Dietary Patterns and Their Association With Symptoms Activity in Inflammatory Bowel Diseases

Background Existing studies on diet and inflammatory bowel disease (IBD) have largely focused on evaluating the effects of single nutrients or whole predesigned diets but not on evaluating the effects of diverse dietary patterns. This study applied unsupervised methods to identify dietary patterns of individuals with IBD and evaluated their association with symptoms activity. Methods This retrospective study of adults with IBD collected current clinical data and typical diet recalled from the time when in clinical remission. Discrete dietary structures were defined by k-means clustering. Multivariable logistic regression evaluated the relationship between diet clusters and the presence of active symptoms, while adjusting for age, sex, disease duration, disease behavior, and medication use. Results Of 691 participants, 36% had Crohn's disease (CD) and 64% had ulcerative colitis (UC) or IBD-unclassified. Five major dietary clusters were identified: 2 resembled a Western diet (WD) (WD1, WD2), 1 resembled a balanced diet, and 2 resembled a plant-based diet (PB) (PB1, PB2). Compared with WD1, PB2 was associated with lower odds of active symptoms for CD (odds ratio [OR], 0.32; 95% confidence interval [CI], 0.12-0.83) and UC (OR, 0.31; 95% CI, 0.15-0.62). PB1 was associated with lower odds of active symptoms for participants with UC (OR, 0.45; 95% CI, 0.23-0.90) but not for participants with CD (OR, 0.95; 95% CI, 0.36-2.51). Conclusions Diets with increased intake of fruits and vegetables, reduction of processed meats and refined carbohydrates, and preference of water for hydration were associated with lower risk of active symptoms with IBD, although increased intake of fruits and vegetables alone did not reduce risk of symptoms with CD. Lay Summary The study used machine learning methods to provide minimally biased classifications of dietary patterns among individuals with inflammatory bowel diseases, followed by an evaluation of the association between the different diet clusters and symptoms activity.