Physical activity and risk for lung cancer in a Danish cohort.

Physical activity might reduce the risk for lung cancer by various mechanisms, but the evidence is inconclusive. We therefore examined the relationship between physical activity and risk for lung cancer in a large population-based Danish cohort with detailed information about number of hours per week spent on specific physical activities as well as lifetime smoking patterns. Between 1993 and 1997, a total of 57,053 persons aged 50-64 years agreed to participate in the cohort. After exclusions of persons with cancer diagnosis before invitation and persons for whom data regarding study variables were missing, 26,070 men and 28,352 women remained for study. By 31 December 2002, lung cancer had been diagnosed in 194 men and 175 women. A questionnaire registered average number of hours per week spent on each of 6 types of leisure time physical activity. Level of occupational physical activity was registered in 5 categories. Cox's proportional hazard model stratified according to age at entry (1-year intervals) was adjusted for smoking, school education, possible occupational exposure to lung carcinogens and intake of fruit and vegetables. No significant association was found between number of hours per week spent on 6 types of physical activity during leisure time and the incidence rate ratio (IRR) for lung cancer. For each type of activity, the IRR of lung cancer was lower for active compared to nonactive women, whereas for men lower IRRs were only observed for sports and gardening. Higher levels of occupational physical activity had no protective effect; the lowest IRR was found for sitting work. Our study shows no convincing protective effect of physical activity on lung cancer risk.