A wearable cryothermal–vibrotactile neuromodulation system for the body side of depression and anxiety.
Depression and anxiety are not only states of mind. They show up in the body — in heart rate variability, skin conductance and breathing rate — and those signals rarely settle just because someone has been told to relax.
Standard care works, slowly and unevenly: therapy waitlists, SSRIs that take weeks to reach full effect and carry side effects, benzodiazepines with dependence risk. What is missing is something titratable, on-demand and non-pharmacological for the autonomic half of the problem.
This is not an argument against medication or therapy. It is an argument for a third, physiological lever — one a person could hold in their own hands, on a bad afternoon, without waiting for an appointment.
Illustrative model of the intended direction of effect. Not patient data.
Vibration on the tragus — the small cartilaginous flap at the front of the ear, carried almost entirely by the auricular branch of the vagus (ABVN). Surface contact only; nothing is inserted.
Afferent signals reach the nucleus tractus solitarius. Efferent vagal fibres route through the splenic nerve toward the celiac ganglion.
Acetylcholine binds α7 nicotinic receptors on macrophages, suppressing NF-κB and downstream TNF-α, IL-1β and IL-6 — inflammatory signalling repeatedly reported in a harder-to-treat subtype of depression.
The splanchnic–celiac route is densest mid-thoracic — hence a T1–T12 array. Drive stays deliberately sub-threshold: strong sympathetic stimulation does the opposite of what we want.
The cholinergic anti-inflammatory pathway is well characterised for electrical VNS. Vibrotactile evidence at this site is earlier and thinner — which is exactly what Aim 2 is scoped to test.
Every claim in this project traces to something published. Where the evidence runs out, the proposal says so.
Implanted vagus nerve stimulation suppressed cytokine production and reduced disease severity in rheumatoid arthritis — proof that vagal stimulation moves human immune signalling.
Vibrotactile stimulation on this nerve branch reduced TNF, IL-6 and IL-1β production and disease severity in a small rheumatoid arthritis cohort — a neighbouring patch of the same branch; we target the tragus.
Bilateral auricular stimulation reduced fatigue and depression scores alongside increased heart rate variability in a 2025 clinical trial.
Eleven studies, more than 3,000 participants: significant reductions in depressive symptoms, anxiety and perceived stress.
During immersion, cold-water exposure produced a measurable decrease in HRV-indexed vagal tone — so the mood benefit may not be vagally mediated.
Low-frequency sound and vibration (20–120 Hz) associated with parasympathetic-shift physiology in pilot work.
Cooling demonstrably helps mood. Whether that runs through vagal activation is contested — and the 2026 HRV finding argues against the simple story. The cooling subsystem's mechanism is therefore treated as an empirical question, not a premise.
The frame itself is unremarkable printed TPU. The cold comes from two 10 mm braided copper straps running its length: four Peltier modules chill the copper, and the copper carries that cold straight to the skin.
Eight nodes run down the middle of the frame, with four more attached outboard to widen coverage. Spacing stays above the two-point tactile discrimination threshold for dorsal skin, so each node reads as a distinct point.
Each channel reflects a distinct physiological system rather than repeating the same signal. Convergence is the evidence, not any single number.
Safety-critical cutoffs sit outside firmware. A software fault cannot defeat them.
A simulated ramped session across all four channels. Watch for convergence — not a single dramatic number.
Simulated for illustration. Real values come from the pilot described in section 06 — no fabricated subject data is presented anywhere on this site.
All four measures at rest, plus STAI-6, GAD-7 and PHQ-9. No device activity.
Cooling and vibration rise from minimum intensity over several seconds. Never sudden onset.
Brief subjective check at the session midpoint.
If GSR rises or HRV falls, intensity is reduced immediately — stress-response trajectory aborted.
Repeat measures and all three questionnaires. Sham condition runs on a separate occasion.
8–10 participants, each comparing device-on against a short sham condition, with state anxiety and mood measured before and after. Within-subject design maximises evidentiary value per participant; this is pilot data, not a powered trial.
Elevated IL-6 and TNF-α are reported in a subset of people with depression. PMA-differentiated THP-1 cells, LPS-stimulated at 100 ng/mL, are treated with carbachol and norepinephrine as proxies for vagal and sympathetic activation, then assayed by ELISA at 6 and 24 hours.
Institutional review pre-approval before any data collection; informed consent; participants told the device is investigational. Anyone screening positive for severe depression or suicidal ideation is excluded and referred to clinical care.
No current to the body. Hardware thermal cutoff independent of software, enforced session maximum, physical kill switch.
THP-1 line, BSL-1, no human tissue, supervised laboratory setting.
No vertebrate animal experiments are proposed.
A complete annotated bibliography is available on request.