Diabetes is a major healthcare crisis affecting more than 422 million people worldwide with 38 million in the USA—15% of which have painful Diabetic Peripheral Neuropathy (DPN). The direct cost of diabetes (including pain) has been steadily increasing—412B in 2022 in the United States (US) alone (ADA 2022 report)—not including the 95 million pre-diabetics and their combined associated co-morbidities costs. At the current rate, the overall cumulative cost of diabetes, pre-diabetes and its associated co-morbidities will consume all US healthcare expenditures within the next 5–10 years (Dyess, 2023; Lin et al., 2018; Cawley et al., 2021). This meteoric rise underscores the failure of current symptom control strategies and the critical need for immediate therapies addressing the root cause of the underlying metabolic dysfunction and associated co-morbidities including neuropathic pain.

Type two diabetes (T2DM) appears to have a genetic predisposition and typically affects those with a poor diet and obesity. Native Americans are 2.3 times more likely to die from complications of diabetes compared to the general American population. DPN is a debilitating neurologic disorder characterized by a distal-to-proximal loss of peripheral nerve function, associated with pain, disability, as well as higher all-cause and cardiovascular mortality (Hicks et al., 2021). Sensory neurons, vulnerable to hyperglycemic damage due to their lack of insulin-regulated glucose uptake can still be affected in well-controlled T2DM (Tomlinson and Gardiner, 2008). Reduced expression of vascular endothelial growth factor-A (VEGF-A) underlies neuronal damage and pain in DPN, as VEGF-A supports both vascular and neuronal integrity (Jerić et al., 2017; Pawson et al., 2010; Carmeliet and Storkebaum, 2002). Since the VEGF family also regulates lipid metabolism (Zhou et al., 2023), additional metabolic risk factors, including dyslipidemia and metabolic syndrome, exacerbate peripheral nerve bioenergetic failure. Notably, the term “diabesity,” coined over 50 years ago, underscores the pathophysiological link between obesity and T2DM as well (Sims et al., 1973).

DPN management in T2DM patients has traditionally centered on lifestyle interventions including weight loss, exercise, systemic and topical medications and implanted single modality neuromodulation therapies. Current therapeutic approaches often fall short in providing sustained symptom relief, necessitating the exploration of alternative modalities that address the root cause of metabolic and neurovascular dysfunction (Barrett et al., 2007; Yang et al., 2022). More recently neuromodulation strategies including spinal cord stimulation and auricular stimulation have been approved by the FDA for diabetic neuropathy pain.

Even with aggressive worldwide symptom control pharmacological strategies, diabesity prevalence continues to rise globally, particularly in countries transitioning from low to middle-income economies (Saeedi et al., 2019). Excessive consumption of sugary and high-fat ultra processed foods, driven by hedonic rather than homeostatic mechanisms, has been compounded by increasingly sedentary lifestyles. Addressing the brain’s role in integrating hedonic and homeostatic signals offers a promising approach to curbing the global diabesity epidemic (Berthoud et al., 2017; Campos et al., 2022). Despite the distinction between hedonic and homeostatic feeding—where hedonic feeding is driven by pleasure and homeostatic feeding fulfilling basic metabolic needs—their overlapping neural circuits make them difficult to separate functionally and anatomically (Rossi and Stuber, 2018; Ahn et al., 2022).

The nucleus tractus solitarius (NTS), a complex structure in the dorsomedial medulla, plays a central role in integrative processes, including feeding. The NTS exhibits remarkable neurochemical diversity, with most neuroactive substances identified in the central nervous system also present in the NTS, where they function as classical neurotransmitters and/or neuromodulators. The NTS receives peripheral inputs from cranial nerves, notably vagal afferent projections, as well as signals from the glossopharyngeal, facial, and trigeminal nerves. Gustatory and somatic afferents from the oropharyngeal region map to the rostral NTS, while visceral afferents from the cardiovascular, digestive, respiratory, and renal systems terminate in the caudal NTS (Jean, 1991).

