An electroencephalographic fingerprint of NREM sleep

The existence of homeostatic and circadian processes regulating human sleep seems to challenge the common notion of individual variations in sleep need or differences in circadian types. The elucidation of some biological mechanisms for the behavioral trait of morningness-eveningness and for the variability of sleep duration in the general population encourages a neurophysiological approach to individual differences in sleep characteristics. Sleep spindles seem a natural candidate, since they are one of the hallmarks of non-REM sleep, one of the few transient EEG events which are unique to sleep, with inter-individually variable incidence in humans and genotype differences in inbred mice. The aim of this study is the assessment of individual differences in antero-posterior EEG power, calculated across the 8.0–15.5 Hz range for NREM sleep episodes. Data from sleep recordings (Fz-A1, Cz-A1, Pz-A1, Oz-A1) of 10 normal males served as database. The subjects participated in a SWS-deprivation study (six consecutive nights: one adaptation, two baselines, two nights without SWS, one recovery). Power values in the 8.00–15.50 Hz range were calculated in a 0.25-Hz resolution. Absolute power was transformed into z scores. The analysis of individual cases showed a high topographical invariance. Five subjects showed one spectral peak in the range of spindle frequencies (range = 12.25–13.25 Hz) on all brain sites, and the others had two peaks on different sites, with one peak higher on Fz (range = 10.00–11.25 Hz) and a second one higher on Cz and Pz (range = 12.50–13.75 Hz). The frequency · EEG power matrices were highly inter-correlated within each individual. Individual mean correlation coefficients ranged from r = 0.78 to r = 0.99 [grand mean r = 0.93 (±0.07)]. On the other hand, the mean of all the possible intercorrelations between different nights/ subjects (n = 1350) yielded a coefficient of r = 0.56 (±0.21). The percentages of variance explained by the within-subject and by the betweensubject similarity were 86.6 and 31.7%, respectively (Z = 10.3; P = 0. 69–24). This invariance in the individual sleep EEG topography appears more noteworthy in the light of the considerable modifications of sleep characteristics across the six nights. Indeed, the homeostatic mechanisms of sleep regulation were dramatically challenged in the two SWS-deprivation nights, in which 328.1 (±167.69) and 739.8 (±314.8) auditory stimuli were respectively delivered, causing a large, regionally-graded increase in slow EEG frequency power during the non-REM sleep of the postdeprivation night. Like a !fingerprint", it may be related to individual differences in genetically determined functional brain anatomy rather than to sleep-dependent mechanisms. Topographic distribution of the EEG along the antero-posterior cortical axis steadily distinguishes each individual during NREM sleep.