When "noise" isn't simply noise: deterministic postural drive during noisy galvanic vestibular stimulation

Age- and disease-related vestibular decline contributes to dizziness and postural instability, motivating interventions, such as noisy galvanic vestibular stimulation (nGVS). nGVS is often applied at imperceptible amplitudes and interpreted through a stochastic resonance (SR) mechanism, but because galvanic stimulation directly modulates vestibular afferents, even imperceptible currents may deterministically influence balance. The present study tested whether low nGVS stimulus amplitudes (<1 mA), commonly used in stochastic resonance paradigms, directly drive postural behavior. Twenty healthy young adults stood on a force plate with feet together and eyes closed under two surface conditions (rigid and 10-cm foam). Participants completed 300-s trials with band-limited (0-30 Hz), zero-mean nGVS at ±0, 0.1, 0.2, 0.3, 0.5, and 0.7 mA in randomized order. Coupling between the stimulation waveform and mediolateral ground-reaction force was quantified using coherence and time-cumulant density. Group mean coherence became significant primarily at higher amplitudes (0.5-0.7 mA) for both surfaces, while time-cumulant density revealed significant time-locked vestibular-evoked response components beginning at much lower amplitudes (down to 0.1 mA), including an early component (∼135-155 ms) and a later prominent component (∼360-410 ms). Individually, significant coherence was prevalent at 0.5-0.7 mA (15-19 of 20 participants, depending on surface and amplitude), whereas cumulant-based responses were observed in a subset as low as 0.1 mA. Responses were more evident on foam, consistent with increased vestibular reliance under reduced somatosensory reliability. These findings show that low-amplitude nGVS can produce stimulus-locked postural output, implying that balance changes during imperceptible nGVS amplitudes may reflect deterministic vestibular drive, SR-like facilitation, or both.NEW & NOTEWORTHY Low-amplitude noisy galvanic vestibular stimulation (nGVS), including levels often imperceptible to the participant, can drive postural output. Time-cumulant analysis detected stimulus-locked vestibular-evoked responses as low as 0.1 mA, whereas coherence became significant mainly at 0.5-0.7 mA, with stronger effects on foam. These findings suggest that balance changes during nGVS may not be attributable to stochastic resonance alone and highlight the value of measuring stimulus-response coupling alongside behavioral outcomes.