P1730 - Comparison of Dietary Patterns in IBD Patients between the Northern and Southern Provinces of the Netherlands

Introduction Evidence-based dietary guidelines are not available for patients with Inflammatory Bowel Disease (IBD). Therefore, patients tend to follow their own “unguided” dietary habits, often leading to unbalanced intakes that may further impact disease course. Hence, we aimed to identify dietary patterns in post-diagnosis habitual dietary intake in 2 separate cohorts of Dutch IBD outpatients. Aims & Methods 489 IBD patients (286 Crohn's Disease (CD), 203 Ulcerative Colitis (UC)) from Groningen and 236 IBD patients (154 CD, 82 UC) from Maastricht with data on demographics and disease phenotype were included. Dietary intake, including nutrients and 22 food groups, was obtained via semi-validated food frequency questionnaires. Dietary differences in macronutrients and food groups, and baseline characteristics were analysed between the cohorts using a students' t-test or X2-test when appropriate. A preliminary principal-components analysis (PCA) was conducted on 22 food groups to identify relevant dietary patterns. Results Compared to the Maastricht cohort, patients in the Groningen cohort had a lower age at inclusion and diagnosis, disease duration, and less men (all p< 0.05); no differences were observed in phenotype, BMI or smoking. The total energy intake including energy from all separate macronutrients was significantly lower in the Groningen cohort. Moreover, patients in Groningen consumed less legumes, grains, red and processed meat, poultry, fish, confectionery, coffee, oils, alcoholic beverages and condiments, but more potatoes, dairy, non-alcoholic beverages and prepared meals than patients in Maastricht. First, a PCA was run in the Groningen cohort to identify dietary patterns. This PCA, with an overall Kaiser-Meyer-Olkin (KMO) measure of 0.624 revealed 8 components with eigenvalue >1. Visual inspection of the scree plot (Cattell, 1966) and interpretability criteria, indicated that two components should be retained, explaining 22.0% of the total variance. The interpretation of the data was in line with the western and prudent components as described before (Stricker, 2013). Condiments, grain products, potatoes, oils, processed and red meat, snacks and confectionery contributed to a western component; whereas fruits, vegetables, fish, tea, eggs and nuts loaded positively, and snacks and non-alcoholic beverages loaded negatively on the prudent component. To confirm the above patterns, a PCA was also run on the Maastricht data, with an overall KMO of 0.602 and revealing 9 components with an eigenvalue >1. Here, visual inspection of the scree plot and interpretability criteria also indicated that two components should be retained. These explained 21.8% of the total variance and were also consistent with the western and prudent component. The same food groups loaded on the western component and prudent component, except for snacks and fruits to the prudent component. This might be due to smaller sample size in the Maastricht cohort. Conclusion Our study shows that IBD patients in the Groningen cohort mainly use a western or prudent-based diet after diagnosis. These findings were confirmed in the geographical distinct Maastricht cohort; as mainly the same patterns were extracted. Since, adapting a western dietary pattern may contribute to potentially unintended effects on disease course, dietary intervention in these patients might be beneficial. References Abstract: V. Peters and C. Spooren are shared first author, D.M. Jonkers and M.J.E. Campmans-Kuijpers are shared last author. Ref: Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276. Stricker, M.D. (2013). Dietary patterns derived from principal component- and k-means cluster analysis: Long-term association with coronary heart disease and stroke. Nutrition, Metabolism & Cardiovascular Diseases, 23, 250e256.