Associated factors, nonlinear risk patterns and effect heterogeneity of severe injuries in recreational skiers and snowboarders

Background: Severe injuries in recreational alpine skiing and snowboarding impose disproportionate clinical and societal burden. Evidence on modifiable countermeasures beyond helmets remains fragmented and may vary across risk profiles and exposure conditions. This study aimed to identify factors associated with severe injuries among recreational skiers and snowboarders, and to examine nonlinear dose-response relationships and effect modification by preventive practices. Methods: We conducted a retrospective cross-sectional injury-severity study of injured adult skiers and snowboarders treated at resort medical clinics and emergency departments at two ski resorts in Zhangjiakou, China, across three winter seasons from 2021 to 2024. Severe injury was defined as an injury severity score (ISS) above 15. We modeled severe injury conditional on injury using Firth-penalized logistic regression. Restricted cubic splines were applied for age, body mass index, temperature, and snow depth to assess nonlinear associations. Prespecified interaction blocks were tested using joint Wald tests with false discovery rate control, and scenario-standardized impact metrics were estimated with bootstrap uncertainty. Results: = 0.011). Dose-response analyses suggested non-linear patterns across continuous variables, with statistically significant overall evidence for snow depth. Warm-up and binding tests showed stronger protection in high-risk subgroups such as beginners and risk-takers, although scenario estimates were imprecise. Conclusion: Severe outcomes among injured snow-sport participants reflect interacting exposure intensity, environment, and modifiable practices. Findings support risk-stratified prevention and condition-aware resort safety operations.