Even with exposure to errors, motor imagery cannot update internal models

Recent evidence suggests that motor imagery is insufficient for updating internal models, essential for predicting and refining overt movement outcomes. The covert nature of motor imagery limits exposure to errors, perhaps preventing the updating of internal models. To explore this, 90 participants were exposed to a prism that shifted vision leftward, completing 20 physical pointing trials followed by either 230 more physical pointing trials [physical practice (PP)], 230 imagined pointing trials [physical practice motor imagery (PP-MI)], 230 unrelated task trials [physical practice control (PP-CTRL)], or no further trials [physical practice none (PP-None)]. We hypothesized that if exposure to errors is needed for motor imagery to update internal models, then PP-MI would exhibit aftereffects, characterized by pointing opposite to the prism shift (i.e., rightwards), similar to PP, but differing from PP-CTRL and PP-None. After prism exposure, all groups showed significant aftereffects (PP: 4.73° ± 2.12°, PP-MI: 2.62 ± 1.61, PP-CTRL: 2.58 ± 1.53, PP-None: 3.11 ± 1.68), however there were no significant differences in the magnitude of aftereffects between PP-MI, and PP-CTRL/PP-None. Our findings demonstrate that motor imagery alone is insufficient for updating internal models, even when participants are initially exposed to errors under a prism shift. This further reinforces that motor imagery is not a direct simulation of overt movement, as proposed by Motor Simulation Theory– the foundation for its use in rehabilitation. Deepening our understanding of how learning occurs through motor imagery is crucial for enhancing its effectiveness in practical applications like rehabilitation.