LAUSR

Objective To determine the association between pre-diagnostic recreational physical inactivity (RPI) and pancreatic cancer (PC) mortality. Methods This analysis included 107 patients seen at Roswell Park Comprehensive Cancer Center diagnosed with PC between 1989 and 1998. Cox proportional hazards models were used to determine hazard ratios (HR) and 95% confidence intervals (CI) for PC mortality associated with self-reported pre-diagnostic RPI. Models were adjusted for known prognostic factors, including age, sex, stage at diagnosis, smoking status, and body mass index (BMI). Results were also stratified by sex, BMI, smoking status, histology, and treatment status. Results We observed a significant association between RPI and PC mortality in all patients (HR = 1.72, 95% CI = 1.06–2.79), as well as among overweight or obese patients (HR = 2.74, 95% 95% CI = 1.42–5.29), females (HR = 2.63; 95% CI, 1.08–6.39), and non-smokers (HR = 1.72; 95% CI, 1.02–2.89). Conclusion These results suggest that RPI prior to PC diagnosis is associated with a higher risk of death. Future studies with larger sample sizes are needed to explore whether this association varies across tumor histology.