LAUSR

INTRODUCTION Metabolic and hormonal signals have been shown to be associated with brain activity in the context of ingestive behaviour. However, this has mostly been seen in studies using external administration of hormones or glucose. We therefore studied endocrine-brain interaction in a physiological setting with hormone levels determined by metabolic conditions such as normal food intake vs. prolonged fasting. METHODS 24 healthy, normal weight men participated in two sessions, one involving a 38-h fasting period and one a non-fasting control condition with standardized meals. Functional magnetic resonance imaging was performed at the end of the experiment with participants being required to rate pictures of food. Brain activation was compared between conditions in predefined regions of interest (ROIs). Multiple blood samples were taken to determine levels of insulin, C-peptide, cortisol, ACTH, glucose and adiponectin. These were used as a predictor variable in a regression analysis on brain activations in the different ROIs. RESULTS Food pictures were rated as more desirable in the fasting condition. Univariate analysis of ROI activations revealed mainly effects of food rating and no significant effects of the metabolic state. Multiple regression analysis revealed associations between orbitofrontal cortex activation and blood glucose in the non-fasting condition. In the fasting condition adiponectin was associated with the signal from the caudate nucleus and insulin and C-peptide were associated with functional activity of orbitofrontal regions. DISCUSSION Associations of endocrine signals and functional neural regions could be demonstrated in a realistic setting without external administration of hormones. As the current approach was correlational, further studies need to address the causal role of hormonal signals.