The relationship between team-level and league-level injury rate, type and location in a professional football league

Objectives: To describe the relationship between team- and league-level variability of injury rate, type, and location over 6 seasons in professional Australian football (A-League).

Design: Prospective epidemiological study.

Methods: Injury incidence, type and location were collected from all A-League teams (n = 10) for 6 consecutive seasons (2012/13 to 2017/18) via a standardised injury surveillance system. Intra-class correlation and coefficient of variation were calculated to assess the between-season variability of injury measures for each team. To determine the relationship between team-level injury variability on league-level injury rates, Marginal Coefficient of Determination to Coefficient of Determination were then calculated from generalised linear mixed models. This allowed determination between season trends, where league-level injury incidence, type- and location rates as the response variables, season as the predictor variable and teams as random intercepts.

Results: The majority of teams showed poor to moderate correlations for between-season injury rates (intra-class correlation: r = 0.319-0.831), but also showed low-moderate variability between-seasons for injury rate (coefficient of variation 34 ± 22%). League injury rates were stable, though were reduced in 2015/16 compared to 2012/13 (β = 0.738; p = 0.011). Joint/Ligament injuries had coinciding significant reduction in 2015/16 (p = 0.001). The model variance showed the reduction of Joint/Ligament injuries was league-wide rather than team-specific (Marginal Coefficient of Determintion = 0.23; Coefficient of Determination = 0.23).

Conclusions: In the A-League, low between-season injury rate variability from teams contributed to a stable league-level injury trend over seasons. A reduction in league injury rate in 2015/16 was mirrored by league-wide Joint/Ligament injury rates, without specific effect by team.