). The results of this study show an increase in metabolically active T3, after the preseason which returned toward baseline as the season progressed. The initial increase at T2 was a unique finding in that a decrease (or no change) in T3 is often seen with an increase in EEE to provide an “energy-sparing” effect (30). These results are contrary to what was found in collegiate female rowers who displayed decreased thyroid markers during periods of high EEE over a 20-week in-season training block (3). Future studies may benefit from the use of a combination of TSH, T3, and T4 to get a complete profile of thyroid function.
Dietary biomarkers can also provide additional information for athlete readiness. Vit-D and n-3FA have significant effects on female athletes with performance implications including bone health, muscle damage, and inflammation (15). In this study, an immediate and sustained decrease in n-3FA after the preseason was found, indicating that there may be insufficient dietary compensation to account for resource recruitment. Furthermore, Vit-D decreased toward the end of the season, which also coincides with the change of seasons: Games and practices are often transitioned to indoor settings or more clothes are worn in November in the northeastern United States. In light of the observed decreases in Vit-D and n-3FA, supplementation may have an impact on the recovery status of the athletes and provide a more favorable physiological environment to maximize recovery (19).
The hematological markers evaluated in this study represent both the Fe and transferrin statuses of these athletes (23,24,28). Fe status consists of total Fe in the blood and the amount stored as Fer, which is mobilized in times of decreased Fe (23). Transferrin status incorporates TIBC which represents the capacity of Fe to bind to transferrin while %Sat represents the occupied iron-binding sites on transferrin (24,28). Changes in these markers show a shift toward a training-induced Fe deficiency or anemia. After the preseason training block (T1–T2), there was a negative change Fe and transferrin status, which persisted throughout the season. Fe reached its lowest values at T5 (56% lower than the initial baseline values). Similarly, Fer demonstrated a 35% decrease by this timepoint, whereas TIBC and %Sat decreased by 9 and 40%, respectively. Although these values never went below the “clinically normal” range (Fe: 8.95–31.32 µmol·L−1, Fer: 10–154 µg·dL−1, TIBC: 44.75–80.55 µmol·L−1, %Sat: 11–50%) (15), the change from baseline likely represents a suboptimal range for performance given the magnitude of change. Additional measurements of Hct showed decreases after the preseason. It is well known that negative changes in iron status result in significant decreases in performance and overall exercise capacity, e.g., reduced VO2 max (15). The results found in this study are consistent with previous research, indicating that a high number of female athletes experience declines in hematological values during the season (1). Furthermore, it is important to recognize differences between clinical ranges and optimal ranges for athletic performance. Clinical ranges are generalized and concrete numbers for diagnosis which do not take into consideration changes from the athlete's baseline.
This study provides a comprehensive evaluation of TL, EEE, and biomarkers in high-level female soccer players throughout the season and during tournament play. We recognize a few limitations with the study, some of which are inherent to working with high-level athletes in a real-world setting. Expectations and demands on the athletes and coaches related to the research must be balanced with the reality of the season. First, the athlete's diet was not measured throughout the season. However, research has indicated that self-reported dietary measures can be highly flawed and can be unreliable and impractical to measure in this population during an already demanding season (5). Secondly, this study did not account for the menstrual cycle or the use of oral contraceptives, although a 28-day period between blood draws was used to provide a degree of “control” for the menstrual cycle. It is the authors' views that because the athletes cannot control for the menstrual cycle during competition, it is essential to evaluate their response in a holistic and real-world analysis. In addition, sleep quantity/quality was not evaluated before blood draws. Although this was impractical to attempt to control these factors when working with this team, future researchers may consider reasonable ways to practically assess this. Future studies would benefit from tracking diet, menstrual cycle, mood disturbances, sleep quality, and additional biomarkers including estrogen, testosterone, and a more comprehensive thyroid panel when possible. It is important to note that including a degree of control as suggested for future studies may prove problematic in free-living athletes if it becomes too invasive and disruptive for coaches and athletes.
Despite the limitations, this study provides much needed observational data on high-level female athletes in a real-world setting. The results of this study show that female athletes experience a culmination of decreased performance and significant physiological disruptions in conjunction with increased cumulative TL/EEE and external stressors during the collegiate competitive season. The preseason/early season training block resulted in a negative change in dietary and hematological markers that persisted throughout the season. As the season progressed, there appeared to be a delayed hormonal response associated with the cumulative stress, which may indicate the onset of NFOR. These results emphasize the importance of tracking TL and biomarkers through a full season to show the cumulative effects to the on- and off-field stressors placed on collegiate athletes. Biomarkers, in conjunction with TL monitoring, provide a more complete profile on athlete readiness and overall health, allowing for better player management. Periodic evaluations provide several opportunities to intervene and potentially mitigate the performance decrements as seen in this study. Possible supplementation of Fe, n-3FA, and Vit-D may prove beneficial for many female athletes to maintain performance throughout the entire season. Monitoring techniques can be used to make in-season adjustments to maximize player performance.
Special thanks to the Rutgers Women's Soccer Team. This study was funded by Quest Diagnostics. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The authors have no conflicts of interest to report.
1. Alaunyte I, Stojceska V, Plunkett A. Iron and the female athlete: A review of dietary treatment methods for improving iron status and exercise performance
. J Int Soc Sports Nutr 12: 38, 2015.
2. Anderson T, Wideman L. Exercise and the cortisol awakening response: A systematic review. Sports Med Open 3: 37, 2017.
3. Baylor LS, Hackney AC. Resting thyroid and leptin hormone changes in women following intense, prolonged exercise training. Eur J Appl Physiol 88: 480–484, 2003.
