STATE DEPENDENT ASPECTS OF COGNITION

点击翻译按钮获取中文摘要

Our program focuses on the functional interactions between sleep and waking. In the work proposed here, we ask two general questions: 1.) How does sleep affect learning and associative memory in subsequent waking; and 2.) How does cognition during waking affect subsequent sleep? In both parts of our program we rely heavily on a new portable sleep monitoring system, the Nightcap. Using this device, we are able to monitor sleep-wake states in the home and either collect mentation reports or carry out cognitive tests after computer-initiated awakenings from REM and NREM sleep. In studies reported here, we have found that performance on a semantic priming task is altered following awakenings from REM sleep, with the relative efficacy of strong and weak primes reversed; under these conditions "weak primes" actually produce more priming than "strong primes" (t=2.51, p=0.01). We now propose new studies to distinguish the effects of strength of prime from the effects of prime-target relationships. We also report new results showing that procedural learning, measured by the overnight improvement on a visual discrimination task (VDT), is highly correlated with (r=0.79, p=0.002) and directly proportional to the amount of REM sleep in the last quarter of the night, 6 - 8 hr post-training, but totally uncorrelated (r=0.02, p=0.96) with the amount of REM sleep in earlier portions of the night, up to 6 hr post-training. The timing of this "critical REM window" corresponds to the time following a 6 hr latent period proposed by Karni and Sagi (1993) when learning of the VDT occurs in waking. We will test the hypothesis that long-term improvement on the VDT in sleep or waking is dependent On the cholinergically driven reinforcement of neocortical memory traces during a critical period 6 - 8 hr post-training. Preliminary studies have further shown that training on the VDT has a reciprocal effect on sleep, leading to a 25 percent increase in the amount of REM sleep during the night following VDT training. We propose to extend these studies to test the hypothesis that the increased REM sleep occurs primarily during the critical REM window and is a response to the presence of recent, unstable, cholinergically sensitive memories. Finally, we propose to determine which aspects of the visual discrimination task actually lead to the REM increase.