Abstract
Slow Wave Sleep (SWS) has been associated with consolidation of declarative memories and visual perceptual learning. This has been inferred from improved performance across an early versus late night sleep interval as the early night has a relatively high proportion of SWS. The direct relationship with SWS is unclear though as homeostatic sleep pressure, circadian phase, sleep inertia, and other physiology also differ between early and late night, and are known to affect cognition. We investigated whether these confounds could be overcome with administration of the circadian hormone melatonin, as it is know to reduce SWS and cause other physiology similar to late night sleep. Further, melatonin has a short one-hour half-life and no "hangover" effects. Thus, melatonin administration affords modulation of sleep EEG architecture at will, while also ensuring participants learn and test under the same circadian, homeostatic, cognitive, and physiological states across electrophysiological conditions. In a visual perceptual learning paradigm, we manipulated the amount of SWS during a 90-minute afternoon nap by administering 3 mg of melatonin or placebo 45 minutes before each nap. We compared the within-subject effects of melatonin versus placebo on consolidation of a classic perceptual learning task (line orientation discrimination) and on sleep EEG. As hypothesized, melatonin reduced the delta (0-4 Hz) EEG power during the nap and reduced performance improvement on the perceptual learning task. In contrast, melatonin increased performance improvement on a control task (letter discrimination) indicating that the detrimental effect of melatonin on the perceptual task is not due to a reduction in vigilance, but rather to perceptual discrimination ability per se. Our findings confirm that an afternoon nap is sufficient to consolidate perceptual learning and that SWS is directly involved in consolidation.
Meeting abstract presented at VSS 2013