Metabolic factors and the risk of small intestine cancers : pooled study of 800 000 individuals in the Metabolic syndrome and Cancer project

To explore the largely unknown etiology of small intestine cancer, we examined metabolic factors and risk of small intestine cancer overall and by subtypes. Amongst 404 220 women and 403 265 men in six European cohorts, we applied Cox regression with adjustment for smoking and body mass index (BMI), to calculate sex-specific hazard ratios (HR) of small intestine cancer by levels of BMI, mean arterial pressure (MAP), and plasma total cholesterol, triglycerides and glucose. We also calculated HRs for these factors combined (metabolic score; MetS) and used Wald test statistics to investigate pairwise interactions between metabolic factors on risk. We also performed analyses separately per subtype (neuroendocrine tumors (NETs) and adenocarcinomas). During a median follow-up of 16.9 years, 144 women and 195 men were diagnosed with small intestine cancer, including 184 NETs and 99 adenocarcinomas. Among men, no main associations or interactions between metabolic factors were observed in relation to the risk of small intestine cancer. Among women, triglycerides were positively and linearly associated with risk (HR per standard deviation [SD]: 1.23, 95% CI 1.04 to 1.46), and a positive association was also observed for the MetS (HR per SD: 1.25, 95% CI 1.02 to 1.52). Positive interactions were observed among women between triglycerides and cholesterol (p=0.0005), and between MAP and glucose (p=0.009), on risk. Glucose was positively associated with adenocarcinomas among women. This large, prospective study suggests that elevated triglycerides, and metabolic factors in interaction, confer an increased risk of small intestine cancer among women, but not among men.