Abstract Introduction A proportion of obstructive sleep apnea patients report persistent non-restorative sleep (NRS) despite effective CPAP therapy with adequate compliance and treatment efficacy. This suggests that factors including sleep architecture disturbances, cortical arousals, and autonomic activation contribute to impaired sleep quality. Using ambulatory sleep EEG technology, this study aims to identify neurophysiological predictors of NRS in CPAP-compliant adults aged 45 years and above. Characterizing these EEG-based biomarkers could enhance risk stratification, guide targeted interventions, and inform adjunctive therapies to optimize sleep quality outcomes beyond respiratory normalization alone. Methods This retrospective observational pilot study analyzed ambulatory sleep EEG data and CPAP compliance reports from 21 patients (n=21; 15 male, 6 female; age ≥45 years) with obstructive sleep apnea. Eligible participants demonstrated adequate CPAP compliance (≥80% nightly usage, ≥70% nights with ≥6 hours use) and treatment efficacy. Sleep architecture parameters, including N3 and REM sleep percentages, cortical arousals per hour, and autonomic activation values, were referenced against mean normative data for age/sex-specific categories from Walsleben et al. (2004) to identify deviations associated with non-restorative sleep. Results Using one-sample t-tests comparing patient values to age- and sex-adjusted normative means, patients demonstrated reduced N3 sleep (61.9% below norm; t = –0.91, p = .37) and reduced REM sleep (57.1% below norm; t = –1.93, p = .07). Cortical arousal index was elevated, with 85.7% above normative expectations (t = –3.30, p = .004). Autonomic activations exceeded normative cutoff (11.4) in 76.2% of patients and were significantly elevated in the 55–69 and 70–91 groups (t = –2.65, p = .045; t = –3.45, p = .007). Findings suggest physiologic hyperarousal despite CPAP treatment. Conclusion CPAP-treated patients with persistent non-restorative sleep demonstrated reduced N3 and REM sleep, elevated cortical arousal burden, and marked autonomic hyperarousal—with over three-quarters exceeding normative reference values—indicating sleep fragmentation. These physiologic patterns suggest identifiable EEG/autonomic signatures of NRS in treated OSA, characterized by inadequate slow-wave/REM expression and increased central arousal. Further research using larger samples is needed to validate these markers and evaluate whether interventions targeting sleep continuity and arousal regulation can improve outcomes in this patient phenotype. Support (if any)
Mayakrishnan et al. (Fri,) studied this question.