Within applied sports science and medicine research, many challenges hinder the establishment and detailed understanding of athletic injury causality as well as the development and implementation of appropriate athletic injury prevention strategies. Applied research efforts are faced with a lack of variable control, while the capacity to compensate for this lack of control through the application of randomised controlled trials is often confronted by a number of obstacles relating to ethical or practical constraints. Such difficulties have led to a large reliance upon observational research to guide applied practice in this area. However, the reliance upon observational research, in conjunction with the general absence of supporting causal inference tools and structures, has hindered both the acquisition of causal knowledge in relation to athletic injury and the development of appropriate injury prevention strategies. Indeed, much of athletic injury research functions on a (causal) model-blind observational approach primarily driven by the existence and availability of various technologies and data, with little regard for how these technologies and their associated metrics can conceptually relate to athletic injury causality and mechanisms. In this article, a potential solution to these issues is proposed and a new model for investigating athletic injury aetiology and mechanisms, and for developing and evaluating injury prevention strategies, is presented. This solution is centred on the construction and utilisation of various causal diagrams, such as frameworks, models and causal directed acyclic graphs (DAGs), to help guide athletic injury research and prevention efforts. This approach will alleviate many of the challenges facing athletic injury research by facilitating the investigation of specific causal links, mechanisms and assumptions with appropriate scientific methods, aiding the translation of lab-based research into the applied sporting world, and guiding causal inferences from applied research efforts by establishing appropriate supporting causal structures. Further, this approach will also help guide the development and adoption of both relevant metrics (and technologies) and injury prevention strategies, as well as encourage the construction of appropriate theoretical and conceptual foundations prior to the commencement of applied injury research studies. This will help minimise the risk of resource wastage, data fishing, p-hacking and hypothesising after the results are known (HARK-ing) in athletic injury research. Key Points Athletic injury research has a large reliance upon observational research to guide applied practice. However, appropriate supporting causal inference tools and structures are generally absent within the existing literature, which is problematic. Owing to various forms of bias, such as confounding and collider-stratification bias, current research approaches may erroneously implicate variables that are protective to athletes with an unchanged or increased injury risk, as well as variables that are harmful to athletes with an unchanged or decreased injury risk. This may facilitate the implementation of injury prevention strategies that are ineffective or, worse, increase the risk of injury and interfere with an athlete`s training process. To help alleviate these concerns, athletic injury research and prevention efforts should shift their attention towards the formation, utilisation, investigation and when necessary, revision or replacement of causal diagrams including theoretical and conceptual causal frameworks and models, DAGs and similar diagrammatical constructs.
© Copyright 2024 Sports Medicine. Springer. Alle Rechte vorbehalten.