Background: Patients diagnosed with pancreatic cancer have a poor prognosis and incidence rates have risen over the past decades. In the absence of screening programs, the only way to decrease the burden of pancreatic cancer is through primary prevention, although aetiological gaps remain. The aim of this study was to describe the association with pancreatic cancer risk for hypothesised and established risk factors.

Method: We used data from the UK Biobank (UKB) longitudinal cohort (N: 474,941, ncases: 1,273). We assessed the association of 27 hypothesized or established risk factors with pancreatic cancer using Cox proportional hazards models.

Results: We found that participants currently exposed to smoking had a 2-fold risk increase for pancreatic cancer (HR:1.90, 95% CI: 1.61-2.24) compared to participants who never smoked. Participants with recent (within the last three years) type 2 diabetes mellitus diagnosis had a 2.10-fold risk increase of pancreatic cancer (HR: 2.10, 95% CI: 1.60-2.77) compared to those without diabetes, whereas participants with long-term diabetes had a 1.60 (HR: 1.57, 95% CI: 1.23-2.01) increase in risk. Each standard deviation increment in BMI was associated with a 1.2-fold risk increase (HR:1.18, 95% CI: 1.11-1.24). Glycated haemoglobin (HbA1c) was associated with 1.10-fold risk increase of pancreatic cancer for each standard deviation increment (HR:1.09, 95% CI: 1.07-1.11). We found little evidence for associations with pancreatic cancer risk for reproductive factors, physical activity and exposure to air pollution. We will provide further evidence from complementary Mendelian Randomization and mediation analyses during the meeting. 

Conclusions: This analysis highlights the importance of some, but not all, risk factors previously implicated or hypothesised in pancreatic cancer aetiology. Specifically, confirms that smoking, body fatness, type 2 diabetes and hyperglycaemia represent important risk factors for pancreatic cancer.