OBJECTIVE:To determine whether abnormal obese-like neural responses to a meal persist in postobese individuals, who achieved and maintained a normal body weight despite a past history of severe obesity. DESIGN AND SUBJECTS: Cross-sectional study of the brain's response to tasting and consuming a satiating meal in 11 postobese (age: 4076 y, body mass index (BMI): 23.671.9 kg/m 2 ), 23 obese (age: 2976 y, BMI: 39.673.8 kg/m 2 ) and 21 lean (age: 3379 y, BMI: 22.872.1 kg/m 2 ) subjects. MEASUREMENTS: Regional cerebral blood flow (rCBF, a marker of neural activity) at baseline (after a 36-h fast), after tasting and after consuming a satiating liquid meal was assessed using positron emission tomography and state-dependent changes (taste-baseline; satiation-baseline), and compared across groups. Subjective ratings of hunger and fullness were measured by a visual analogue scale and body fatness by dual-energy X-ray absorptiometry. RESULTS: In response to tasting the liquid meal, changes in rCBF were different in the obese as compared to the lean individuals (Po0.05, corrected for multiple comparisons) in the middle insula (peak voxel, x ¼ À41, y ¼ 1, z ¼ 8; Montreal Neurological Institute coordinates) and posterior cingulate cortex (peak voxel, x ¼ 17, y ¼ À47, z ¼ 40). The middle insular cortex exhibited a similar increase of neural activity in the obese and postobese subjects, whereas in the lean subjects the regional activity did not change. In the posterior cingulate cortex, the changes in rCBF in the postobese subjects were not different from those in the other groups. In response to a satiating amount of the same liquid meal, changes in rCBF were different in the obese as compared to the lean individuals (Po0.05, corrected for multiple comparisons) in the posterior hippocampus (peak voxel, x ¼ 21, y ¼ À45, x ¼ 4), posterior cingulate cortex (peak voxel, x ¼ 17, y ¼ À47, z ¼ 40), and amygdala (peak voxel, x ¼ 27, y ¼ 1, z ¼ À24). The posterior hippocampus exhibited a similar decrease of neural activity in the obese and postobese subjects, whereas in the lean subjects the regional activity increased. In the posterior cingulate cortex and amygdala, the changes in rCBF were not different between the postobese and lean individuals. None of the changes in neural activity were correlated with the age of the individuals, the subjective ratings of hunger and fullness, or the meal induced-changes in plasma glucose, insulin, or serum free fatty acids. CONCLUSION: Persistence of abnormal neural responses to a meal in the postobese individuals, a group at high risk for relapse, indicates that a predisposition to obesity may involve areas of the brain that control complex aspects of eating behavior including anticipation and reward, chemosensory perception, and autonomic control of digestion (insular cortex), as well as enteroception and learning/memory (hippocampus).