Endocrine Abstracts

JOINT3860

Introduction: Sleep duration is increasingly recognised as a modulator of metabolic health, yet its acute effects on glucose regulation in healthy populations remain poorly understood. This study explores the relationship between sleep duration and post-sleep glucose metrics in healthy 18-year-olds.

Methods: This study included 206 participants (112 females, 94 males) from the COPSAC2000 cohort, with a mean age of 17.7 years. Sleep duration was assessed using actigraphy over a median of 13 nights per participant (IQR: 3), totaling 2,246 individual nights of sleep data. Concurrent glucose metrics (mean, median, variability) were derived from continuous glucose monitoring (CGM) during the waking period following each sleep measurement. Multivariate linear mixed-effects models, adjusted for age, sex, physical activity, social circumstances, and other confounders, were used to assess associations. Subanalyses focused on the 4-hour post-waking and 2-hour pre-waking periods to account for dietary confounding and the cortisol awakening response.

Results: Females slept longer on average than males (6:56 vs. 6:25 h,P < 0.001) and had earlier sleep onset times (00:10 vs. 00:50,P < 0.001). In multivariable modelling, longer sleep duration (per hour of sleep) was associated with higher mean (β = 0.28, 95% CI: 0.05–0.52) and median glucose levels (β = 0.39, 95% CI: 0.15–0.63) and reduced glucose variability (SD: β = -0.12, 95% CI: -0.23 to -0.01; CV: β = -0.16, 95% CI: -0.28 to -0.04). Associations for glucose levels were stronger in the 4-hour post-waking window (mean: β = 0.56, 95% CI: 0.29–0.83; median: β = 0.66, 95% CI: 0.37–0.94). In the 2-hour pre-waking period, glucose levels rose at a population level by 10mg/dL (0.6mmol/l), consistent with the cortisol awakening response. This rise in glucose may explain why longer sleep duration was associated with higher pre-waking glucose variability (SD: β = 0.27, 95% CI: 0.09–0.45; CV: β = 0.29, 95% CI: 0.09–0.48). Mediation analysis revealed that the morning glucose slope partially mediated the association between sleep duration and median glucose levels (indirect effect: 0.019, 95% CI: 0.0015–0.0448). Effect sizes were modest and unlikely to be clinically meaningful.

Conclusion: Sleep duration is associated with glucose levels and variability in healthy adolescents, with longer sleep linked to higher waking glucose and lower variability. The pre-waking glucose rise underscores the role of circadian rhythms in glucose dynamics. While these findings provide mechanistic insights, the modest effect sizes suggest limited clinical relevance in this healthy cohort.