4. Bosquet L, Merkari S, Arvisais D, Aubert E. Is heart rate a convenient tool to monitor over-reaching? A systematic review of the literature. Br J Sports Med 42: 709–714, 2008.
5. Burke LM, Lundy B, Fahrenholtz IL, Melin AK. Pitfalls of conducting and interpreting estimates of energy availability in free-living athletes. Int J Sport Nutr Exerc Metab 28: 350–363, 2018.
6. Ceesay SM, Prentice AM, Day KC, Murgatroyd PR, Goldberg GR, Scott W. The use of heart rate monitoring in the estimation of energy expenditure: A validation study using indirect whole-body calorimetry. Br J Nutr 61: 175–186, 2018.
7. Datson N, Hulton A, Andersson H, et al. Applied physiology of female soccer: An update. Sports Med 44: 1225–1240, 2014.
8. Dempster P, Aitkens S. A new air displacement method for the determination of human body composition. Med Sci Sports Exerc 27: 1692–1697, 1995.
9. Gaskill SE, Ruby BC, Walker AJ, Sanchez OA, Serfass RC, Leon AS. Validity and reliability of combining three methods to determine ventilatory threshold. Med Sci Sports Exerc 33: 1841–1848, 2001.
10. Halson SL. Monitoring training load
to understand fatigue in athletes. Sports Med 44(suppl 2): S139–S147, 2014.
11. Heisterberg MF, Fahrenkrug J, Krustrup P, Storskov A, Kjær M, Andersen JL. Extensive monitoring through multiple blood samples in professional soccer players. J Strength Cond Res 27: 1260–1271, 2013.
12. Kraemer WJ, French DN, Paxton NJ, et al. Changes in exercise performance
and hormonal concentrations over a big ten soccer season in starters and nonstarters. J Strength Cond Res 18: 121–128, 2004.
13. Kreher JB, Schwartz JB. Overtraining syndrome: A practical guide. Sports Health 4: 128–138, 2012.
14. Kunz M. 265 Million playing football. FIFA Mag 11–13, 2007.
15. Lee EC, Fragala MS, Kavouras SA, Queen RM, Pryor JL, Casa DJ. Biomarkers in sports and exercise: Tracking health, performance
, and recovery in athletes. J Strength Cond Res 31: 2920–2937, 2017.
16. Mann JB, Bryant K, Johnstone B, Ivey P, Sayers SP. The effect of physical and academic stress on illness and injury in Division 1 college football players. J Strength Cond Res 30: 20–25, 2015.
17. Meeusen R, Duclos M, Foster C, et al. Prevention, diagnosis, and treatment of the overtraining syndrome: Joint consensus statement of the european college of sport science and the American College of Sports Medicine. Med Sci Sports Exerc 45: 186–205, 2013.
18. Meyer T, Meister S. Routine blood parameters in elite soccer players. Int J Sports Med 32: 875–881, 2011.
19. Mickleborough TD. Omega-3 polyunsaturated fatty acids in physical performance
optimization. Int J Sport Nutr Exerc Metab 23: 83–96, 2013.
20. Mougios V. Reference intervals for serum creatine kinase in athletes. Br J Sports Med 41: 674–678, 2007.
21. Otis C, Drinkwater B, Johnson M, Loucks A, Wilmore J. American College of Sports Medicine position stand: The female athlete triad. Med Sci Sports Exerc 29: 1–9, 1997.
22. Pedersen BK, Febbraio MA. Muscle-derived interleukin-6: A possible link between skeletal muscle, adipose tissue, liver, and brain. Brain Behav Immun 19: 371–376, 2005.
23. Pedlar C, Brugnara C, Bruinvels G, Burden R. Iron balance and iron supplementation for the female athlete: A practical approach. Eur J Sport Sci 18: 295–305, 2017.
24. Podgórski T, Kryściak J, Konarski J, et al. Iron metabolism in field hockey players during an annual training cycle. J Hum Kinet 47: 107–114, 2015.
25. Purvis D, Gonsalves S, Deuster PA. Physiological and psychological fatigue in extreme conditions: Overtraining and elite athletes. PM R 2: 442–450, 2010.
26. Silva JR, Rebelo A, Marques F, et al. Biochemical impact of soccer: An analysis of hormonal, muscle damage, and redox markers during the season. Appl Physiol Nutr Metab 39: 432–438, 2014.
27. Skarpańska-Stejnborn A, Basta P, Trzeciak J, Szcześniak-Pilaczyńska Ł. Effect of intense physical exercise on hepcidin levels and selected parameters of iron metabolism in rowing athletes. Eur J Appl Physiol 115: 345–351, 2014.
28. Smith DJ, Roberts D. Effects of high volume and/or intense exercise on selected blood chemistry parameters. Clin Biochem 27: 435–440, 1994.
29. Smith LL. Cytokine hypothesis of overtraining: A physiological adaptation to excessive stress? Med Sci Sport Exerc 32: 317, 2000.
30. Steinacker J, Brkic M, Simsch C, Kresz A, Prokopchuk O, Liu Y. Thyroid hormones, cytokines, physical training and metabolic control. Horm Metablic Res 37: 538–544, 2005.
31. Suay F, Salvador A, González-Bono E, et al. Effects of competition and its outcome on serum testosterone, cortisol and prolactin. Psychoneuroendocrinology 24: 551–566, 1999.
32. Todd A, Jerome J, Alfred A. Seasonal changes in VO2max among Division 1A collegiate women soccer players. J Strength Cond Res 21: 48–51, 2007.
33. Wyatt F, Donaldson A, Brown E. The overtraining syndrome: A meta-analytic review. J Exerc Physiol 16: 12–23, 2013.