Human risk taking and metabolic state: No credible evidence for desacyl-ghrelin modulation of neural or behavioural...

Risk taking has often been proposed to depend on homeostatic systems. However, empirical evidence remains inconsistent, and the underlying mechanisms remain debated. Changes in risky decision-making, which are associated with homeostatic alterations, may be affected by underlying ghrelin signaling, a peptide hormone that also interacts with dopaminergic functions. The deacylated form of ghrelin has recently received increased attention. Across two studies, we examined whether experimental interventions known to alter desacyl-ghrelin levels influence risky choice in healthy male participants. In study 1, participants underwent a brief fasting period ( $N=37$ ; $N=26$ for fMRI analyses), whereas in study 2, participants experienced one night of total sleep deprivation ( $N=40$ ; $N=36$ for fMRI analyses). Risky decision making was assessed using a standard choice paradigm, and behaviour was analysed using computational modelling. Across both studies, we found no credible effects of the experimental manipulations on the proportion of risky choices. Computational modelling indicated that the standard prospect theory (which describes how participants may overestimate small probabilities and underestimate large ones), without accounting for choice repetition, best described the observed behaviour. However, model parameters were not systematically influenced by the experimental manipulations, and changes in desacyl-ghrelin levels were not robustly associated with decision parameters. Functional MRI analyses revealed no effects of state manipulation on neural representations of subjective value or choice in a priori defined regions of interest, including the ventromedial prefrontal cortex, ventral striatum, posterior cingulate cortex, anterior cingulate cortex, and anterior insula. Overall, these findings suggest that state-dependent influences on risky decision making, and their modulation by desacyl-ghrelin, may be weaker than previously thought.