Changes in physical activity and mortality risk among Korean cancer survivors: a population-based cohort study.

Background The impact of changes in physical activity (PA) post-cancer diagnosis on prognosis remains unclear. This study evaluated mortality risks according to changes in PA from pre-diagnosis to post-diagnosis among cancer survivors. Methods This population-based retrospective cohort study used the Korean National Health Insurance Service database. The study included 215,191 participants (125,756 men and 89,435 women) diagnosed with cancer between 2009-2017. PA, measured as the total of various light-, moderate-, and vigorous- intensity activities, was assessed pre- and post-diagnosis. Deaths were ascertained between 2009-2019. All-cause and cancer-specific mortality risks were assessed according to PA changes using Cox proportional hazards regression. Results Post-cancer diagnosis, active patients accounted for 63.30% of men and 55.29% of women, increasing from 54.04% and 43.35% pre-diagnosis. Compared with the consistently inactive group, all-cause mortality risks were significantly lower in patients who became active post-diagnosis (adjusted hazard ratio [95% confidence intervals]: men, 0.82 [0.79, 0.85]; women, 0.87 [0.82, 0.93]) and in the consistently active group (men, 0.77 [0.74, 0.80]; women, 0.81 [0.76, 0.86]). Lower mortality risks were observed across cancer stages in men and localized/regional stages in women. PA and all-cause mortality had a dose-response association. PA was inversely associated with all-cause or cancer-specific mortality in men with gastric, colorectal, liver, and lung cancers and women with colorectal cancer. Conclusions Being physically active post-diagnosis is associated with reduced all-cause mortality among cancer survivors in a dose-response manner, regardless of pre-diagnosis PA levels. Impact: PA should be promoted as a standard component of cancer care to improve prognosis.