0098 Local Spindle Increase is Correlated with Sleep-Dependent Memory Consolidation of Motor Sequence Task

Abstract

Introduction: There is converging evidence that motor-skill learning is consolidated by sleep spindles during non-rapid eye movement stage 2 (N2) sleep, resulting in improved post-sleep performance. We localized learning-related changes in spindle activity with source localization of simultaneous electroencephalogram (EEG) and magnetoencephalography (MEG) recordings of motor performance and the sleep that followed to identify cortical regions involved in consolidating a finger tapping motor sequence task (MST). Methods: Fifteen healthy individuals (age 29.2 ± 4.6 years, 11 males) had a structural MRI and underwent three EEG/MEG sessions, including an adaptation, a non-learning baseline and an MST nap. During the MST visit, participants trained on the MST with their left hand and were tested after a 90 min nap. MST task-activated areas were identified based on mu (8-13 Hz) suppression during tapping compared to rest. We reconstructed cortical source signals during N2 sleep using MRI-anatomically constrained dynamic statistical parametric mapping, and localized spindles in the source space. We then compared spindle density during baseline and MST naps, and the relations of spindle changes to overnap MST improvement. Results: MST-training activated different cortical areas than postsleep MST-testing. While training activated bilateral premotor, primary motor (M1) and primary sensory (S1) cortices, testing activated left M1 and right parietal and prefrontal cortex. Compared to the baseline nap, during the MST nap spindle density increased significantly in MST task-activated (training + testing) areas. In addition, in two regions showing MST activation, pre-supplementary motor area (pre-SMA) and dorsal anterior cingulate cortex (dACC), the change in spindle density between MST and baseline visits correlated with overnap improvement. In contrast, in right premotor and S1, the spindle density change inversely correlated with overnap improvement. Conclusion: Motor learning enhances N2 spindle activity in cortical regions involved in task performance and these changes predict sleep-dependent performance improvement. In the right hemisphere, spindle increases in pre-SMA and dACC are associated with task improvement, while spindle increases in premotor and S1 are associated with decreased improvement. Finally, sleep-dependent motor-skill consolidation changes the network involved in task performance.