OBJECTIVE: Several large studies have shown a negative association between Helicobacter pylori (H. pylori) infection and esophageal adenocarcinoma. Diminution of gastric ghrelin secretion by H. pylori could protect against esophageal malignancy by decreasing appetite, food intake, and acid production, thereby decreasing weight and gastroesophageal reflux. METHODS: We evaluated the association of ghrelin with esophageal adenocarcinoma using a population from a previous nested case-control study. Among 128,992 enrolled in a multiphasic health checkup (MHC) between 1964 and 1969, 52 patients developed esophageal adenocarcinoma by the year 2000. Three random controls from the MHC cohort were matched to each case by age, sex, race, and the date and site of their MHC. Serum samples collected at the MHC had been previously tested for IgG antibodies against H. pylori and the CagA protein. Serum ghrelin concentrations were determined by a commercial EIA on 52% of the initial subjects (31 cases and 79 controls). RESULTS: A concentration of ghrelin greater than 3,200 pg/mL at MHC (fourth quartile) was associated with a lower risk of esophageal cancer (H. pylori and body mass index [BMI] adjusted OR=0.18 [CI 0.04-0.78]). This inverse association was seen only in overweight subjects (BMI>or=25, P value for interaction=0.09). The effects of H. pylori and ghrelin were independent. CONCLUSION: Contrary to the original hypothesis, high rather than low serum ghrelin was associated with protection against esophageal adenocarcinoma but only among overweight subjects.