Female Athlete Research Camp: A Unique Model for Conducting Research in High-Performance Female Athletes.

The FARC 1.0 was designed to address a variety of researcher and participant challenges thought to contribute to the poor quality and low quantity of SES research involving female athletes as participants. Table 1 provides a summary of the global issues/challenges, the approach taken by our research team to address these issues within the specific design of this project, and a general summary of the outcome. This summary is presented to facilitate an appreciation of the FARC 1.0 rationale and implementation within the larger goal of enhancing the participation of female athletes in high-quality SES research ( 9 ). Specific details of the FARC 1.0 protocols related to tracking and assessment of MC function, MC phase, as well as symptomology, are presented in greater detail below. Summary of the global and specific issues facing athletes and research staff when undertaking female athlete research, as well as our approach to addressing these issues within the specific design of this project, and a general summary of the outcome. AIS FPHI, Australian Institute of Sport Female Performance and Health Initiative; MMIHR, Mary MacKillop Institute for Health Research; ACU, Australian Catholic University; NRL, National Rugby League; WTHPA FAH, Wu Tsai Human Performance Alliance Female Athlete Hub, Boston Children’s Hospital. Australian First Nation athletes from the National Rugby League Indigenous Women’s Academy pathways program were recruited for this study. Athletes were highly trained (Tier 3 per McKay et al. criteria; ( 7 )) and were nominated for participation by the National Sporting Organization. No exclusion criteria based on MC function were implemented, and all athletes were deemed eligible if they were able and willing to complete the study requirements. Although 43 athletes provided written informed consent, barriers to participation, meant only 25 athletes attended FARC 1.0. One athlete was required to return home during the study, leaving 24 complete data sets for analysis. The recruitment process, barriers to participation and athlete characteristics have been summarized in Figure 1 . Ethics approval to conduct this study was obtained from the Australian Catholic University’s Human Research Ethics Committee (2021-285HI). Importantly, athletes were not disadvantaged if they did not wish to partake in the research study and were still offered the opportunity to attend the camp. Flow chart which outlines the recruitment process, reasons for exclusion and the final athlete cohort’s MC function classification. This study consisted of an 11-wk period of MC characterization followed by a 5-wk residential training camp held at the Australian Institute of Sport (Canberra, Australia). Throughout the duration of the study, athletes completed a daily survey upon waking and reported their MC charateristics and any associated symptoms. During the training camp, athletes reported to the laboratory on three occasions determined by each athlete’s individual MC and provided a venous blood sample. The resistance-exercise and skills-based training undertaken throughout the camp was prescribed by the Academy coaches and comprised of three gym sessions and two field sessions per week. Before FARC 1.0 commenced, athletes were asked to complete a survey which documented their MC history, including any current or previous HC use (type, formulation), the length and frequency of their MC (including determination of any primary and secondary amenorrhea) and prevalence of known menstrual diagnoses (e.g., polycystic ovary syndrome [PCOS], endometriosis; Supplemental Digital Content, Preliminary female athlete questionnaire, http://links.lww.com/MSS/C975 ). One athlete reported a prior diagnosis of PCOS, with no other MC dysfunction was reported by other athletes. Subsequently, for the duration of the 16-wk period (11-wk precamp tracking and 5-wk training camp), athletes received a text message at 8:00 am every day with a link to an electronic survey (Supplemental Digital Content, Female athlete research camp daily monitoring, http://links.lww.com/MSS/C975 ). This survey was customized to suit either a naturally cycling athlete (athleteNC) or athlete using HC (athleteHC). All athletes were asked to report on the presence and heaviness of menstruation, the occurrence of any symptoms and their use of medications across the previous 24 h. In addition, athleteNC were asked to use a dual hormone ovulation kit (Advanced Digital Ovulation Test, Clearblue, Geneva, Switzerland) from MC day 8 until ovulation occurred. If ovulation was not detected, athletes were told to stop using their ovulation kits on day 17; however, additional days were often requested according to individual MC characteristics to increase the chances of capturing ovulation. During FARC 1.0, each athlete attended the laboratory on three occasions specific to her MC function and MC phase. The precamp MC monitoring was used to identify potential testing