EXPLORING THE USE OF VAGAL NERVE STIMULATION IN THE ACQUISITION AND EXTINCTION OF BODILY SYMPTOMS.

Inspired by promising results in rats, researchers have investigated whether vagus nerve stimulation (VNS) in humans improves the ability to learn changes in the relationship between stimuli and outcomes. The hypothesized working mechanism is that VNS increases central noradrenergic (NA) activity and, in turn, facilitates learning contingency changes. However, the available evidence is inconclusive, and little is known about the optimal stimulation parameters. The goal of this PhD project was to further investigate the NA mechanism and cognitive effects of non-invasive VNS in humans. We mainly focused on transcutaneous auricular vagus nerve stimulation (taVNS), but also explored slow deep breathing (SDB). We addressed three research questions. First, do taVNS and SDB improve learning contingency changes? Second, does taVNS increase biomarkers of tonic and phasic NA activity? Third, does taVNS modulate NA activity as a function of the stimulation parameters? In the first study (chapter 2), we tested whether taVNS with commonly used parameters (30s ON/30s OFF; 0.5mA) improves reversal learning. We also investigated the NA mechanism of taVNS using indirect markers of tonic NA activity (i.e., tonic pupil size at rest, cortisol, and salivary alpha-amylase - sAA). Contrary to our hypotheses, taVNS did not facilitate reversal learning and did not reliably increase the tested markers of tonic NA activity. In parallel, we conducted a second study (chapter 3) investigating whether SDB improves reversal learning in the same task used in chapter 2 and similarly found no effect of SDB. Given the inconclusive evidence in favor of a NA mechanism of taVNS, we ran a third study (chapter 4) to test whether taVNS increases markers of phasic (i.e., event-related pupil dilation) and tonic NA activity (i.e., two pupil indexes of tonic NA activity, sAA, and cortisol) in the context of an auditory oddball task. Assuming that the low intensity of taVNS in the first study contributed to the zero findings, we administered taVNS continuously with the maximum intensity below the pain threshold (chapter 4). Nevertheless, again taVNS did not modulate any of the phasic or tonic NA biomarkers. The reported zero findings on taVNS (chapters 2 & 4) prompted us to systematically investigate the parameters of taVNS. For this purpose, we set up a fourth study (chapter 5) to test whether short bursts of taVNS increase markers of phasic NA activity (i.e., two indexes of evoked pupil dilation) as a function of the pulse width and intensity. We found that a higher charge per pulse (intensity*pulse width) was associated with a larger evoked pupil dilation. In line with our expectations, this effect was stronger in the taVNS condition compared to sham stimulation. Our results suggest that non-invasive VNS in humans has no (robust) effects on learning contingency changes and challenge the hypothesis that taVNS with commonly used parameters increases NA activity. The observed parametric dependent effect of taVNS indicates that suboptimal stimulation parameters might underlie the observed zero findings. More research is thus needed to establish whether administering short bursts of taVNS with a higher charge per pulse improves learning contingencies changes.