Circadian Misalignment in Mood Disturbances Alfred J. Lewy, MD, PhD Corresponding author Alfred J. Lewy, MD, PhD Oregon Health & Science University, 3181 Southwest Sam Jackson Road, Portland, OR 97239, USA. E-mail: lewy@ohsu.edu Current Psychiatry Reports 2009, 11:459–465 Current Medicine Group LLC ISSN 1523-3812 Copyright © 2009 by Current Medicine Group LLC Recent refnements in methodology allow chro- nobiological researchers to answer the following questions: is there circadian misalignment in sleep and mood disturbances, and, if so, is it of the phase- advance or phase-delay type? Measurement of the dim light melatonin onset-to-midsleep interval, or phase-angle difference, in sleep and mood disor- ders should answer these questions. Although the phase-advance hypothesis of affective disorders was formulated three decades ago, recent studies suggest that many, if not all, mood disturbances have a cir- cadian misalignment component of the phase-delay type, operationally defned as a delay in the dim light melatonin onset relative to the sleep/wake cycle. Phase-delayed disorders can be treated with bright light in the morning and/or low-dose melatonin in the afternoon/evening. Phase-advanced disorders can be treated with bright light in the evening and/or low-dose melatonin in the morning. Introduction Although the phase-advance hypothesis of affective disorders was proposed three decades ago [1–3], sup- portive evidence has not been forthcoming in recent years. In fact, it has become increasingly apparent that the circadian phase disturbance in depressive disorders may actually be a phase delay, not a phase advance [4]. Now that investigators are interested in both types of phase disturbances, the feld is poised for an upsurge in research activity, particularly with recent methodologic advances. Progress has been slow, in part because it has taken time to sort out key issues, several of which have recently been resolved or are in the process of being resolved. These issues can be framed as questions and are addressed below. What Is the Best Marker for Circadian Phase Position? As expostulated in the phase-advance hypothesis [1–3], there seem to be two sets of circadian rhythms that can become misaligned. One set of (mostly metabolic) rhythms is tightly coupled to the endogenous circadian pacemaker. The second set comprises circadian rhythms that are less tightly coupled to the endogenous circadian pacemaker and are related to—and are evoked responses of—the sleep/wake cycle—that is, the actual times of sleep and wake, which are infuenced by stress and schedule demands. (Note: the circadian rhythm of sleep propensity belongs to the frst set of rhythms.) The preeminent marker for the frst set of rhythms— which includes cortisol, temperature, and melatonin—is recognized as the circadian rhythm in melatonin production [5•,6], most conveniently assessed using the dim light mela- tonin onset (DLMO). The DLMO, which is obtained by measuring melatonin levels collected every 30 to 60 minutes from about 6 pm until bedtime, is the most practical marker for the endogenous circadian pacemaker and probably the most precise [7,8], which accounts for its widespread use. Compared with the rectal temperature measured under a constant routine in which individuals must remain awake while supine and restricted to hourly isocaloric meals [9], the DLMO protocol is less tedious, with the most arduous part being the need for sampling conditions of very dim light in the evening because light suppresses melatonin production [10]. Now that it is known that this response is mediated by blue light [11], orange lenses or goggles may obviate the need for very dim light [12]. Thus, the DLMO becomes even more convenient. The salivary DLMO can be done at home or in the clinical or sleep laboratory. General use of the DLMO required the resolution of several issues. As late as the mid-1980s, there was disagreement over the number of endogenous circadian pacemakers in the brain. Although some investigators thought there were two pacemakers [13], research now indicates that there is one master pacemaker located in the suprachiasmatic nucleus of the hypothalamus. The possible dual nature of the pacemaker was then construed by some investigators as a complex or two- oscillator pacemaker [14,15]; in the case of melatonin production, a second oscillator was hypothesized to con- trol the offset of melatonin production (in addition to the