0988 Insomnia and COMISA Are Key Drivers of Nocturnal Glycemic Variability in Well-Controlled T2DM: Insights from Multi-Night Sleep and CGM Monitoring

Abstract Introduction Glycemic variability (GV) independently predicts diabetic complications, with fluctuations promoting oxidative stress, inflammation, and endothelial dystunction. Insomnia and obstructive sleep apnea (OSA) are highly prevalent in Type 2 Diabetes (T2D) and may exacerbate metabolic dysregulation, but their impact on nocturnal GV in treated patients with apparently stable glycemic control remains unclear. This study evaluated whether insomnia, OSA, and their co-occurrence (COMISA) have distinct nocturnal GV profiles in a real-world cohort of well-controlled T2D. Methods Fifty-nine patients with T2D (median age 62.0 years) underwent up to 15 consecutive nights of continuous glucose monitoring (CGM; Abbott Libre 2) paired with objective sleep assessment via Belun Ring (BR). Participants were classified as OSA (BR-AHI3% ≥15; N=19), insomnia (BR sleep efficiency < 80% with intake interview; N=11), or COMISA (meeting criteria for both disorders; N=29). Nocturnal GV was quantified using sleep-time coefficient of variation (SLP-CV) and sleep-time continuous overall net glycemic action (SLP-CONGA). Groups differences in median GV were tested, followed by generalized linear mixed-effects models (GLMMs) adjusted for age, sex, BMI, and HbA1c, incorporating participant-level random intercepts to account for night-to-night variability. Results COMISA and insomnia exhibited higher SLP-CV (12.3% and 13.3%) than OSA (8.9%, both p< 0.05), despite comparable glycemic control across groups (median HbA1c 7.3 vs 7.2% vs 7.1% for COMISA, insomnia, and OSA; p=0.81; 83% on metformin; 19% on insulin). Males exhibited higher nocturnal GV than females for both SLP-CV (p=0.07) and SLP-CONGA (p< 0.05). In adjusted GLMMs, COMISA was associated with a 39% increase in nocturnal SLP-CV (β=0.33, p< 0.05) and insomnia with a 43% increase (β=0.36, p< 0.05) compared with OSA. Moreover, male sex (β=-0.24, p< 0.05) and a higher REM% (β< 0.01, p< 0.05) were independently associated with higher SLP-CV. Conclusion This pioneering study identifies Insomnia and COMISA as independent drivers of elevated nocturnal GV in pharmacologically treated T2D, a metabolic risk factor linked to microvascular and macrovascular complications that is not reflected by HbA1c. Multi-night CGM integrated with sleep monitoring reveals hidden nocturnal metabolic instability, suggesting T2D patients with insomnia-related phenotypes and male sex, represent a high-risk subgroup who may benefit from personalized sleep and metabolic interventions. Support (if any)