BACKGROUND: The purpose of the study was to evaluate, using meta-analysis techniques, whether enteral nutrition is effective in inducing clinical remission in active Crohn's disease. METHODS: Randomized trials either comparing enteral nutrition with steroids or comparing elemental (amino acid-based) with nonelemental diets were selected using MEDLINE (1984 to 1994), reference lists from published articles, reviews, and abstracts from major gastrointestinal meetings. Sixteen studies fulfilled the inclusion criteria (four published as abstracts). Crude rates for induction of remission were collected on an intention-to-treat basis by three independent observers. Each study was given a quality score, based on predetermined criteria. RESULTS: The pooled odds ratio (OR) for all type of enteral diets compared with steroid therapy was 0.35 (95% CI, 0.23 to 0.53). This result was similar for the best studies (by quality score) combined, for trials using tube feeding combined, and when noncompliant patients were withdrawn. Further subgroup analyses were conducted on the basis of the type of diet administered. Peptide-based diets were significantly inferior to steroids (pooled OR, 0.32; CI, 0.20 to 0.52). There was a trend to lower remission rate after elemental diets than after steroids (pooled OR, 0.44; CI 0.17 to 1.12). On the other hand, pooled OR for whole protein-based diets compared with elemental diets was 1.28 (CI, 0.40 to 4.02). CONCLUSIONS: Data available to date show that steroids are better than enteral nutrition to induce remission in active Crohn's disease. These results are more evident when peptide-based diets are administered, but they are not conclusive when either elemental or whole protein-based diets are used.
BACKGROUND: The purpose of the study was to evaluate, using meta-analysis techniques, whether enteral nutrition is effective in inducing clinical remission in active Crohn's disease. METHODS: Randomized trials either comparing enteral nutrition with steroids or comparing elemental (amino acid-based) with nonelemental diets were selected using MEDLINE (1984 to 1994), reference lists from published articles, reviews, and abstracts from major gastrointestinal meetings. Sixteen studies fulfilled the inclusion criteria (four published as abstracts). Crude rates for induction of remission were collected on an intention-to-treat basis by three independent observers. Each study was given a quality score, based on predetermined criteria. RESULTS: The pooled odds ratio (OR) for all type of enteral diets compared with steroid therapy was 0.35 (95% CI, 0.23 to 0.53). This result was similar for the best studies (by quality score) combined, for trials using tube feeding combined, and when noncompliant patients were withdrawn. Further subgroup analyses were conducted on the basis of the type of diet administered. Peptide-based diets were significantly inferior to steroids (pooled OR, 0.32; CI, 0.20 to 0.52). There was a trend to lower remission rate after elemental diets than after steroids (pooled OR, 0.44; CI 0.17 to 1.12). On the other hand, pooled OR for whole protein-based diets compared with elemental diets was 1.28 (CI, 0.40 to 4.02). CONCLUSIONS: Data available to date show that steroids are better than enteral nutrition to induce remission in active Crohn's disease. These results are more evident when peptide-based diets are administered, but they are not conclusive when either elemental or whole protein-based diets are used.