THU341 Associations Of Longitudinal Nighttime Sleep Characteristics With Nighttime/Daytime Glucose Values As Assessed By Continuous Glucose Monitoring In Type 2 Diabetes

Abstract

Abstract Disclosure: Y. Li: None. K.K. Danielson: None. J.C. Park: None. J. McAnany: None. F.Y. Chau: None. B. Prasad: None. E.C. Hanlon: None. S. Reutrakul: None. Background: Sleep disturbances have been shown to be associated with incident diabetes and poor glycemic control. Recently, evidence emerged that nightly sleep may affect glucose levels both during that night and the following day. In this study, we explored the association of serial measures of nightly sleep parameters with night/day glucose levels in patients with type 2 diabetes (T2D). Methods: Baseline data of 28 T2D patients participating in an ongoing study exploring the relationship between diabetes retinopathy (DR) and sleep (NCT 04547439) were included. Sleep was objectively recorded using 14-days actigraphy. Sleep parameters nightly included sleep duration, sleep efficiency, wake time after sleep onset (WASO), and mid-sleep time (MST, time between sleep time and waketime). Glucose was recorded using blinded continuous glucose monitors (CGM). CGM data were divided into nighttime and daytime glucose values, utilizing the sleep time from actigraphy. CGM parameters included mean glucose, percent time spent in range (TIR%; glucose values 70-180 mg/dL), percent time above range (>180 mg/dL), and percent time below range (<70 mg/dL). Glucose variability indices included percent coefficient of variation (%CV) and Continuous Overall Net Glycemic Action (CONGA). Longitudinal mixed linear regression models, adjusting for age, sex, insulin use and DR status, were used to determine the associations between sleep parameters and nighttime/subsequent daytime CGM values. Associations were considered statistically significant at p<0.01 due to the multiple comparisons. Results: Mean (SD) age was 55.0 (6.5) years, A1C was 7.8 (1.5)%, 57% of the participants were female and 60.1% had DR. Average sleep duration was 6.1 (1.0) hours, sleep efficiency was 79.8 (9.6)%, WASO was 49.0 (27.1) minutes and MST was 3:25 (1:23) AM. There were a total of 281 nights and 254 days of CGM recordings. For nighttime CGM analyses, sleep duration (mins) was significantly longer in %TIR of ≥70% vs. <70% (β=57, SE=13, p=0.0005), but there were no associations of duration with other CGM parameters. WASO (log) was positively associated with CONGA (β=2.3, SE=0.9, p=0.008). Earlier MST was associated with greater glucose variability including higher CONGA (β=−1.6, SE=0.4, p<0.0001), and %CV (β=−1.6, SE=0.4, p<0.0001). For the following day CGM analyses, only the association between earlier MST and greater time spent below range (β=−0.7, SE=0.2, p=0.006) was significant. Conclusion: Longer nightly sleep duration was associated with higher nighttime %TIR, while poor sleep continuity was associated with higher nighttime glucose variability. In contrast to previous findings, earlier sleep timing was associated with unfavorable glucose profiles both during nighttime and daytime. These pilot data highlighted the associations between nighttime sleep and metabolic control, and should be explored in a larger cohort. Presentation: Thursday, June 15, 2023