dates for each athlete for camp planning, with confirmation of actual testing days occurring 24 h prior. In athleteNC, athletes tested in phase 1 (low estradiol and low progesterone), phase 2 (elevated estradiol and low progesterone), and phase 4 (elevated estradiol and elevated progesterone) of the MC, as described by Elliott-Sale et al. ( 9 ). Although phase 3 is also of interest, pragmatic challenges around resources and athlete availability limited data collection to three phases. Phase 2 was selected in preference to Phase 3 due to the larger differential between sex hormones (elevated estradiol and low progesterone) and the scarcity of research examining this phase. Phase 1 testing generally occurred on day 2, after the athlete reported their menstrual period to the research team or as close to this time as possible. Phase 2 was determined by a “flashing smile” icon on the urinary ovulation kits which indicated elevated estrogen metabolite concentrations and preceded the rise in luteinizing hormone (LH). To identify Phase 2, ovulation kit results were reported in the daily survey and/or communicated directly to the lead researcher. Because serial blood samples were not taken in the days before Phase 2, the “peak” in estrogen concentrations could not be confirmed. However, analysis of venous blood samples verified the absence of late collections (i.e., Phase 3). Phase 4 was determined as 7 d after ovulation, and in cases where ovulation was not determined, an arbitrary “day 21” was used for testing instead. For athleteHC, three arbitrary testing dates were selected, each separated by 7 to 10 d to replicate the pattern of blood collection from the naturally cycling athletes. We had originally intended that athletes taking the oral contraceptive pill (OCP) would be tested twice during the active pill-taking days (i.e., days 8–28) and once during the withdrawal period (days 1–7). However, without instruction from the research team, all athletes taking the OCP purposefully manipulated their cycles to experience a withdrawal bleed in the days before arrival in camp and expressed their intention to skip further withdrawal bleeds while in camp. Accordingly, we chose to instead capture three active pill-taking days, with the timing of pill ingestion standardized before each testing day. Accordingly, all athleteHC were tested at time points where an assumed, consistent supply of exogenous hormone provision occurred. During each laboratory visit, athletes arrived in a fasted and rested state at the same time of day (±15 min). An 8.5-mL venous blood sample was then collected from an antecubital vein into a serum separator tube by a trained phlebotomist. Blood tubes were left to clot at room temperature for 30 min, before centrifugation at 2200 G for 10 min at 4°C. Remaining serum was split into aliquots and stored at −80°C until batch analysis could occur. Estradiol, progesterone, LH, follicle stimulating hormone (FSH), and growth hormone (GH) were measured via an Access 2 Immunoassay System (Beckman Coulter, Brea, CA). Finally, total testosterone and sex hormone binding globulin (SHBG) were analyzed using liquid chromatography-tandem mass spectrometry (Waters UPLC-TQX S, Waters Corp., Wilmslow) and free testosterone was subsequently calculated. The retrospective analysis of hormone concentrations was used to confirm cycle phase. All statistical analysis was performed using R Studio (v4.0.2; R Foundation for Statistical Computing, Vienna, Austria). Hormonal data were analyzed using linear mixed models with restricted maximal likelihood estimates. Normality was assessed via visual inspection of residual and q-q plots, where progesterone, estradiol, and LH were deemed to be nonnormally distributed. Accordingly, log-transformation of these three variables occurred before analysis. Fixed effects for MC function (athleteNC or athleteHC) and cycle phase ( 1 , 2 , 4 ) were included, with a random effect of subject ID. Statistical significance of fixed effects occurred using type II Wald tests with Kenward–Roger degrees of freedom, with post hoc analysis performed using Tukey adjustments. To assess the frequency of symptoms reported daily, cumulative link mixed models were employed, which account for ordered, ordinal data. Menstrual cycle function (athleteNC or athleteHC) and presence of bleeding (yes vs no) were included as fixed effects, with subject ID included as a random effect. In this instance, bleeding refers to either menstrual blood loss (athleteNC) or breakthrough/withdrawal bleeding experienced by athleteHC. Model estimates were used to calculate odds ratios (OR) and 95% confident intervals (CI). Finally, each symptom that had an incidence rate of >2.5% was individually analyzed using a general linear mixed effects model approach for binary data. Symptoms that did not reach this threshold included bladder incontinence, diarrhea, nausea, night sweats and pelvic pain. The aforementioned fixed and random effects were included in each model. Significance for all analyses was accepted at P < 0.05.

