Ultradian Rhythms and Temporal Coherence in Sleep EEG in Depressed Children and Adolescents Roseanne Armitage, Graham J. Emslie, Robert F. Hoffmann, Warren A. Weinberg, Robert A. Kowatch, Jeanne Rintelmann, and A. John Rush Background: It has been suggested that a primary ultra- dian (80 –120 minute) rhythm disturbance in EEG under- lies sleep abnormalities in adults with depression. The present study evaluated ultradian rhythm disturbances in childhood and adolescent depression. Methods: Sleep macroarchitecture and temporal coher- ence in quantitative EEG rhythms were investigated in 50 medication-free outpatients with major depression (25 children and 25 adolescents) and 15 healthy normal controls (5 children and 10 adolescents). Results: Few of the macroarchitectural measures showed significant group effects. In fact, age and sex effects were stronger than disease-dependent components. Temporal coherence of EEG rhythms during sleep did differentiate those with MDD from controls. Both depressed children and adolescents had lower intrahemispheric coherence, whereas interhemispheric was only lower in depressed adolescents in comparison with controls. Gender differ- ences were evident in adolescents, but not children, with MDD with lowest interhemispheric coherence in adoles- cent girls. Conclusions: These findings are in keeping with in- creased risk for depression in females beginning at ado- lescence and extending throughout adulthood. It was suggested that low temporal coherence in depression reflects a disruption in the fundamental basic rest-activity cycle of arousal and organization in the brain that is strongly influenced by gender. Biol Psychiatry 2000;47: 338 –350 © 2000 Society of Biological Psychiatry Key Words: Sleep EEG, ultradian rhythms, depression, children, coherence Introduction M ajor depressive disorders (MDD) are associated with significant morbidity and mortality throughout the life cycle (Cantwell 1992; Fleming and Offord 1990). Although the intimate connection between sleep distur- bance and MDD is reasonably clear in adulthood, it has been more difficult to establish in childhood and adoles- cent depression. Studies in the sleep laboratory have shown that sleep macroarchitecture, based on visual stage scoring of the polysomnogram, is characterized by pro- longed sleep onset, short latency to the first rapid eye movement (REM) sleep period (65 minutes), decreased slow-wave sleep (Stages 3 + 4), and increased sleep fragmentation in 40 – 60% of adult outpatients with MDD (Armitage 1995; Gillin et al 1982; Kupfer 1976; Mendel- son et al 1987; Reynolds and Kupfer 1987; Reynolds et al 1990). In addition, persistent sleep disturbance is associ- ated with increased risk of relapse and recurrence (Giles et al 1989) and suicide in adults with MDD (Fawcett et al 1990; Ford and Kamerow 1989; Wingard and Berkman 1983). Sleep microarchitecture, based on computerized analy- ses of sleep EEG frequencies, has also revealed significant differences between adults with MDD and healthy control subjects. A number of studies have shown elevated fast- frequency EEG activity, most notably in the beta (16 Hz) band, and either a reduction in the amplitude or timing of delta (4 Hz) activity during sleep as characteristic of adult MDD (Armitage 1995; Armitage et al 1992a, 1992b, 1995; Borbe ´ly et al 1984; Beersma et al 1985; Kupfer et al 1984a, 1984b, 1986; Van den Hoofdakker and Beersma 1986). Some of this work has led to the speculation that both macro- and microarchitectural sleep abnormalities in adults with MDD can be explained by a single ultradian rhythm disturbance (Beersma et al 1984, 1985). Several recent studies provide strong support for this view, dem- onstrating significantly lower temporal coherence of ultra- dian, 90-minute EEG rhythms in those with MDD com- pared to healthy controls (Armitage 1995; Armitage and Hoffmann 1997; Armitage et al 1992b, 1993, 1997, in press-a, in press-b). This work has been driven by Kleit- man’s theory that a fundamental, ultradian, basic rest- activity cycle (with an approximate 90 minute periodicity in adults) underlies the organization of REM/NREM sleep cycles, physiology, mood and behavior throughout 24 From The University of Texas Southwestern Medical Center at Dallas, Dallas. Address reprint requests to Roseanne Armitage, Ph.D., Director, Sleep Study Unit, Department of Psychiatry, The University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, Texas, 75235-9070. Received January 19, 1999; revised May 4, 1999; accepted May 6, 1999. © 2000 Society of Biological Psychiatry 0006-3223/00/$20.00 PII S0006-3223(99)00129-8