The Utility of Mixed Models in Sport Science: A Call for Further Adoption in Longitudinal Data Sets

Optimizing recovery: the impact of training load in elite volleyball players

This study aimed to analyze the relationships between training load metrics and various dimensions of recovery over the course of 105 days during the competitive season. This study included 14 elite male volleyball players who participated in 95 training sessions and 18 matches. The session rating of perceived exertion and inertial measurement units were utilized to measure internal and external training loads, respectively. Biological recovery was assessed using the countermovement jump test, while social and psychological recoveries were measured through the RESTQ-Sport. Relationships were examined using linear mixed-effect models to account for repeated measures across players and time points. The results revealed that chronic external training loads negatively influenced the biological recovery, with increases in chronic external training loads associated with a longer time to take-off (β=- 0.008, p=0.03) and a reduced modified reactive strength index (β=- 0.008, p=0.047). Social recovery was influenced by internal training load metrics, with acute internal training loads increasing social stress (β=- 3.512, p=0.034). For psychological recovery, chronic internal training loads and the acute:chronic internal training load ratio were linked to higher emotional exhaustion (β=- 0.013, p=0.029; β=0.021, p<0.001, respectively). These findings highlight the critical role of external load management in the biological recovery and the relevance of subjective internal load measures in monitoring social and psychological recovery dimensions. Coaches are encouraged to integrate time-based neuromuscular metrics and perceptual load assessments into their recovery monitoring protocols to optimize athletes' well-being and performance.

"Outlier detection: Underused in sport science despite outliers' impact on inference and prediction"

Outlier detection is critical in statistical analysis and predictive modelling, but it is often overlooked in research, leading to potentially inaccurate conclusions. This study aimed to (1) assess the prevalence of outlier detection strategies in sport science publications, (2) examine the effect of outliers on statistical inference using general linear mixed-effects models with longitudinal data, and (3) evaluate the impact of outliers on regression predictive models using the same dataset. A systematic literature search of 4,622 articles published in 2023 in Q1 journals in "sport science" category found that only 9.05% (99% CI: 4.87%-14.99%) reported employing outlier detection practices. A comprehensive outlier detection framework was designed to enhance data quality before formal statistical analysis using general(ised) linear mixed-effects models. The framework was applied to publicly available longitudinal data from rugby union small-sided games. Inclusion of outliers resulted in inaccurate maximum likelihood estimates, inflated standard errors, and biased conclusions regarding model parameters. Removing error outliers significantly improved the accuracy of predictive models, as evidenced by reduced root mean square errors. These findings underscore the importance of outlier detection in sport science research and demonstrate that appropriate handling of outliers enhances the validity of statistical inferences and predictions.

Examination of sex differences in fatigability and neuromuscular responses during continuous, maximal, isometric leg extension

Objective.This study examined sex-related differences in fatigability and neuromuscular responses using surface electromyographic (sEMG) and mechanomyographic (sMMG) amplitude (AMP) and frequency (MPF) during fatiguing, maximal, bilateral isometric leg extensions.Approach.Twenty recreationally active males and females with resistance training experience performed continuous, maximal effort, bilateral isometric leg extensions until their force reduced by 50%. Linear mixed effect models analyzed patterns of force, sEMG, and sMMG AMP and MPF responses in the dominant limb. An independent samples t-test compared time-to-task failure (TTF) between sexes.Main Results.There were no significant differences in TTF between males and females. However, males experienced a greater rate of force loss compared to females. Furthermore, sEMG AMP and MPF and sMMG AMP responses followed similar linear trends for both sexes, while sMMG MPF showed non-linear responses with sex-dependent differences.Significance.These data suggest that although TTF was similar, males had a higher rate of force reduction, likely due to greater absolute strength. Furthermore, despite parallel changes in sEMG AMP and MPF, as well as sMMG AMP, the divergent responses observed in sMMG MPF highlight sex-dependent differences in how males and females experience changes in the firing rates of active motor units during sustained maximal contractions.

