Quantification of cardiac and respiratory modulation of axonal activity in the human vagus nerve

We recently documented the first microelectrode recordings from the cervical vagus nerve in awake humans. Here we aimed to quantify cardiac and respiratory modulation of vagal activity to assess the feasibility of targeting axons supplying the heart and airways. Multi-unit activity was recorded from 43 sites in 19 healthy participants in the left (n = 10) and right (n = 9) vagus nerves with ECG, continuous non-invasive blood pressure and respiration. Cross-correlation histograms were constructed between axonal spikes and the R-waves or the peaks of inspiration. The latencies for the peak in cardiac modulation showed a bimodal distribution: while the majority of sites (72%) had peak latencies that preceded the R-wave by up to 550 ms (mean ± SD, -300 ± 178 ms), 12 sites had latencies of up to 250 ms following the R-wave (64 ± 87 ms). Interestingly, the majority of sites with negative latencies (68%) were found in the left nerve whereas most of those with positive latencies (75%) were found in the right. Conversely, on average the peak of respiratory modulation straddled the peak of inspiration. Sites showing respiratory modulation were more prevalent and showed stronger modulation than those with cardiac modulation: calculated for sites with modulation indices ≥15%, the median cardiac and respiratory modulation indices were 23.4% (n = 17) and 44.5% (n = 35), respectively. We conclude that, despite the fact that much of the vagus nerve supplies the gut, cardiac and respiratory modulation of vagal nerve activity can be identified through invasive recordings in awake humans. KEY POINTS: Intraneural recordings from the cervical vagus were obtained in awake humans via tungsten microelectrodes inserted into the nerve through ultrasound guidance. Cross-correlation analysis of multi-unit vagal activity revealed cardiac and respiratory modulation, from which the amplitude and latency of the peaks could be computed. The magnitude of the cardiac modulation (23%) was weaker than that of the respiratory modulation (45%). The latencies for the peak in cardiac modulation showed a bimodal distribution: the majority of sites (72%) had peak latencies that preceded the R-wave, while the remainder had latencies that followed the R-wave. The majority of sites with negative latencies (68%) were found in the left nerve whereas most of those with positive latencies (75%) were found in the right. On average the peak of respiratory modulation coincided with the peak of inspiration.