Load and recovery monitoring in Swiss top-level youth soccer players: Exploring the associations of a new web application-based score with recognised load measures

Introduction Systematic assessment of load and recovery in athletes is essential for effectively adjusting various training demands and their corresponding recovery measures (Kellmann et al., 2018), thereby reducing the risk of nonfunctional overreaching, overtraining, and potential subsequent injuries and illnesses (Bourdon et al., 2017; Kellmann et al., 2018; Taylor et al., 2012). The information obtained from the assessment can support athletes, coaching staff, and their medical teams in the tightrope act between performance optimisation and injury risk reduction. The expert consensus in the field of load and recovery monitoring and other research emphasises the importance of employing a multivariate approach for assessing load and recovery (Bourdon et al., 2017; Kellmann et al., 2018). Various physiological and psychological measures should be used for this purpose (Heidari et al., 2019). In team sports, it is also required that these assessments be carried out quickly, non-invasively, and with minimal added burden on the athletes (Thorpe et al., 2017). In this research project, we developed a web application-based Load and Recovery Score (LRS) and evaluated its relationship with established load parameters. It is assumed that specific training and match load variables correlate negatively with the following day’s LRS when controlled for intra-subject variability. Methods 78 female and male athletes from the U18, U19 and U21 teams of the Swiss soccer club “BSC Young Boys” were selectively recruited. 71 players (32.4% female) with an average age of 17.9 years (SD = 1.2) were monitored over a minimum period of 35 days. A repeated-measure design by means of a five-to-seven-week prospective longitudinal data collection was used in this study. The dependent variable (LRS) and four other independent load variables were repeatedly measured over time in the same athletes. The LRS comprises eight subscales integrated into an interval-scaled score ranging from 0 to 120. A higher score indicates a better recovery state and lower loads. The players recorded values for these eight different subscales daily using the web application. The subscales include questions drawn from various previously validated questionnaires related to the player’s 1) Physical capability, 2) General state of regeneration, 3) Muscular stress, 4) Fatigue, 5) Mood, and 6) Sleep quality, contributing to the recovery component of the score. Additionally, there are two load subscales pertaining to the player’s 7) Heart Rate Variability (HRV) and their 8) Acute:Chronic Workload Ratio (ACWR). The entries are either directly recorded on an ordinal scale (0-6) or are converted to conform to this scale level. Daily logs are incorporated into the different subscale values using a specific algorithm. The algorithm is informed by current research recommendations and is a proprietary business secret. The independent variables included the subjective Player- and Trainer – Session Rating of Perceived Exertion (PSRPE/TSRPE), as well as two GPS and accelerometry-based parameters: Total distance covered (TD) and Total distance > 20km/h (TD20). To examine direction and strength of the relationship between the LRS and the above-mentioned measures of training and match load, various linear mixed-effects models (LMM) were fitted via restricted maximum likelihood (REML). Random intercepts were defined for each player to account for the repeated within-subject measurements (Fisher et al., 2018; Molenaar & Campbell, 2009; Neumann et al., 2021), and the demographic control variables Height, Body mass and Sex were included in the models. Furthermore, the variance explained by the random effects was calculated using Nakagawa’s marginal and conditional R2 for mixed models. Results All training and match load parameters demonstrated significant negative correlations with the subsequent day’s LRS. In the linear mixed-effects model analysis PSRPE and TSRPE showed similar fixed effects (-0.013, 95% CI [-0.017, -0.010], p < .001 versus -0.008, 95% CI [-0.011, -0.006], p < .001), while TD exhibited stronger associations (-0.668, 95% CI [-0.979, -0.355], p < .001) than TD20 (-0.009, 95% CI [-0.012, -0.006], p < .001). The addition of control variables did not significantly influence direction or magnitude of the model’s effects. Variance explained by the residual factor ID (defining each individual) was high (≥ 0.444) in all of the analyses and post-hoc analyses on the influence of the variables Playing position and Sex showed high variation between these subgroups. Discussion/Conclusion The results show that the LRS has significant negative