Examining the relationship of training load and menstrual cycle on a web application-based load and recovery score: A longitudinal observational study in elite women's soccer

Abstract

Monitoring training load and recovery is crucial in team sports. Despite the increasing prevalence of professional women's sports, most monitoring methods fail to account for the menstrual cycle (MC). Given the potential influence of the MC on sports performance, training responses, and recovery, incorporating it into monitoring practices is essential. This study assessed the relationship between training load and a newly developed web application-based Load and Recovery Score (LRS) among elite female soccer players. Furthermore, it explored the potential influence of the MC on the individual training responses. Forty-nine female elite soccer players were recruited for this 6-week observational study. The LRS was recorded daily using a web application, and a subgroup of 34 players recorded their basal temperature every morning, using a digital thermometer. Training load was assessed based on trainer-intended ratings of perceived exertion. A significant negative relationship was found between training load and players’ LRS (Est. = −0.009, 95% CI −0.011, −0.007, p < .001). When controlling for training load, no significant effect of the MC on players’ LRS was found (F = 1.274, p = .283). The explained variance of both models was 46.59% and 50.07%, respectively, with a high proportion of variance attributed to random effects (43.67%, and 47.43% respectively). The LRS represents a multifactorial tool that depicts training load in terms of Trainer Session Rating of Perceived Exertion, helping coaches identify athletes’ responses to training stimuli. While the MC did not show a significant effect on the LRS, systematic monitoring should still be considered, as its irregularity may indicate serious health problems.

Keywords

Athlete monitoring heart rate variability hormonal contraception luteal phase ovulation rating of perceived exertion

Introduction

In recent years, the monitoring of training load and recovery in team sports has become increasingly important.¹ Systematic monitoring of players’ load and recovery status provides valuable insights for coaches and medical staff, enabling adjustments to be made to the training program, thereby reducing the risk of injury, nonfunctional overreaching, and overtraining.^1,2 The interindividual differences in recovery potential, load tolerance, and exposure to various stressors, alongside the training stimulus, result in varying vulnerabilities among players, even when exposed to identical training conditions. This underscores the importance of individualized monitoring.²

Selecting an appropriate method for monitoring load and recovery presents a challenge. Several commonly employed methods have faced criticism for being impractical (e.g., maximal strength testing), expensive (e.g., biochemical/hematological assessments), or for assessing only single aspects of recovery and load (e.g., accelerometry).¹ Furthermore, the practical approach of self-assessment bears the risk of biased responses, influenced by fear of repercussions or lack of self-awareness.³ As no gold standard has yet been established,³ the expert consensus emphasizes the importance of a multivariate approach that amalgamates different subjective and objective load and recovery parameters.⁴

Despite the increasing prevalence of professional women's sports, most load and recovery monitoring methods do not account for the menstrual cycle (MC).⁵ This is surprising, considering that current literature highlights the potential influence of the MC on sports performance, individual training responses, and recovery.^6–9 In naturally menstruating females, fluctuations in endogenous sex hormone levels, especially estrogens and progesterone, likely drive these effects,⁵ exerting distinct influences on energy metabolism. Estrogens appear to impact the breakdown of energy sources by enhancing the rate of carbohydrate breakdown,¹⁰ and increasing the use of glycogen stores,¹¹ ultimately supporting endurance performance.¹² Conversely, progesterone, acting as an antagonist to estrogen, inhibits carbohydrate breakdown, resulting in increased breakdown of proteins and higher rates of amino acid use.¹³ This becomes evident during the luteal phase, characterized by elevated progesterone levels, where an increase in the preservation of muscle glycogen and a greater reliance on fat metabolism has been reported.^11,14 These metabolic changes may influence an individual's readiness for training and their overall responses to training. Specifically, approximately 50% of eumenorrheic athletes who do not use hormonal contraceptives may encounter significant physiological and psychological alterations during their MC.¹⁵ Therefore, conducting more comprehensive research on this topic and incorporating continuous monitoring of the MC into regular training practices appears imperative to facilitate a deeper comprehension of its potential impact on recovery and performance.¹⁵

As part of a collaborative project involving the University of Bern (Institute of Sport Science) and the company Commbuddy, a web application called “hieros” was developed to monitor the load and recovery of athletes in team sports. With this web application, players can track their physical performance capability, overall recovery, muscular stress, fatigue, mood, and sleep quality via a daily digital query. Additionally, heart rate variability (HRV), and acute: chronic workload ratio (ACWR) is determined. The results from the daily assessment are then aggregated into an overall load and recovery score (LRS). The LRS and its subscales may serve as practical indicators, providing coaches and medical staff with insight into the individual stress and recovery status of their players.

