800 Similarities of Sleep Macrostructure in Cognitively Normal Elderly and Patients with Traumatic Brain Injury

Abstract

Abstract Introduction The stability of sleep architecture and breathing across nights can depend on factors relating to the integrity of the nervous system. Traumatic brain injury (TBI) represents a sudden-onset dysfunction of the nervous system while normal aging is associated with more gradual changes to the nervous system. While normal aging and history of TBI are both associated with sleep complaints, less is known about the stability of sleep physiology variables in these populations. Therefore, the aims of our study are to determine which sleep variables have greater night-to-night stability in separate populations of individuals with TBI and in cognitively normal older individuals. Methods All volunteers completed 2 consecutive in-laboratory nocturnal polysomnograms (NPSG). The TBI sample (N=35) comprised 71% women and 26% men (average age of 47.3 years). The cognitively normal older sample (N=78) included 74% women and 25% men (average age of 66.4 years). Descriptive statistics and intra-class correlations (ICCs) were calculated for sleep macrostructure variables (total sleep time (TST), sleep efficiency (SE), arousal index (ArI), rapid eye movement (REM), non-REM 1 & 2 (N1, N2), slow-wave sleep (SWS)), and sleep apnea including stage-specific apneas (i.e., AHI4%, AHI3A). Results Among volunteers with TBI, ICCs for sleep architecture variables were: TST (0.68), SE (0.65), ArI (0.92), %SWS (0.77), %REM (0.50), %N1 (0.83), %N2 (0.62). ICC’s for sleep apnea variables were: AHI4% (0.86), AHI3A (0.86), REM AHI4% (0.63), REM AHI3A (0.65). Among cognitively normal older volunteers, ICCs for sleep architecture variables were: TST (0.26), SE (0.29), ArI (0.80), %SWS (0.68), %REM (0.39), %N1 (0.66), %N2 (0.49). ICC’s for sleep apnea variables were: AHI4% (0.91), AHI3A (0.92), REM AHI4% (0.85), REM AHI3A (0.83). All ICCs were statistically significant in both groups, except for %N1 among cognitively normal older volunteers. Conclusion In both populations, ICC’s for arousal index were greater than for TST or SE. Likewise, ICC’s were higher in %SWS and %N1 than for %N2 or %REM. Breathing variables were more stable than architecture variables. REM-specific breathing variables showed comparatively less consistency, possibly the product of lower ICC’s for %REM sleep versus other sleep stages. Support (if any) 5T32HL129953-04, R01AG056682, R01AG066870, R21AG059179, 1RF1NS115268-01, K24HL109156, P30AG059303, P30AG066512, AASM 231-BS-20, R25HL105444