Reference
Levine A, Wine E, Assa A, et al. Crohn’s disease exclusion diet plus partial enteral nutrition induces sustained remission in a randomized controlled trial. Gastroenterology. 2019;157(2):440-450.
Design
Prospective, randomized clinical trial
Participants
Study participants included 78 children and adolescents aged 4-18 years with mild to moderate active luminal Crohn’s disease as determined by the Pediatric Crohn’s Disease Activity Index (PCDAI). The PCDAI is calculated based on symptoms, a physical exam, and laboratory markers (hematocrit, erythrocyte sedimentation rate, and serum albumin). The authors ran 2 arms of the study, one in Canada and one in Israel.
Exclusion criteria included recent use of steroids or biologics, a recent change or initiation of immunomodulators, and rectal or anal disease. Participants were allowed to use antibiotics as needed (excluding quinolones or metronidazole) for the first 10 days of treatment. They were also allowed to continue a stable dose of immunomodulators while enrolled in the study.
Interventions
The Crohn’s disease exclusion diet (CDED), partial enteral nutrition (PEN), and exclusive enteral nutrition (EEN). PEN and EEN consisted of an elemental diet supplement (Modulen by Nestlé).
Study Parameters Assessed
Participants were randomized to 2 groups: Group 1 participants received 50% of their calories from CDED and 50% from PEN for weeks 1-6. For weeks 7-12 this group continued phase II of CDED for 75% of their calories, and received 25% of their calories from PEN. Group 2 received EEN for the first 6 weeks and then no dietary restrictions plus 25% of their calories from PEN for weeks 7-12.
Primary Outcome Measures
The primary outcome measure in this study was tolerability of the diet. Other studies have shown EEN to be effective at inducing remission, but it has poor tolerability and requires tube feeding in 50% to 60% of cases.
Secondary outcome measures looked at response to the treatment. These measures included a drop in the PCDAI of 12.5 points or remission; a decrease or normalization of inflammatory markers (CRP, ESR, calprotectin) at week 6; the lactulose/mannitol ratio (L/M ratio); and adherence.
Researchers also evaluated changes in the gut microbiome over the course of the study.
Key Findings
Primary outcome
Of the 78 patients who started the study, 4 randomized to EEN withdrew within 48 hours for refusal to take Modulen orally. Following the first 48 hours, 39 of 40 participants (97.5%) in the CDED+PEN arm (Group 1) successfully tolerated their regimen, while 28 of 38 participants (73.7%) in the EEN arm (Group 2) were successful.
Secondary outcomes
At week 3, the L/M ratio had improved for CDED+PEN patients, but there was no change in the L/M ratio for EEN patients.
At week 6, there was no statistical difference in response to treatment between the 2 groups; 85% of participants in both groups were in corticosteroid-free remission (as determined by the PCDAI score), though rates of remission were strongly associated with good compliance to the regimen. A normal CRP was present in 51.3% of Group 1 participants and 55.8% of Group 2 participants. Calprotectin levels also dropped significantly for both groups. Analysis of the microbiome in patients who achieved remission, in both groups, showed a specific pattern in certain species that increased and decreased over 6 weeks.
At week 12, 75.9% of Group 1 patients had a normal CRP, versus 47.6% of Group 2 patients. Of those who were in remission at week 6, 87.5% of Group 1 patients and 56% of Group 2 patients were still in remission at week 12. Calprotectin levels continued to decline between weeks 7 and 12 for Group 1 but rose slightly for Group 2. The microbiome of those who achieved remission in Group 1 continued to change in a pattern similar to the first 6 weeks, whereas the microbiome of those who achieved remission in Group 2 rebounded to pretreatment levels.
Practice Implications
Several other studies have investigated the efficacy of exclusively elemental diets (EED) for inducing and sustaining remission of Crohn's disease. EED has been shown to decrease inflammation, increase absorption, and prevent recurrence.2-5 EED, however, is restrictive, generally unpalatable, and costly. For example, for 1,500 calories/day, a typical EED program costs more than $1,000/month. Modulen is not readily available in the United States, but a 1-month supply through Amazon appears to cost nearly $2,000. For these reasons, compliance over the long term tends to be low. Finding dietary strategies that yield the clinical benefit of EEN while giving the patient real food options would greatly improve quality of life as well as compliance.
