INTRODUCTION: A restrictive spirometry pattern is associated with high morbidity and mortality. Whether regular physical activity (PA) protects against this pattern has never been studied.
OBJECTIVE: To assess if PA is associated with the development of restrictive pattern.
METHODS: Lung function and PA were assessed in the second and third follow-up of the ECRHS (n=2757, 39-67 years) and SAPALDIA (n=2610, 36-82 y) cohorts. Subjects with restrictive or obstructive pattern at baseline were excluded. We assessed the association of being active at baseline (defined as exercising vigorously >2-3 times/wk for >1 h) and restrictive pattern at follow-up (defined as a post-bronchodilator FEV1/FVC≥LLN and FVC<80% pred.) using modified Poisson regression, adjusting for age, sex, smoking and asthma. We explored the impact of adjusting for baseline FVC. Additionally, models were repeated stratified by BMI.
RESULTS: After 10 years follow-up, 3.7% and 2.8% of participants developed a restrictive pattern, in ECRHS and SAPALDIA respectively. In both cohorts, being physically active was associated with lower risk of a restrictive pattern (meta-analysed RR 0.65, 95% CI 0.47-0.89). This association was stronger in overweight (0.41, 0.23-0.75) and obese (0.42, 0.17-1.05) than in normal weight subjects, but was attenuated when adjusting for baseline FVC (0.77, 0.58-1.04).
CONCLUSION: In two large European studies, adults who reported more PA were at lower risk of developing a restrictive spirometry pattern. Lung function at baseline seemed to explain part of the observed association, stressing the need of adequate method to take into account both horse-racing and regression-to-the-means effects.
