0985 Comparison of Various Sleep Measurement Devices in Psychiatric Patients

Abstract Introduction Sleep disturbances are key symptoms in psychiatric disorders. Despite the development of various methods to assess sleep patterns and their quality, the concordance between these methods in those with psychiatric disorders is unclear. In particular, it is not well understood how comorbid symptoms of depression and obstructive sleep apnea (OSA), which are common in psychiatric disorders, relate to the concordance of sleep measurements. Methods In this study, 62 individuals with psychiatric disorders from Nagoya University Hospital were included as participants. Their sleep was monitored using several methods polysomnography (PSG), a portable EEG device called Zmachine, Fitbit, and sleep logs. The morning after the sleep measurement with these objective sleep devices for a single night, participants estimated their sleep duration subjectively. To evaluate sleep parameters, Intraclass Correlation Coefficients (ICCs) were used. This research was carried out following the approval of the Ethics Review Committees at Nagoya University Hospital, and written consent was acquired from all participants. Results PSG and Zmachine demonstrated moderate concordance in Total Sleep Time (TST: ICC=0.46), Wake After Sleep Onset (WASO: ICC=0.39), and Sleep Efficiency (SE: ICC=0.40). However, there was a lack of agreement between PSG and Zmachine parameters with Fitbit, and between Fitbit and sleep logs. Sleep logs showed minimal correlation with PSG and Zmachine for TST and WASO, and a slight agreement for SE with Zmachine. When participants with severe OSA were excluded, ICCs for TST, WASO, and SE between PSG and sleep logs improved marginally (TST: ICC=0.27; WASO: ICC=0.38; SE: ICC=0.22). Excluding participants with severe depression, the TST ICC for PSG and sleep logs were slightly elevated (TST: ICC=0.24). However, even when participants with severe depression and OSA were excluded, ICCs between Fitbit and other sleep measurement methods did not elevate. Conclusion Our results indicated that Fitbit has a discrepancy when comparing data from PSG and sleep logs in individuals with psychiatric disorders. Although the growing adoption of Fitbit for sleep measurement, it is essential to acknowledge its weakness in the assessment of individuals with psychiatric disorders. Support (if any)