FIGURE 4. Enhancement of performance after sleep and its relationship to SWA.

a, Post-sleep performance improvement. Two groups of subjects (sleep group and wakefulness group) underwent rotation adaptation as described in Fig. 1. The extent of rotation adaptation was tested 10 min after the end of training using an imposed rotation of 60°. Subjects in the sleep group (n = 11, mean age 25.8 1.8 yr) trained in the evening and were re-tested after 7–8 h of sleep (one subject did not complete the re-test and was dropped from the analysis). Subjects in the wakefulness group (n = 10, mean age 28.2 2.9 yr) trained in the morning and were re-tested after being engaged in their normal waking activities for 8 h. A repeated measure analysis of variance and post-hoc tests showed that the two groups had similar performance when tested immediately after training; however, at re-test approximately 8 h later, mean directional error was significantly reduced in the sleep group (P < 0.006) but not in the wakefulness group. The two groups differed significantly in the extent of test–re-test performance change (P < 0.002). b, Positive correlation between peak SWA increase during sleep within the six electrode cluster and decrease of mean directional error after sleep (n = 10 subjects). There was no correlation between the decrease of the mean directional error after sleep and the amount of total sleep or individual sleep stages during the first sleep cycle. Dotted lines indicate the 95% confidence interval. c, Positive correlation between the change in standard deviation of directional error during rotation adaptation (block 1 minus block 10; see Fig. 1b) and the peak SWA increase in the subsequent non-REM sleep episode (n = 10 subjects).