Adequate enteral nutritional support is often limited by gastrointestinal (GI) side effects. In this pilot clinical trial we compared an enteral nutrition formula based on soy hydrolysate (study formula, SF) against a widely used intact casein formula (control formula, CF) for the incidence of GI side effects in a completely randomized double blind design. Twenty-three nonsurgical hospitalized patients requiring enteral nutritional support and free of GI symptoms were randomly assigned to receive either the CF or the SF for 6 days continuously. Both formulas were isotonic, low in residue, lactose free and isocaloric, but differed in the type and concentration of protein and the concentration of medium-chain triglycerides. After randomization both groups were comparable in demographic characteristics, and nutritional status, but there were more patients on antibiotics in the CF group. The amount of formula infused per day and the route of administration were equivalent. The number of bowel movements per day was 1.0 +/- 0.5 for the CF group and 0.6 +/- 0.3 for the SF group (p less than 0.05). The incidence of diarrhea was 10.8% days for the CF group and 6.2% for the SF group (p = NS). High gastric residuals occurred in 16.9% of days in the CF group and 3.3% in the SF group (p less than 0.05). Vomiting incidence was 10.8% in the CF group and 1.5% in the SF group (p less than 0.05). After adjustment for the use of antibiotics as a covariate, the differences in number of bowel movements, vomiting and incidence of high residuals became less significant (p less than 0.10).(ABSTRACT TRUNCATED AT 250 WORDS)