Abstract
OBJECTIVE: This study examines the dose-response relationship between physical activity (PA) and all-cause mortality across different severities of airflow limitation, identifying threshold effects that yield new insights into the PA-mortality association. DESIGN: A prospective cohort study with a 5-year follow-up (2018-2023), employing multivariate Cox models and penalized spline smoothing to assess non-linear associations. SUBJECTS/PATIENTS: A total of 2,975 individuals from a cohort categorized by airflow limitation severity (normal, GOLD 1-4). METHODS: PA levels were quantified in metabolic equivalent hours per week (MET·h/week). Cox proportional hazards models were used to evaluate PA-mortality associations, with penalized spline analysis detecting threshold effects. RESULTS: Identified thresholds were 41.50 MET·h/week (95% CI: 23.03-64.22) for normal lung function and 13.21 MET·h/week (95% CI: 9.67-16.14) for GOLD 1. Below these thresholds, higher PA levels were associated with a significant reduction in mortality risk (HR = 0.66, HR = 0.41, respectively). In GOLD 2, PA levels below the threshold were associated with a lower mortality risk (HR=0.85), whereas PA exceeding the threshold was associated with a higher mortality risk (HR = 1.23). No significant associations were observed in GOLD 3-4. CONCLUSION: PA demonstrates a non-linear, threshold-dependent association with mortality. These findings underscore the importance of individualized PA recommendations for optimizing health outcomes in individuals with chronic respiratory conditions.