0459 Restriction of Breathing in OSA Does Not Affect Sleep Spindle Density

Christian Harding,Breanna Holloway,Christopher Schmickl,Pamela DeYoung,Crystal Kwan,Sonia Ancoli-Israel,Sarah Banks,Atul Malhotra

DOI: https://doi.org/10.1093/sleep/zsae067.0459

IF: 6.313

2024-04-20

SLEEP

Abstract:Abstract Introduction Obstructive sleep apnea (OSA) is associated with major neurocognitive sequalae including impaired memory. OSA patients are deficient in NREM sleep spindles (11 Hz to 16 Hz electroencephalography (EEG) sigma oscillations) which have been shown to promote memory. This finding is a result of less time spent in deeper sleep states in which most spindles occur but also a reduction in spindle density, particularly during the later portion of the night. We tested the hypothesis that spindle density deficiencies are driven by restriction of breathing preventing spindle generation. Methods 109 adults (ages 65-83, 53% female) with normal cognition underwent overnight polysomnography (PSG). Presence of sleep apnea was defined using AASM recommended criteria (apnea-hypopnea index >15, n=81[OSA] vs 28[controls]). The right occipital channel EEG was bandpass filtered between 0.3-35Hz. Following rejection of movement artifacts, spindles were detected in NREM sleep stages 2 and 3 (N2/N3) using complex demodulation to extract high sigma activity signals from background sigma noise. Spindles were classified depending on whether the patient was in eupnea or apnea-hypopnea. Results Overall spindle density in OSA patients did not differ from overall spindle density in controls (2.86 vs 3.17 spindles/minute, t(57.4)=0.49, p=0.14). Apnea-hypopnea spindle density in OSA patients did not differ from eupnea spindle density in controls (3.10 vs 3.17 spindles/minute, t(70.8)=0.31, p-value=0.76). Eupnea spindle density in OSA patients was significantly lower than eupnea spindle density in controls (2.57 vs 3.17 spindles/minute, t(54.4)=2.94, p-value< 0.01). Conclusion Contrary to our hypothesis, sleep spindle density was not reduced in OSA patients during apnea-hypopnea compared to healthy controls during eupnea, suggesting restriction of breathing does not affect spindle generation, but was reduced in OSA patients during eupnea. Whilst we did not observe an overall OSA-related spindle density deficit, our results suggest that previous findings of lower spindle density in OSA patients may in fact result from changes in spindle behaviour during eupnea. This observation raises the interesting possibility that using OSA interventions to prevent apnea-hypopnea occurring may not recover normal spindle density. Alternatively, our results could suggest that undetected N2/N3 sleep disruptions are occurring when OSA patients are breathing, skewing eupnea spindle density estimates. Support (if any) 1R01AG063925-01A1

neurosciences,clinical neurology