The present thesis focuses on the association between sleep and cognitive functions in the developing human brain during the childhood and adolescent years. The importance of sleep for learning and memory processes has been established firmly in a large number of studies in adults. These studies have shown that sleep contributes to efficient consolidation of both declarative memory— the memory for facts and events—and procedural memory—the memory for skills and procedures, and that sleep also supports cognitive functions other than memory, as well as emotional functioning. Furthermore, experimentally restricting sleep in healthy adult participants resulted in a host of negative consequences for cognitive, emotional and motor performance.
  In contrast to the large number of studies on sleep and cognition in adults, a much smaller number of studies investigated the role of sleep and its associated oscillations in cognition across childhood. Yet there are valid applied and fundamental scientific reasons to investigate this relationship earlier in development. The work presented in this thesis therefore addresses the following questions:

These questions we addressed by means of a meta-analysis on existing studies (Chapter 2), an observational study (Chapter 3), and an experimental study (Chapters 4 and 5). We briefly summarise the answers to these questions below.

Chapter 2 describes a meta-analysis on all past scientific studies relating children’s sleep (duration or efficiency) to cognition and behavioural problems. A total of 86 studies on 35,936 children (5–12 years old) was found suitable for inclusion and provided a number of insights, of which the most important are summarised here. Children who sleep longer show slightly, but significantly, better cognitive performance. This association is most prominent for cognitive tasks that required executive functioning, for tasks that addressed multiple cognitive domains, and for school performance. Children who sleep longer, moreover, show slightly less internalising and externalising behavioural problems. Quite unlike typical findings in adults, sleep duration is not associated with sustained attention or memory performance, whilst these are the two domains most severely affected by sleep restriction in adults. Both methodological issues and brain developmental immaturities were proposed to underlie the marked differences between the findings in children and adults. A practical implication of these findings is that it may be valuable to research whether interventions aimed at increasing sleep duration in school-age children can improve complex cognitive functions, executive functions, and school performance, whilst ameliorating internalising and externalising behavioural problems.

Chapter 3 examines the interaction between sleep restriction and stress in adolescents. When sleep duration is restricted in adults, a homeostatic response ensures that sleep will become more consolidated, more efficient. Chapter 3 aims to determine whether sleep restriction will lead to a similar compensatory sleep efficiency response in adolescents. Furthermore, it aims to investigate whether this compensatory mechanism can persist when faced with chronic stress? These questions were addressed in a naturalistic ecologically valid quasi-experimental repeated-measures study, in which we evaluated sleep during a week’s holidays (low-stress extended-sleep), during a regular week of school (low-stress restricted-sleep), and during stressful examination weeks (high-stress restricted-sleep). The findings suggest that when adolescents’ sleep is challenged—as a consequence of school attendance—by a reduction in its duration, it responds by an increase in its efficiency. However, when adolescents experience chronic stress—due to examinations—in addition to sleep restriction, they fail to maintain this seemingly compensatory increase in efficiency. A practical implication of these findings is that it might prove valuable to investigate whether a more dispersed schedule of examinations would interfere less with sleep and its supporting role for cognitive performance.

Chapter 4 evaluates the possibility to investigate the supportive role of sleep for memory consolidation, without having to rely on invasive and costly methodologies. Animal studies, as well as a handful of intracranial and magnetoencephalograpy studies in humans, have shown that during the slow waves of sleep bouts of high-frequency (gamma band) electrical activity occurs. These bouts resemble the cortical activity underlying cognition in wakefulness. They have therefore been proposed to represent very brief periods of wakefulness to support cognition. Although it would be most interesting to study the role of phasic gamma-band activity in relation to daytime cognition, gamma oscillations are of such small amplitude that they are difficult to measure reliably in the electroencephalography (EEG) of adults. We noted that sleep-EEG studies in 11-year-olds might provide an interesting opportunity to study this phenomenon, as during this developmental stage oscillations are most pronounced and thus lead to a better signal-to-noise ratio than is the case in adults. Indeed, using time-frequency analyses on the sleep EEG obtained in 30 children, we found a remarkable modulation of gamma power along the time course of a slow wave. Furthermore, for the first time in children, we found a direct link between sleep’s slow waves and spindles. A practical implication of these findings is that children provide a unique opportunity to conduct non-invasive and affordable investigations into the role of gamma—during sleep—for daytime cognition.

Chapter 5 focuses on the association of motor skill performance and sleep in children. Similar to that previously found in adults, the accuracy of children’s motor skill performance increased only if the consolidation period includes a period of sleep. However, children increased the speed of their performance no matter whether the interval included a period of sleep or wakefulness. Moreover, we showed that the dominant frequencies of the two most characteristic sleep- EEG events (i.e., spindles and slow waves) were strongly predictive of individual motor skill performance levels. Children with a lower density of slow spindles, a higher density of fast spindles, and a faster slow wave frequency perform better. Those children with a higher density of slow sleep spindles and a slower average frequency of slow waves show lower initial and lower overall performance, yet the greatest overnight improvement in accuracy. Slower spindle and slow wave frequencies may thus reflect immaturity of the neuronal networks involved in motor skill learning. A first practical implication of the findings is that studies on the role of spindles in overnight memory consolidation should be aware of the confounding effects of initial differences in baseline performance onto the investigated parameters. An intriguing second implication of these findings is that it would be of great value to study why children are able to increase their motor skill speed without training, and why this capacity disappears in adulthood.
In summary, the studies described in this thesis have added a valuable contribution to our knowledge of the role of sleep in cognition and behavioural problems in children. This thesis shows it is important to consider sleep in our understanding of individual differences in cognition and behavioural expressions in children. Chapters 2 and 3 suggest that it is timely to evaluate whether interventions aimed at improving sleep in children may improve cognitive performance—including school performance—and ameliorate behavioural problems. Chapters 4 and 5 indicate that it may be of particular relevance to study the role of sleep in cognitive performance across different developmental stages, and not just in young adults. It appears timely to consider large-scale multivariate follow-up studies to disentangle individual traits from developmental aspects in the supportive role of sleep for cognition and behaviour.