Does vagus nerve stimulation strengthen auditory steady-state responses in rat cortex?

VNS as a booster for rhythmic sound

Gamma entrainment using sensory stimulation (GENUS) delivers rhythmic sensory input in the hope of driving frequency-specific cortical activity, but how strongly the cortex follows the stimulus varies. The paper asks a narrow, testable version of that problem: can VNS make cortical responses to rhythmic sound stronger and more consistently locked to the stimulus?

The rationale comes from prior work cited in the paper: vagal afferents recruit noradrenergic, cholinergic, and serotonergic systems that shape cortical state, and VNS has been reported to amplify sound-evoked beta–gamma power in rat auditory cortex through cholinergic mechanisms. What remained unknown was whether VNS strengthens responses at the specific beta and gamma rates used in rhythmic stimulation, and whether any effect depends on which cortical region is recorded.

Four regions, three ways to measure locking

The team recorded electrocorticography (ECoG) from four right-hemisphere regions in seven male Wistar rats under isoflurane anesthesia: core auditory cortex (ACcore), the surrounding belt (ACbelt), the insular auditory field (IAF), and anterior insular cortex (AIC). The regions were chosen to separate auditory-driven tissue from tissue without a clear stimulus-driven auditory response: ACcore and IAF show distinct click-evoked P1 peaks, whereas ACbelt and AIC lack a comparable onset response.

Auditory stimuli were 500-ms click trains at 20 Hz and 40 Hz, plus 1-Hz clicks used to measure the transient P1–N1 evoked potential. Recordings were made before VNS and again 90 minutes after starting 30-Hz cervical VNS (500-µA biphasic pulses, 30 s on, 5 min off).

Three measures answer different questions and should not be treated as interchangeable. Inter-trial phase clustering (ITPC) measures how consistently response phase aligns with the stimulus across trials. Event-related spectral perturbation (ERSP) measures change in spectral power in decibels. The debiased weighted phase-lag index (WPLI) measures non-zero-lag phase coupling between recording sites. A stronger ITPC means tighter timing; a larger ERSP means more power; a higher WPLI means more coordinated phase relationships between electrodes.

The transient click response rose everywhere

VNS increased the normalized P1–N1 amplitude of the transient click-evoked response. Across regions, the estimated marginal mean rose from 0.47 to 0.77, a difference of 0.30, with a main effect of timing but no significant region-by-timing interaction. In other words, the transient response grew overall without evidence that the growth differed by region.

Baseline amplitudes did differ by region: ACcore (1.19) exceeded IAF (0.89), and both exceeded ACbelt (0.31) and AIC (0.08). This regional ordering matters for reading the steady-state results, because the regions that responded most strongly to single clicks were also the ones where sustained rhythmic responses were most affected.

40-Hz gains stayed inside auditory core

During 40-Hz trains, VNS increased both 40-Hz ITPC and 40-Hz ERSP only in ACcore. ITPC rose from 0.134 to 0.284, and ERSP rose from −1.32 to 0.58 dB. No other region changed significantly on either measure.

The analysis window was restricted to 100–500 ms of each train to emphasize sustained responses rather than the transient onset. The paper reports that the broadband onset response looked similar before and after VNS, with traces diverging afterward through train offset. That detail supports reading the effect as a change in sustained stimulus locking rather than a general boost to the onset response.

20-Hz locking spread beyond core

During 20-Hz trains, the pattern split by measure. ERSP at 20 Hz increased only in ACcore, from −0.49 to 1.55 dB. ITPC at 20 Hz, however, increased in ACcore (0.117 to 0.314), IAF (0.091 to 0.164), and ACbelt (0.081 to 0.119), but not in AIC.

So phase locking became more consistent across three auditory-driven regions, while spectral power rose only in core. The 40-Hz harmonic of the 20-Hz trains behaved like the 40-Hz condition: ITPC and ERSP increases were confined to ACcore. The authors read this as response frequency shaping the spatial extent of the effect, with 40-Hz enhancement staying in core at both train rates.

Coupling rose at driven frequencies, not alpha

Within the auditory array, WPLI increases concentrated at frequencies present in each click train. During 20-Hz trains, 20-Hz WPLI increased in ACcore–ACcore, ACcore–ACbelt, and ACbelt–ACbelt pairs; at the 40-Hz harmonic, only ACcore–ACcore pairs reached significance. During 40-Hz trains, the largest estimated increase was in ACcore–ACcore pairs, but it did not survive correction across pair types.

Two negative results constrain the interpretation. Alpha-band WPLI showed no significant increase during either train rate, and all estimated changes were negative. Cross-array pairs showed no significant change at driven frequencies, apart from a decrease in alpha-band coupling for ACbelt–IAF pairs during 40-Hz trains. The authors use this pattern to argue against a generalized rise in cortical synchrony.

A plausible cholinergic route, not a tested one

The paper's proposed mechanism is cholinergic modulation. Prior work cited in the paper shows that cholinergic manipulations affect auditory steady-state responses, particularly at 20 and 40 Hz, and that cholinergic blockade prevents VNS-induced enhancement of sound-evoked beta–gamma power. The authors present this as a plausible route, not a demonstrated one: the cholinergic mechanism was not manipulated directly in this study.

Regional selectivity also limits how far the result can be pushed. Enhancement appeared in ACcore and, at 20 Hz, in ACbelt and IAF, but not in AIC. The authors note that AIC differs from the other regions in laminar organization, cell composition, and baseline dynamics, so auditory drive is not established as a causal requirement. A granular, thalamorecipient circuitry is offered as a plausible contributor, since IAF occupies granular posterior insula and receives auditory thalamic input while AIC is agranular.

Target engagement is not therapeutic efficacy

The most decision-relevant comparison is between frequency matching and non-matching. Forty-hertz phase locking, power, and coupling increased with 30-Hz VNS, so matching the stimulation rate to the desired cortical rhythm was unnecessary under these conditions. The 20-Hz effects suggest relevance to beta-rate protocols, and the authors connect them to disorders with impaired 20-Hz auditory steady-state responses.

Several boundaries qualify that reading. The rats were healthy and anesthetized; isoflurane suppresses rapid auditory steady-state responses, and awake-state effects may differ. There was no concurrent sham group, although the authors note that a prior comparable preparation found stable P1 amplitudes over time without VNS, and they argue that the regional and frequency selectivity makes a global time or anesthesia effect less likely. Only 20- and 40-Hz trains were tested, so the boundary between this enhancement and suppression at 6–8 Hz is unknown. Finally, ITPC and ERSP cannot distinguish entrainment of an endogenous oscillator from repeated evoked responses, so the data support strengthened stimulus-locked responses more directly than entrainment of intrinsic gamma oscillations.

The paper's own priority is therefore specific: VNS may strengthen auditory cortical target engagement during GENUS when disease-related dysfunction limits auditory responses, and the effect is regionally and frequency selective rather than a global increase in synchrony. Stronger target engagement is not the same as therapeutic efficacy, because no disease pathology or behavioral outcome was measured here.

Sources used in this answer