Extended Data Fig. 8: Additional analyses of cortical and hippocampal sleep needs.

a, Time points (1 h each) used for comparison (as in Fig. 4). b, Lack of significant correlation between SPW-ripple and DS rebounds between early recovery sleep and matched circadian time in baseline (n = 28 sessions of 3 kinds from 16 animals). c, Lack of correlation between cortical SWA rebound and changes in most hippocampal parameters (ripple frequency being the exception) during recovery sleep (n = 28 sessions of 3 kinds from 16 animals). In this and the following panels, the asterisk on the x-axis indicates that the hippocampal parameter did not show a significant difference across the interval being considered (for example, early and late recovery sleep in c). d, As in c, but for the baseline day (n = 28 sessions of 3 kinds from 16 animals). e,f, Same as c and d but, analogously to Fig. 4, expanded to include additional cortical regions (64 observations across 13 cortical areas) and the possibility of an interaction with neocortical region—primary visual (V1; n = 6), secondary visual (V2; n = 14), primary motor (M1; n = 4), secondary motor (M2; n = 10), medial parietal association (mPPC; n = 8), lateral parietal association (lPPC; n = 4), prelimbic (PrL; n = 5), cingulate (Cg1; n = 4), infralimbic (IL; n = 3), ventral orbital (VO; n = 3), medial orbital (MO; n = 1), entorhinal (EC; n = 1), retrosplenial (RS; n = 1). For analysis, areas were grouped (see Methods for details)—M1 and M2 into motor cortex (n = 14), V1 and V2 into visual cortex (n = 20), mPPC and lPPC into parietal association cortex (n = 12), VO, MO, IL, PrL, and Cg1 into frontal cortex (n = 16), and EC and RS into parahippocampal cortex (n = 2). Solid gray and gold lines indicate significant linear relationships in the presence of an interaction between experiment and the covariate (not region), and black solid lines indicate the presence of a significant linear relationship in the absence of an interaction. The solid green line in e indicates a negative correlation between DS amplitude decline and cortical SWA decline only in visual cortex following locomotion. Interactions and main effects were assessed using asymptotic likelihood ratio tests, and post hoc tests were performed using general linear hypothesis tests. All tests were performed on linear mixed effects models, were two-sided and corrected for multiple comparisons. Linear relationships and 95% confidence intervals are model expectations estimated using a marginal effects method (see Supplementary Table 1 and Methods for details).