Abstract
Lightly touching a solid object reduces postural sway. Here, we determine the effect of artificially modifying haptic feedback for balance. Participants stood with their eyes closed, lightly gripping a manipulandum that moved synchronously with body sway to systematically enhance or attenuate feedback gain between +2 and -2, corresponding to motion in the same or opposite direction to the body, respectively. This intervention had a systematic effect on postural sway, which exhibited an asymmetric u-shape function with respect to haptic feedback gain. Sway was minimal around zero gain, corresponding to a static object. Sway increased slightly at gains below -0.25 but increased greatly at gains above +0.25. At +2, it was approximately double that of a no-touch condition. Mean interaction force between the hand and manipulandum remained < 0.9 N throughout, although it increased slightly at extreme gains. Cross-correlations between hand force and trunk position were highest during conditions of least sway, suggesting that higher quality haptic feedback is associated with greater sway reduction. We successfully replicated the sway behaviour using a feedback control model that attenuated haptic feedback signals when the discrepancy between haptic and proprioceptive signals reached a threshold. Our findings suggests the CNS can utilise augmented haptic feedback for balance, but only with relatively small changes to natural feedback gain. In healthy volunteers, it offers minimal benefit over a static object. Haptic feedback is therefore optimal when motion is physiologically realistic and subtle enough to be misinterpreted as self-motion.