The sense and control of standing balance in the presence of motor noise

This study aimed to characterize motor noise in human standing balance and uncover mechanisms that enable the nervous system to robustly sense and control upright posture despite this variability. We conducted three experiments using a robotic balance simulator. First, we quantified the natural variability of ankle torques, revealing that torque variability was stable within preferred postures and increased only at more extreme orientations. The range of constant variability may be explained by passive mechanisms that contribute to plantarflexion torque along with the bilinear scaling of motor noise with torque. Together, these factors help maintain stable motor variability, despite the substantial increase (6% of mgL) in gravitational torque load. Second, we investigated perceptual thresholds for detecting the sensory consequences of artificially imposed torque noise, designed to replicate and amplify the variability of ankle torque during quiet standing. Participants detected the consequences of imposed torque when it exceeded ~50% of baseline torque variability, which was accompanied by a two-fold increase in whole-body angular velocity variability. Third, we assessed the impact of artificially imposed torque noise on the vestibular control of balance. We found that the threshold for generating corrective responses remained unchanged, as increased variability in balance-correcting responses was accompanied by proportional increases in vestibulomotor gain. Collectively, these findings reveal that the nervous system accommodates motor noise by leveraging passive stiffness and the minimal scaling of motor noise at ankle torques near preferred postures, while also engaging robust error correction mechanisms – such as increased vestibulomotor gain – that operate outside conscious awareness.