Vagus nerve stimulation has quietly become one of the most intriguing tools in modern neurorehabilitation. By delivering brief electrical pulses to the vagus nerve—the body’s longest cranial nerve—clinicians can trigger the release of brain chemicals that help the nervous system rewire itself. When these pulses are paired precisely with rehabilitation exercises, sounds, or touch, the technique, known as targeted plasticity therapy, has shown promise for stroke recovery, spinal cord injury, tinnitus, and even post-traumatic stress disorder. But a new review from researchers at the University of Miami argues that the therapy’s next leap forward may depend not just on how strongly the vagus nerve is stimulated, but on when—and on what state the patient’s brain and body happen to be in at that exact moment.

The review, published in Frontiers in Control Engineering, examines two major factors that could dramatically influence how well vagus nerve stimulation works in practice: altered cognitive states such as pain, arousal deficits, and diminished motivation, and the ongoing rhythms of the cardiovascular and respiratory systems. Both factors, the authors contend, directly modulate the release of neuromodulators—acetylcholine, norepinephrine, serotonin, and dopamine—that form the biological foundation of targeted plasticity therapy. Understanding these influences, they propose, could pave the way for personalized, state-dependent stimulation paradigms that adapt in real time to each patient’s unique physiology.

At the heart of the therapy lies a deceptively simple mechanism. During a typical session, a brief burst of stimulation—often just half a second long—activates the nucleus tractus solitarius in the brainstem, which then engages subcortical structures including the nucleus basalis of Meynert, the locus coeruleus, the dorsal raphe nucleus, and the ventral tegmental area. These structures serve as the principal sources of the brain’s key neuromodulators. Even a single brief burst releases these chemicals diffusely across the cortex, enhancing neuronal excitability and improving the signal-to-noise ratio of salient events. Crucially, the stimulation alone is insufficient to strengthen synapses; it must be paired with a relevant behavioral event in a temporally precise manner, leveraging what neuroscientists call the synaptic eligibility trace—a short-lived window during which neuromodulators can reinforce Hebbian plasticity, the spike-timing-dependent process that dictates whether connections between neurons are strengthened or weakened.

Decades of preclinical research have converged on what appears to be an optimal stimulation dose: roughly 0.8 milliamperes, 30 hertz, 100-microsecond pulse width, half-second train duration, and biphasic pulses. Yet the relationship between dose and plasticity follows a non-monotonic inverted U curve—too little stimulation fails to engage the neuromodulatory systems, while too much can actually impede recovery. Increasing stimulus intensity and pulse width reliably increases noradrenergic activity in the locus coeruleus and cholinergic transmission from the nucleus basalis, but increasing frequency alters the timing rather than the total amount of VNS-driven neuronal activity. The number of stimulation pairings per session, the intervals between them, and the total therapy duration across weeks all influence the extent of neuroplasticity and functional recovery.

Despite this robust preclinical characterization, clinical responder rates have varied significantly. In the pivotal VNS-REHAB trial for stroke, only 47 percent of participants responded to the therapy; in a tinnitus study pairing stimulation with tone therapy, the responder rate was 50 percent. The Miami researchers argue that this gap between animal models and human trials may stem partly from patient-specific physiology—particularly the comorbidities that accompany neurological disease. Spinal cord injury often brings neuropathic pain; parkinsonism can alter arousal levels and induce sudden daytime sleepiness; stroke patients frequently exhibit apathy. Each of these altered cognitive states modifies neuromodulator transmission and cortical excitability, potentially undermining the very biological processes that targeted plasticity therapy relies upon.

Recent preclinical evidence underscores the stakes. In one study, vagus nerve stimulation failed to improve motor recovery in the presence of neuropathic pain, possibly because the pain reduced the animals’ behavioral engagement in the rehabilitation task. Pain also alters cholinergic signaling in the basal forebrain and can exacerbate noradrenergic transmission when co-occurring with chronic depression. Arousal state presents a similar challenge. The century-old Yerkes-Dodson law reveals an inverted U relationship between arousal and performance, with moderate arousal being optimal and hyperarousal—whether from stress or pain—being detrimental. The authors hypothesize that hyperarousal resulting from a higher stimulation dose could impede targeted plasticity therapy, potentially explaining the inverted U relationship between stimulation dose and therapeutic outcome. Diminished motivation, common after brain injury, further complicates matters by impairing endogenous dopaminergic engagement, though intriguingly, vagus nerve stimulation itself has been shown to boost the drive to work for rewards.

