Results of a complete-case multilevel, multivariable logistic regression with a random effect for team, estimating the effect of mouthguards on the odds of concussion adjusted for covariates (n=513)
| Covariates | OR | 95% CI |
| Mouthguard | ||
| No mouthguard | (Ref) | |
| Mouthguard | 0.36 | (0.17 to 0.73) |
| Level of play | ||
| Lower level | (Ref) | |
| Higher level | 0.77 | (0.37 to 1.59) |
| Age group | ||
| Pee Wee | (Ref) | |
| Bantam | 0.85 | (0.26 to 2.73) |
| Midget | 0.47 | (0.12 to 1.78) |
| Position | ||
| Forward | (Ref) | |
| Defence | 1.17 | (0.74 to 1.87) |
| Goalie | 1.28 | (0.44 to 3.72) |
| Concussion history | ||
| No | (Ref) | |
| Yes | 1.44 | (0.94 to 2.23) |
| Mechanism of injury | ||
| Body checking | (Ref) | |
| Unintentional or intentional contact with player or equipment | 0.44 | (0.25 0.74) |
| Contact with environment or no contact | 0.14 | (0.07 to 0.27) |
| Cohort | ||
| 2013–2016 | (Ref) | |
| 2011–2012 | 0.77 | (0.37 to 1.59) |
| Body checking league | ||
| No | (Ref) | |
| Yes | 0.59 | (0.2 to 1.73) |
| Session type | ||
| Practice | (Ref) | |
| Game | 0.56 | (0.25 to 1.2) |