The NTS in turn maintains extensive connections with multiple central structures, directly projecting to regions involved in affective, appetitive, cardiorespiratory, and neuroendocrine functions (Beckstead et al., 1980; Geerling et al., 2006; Herbert et al., 1990; Holt, 2022; Loewy and Burton, 1978; Quillet et al., 2023; Rinaman, 2010; Horst et al., 1989). These connections include short pathways to bulbo-ponto-mesencephalic structures (e.g., the parabrachial nucleus, cranial nerve motor nuclei, ventrolateral reticular formation, and raphe nuclei) and longer pathways to the spinal cord, hypothalamus, and limbic structures (Jean, 1991).

Additionally, the NTS receives central top-down inputs, including direct projections from the cerebral cortex (Gasparini et al., 2020; Holt et al., 2019; Van Der Kooy et al., 1984; Whitehead et al., 2000). Most structures receiving projections from the NTS reciprocally project back, forming a closed-loop regulatory system. This design highlights the NTS’s pivotal role in autonomic and neuroendocrine functions, integrating somatic and autonomic responses to influence and regulate moment to moment behavior to and from all major organ systems.

Finally, depending on their topography within the NTS, distinct subpopulations of neurons can either potently reduce food intake (caudal ventral NTS) with or without the adverse hedonic affect, or increase hunger (rostral NTS neurons) (Gasparini et al., 2024). Interestingly, multiple NTS neuronal populations can cumulatively suppress food intake (Qiu et al., 2023), and oral, as well as visceral feedback to NTS can induce sequential appetite suppression (Ly et al., 2023).

The NTS overlaps sympathetic and vagal afferents (Tjen-A-Looi et al., 1997), both involved in type 2 diabetes pathophysiology (Zhang et al., 2024; Zsombok et al., 2024; Chen et al., 2024), as well as spinal cord connections through the cervical plexus (Stornetta et al., 2016). Therefore, optimal neuromodulation of the NTS through combined stimulation of the vagus nerve (caudal NTS) and the trigeminocervical complex (rostral NTS), as well as upper cervical spinal roots warrants assessment in optimizing clinical outcomes of type 2 diabetic patients and their associated co-morbidities—notably if mediated via minimally invasive modalities (Guglielmi et al., 2023). Interestingly, the outer ear is the only superficial organ anatomically innervated by the vagus cranial nerve (auricular peripheral branch nerve or ABVN), the trigeminal cranial nerve (via the auriculotemporal peripheral nerve or ATN), and the distal branches of the upper C2 and C3 cervical spinal roots (namely lesser occipital and greater auricular peripheral nerves respectively, LON and GAN) (Mercante et al., 2018). Moreover, aVNS stimulation alone is already a safe emerging minimally invasive neuromodulation therapy being explored in a wide range of diseases (Gerges et al., 2024; Farmer et al., 2021).

Given the interplay between autonomic dysfunction, metabolic imbalance, and chronic inflammation in T2DM, a neuromodulatory approach targeting both vagal and trigeminocervical pathways holds significant therapeutic promise. By simultaneously engaging these systems, this dual-modality stimulation may provide superior clinical benefits compared to single-pathway interventions.

Indeed, a longitudinal single center RCT by Madhuchander et Gurunath (NCT03540446; abstracts from the 22nd Annual Meeting of the North American Neuromodulation Society 2019) validated that auricular Percutaneous Electrical Neuro-Stimulation (PENS) stimulating the ATN, ABVN and LON peripheral cranial nerves could successfully improve DPN in all active-PENS patients who completed the 12 week-study (alternating stimulation on/off every 2 h—on a 1 week on/off basis). As the patients’ resilience in participating may explain this success (26 out of 89 patients dropped-out), we launched an observational study to question the efficiency of a similar combined peripheral stimulation, implanted during a shorter duration (less than 1 month), among the most severe, underserved DPN populations: Native Americans. Given the growing burden of T2DM and its complications, exploring innovative, non-pharmaceutical interventions is crucial. This pilot observational study pioneers the assessment of short-lived minimally invasive auricular neuromodulation in patients with severe DPN and T2DM.