Of the 24 athletes who participated in FARC 1.0, 11 athletes were naturally cycling (athleteNC) and 13 were using hormonal contraception (athleteHC). In the athleteNC cohort, only one athlete reached criteria for classification as eumenorrheic defined as having a MC length between 21 and 35 d, ≥9 periods/year, no HC use in the previous 3 months, and observation of the expected hormonal profile ( 9 ). In the case of five athletes, retrospective blood sampling was unable to confirm the expected hormonal profile, instead finding that estradiol concentrations were higher during Phase 4 than phase 2. A further five athletes had evidence of MC dysfunction, including an anovulatory cycle ( n = 1), oligomenorrhea ( n = 3) and luteal phase deficiency ( n = 1) as described by Elliott-Sale et al. ( 9 ). Two of the oligomenorrheic athletes had hormonal profiles that were deemed atypical; one due to recent HC use (~ 4 months postcessation) and the other for unknown reasons. Amenorrhea was not evident in this cohort. Of the athleteHC ( n = 13), four used combined-monophasic formulations of the OCP, eight had contraceptive implants and the remaining athlete used contraceptive injections. When athletes were asked about their reasons for using HC, answers varied from “prevention against pregnancy” ( n = 7; 54%), “to reduce bleeding” ( n = 4; 31%), “to reduce pain” ( n = 4; 31%), “to regulate MC for performance” ( n = 3; 23%), and “to control acne” ( n = 1; 13%). A summary of the MC status of participants has been provided in Figure 1 . In athleteNC ( n = 11), testing of Phases 1, 2, and 4 occurred on day 1.8 ± 0.4, 11.4 ± 1.4 and 20.8 ± 1.6 of the MC, respectively. Of the 10 athletes who ovulated, ovulation occurred on day 14.9 ± 2.5, as evident by a surge in LH on a urinary ovulation kit. Hormone concentrations are presented in Table 2 . A significant interaction for progesterone occurred ( P < 0.001), where concentrations increased during Phase 4 compared with Phases 1 ( P < 0.001) and 2 ( P < 0.001) in athleteNC. A similar interaction for estradiol was seen ( P < 0.009), where concentrations were elevated during Phase 4 compared with Phase 1 in athleteNC ( P = 0.001), with a trend toward higher concentrations in Phase 2 compared with Phase 1 ( P = 0.064). No changes in either progesterone or estradiol were evident across phases in athleteHC (all P > 0.05). A main effect of phase was evident for FSH ( P = 0.004), where both groups showed reduced concentrations during Phase 4, comparative to Phase 1 ( P = 0.008) and phase 2 ( P = 0.010). No changes in GH, LH, total testosterone or SHBG were evident between groups or phases ( P > 0.05). Calculated free testosterone was significantly higher in athleteNC compared athleteHC ( P = 0.041). Hormone concentrations for the naturally cycling athletes (AthleteNC; n = 11) and athletes using hormonal contraception (AthleteHC; n = 13) during estimated phases 1, 2, and 4. a Significantly different to phase 1. b Significantly different to phase 2. c Significantly different to HC users. Note, data included here is for all 24 athletes involved in the study, and therefore includes some naturally menstruating athletes who did not achieve the hormonal profile for eumenorrheic classification of phase 2 or phase 4. A total of 2397 daily surveys were completed across the 16-wk period by 24 athletes, which reflects a compliance rate of 99.3%. A total of 1465 symptoms were reported across this period. The prevalence of symptoms on any given day was similar between athleteNC (33.7%) and athleteHC (22.9%; OR: 2.1 [95% CI: 0.31, 10.8]; P = 0.376). However, of the 1465 symptoms reported, 949 symptoms occurred in athleteNC (86/athlete) and 516 in athleteHC (40/athlete), suggesting that when athleteNC did reported symptoms, they were more likely to report multiple symptoms. Regardless of MC function, there was a significant, positive association between bleeding and symptoms (OR: 12.5 [9.6, 16.5], P < 0.001), where athletes were more likely to report one or more symptoms on bleeding (50.1%) compared with nonbleeding days (22.0%). The total number of bleeding days across the 16-wk study period were 20.9 and 19.5 d/athlete for the athleteNC and athleteHC, respectively. The athleteNC cohort classified 16% of bleeding days as heavy, with 53% and 31% as moderate and light, respectively. Furthermore, in athleteHC 12% of bleeding days were heavy, with 58% defined as moderate and 30% as light. The most reported symptoms were abdominal cramps (15.8%), followed by fatigue (12.9%), mood changes (10.8%), bloating (10.1%), and acne (10.0%). When comparing athletes of different MC functioning, acne was the only symptom that differed between the groups ( P = 0.047), with a higher incidence of symptoms reported in athleteNC (Table 3 ). For nearly all symptoms with an incidence rate >2.5%, there was a greater frequency of symptoms reported on bleeding, compared with nonbleeding days ( P < 0.05; Table 3 ). The exceptions were acne ( P = 0.442), constipation ( P = 0.196) and sleep changes ( P = 0.075), where the frequencies were similar between bleeding and nonbleeding days. Incidence of menstrual cycle symptoms reported in hormonal contraceptive users (AthleteHC) and naturally cycling athletes (AthleteNC) and on bleeding and nonbleeding days. Only symptoms with an incidence of >2.5% are reported here. a Significant effect.

