Abstract
Brain oscillations of different frequencies characterize the electroencephalogram (EEG) during distinct cognitive and vigilant states. Theta oscillations (4-8 Hz) are unusual because they have been found in the near-opposite conditions of sleepiness and alert cognitive control. Most neuroscience research fo-cuses exclusively on the latter, leaving this paradox unresolved. With this thesis, I focus instead on the-ta during sleep deprivation (sdTheta), which has been hypothesized to reflect intrusions of local slow wave sleep on wake, based on a study in rats. The goal was to determine whether sdTheta in humans could also be considered a form of local sleep in wake, or if it was a manifestation of more typical cog-nition-related theta. I collected high-density EEG data from young healthy adults undergoing sleep deprivation to observe how sdTheta is affected by time awake, time of day, different tasks, and condi-tions. To independently track the effects of sleep deprivation, I also conducted extensive questionnaires and collected pupillometry data. I found that sdTheta can be widespread across the brain, although the specific sources depend on the ongoing task. Curiously, theta mostly originated from areas not critical for the task. I found that sdTheta occurs in bursts, making it unlike the isolated theta events thought to reflect local sleep. Furthermore, I found that independently from changes in the occurrences of such bursts, wake oscillation amplitudes increase with time awake, following a homeostatic trajectory. This supports the hypothesis that neuronal connectivity increases with time awake, which is what underlies sleep need. Unexpectedly, I found that the wake maintenance zone, a time before habitual bedtime when it is difficult to fall asleep, can mask these homeostatic changes in oscillation amplitudes. How-ever, the wake maintenance zone only minimally affects the presence of sdTheta bursts. Finally, I could not find any evidence that theta bursts were the cause of behavioral lapses nor compensating for sleep loss, supporting the previous finding of sdTheta originating from task-unrelated areas. Therefore, I ten-tatively propose that sdTheta bursts are a manifestation of unneeded parts of the brain at rest, alt-hough not necessarily “local sleep.” If this means that sdTheta is a different type of oscillation from theta involved in cognition, then care will be needed to dissociate the two types. Regardless of what it does, theta makes for a robust marker of sleep need and can have many clinical diagnostic applications, especially when analyzed effectively.
Snipes, Sophia