Training loads and microcycle periodisation in Italian Serie A youth soccer players

Microcycles are fundamental structures for training prescription and load management, helping to optimise training effects and performance. This study quantified external and internal loads of Italian Serie A youth soccer players across competitive weeks and their periodisation within microcycles. Data were collected from 90 players belonging to four age groups (under-19, -17, -16, -15) across a season. Methods of monitoring external [duration and global navigation satellite systems (GNSS)] and internal load [heart rate (HR) and rating of perceived exertion (RPE)] were employed. Linear mixed models determined differences in training loads across age groups, training days and player positions. Under-19 and under-17 players trained five times per week, while younger players trained four times. Late-stage academy players (under-19 and -17) demonstrated higher weekly accumulated external and sRPE training load compared to their younger counterparts (p < 0.05 between groups). Weekly accumulated HR internal loads were higher in under-15 players (p < 0.05 between groups). Marked fluctuations of daily load were observed across microcycles in under-19 and under-17 groups (p < 0.05 between days). These findings highlight progressive increases in training load throughout the development pathway, with late-stage academy players training with higher frequency, volume and marked periodisation compared to younger players.

Effect of Data-Processing Methods on Acceleration Summary Metrics of GNSS Devices in Elite Australian Football

This study aimed to measure the differences in commonly used summary acceleration metrics during elite Australian football games under three different data processing protocols (raw, custom-processed, manufacturer-processed). Estimates of distance, speed and acceleration were collected with a 10-Hz GNSS tracking technology device from fourteen matches of 38 elite Australian football players from one team. Raw and manufacturer-processed data were exported from respective proprietary software and two common summary acceleration metrics (number of efforts and distance within medium/high-intensity zone) were calculated for the three processing methods. To estimate the effect of the three different data processing methods on the summary metrics, linear mixed models were used. The main findings demonstrated that there were substantial differences between the three processing methods; the manufacturer-processed acceleration data had the lowest reported distance (up to 184 times lower) and efforts (up to 89 times lower), followed by the custom-processed distance (up to 3.3 times lower) and efforts (up to 4.3 times lower), where raw data had the highest reported distance and efforts. The results indicated that different processing methods changed the metric output and in turn alters the quantification of the demands of a sport (volume, intensity and frequency of the metrics). Coaches, practitioners and researchers need to understand that various processing methods alter the summary metrics of acceleration data. By being informed about how these metrics are affected by processing methods, they can better interpret the data available and effectively tailor their training programs to match the demands of competition.

A framework for test measurement selection in athlete physical preparation

Preparing athletes for competition requires the diagnosis and monitoring of relevant physical qualities (e.g., strength, power, speed, endurance characteristics). Decisions regarding test selection that attempt to measure these physical attributes are fundamental to the training process yet are complicated by the myriad of tests and measurements available. This article presents an evidenced based process to inform test measurement selection for the physical preparation of athletes. We describe a method for incorporating multiple layers of validity to link test measurement to competition outcome. This is followed by a framework by which to evaluate the suitability of test measurements based on contemporary validity theory that considers technical, decision-making, and organisational factors. Example applications of the framework are described to demonstrate its utility in different settings. The systems presented here will assist in distilling the range of measurements available into those most likely to have the greatest impact on competition performance.

Optimizing wheelchair basketball lineups: A statistical approach to coaching strategies

This study was designed to support the tactical decisions of wheelchair basketball (WB) coaches in identifying the best players to form winning lineups. Data related to a complete regular season of a top-level WB Championship were examined. By analyzing game-related statistics from the first round, two clusters were identified that accounted for approximately 35% of the total variance. Cluster 1 was composed of low-performing athletes, while Cluster 2 was composed of high-performing athletes. Based on data related to the second round of the Championship, we conducted a two-fold evaluation of the clusters identified in the first round with the team's net performance as the outcome variable. The results showed that teams where players belonging to Cluster 2 had played more time during the second round of the championship were also those with the better team performance (R-squared = 0.48, p = 0.035), while increasing the playing time for players from Classes III and IV does not necessarily improve team performance (r2 = -0.14, p = 0.59). These results of the present study suggest that a collaborative approach between coaches and data scientists would significantly advance this Paralympic sport.

