AN INDIVIDUAL FINGERPRINT OF HUMAN NREM SLEEP: THE ANTERO-POSTRIOR TOPOGRAPHY OF THE MIDDLE EEG FREQUENCIES.

Introduction: Recent topographic analyses of sigma EEG activity during NREM sleep pointed to the existence of two distinct sleep spindle (SS) types: “slow” spindles at ~12 Hz more pronounced over the frontal sites, and “fast” SS at ~14 Hz more pronounced over centro-parietal sites; the former declining and the latter increasing over consecutive sleep episodes. Functional dissociations have also been reported, as a function of age and maturation, homeostatic and circadian factors, menstrual cycle phase, pregnancy and pharmacological agents. These results have been interpreted according to two functionally separated spindle generators.On the other hand, lower and higher frequencies of SS have been attributed to a single mechanism, namely, the duration of hyperpolarization- rebound sequence in thalamocortical neurons; and cortical areas on which “slow” SS have been detected are related to those thalamic nuclei in which relay cells display long hyperpolarizations.In fact, most studies showing the two cortically-independent SS bulk the bounds of sigma band from the traditional 12-14 Hz range, including frequency bins traditionally considered within the alpha band. Secondly, there are great inter-individual differences, and topographical analyses on between-subject averaged data could confound individual (different) cortical topographical distributions with the existence of different topographically- specific SS. Finally, spindle frequency could change as a function of the same variables differently affecting the two types of SS. Topographical EEG analysis in single subjects could allow to disentangle the question. Methods: 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). Power values were calculated across a 8.00-15.50 Hz range in a 0.25-Hz resolution for NREM episodes of each subject. Absolute power was transformed into z scores. Results: Four subjects had one spectral peak in the range of SS frequencies on all derivations (12.25-13.25 Hz); three others had two spectral peaks on different derivations, with one peak higher on Fz (10.25- 11.25 Hz) and another peak higher on Cz-Pz (13.50-13.75 Hz); the last three subjects had two spectral peaks discernible only on Fz (respectively at 10.00-10.50 and 12.50-13.00 Hz). Conclusions: Individual data show one kind of SS, ranging within the 12.25-13.75 Hz, and in some subjects alpha activity peaking on more anterior brain sites. However, the most impressive result is the high spectral invariancy of individual antero-posterior topography in the 8.00-15.50 Hz frequency range, notwithstanding wide changes in sleep architecture across the six nights (one adaptation, two baselines, two nights without SWS, one recovery). This invariancy of EEG topography may be related to brain anatomy, and studies comparing monozygotic and dizygotic twins will allow to demonstrate this hypothesis.