0973 Sleep Disturbance and Emotion Regulation Dysfunction in Depression: Self-Report and Neural Evidence

Abstract

Introduction: Almost 60-90% of major depressive disorder (MDD) patients suffer some sort of sleep disturbances such as difficulty of initiating sleep, nonrestorative sleep or increase in wake after sleep onset. Polysomnographic (PSG) studies also suggested that several sleep-architecture abnormalities in EEG profiles of MDD patients. A recent meta-analysis revealed that REM density, REM latency and reduced slow-wave sleep are potential candidates for consistent MDD biomarker. However, there remain some cautions that publication , study-selection, and population biases could dissociate from real-world clinical situation for MDD diagnosis. Thus, we conducted a multicenter exploratory prospective study for developing practical use of a single channel sleep EEG device for MDD diagnosis. Methods: We adopt a laptop single channel EEG device enabling facile sleep EEG measure at individual patients' home. Two-hundred patients with MDD and the other psychiatric disorders including insomnia was recruited. EEG is measured two times form an electrode attached on participants' forehead during two nights of sleep before and after the treatment for their psychiatric disorders. In addition to the sleep EEG parameters described above, we focus on the alpha and sigma spindle, and delta activities during NREM sleep periods, and the beta activity during REM sleep periods. Patients are diagnosed by a psychiatrist using DSM-5 criteria. Results: A receiver operating characteristic curves was carried out to evaluate the contribution of sleep EEG parameters to predictive diag-nostics of MDD. The area under the curve (AUC) was used to evaluate the predictive diagnostic ability. A combination with REM latency, wake after sleep onset, and NREM-to-REM ratios in average power spectral density of alpha and sigma and delta waves showed a sensitivity of about 80%, a specificity of about 70%, an AUC of about 0.8, with adjustment for age.