Objective: We studied mortality during long-distance ski races in Sweden.
Design: The cohort comprised all subjects who participated in ski races during the Vasaloppet week from 1970 through 2005. All deaths during the races were identified and compared with the corresponding expected numbers based on national death rates.
Results: In total, 698 102 starters generated altogether 581 person-years of skiing. Overall, 13 deaths occurred compared with 1.68 expected during the skiing time, yielding a standardized mortality ratio (SMR) of 7.7 [95% confidence interval (CI) 4.1-13.2]. Twelve of the deaths were caused by cardiovascular disease (SMR=22.1, 95% CI 11.4-38.6). Interviews of relatives revealed that two to four victims had medical conditions of the sort that they probably would have been recommended to avoid participation for medical reasons had they consulted a physician.
Conclusion: Our study showed a clearly increased risk of acute mortality during long-distance skiing. However, as physically active people spend only a minority of their time in races, in that population the short-term excess mortality during these periods is by far outweighed by the long-term protective effects of exercise on mortality.