P628 Processed and ultra-processed food and risk of Inflammatory Bowel Disease in the European Prospective Investigation into Cancer and Nutrition cohort (EPIC-IBD)

Dong, C.(1,2);Casagrande, C.(3);Chan, S.(4,5);Huybrechts, I.(3);Nicolas, G.(3);Boutron-Ruault, M.C.(2);Carbonnel, F.(1,2);

(1)University Hospital of Bicêtre- Assistance Publique-Hôpitaux de Paris and University Paris-Saclay, Department of Gastroenterology, Le Kremlin Bicêtre, France;(2)Institut Gustave Roussy-University Paris Sud, INSERM- Centre for Research in Epidemiology and Population Health- U1018- Team 9, Villejuif, France;(3)World Health Organization, International Agency for Research on Cancer, Lyon, France;(4)Norfolk and Norwich University Hospital NHS Trust, Department of Gastroenterology, Norwich, United Kingdom;(5)University of East Anglia, Department of Medicine, Norwich, United Kingdom;


During the past decades, human diet has evolved towards higher intakes of processed and ultraprocessed food. We have investigated the association between these food groups and the risk of inflammatory bowel disease (IBD) in the European Prospective Investigation into Cancer and Nutrition.


413 590 participants from 8 European countries were included. Dietary data were collected at baseline from validated food frequency questionnaires. All EPIC food items were expressed as g/day and categorized according to the NOVA classification: “group 1—unprocessed or minimally processed foods”; “group 2—processed culinary ingredients”; “group 3—processed foods”; and “group 4—ultra-processed foods (UPFs)”. Because the EPIC dietary questionnaires were conceived before the NOVA classification, the food items were retrospectively categorized into this classification. Some food items collected in the EPIC questionnaires were difficult to categorize into group 3 or 4. We therefore merged them into a single category. We tested three scenarios for the estimation of the dietary content of this merged category: lower, middle and upper contents of groups 3 + 4. The association between the proportion of each NOVA group in the diet and IBD were estimated using Cox proportional hazard models to obtain Hazards Ratios (HRs) and 95% confidence intervals. We adjusted the HR for smoking status, educational level, physical activity, BMI, alcohol consumption and energy intake and stratified by centre, age at baseline (1-y interval), and sex.


After a mean follow-up of 16 years, 179 Crohn's disease (CD) and 431 ulcerative colitis (UC) cases were identified. NOVA group 1 was negatively associated with CD risk (adjusted HR for the fourth vs. the first quartile = 0.58; 95% CI = 0.35-0.95; P-trend = 0.03). Within group 1, fruit intake was the only food item to be negatively associated with the risk of IBD (adjusted HR for the fourth vs. the first quartile = 0.42; 95% CI = 0.24-0.73; P-trend = 0.003). We found numerical associations between group 3+4 and the risk of CD with the middle or upper scenarios (adjusted HRs for the fourth vs. the first quartile = 1.50; 95% CI = 0.92-2.46; P-trend = 0.12 and 1.41; 95% CI = 0.88-2.27; P-trend = 0.09 respectively). There was no association between any category of the NOVA classification and UC risk.


In the EPIC cohort, consumption of non-processed food was associated with a lower risk of CD while consumptions of processed or ultra-processed food were numerically associated with increased risk of CD. We found no association with UC.