Non-invasive vagus nerve stimulation modulates Pavlovian bias in a state-dependent manner

The vagus nerve transmits vital signals between organ systems of the body and the brain. Despite growing interest in non-invasive transcutaneous vagus nerve stimulation (tVNS) for treatments of various disorders, it is not known whether its modulatory effects on behavior are dependent on bodily states. Here, we used a single-blind randomized crossover design with two within-participant factors, stimulation (right tVNS vs. sham) and metabolic state (water vs. caloric load), to investigate 54 healthy participants (210 sessions). To evaluate state-dependent changes in reinforcement learning, we used a valenced go/no-go task. Performance on this task can be used to estimate a Pavlovian choice bias, which captures improved performance for valence-congruent actions (e.g., go to win rewards) versus valence-incongruent actions (e.g., no-go to win rewards) and has been associated with dopaminergic neurotransmission. In line with theorized state-dependency of tVNS-induced changes, we observed that tVNS decreased Pavlovian response biases (Go x Win) after the caloric load (Stimulation x Load: p =.038). Taking into account individual changes in hunger but not satiety ratings best captured behavioral modulation, indicating that this modulatory effect could be driven by motivational effects related to metabolic interoception (sensed "need"). We conclude that right tVNS decreases Pavlovian response biases if the body is in a postprandial, but not a fasting state. Since fasting and Pavlovian biases have been previously associated with elevated dopamine neurotransmission, our results call for additional research on state-dependent motivational effects of tVNS to optimize interventions.