Dose-Response Relationship for External Workload and Neuromsuclar Performance Over a Female, Collegiate, Basketball Season

ABSTRACT

Philipp, NM, Cabarkapa, D, Blackburn, SD, and Fry, AC. Dose-response relationship for external workload and neuromsuclar performance over a female, collegiate, basketball season. J Strength Cond Res 38(5): e253-e263, 2024-The aim of this study was to investigate the relationship between external workload exposure and changes in countermovement jump force-time characteristics over the course of an entire basketball season, in a sample of National Collegiate Athletic Association Division I, female, basketball players. Data for 12 players were retrospectively analyzed, with external workload being quantified by means of an exponentially weighted, acute, and chronic workload, as well as an acute:chronic workload ratio derived from an inertial measurement unit-based system worn by athletes for all practices and games during the regular season. Countermovement jumps were performed on a total of 26 test days over the span of the in-season competitive period. To statistically analyze these relationships, and to account for multiple observations of the same athletes in a data set, linear mixed-effects models with athlete identity (ID) intercept as the random effect were used. Study findings suggested that associations between external workload exposure and respective force-time characteristics after controlling for the random effect of athlete ID were dependent on the specific metric or metric subgroup used, as well as the type of workload exposure (e.g., acute vs. chronic). Force-time signatures from the braking phase (e.g., average braking force) seemed to be particularly associated with higher degrees of acute workload exposure, whereas strategy-based metrics such as countermovement depth showed significant associations with chronic workload exposure. Furthermore, model results suggested the importance of analyzing neuromuscular responses to external workload on an individual basis, rather than across an entire team. Findings might help practitioners in their selection process related to metrics of interest in monitoring neuromuscular fatigue and readiness.

The effect of advanced footwear technology on elite male marathon race speed

PURPOSE

The aim of this study was to explore the ergogenic effect of advanced footwear technology (AFT) upon world-class male marathon running speed.

METHOD

A retrospective analysis of 99 world-class male marathon runners' performances between 2012 and 2021 was undertaken, providing a sample size of 971 performances, split into two footwear groups: AFT (n = 299) and traditional (n = 672). Additionally, details regarding the year of the marathon performance and racecourse were extracted. A mixed model for repeated measures (MMRM) analysis were undertaken identifying athlete (Wald Z = 2.821; p = .005) and course (Wald Z = 4.111; p < 0.001) as significant contributors to the variance in marathon running speed and as such were included as random factors with footwear type set as a fixed factor.

RESULTS

World-class male marathon running speeds were significantly faster (p < 0.001) when running in AFT (5.441 m.s-1) when compared with traditional shoes (5.386 m.s-1) with a mean difference of 0.055 m.s-1 (95% CI 0.039-0.071 m.s-1), translating to an improvement in marathon speed of 1.0% or a 79 s improvement in marathon race time.

CONCLUSION

Our findings demonstrate an improvement in world-class male marathon running speed of 1% when running in AFT, a near identical degree of improvement to the male marathon world record ran in AFT. Whilst a 1% improvement in marathon running times associated with AFT is smaller than previously predicted utilizing laboratory-based models, this still reflects a significant degree of improvement at the elite level.

An 8-week injury prevention exercise program combined with change-of-direction technique training limits movement patterns associated with anterior cruciate ligament injury risk