This article summarizes the implementation of a novel research protocol that investigated MC characteristics of highly trained female athletes. FARC 1.0 was implemented as a 5-wk residential research-embedded training camp, preceded by tracking ~3 full MCs. The 25 participants represented the National Rugby League’s Indigenous Women’s Academy and completed a structured resistance-exercise and skills-based training program that also integrated a battery of standardized tests at specific phases of interest within each participant’s specific MC. This project was a response to improve the quality and quantity of SES research involving highly trained female athletes. The paper summarizes strategies to address the challenges of undertaking research with female athletes, noting global issues and the protocols that were employed in FARC 1.0 to address them (see Table 1 ). It demonstrates a novel strategy to characterize and investigate the effect of MC function and MC phase on various metrics of interest: implementing a test battery during MC Phases 1, 2, and 4 in athleteNC (or compatible occasions in athleteHC) as they occurred specifically for each participant over the 5-wk camp duration. The important information provided in this article include the following: 1) Insights from the study planning and implementation that will assist other research groups to conduct their own research with female athletes. 2) The backdrop of the MC characteristics of the current cohort of First Nation female Rugby League players against which data from future articles will be compared. 3) Summary of the authentic findings from an investigation of the MC characteristics of a cohort of female athletes of similar background and training experience, noting the reality of variability in the profiles of female reproductive hormones from “textbook” MC and commentary on achieving optimal characterization and standardization of the MC in SES research. Hormonal contraception was used by 54% of our study participants; a figure comparable to previous reports in highly trained athlete cohorts ( 10 , 11 ). However, contraceptive choices among the present cohort were disparate from previous reports of Australian athletes, with the majority selecting a hormonal implant (62%), followed by the OCP (31%) and hormonal injections (8%). In a study by McNamara et al. ( 10 ), elite Australian female athletes reported OCP (62%) as the most frequently used contraceptive, followed by the Mirena hormone-releasing intrauterine device (34%). Although this discrepancy may be attributed to availability and cultural differences associated with our Australian First Nation’s cohort ( 12 ), we note that the self-reported reasons for using contraception were similar to those cited by previous investigations of Australian athletes ( 13 ), which include for contraception purposes, to reduce pain and to reduce bleeding. In our athleteNC cohort, 6/11 athletes (55%) had MC dysfunction and/or diagnosed menstrual conditions. Amenorrhea was not detected in this cohort, whereas “silent” conditions such as luteal phase deficiency and anovulation were prevalent. This highlights the importance of tracking and monitoring MC in athletes, as regular bleeding does not always imply a healthy MC. Overall, there was a relatively low incidence of symptoms reported throughout the 16-wk monitoring period, and the total number of days where symptoms were reported was similar between athleteNC and athleteHC. This was somewhat unexpected given that contraception is commonly chosen by athletes as a strategy to minimize pain/negative side-effects associated with the MC ( 10 , 13 ). However, the presence of bleeding was strongly associated with increased symptomology, and a similar number of bleeding days was reported in athleteNC and athleteHC, despite some OCP users manipulating their cycle in camp. Nevertheless, more symptoms were reported in athleteNC across the study period, suggesting that it was common for multiple symptoms to be reported in comparison to athleteHC. Interestingly, a recent study in recreationally women found similar outcomes, where no differences in symptomology were apparent between naturally menstruating participants and OPC users, and symptom magnitude was increased while bleeding; this was associated with a perceived reduction to exercise performance and recovery ( 14 ). Collectively, these findings highlight the need to track the perceived effects of the MC in all female athletes, regardless of their reproductive hormone profile. In the absence of consensus on, or validation of, a preferred approach or tool for MC tracking, we used a list of symptoms associated with popular MC smart phone applications in our daily electronic survey. Important to note is that we did not assess severity of symptoms, and therefore have limited opportunity to measure the impact of the symptoms on performance and health. Furthermore, many of these symptoms are not unique or specific to the MC. For instance, some of the most frequently reported symptoms were “Fatigue” and “Mood Changes,” both of which are prevalent among other conditions [e.g., iron deficiency ( 15 )] and psychological states [e.g., stress ( 16 )]. Furthermore, high training loads will also impact many of these symptoms, and dissociating the impact of training from the MC is