The relationship between the LRS and training load has not been studied. Thus, the current study aims to: (1) examine the magnitude and direction of the relationship between training load and the subsequent day's expression of LRS in elite women's soccer. Additionally, considering the purported impact of MC-related changes on training response and recovery, the study aims to: (2) ascertain whether the MC influences players’ LRS beyond the impact of training load. We hypothesize that there is a significant negative relationship between training load and players’ LRS, and that the menstrual cycle will have a significant effect on the LRS.

Method

Subjects

Forty-nine female players from the “AXA Women's Super League” team (AWSL team) and the U19 junior team of a swiss soccer team who met the inclusion criteria consented to participate in the study. The inclusion criteria were as follows: (1) being an active member of the AWSL team or the U19 junior team, (2) being at least 16 years old, (3) having no illness or injury that would prevent participation in training, and (4) not being pregnant. An onboarding session was conducted to inform participants and obtain written informed consent. For players under the age of 18, written consent of the legal guardian was obtained. We included a convenient sample of 49 players (age: 19 ± 2 years). Eight players were excluded from the final analysis due to missing LRS data or missed training sessions. For the second study aim, which investigated the effect of the MC on LRS, only players who were not currently using hormonal contraception (n = 35) were included. For the final analysis, 10 additional players were excluded because their MC phases could not be determined, leaving a final sample of 25 players included in the model (see flowchart, Figure 1).

The present study was approved by the Ethics Committee of the Faculty of Humanities and Social Sciences of the University of Bern (Nr. 2022-06-00008).

Study design

The present study was conducted as a longitudinal observational study with prospective data collection over a six-week period in July and August 2022, spanning the pre-season and the beginning of the competition phase. The variable “menstrual cycle” was recorded in the subsample (n = 35) every morning under standardized conditions, starting in week one (t = 42 time points), by tracking oral basal temperature and menstrual bleeding using the Sensiplan cycle sheet. This ensured that a complete MC could be recorded. Furthermore, HRV and ACWR were monitored to establish baseline values. Starting from week two, all other LRS subscales and the variable “training load” were assessed daily (t = 35 time points). To provide natural training and competition conditions, coaches made no decisions based on the collected LRS data.

Demographic data

Demographic and MC characteristics (regularity, bleeding phases, use of contraception, type of contraception), as well as employment status, were collected prior to the observation period through the online survey application LimeSurvey (LimeSurvey GmbH, Hamburg, Germany).

Load and recovery score

The players utilized the web application (hieros) to record daily load and recovery values across eight distinct subscales assessing the players’ physical performance capability, overall recovery, muscular stress, fatigue, mood, and sleep quality. Additionally, HRV, and ACWR¹⁶ were obtained.

Physical performance capability, overall recovery, and muscular stress were assessed by having players rate the extent to which a set of adjectives related to each aspect applied to them at the moment. Players provided responses on a scale ranging from 0 (does not apply at all) to 6 (fully applies) for each of these three subscales.² Fatigue was evaluated using the question “Do you feel exhausted?” (translated from German). Players responded to this query on a Likert scale ranging from 0 (Fully applies) to 6 (Does not apply at all). Mood was assessed through the question “How do you feel today?” (translated from German). Players provided their responses on a Likert scale ranging from 0 (Sad) to 6 (Happy). Sleep quality was determined by asking players to rate the quality of their sleep from the previous night using a scale ranging from 0 (Very bad) to 6 (Very good).

Upon waking, the players who voluntarily performed daily HRV measurements (n = 5) recorded their resting HRV using a heart rate monitor and a chest strap (Polar H10, Polar Electro OY, Kempele, Finland) in conjunction with the Elite-HRV application (Elite HRV Inc, Asheville, United States).¹⁷ In the first week of the observation period, players tracked their root mean square of successive RR interval differences (RMSSD) values on a daily basis for one week and entered it into the application to establish baseline values. During the following weeks, any deviation of the RMSSD values from this baseline by more than 5% over three consecutive days resulted in a reduction of 1 point in the HRV-Subscale-Score. If the values remained below the baseline for two additional days, the score was further reduced by 1 point. Upon the RMSSD returning to baseline and remaining within the baseline range for two consecutive days, the subscale score increased by 1 point. If the RMSSD stayed within the baseline value for 3 consecutive days, the score returned to 6.