Knee ligament sprains are common during change-of-direction (COD) maneuvers in multidirectional team sports. This study aimed to compare the effects of an 8-week injury prevention exercise program containing COD-specific exercises and a similar program containing linear sprint exercises on injury- and performance-related variables during a 135° COD task. We hypothesized that the COD-specific training would lead to (H1) stronger reductions in biomechanical variables associated with anterior cruciate ligament (ACL) injury risk during COD, i.e. knee abduction moment and angle, hip internal rotation angle and lateral trunk lean, and (H2) more effective improvements in COD performance according to the COD completion time, executed angle, ground contact time, and approach speed. Twenty-two sports science students (40% female) completed biomechanical assessments of COD movement strategies before and after participating in two supervised 25-min training sessions per week over 8 weeks. We observed significant 'training x group' interaction effects in support of H1: the COD-specific training but not the linear sprint training led to reduced peak knee abduction moments (interaction, p = 0.027), initial knee abduction (interaction, p < 0.001), and initial lateral trunk lean angles (interaction, p < 0.001) compared to baseline. Although the COD-specific training resulted in sharper executed angles (interaction, p < 0.001), the sprint-specific training group showed reduced COD completion (interaction, p = 0.037) and ground contact times (interaction, p < 0.001). In conclusion, a combination of generic and COD-specific injury prevention training resulted in COD technique adaptations that can help to avoid ACL injury-prone COD movements but may negatively affect COD speed.

The impact of sprint, middle-distance, and long-distance orienteering races on perceived mental fatigue in national level orienteers

Experiencing mental fatigue (MF) before an orienteering race can lead to a slower completion time. This study aimed to explore the changes in perceived MF, mood and other psychological responses during an orienteering competition. Sixteen national level orienteering athletes (20.8 ± 4.9 years) provided informed consent and completed the online surveys, before and immediately after each race, and 24- and 48-hours post competition (48POST). This study measured MF, physical fatigue, stress, tiredness and motivation using 0-100 Visual Analogue Scale, and the mood was assessed using The Brunel Mood Scale (BRUMS). A moderate to large increase in MF (ES = 0.93 [0.54 to 1.31]), BRUMS fatigue (ES = 0.61 [0.3 to 0.92]), and PF (ES = 1.21 [0.81 to 1.61]) was reported following orienteering races. A small increase in tiredness and BRUMS confusion, and a small decrease in motivation, stress and BRUMS vigour was also reported. There was a delay in recovering from the MF elicited by competition, with a small increase in MF (ES = 0.54 [0.08 to 1.15]) at 48POST compared to the pre-competition value. This study found that orienteers experience MF during competition and have a delayed recovery that can last up to two days after the competition.

Rethinking the Statistical Analysis of Neuromechanical Data

Researchers in neuromechanics should upgrade their statistical toolbox. We propose linear mixed-effects models in place of commonly used statistical tests to better capture subject-specific baselines and treatment-associated effects that naturally occur in neuromechanics. Researchers can use this approach to handle sporadic missing data, avoid the assumption of conditional independence in observations, and successfully model complex experimental protocols.

Season Match Loads of a Portuguese Under-23 Soccer Team: Differences between Different Starting Statuses throughout the Season and Specific Periods within the Season Using Global Positioning Systems

This study aimed to quantify the external match loads (EMLs) of a Portuguese u-23 soccer team, competing at the highest national level for the age group, comparing players with different starting status throughout a competitive season and specific blocks. Thirty-five outfield soccer players were split into three groups for the entire season analysis and for each 3-month block, based on the percentage of games played as a starter. The three groups consisted of “starters” (≥55% of the games as a starter), “fringe” (30−54%), and “non-starters” (<30%). EMLs were recorded using 10 Hz GPS technology throughout the whole season (26 matches). Differences (p < 0.05) were found for total distance (TD), exposure time, and the number of accelerations and decelerations between starters and non-starters throughout the season (d = 0.73 to 1.08), and within each block (d = 0.59 to 1.68). Differences were also found between starters and fringe players for the number of accelerations in Block 2 (p = 0.03; d = 0.69), and TD (p = 0.006; d = 1) and exposure time (p = 0.006; d = 0.95) in Block 3. Differences in the EML were almost always accompanied by large differences in game time. Our results highlight the differences in the EML of starters and non-starters, emphasizing the need for compensatory training, especially with players that obtain significantly less playing time, to prepare the players for match demands (e.g., high-intensity efforts such as sprinting, accelerations, and decelerations).