A Randomized Sham-Controlled Cross-Over Study on the Short-Term Effect of Non-Invasive Cervical Vagus Nerve Stimulation on Spinal and Supraspinal Nociception in Healthy Subjects

Objective The aim of the present study was to test the effects of vagus nerve stimulation (VNS) on the descending pain inhibition, quantified by the nociceptive flexor (RIII) reflex and the conditioned pain modulation (CPM) paradigm, and on supraspinal nociceptive responses, assessed by pain intensity and unpleasantness ratings and late somatosensory evoked potentials (SEPs), in healthy subjects. Background Non-invasive vagus nerve stimulation (nVNS) showed promising effects on headache and pain treatment. Underlying mechanisms are only incompletely understood but may include the activation of the descending pain inhibitory system and/or the modification of emotional responses to pain. Methods Twenty-seven adult, healthy, and pain-free subjects participated in this double-blind cross-over study conducted at a university research center. They received 4 minutes of cervical nVNS or sham stimulation in randomized order. RIII reflexes, pain ratings, and SEPs were assessed before, during, and 5, 15, 30, and 60 minutes after nVNS/sham stimulation, followed by CPM testing. The primary outcome was the nVNS effect on the RIII reflex size. Three subjects were excluded after the preparatory session (before randomization), 1 subject was excluded after outlier analysis, leaving 23 for analysis. Results RIII reflex areas were 917.1 +/- 563.8 mu V x ms (mean +/- SD) before, 952.4 +/- 467.4 mu V x ms during and 929.2 +/- 484.0 mu V x ms immediately after nVNS and 858.4 +/- 489.2 mu V x ms before, 913.9 +/- 539.7 mu V x ms during and 862.4 +/- 476.0 mu V x ms after sham stimulation, revealing no differences between the immediate effects of nVNS and sham stimulation (F-[3,F-66] = 0.67,P = .574). There also were no effects of nVNS over sham on RIII reflex areas up to 60 minutes after nVNS (F-[1.7,F-37.4] = 1.29,P = .283). Similarly, there was no statistically significant effect of nVNS on pain intensity ratings and thresholds, RIII reflex thresholds, late SEP amplitudes, and the CPM effect, compared to sham. Pain unpleasantness ratings statistically significantly decreased from 4.4 +/- 2.4 (NRS 0-10) to 4.1 +/- 2.5 during nVNS compared to sham stimulation (F-[1,F-22] = 8.74,P = .007), but there were no longer lasting effects (5-60 minutes after stimulation). Conclusions The present study does not support an acute effect of nVNS on descending pain inhibition, pain intensity perception or supraspinal nociception in healthy adults. However, there was a small effect on pain unpleasantness during nVNS, suggesting that nVNS may preferentially act on affective, not somatosensory pain components.