The investigators hypothesize that rates of gastroesophageal reflux will be lower in blenderized diets compared to formula. The investigators will determine if reflux burden as measured by multichannel intraluminal impedance/esophageal pH monitoring performed for clinical purposes differs between low viscosity conventional formula, medium viscosity commercial blenderized diet, and high viscosity commercial blenderized diet in 45 children receiving conventional formula via gastrostomy. All participants will receive each of the three test diets in random order in a 3-way crossover design. Results of gastric emptying scans will be recorded if done for clinical purposes.
