Physical activity, sedentary behaviour, and exocrine and endocrine pancreatic diseases: results from a prospective cohort study and Mendelian randomization

Methods

The prospective study included 437,131 UK Biobank participants free of pancreatic diseases at baseline. We used moderate-to-vigorous physical activity (MVPA) to indicate physical activity and leisure screen time (LST) to indicate sedentary behaviour. Outcomes included five diseases related to pancreatic secretory functions.

We conducted Cox proportional hazards regression and mediation analyses to examine the mediating role of body mass index (BMI). MR analysis was used to assess the robustness of the observational findings.ResultsDuring a mean follow-up of 14 years, we identified 2,548 acute pancreatitis (AP), 638 chronic pancreatitis (CP), 1,711 pancreatic cancer (PC), 1,705 type 1 diabetes mellitus (T1DM), and 19,371 type 2 diabetes mellitus (T2DM) cases.

Increasing MVPA was associated with AP, CP, T1DM, and T2DM risk in a U-shaped manner, with the lowest risk at ∼3,000 metabolic equivalent tasks-minutes/week (all P-nonlinearity < 0.05). BMI partially mediated associations of MVPA with AP, T1DM, and T2DM. LST was associated with an increased risk of AP, CP, PC, T1DM, and T2DM, with significant linear trends for CP, PC, and T1DM (all P-nonlinearity > 0.05).

BMI partially mediated the associations of LST with AP, PC, T1DM, and T2DM. MR confirmed the associations of MVPA with AP and CP and of LST with AP, CP, and T2DM.ConclusionModerate MVPA and strict restriction of LST can serve as effective and practical strategies for preventing AP, CP, and T2DM.