RESUMEN
Introduction: Measurement of training in water polo goalkeepers has focused first on psycho-physiological variables, but also on external volume estimated with wearable sensors. However, there are limited studies exploring training monitoring in water polo goalkeepers longitudinally. Methods: Three female senior national team goalkeepers participated in this study from May to August 2021. Internal loads were defined using session rating of perceived exertion (sRPE). Tri-axial accelerations and angular velocities were measured with an inertial measurement unit (IMU) placed on the lower back to measure external loads. Relationships between self-reported and IMU-derived metrics were explored using Spearman correlations. Two-way ANOVAs were used to assess differences between session types and between athletes. Results: In total, 247 sessions were collected (159 practices, 67 matches and 21 game warm up), with 155 sessions having complete data. IMU metrics, such as number of kicks, number of jumps or player-load showed high correlation with each other (ρ = 0.80-0.88). There was also a moderate correlation (ρ = 0.47, 95% CI = 0.33-0.58) between sRPE and player-load measured with the IMU. ANOVA tests showed that there were significant differences between athletes for sRPE (p < 0.01) but not for player load (p = 0.47). There were no interactions between athletes and training types, except for index score (p < 0.01). Conclusions: This study shows that monitoring of training loads can be performed successfully in water polo goalkeepers using a combination of self-reported and IMU measures. Self-reported outcomes can be expected to vary significantly across athletes within the same session, while IMU metrics vary across training situations. Finally, coaches should be mindful of missing data, as they can skew the interpretation of training loads.
RESUMEN
Injuries limit the athletes' ability to participate fully in their training and competitive process. They are detrimental to performance, affecting the athletes psychologically while limiting physiological adaptations and long-term development. This study aims to present a framework for developing random forest classifier models, forecasting injuries in the upcoming 1 to 7 days, to assist the performance support staff in reducing injuries and maximizing performance within the Canadian National Female Short-Track Speed Skating Program. Forty different variables monitored daily over two seasons (2018-2019 and 2019-2020) were used to develop two sets of forecasting models. One includes only training load variables (TL), and a second (ALL) combines a wide array of monitored variables (neuromuscular function, heart rate variability, training load, psychological wellbeing, past injury type, and location). The sensitivity (ALL: 0.35 ± 0.19, TL: 0.23 ± 0.03), specificity (ALL: 0.81 ± 0.05, TL: 0.74 ± 0.03) and Matthews Correlation Coefficients (MCC) (ALL: 0.13 ± 0.05, TL: -0.02 ± 0.02) were computed. Paired T-test on the MCC revealed statistically significant (p < 0.01) and large positive effects (Cohen d > 1) for the ALL forecasting models' MCC over every forecasting window (1 to 7 days). These models were highly determined by the athletes' training completion, lower limb and trunk/lumbar injury history, as well as sFatigue, a training load marker. The TL forecasting models' MCC suggests they do not bring any added value to forecast injuries. Combining a wide array of monitored variables and quantifying the injury etiology conceptual components significantly improve the injury forecasting performance of random forest models. The ALL forecasting models' performances are promising, especially on one time windows of one or two days, with sensitivities and specificities being respectively above 0.5 and 0.7. They could add value to the decision-making process for the support staff in order to assist the Canadian National Female Team Short-Track Speed Skating program in reducing the number of incomplete training days, which could potentially increase performance. On longer forecasting time windows, ALL forecasting models' sensitivity and MCC decrease gradually. Further work is needed to determine if such models could be useful for forecasting injuries over three days or longer.
RESUMEN
High-Intensity Interval Training (HIIT) is a time-efficient training method suggested to improve health and fitness for the clinical population, healthy subjects, and athletes. Many parameters can impact the difficulty of HIIT sessions. This study aims to highlight and explain, through logical deductions, some limitations of the Skiba and Coggan models, widely used to prescribe HIIT sessions in cycling. We simulated 6198 different HIIT training sessions leading to exhaustion, according to the Skiba and Coggan-Modified (modification of the Coggan model with the introduction of an exhaustion criterion) models, for three fictitious athlete profiles (Time-Trialist, All-Rounder, Sprinter). The simulation revealed impossible sessions (i.e., requiring athletes to surpass their maximal power output over the exercise interval duration), characterized by a few short exercise intervals, performed in the severe and extreme intensity domains, alternating with long recovery bouts. The fraction of impossible sessions depends on the athlete profile and ranges between 4.4 and 22.9% for the Skiba model and 0.6 and 3.2% for the Coggan-Modified model. For practitioners using these HIIT models, this study highlights the importance of understanding these models' inherent limitations and mathematical assumptions to draw adequate conclusions from their use to prescribe HIIT sessions.