To calculate the ACWR, the training intensity was determined after each session (using a scale from 0 to 10) and multiplied by the session's duration (in minutes). The total value for a week was then compared with the average of the last 1–4 weeks. Deviations greater than 1.5 resulted in a reduction of one point in the ACWR-Subscale-Score, while deviations greater than 1.8 resulted in a reduction of two points in the score. If the deviation is less than 1.5, the score returned to its baseline value of 6. Subsequently, the daily subscale scores were combined to generate the LRS (ranging from 0 to 120), with a higher LRS score indicating better recovery.

Training load

The Trainer Session Rating of Perceived Exertion (Trainer-SRPE) was selected as the method for assessing training load and match load. The trainer-SRPE has been established as a non-invasive, accessible, valid, and reliable method to monitor training load.¹⁸ In the present study coaches determined the trainer-SRPE before each training session for the players. This rating is calculated by multiplying the coaches’ assessment of intended training intensity (using a scale from 0 to 10) by the session's duration (in minutes).

Determination of the menstrual cycle phase

Those players not using hormonal contraception (n = 35) recorded their oral basal temperature every morning for six weeks, using a digital thermometer (Braun digital thermometer PRT 1000, Germany), and logged the temperature, date, and menstrual bleeding on the Sensiplan cycle sheet. The Sensiplan is a validated tool for identifying ovulation.¹⁹ A combination of different methods were employed to determine the MC phases. The ovulation phase was identified by interpreting the basal temperature curve using the procedure outlined by Holt and Döring.¹⁹ The luteal phase was determined by counting back 14 days from the first day of bleeding and aligning it with the temperature curve.¹⁶ The follicular phase, including the menstrual phase (characterized by menstrual bleeding), starts on the first day of bleeding and ends on the last day before the ovulation phase.¹⁶ Identification of the menstrual phase was conducted by a senior physician, a resident in gynecology, and a midwife. If there was no consensus regarding the menstrual phase, players were excluded from the final analysis, leaving a final sample of 25.

Statistical analysis

The data analysis was conducted using the open-source statistical software R (version 4.2.3., R Core Team, 2022, Vienna) and the integrated development environment RStudio. The effects and model variances of training load and MC on LRS were tested using the functions lmer() and rsq() from the packages “lme4”, “lmerTest”, and “rsq” and visualized with plot() and the function ggpredict() from the package “ggeffects”. Descriptive statistics were calculated for each demographic and characteristic variable, including means, standard deviations (SD), and relative frequencies. The normal distribution was visually assessed using Quantile-Quantile plots.²⁰ A priori cut-off missing data were set at 15% (=5 time points) for the LRS.¹⁶ Expert opinions were relied on for cycle-specific analysis. Linear mixed models were created for the inferential analysis, which provided information on the strength and direction of the relationship between trainer-SRPE and the LRS of the following day.²¹ The significance level was set at p < .05 for all model calculations and model assumptions were visually checked.²⁰ Linear mixed models were chosen to account for the correlated data associated with the repeated measurement design.^21,22 Unlike the trainer-SRPE, visual assessment of the distribution of the LRS did not reveal any deviations from the normal distribution, justifying the use of the linear mixed models.²¹

The training load model incorporated the trainer-SRPE as a fixed effect and the individual player response of all participants, irrespective of their contraceptive methods (n = 41), as a random effect (random intercept for player). A negative correlation between training load and LRS was assumed in the analysis. Estimates, their corresponding 95% confidence intervals (95% CI), and the model's variance components (R2) were calculated using the Restricted Maximum Likelihood method. To investigate the effect of the MC on LRS, controlled for training load, a within-subject design with covariate was chosen.²³ The model included the MC (a factor with four levels: menstrual phase, follicular phase, ovulation phase, and luteal phase) of the players (who were not currently using hormonal contraception and for whom the cycle phases could not be determined (n = 25)) and the trainer-SRPE as fixed effects, and a random intercept for Player. A Type III analysis of variance was performed using Satterthwaite's method to test the main effects of the menstrual cycle and training load on LRS of the following day, and the corresponding metrics were recorded (sum of squares, p-value, and variance components of the model).

