RECENT PRESENTATIONS

McEvoy, L.K., Smith, M.E. & Gevins, A. (2003). Assessing the cognitive consequences of drowsiness with EEG and working memory task performance measures. Associated Professional Sleep Societies 17th Annual Meeting. June, Chicago, IL.

ABSTRACT

Introduction: Problematic drowsiness is typically assessed with subjective ratings or objective tests of sleep tendency. An objective test of functional alertness - the ability not just to remain awake but to perform cognitively at one's normal capacity - could be a valuable tool in the assessment of pathological drowsiness and its treatments. EEGs recorded during performance of cognitive tasks may provide useful information in this context because they are sensitive to changes in attentional processes underlying task performance (Gevins et al Cerebral Cortex, 1997; Neuropsychopharm, 2002; Smith et al, Sleep, 2002). Here we investigated neurophysiological signals of drowsiness in a task that requires sustained focused attention and working memory, abilities that are fundamental to higher cognitive functioning. We also investigated the added value of combining EEG with behavioral data to better discriminate alert from drowsy states.
Method: Sixteen healthy young adults participated in several experimental sessions, including one in which they remained awake overnight performing tasks at periodic intervals from 11:00 PM until 6:00 AM. In each interval, subjects completed subjective sleepiness questionnaires and EEGs were recorded while they performed easy and difficult versions of a spatial working memory task, and while they performed other tasks or rested quietly. Data from the overnight session was compared to data obtained from alert daytime sessions.
Results. Subjective drowsiness increased linearly throughout the overnight session, as did electrophysiological measures of drowsiness obtained from the resting conditions. Neurophysiological indicators of sustained focused attentional effort, such as the difference in frontal midline theta band power and in parietal alpha band power between the easy and difficult levels of the working memory task, did not significantly change across the night. In contrast, working memory task performance and components of the event-related potentials sensitive to transient allocation of attention to task stimuli significantly decreased from 11:00 PM until 1:30 AM, and remained at that depressed level throughout the rest of the night. Stepwise linear discriminant analyses performed to differentiate data in each overnight interval from data obtained in alert, daytime sessions showed that functions combining EEG and behavioral measures produced greater test-set classification accuracy (94%; p<.0001) than did functions comprising behavioral variables alone (81%; p<.01; maximum of 4 variables in both analyses).
Conclusions: Drowsiness induced by a period of prolonged wakefulness significantly disrupts neurophysiological circuits underlying transient focusing of attention, resulting in impaired task performance even when subjects maintain a high level of attentional effort. The ability to discriminate states of drowsiness from states of alertness is improved by the addition of EEG variables. This approach, in which task-related EEG is analyzed in conjunction with behavioral data obtained during the performance of demanding cognitive tasks, may aid in research on cognitive consequences of sleep disorders and may provide a useful basis for an objective clinical test of functional alertness. Supported by grants from NINDS, NIMH, NIAAA, NASA and AFRL.