Ninety-one Native American patients with severe diabetic peripheral neuropathy pain (meeting the inclusion/exclusion criteria) underwent implantation, without any significant AE reported through the 90-day follow-up. Five patients experienced mild postural intensity changes at the electrode implantation site that resolved completely within 24–48 h. One patient experienced topical sensitivity to the surgical dressing which resolved within 24 h of implantation when the dressing was changed.

Eight patients were lost to follow-up, so statistical analysis was conducted on the remaining 83 patients (male 34, Female 49). All 8 lost-to-follow-up were equivalent responders systematically through explant (day 19). However, during the follow-up period one patient moved off reservation, and the remote geography and limited resource infrastructure contributed to the remaining seven patients being lost to a combination of: an inability to communicate (loss of cell phone/coverage without landlines); an unknown change in contact information; a loss of all transportation; and an inability to overcome financial and/or untenable, unspecified family constraints.

Average improvement at 30-, 60-, and 90-days after application of the aVNS and TCC peripheral nerve stimulator were followed for up to 8 months, or an average of 227 days. Medication usage over time was assessed by a retrospective follow-up call to all 83 patients asking the same single yes or no question (“Are you off all of your pain and diabetes medications?”). Unfortunately, only 45 of the 83 actually (randomly) answered the phone and 36 of those answered affirmatively. Hence, among the random subset of 45 patients, 80% reported discontinuing their medications after receiving combined aVNS and TCC neurostimulation treatment. Both the pain (DPN) and blood sugar improvements decreased most significantly within the first 90 days, and then remained at those lower levels through the end of the follow up period—reported as an average of 7.85 months of all 83 patients.

DPN pain levels decreased, respectively, by 78, 83 and 87% at 30-, 60-, and 90-day follow ups (as shown in Table 1). As a consequence, the level and percent improvements in pain scores at 30, 60, and 90 days are remarkably statistically significantly different from baseline at the 5, 1, and 0.1% levels (Table 2). This dramatic 87% reduction of pain (from 7.92 on day 0 to 1.04 on day 90) is visually represented in Figure 2.

Regarding glycemia, glucose levels and percentage improvements for 30-, 60-, and 90-day follow-ups, respectively 19, 24, and 37%, are shown in Table 3. Then, Table 4 describes one-sample, two-sided student’s t-test results for the level and percent improvement in blood glucose measurements. In terms of both level and percentage improvements, results are statistically significant at the 5, 1, and 0.1% levels. Furthermore, by 90 days, the glucose mean was 121 mg/dL which is within normal glucose levels. As such, by 90 days, many patients no longer experienced hyperglycemia. Finally, Figure 3 displays the glycemic drop from 8.9% or 209 mg/dL on day 0 to 5.8% or 121 mg/dL at 90 days, a 37% overall reduction.

DPN is characterized by burning, tingling, or shooting pain that significantly impacts quality of life and increases healthcare utilization (Higgins et al., 2022). The medical management of DPN has primarily focused on symptom control and improved patient functionality. However, the complexity of this condition and variability in patient response necessitate a multimodal approach tailored to individual needs.

The first-line pharmacologic agents for managing DPN include antidepressants, such as serotonin-norepinephrine reuptake inhibitors (SNRIs) and tricyclic antidepressants (TCAs), as well as anticonvulsants like gabapentinoids (Rafiullah and Siddiqui, 2022; Attal, 2019). Duloxetine, an SNRI, has been widely studied and is approved by the U.S. Food and Drug Administration (FDA) for DPN (Rastogi and Jude, 2021). It works by modulating descending pain pathways in the central nervous system (CNS) (Boomershine et al., 2011). Similarly, pregabalin, a gabapentinoid, is FDA-approved for DPN and reduces neuropathic pain by binding to voltage-gated calcium channels, thereby attenuating neuronal excitability (Alles et al., 2020; Paul et al., 2022). Despite their efficacy, these agents often have dose-limiting side effects, such as dizziness, sedation, and gastrointestinal upset, which can reduce patient adherence (McCarberg et al., 2012).

