An interdisciplinary examination of stress and injury occurrence in athletes

Abstract

This thesis examined the multifaceted relationship between stress and sports injury. Study 1 explored the relationships between psychological sources of stress (major life events and personality characteristics) and stress-related physiological markers (heart rate variability, muscle stiffness and postural control) using a prospective, repeated measures design. Two Bayesian networks were used to perform the analysis and provided probabilistic statements regarding the effect of different combinations of variables in the network on injury occurrence. The first network revealed that “High” levels of muscle stiffness resulted in the greatest probability (Pr) of injury (Pr = 0.31). However, there was no meaningful difference between “Low” and “High” levels of negative life events on the probability of sustaining an injury (“Low” Pr = 0.24, “High” Pr = 0.26), despite a large body of research finding evidence to the contrary. The second network explicitly modelled changes between time points and found that the combination of increases in muscle stiffness and negative life events resulted in the greatest probability of sustaining an injury (Pr = 0.71). Study 2 addressed a number of research questions that built on those of Study 1, including; whether additional measures such as the stress hormone cortisol was associated with major life events and injury; whether an alternative method of scoring major life events would be related to injury; and how these variables related to both injury occurrence and severity. A subsample from the first study of male football and male rugby players were recruited for the study. Both Bayesian hurdle regression and Bayesian linear regression models were used to analyse the data. Findings revealed that higher levels of both average negative life event score and muscle stiffness increased the probability of injury occurrence and the number of days lost due to injury, although large credible intervals (CrI) were present. The relationship between cortisol and injury was less clear, with each of the two teams involved in the study demonstrating a different response (football, estimate = 0.10, 95% CrI = [-0.43, 0.62]; rugby, estimate = 0.54, 95% CrI = [0.05, 1.05]). The thesis concludes with a discussion of conceptual and theoretical issues, practical implications, strengths and limitations, and directions for future research.