Background Shoulder injuries are a significant risk in rugby union football. However, the aetiology and the risk factors associated with shoulder injuries are not well understood.

Objective To analyse the incidence and risk factors for shoulder injuries in rugby union football.

Design A prospective shoulder injury study was carried out in 2009 using a recursive injury model framework.

Setting Community level adult male rugby union players.

Participants 75 players at the start and 54 players at the end.

Assessment of risk factors The effects of player size, age, injury history and usage of shoulder padding were assessed.

Main outcome measurements Shoulder injury incidence rate (IIR). Performance on shoulder assessments using the Constant Score Protocol was measured pre-, mid- and end-season. Weekly training and shoulder impairment was measured.

Results A total of 1138.1 player participation hours (pph) were recorded. The overall shoulder IIR was 4.4 injuries per 1000 pph with an average severity of 1.0 missed game. Previous injury and shoulder pads were the main observed intrinsic and extrinsic injury risk factors, respectively. An IIR of 3.5/1000 pph were recorded in players with no history of shoulder injury in the last 3 years (80%). There was a non-significant reduction in injury incidence for players wearing shoulder pads during a match. The tackle has been identified as the intrinsic factor that contributed to 67% of the total time loss to injury. Shoulder pain and impairment were reported by 41% of the players throughout the season but none caused the players to miss any games.

Conclusion A recursive injury model was applied however shoulder pain and impairment did not appear to be correlated with injury risk. Recursiveness may not be relevant to the shoulder injuries observed in this cohort. Pre-season screening and appropriate rehabilitation may prevent repeat shoulder injury.