LAUSR

It is not known how physical exercise affects the risk of different types of fractures, especially in highly active individuals. To investigate this association, we studied a cohort of 118,204 men and 71,757 women who from 1991 to 2009 participated in Vasaloppet, a long‐distance cross‐country skiing race in Sweden, and 505,194 nonparticipants frequency‐matched on sex, age, and county of residence from the Swedish population. Participants ranged from recreational exercisers to world‐class skiers. Race participation, distance of race run, number of races participated in, and finishing time were used as proxies for physical exercise. Incident fractures from 1991 to 2010 were obtained from national Swedish registers. Over a median follow‐up of 8.9 years, 53,175 fractures of any type, 2929 hip, 3107 proximal humerus, 11,875 lower leg, 11,733 forearm, and 2391 vertebral fractures occurred. In a Cox proportional hazard regression analysis using time‐updated exposure and covariate information, participation in the race was associated with an increased risk of any type of fracture (hazard ratio [HR], 1.02; 95% CI, 1.00 to 1.05); forearm fractures had an HR, 1.11 with a 95% CI, 1.06 to 1.15. There was a lower risk of hip (HR, 0.75; 95% CI, 0.67 to 0.83), proximal humerus (HR, 0.90; 95% CI, 0.82 to 0.98), and lower leg fractures (HR, 0.93; 95% CI, 0.89 to 0.97), whereas the HR of vertebral fracture was 0.97 with a 95% CI, 0.88 to 1.07. Among participants, the risk of fracture was similar irrespective of race distance and number of races run. Participants close to the median finishing time had a lower risk of fracture compared with faster and slower participants. In summary, high levels of physical exercise were associated with a slightly higher risk of fractures of any type, including forearm fractures, but a lower risk of hip, proximal humerus, and lower leg fractures. © 2018 American Society for Bone and Mineral Research.