Sebastian C. Holst, PhD Submitted: Session description In Angelman syndrome (AS), qualitative caregiver reports often highlight sleep as a top priority for intervention. Nevertheless, objective sleep assessments remain limited, which is particularly true in multi-site AS trials, where home-setting variability complicates recording quality and AS-specific high-amplitude delta oscillations challenge visual sleep staging. Avoiding such analysis bias requires a specialized processing approach to accurately determine sleep-wake physiology. Methods: We analyzed overnight home polysomnographic EEG recordings from two pediatric cohorts, FREESIAS and TANGELO (AS n=41; TDC n=13). To ensure valid comparisons despite variable recording start times and fragmented sleep, we implemented a Major Sleep Episode (MSE) detection to more accurately compare night time sleep across AS and controls. We also classified sleep spindles using a semi-automated approach. Between-group comparisons were performed using non-parametric Mann-Whitney U tests. Results: Total sleep time was ~30% (2.5-3 hours) shorter in AS compared to TDC (p<0.001), regardless of whether whole-night or MSE-restricted windows were used. In contrast, sleep efficiency appeared reduced in whole-night analyses but was comparable between groups within the MSE (AS: 89% vs. TDC: 95%; p=0.30). As expected, spectral analysis revealed a profound, ~5-fold elevation in absolute delta (0.5–4 Hz) power in AS (p<0.001). However, normalization to pre-sleep wake baselines demonstrated that homeostatic sleep pressure dissipation is remarkably preserved, with both groups showing comparable overnight declines (AS: 48% decline [Interquartile Range: 40–57%] vs TDC: 55% [Interquartile Range: 48–67%]; p=0.25). Instead, AS was characterized by significant abnormalities in sleep spindles, which were fewer, shorter, and higher in spindle oscillation frequency than those of TDCs (p<0.001). Conclusions: Our at-home EEG analysis revealed a fundamental dissociation in AS sleep. While sleep was generally shorter in AS, the homeostatic mechanisms governing sleep pressure (Process S) appear to remain intact, the circuits controlling spectral architecture and sleep spindle generation appear abnormal.