For the subjects in this study who used the CDED+PEN, inflammatory markers, disease severity, and microbiome were all positively impacted more than in those who used EEN and then PEN plus an unrestricted diet, particularly in weeks 7-12. The CDED is based on the premise that a diet high in animal fats, sugar, gluten, emulsifiers, and thickeners and low in fiber depletes the mucus layer of the gastrointestinal tract, allowing bacteria to invade the lining of the gut and cause inflammation and intestinal permeability. A number of studies have used this protocol and shown positive results.6,7 The diet removes elements that increase inflammation and microbial permeability, and includes foods that improve microbial diversity and decrease inflammation. (See the dietary protocol.) The specific components of the diet plan include the following.
Fiber and pectins from fruits and vegetables, which help to produce butyrate and other short-chain fatty acids, reducing inflammation in the gut.6 A low-fiber diet, in contrast, promotes the invasion of pathogenic bacteria into the lining of the gut.
Extremely limited animal fats and moderate intake of fats from vegetable sources. A diet high in animal fat is associated with a higher incidence of inflammatory bowel disease (IBD), whereas a diet high in omega-3 fats decreases risk.8 High-fat diets also encourage the accumulation of secondary bile acids, which inhibit the growth of healthy bacteria.6
Exclusion of gluten-containing grains. The specific carbohydrate diet, which excludes all grains, has been a popular option for Crohn’s patients seeking dietary treatment and has shown some clinical success.9 The authors of this study specifically describe gluten and wheat as alpha-amylase/trypsin inhibitors that encourage inflammation in the gut and promote intestinal permeability.6
Exclusion of processed foods. A number of food additives have been shown to impact intestinal permeability and the integrity of GI mucosa. This includes 2 common emulsifiers, carboxymethylcellulose (CMC) and polysorbate-80, and thickeners such as maltodextrin and carrageenan.10 Martino JV et al write, “Animal studies consistently report that carrageenan and CMC induce histopathological features that are typical of IBD while altering the microbiome, disrupting the intestinal epithelial barrier, inhibiting proteins that provide protection against microorganisms, and stimulating the elaboration of pro-inflammatory cytokines.“11
L/M ratio is a test for gut permeability whereby subjects are given a loading dose of both lactulose and mannitol, and then the urine is tested for clearance of these sugars. Lactulose is a large sugar and will not be absorbed by an intact gut; appearance in the urine will then signal greater permeability in the lining of the GI tract. Mannitol will be better absorbed the more intact the villi in the gut are and will, thus, be present in greater amounts as the gut heals.1 The intestinal permeability of the gut improved only in Group 1 as indicated at week 3.
While it was not a primary outcome measure, this study looked into the role of the makeup of the microbiome in Crohn’s disease. The results showed a significant change in the microbiome between the beginning of the study and week 6 for those subjects who went into remission. By week 12, the microbiome reverted to original levels for subjects who reintroduced an unrestricted diet. Subjects with active disease show a greater number of species such as pathogenic Escherichia coli spp, Bilophila spp, and several species in the phyla Proteobacteria. They also exhibit fewer healthy species such as Bifidobacteria spp and those from the Firmicutes phyla.12 However, trials using probiotics, including fecal transplantation, and antibiotics with Crohn’s disease have had mixed results at best.13-17 This leads to a fascinating series of questions about whether the makeup of the microbiome is causative of symptoms or merely incidental; whether adding the appropriate probiotic bacteria or killing dysbiotic bacteria is useful; and whether the food we eat is the most important factor that influences the microbiome in the pathogenesis of Crohn’s disease.
I am also left with a couple questions about the study itself. While the diet specifically excludes all dairy and butterfat, the protein source in Modulen is casein, and the primary fat source is milk fat. This leads me to wonder how dairy products impact Crohn’s, and whether products with more hydrolyzed forms of protein would yield different clinical results. The primary funding for this study came from Nestlé, the manufacturer of Modulen, and the primary researchers receive compensation for consulting for Nestlé.
