objective: To investigate the relation between ghrelin responses and meal initiation and the effects of BMI and energy status on this. Design: The experiment had a randomised, cross-over design. Setting and subjects: Nine normal-weight (age: 33.2 +/- 4.8 y. BMI: 23.2 +/- 0.5 kg/m(2)) and eleven obese (age: 40.8 +/- 4.7 y, BM L 33.2 +/- 0.8 kg/m(2)) healthy men were recruited from a pool of volunteers and by advertisements. Interventions: Subjects followed a three-day energy restrictive and a three-day energy balanced diet separated by one month. Each diet was followed by a time-blinded (overnight) stay at the research facility. Subjects received a breakfast (preload) and were instructed to ask for lunch when they felt hungry. Ghrelin, insulin, glucose. free fatty acids, appetite, IMI and energy intake during lunch were assessed. Results: Postprandial decreases in ghrelin (r=-0.54; p<0.05) and the AUC of the ghrelin response (r=-0.57, p=0.01) were associated with the intermeal interval, independent of diet, but in normal weight subjects only. Lunch request was preceded by an increase in ghrelin, reaching at least 93% of fasting values. These preprandial increases in ghrelin were correlated with IMI, after energy restriction only. Ghrelin concentrations but not changes in ghrelin were correlated with appetite. Conclusion: Meal-related changes in ghrelin are correlated with the IMI in normal weight subjects only, independent of diet. Ghrelin concentrations may need to reach a certain threshold level before the next meal is initiated. Sponsorship: Supported by Dutch Ministry of Education, Culture and Science. Dutch Ministry of Health, Welfare and Sport and Danone Research. (C) 2009 Published by Elsevier Inc.