Does extensive motor learning trigger local sleep?

After prolonged learning we all have experienced a reduction of alertness, resulting in errors that we would normally not make. Despite this being a common situation in everyday life, the reasons for this phenomenon are unclear. A possible explanation is that the regions of the brain which are involved in the learning, go off-line trying to partially recover. This event is defined as local sleep and it has been detected in animals and sleep-deprived humans performing learning tasks. Local sleep is a sleep-like electrophysiological activity occurring locally, while the rest of the brain is fully awake, and producing performance deterioration. However, since all the studies included both lack of sleep and learning, it is uncertain whether such phenomenon is related to sleep deprivation or if it is the consequence of prolonged learning. Further, local sleep has not been related to electrophysiological changes occurring during the task. This thesis aimed to assess, for the first time in well rested subjects, whether local sleep and performance decline occur because of prolonged learning. Specifically, the goal was to discriminate between sustained practice and learning, as to determine whether learning is required to cause local sleep. Also, a 90-minute nap was evaluated to establish whether sleep is necessary to counterbalance neuronal fatigue and performance decrease. The starting hypothesis was that local sleep is a plasticity-related phenomenon affecting performance and requiring learning to be triggered. Consequently, sleep would be a prerequisite to counterbalance performance and electrophysiological changes. High-Density EEG and behavioral data of 78 healthy young subjects were collected during and after two learning tasks performed for three hours: a visual sequence learning task, and a visuo-motor rotation task, randomly selected. Afterward, subjects were divided in two groups: those who slept for one hour and a half and those who remained awake and quietly rested for the same amount of time before being tested for electrophysiological and behavioral changes. Moreover, to discriminate between the effects of prolonged learning and practice, 11 additional subjects performed a control condition consisting in planar upper limb reaching movements instead of the above-mentioned learning tasks. In detail, the power spectrum of the EEG activity during the task and at rest with eyes opened was divided into five ranges to determine frequency changes of the EEG activity: delta 1 to 4 Hz; theta 4 to 8 Hz; alpha 8 to 13 Hz, beta 13 to 25 Hz, gamma 25 to 55 Hz. Additionally, movement-related beta activity of 35 young subjects was analyzed to find a relationship between task related oscillations and performance indices, as the modulatory activity during practice may reflect plasticity-related phenomena that can describe the occurrence of local sleep. Finally, 13 young subjects were compared to a dataset of 13 older participants who performed planar upper limb reaching movements to determine whether beta oscillations were affected by age. Specifically, beta activity was assessed during reaching movements in different brain regions, in terms of topography, magnitude, and peak frequency. Results demonstrated that sustained learning produced electrophysiological changes both at rest and during the task. In fact, resting state was characterized by a progressive slowing of the EEG activity over areas overlapping with those engaged during the task. Precisely, we detected task-related activity mainly in the high-frequency ranges (gamma and beta right temporo-parietal activity for the visual sequence learning task; alpha and beta activity over a fontal and left parietal areas for the visuo-motor rotation); the same areas were characterized by a progressive increase of the low frequency EEG activity at rest ranging from alpha, beta after one hour of practice, to theta after three one-hour blocks. The control task did not trigger such EEG slowing, as reaching movements without learning did only left an alpha, beta trace in the resting state over a cluster reflecting the motor area contralateral to the movement. Further, continuous learning triggered performance deteriorations only in tests sharing the same neural substrate of the previously performed task. In other words, the visuo-motor learning task only affected performance in a motor test consisting in random reaching movements; conversely, visual sequence learning altered performance on a visual working memory test, but did not influence reaching movements. Also, the control condition did not affect performance in any of the two exercises. Performance decline, learning ability and local sleep were partially renormalized by a 90-minute nap but not by an equivalent period of wake. As such, the global EEG activity, computed as the mean power of all the electrodes, was not affected by either 90 minutes of sleep or quiet wake. However, the regions characterized by low frequency at rest benefited from the sleep period, as the low frequencies content partially decreased after the nap but not after quiet wake. Task related beta activity during motor practice presented similar magnitude and timing patterns in different brain areas, with a progressive increase with practice, in both young and older subjects, despite the older subjects performing slower, less accurate movements. Intriguingly, the motor areas showed a post movement beta synchronization having a peak between 15 and 18 Hz, as opposed to a frontal area that has it between 23 and 29 Hz. Finally, results did not reveal any direct relationship between EEG beta oscillations and performance indices. Altogether, these results indicate that local sleep and performance decrease can be triggered by prolonged learning in well rested subjects; furthermore, some amount of sleep can partially renormalize learning ability, EEG activity and performance. Also, differences in the brainnoscillations during motor activity can express separate processes underlying motor planning, execution and skills acquisition. The present study adds some important knowledge in the field of local sleep; in fact, it suggests that such phenomenon is triggered by sustained learning rather than sleep deprivation, thus being a plasticity-related phenomenon. Finally, the role of sleep on counterbalancing local sleep has been proved, despite additional studies are required to establish whether a full night of sleep rather than a specific amount of time is needed to fully restore learning ability and electrophysiological activity. In conclusion, the present findings are of importance in all the fields where sustained learning is required, such as rehabilitative programs, sport and military trainings, and must be taken into account when plasticity plays a fundamental role in the acquisition of new skills.