ABSTRACT
INTRODUCTION: Evidence about specific carbohydrate diet (SCD) for inflammatory bowel disease (IBD) is limited. We conducted 54 single-subject, double-crossover N-of-1 trials comparing SCD with a modified SCD (MSCD) and comparing each with the participant's baseline, usual diet (UD). METHODS: Across 19 sites, we recruited patients aged 7-18 years with IBD and active inflammation. Following a 2-week baseline (UD), patients were randomized to 1 of 2 sequences of 4 alternating 8-week SCD and MSCD periods. Outcomes included fecal calprotectin and patient-reported symptoms. We report posterior probabilities from Bayesian models comparing diets. RESULTS: Twenty-one (39%) participants completed the trial, 9 (17%) completed a single crossover, and 24 (44%) withdrew. Withdrawal or early completion occurred commonly (lack of response [n = 11], adverse events [n = 11], and not desiring to continue [n = 6]). SCD and MSCD performed similarly for most individuals. On average, there was <1% probability of a clinically meaningful difference in IBD symptoms between SCD and MSCD. The average treatment difference was -0.3 (95% credible interval -1.2, 0.75). There was no significant difference in the ratio of fecal calprotectin geometric means comparing SCD and MSCD (0.77, 95% credible interval 0.51, 1.10). Some individuals had improvement in symptoms and fecal calprotectin compared with their UD, whereas others did not. DISCUSSION: SCD and MSCD did not consistently improve symptoms or inflammation, although some individuals may have benefited. However, there are inherent difficulties in examining dietary changes that complicate study design and ultimately conclusions regarding effectiveness.
Subject(s)
Colitis, Ulcerative , Crohn Disease , Leukocyte L1 Antigen Complex , Adolescent , Bayes Theorem , Child , Colitis, Ulcerative/complications , Colitis, Ulcerative/diet therapy , Crohn Disease/complications , Crohn Disease/diet therapy , Diet , Feces/chemistry , Humans , Inflammation/complications , Inflammation/diet therapy , Inflammatory Bowel Diseases/complications , Inflammatory Bowel Diseases/diet therapy , Leukocyte L1 Antigen Complex/analysis , Precision MedicineABSTRACT
Inflammatory bowel disease (IBD) affects 3 million children and adults in the US. Treatment involves medications with considerable risk profiles. Dietary modification, such as the specific carbohydrate diet (SCD), may be helpful in treating IBD, but there is insufficient evidence of its effectiveness. N-of-1 trials are ideal for addressing this important research question. The Personalized Research on Diet in Ulcerative Colitis and Crohn's Disease (PRODUCE) study employs a series of 50 individual N-of-1 trials that compare the SCD to a modified SCD. Treatment periods are assigned in blocks of two, with each patient completing two balanced treatment blocks. Patients are randomized to start with the SCD or modified SCD and alternate between conditions for four eight-week periods. A mobile app guides collecting and viewing data, transitioning diets, and reviewing personal results. Primary outcomes include patient reported outcomes (PROs) of stool frequency, stool consistency, pain interference, and gastrointestinal (GI) symptom severity. We examine changes in inflammation via fecal calprotectin. Participants will receive a personalized answer regarding comparative effectiveness between the SCD and a less restrictive diet option (modified SCD), as well as compared to their baseline diet. We will aggregate the results of completed N-of-1 trials across patients to estimate population level comparative effectiveness of these treatments and the effectiveness of each diet.
