Title: CHANGES IN SUBJECTIVE AND OBJECTIVE SLEEP MEASURES DURING RECOVERY IN PEDIATRIC CONCUSSION

Background: Sleep problems after concussion have been associated with longer recovery times and more severe symptoms. Early research has primarily evaluated sleep via subjective measures; however, actigraphy data may help objectively understand sleep patterns after pediatric concussion. Purposes: Our study examined changes in subjective and objective sleep measures during pediatric concussion recovery. We also evaluated whether sleep measures at initial assessment predict concussion symptom resolution time. Methods: Pediatric patients evaluated within 14 days of concussion were given a commercial activity tracker to be worn full-time, including during sleep, through return-to-play (RTP) clearance. During the initial and RTP assessments, patients completed the Pittsburgh Sleep Quality Index (PSQI), self-reporting sleep duration, time spent in bed, and sleep quality. PSQI score >5 indicates poor sleep quality. Activity tracker-recorded sleep duration and time spent in bed from the first 3 nights after initial assessment and 3 nights prior to the RTP clearance assessment were also used for analysis. Results: Patients (n=32, 15.1±1.7 years, 41% female) completed initial and RTP clearance assessments (Table 1). On the PSQI, patients reported significantly worse overall sleep quality despite longer sleep duration and more time in bed at initial assessment relative to RTP assessment (Table 2). A majority of patients reported PSQI score >5 at both initial and RTP assessments (Table 2). There were no significant differences across time for sleep duration or time in bed as measured by actigraphy (Table 2). Patients perceived that they spent longer in bed compared to objective measures with initial assessment, and perceived longer sleep duration compared to objective measures for both initial and RTP assessments (Table 3). In each regression model, higher symptom severity was significantly associated with longer symptom resolution time (Table 4). In addition, longer time in bed after the initial assessment measured by actigraphy was significantly associated with longer symptom resolution time (Table 4). Conclusion: Patients reported obtaining more sleep and spending longer time in bed at the initial assessment compared to actigraphy measures, and a majority of patients reported poor sleep quality throughout recovery. Concussion symptom duration increased by 1 day for every 13 additional minutes spent in bed, as measured by actigraphy. Increased time in bed may indicate poor sleep habits during recovery. Both subjective and objective assessment of sleep following concussion offer valuable, yet differing, information. Future research assessing the impact of sleep intervention after concussion on symptom resolution should be considered. [Table: see text][Table: see text][Table: see text][Table: see text]