Autonomic responses to proprioceptive and visual errors during single-trial reach adaptation

Recent evidence suggests that not only the implicit but also explicit cognitive learning processes play a significant role in acquiring sophisticated motor skills. How these processes interact with each other, or how we can quantify them without bias, are, however, still elusive. To tackle these issues, here we employ simultaneous measurement of peripheral autonomic signals during the short-term motor learning paradigm. Through extensive research in the cognitive learning field, autonomic measures, such as pupil diameter, have been shown to reflect various internal states associated with the adjustment of learning behavior, suggesting that they could also be an effective tool to assess cognitive processes during motor learning. In a series of experiments, we measured the multiple autonomic signals, including pupil diameter, skin conductance, and heart rate, while human participants of both sexes learned to reach under occasional proprioceptive or visual disturbances. We characterized the phasic autonomic responses to errors and evaluated their influence over motor learning, which was quantified by comparing the trials immediately before and after the error event. The results demonstrated the dose-dependent increase in the phasic autonomic responses to errors of both modalities, consistent with higher cognitive demand in large errors. Using a latent factor analysis, which combines the multimodal autonomic response data, we also found a statistical relationship between the latent autonomic states and the motor learning rate, the suppression of implicit motor adaptation at higher error-induced sympathetic state. These results provide a novel insight into how internal state change affects motor learning.