1174 SLEEP ARCHITECTURE FOLLOWING TRAUMATIC BRAIN INJURY: A SYSTEMATIC REVIEW AND META-ANALYSIS

Abstract Introduction: Sleep architecture alterations are present soon after traumatic brain injury (TBI), but what remains less clear is whether these alterations persist chronically (long-term) after injury. It is important to identify whether sleep architecture differs after chronic TBI, as sleep may serve as both a marker of recovery and a point of intervention. We sought to address this question using the Meta-analysis of Observational Studies in Epidemiology technique. Methods: We performed two independent searches (MEDLINE and EMBASE) and identified 5556 potentially relevant studies. Fifteen of these studies assessed sleep with at least one night of PSG in both a TBI and a control group. Statistical analyses were performed using Comprehensive Meta-Analysis software using standard mean differences (SMD). Data were pooled using inverse variance weighting and random effects model. Results: Overall, the TBI group had significantly more SWS than controls (SMD: .44; CI: .06, .82). There were no significant differences in N1, N2 or REM between TBIs and controls. Injury severity was included as a moderator (comparing mild TBI and moderate-severe TBI to controls). The moderate-severe group had significantly more SWS than controls (SMD: .65; CI: .08, 1.2). On the other hand, the mild TBI group had significantly less REM than controls (SMD = -.38; CI: -.72, -.04). There were no differences in N1 or N2 between groups. The use of an adaptation night was also included as a moderator. The TBI adaptation night group had significantly more N1 (SMD: .57; CI: .34, .79) and more REM than controls (SMD: -.40; CI: -.74, -.06). Lastly, the non-adaptation TBI group had significantly more SWS than controls (SMD: .87; CI: -.26, .61). No differences in N2 were observed. Conclusion: Post-TBI sleep staging is impacted by injury severity and the use of an adaptation night prior to data collection. These data unify the findings of previous investigations that had found differing sleep staging in chronic TBI. These data also provide guidelines for future work on sleep following chronic TBI. Support (If Any): NRC Research Associateship Program.