Brain-Heart Interaction During Transcutaneous Auricular Vagus Nerve Stimulation

Objectives Transcutaneous auricular vagus nerve stimulation (taVNS) modulates brain activity and heart function. The induced parasympathetic predominance leads to an increase of heart rate variability (HRV). Knowledge on the corresponding cortical activation pattern is, however, scarce. We hypothesized taVNS-induced HRV increases to be related to modulation of cortical activity that regulates the autonomic outflow to the heart. Materials and Methods In thirteen healthy subjects, we simultaneously recorded 64-channel electroencephalography and electrocardiography during taVNS. Two taVNS stimulation targets were investigated, i.e., the cymba conchae and inner tragus, and compared to active control stimulation in the anatomical vicinity, i.e., at the crus helicis and outer tragus. We used intermitted stimulation bursts of 25 Hz applied at a periodicity of 1 Hz. HRV was estimated with different time-domain methodologies: standard deviation of RR (SDNN), the root mean squares of successive differences (RMSSD), the percentage of RR-intervals with at least 50 ms deviation from the preceding RR-interval (pNN50), and the difference of consecutive RR intervals weighted by their mean (rrHRV). Results The stimulation-induced HRV increases corresponded to frequency-specific oscillatory modulation of different cortical areas. All stimulation targets induced power modulations that were proportional to the HRV elevation. The most prominent changes that corresponded to HRV increases across all parameters and stimulation locations were frontal elevations in the theta-band. In the delta-band, there were frontal increases (RMSSD, pNN50, rrHRV, SDNN) and decreases (SDNN) across stimulation sites. In higher frequencies, there was a more divers activity pattern: Outer tragus/crus helicis stimulation increased oscillatory activity with the most prominent changes for the SDNN in frontal (alpha-band, beta-band) and fronto-parietal (gamma-band) areas. During inner tragus/cymba conchae stimulation the predominant pattern was a distributed power decrease, particularly in the fronto-parietal gamma-band. Conclusion Neuro-cardiac interactions can be modulated by electrical stimulation at different auricular locations. Increased HRV during stimulation is correlated with frequency-specific increases and decreases of oscillatory activity in different brain areas. When applying specific HRV measures, cortical patterns related to parasympathetic (RMSSD, pNN50, rrHRV) and sympathetic (SDNN) modulation can be identified. Thus, cortical oscillations may be used to define stimulation locations and parameters for research and therapeutic purposes.