TY - JOUR
T1 - Assessing the role of ultra-processed foods in colorectal cancer incidence
T2 - insights from the EPIC cohort
AU - Berden, Jeroen
AU - Langselius, Oliver
AU - González-Gil, Esther M.
AU - Lopez, Jessica Blanco
AU - Nicolas, Geneviève
AU - Srour, Bernard
AU - Millett, Christopher
AU - Vamos, Eszter P.
AU - Rumgay, Harriet
AU - Chang, Kiara
AU - Touvier, Mathilde
AU - Heath, Alicia K.
AU - Kyrø, Cecilie
AU - Marques, Chloé
AU - Jannasch, Franziska
AU - Schulze, Matthias B.
AU - Franco, Matteo
AU - Zabala, Ana M.Jimenez
AU - Tjønneland, Anne
AU - van Gils, Carla H.
AU - Le Cornet, Charlotte
AU - Cadeau, Claire
AU - Guevara, Marcela
AU - Sanchez, Maria José
AU - Hajji-Louati, Mariem
AU - Crous-Bou, Marta
AU - Tumino, Rosario
AU - Pala, Valeria
AU - Katzke, Verena
AU - Skeie, Guri
AU - Vineis, Paolo
AU - Ferrari, Pietro
AU - Viallon, Vivian
AU - Huybrechts, Inge
N1 - Publisher Copyright:
© The Author(s) 2026.
PY - 2026/6/4
Y1 - 2026/6/4
N2 - Background: Studies have reported associations between ultra-processed food (UPF) consumption and colorectal cancer (CRC) incidence, yet the magnitude of these associations remains unclear. This study estimated the cohort-associated fraction (CAF), adapted from the population-attributable fraction metric, between UPF consumption and CRC incidence and explored the potential benefits of substituting UPFs with unprocessed or minimally processed foods. Methods: Using data from 416,081 middle-aged adults in the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort, including 5,845 incident CRC cases, Cox proportional hazards models were applied to estimate associations between UPF consumption (grams/day) and CRC, colon and rectal cancer risks. Substitution analyses were also carried out to estimate associations of replacing UPFs with unprocessed or minimally processed foods, to reflect potentially more interpretable estimates. CAFs quantified the magnitude of these associations. Results: UPF consumption was positively associated with colon cancer (HR1-SD Increment: 1.04 [95%CI: 1.01; 1.08]) and CRC (HR1-SD Increment: 1.04 [95%CI: 1.01; 1.07]) but not with rectal cancer incidence. The CAF for UPF consumption was 5.58% (95%CI: 0.55; 10.50) for colon, and 5.10% (1.10; 9.02) for CRC incidence. Substituting 1-SD of UPFs with unprocessed or minimally processed foods was inversely associated with colon cancer (HR: 0.93, [95%CI: 0.89; 0.97]) and CRC incidence (HR: 0.94, [95%CI: 0.92; 0.98]), in energy- and nutrient-adjusted models. In substitution models, the CAF increased to 9.67% (4.36; 14.85) for colon cancer and 7.30% (3.05; 11.48) for CRC incidence. Conclusion: We estimated the magnitude of associations between UPF consumption and CRC incidence, highlighting the potential public health relevance of reducing UPF intake. Given the rising consumption of UPFs and burden of CRC incidence, public health interventions and continued research are needed to address this growing issue.
AB - Background: Studies have reported associations between ultra-processed food (UPF) consumption and colorectal cancer (CRC) incidence, yet the magnitude of these associations remains unclear. This study estimated the cohort-associated fraction (CAF), adapted from the population-attributable fraction metric, between UPF consumption and CRC incidence and explored the potential benefits of substituting UPFs with unprocessed or minimally processed foods. Methods: Using data from 416,081 middle-aged adults in the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort, including 5,845 incident CRC cases, Cox proportional hazards models were applied to estimate associations between UPF consumption (grams/day) and CRC, colon and rectal cancer risks. Substitution analyses were also carried out to estimate associations of replacing UPFs with unprocessed or minimally processed foods, to reflect potentially more interpretable estimates. CAFs quantified the magnitude of these associations. Results: UPF consumption was positively associated with colon cancer (HR1-SD Increment: 1.04 [95%CI: 1.01; 1.08]) and CRC (HR1-SD Increment: 1.04 [95%CI: 1.01; 1.07]) but not with rectal cancer incidence. The CAF for UPF consumption was 5.58% (95%CI: 0.55; 10.50) for colon, and 5.10% (1.10; 9.02) for CRC incidence. Substituting 1-SD of UPFs with unprocessed or minimally processed foods was inversely associated with colon cancer (HR: 0.93, [95%CI: 0.89; 0.97]) and CRC incidence (HR: 0.94, [95%CI: 0.92; 0.98]), in energy- and nutrient-adjusted models. In substitution models, the CAF increased to 9.67% (4.36; 14.85) for colon cancer and 7.30% (3.05; 11.48) for CRC incidence. Conclusion: We estimated the magnitude of associations between UPF consumption and CRC incidence, highlighting the potential public health relevance of reducing UPF intake. Given the rising consumption of UPFs and burden of CRC incidence, public health interventions and continued research are needed to address this growing issue.
KW - Cancer epidemiology
KW - Cohort-attributable fraction (CAF)
KW - Colorectal cancer (CRC)
KW - EPIC cohort
KW - Food processing
KW - Ultra-processed foods (UPF)
UR - https://www.scopus.com/pages/publications/105041070459
U2 - 10.1186/s12889-026-27225-3
DO - 10.1186/s12889-026-27225-3
M3 - Article
C2 - 42021246
AN - SCOPUS:105041070459
SN - 1471-2458
VL - 26
JO - BMC Public Health
JF - BMC Public Health
IS - 1
M1 - 1800
ER -