활동기록기 기반 수면 지표를 이용한 우울증 예측

Prediction of Depression Using Actigraphy-Based Sleep Parameters

Objective: This study aimed to examine which actigraphy-based sleep parameters could effectively discriminate individuals with depression from those without depression, using receiver operating characteristic (ROC) analysis. Methods: We recruited outpatients aged 50 years or older from the Department of Psychiatry at Chosun University Hospital. Participants completed the Korean version of the Patient Health Questionnaire-9 (PHQ-9) and the Montgomery–Åsberg Depression Rating Scale. Those with PHQ-9 scores ≥10 were classified into the depression group (DG). Actigraphy (Actiwatch 2; Philips Respironics) was used at home for two consecutive days to assess sleep parameters, including bedtime (BT), time in bed (TIB), total sleep time, sleep onset latency, sleep efficiency, wake after sleep onset, and fragmentation index. Group comparisons were performed using the Mann–Whitney U test, and discriminative validity was assessed with ROC analysis. Results: Nineteen participants were included (DG: 13, non-depression group [NDG]: 6). Compared with the NDG, the DG showed significantly earlier BT and longer TIB (p<0.05). Among sleep parameters, TIB was identified as a significant discriminator of depression (area under the curve=0.80, cutoff=9 h 58 min, sensitivity=0.76, specificity=0.75). Conclusion: Individuals with depressive symptoms showed earlier bedtimes and significantly longer TIB compared to nondepressed participants. Prolonged TIB, exceeding two hours, may reflect reduced physical activity and serve as an objective marker for depression. Importantly, TIB was identified as a predictor of depression based on actigraphy-derived measures, suggesting potential clinical utility for early screening and intervention.