TABLE 1.

Summary of dietary treatments for IBD tested in human trials

Diet(s) tested (reference)	Population tested	Study design	No. of subjects included	Duration of the intervention	Outcomesa
EEN (polymeric diet) vs corticosteroids (73)	Children with active, naive CD	Prospective, randomized, open-label trial	n = 37, 19 on the polymeric diet and 18 on corticosteroids	10 wk	Remission in 79% of patients on EEN group compared to 67% of patients on corticosteroid group. Mucosa healing was significantly higher in the EEN group, 74%, than in the corticosteriod group, 33%.
EEN (polymeric diet, semi-elemental diet, and elemental diet) vs corticosteroids (74)	Children with newly diagnosed active CD	Retrospective	n = 47, 37 on EEN (12 polymeric; 13 semi-elemental; 12 elemental diet) and 10 on corticosteroids	8 wk	Remission in 86.5% patients on EEN vs 90% treated with corticosteroids. Mucosa healing was higher in the EEN group, 64.8%, than in the corticosteroid group, 40%. Compared to group receiving corticosteroids, the duration of clinical remission was longer in the EEN groups, without differences among the three different formulas.
EEN (79)	Children with active CD	Prospective, nonblinded observational case study with both groups receiving the intervention	n = 44, 23 patients with CD and 21 controls	8 wk	Remission in 62% of patients with CD. Reduction of bacterial diversity. Reduction of Bifidobacterium, Ruminococcus, and Faecalibacterium.
EEN (80)	Children newly diagnosed with CD	Prospective, nonblinded observational case study with only CD patients receiving the intervention. Stool sample collected prior to, during, and after EEN.	n = 43	6 wk	During the intervention, there was a decrease in microbiota diversity and a reduction of amino acids. Also, an increase in microbial metabolism of bile acids. Prior to EEN, microbiota and metabolome are different between responders and nonresponders
SCD (41)	Children with CD (PCDAI > 10)	Retrospective	n = 7 patients with CD	5–30 mo	All symptoms resolved 3 mo after SCD. Serum albumin, C-reactive protein, hematocrit, and stool calprotectin either normalized or significantly improved during follow-up clinic visits.
SCD vs mSCD vs whole foods (36)	Children with CD, mild to moderate	Randomized trial	n = 14, 5 on the SCD, 5 on the mSCD, and 4 on the whole food diet. Only 10 completed the study.	12 wk	100% of patients on any of the diets achieved remission. At wk 12, 100% of participants who had elevated CRP at enrollment (n = 8) and completed the study had normal CRP, and 80% (n = 8 out of 10) of participants had a decrease in ESR.
CD-TREAT (42)	Children with CD, mild to moderate active luminal disease	Prospective, open-label trial	n = 5 (only 4 completed the study)	4 wk, with additional 4 wk (follow-up)	3 out of 4 children achieved clinical response, 2 achieved remission (4 wk). All patients achieved clinical response, 3 achieved remission (8 wk).
CDED+PEN or CDED alone (44)	Children and young adults with active disease defined by a pediatric Crohn's disease activity index of >7.5 or Harvey-Bradshaw index of ≥4	Prospective, open-label trial	n = 47 (7 used CDED without PEN)	6 wk	Response and remission were obtained in 37 (78.7%) and 33 (70.2%) patients, respectively. Remission was obtained in 70% of children and 69% of adults. Normalization of previously elevated CRP occurred in 21 of 30 (70%) patients in remission.
CDED+PEN vs EEN (34)	Children with CD with short duration of mild to moderate activity, mostly naive to treatment and with small bowel involvement (noninflammatory stricture or resection)	Randomized, nonblinded	n = 74. Group 1: 40 received CDED + 50% PEN for 6 wk (stage 1), followed by CDED + 25% PEN from wk 7 to 12 (stage 2). Group 2: 34 received EEN in stage 1, followed by a free diet with 25% PEN in stage 2.	12 wk	At wk 12, in group 1, 75.5% achieved remission, and of those, 75.9% had a normal CRP, 87.5% sustained remission, and microbiome exhibited dominance of Clostridia and decrease of Proteobacteria. At wk 12, in group 2, 45.1% achieved remission, and of those, 47.6% had a normal CRP, 56% sustained remission, and microbiome reverted to the pretreatment composition. In both groups, fecal calprotectin levels dropped significantly.
SCD vs Mediterranean diet (43)	Adult patients with CD, mild to moderate	Randomized trial	n = 191, 99 on the SCD and 92 on the MD	12 wk	At 6 wk, 47% on the SCD and 44% on the MD achieved symptomatic remission with up to 35% showing reduction of fecal calprotectin levels. At wk 12, 42% and 40% on the SCD and MD, respectively, achieved or maintained symptomatic remission. Fecal calprotectin response was observed only in 26% and 8% of patients on the SCD and MD, respectively. No significant change in alpha diversity was observed. In both groups, reduction of Faecalibacterium prausnitzii, Eubacterium eligens, and Eubacterium rectale was detected along with increased abundance of Bacteroides vulgatus and Enterobacteriaceae.
Low FODMAP diet (100)	Adults with IBD in remission or with mild disease	Randomized trial		6 wk	Reduced disease activity and fecal calprotectin along with improvement of self-reported quality of life in patients on low FODMAP diet compared to patients on standard diet.
Low FODMAP vs sham control diet (98)	Adults with quiescent IBD	Randomized single-blind trial, placebo control trial	n = 52 IBD patients, 27 on low FODMAP diet, 25 on placebo diet	4 wk	Higher health-related quality of life scores and reduction of symptoms in half of patients on low FODMAP diet compared to control diet. Low abundance of Bifidobacterium adolescentis, Bifidobacterium longum, and Faecalibacterium prausnitzii, along with reduction of total concn of SCFAs in low FODMAP diet-treated patients.
Low-fat, high-fiber diet (LFD) vs improved standard American diet (iSAD) (97)	Patients with UC in remission or with mild disease	Parallel-group, crossover study. Patients were randomized to an LFD (10% of calories from fat) or an iSAD (35%–40% of calories from fat) for the first 4-wk period, followed by a 2-wk washout period, and then switched to the other diet for 4 wk.	n = 17	4 wk on each diet	All patients remained in remission throughout the study. Both diets increased self-reported quality of life. Patients showed decreased abundance of Actinobacteria, whereas the relative abundance of Bacteroidetes increased. LFD favored the abundance of Faecalibacterium prausnitzii.
IBD-AID (37)	Patients with CD or UC, mild to severe activity	Retrospective	n = 11	4 wk or more	All patients discontinued at least one of their prior IBD medications, and all patients had symptom reduction. Disease activity scores decreased significantly.
IBD-AID (120)	Patients with IBD in remission or with mild to severe disease	Prospective, open-label trial. First, 6-wk baseline period, followed by an 8-wk intervention period.	n = 19	8 wk	Consumption of prebiotics, probiotics, and beneficial foods correlated with increased abundance of Clostridia (namely, Roseburia hominis, Faecalibacterium prausnitzii, Eubacterium eligens, Fusicatenibacter saccharivorans) and Bacteroides (namely, Bacteroides dorei, Bacteroides ovatus, and Bacteroides vulgatus) species. Prebiotics and adverse foods have an inverse impact on cytokine levels (i.e., IL-6, IL-8, TNF-α): negatively and positively correlating with proinflammatory cytokines, respectively.

^aCRP, C-reactive protein: ESR, erythrocyte sedimentation rate.