Maturation-dependent changes in cortical and thalamic activity during sleep slow waves: Insights from a combined EEG-fMRI study

Studies using scalp EEG have shown that slow waves (0.5–4 Hz), the most prominent hallmark of NREM sleep, undergo relevant changes from childhood to adulthood, mirroring brain structural modifications and the acquisition of cognitive skills. Here we used simultaneous EEG-fMRI to investigate the cortical and subcortical correlates of slow waves in school-age children and determine their relative developmental changes. We analyzed data from 14 school-age children with self-limited focal epilepsy of childhood who fell asleep during EEG-fMRI recordings. Brain regions associated with slow-wave occurrence were identified using a voxel-wise regression that also modelled interictal epileptic discharges and sleep spindles. At the group level, a mixed-effects linear model was used. The results were qualitatively compared with those obtained from 2 adolescents with epilepsy and 17 healthy adults. Slow waves were associated with hemodynamic-signal decreases in bilateral somatomotor areas. Such changes extended more posteriorly relative to those in adults. Moreover, the involvement of areas belonging to the default mode network changes as a function of age. No significant hemodynamic responses were observed in subcortical structures. However, we identified a significant correlation between age and thalamic hemodynamic changes. Present findings indicate that the somatomotor cortex may have a key role in slow-wave expression throughout the lifespan. At the same time, they are consistent with a posterior-to-anterior shift in slow-wave distribution mirroring brain maturational changes. Finally, our results suggest that slow-wave changes may not reflect only neocortical modifications but also the maturation of subcortical structures, including the thalamus. Low spatial resolution of conventional EEG has widely limited the possibility of accurately determining in humans the role of different brain structures in the expression of sleep slow waves throughout development. Here, for the first time, we took advantage of a simultaneous EEG-fMRI approach to accurately describe the cortical and subcortical hemodynamic correlates of sleep slow waves in a sample of 14 school-aged children. In order to elucidate age-dependent changes, we qualitatively compared present findings with those previously obtained in a group of adults. Overall our results have important implications for the understanding of how cortico-cortical and subcortico-cortical interactions shape sleep slow waves acros