After a prolonged period stepping in circles, people walk with a curved trajectory when attempting to walk in a straight line without vision. Podokinetic adaptation shows promise in clinical populations to improve locomotor turning, however the adaptive mechanisms involved are poorly understood. The first phase of this study asks: how does the podokinetic conditioning velocity affect the response velocity and how quickly can adaptation occur? The second phase of the study asks: can a mathematical feedback model account for the rotation trajectories across different conditioning parameters and different datasets?

Twelve healthy participants stepped in place on the axis of a rotating surface ranging from 4 to 20 deg s^{−1 for durations of one to ten minutes, while using visual cues to maintain a constant heading direction. Afterward on solid ground, participants were blindfolded and attempted to step without rotating.}

Participants unknowingly stepped in circles opposite to the direction of the prior platform rotation for all conditions. The angular velocity of this response peaked within one minute and the ratio of the stimulus‐to‐response peak velocity fitted a decreasing power function. The response then decayed exponentially. The feedback model of podokinetic and vestibular adaptive processes had a good fit with the data and suggested that podokinetic adaptation is explained by a short (141 s) and a long (27 min) time constant.

The podokinetic system adapts more quickly than previously thought and slower rotation is more readily adapted than faster rotation. These findings will have implications for clinical applications of the technique.