Results

Preliminary analysis

The demographic characteristics of the players are summarized in Table 1. On average, the players had five training sessions per week, and the majority of them were employed at a moderate to high level. All but two players (who applied hormonal contraceptives) from the total cohort had regular occurrences of menstrual bleeding. Descriptive statistics for individual variables are presented in Table 2. The LRS overall mean of 1382 observations during the study was 88.6 ± 10.38 points, indicating an average recovery state of 73% out of 100%. The LRS had negligible missing values (1.6%–6.9%) depending on the analysis. Model assumptions (linearity, homoscedasticity, and normality of residuals) were visually inspected and found to be satisfactory.

Table 1.

Players’ demographic characteristics.

Characteristic	N = 41^a
Team
AWSL	20 / 41 (49%)
U19	21 / 41 (51%)
Age (years)	20 (2)
Height (cm)	166 (5)
Body mass (kg)	60 (5)
Training Sessions per Week	5 (0.9)
Team Training Sessions per Week	4 (0.5)
Regular Cycle^b	34 / 41 (83%)
Menstrual Bleeding	39 / 41 (95%)
Currently using Hormonal Contraception
None	35 / 41 (85%)
Hormonal Contraception	6 / 41 (15%)
Type of Hormonal Contraception
Intrauterine Device	1 / 6 (17%)
Pill	5 / 6 (83%)
Employment Degree
Full-Time	19 / 41 (46%)
Part-Time	19 / 41 (46%)
Unemployed	3 / 41 (7%)

n / N (%); Mean (SD).

The regularity of menstrual cycle in terms of bleeding frequency, intensity, and duration was based on the players’ perception and not measured. Abreviation: AWSL: “AXA Women's Super League” team.

Table 2.

Descriptive statistics of the variables.

Variable	n	Obs.	Mean (SD) of LRS	Mean (SD) of Trainer-SRPE	NA of LRS (%)	NA of Phase (%)
MC^a	25	861	89 (9.90)	525.08 (233.18)	3.5	1.6
Menstrual phase	25	131	89.85 (10.38)	553.98 (259.27)	1.5	-
Follicular phase	25	329	88.90 (9.73)	526.09 (223.64)	4.3	-
Luteal phase	25	371	88.83 (10.11)	509.03 (231.85)	3	-
Ovulatory phase	25	30	88.07 (8.32)	542.75 (220.42)	3.3	-
Variable	n	Obs.	Mean of Variable	Mean (SD) of LRS	SD	Min.	Max.	NA of LRS (%)
LRS	41	1382	88.6		10.38	54	118	3.7
Trainer-SRPE	41	847	527.57	88.60 (10.38)	232.16	90	900	3.2
AWSL	20	655	490.43	87.2 (10.7)	248.4	90	900	6.9
U19	21	727	568.25	89.8 (9.94)	217.63	90	900	1.1

MC = Menstrual cycle (Factor with 4 Levels).

Abbreviations: AWSL: “AXA Women's Super League” team; LRS: Load and recovery score; NA: Not Available; n: Number of players; Obs.: Observations; SD: Standard Deviation; trainer-SRPE: trainer-Session Rating of Perceived Exertion.

Relationship between training load and LRS

The analysis of the relationship between the trainer-SRPE and the LRS included 847 observations from 41 participants, excluding training free days. Eight players had to be excluded from the analysis due to a significant amount of missing LRS data. A significant negative relationship was found between trainer-SRPE and LRS (Est. = −0.009, 95% CI −0.011, −0.007, p < .001) (Figure 2). Table 3 shows that the expected value of LRS at trainer-SRPE equal to zero was 92.23 points (95% CI 89.693, 94.772). Therefore, an increase in trainer-SRPE from 0 to 1 was associated with a decrease in LRS by −0.009 points. A game scored with a trainer-SRPE of 900 leads to an 8-point decrease in the LRS.

Table 3.

Relationship between training load and the load and recovery score (LRS).

