Slow Waves and Learning: Beyond Correlations

摘要: ELECTROENCEPHALOGRAPHIC (EEG) SLOW WAVES ARE THE EPITOME OF DEEP SLEEP. PREVALENCE AND AMPLITUDE IN SLEEP EEG are typically quantified as slow-wave activity (SWA; power density in the 0.75 - 4.5-Hz range). SWA appears to be a marker of sleep need and an expression homeostatic sleep-regulatory process.1 But it is unclear whether simply epiphenomenon or directly serves important functions. In this issue SLEEP, Landsness et al2 tested hypothesis that responsible for consolidation visuomotor learning. Many recent studies have linked learning synaptic plasticity. not only aspect has been related such processes, but specifically implicated declarative memory,3 well overnight gains perceptual4 performance.5 use dependent cortical circuits were particularly active during wakefulness produce more subsequent sleep,5,6 whereas underused less SWA.7 Increases induced either by practicing task prior wakefulness5 transcranial direct-current stimulation sleep3 found correlate positively with magnitude sleep-dependent effects. Whereas much earlier evidence was correlative, study using acoustic SWA-suppression paradigm provided support causal relationship between perceptual learning.8 The that— if applied carefully—it allows sizable reduction SWA, while total time REM remain unaffected. Using type paradigm, examined role rotation adaptation, well-characterized form learning. authors show average 31% prevented improvement performance changes over right parietal cortex, measured high-density EEG, correlated performance. Despite its strength, also limitations: difficult know behavioral effects stimuli attributable solely decrease additionally increase arousals. Well aware these limitations, systematically pursued several new strategies overcome them. First, they included control condition (CAS), which subjects exposed when slow waves absent. contrast wave deprivation (SWD), exhibited significant CAS condition, thus ruling out possibility procedure per se Next, compared number clinically defined Although arousal index slightly higher SWD than conditions, multiple regression analysis independent variables revealed predicted next-day Finally, introduced automated based on event-related spectral perturbation response auditory stimuli. basic idea identify responses may meet criteria arousals represent subtle kind fragmentation differ conditions. At end exhaustive analysis, demonstrated, however, indeed lightening performance.2 validated strengthened interpretational power. In summary, colleagues2 provide strong clarity results, one should keep mind some other failed find suppression learning9,10 waking functions.11,12 It possible methodologic differences played role, and, therefore, will future examine systematically. reasonable assume functions degree depend SWA. investigation reveal clues about functional fact largest so far learning2,8 consistent