Risk of pancreatic cancer and non-Hodgkin lymphoma associated with healthy lifestyle behaviors in the EPIC study

Abstract

Introduction Pancreatic cancer (PC) and non-Hodgkin lymphoma (NHL) are both rare cancers which are particularly prevalent in Western countries and for which increasing incidence rates have been observed for the past decades. Moreover, both of them have unclear etiology and only few risk factors identified so far. While PC has been firmly associated with cigarette smoking, type II diabetes, obesity, as well as chronic pancreatitis, NHL has been mainly related to immune disorders and occupational exposure like benzenes. Roles of lifestyle factors in PC and NHL etiologies are still largely undefined, but recent evidences suggested that lifestyle, as well as high levels of alcohol consumption may be related to PC and NHL risks. Therefore, associations between modifiable lifestyle factors and either PC risk or NHL risk were investigated using an index expressing healthy lifestyle behaviors in the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Methods Within 14 years of follow-up, 1113 incident PC cases and 2244 incident NHL cases (1845 B-NHL, 372 DLCBL, 415 CLL/SLL, 278 FL, and 535 PCN/MM) were diagnosed in the EPIC study. A score named the Healthy Lifestyle Index (HLI) ranging from 0 to 20 was built combining lifestyle factors assessed at baseline, i.e. a score for diet, physical activity, smoking habits, alcohol intake, and BMI. A score of 20 corresponding to the highest level of adherence to primary prevention recommendations was meant to reflect a healthy lifestyle. HLI was modelled in categories ( 15) and in continuous for a 1-category increase. Cox proportional hazard models with age as primary time variable were used to estimate PC and NHL hazard ratios (HR) and their 95% confidence intervals (95% CI) in relation to the HLI, adjusting for potential confounding factors. Results The combination of healthy lifestyle behaviors into an increasing HLI was inversely associated with PC with a 29% (95% CI: 0.64, 0.78) decreasing risk for a 1-category increase of HLI among EPIC participants. Modifiable lifestyle factors driving PC risk's reduction were smoking abstinence (pWald Conclusions Findings from the EPIC study suggested that the risk of PC was inversely associated with healthy lifestyle profiles, whereas no association was found in relation to NHL risk among EPIC participants. The results support adhering to primary prevention guidelines in order to decrease PC incidence and should encourage policy makers to promote healthy lifestyle in general population, in particular for cancers with no available screening strategy. In addition, they provide more evidence in favor of investigating non-modifiable risk factors to further understanding NHL's etiology.