Soccer is a complex, high-intensity sport and is associated with high injury risk. Prediction models to reduce injuries have received considerable attention over the last few years, but their success in reducing injuries has been variable. This thesis aimed to investigate injury prediction research that incorporates the physical profile of the player and functional screening tests, as well as coach-rated perceptions. Study one reviewed the current literature regarding epidemiology and injury mechanisms for adult male soccer players. Findings identified that lower limb injury rates in soccer players had remained an issue over the last ten years. Evidence collated in study one indicated that prevention and rehabilitation of lower limb injuries should be a focus in adult male soccer, with particular emphasis on injuries of the hip/groin, thighs, knees, and ankles. Study two showed the outcome of a comprehensive systematic review and meta-analysis that investigated the risk factors for groin and hamstring injuries among adult (≥18 years) male professional soccer players. Findings demonstrated a previous history of injuries, increasing age, and playing specific positions as significantly associated with groin and hamstring injuries, and found that increased countermovement jumping distance was significant to groin injuries, while higher BMI was a risk factor for hamstring injuries. Strength may play an important role in groin and hamstring injuries, but consistent measures are needed to determine the effect. Study three showed the outcome of a comprehensive systematic review and meta-analysis that investigated the risk factors for knee and ankle injuries among adult (≥18 years) male professional soccer players. Study three demonstrated that a previous history of injuries is significantly associated with knee and ankle injuries. Training and playing matches on an artificial turf pitch was the single modifiable risk factor associated with sustaining an ankle injury, and further studies should consider this. Intrinsic performance tests were not associated with knee or ankle injury risk. Muscle strength and neuromuscular control factors may play a vital role in knee and ankle injuries. However, data measures were not sufficiently consistent to measure the influence. Study four assessed the associations between the history of lower limb injury and performance in physical tests during the pre-season and the association between pre-season fitness and injury events during the season. Findings confirmed that the machine learning approach is beneficial as a predictive mechanism to monitor the most prevalent lower limb injuries in footballers. In the cohort examined, a Naive Bayes model suggested that previous injuries, anthropometric data, and pre-season functional screening tests have 80-88% classification accuracy to injury risk in professional male soccer players. Overall, pre-season functional screening tests and identified injury risk factors can be used to recognise players at risk of injury. Results from the used model can be considered cut-off scores to distinguish who is at risk from who is not. Subsequently, a suitable intervention can be chosen and implemented to lower the injury risk. Study five evaluated whether the pre-season fitness tests were associated with the coaches’ rating of player performance during the season. Findings showed that coaches' qualitative assessment of physical skill closely matched the independently measured functional performance tests. Given the importance of the central role of the coaches in the team, specifically the players. Their close monitoring and judgment of the players' performance could assist in the selection, preparation criteria, and player return to play post-injury. It is recommended that future studies standardise data collection by adhering to current consensus statements on injury definition and data collection procedures and analysing the specific diagnoses of the injury. Future studies could consider multivariate models using standardised clinical and functional measures and incorporating these features to understand further injury prediction and prevention in adult male soccer players. Predictions of potential sporting injuries are growing, and advancements in this area need to be promoted, as machine learning methods have substantial potential. Further validation of implemented models in subsequent seasons, and with other teams is required to improve confidence in these predictive models.
Risk assessment and predictive measures associated with the most prevalent injuries in male professional soccer players
Alfarraj, F. (Author). 2023
Student thesis: Doctoral Thesis