In cases where first-line therapies fail to provide adequate relief, second-line agents, including opioids and topical treatments, are considered. Tapentadol, a dual-action opioid and norepinephrine reuptake inhibitor, has demonstrated efficacy in clinical trials but is associated with concerns about dependence and tolerance (Alshehri, 2023). Topical agents like 8% capsaicin patches and 5% lidocaine patches offer localized pain relief with minimal systemic side effects, making them particularly useful in patients unable to tolerate systemic medications (Elbeddini et al., 2024).

A new pharmacological class mimicking the actions of natural incretin hormones (principally glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP)), potentiating both meal-stimulated insulin and glucagon secretion (Drucker and Holst, 2023), arouses keen interest for the treatment of T2DM and obesity. Besides GLP-1 receptor agonists (A-GLP1), on the market since 2005, newer multi-agonist drugs that use one molecule to activate multiple receptors are drawing attention, like tirzepatide (that stimulates both the GLP-1 and the GIP receptor), currently approved for T2DM. Studies on these multi agonist drugs suggest that activation of multiple receptors is more effective than activating A-GLP1 alone (Karagiannis et al., 2024). They could even prevent the onset of T2DM in people with excess weight or obesity (Jastreboff et al., 2025). The results of the SURMOUNT-1 trial, recently published in The New England Journal of Medicine found that after more than 3 years, nearly 99% of all participants who took tirzepatide were able to prevent type 2 diabetes, and 94% of participants with prediabetes taking tirzepatide were able to reverse their prediabetes and return their glucose levels to normal. In addition, research demonstrated that participants taking tirzepatide saw up to 23% body weight reduction, increased levels of HDL cholesterol, and reductions in waist circumference, blood pressure, and triglycerides throughout the study.

These results are promising, with the caveat that tirzepatide and other drugs in the same class offer risks along with benefits and may not be suitable for everyone. Indeed, for instance, the increasing use of A-GLP1 has recently been linked to several side effects including not only gastrointestinal disorders, hypotension, syncope, arthritis disorders, nephrolithiasis, interstitial nephritis drug-induced pancreatitis (Xie et al., 2025), but also severe ophthalmological complications (Katz et al., 2025) compared to usual care. As retrospective studies cannot establish a causal link between exposure to A-GLP1 and the complications observed, if this link exists, it is likely to result rather from a rapid correction of hyperglycemia than from a direct toxic effect of these medications. By comparison, combined aVNS and TCC stimulation is likely to be a safer option since it takes 90 days to normalize glucose blood levels.

Spinal cord stimulation offers promising results in refractory cases (Wang et al., 2022; Henson et al., 2021) but is relatively costly and carries the risk of operating on the spinal column. Therapies such as cognitive-behavioral therapy (CBT) and physical therapy have demonstrated safety and varying degrees of effectiveness in improving pain and functionality (Higgins et al., 2022; Shi and Wu, 2023; Li et al., 2023; Akyuz and Kenis, 2014; Jahantigh Akbari et al., 2020), albeit with less pronounced benefits in DPN than in our pilot.

