Objective sleep characteristics and continuous glucose monitoring profiles of type 2 diabetes patients in real-life settings

BACKGROUND: Given the significant role of sleep in glycemic control, the association between sleep and glycemic variability determined by continuous glucose monitoring (CGM) is worth investigating among patients with type 2 diabetes (T2D). METHODS: CGM was carried out among 28 T2D patients (aged 62.3±4.8 years, 57% women). Sleep characteristics were assessed by actigraphy within the CGM period. CGM-derived outcomes included glucose level, percentages of time in range (TIR) and time above range (TAR) during the monitoring period. Associations between intraindividual night-to-night variations of sleep characteristics and overall CGM outcomes were analyzed by linear regression. Associations between sleep characteristics in each night and time-matched CGM outcomes were analyzed by linear mixed models. RESULTS: A total 249 person-days of CGM coupled with 221 nights of sleep characteristics were documented. Greater standard deviation (SD) of objective sleep duration (minutes) between measurement nights was associated with higher glucose level (mmol/L, Coefficient [95% CI]: 0.018 [0.004, 0.033], P=0.017), smaller proportion of TIR (% in observation period, -0.20 [-0.36, -0.03], P=0.023), and greater proportion of TAR (0.22 [0.06, 0.39], P=0.011). Later sleep midpoint (minutes from 00:00) was associated with greater SD of glucose during the same sleep period (0.002 [0.0001, 0.003], P=0.037), longer nocturnal sleep duration was associated with smaller coefficient of variance of glucose level in the upcoming day (%, -0.015 [-0.03, -0.001], P=0.041). CONCLUSION: Objectively-determined sleep duration and sleep midpoint, as well as their daily variability, are associated with CGM-derived glucose profiles in T2D patients