	Trainer-SRPE Model
	Estimate	Std. Error	t value	Pr(>\|t\|)	95% CI
Intercept^a	92.231	1.29	71.507	<.001	89.693	94.772
Trainer-SRPE	−0.009	0.001	−7.645	<.001	−0.011	−0.007
Variance analysis	%
R² Model	46.59
R² Fixed	3
R² Random	43.67

Intercept: Mean value of LRS at training load zero.

Abbreviations: CI: Confidence Interval; R²: R-Squared; SRPE: Session Rating of Perceived Exertion: Std.Error, Standard Error.

Effect of the MC on the LRS

Next, we examined whether the MC, controlled for trainer-SRPE, affected the players’ LRS of the following day. A total of 861 cycle phase observations were collected from 25 non-hormonal contraceptive players, 329 of which were measured in the follicular phase, 371 in the luteal phase, 131 in the menstrual phase, and 30 in the ovulation phase (missing 1.6% time points). Table 4 presents the results of the MC model. Testing the main effect showed no significant effect of MC on the following day's LRS (Sum Sq = 193.98, F = 1.274, P = .283) (Figure 3). The model explained 50.07% of the variance, with 47.43% explained by the random effects. Post hoc fixed effects analysis revealed that the LRS of a player in the menstrual phase and training load equal to zero was on average 1.8 points higher than the LRS of a player in the follicular phase, although this effect was not significant (Est. = 1.865, 95% CI −0.086, 3.813, P = .062).

Table 4.

Effect of the menstrual cycle on the load and recovery score (LRS) of the following day.

Testing Main Effect
	Sum Sq	Mean Sq	F value	Pr(>F)	R² Model	R² Fixed	R² Random
MC	193.98	64.66	1.274	0.283	50.07%	2.64%	47.43%
Testing Fixed Effects
	Estimate	Std. Error	t value	Pr(>\|t\|)	95% CI
Intercept^a	91.902	1.671	55.011	<.001	88.612	95.197
Luteal phase	0.351	0.767	0.458	.647	−1.148	1.85
Menstrual phase	1.865	0.997	1.872	.062	−0.086	3.813
Ovulatory phase	−0.027	1.705	−0.016	.988	−3.364	3.305

Intercept: mean value of LRS of someone in follicular phase controlled for training load.

Abbreviations: MC: Menstrual cycle; Mean Sq: Mean sum of squares; Sum Sq: sum of squares.

Discussion

The aim of the present study was to examine the relationship between training load assessed by trainer-SRPE and the LRS of the following day. Furthermore, the study investigated whether the MC modulated the LRS independently of the training load. The results indicated (1) a significant relationship between trainer-SRPE and the LRS of the following day, and (2) no effect of the MC on the LRS of the following day beyond the impact of training load.

The observation that a high trainer-SRPE resulted in a decrease in the LRS on the following day supports the assumption that the LRS is sensitive enough to sufficiently reflect the impact of a previous training session. Therefore, the LRS can provide coaches with valuable insights into whether players respond appropriately to the intended training stimuli. While the trainer-SRPE has been established as a valid and reliable parameter for quantifying training load,^18,24 it is important to note that this one-dimensional parameter may not adequately capture individual responses across the various dimensions of recovery. This is where the primary advantage of the LRS comes into play, with its subscales designed to offer additional insights into both training- and non-training-related influences. This can assist coaches and medical staff in more accurately detecting and comprehending individual training responses on a multidimensional level.

The results of the analyses of variance showed that individual differences between players in LRS responses explained a large proportion of the model variance, which is comprehensible considering the multidimensional construct of load and recovery. This aligns with the current literature and underscores the importance of monitoring the load and recovery of team athletes on an individual level.²⁵

The present study did not reveal any discernible impact of the MC on the players’ LRS beyond the influence of training load. It is noteworthy that the percentage of players not using hormonal contraception in this study was notably higher than in other studies.²⁶ Considering the high prevalence of menstrual-related symptoms in those not using hormonal contraceptives, and in light of recent research showing an influence on various LRS subscales (e.g., sleep quality, mood, fatigue, perceived exertion, and muscle soreness),²⁷ it was hypothesized that an effect on the players’ LRS could be evident. However, it appears that the LRS and its subscales are not sensitive enough to adequately represent the MC. On the other hand, the results may indicate that the MC doesn't have a strong effect on recovery, or only have a very individual-level impact. This is supported by literature reporting that the effects of the MC on recovery and performance remain inconclusive.^27–29