Indeed, notably through GLP-1 signaling, VNS by itself can improve T2DM, weight gain and DPN (Longo et al., 2023; Dai et al., 2020; Dhanapalaratnam et al., 2024). It is well-recognized that VNS, with appropriate parameters, can improve diabetes mellitus in animal models (Payne et al., 2022; Waataja et al., 2023). The vagus nerve, a key regulator of glucose homeostasis, promoting insulin secretion and improving tissue insulin sensitivity by counteracting sympathetic overactivity and reducing systemic inflammation (Sorski and Gidron, 2023)—a major driver of insulin resistance and DPN, is both anatomically (Xiong et al., 2023) and functionally (Yu et al., 2020) down-impacted in T2DM patients with DPN. Additionally, VNS is known to reduce food intake and body weight by delaying gastric emptying. Interestingly, diet-induced shift in the gut microbiome is likely to directly induce DPN, according to a recent bi-directional two sample Mendelian randomization study (Xie et al., 2024). Given that electrical stimulation of the AVBN is able to both change gut microbiota composition and improve “abdominal pain, functioning and catastrophizing” in adolescents (Castillo et al., 2023) as well as to activate enteric neurogenesis and synaptic efficiency in a rodent model (Soylu et al., 2025), it is likely that a limited duration of stimulation of this essential component of the microbiota-gut-brain axis is sufficient to long lastingly reshape and optimize neural networks involved in metabolism and body weight (Obst et al., 2022; Vaughn et al., 2017), notably via the endocannabinoid system (Wood et al., 2024; DiPatrizio, 2021; Patterson et al., 2016). Finally, it is not surprising that VNS could also improve DPN in our trial since it can stimulate VEGF-A synthesis (Lv et al., 2018) and even promote oligodendrogenesis, facilitating the restoration of myelin sheaths in animal models (Huang et al., 2024).

Transcutaneous auricular VNS (taVNS) seems efficient by itself, since a single session of 1 h-stimulation was able to improve both neuropathic pain (subjectively) and HRV (objectively) (Wong et al., 2025). In this controlled pilot trial, the sham stimulation was applied on the ear lobe, the latter not being supplied by the auricular branch of the vagus nerve but by the GAN and LON, originating from the C2 and C3 spinal nerves (Peuker and Filler, 2002; Mercante et al., 2018). Surprisingly and very interestingly, all symptoms decreased post active and sham taVNS, with greater mean change for the active condition, but the difference was not statistically significant between conditions (p values >0.11 across all symptoms). HRV increased during sham taVNS, a stimulation of the trigeminocervical complex, indicating that this was not a true sham stimulation, but an active stimulation of a portion of the TCC complex.

Remarkably, our findings surpass outcomes reported by a recent longer randomized controlled study assessing the efficiency of transcutaneous cervical VNS alone on continuous glucose monitoring in 145 diabetic patients with DPN (including 75 T2DM) for either 1 week (4 daily stimulation of 2 min) or 8 weeks (2 daily stimulations of 2 min) (Kufaishi et al., 2025). These discrepancies of efficiency between the two noninvasive vagus nerve stimulation techniques (i.e., cervical and auricular) could possibly result only from the differences between the protocols used, corroborated by the significant difference in number of vagal fibers between the ear and the neck.

Nevertheless, targeting both the vagus and trigeminal nerve pathways, as in our trial, is likely to yield a more robust therapeutic effect. This synergy may be attributed to the broader engagement of central autonomic and inflammatory control centers. The concurrent stimulation of the TCC may further amplify these effects, as this pathway interacts with the NTS, a central hub for autonomic and metabolic regulation. Notably, cross-talk between vagus and trigeminal nerve stimulation have been reported (Möller et al., 2020; Peng and May, 2022). Both VNS (Brougher et al., 2021; Diepenbroek et al., 2024; DiPatrizio, 2021) and TCC stimulation (Xu et al., 2023; Leimuranta et al., 2018; Zorrilla et al., 2024) can modulate human motivation and behaviors by impacting the dopaminergic and endocannabinoid systems. Moreover, stimulating the TCC could counteract the impact of the estrous cycle on food intake in females without modifying their hormonal status by potentiating the estrogen-dependent signaling (Benchoula et al., 2024; Klen and Wolkoff, 2004) involved in both antinociception (Spekker et al., 2022; Nag and Mokha, 2016) and satiety signal (Huang and Raybould, 2020 for a review).

