Merrifield HH, Cowan RFJ (1973) Groin strain injuries in ice hockey. A disparity in muscle strength between both hip joint adductor muscle groups was found to be a contributing factor in groin strain injuries. Am J Sports Med 1: 41-42

Purpose: The objective of this cohort study was to determine the level of off-season sport specific activity, peak isometric adductor torque, and hip abduction flexibility that are predictive of groin or abdominal strain injury in the National Hockey League (NHL).

Methods: The subjects were 1292 consenting NHL players. Estimated relative risks of injury are reported using the following exposures: 1) level of sport specific training in the off-season, 2) peak isometric adductor torque, 3) total hip abduction flexibility, 4) previous injury, 5) years of NHL experience, and 6) skate blade hollow measurement. Estimates of probability of injury are predicted for various levels of exposures on the basis of logistic regression analysis.

Results: During training camp, players who reported less than 18 sessions sport specific training in the off-season were at greater than three times the risk of injury than those who did not (relative risk (RR); 3.38 95% confidence interval (CI), 1.45-7.92). Players who reported previous history of this injury were at more than two times the risk of injury than those who did not (RR, 2.88; 95% CI, 1.33-6.26). Veterans were at greater than five times the risk of injury than rookies (RR, 5.69; 95% CI, 2.05-15.85). Peak isometric adductor torque, total abduction flexibility, and skate blade hollow measurement were not predictive of injury. There is evidence of a dose-response gradient as predicted probability of injury decreases with increasing levels of sport specific training. In the regular season, sport specific training was not as strong a risk factor (RR, 2.32; 95% CI, 1.0-5.39).