A crucial aspect to consider when interpreting the results is the complexity of detecting MC phases in these relatively young athletes³⁰ and the potential limitations of the MC monitoring method used in this study.³¹ Despite enlisting specialists to interpret the players’ MC phases, we had to exclude ten players from the analysis due to challenges in accurately determining their menstrual phases. In this regard, the more reliable three-step verification method, including serum/plasma hormone analysis,^32,33 should have been applied. This approach would also have reduced the risk of possible misclassification of MC phases, which may have introduced bias in the results obtained in this study. However, this method is less feasible for a long-term study due to ethical concerns, as well as the time, cost, and overall burden on participants, particularly for daily monitoring. Nonetheless, it would be worth considering validating the present study's results using this method.

An interesting observation was that, regardless of training intensity, the average LRS values were consistently higher in the menstrual phase compared to the other phases. Current literature indicates that physical performance, fatigue, and sleep quality are more significantly affected during the menstrual phase than during other MC phases.^27,34 These results may indicate an enhanced recovery during the menstrual phase, as a previous strength training study reported.³⁵

While the current study did not establish a direct impact of the MC on the LRS, consistent monitoring of the MC in athletes is strongly advised as it serves as a significant health indicator among eumenorrheic players.³⁶ The ‘female athlete triad’ represents a spectrum of interconnected conditions, encompassing low energy availability, reduced bone mineral density, and menstrual irregularities.³⁶ These factors collectively contribute to an increased susceptibility to injuries.^30,36 Monitoring the MC can help identify irregularities or absence of bleeding, helping to prevent conditions like the ‘female athlete triad’.³⁶

Regarding the web application, exploring the explicit inclusion of the MC in the LRS should be considered, as it is currently not represented. Integrating the MC into daily monitoring could further contribute to fostering players’ awareness, openness, and comprehension of the MC within the athletic environment. Additionally, it could assist coaches in offering personalized training recommendations and adjustments for individual athletes. However, accurate monitoring remains an issue due to the present financial and time burden. Nevertheless, given the rapidly evolving landscape of biomedical technologies, it is expected that cost-efficient and feasible wearable devices will be made available in the near future, which can be utilized.

Limitations

Some limitations should be considered when generalizing the results of this study. Firstly, the results of this convenience sample of high-level athletes may not be unreservedly applicable to other demographics due to differences in age, performance level, and sport. Secondly, the absence of a gold standard for determining training load in team sports may have led to the inclusion of a suboptimal predictor in the linear mixed models, potentially introducing bias into the results and their generalizability. While the trainer-SRPE stands as an established and reliable load parameter, particularly among experienced coaches, it remains inherently subjective. Hence, additional research aimed at evaluating the impact of other, more objective load parameters on the LRS is warranted. Thirdly, the study lasted only 6 weeks. Although this ensured that each participant completed a full cycle, extending the duration of the study and monitoring over multiple cycle phases would have provided more comprehensive and meaningful data.

Fourthly, MC determination relied solely on calendar-based counting and basal temperature, which may not accurately reflect the MC of each participant. Even though basal temperature assessments have been shown to be suitable for retrospectively confirming ovulation and even comparable to urinary ovulation kit testing^37,38 other studies have questioned the accuracy of this method³¹ and recommend a three-step verification approach, including serum/plasma hormone analysis.^32,33 Finally, even though the players included in the analysis did not report any MC irregularities, it would have been advisable to include regular MCs as an inclusion criterion. This ensures more accurate phase identification, minimizes hormonal variability, and reduces confounding factors, allowing for more reliable assessments of the MCs impact on sports performance and recovery.^31,38

Conclusion

The LRS appears to be a practical tool for monitoring load and recovery status in women's soccer. Its ability to depict training load at an individual and multifactorial level can assist coaches in recognizing players’ responses to training stimuli. Detecting and understanding fluctuations in player's LRS allows coaches to tailor training on an individual basis, thereby reducing the risk of injuries, non-functional overreaching, and overtraining. However, further studies that determine the relationship between the LRS and other load and recovery parameters are highly warranted. Even though previous research indicates that the MC can impact players’ recovery and performance, this study found no significant effect of the MC on LRS beyond the influence of training load. Therefore, the sensitivity of the LRS to MC phases and the general impact of the MC on recovery processes must be questioned.

