Differential effects of vision upon the accuracy and precision of vestibular-evoked balance responses

Vestibular information must be transformed from head-to-foot-centred coordinates for balance control. This transformation process has previously been investigated using electrical vestibular stimulation (EVS), which evokes a sway response fixed in head coordinates. The craniocentric nature of the response has been demonstrated by analysing average responses to multiple stimuli. This approach misses any trial-by-trial variability which would reflect poor balance control. Here we performed single-trial analysis to measure this directional variability (precision), and compared this to mean performance (accuracy). We determined the effect of vision upon both parameters. Standing volunteers adopted various head orientations (0, +/- 30 and +/- 60 deg yaw) while EVS-evoked response direction was determined from ground reaction force vectors. As previously reported, mean force direction was orientated towards the anodal ear, and rotated in line with head yaw. Although vision caused a similar to 50% reduction in response magnitude, it had no influence on the direction of the mean sway response, indicating that accuracy was unaffected. However, individual trial analysis revealed up to 30% increases in directional variability with the eyes open. This increase was inversely correlated with the size of the force response. The paradoxical observation that vision reduces the precision of the balance response may be explained by amulti-sensory integration process. As additional veridical sensory information becomes available, this lessens the relative contribution of vestibular input, causing a simultaneous reduction in both the magnitude and the precision of the response to EVS. Our novel approach demonstrates the importance of single-trial analysis in revealing the efficacy of vestibular reflexes.