Athlete Workloads During Collegiate Women's Soccer Practice: Implications for Return to Play

This was a retrospective, observational study to evaluate biometric load variables in female collegiate soccer players during specific practice activities. The key metrics evaluated were total distance (m), total PL (arbitrary units [AU]), total distance per minute (m/min), and PL per minute (AU/min). Data collection and methods were approved by our institutional review board.

This study involved 32 female collegiate soccer players (age = 20 ± 1 years, height = 168.75 ± 4.28 cm) from an NCAA Division I university. Participants were classified as defenders, forwards or strikers, or midfielders by the coaching staff at the beginning of the season (Table 1). Positions within these 3 categories can vary, but these categories are consistent with those used by previous researchers^8,15  in soccer studies. Goalkeepers were excluded from this study as a result of the unique nature of their position. Practices and games were monitored from preseason (August) through postseason (November). A total of 50 practices and 21 games were recorded. Our university's institutional review board deemed this study exempt due to the retrospective nature of the work and because all data were collected as part of routine athlete monitoring already occurring within the team. Individual player data were deidentified before analysis to protect participant privacy.

The PL is, therefore, a volume-based metric representative of cumulative workload. Both total distance per minute and PL per minute are intensity metrics for which the accumulated measure completed by the athlete is divided by the number of minutes the athlete participated in the drill or activity. For PL per minute, the metric gives an estimate of how much cumulative acceleration, deceleration, change of direction, and vertical displacement (using the Equation) occurred per unit of time.

We intentionally chose not to analyze the distance covered at specific velocity bands for running activities because the predetermined velocity bands in the proprietary software were derived from normative data from male athletes and were not appropriate for classifying the running velocities of the collegiate women's soccer players in this study.

Practice segments were labeled in real time by the team's strength and conditioning coach using specific drill names. For analysis, these specific drill names were then grouped into physical skills, technical skills, tactical skills, large-sided competitions, small-sided competitions, and miscellaneous categories. These categories were based on the work of earlier investigators²⁰  in soccer. Individual sessions and institution-specific drills that did not fit in specific categories were classified as miscellaneous. Descriptions of these drill structures are presented in Table 2. Game data were captured to use as a reference point for practice volumes and intensities.

Given that this was a preliminary assessment of women's soccer data, we made a preemptive decision to use descriptive assessments in lieu of inferential statistics. This approach supports our primary purpose of providing ATs and other clinicians with details on practice structures for collegiate women's soccer athletes. Therefore, total distance, total PL, total distance per minute, and PL per minute were assessed by drill group and player position. Descriptive analyses were completed using Jamovi (version 1.2; The Jamovi Project, Sydney, Australia)²⁷  with graphical analysis using both Jamovi and Tableau (version 2019.2.9; Tableau Software, LLC, Mountain View, CA). Descriptive analyses for game data by position were similarly assessed for all outcomes, and practice data were contextualized as a percentage of game-play demands. Both total distance and PL were highly skewed (>1); thus, we reported the median and interquartile range.

The median and interquartile range for each drill group based on position are provided in Table 3; the distribution graphs for Table 3 are presented in Figures 1 through 4. Additionally, stacked bar graphs that represent the percentage contribution of each drill group during an average practice separated by total distance, total distance per minute, total PL, and PL per minute are displayed in Figures 5 through 8, respectively. For all measures, large-sided competitive drills had the greatest percentages (≥19.81%) for each position group. Across the 3 position groups, defenders had the highest medians for all measures during an average practice session. For reference, in this cohort, midfielders had the highest medians for all measures during an average game (Figures 9A through 9D). Total distance and PL were twice as high during practices than during game play (practice total distance = 204% of game play, practice total PL = 193% of game play). Total distances for physical and tactical skills drills accounted for 22% of total game distance, whereas technical skills drills constituted 16% of game volume. Large- and small-sided competition drills were higher-volume practice components, responsible for 42% and 24% of game volume, respectively. Players exhibited PL at approximately 50% of game play for physical, technical skills, and tactical skills drills during practices, and simulated competitions nearly mimicked game play, with 92% volume for large-sided competitions and 70% for small-sided competition drills.

Our results highlighted several important characteristics of the drill structure and female soccer PL that sports medicine clinicians, sports scientists, and coaches may consider in future practice. Given that past athlete-monitoring investigations^{15–17,20,22}  of women's collegiate soccer players have largely focused on cumulative practice and game data, we added valuable insights into practice-specific demands by drill structure. Clinicians may use this information to facilitate rehabilitation and return-to-play decision making, and coaches and sports scientists may consider this information with respect to training demands in female soccer athletes.

Descriptive assessments reflected that defenders had high median total distance, total distance per minute, total PL, and PL per minute outcomes. These outcomes were consistent with the relative median workloads during game play and, therefore, reinforce the fact that practices mimicked playing style for this team across the season. Previous assessments of game data for female soccer athletes have been mixed; some researchers observed higher demands on defenders than on players at other positions¹⁵  and others noted the highest density of loading on forwards and strikers.^16,22  Our findings are likely attributable to player rotation for the team we analyzed, as defenders tended to play more minutes per game than other position players. Furthermore, these mixed results in the literature may have been influenced by the style of game play based on the level of competition and by different teams, highlighting the need to implement athlete monitoring for individual team and player insights.²  The breakdown of training demands by drill type in our study was consistent across player position, regardless of the session totals. Therefore, ATs and other key stakeholders may gain the best insights into activity modifications and return-to-play decision making by assessing the demand by drill structure rather than by assuming that some specific drills are “safer” during the gradual implementation of stress during this time.

