Introduction. Evidence suggests that sleep facilitates memory consolidation, and certain sleep states appear to preferentially facilitate the consolidation of certain types of memory. Performance on two memory tasks during a period of sleep restriction (SR) was examined to determine: (1) the effect of SR on encoding and (2) the effect of a 3-hour sleep period, with or without pharmacologically enhanced slow wave sleep (SWS), on sleep-dependent memory consolidation.

Methods. Following screening and baseline (Day 2) assessments, subjects underwent two consecutive nights without sleep, each followed by a 3-hour daytime (0800-1100) sleep opportunity (Days 3 and 4) with either 3.5 gram sodium oxybate (n = 30; mean age: 27.1) or placebo (n = 28; mean age 27.1). An initial test to measure encoding was conducted on Day 2 at 0830 and on Days 3 and 4 at 0700 (prior to drug administration and sleep) for a declarative word pair task (WPT) and a nondeclarative finger-tapping task (FTT). Retest occurred 5 hours later, and change from initial test to retest assessed memory consolidation.

Results. During daytime sleep, the sodium oxybate group had more SWS (p<0.001), while the placebo group had more rapid eye movement (REM) sleep (p<0.001). Encoding was decreased by sleep loss, compared to baseline (p<0.031 for WPT, p<0.003 for FTT). For the placebo group, improvement from initial test to retest (consolidation) on the FTT was greater during SR (Days 3 and 4), with intervening sleep, compared to baseline (Day 2), without intervening sleep (p<0.005 for both). Consolidation on the FTT during SR did not differ significantly from baseline for the sodium oxybate group. WPT consolidation was similar on all days for both placebo and sodium oxybate.

Conclusion. Sleep loss impaired encoding on both tasks. A 3-hour sleep period following sleep loss appears to be insufficient to improve memory consolidation for a declarative memory task, even with SWS enhancement. However, three hours of sleep following sleep loss appear to facilitate nondeclarative memory consolidation (in the absence of slow wave sleep enhancement), perhaps because of relatively more REM sleep.