Footnotes

Acknowledgments

We would like to thank the players and coaches for their cooperation during the whole study. Additionally, special thanks are due to Dr Nouk Schori and Dr Rahel Schmid (Department of Gynecology and Obstetrics at Regional Hospital Emmental) for their invaluable professional support during the data evaluation process.

Consent to participate

All participants involved in the study provided written informed consent.

Data availability

The data presented in this study are available upon request from the corresponding author.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical considerations

The present study was approved by the Ethics Committee of the Faculty of Humanities and Social Sciences of the University of Bern (Nr. 2022-06-00008).

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Jan Anderegg

Claudio R Nigg

Sascha Ketelhut

References

Bourdon

Cardinale

Murray

, et al. Monitoring athlete training loads: consensus statement. Int J Sports Physiol Perform 2017; 12: 161–170.

Hitzschke

Kölling

Ferrauti

, et al. Entwicklung der kurzskala zur erfassung von erholung und beanspruchung im sport (KEB). Zeitschrift fur Sport 2015; 22: 146–161.

Soligard

Schwellnus

Alonso

, et al. How much is too much? (part 1) international Olympic committee consensus statement on load in sport and risk of injury. Br J Sports Med 2016; 50: 1030–1041.

Kellmann

Bertollo

Bosquet

, et al. Recovery and performance in sport: consensus statement. Int J Sports Physiol Perform 2018; 13: 240–245.

McNulty

Elliott-Sale

Dolan

, et al. The effects of menstrual cycle phase on exercise performance in eumenorrheic women: a systematic review and meta-analysis. Sport Med 2020; 50: 1813–1827.

Meignié

Duclos

Carling

, et al. The effects of menstrual cycle phase on elite athlete performance: A critical and systematic review. Front Physiol 12. Epub ahead of print 2021. DOI: https://doi.org/10.3389/fphys.2021.654585

Balachandar

Marciniak

Wall

, et al. Effects of the menstrual cycle on lower-limb biomechanics, neuromuscular control, and anterior cruciate ligament injury risk: a systematic review. Muscles Ligaments Tendons J 2017; 7: 136–146.

Hottenrott

Möhle

Ide

, et al. Recovery from different high-intensity interval training protocols: comparing well-trained women and men. Sports 2021; 9: 1–12.

Hottenrott

Ketelhut

Schneider

, et al. Age- and sex-related differences in recovery from high-intensity and endurance exercise: a brief review. Int J Sports Physiol Perform 2021; 16: 752–762.

10.

Zderic

Coggan

Ruby

. Glucose kinetics and substrate oxidation during exercise in the follicular and luteal phases. J Appl Physiol 2001; 90: 447–453.

11.

Devries

. Sex-based differences in endurance exercise muscle metabolism: impact on exercise and nutritional strategies to optimize health and performance in women. Exp Physiol 2016; 101: 243–249.

12.

Bernstein

Behringer

. Mechanisms underlying menstrual cycle effects on exercise performance: a scoping review. Women Sport Phys Act J 2023; 31: 1–18.

13.

Boisseau

Isacco

. Substrate metabolism during exercise: sexual dimorphism and women’s specificities. Eur J Sport Sci 2022; 22: 672–683.

14.

Willett

Koltun

Hackney

. Influence of menstrual cycle estradiol-β-17 fluctuations on energy substrate utilization-oxidation during aerobic, endurance exercise. Int J Environ Res Public Health 2021; 18: 10–15.

15.

Findlay

MacRae

EHR

Whyte

, et al. How the menstrual cycle and menstruation affect sporting performance: experiences and perceptions of elite female rugby players. Br J Sports Med 2020; 54: 1108–1113.

16.

de Jonge

Thompson

Ahreum

HAN

. Methodological recommendations for menstrual cycle research in sports and exercise. Med Sci Sports Exerc 2019; 51: 2610–2617.

17.

Perrotta

Jeklin

Hives

, et al. Validity of the elite HRV smartphone application for examining heart rate variability in a field-based setting. J Strength Cond Res 2017; 31: 2296–2302.

18.

Inoue

dos Santos Bunn

do Carmo

, et al. Internal training load perceived by athletes and planned by coaches: A systematic review and meta-analysis. Sport Med - Open 8. Epub ahead of print 2022. DOI: https://doi.org/10.1186/s40798-022-00420-3

19.