Thus, combining trigeminocervical and vagal neuromodulation stimulates multi-level interplaying closed loops, potentiating their single action, which might explain the impressive changes observed in our observational study, long after stopping the stimulation. There are two closed loop feedback loops noted. The first is the natural closed loop physiology that is being activated by the combined TCC and VNS within the brainstem and more precisely in the NTS. That is, by stimulating these three entries (auricular vagus nerve, trigeminal and cervical neural complexes) instead of one (the vagus nerve alone), we take advantage of the innate closed loop to regulate more efficiently and for a long-term period blood sugar level and DPN (discussed in the discussion). The second is the proprietary closed loop impedance drop software platform referenced above within the PTU (Figure 1) that objectively localizes each of the individual nerves of interest to be implanted. However, as our combined stimulation of aVNS and TCC may only partly induce the large size improvements reported due to several confounding factors, (namely the placebo effect, the regression to the mean, concomitant care changes including medication during the study, the natural history of the cohort), further double-blind, controlled, randomized trials are definitely needed to assess the potential causal inference.

This dual-pathway approach also addresses the underlying vascular and bioenergetic dysfunction associated with DPN. By enhancing mitochondrial efficiency and restoring vascular endothelial growth factor (VEGF-A) expression, neuromodulation improves peripheral nerve function and reduces pain. The rapid reduction in neuropathic pain observed in our analysis, along with sustained glucose control, may reflect the synergy between metabolic and neurological improvements.

Provided this condition is met, neurostimulation could become transformative in T2DM associated with obesity management, for cost-effectiveness purposes. Diabetes is a major healthcare crisis affecting more than 422 million people worldwide with 38 million in the USA—15% of which have painful diabetic neuropathy. Painful DPN management alone increases both the direct cost of treatment up to threefold, and leads to work absence, change in employment and disability (Tesfaye et al., 2010). Furthermore, only 32.15% (8.2 million) of 25.5 million adults with diagnosed diabetes are employed. This equates to $28.3 billion in lost employment, $35.8 billion from presenteeism, $32.4 billion due to premature mortality, $5.4 billion in work absences, and $4.4 billion from lost productivity. At these current trends, the overall cumulative cost of diabetes, pre-diabetes and its associated co-morbidities will consume all national healthcare expenditures within the next 5–10 years (Dyess, 2023; Lin et al., 2018; Cawley et al., 2021). Remarkably, the minimally invasive nature of auricular neurostimulation offers a safer, more accessible and cost-effective therapeutic alternative, potentially increasing patient adherence and treatment scalability. The significant improvements in pain and glucose control—which lasted an average of 227 days after only 19 days of treatment (in comparison to lifelong drug treatment)—along with the associated physical and mental relief, may support reintegration into the workforce, thereby reducing both personal, mental health, and societal economic burdens related to DPN and T2DM. Examining the cost-effectiveness of this therapy, particularly in populations disproportionately affected by T2DM, such as Native American communities, could support its broader adoption and has to be assessed in future double-blind, randomized, placebo-controlled trials, using lab-based HbA1c, validated diabetes control questionnaires, and rigorously tracking medication changes ideally with prescribing data and structured interview.

The significant improvements observed in our observational study include an 87% reduction in pain, a 37% decrease in mean blood glucose, a reduction in HbA1c from 8.9 to 5.8% and a random subset that discontinued 80% of all pain and diabetes medications; all sustained for an average of 7.85 months; surpassing prior outcomes achieved with cervical VNS alone.

Thus, this open label pilot positions combined minimally invasive aVNS and TCC stimulation as a transformative, non-pharmacologic alternative capable of addressing both the DPN and diabesity crisis, as well as the intertwined pathophysiology of diabetes, neuropathy, and metabolic dysfunction. As such, this approach has the potential to restore autonomic balance and redefine the standard of care for diabetic patients by reducing the long-term societal, mental health, and economic burdens of the disease. Further double-blind randomized controlled trials are needed to confirm, better elucidate the mechanisms and biomarkers (i.e., Heart Rate Variability, Laser Evoked Potentials), and establish standardized protocols for broader clinical application.