Sleep Efficiency Predicts Next-Day Glycaemia and Daytime Glycaemia Influences Sleep in Free-Living Adults at Risk of Type 2 Diabetes

Aim: To investigate how sleep features influence next-day diurnal glucose homeostasis and vice versa in free-living adults at risk of type 2 diabetes. Materials and Methods: This observational repeated-measures study included 388 adults aged 25–65 years (50% women) with overweight or obesity (BMI ≥ 25.0–< 40.0 kg/m²). Sleep and glucose homeostasis were simultaneously assessed over 14 days using wrist-worn accelerometers and continuous glucose monitors. Linear mixed models evaluated day-level associations between sleep metrics—wake-up time, sleep period time (i.e., time from sleep onset to wake-up) and sleep efficiency ([total sleep time/sleep period time] × 100)—and diurnal glucose metrics, including mean glucose and its standard deviation (glycaemic variability). Results: We analysed 3942 valid person-days. Each 1% increase in sleep efficiency was associated with lower next-day mean glucose (B [95% CI] = −0.05 [−0.08, −0.01] mg/dL; p = 0.007). Each 1 h delay in wake-up time was linked to reduced next-day glucose variability (−0.24 [−0.38, −0.10] mg/dL; p = 0.001). Conversely, each 1 mg/dL increase in daytime mean glucose was associated with later wake-up time (0.008 [0.002, 0.014] h; p = 0.008), longer sleep period time (0.006 [0.000, 0.012] h; p = 0.039) and lower sleep efficiency (−0.05% [−0.08%, −0.01%] %; p = 0.005) the subsequent night. Each 1 mg/dL increase in glucose variability was associated with earlier wake-up time (−0.02 [−0.03, −0.01] h; p < 0.001). Conclusions: This study provides evidence that sleep and glucose dynamics are temporally associated in free-living adults at risk of type 2 diabetes. These findings underscore the potential of combining sleep and glucose metrics to inform cardiometabolic risk prevention strategies.