Article
作者: Keski-Rahkonen, Pekka ; Gunter, Marc ; Ferrari, Pietro ; Freisling, Heinz ; Sánchez, Maria-Jose ; Robinot, Nivonirina ; Delpierre, Cyrille ; Sieri, Sabrina ; Nøst, Therese Haugdahl ; Tsilidis, Kostas ; Borch, Kristin Benjaminsen ; Tjønneland, Anne ; Braaten, Tonje ; Fordellone, Mario ; Dahm, Christina C ; Tumino, Rosario ; Ricceri, Fulvio ; Wedekind, Roland ; Matta, Komodo ; Schulze, Matthias B ; Fortner, Renée Turzanski ; Trobajo-Sanmartín, Camino ; Colorado-Yohar, Sandra ; Weiderpass, Elisabete ; Harewood, Rhea ; Johansson, Mattias ; Viallon, Vivian ; Vaccarella, Salvatore ; Lamy, Sebastien ; Mokoroa-Carollo, Olatz ; Marques, Chloé ; Rostgaard-Hansen, Agnetha Linn ; Truong, Therese ; Frenoy, Pauline ; Riboli, Elio ; Chatziioannou, Anastasia Chrysovalantou ; Kaaks, Rudolf ; Papier, Keren
BACKGROUND:Colon cancer is strongly influenced by lifestyle factors. Sociodemographic factors like sex and socioeconomic position (SEP) might modulate the relationship between lifestyle and colon cancer risk. Metabolomics offers potential to uncover biological mechanisms linking lifestyle and colon cancer.
METHODS:Lifestyle and untargeted metabolomic data were available from a nested case-control study within the European Prospective Investigation into Cancer and Nutrition (EPIC), including 1,067 colon cancer cases and 1,067 controls matched on age, sex, study centre, and blood collection time. Serum samples were analyzed using liquid chromatography-mass spectrometry. The Healthy Lifestyle Index (HLI) score was derived from smoking habits, alcohol intake, body mass index (BMI), physical activity, and diet. Penalised regression was applied in controls to derive metabolic signatures for the HLI and the lifestyle components. Associations of lifestyle factors and the metabolic signatures with colon cancer risk were estimated in conditional logistic regression models, overall and by sex and SEP.
RESULTS:The HLI score was inversely associated with colon cancer risk, with an odds ratio (OR) per 1-standard deviation (SD) increment equal to 0.79; 95% CI: 0.71, 0.87. The metabolic signature of HLI, comprising 130 features, was moderately correlated with HLI (r = 0.59; 94% CI: 0.56, 0.61), and was inversely associated with colon cancer risk (OR = 0.86; 95% CI: 0.78, 0.95). After adjustment for the HLI score, the association of the metabolic signature of HLI and colon cancer risk was null (OR = 1.00, 95% CI 0.88, 1.13). Associations of lifestyle factors and the metabolic signature with colon cancer risk were consistently stronger for men than for women and did not differ by SEP.
CONCLUSIONS:In this study across seven European countries, healthy lifestyle was inversely associated with colon cancer risk, with stronger associations in men than women and no differences across SEP. However, the serum metabolic signatures after adjustment for lifestyle factors were not found to be associated with colon cancer risk, suggesting that lifestyle impacts colon cancer through mechanisms not captured by the signatures.