associations when controlled for repeated within-subject measurements with different subjective and objective training and match load measures, such as the PSRPE, the TSRPE, TD, and TD20. Therefore, it can track the effect of those variables whilst also being an indicator of different recovery parameters. All training and match load variables behave according to the a priori assumption and correlate negatively with the following day’s LRS. This is in line with the available literature, where it has already been shown that certain parameters, which are also part of the score, show good moderate to strong evidence for associations with different load indicators. The fact that the variance explained by the residual factor ID and the influence of grouping variables (Playing position/Sex) was high in all the analyses is consistent with current research (Hader et al., 2019; Neumann et al., 2021), where the impact of the different load parameters on recovery varied across groups and individuals. No single marker can provide global information (Temm et al., 2022) regarding an athlete’s recovery. The comprehensive LRS offers a solution to that problem because it can track different load parameters in elite youth soccer players and present multiple accepted recovery and load measures separately and on an individual level so that athletes, coaches and staff can use it to enhance their knowledge of responses (Bourdon et al., 2017) and determine future training and match load as well as suited means of recovery. By doing this, injury risk could be reduced and performance optimised. The ultimate decision of which monitoring tools to work with should remain with the sports professionals. It is essential that the protocol has reasonable practicability and uses an individualised (Temm et al., 2022), and multimodal approach, including biological and social aspects (Heidari et al., 2019). References Bourdon, P. C., Cardinale, M., Murray, A., Gastin, P., Kellmann, M., Varley, M. C., Gabbett, T. J., Coutts, A. J., Burgess, D. J., Gregson, W., & Cable, N. T. (2017). Monitoring athlete training loads: Consensus statement. International Journal of Sports Physiology and Performance, 12(Suppl 2), S2161–S2170. https://doi.org/10.1123/IJSPP.2017-0208 Fisher, A. J., Medaglia, J. D., & Jeronimus, B. F. (2018). Lack of group-to-individual generalizability is a threat to human subjects research. Proceedings of the National Academy of Sciences, 115(27), E6106–E6115. https://doi.org/10.1073/pnas.1711978115 Hader, K., Rumpf, M. C., Hertzog, M., Kilduff, L. P., Girard, O., & Silva, J. R. (2019). Monitoring the athlete match response: Can external load variables predict post-match acute and residual fatigue in soccer? A systematic review with meta-analysis. Sports Medicine - Open, 5(1), Article 48. https://doi.org/10.1186/s40798-019-0219-7 Heidari, J., Beckmann, J., Bertollo, M., Brink, M., Kallus, W., Robazza, C., & Kellmann, M. (2019). Multidimensional monitoring of recovery status and implications for performance. International Journal of Sports Physiology and Performance, 14(1), 2-8. https://doi.org/10.1123/ijspp.2017-0669 Kellmann, M., Bertollo, M., Bosquet, L., Brink, M., Coutts, A. J., Duffield, R., Erlacher, D., Halson, S. L., Hecksteden, A., Heidari, J., Kallus, K. W., Meeusen, R., Mujika, I., Robazza, C., Skorski, S., Venter, R., & Beckmann, J. (2018). Recovery and performance in sport: Consensus statement. International Journal of Sports Physiology and Performance, 13(2), 240–245. https://doi.org/10.1123/ijspp.2017-0759 Molenaar, P. C. M., & Campbell, C. G. (2009). The new person-specific paradigm in psychology. Current Directions in Psychological Science, 18(2), 112–117. https://doi.org/10.1111/j.1467-8721.2009.01619.x Neumann, N. D., Van Yperen, N. W., Brauers, J. J., Frencken, W., Brink, M. S., Lemmink, K. A. P. M., Meerhoff, L. A., & Den Hartigh, R. J. R. (2021). Nonergodicity in load and recovery: Group results do not generalize to individuals. International Journal of Sports Physiology and Performance, 17(3), 391–399. https://doi.org/10.1123/ijspp.2021-0126 Taylor, K.-L., Chapman, D., Cronin, J., Newton, M., & Gill, N. (2012). Fatigue monitoring in high performance sport: A survey of current trends. Journal of Australian Strength and Conditioning, 20, 12–23. Temm, D. A., Standing, R. J., & Best, R. (2022). Training, wellbeing and recovery load monitoring in female youth athletes. International Journal of Environmental Research and Public Health, 19(18), 11463. https://doi.org/10.3390/ijerph191811463 Thorpe, R. T., Atkinson, G., Drust, B., & Gregson, W. (2017). Monitoring fatigue status in elite team-sport athletes: Implications for practice. International Journal of Sports Physiology and Performance, 12(Suppl 2), S227–S234. https://doi.org/10.1123/ijspp.2016-0434