difficult. Collectively, there is a need to develop a validated and reliable athlete specific scale for measuring the impact of the MC on both performance and health. Such a resource will be invaluable for coaches and practitioners working with female athletes. One objective of the present study was to capture three key phases of the MC in a scenario of high ecological validity for athletes. The most challenging phase to capture was Phase 2, as it is short in duration and it has been suggested that daily hormone samples are necessary to isolate the period of desired differential between estrogen and progesterone concentrations ( 9 ). As a result of these methodological difficulties, this phase has been largely ignored by researchers. Our team implemented numerous strategies to isolate Phase 2 in a noninvasive way. First, we used daily MC tracking, inclusive of urinary ovulation kits, for 11 wk before study commencement, which provided ~3 cycles of data. It was hypothesized that some variability within individuals would exist, yet this information would still be useful in predicting the timing of key MC phases during camp. However, as evident in Figures 2 and 3 , the within-athlete variability in bleed length, ovulation day, and overall cycle length was substantial. This meant that although most MC patterns could be identified by tracking, these data are limited in their ability to accurately identify specific MC phases. Our next strategy for attempting to identify phase 2 was to use dual hormone urinary ovulation kits, which claim to distinguish a rise in estrogen from a rise in LH. Although purported to identify the phase of the MC commensurate with high estrogen concentrations, the threshold of estrogen concentration for this method was unknown. Within our cohort, some athletes failed to show a rise in urinary estrogen before their ovulation day, whereas others had 12 d with elevated estrogen (see Fig. 2 ). Such findings highlight the absolute differences associated with the normal variability in peak estrogen concentrations between athletes ( 17 ). Menstrual cycle characteristics of naturally cycling athletes ( n = 8). Each row represents a complete menstrual cycle. Red squares represent bleeding days, yellow squares represent a rise in urinary estrogen (as claimed by the manufacture), and green squares represent ovulation. A “?” is placed where an ovulation kit was not used, but a rise in estrogen concentrations was expected based on prior/subsequent day measures. Gray bars represent the point at which tracking was ceased during the last cycle. n = 3 naturally cycling athletes were removed from this figure: two subjects collected <2 cycles worth of data (see Fig. 1 ) and 1 athlete had indissociable menstrual cycle characteristics due to recent HC use. Individual hormonal responses from four athletes defined as naturally menstruating, with figures made to recreate those published by Elliot-Sale et al. ( 6 ). Each dot represents the measured hormone concentrations, with the dotted lines representing the predicted cyclical nature of the hormones across the menstrual cycle between testing days. The yellow line is estradiol, green is progesterone, blue is FSH and purple is LH. Each red line represents ovulation measured using urinary ovulation kits and the red shaded area represents the bleeding period. Although we were able to consistently identify a phase of the athletes’ MC with elevated estrogen and low progesterone concentrations prior to ovulation, we were unable to confirm this as true Phase 2 accordingly to Elliott-Sale et al., ( 9 ). Although elevated, serum estrogen concentrations measured during this phase unlikely represented “peak” estrogen concentrations in many participants, as estrogen concentration was higher in Phase 4 compared with Phase 2 (Fig. 3 ). This suggests that we were either too early or late in our selected testing date for Phase 2, choosing a day when estrogen concentrations were still rising or declining. However, our philosophy was to sample earlier, rather than later, to avoid missing Phase 2 and instead capturing Phase 3, where a drop in estrogen and rise in LH occurs. Phase 2 (~14–26 h before ovulation); ( 9 ) represents a very small window in which research protocols can be implemented, and in practical terms, is very difficult to pinpoint in athletes with highly variable MCs. Thus, we question the relevance of this phase in elite sport, and whether it is as important to the health and performance of female athletes as other hormonal profiles/phases. Nevertheless, Phase 2 provides an interesting physiological model, as it is the only time throughout the MC where estrogen is high and progesterone is low. Interestingly, these two hormones can exert opposing actions: estrogen shows anabolic effects ( 18 ) and regulates substrate metabolism to favor fat oxidation pathways ( 19 ) whereas, progesterone has antiestrogenic effects ( 20 ). We note that the majority of SES literature has examined scenarios where both hormones are high (Phase 4) or low (Phase 1), potentially having opposed and nullifying actions. Therefore, testing in the late follicular phase (i.e., Phase 2) does provide a hormonal concentration differential that may shed light on the influence of ovarian hormones on different parameters and may be of increased importance in