Multidimensional sleep health in type 2 diabetes: The role of sleep variability in glycemic control

Sleep health consists of multiple dimensions, which may contribute differently to glycemic control in persons with type 2 diabetes (T2D). This study evaluated the associations between sleep dimensions and glycemic control in T2D.

Seventy-five participants with T2D were enrolled. Sleep duration, efficiency, sleep midpoint, and sleep variability (standard deviation, SD, of sleep duration) were obtained from 14-day actigraphy recordings. Sleep satisfaction and daytime sleepiness (Epworth Sleepiness Scale, ESS) were obtained from questionnaires. Multidimensional sleep health score (MSH) was created from a sum of "good" sleep scores in each dimension. Hemoglobin A1C (A1C) was obtained. A stepwise regression analysis was performed to investigate independent association(s) between sleep and A1C (natural log transformed).

Participants' mean (SD) age was 54.0 (6.0) years, median (IQR) A1C was 7.6 (6.6, 8.7)% and 57.3 % were using insulin. The mean sleep duration was 6.2 (1.0) hours, median sleep efficiency was 82.2 (77.1, 86.8)%, median sleep midpoint was 03:17 (02:32, 04:10) AM, and median SD of sleep duration was 1.15 (0.92, 1.49) hours. The median ESS was 7 (4,10). The median MSH score was 4 (3, 4). A stepwise regression analysis with additional assessment for potential confounders demonstrated that sleep variability was positive and independently associated with A1C (B = 0.116, p = 0.006; both variables ln transformed), while other sleep dimensions and MSH score were not related to A1C.

Among sleep dimensions, greater variability of sleep duration may have a negative impact on glycemic control in this study. Enhancing regularity of sleep duration could potentially benefit glycemic control in T2D.