Outcome variable compared with control Matched odds ratio (95% CI)* Partially adjusted matched odds ratio (95% CI) Ideal matched odds ratio (95% CI) Final adjusted matched odds ratio (95% CI) Any head injury 0.81 (0.64 to 1.02) 0.78 (0.61 to 1.0) 0.73 (0.49 to 1.08) 0.71 (0.55 to 0.92) Potentially severe head injury¶ 0.67 (0.40 to 1.11) 0.59 (0.34 to 1.0) — 0.44** (0.24 to 0.81) Any neck injury 1.11 (0.67 to 1.83) 0.96 (0.56 to 1.66) — 0.62 (0.33 to 1.19) Potentially severe neck injury¶ 1.29 (0.41 to 4.04) — — —

Age=<15, 15-25, 26; days of participation that season to day of injury=1, 2-10, 11)

^* Helmet use only.

Helmet use, age, and sex; adjusted for activity through matching.

Adjusted for age, sex, ability, seasons of participation, days of participation that season to day of injury, lessons (yes, no), education (high school or less, college or professional diploma, university or graduate degree), previous head or neck injury, hours of participation on day of injury (1, 2-5, 6), damage to non-helmet equipment (yes, no), self reported speed in relation to average (slower, average, faster), mechanism of injury (collision or jump, fall), type of participation on day of injury (recreation, lessons, or school outing), run difficulty (easy, difficult, very difficult to extreme), other protective equipment besides helmet (yes, no), visibility (good, average to fair), snow conditions (groomed to hard pack ice, other), temperature (<–10°C, –10°C to –1°C, 0°C); adjusted for activity through matching; interaction of age and helmet use (<15 and 15-25) initially tested in full model; age and sex forced into model.

Adjusted for age, sex, days of participation that season to day of injury adjusted for activity through matching; age and sex forced into model; backward deletion modelling strategy.

^¶ Evacuated by ambulance.

^** Adjusted for age, sex, days of participation that season to day of injury, and wearing other protective equipment besides helmet at time of injury (yes, no); age and sex forced into model; forward selection modelling strategy.

Adjusted for age, sex, and days of participation that season to day of injury; age and sex forced into model; forward selection modelling strategy.