BACKGROUND Physical activity (PA) is important for improving life expectancy and is suggested as a prognostic factor for various diseases. However, the association between PA and mortality outcomes in survivors of lung cancer remains unclear. Therefore, this study aimed to determine the association of PA levels, including changes before and after diagnosis, with mortality outcomes among survivors of lung cancer.

METHODS We conducted a nationwide cohort study involving 23 257 individuals diagnosed with lung cancer between 1 January 2010 and 31 December 2016, who attended the National Health Screening Program within 2 years before and after diagnosis. Pre- and postdiagnosis leisure-time PA levels and changes in PA were assessed using self-administered questionnaires. Individuals who reported engaging in moderate-intensity exercise ≥ 5 days/week or vigorous-intensity exercise ≥ 3 days/week were classified as physically active. The amount of PA was calculated as the metabolic equivalent of task (MET)-min/week for each individual. All participants were followed from the date of diagnosis to 31 December 2022, for the outcome of mortality.

RESULTS During the follow-up period of 165 344.0 person-years, 9094 deaths occurred (6633 lung cancer-specific and 2461 non-lung cancer deaths). Multivariable analyses revealed that both pre- and postdiagnosis PA were associated with significantly reduced risk of all-cause (aHR = 0.92, 95% CI = 0.88-0.97 for prediagnosis and aHR = 0.85, 95% CI = 0.81-0.89 for postdiagnosis) and lung cancer-specific (aHR = 0.93, 95% CI = 0.88-0.99 for prediagnosis and aHR = 0.89, 95% CI = 0.84-0.94 for postdiagnosis) mortality compared with inactivity. Significant dose-response relationships were observed between PA levels and mortality risk reduction. Compared with individuals who were consistently inactive before and after diagnosis, significant mortality risk reduction was seen in those who maintained PA (aHR = 0.77, 95% CI = 0.71-0.83 for all-cause and aHR = 0.81, 95% CI = 0.75-0.89 for lung cancer-specific mortality) and those who promoted PA after lung cancer diagnosis (aHR = 0.91, 95% CI = 0.85-0.96 for all-cause and aHR = 0.94, 95% CI = 0.88-1.00 for lung cancer-specific mortality). However, individuals who were active before diagnosis but became inactive after diagnosis showed no significant difference in survival outcomes compared with those who were consistently inactive.

CONCLUSIONS Both pre- and postdiagnosis PA are associated with reduced mortality in survivors of lung cancer in a dose-response manner. Maintenance and promotion of PA after diagnosis are key to achieving optimal benefits in overall survival.