answering research questions with a hormone-driven hypothesis. Historically, research examining MC phase has treated data categorically (i.e., as discrete phases), which might be limiting due to the aforementioned difficulties in sampling phases correctly and the inter-individual variation. In some scenarios, two athletes could also have similar estrogen yet different progesterone concentrations, with different outcome measures depending on the dominate hormonal influence ( 17 ). Therefore, analyzing such data in a continuous manner using the estradiol/progesterone (E/P) ratio maybe more appropriate to capture the hormonal differential ( 17 ). This approach has been used successfully by others ( 19 ) and considers the synergistic effects of both hormones. Furthermore, it overcomes issues with not capturing “peak” estrogen in Phase 2, and allows ecologically valid study designs to be employed, where high levels of scientific control can be balanced with the practicalities of working with real-world athletes. The classification of MC function in our cohort after the retrospective hormonal analysis was another challenging aspect of the study. One of the required criteria to achieve eumenorrheic status is a “correct hormone profile” ( 9 ), which could be interpreted in many ways. For example, Athlete A-C (Fig. 3 ) appears to have regular MCs according to our precamp MC tracking metrics. However, due to sample collection timing during the camp, we found higher estrogen concentrations in Phase 4 compared with Phase 2; this represents an “incorrect” hormonal profile and would preclude classification as eumenorrheic. Importantly, if we had only measured two time points (e.g., Phase 1 and Phase 4), we would have classified these athletes as eumenorrheic. Similarly, the identification of luteal phase deficiency in one athlete from our cohort would have been missed if hormones were only sampled in Phase 1 and 2, and accordingly she would have been defined at eumenorrheic. Promoting a minimum number of samples per cycle, or repeated samples across two cycles may help improve clarity and classification of MC function. However, this requirement could introduce further barriers for female athlete research that may be unnecessary. Nevertheless, it is important to acknowledge and highlight the nuances and ambiguities in data interpretation and application into practice. Finally, the rationale for undertaking this study and the explanation for the provision of such detailed insight into protocol planning and implementation should be considered. This project was initiated to directly address the recent focus on the underrepresentation of female athletes in SES research ( 21 , 22 ). We note the particular lack of research in elite female athletes, high quality control or characterization of MC function and phase, and outcomes with practical and real-world application. It is easy to understand the preference for undertaking laboratory-based research in recreationally active individuals ( 7 ), as invasive research designs, large time commitments and interference with training and/or competition programs can be an obstacle to participation of high-performance athletes across sexes and genders. However, as we have summarized in Table 1 , there are numerous challenges, both for female athletes and the research team, in implementing worthwhile observational or interventional research. Our research group has had extensive experience with applied research with high-performance athletes, and in numerous scenarios have collaborated with sporting organizations to create projects with a “win-win” scenario, in which relevant research questions are answered while athletes gain direct benefits from a high-quality training environment ( 4 – 6 ). In developing FARC 1.0, we further evolved our training camp model with attention to insights gained from the planning and implementation of the project, as well as the outcomes. Table 1 was intentionally created to provide robust insights to other research groups, to enable others to follow in our footsteps, or to gain ideas and skills that might improve their confidence and willingness to embark on female athlete research projects of different types. We hope that sharing the considerations, intentions and actual outcomes from various aspects of this project with enhanced transparency and will increase the interest in female-focused research, noting the range of benefits to both researchers and participants beyond the traditional data outputs and publications.

Excellent resources outlining nomenclature and research methodologies for conducting SES research with female participants have recently been published ( 9 , 22 , 23 ). However, further work is needed to address the range of special challenges of studying female athletes, particularly high-performance athletes in real-world situations. This article shares insights into our experiences and recommendations for conducting research in the daily training environment of female athletes. We provide information on the MC characteristics of our cohort of highly trained First Nations female Rugby League players both to demonstrate the challenges of working in this area, as well as to provide a background to the series of individual data sets that were collected within the FARC 1.0 research-embedded training camp.