A causal inference method for athletic injuries based on quantile threshold functions and latent Gaussian DAG models

IntroductionCausal inference of athletic injuries provides the critical foundations for the development of effective prevention strategies. In recent years, the directed acyclic graph model (DAG) has established itself as an indispensable tool in the study of athletic injuries.MethodsThis study proposes a quantile threshold function (QTF) and integrates it with the causal inference framework within the latent DAG model for ordinal variables. This process begins by transforming continuous variables into ordinal variables to construct a DAG, which is analyzed using the latent causal inference framework to estimate ordinal causal effects (OCE).ResultsTesting this approach on real-world data showed clear differences between groups (F > 52,000, P < 0.05). The analysis also revealed three direct paths and two indirect paths related to athletic injuries, based on the DAG.DiscussionWe obtained the OCE by intervening on variables that directly or indirectly influence athletic injuries. DAG path analysis further elucidated the impact of causal pathways on the risk of injury. The approach proposed in this study provides novel theoretical and methodological insights into athletic injuries and serves as a crucial basis for optimizing training programs and mitigating injury risk.