Abstract
Oscillatory neural activity during sleep, such as that in the delta and sigma bands, is important for motor learning consolidation. This activity is reduced with typical aging, and this reduction may contribute to aging-related declines in motor learning consolidation. Evidence suggests that brain regions involved in motor learning contribute to oscillatory neural activity during subsequent sleep. However, aging-related differences in regional contributions to sleep oscillatory activity following motor learning are unclear. To characterize these differences, we estimated the cortical sources of consolidation-related oscillatory activity using individual anatomical information in young and older adults during non-rapid eye movement sleep after motor learning and analyzed them in light of cortical thickness and pre-sleep functional brain activation. High-density electroencephalogram was recorded from young and older adults during a midday nap, following completion of a functional magnetic resonance imaged serial reaction time task as part of a larger experimental protocol. Sleep delta activity was reduced with age in a left-weighted motor cortical network, including premotor cortex, primary motor cortex, supplementary motor area, and pre-supplementary motor area, as well as non-motor regions in parietal, temporal, occipital, and cingulate cortices. Sleep theta activity was reduced with age in a similar left-weighted motor network, and in non-motor prefrontal and middle cingulate regions. Sleep sigma activity was reduced with age in left primary motor cortex, in a non-motor right-weighted prefrontal-temporal network, and in cingulate regions. Cortical thinning mediated aging-related sigma reductions in lateral orbitofrontal cortex and frontal pole, and partially mediated delta reductions in parahippocampal, fusiform, and lingual gyri. Putamen, caudate, and inferior parietal cortex activation prior to sleep predicted frontal and motor cortical contributions to sleep delta and theta activity in an age-moderated fashion, reflecting negative relationships in young adults and positive or absent relationships in older adults. Overall, these results support the local sleep hypothesis that brain regions active during learning contribute to consolidation-related neural activity during subsequent sleep and demonstrate that sleep oscillatory activity in these regions is reduced with aging.
Highlights
Sleep is important for consolidating motor sequence learning
Young adults had increased contributions to theta activity during the nap from multiple cortical regions compared to older adults, which can be conceptually grouped into three clusters (Figures 3, 4 and Supplementary Table 4): (1) a left-weighted motor cortical network consisting of caudal superior frontal gyrus, caudal middle frontal gyrus, left inferior frontal gyrus, left precentral gyrus, and paracentral lobule, (2) a prefrontal network consisting of rostral superior frontal gyrus and rostral middle frontal gyrus, and (3) cingulate gyrus
Young adults had increased contributions to sigma activity during the nap from multiple cortical regions compared to older adults, which can be conceptually grouped into three clusters (Figures 3, 5 and Supplementary Table 5): (1) left precentral gyrus, (2) a right-weighted prefrontal-temporal network consisting of rostral superior frontal gyrus, rostral middle frontal gyrus, inferior frontal gyrus, lateral orbitofrontal cortex, medial orbitofrontal cortex, frontal pole, right temporal pole, right superior temporal gyrus, and right insula, and 3) cingulate gyrus
Summary
Sleep is important for consolidating motor sequence learning. Performance improves over a night of sleep or a nap, but not an equivalent interval of wakefulness (Fischer et al, 2002; Walker et al, 2002; Backhaus and Junghanns, 2006; Korman et al, 2007; Doyon et al, 2009b; Debas et al, 2010). Non-rapid eye movement (NREM) sleep features, including oscillatory neural activity in the delta frequency band (0.5–4 Hz; e.g., slow waves, K-complexes), in the theta frequency band (4–8 Hz), and in the sigma frequency band (12–16 Hz; e.g., sleep spindles), have been linked to motor learning consolidation during sleep (Nishida and Walker, 2007; Tucker and Fishbein, 2009; Barakat et al, 2012; Albouy et al, 2013; Tamaki et al, 2013; Fitzroy et al, 2021a). Consolidation involves synaptic changes in learning-related cortical circuits, and learning induces local, use-dependent changes in sleep oscillatory neural activity apparent in the scalp electroencephalogram (EEG) (“local sleep”; e.g., Huber et al, 2004). Source estimation, a method of identifying the neural generators of the scalp-recorded EEG and near-scalprecorded magnetoencephalogram (MEG), suggests that regions involved in learning contribute to subsequent delta and sigma activity. Increased slow wave activity is observed in premotor cortex during NREM sleep following visuomotor learning (Murphy et al, 2011), and increased slow sigma activity is observed in visual cortex during NREM sleep following visual perceptual learning (Bang et al, 2014)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
