Predictive encoding of pain location in the motor system indexed by somatomotor alpha and corticospinal excitability

The pain network must not only respond to pain but also predict its occurrence to proactively guide defensive response. Yet how pain-related predictions are learned and updated in the nervous system remains unclear. Using a Pavlovian threat conditioning task, participants learned that two different visual cues predicted painful shocks either to the left or right arm, while a third cue never predicted pain. After acquisition, cue-pain contingencies were reversed to assess prediction updating. Neural and autonomic responses were recorded during pain anticipation (i.e. cue presentation) using electroencephalography, skin conductance response, and double-coil transcranial magnetic stimulation to measure corticospinal excitability both contralateral and ipsilateral to the anticipated pain location. We found that skin conductance response increased bilaterally and irrespective of the predicted pain location, reflecting a general state of threat preparedness by the autonomic system. In contrast, the suppression of somatomotor alpha power and reduction of corticospinal excitability were lateralized and functionally coupled, suggesting precise motor tuning to the predicted pain location. During reversal, an increase in frontocentral theta power emerged, supporting the updating of previously learned pain predictions. Connectivity analyses further corroborated this, revealing dynamic interactions between somatomotor and frontocentral oscillatory activity. Together, these findings advance the understanding of the neurocognitive mechanisms of nociception by showing that pain anticipation is subserved by dissociable yet interacting networks supporting general threat detection and context-sensitive defensive response.