The relationship between physical activity, obesity, and lung cancer risk by smoking status in a large prospective cohort of US adults

Physical activity has been associated with lower lung cancer risk in numerous studies with estimates ranging from 20 to 50% lower risk in the most versus the least active study participants. Underweight and obesity have also been associated with lower lung cancer risk, with a nonlinear, inverted U-shaped relationship. However, associations of physical activity and obesity with lung cancer are likely significantly confounded by smoking since individuals who smoke are generally less active and leaner than non-smokers, but few studies have examined these associations stratified by smoking status. Using data from 162,679 men and women who were cancer-free at enrollment (1992-1993) in the American Cancer Society Cancer Prevention Study-II Nutrition Cohort, we examined associations of baseline recreational physical activity (MET-hours per week; none, 0.1 to < 8.75 (reference), 8.75-17.4, 17.5+ MET-hours/week), baseline body mass index (BMI, weight (kg)/height (m(2)); < 18.5, 18.5-22.0 (reference), 22.1-24.9, 25.0-29.9, 30.0+ kg/m(2)), and waist circumference (measured in 1997; sex-specific quartiles) in relation to lung cancer risk stratified by smoking status and years since quitting among former smokers (never, current, former < 10 years, former, 10-19 years, former 20+ years). Cox proportional hazards modeling computed hazard rate ratios (RR) and 95% confidence intervals (CI) while adjusting for potential confounders. During 2,384,546 person years of follow-up time, 4,669 men and women were diagnosed with lung cancer (453 among never smokers; 1,452 among current smokers; 1,194 among former smokers < 10 years since quitting; 725 among former 10-19 years; and 845 among former 20+ years). Physical activity was not associated with lung cancer risk within any of the smoking strata except in former smokers less than 10 years since quitting (RR = 0.77; 95% CI 0.67-0.90 for 17.5+ MET-hours/week). Similarly, BMI was inversely associated with lung cancer in former smokers less than 10 years since quitting (RR = 0.68; 95% CI 0.55-0.84 for 30+ kg/m(2)) and more modestly in former smokers who quit 10-19 and 20+ years ago. Waist circumference was not associated with lung cancer risk in any smoking category. While being physically active and maintaining a healthy body weight are important for prevention of various chronic diseases, including several types of cancer, our results suggest that physical activity, BMI, and waist circumference are not associated with lung cancer risk, regardless of smoking status.