Medications add yet another layer of complexity. Beta blockers can alter the inverted U relationship between stimulus intensity and neuroplasticity, while alpha blockers attenuate activity-dependent plasticity during rehabilitation. The authors emphasize that pharmacological agents affecting neuromodulator systems must be carefully considered when designing stimulation strategies that depend on those same systems. Quantifying these cognitive states in the clinic may soon become feasible: visual analog scales can score pain perception, pupillometry can serve as a biomarker of arousal linked to noradrenergic signaling in the locus coeruleus, and EEG measures such as the ratio of high-frequency to low-frequency spectral power or the P300 event-related potential can quantify correlates of attention and noradrenergic signaling.

Beyond cognitive states, the review highlights the profound influence of cardiorespiratory rhythms on neuromodulator signaling. The vagus nerve is primarily a sensory conduit—roughly 80 percent of its cervical fibers are afferent, relaying visceral information from the cardiovascular and respiratory systems to the brain. Baroreceptors in the aortic arch and carotid sinus signal blood pressure changes, pulmonary stretch receptors signal lung inflation, and peripheral chemoreceptors in the carotid and aortic bodies signal blood gas content. All of these inputs converge on the nucleus tractus solitarius, the same brainstem structure that vagus nerve stimulation engages. Because these visceral signals interact and modulate overall nucleus tractus solitarius activity, the neuromodulator release induced by stimulation may differ depending on whether it is delivered during systole or diastole of the cardiac cycle, or during inspiration or expiration of the respiratory rhythm.

Evidence for this phase dependence is mounting. Neurons in the locus coeruleus are more likely to be activated during the diastolic rather than the systolic phase of the cardiovascular rhythm, and respiratory rhythms entrain locus coeruleus neuronal activity, making it more responsive to specific phases of the respiratory cycle. A recent functional MRI study demonstrated greater locus coeruleus activation when auricular vagus nerve stimulation was delivered during exhalation compared to inhalation, suggesting the nucleus tractus solitarius may be more sensitive to afferent input during expiration. Cortical excitability in response to sensory stimuli is also preferentially modulated by cardiac phase: somatosensory evoked potentials and pain perception are higher during diastole than systole, a phenomenon that appears across different sensory modalities. Heart rate variability, quantified as the ratio of low-frequency to high-frequency activity, offers a marker of sympathetic and parasympathetic balance, though its sensitivity to recording length, breathing, body position, and emotional state poses challenges for adaptive paradigms requiring second-level timescales.

The Miami team proposes that a state-dependent targeted plasticity therapy system could integrate these multiple signals—EEG for arousal, pupillometry for basal neuromodulator tone, electrocardiography and pneumography for cardiorespiratory phase, and wearable sensors for longer-term functional tracking—into a closed-loop controller that adapts stimulation parameters in real time. Existing neuromodulation devices already span a spectrum of control schemes, from open-loop systems with fixed parameters to closed-loop feedforward and feedback controllers that monitor variables such as electrocorticography, evoked compound action potentials, or respiratory sensors. Machine learning frameworks, including Bayesian optimization, deep neural networks, and ensemble learning, may be needed to fine-tune stimulation parameters and decipher the complex relationships between state-dependent signals and optimized neuroplasticity. Bayesian optimization, which iteratively converges on ideal settings while balancing exploration and exploitation, may be especially effective for individualizing therapy to each patient’s unique physiological profile. The authors acknowledge that extensive work remains to test these hypotheses, particularly the combinatorial effects of cognitive states and autonomic rhythms on neuromodulator signaling, plasticity, and functional improvement. But if the framework holds, the future of vagus nerve stimulation may look less like a fixed prescription and more like a therapy that listens to the body’s own rhythms—delivering its plasticity-enhancing pulses only when the brain is primed to receive them.

Subject of Research: State-dependent vagus nerve stimulation for targeted plasticity therapy: how cognitive states and cardiorespiratory rhythms shape neuromodulator signaling and neuroplasticity

Article Title: State dependent vagus nerve stimulation for targeted plasticity therapy: challenges and considerations

Article References: Nandakumar, B., Mylavarapu, R. V., Harris, R., Albuquerque, E. R., Yan, Z., Herter, C., McMillan, D. W., Kanumuri, V. V., & Ganzer, P. D. (2024). State dependent vagus nerve stimulation for targeted plasticity therapy: challenges and considerations. Frontiers in Control Engineering, 5, Article 1452442. https://doi.org/10.3389/fcteg.2024.1452442

Image Credits: AI Generated

DOI: 10.3389/fcteg.2024.1452442

Keywords: vagus nerve stimulation, targeted plasticity therapy, neuromodulation, neuroplasticity, stroke rehabilitation, cardiorespiratory rhythms, personalized medicine, closed-loop systems, State, dependent, vagus, nerve