Frank-Herrmann

Baur

Freundl

, et al. Natürliche Familienplanung – aktueller stand. Gynakologe 2015; 48: 657–666.

20.

Portney

. Foundations of clinical research applications to evidence-based practice. 4th ed. Philadelphia: F.A. Davis Company, 2020.

21.

Liu

Rovine

Molenaar

PCM

. Selecting a linear mixed model for longitudinal data: repeated measures analysis of variance, covariance pattern model, and growth curve approaches. Psychol Methods 2012; 17: 15–30.

22.

Iannaccone

Conte

Cortis

, et al. Usefulness of linear mixed-effects models to assess the relationship between objective and subjective internal load in team sports. Int J Environ Res Public Health 2021; 18: 1–11.

23.

Nguyen

Şentürk

Carroll

. Covariate-adjusted linear mixed effects model with an application to longitudinal data. J Nonparametr Stat 2008; 20: 459–481.

24.

Paul

Read

Farooq

, et al. Factors influencing the association between coach and athlete rating of exertion: a systematic review and meta-analysis. Sport Med - Open 2021; 7: 1–20.

25.

Wiig

Andersen

Luteberget

, et al. Individual response to external training load in elite football players. Int J Sports Physiol Perform 2020; 15: 696–704.

26.

Parker

Elliott-Sale

Hannon

, et al. An audit of hormonal contraceptive use in women’s super league soccer players; implications on symptomology. Sci Med Footb 2022; 6: 153–158.

27.

Carmichael

Thomson

Moran

, et al. The impact of menstrual cycle phase on athletes’ performance: a narrative review. Int J Environ Res Public Health 2021; 18: 1–24.

28.

Kubica

Ketelhut

Querciagrossa

, , et al. Effects of a training intervention tailored to the menstrual cycle on endurance performance and hemodynamics. J Sport Med Phys Fitness. Epub ahead of print 2023. DOI: doi.org/10.23736/s0022-4707.23.15277-7

29.

Kubica

Ketelhut

Nigg

. Polarized running training adapted to versus contrary to the menstrual cycle phases has similar effects on endurance performance and cardiovascular parameters. Eur J Appl Physiol. Epub ahead of print 2024. DOI: https://doi.org/10.1007/s00421-024-05545-9

30.

Temm

Standing

Best

. Training, wellbeing and recovery load monitoring in female youth athletes. Int J Environ Res Public Health 19. Epub ahead of print 2022. DOI: https://doi.org/10.3390/ijerph191811463

31.

Elliott-Sale

Minahan

de Jonge

XAKJ

, et al. Methodological considerations for studies in sport and exercise science with women as participants: a working guide for standards of practice for research on women. Sport Med 2021; 51: 843–861.

32.

Schaumberg

Jenkins

Janse de Jonge

XAK

, et al. Three-step method for menstrual and oral contraceptive cycle verification. J Sci Med Sport 2017; 20: 965–969.

33.

Johnson

Marriott

Zinaman

. Can apps and calendar methods predict ovulation with accuracy? Curr Med Res Opin 2018; 34: 1587–1594.

34.

Solli

Sandbakk

Noordhof

, et al. Changes in self-reported physical fitness, performance, and side effects across the phases of the menstrual cycle among competitive endurance athletes. Int J Sports Physiol Perform 2020; 15: 1324–1333.

35.

Statham

. Understanding the effects of the menstrual cycle on training and performance in elite athletes: a preliminary study. Prog Brain Res 2020; 252: 25–58. DOI: https://doi.org/10.1016/bs.pbr.2020.05.028

36.

Mountjoy

Sundgot-Borgen

Burke

, et al. IOC Consensus statement on relative energy deficiency in sport (RED-S): 2018 update. Br J Sports Med 2018; 52: 687–697.

37.

Martinez

van Hooff

MHA

Schoute

, et al. The reliability, acceptability and applications of basal body temperature (BBT) records in the diagnosis and treatment of infertility. Eur J Obstet Gynecol Reprod Biol 1992; 47: 121–127.

38.

Schmalenberger

Tauseef

Barone

, et al. How to study the menstrual cycle: Practical tools and recommendations. Psychoneuroendocrinology 123. Epub ahead of print 2021. DOI: https://doi.org/10.1016/j.psyneuen.2020.104895