Haptic feedback goes on the forearm or wrist for a reason that has nothing to do with physiology: that is where people already wear things. The arm is “the most common haptic wearable site” largely because the watch got there first.
A paper from Yoon Kyung Shon and Pablo E. Paredes at the University of Maryland, submitted on 15 October, tests what that convention costs. The answer, for one specific application, is everything.
Two questions nobody had separated
Transcutaneous electrical stimulation for stress modulation has concentrated almost entirely on vagus nerve stimulation at the neck, for an anatomical reason — proximity to the cervical vagus nerve. The authors identify two things that research left unexamined:
- How users actually experience electrotactile sensation at the neck, a sensitive site
- Whether electrotactile feedback by itself can work as a haptic stress intervention
The second is the sharper question. If stimulating the neck reduces stress, is that the vagus nerve being engaged, or is it just that a pleasant buzzing sensation is calming wherever you put it? Those have completely different design implications, and you cannot tell them apart by studying the neck alone.
The design
An 8-node TES array was applied to two body sites during a cognitively demanding 2-back task:
- The neck — used in prior vagal TES research for its proximity to the cervical vagus nerve
- The arm — the common haptic wearable site, which “primarily provides cutaneous electrotactile sensation without vagal access”
The arm is the control for the mechanism. It delivers the tactile experience without the anatomical access, so if the effect is sensory rather than neural, the arm should reproduce it.
Two studies ran. A usability study assessed how the stimulation felt at each site. A physiological study compared active versus inactive stimulation under identical electrode attachment — meaning the electrodes were on the participant either way, and only the current differed.
That second control is the one that makes the result worth reporting. Wearing an electrode array on your neck is itself an experience; comparing active neck stimulation against nothing at all would confound the current with the apparatus.
What they found
Usability: both sites were rated comparably comfortable, and neither disrupted task performance. But the same stimulation was “interpreted differently across sites” — identical current, different subjective reading depending on where it landed.
Physiology: only neck TES produced cortical and autonomic changes consistent with stress modulation. Arm TES “did not differ from inactive conditions.”
So the arm delivered a sensation that was comfortable, noticeable and interpreted — and physiologically inert.
Why this is a useful negative result
Two conclusions, and the first is the surprising one.
The neck is acceptable. The assumption that users will not tolerate electrical stimulation on the throat appears to be wrong — it rated as comfortable as the arm, without hurting performance on a demanding task. For a site that gets avoided largely on the presumption of squeamishness, measuring it rather than assuming is worth a paper on its own.
The effect is not the sensation. Because the arm produced a comparable subjective experience and no physiological change, the neck result is not reducible to “a nice buzz is relaxing.” The authors’ framing is that “both the physiological effects and the sensory experience of TES depend on placement.”
That is bad news for a plausible and attractive product idea. A stress-reduction band on the wrist, using electrotactile patterns, would be easy to build and easy to sell, and this study says the mechanism it would claim is not available at that site. You would be shipping a sensation.
The caution worth keeping
The honest reading cuts in a second direction too. A cortical and autonomic change “consistent with stress modulation” is a physiological signature, not a demonstration that anyone felt better or performed better under stress over time. The 2-back task establishes that stimulation does not interfere with cognitive work; it does not establish a therapeutic outcome.
And the generalisation is about placement, not about devices. Eight nodes, two sites, one task, one lab — what it licenses is “test your site” rather than “put it on the neck.”
For anyone building body-worn interfaces, that is still the transferable lesson, and it is more general than TES. The sites we put wearables on were chosen by watch straps, headphone bands and phone pockets. When the thing you are delivering is information, that convention is fine. When it is a physiological intervention, the anatomy gets a vote, and this is a clean demonstration of what happens when it is not consulted.
The paper is 25 pages with six figures, and is due to appear in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, Vol. 10, No. 4, December 2026.
Sources
- ElectroCalm: Investigating User Experience and Physiological Responses to Multi-Node Transcutaneous Electrical Stimulation for Stress Modulation — arXiv:2610.22748
- arXiv PDF
- DOI: 10.1145/3857961