Physical drills, including warm-ups and conditioning for sustained aerobic exercise, constituted approximately 20% of all external-loading metrics during practices. This is important, as physical drill distance accounted for a median of 660 m for defenders, 622 m for forwards and strikers, and 640 m for midfielders per practice (Table 3). Athletic trainers may use these data to inform alternative forms of physical conditioning for athletes recovering from overuse lower extremity injuries in order to reduce the overall load during practices while attaining similar training benefits. For example, athletes recovering from stress fractures may use other cardiovascular training equipment, such as a bicycle or elliptical machine, to warm up or complete conditioning drills. This tactic would reduce the cumulative impact load imposed on the athletes by approximately one-fifth of a regular practice while reserving the allowed total distance for the practice to be used for more soccer-specific or team-specific drills.

Technical and tactical skills drills both require athletes to perform activities with the soccer ball to hone in-play skills; however, the key difference is that technical work focuses more on ball handling and possession and tactical work involves a performance-oriented outcome (ie, corner kick to score or other restart scenarios for either offense or defense).²²  The descriptive assessments showed that tactical skills drills had a high associated training intensity during play, as reflected in the total distance per minute and PL per minute (approximately 16%–17%), and players exhibited about 50% volume during practices relative to game play. These results are important to consider given that PL is calculated using triaxial accelerometry loading metrics and takes into account acceleration, deceleration, change of direction, and vertical displacement in the PL metrics. Tactical skills drills required high workload intensity, so these drill structures may be limited or closely monitored using wearable technology during early return to play, particularly for players with lower extremity joint injuries, such as anterior cruciate ligament tears and lateral ankle sprains, with rotational mechanisms of injury.^28,29 

Simulated game-play drill structures imposed the greatest demand on the athletes and were similar to game-play demands across all athlete monitoring outcomes, regardless of player position. Also, competition sizes were comparable with one another despite the difference in the number of athletes involved in play. This finding was somewhat surprising because, in previous evaluations of different sizes of game play, researchers³⁰  found that smaller-sided competition drills imposed increased levels of play intensity. However, these outcomes were largely influenced by the area of play rather than the number of players participating; large-sided competition drills enable athletes to rely on teammate support and reduce the workload on individual players.³⁰  Given the retrospective nature of this investigation, the dataset provided no information on the size of the field of play during the simulated game drill structures (ie, full versus half field). It is plausible that the area of play was modified based on the size of the competition, thereby adjusting for distance and intensity across these drill structures. Future authors should delve into simulated game-play structures to elucidate these demands on female athletes. Still, our data can be used to help inform rehabilitation progressions across a spectrum of injuries. Based on an athlete's specific phase of the return-to-sport progression during the recovery from injury, the clinician may have the athlete pursue certain parts of practice that align with the workload volume and intensity goals of the personalized rehabilitation plan. Clinicians can use these objective findings to promote dialogue with coaches and athletes in order to facilitate a safe return to play.

As several of the drills recorded during practice were specific to the team studied, we binned a small subset of observations into a miscellaneous category to increase the external applicability of this dataset to other female soccer teams. We should, as previously discussed, incorporate athlete monitoring via wearable technology into competitive sports to make the best team- and athlete-specific decisions.²  Nevertheless, our results can be applied to clinical practice because general drill structures that uphold the principles of soccer play were assessed in these athletes.²⁰ 

Historically, sports medicine clinicians and other performance specialists have largely advanced return-to-play protocols subjectively; however, the ability to measure and analyze objective workload data as injured athletes advance through rehabilitation allows for a more objective approach to designing rehabilitation programs and making return-to-sport decisions. Incorporating wearable sensors in sport activities, particularly for athletes in field-based sports that do not have obviously calculable loading outcomes (eg, pitch counts in baseball pitchers), creates opportunities for more effective workload monitoring and management. Leveraging these data can help clinicians optimize rehabilitation programs and creates the opportunity for data-informed dialogue among the key stakeholders to return athletes safely to activity after injury. Although assessing injury risk was beyond the scope of our study, accumulating more workload data on athletes during practices and games will allow future researchers to prospectively assess the relationships between external-load accumulation and injury risk.

This dataset was based on a single season of play by a single NCAA Division I women's soccer team, so the results may not be fully generalizable to all teams or levels of play. All of the athletes studied were injury free, and loading outcomes may differ for injured soccer players. In addition, styles of coaching and play may have factored into these outcomes, and the exact session breakdown may not transfer to other settings. However, our main purpose was to obtain basic descriptive information on drill structures to provide ATs and other stakeholders with a starting point for interpreting the athletic demand on female soccer athletes during specific drills. We only assessed distance and PL metrics because other common sensor-derived outcomes, such as those using running velocity bands, were not appropriate to the female athletes assessed in this investigation. Future iterations of sensor algorithms should be tailored to female populations to increase the applicability of the outcomes to these athletes. Finally, our analyses combined data from all practices, and underlying differences among practices likely depended on when they occurred during the season and in relation to games. Further analysis of such patterns is warranted and can help clinicians design rehabilitation programs and tailor return-to-play progressions of individual athletes.

Overall, this descriptive analysis of female soccer athletes' workloads based on player position across practice drill structures reflected that defenders had the highest overall demand during practices. Across all positions, simulated game-play data imposed the highest load on athletes and was prescribed similarly to game play, and tactical skills drills had higher PL intensity. Athletic trainers and other health care professionals may use this information to develop more objective return-to-play guidelines and engage coaches and sports scientists in dialogue on athlete-loading adjustments.