MS 202.qxp

IN HISTORICAL REFERENCES TO WHAT TODAY WOULD BE CONSIDERED MAJOR DEPRESSION, MOOD DISORDERS HAVE LONG BEEN ASSOCIATED WITH CHANGES IN SLEEP PATTERNS. Using the term melancholia, Hippocrates (460-377 BC) already described many features, including “loss of sleep,” by which we define major depressive disorder (MDD) today. Subjective sleep complaints are frequently present in depressed patients,1,2 and insomnia or hypersomnia are among 9 diagnostic criteria for MDD.3 Patients with depression typically complain of prolonged sleep latency, frequent arousals with early-morning awakenings, short and nonrestorative sleep, and disturbing dreams.1 These subjective symptoms may be accompanied with objective sleep alterations such as disturbed sleep continuity, abnormalities of rapid-eye-movement (REM) sleep, loss of deep non-REM (NREM) sleep (ie, slow-wave sleep), and changes in variables derived from quantitative sleep electroencephalogram (EEG) analysis (eg4-8). Several models have been proposed to account for the relationship between depressed mood and disturbed sleep in patients with MDD.1 For example, it has been suggested that REM sleep is excessively increased,9 the balance between cholinergic and monoaminergic neurotransmission is abnormal,10,11 and the phase-relationship between sleep and the circadian system is altered12 in depressed patients. Moreover, Borbély and WirzJustice13 hypothesized that depression is associated with a deficient build up of a sleep-regulatory process that accumulates during waking and dissipates during sleep. This process reflects NREM sleep intensity, depending on the history of prior sleep and wakefulness, and has been referred to as the homeostatic Process S in the 2-process model of sleep regulation.14 The dynamics of Process S can be reliably tracked by the time course of slow-wave (or delta) activity (SWA)(~ 0.5-5 Hz) in the NREM sleep EEG. In healthy adults, delta activity declines ROUGHLY exponentially during a sleep episode, is enhanced after sleep deprivation, and is reduced after a daytime nap (see15 for recent overview). In early studies of untreated depressed patients and healthy controls, the level of delta activity in the initial part of sleep was reduced, and its decline across consecutive NREM sleep episodes was altered in accordance with the S-deficiency hypothesis.5,16 These changes, however, were not confirmed in all further studies. They are present in some patients, but not in others (eg8,17,18; see19 for discussion). Thus, the question whether Process S is deficient or not in individuals with depression remained unanswered. A novel approach to detect subtle changes in sleep regulation is the investigation of fronto-occipital EEG power gradients dur-