Effects of Transauricular Vagus Nerve Stimulation on Heart Rate Variability: Wearable Sensor Data in Healthy Volunteers

Aims Surgically implanted vagus nerve stimulation (VNS) is a recognised treatment for depression. The vagus nerve can also be stimulated non-invasively via its auricular branch, using transauricular vagus nerve stimulation (taVNS). Heart rate variability (HRV) is a putative biomarker of autonomic nervous system (ANS) engagement. We aimed to test the impact of taVNS on the ANS of healthy volunteers by measuring HRV using a double-blind, sham-controlled, longitudinal design to acquire data over 7 days using wearable cardiac sensors. Methods taVNS was delivered to the left ear of healthy volunteers using a transcutaneous electrical nerve stimulation (TENS) device via a custom clip electrode (developed at Newcastle University). All participants were stimulated at 10 Hz, with pulse widths of 300 ms and variable current outputs, depending on perceptual thresholds. We delivered double-blinded active and sham taVNS for hour-long periods, in the morning and evening. We also recorded an electrocardiogram (ECG) lead I using a VitalPatch for 7 consecutive days. Python scripts were developed to produce HRV timeseries and plot data. ECG frequency domain parameters – low- (LF) (0.05–0.15 Hz) and high-frequency (HF) power (0.15–0.4 Hz) – were calculated for each stimulation period. The LF/HF ratio was used as a marker of autonomic modulation. The Wilcoxon signed-rank test was used to compare LF/HF ratio distributions. Results Initial data from the wearable sensors were used to develop interpolation scripts to improve the processing of noise, missed R waves and ectopic beats, to reduce errors when estimating HRV from the heart rate signal. Initial results from 97 individual 1-hour long stimulation periods, from 18 participants, show that active stimulation in the morning, when compared with sham stimulation in the same period, significantly reduces the LF/HF ratio. The median and interquartile range (IQR) of the LF/HF ratio for the active and sham periods was, respectively, 1.72 (1.99) and 2.75 (2.82), a statistically significant difference (p = 0.043). Conclusion taVNS modulates HRV frequency domains, suggestive of vagal cardiac effects, and replicates findings from previous taVNS studies. Reductions in the LF/HF ratio are suggestive of increased parasympathetic tone. As the auricular branch of the vagus does not have any direct cardiac efferents, this suggests central ANS modulation using taVNS. Secondly, it suggests that cardiac ANS modulation could be used as a proxy measure of afferent vagal stimulation, which could be of clinical utility. These effects warrant exploration in a larger cohort study, including wider demographics (including age range) and improved processing pipelines.