A Randomized Crossover Study Investigating the Effect of Sleep Manipulation on Insulin Sensitivity and Cognitive Functions in Adolescents at Risk of Type 2 Diabetes: The Sleep Manipulation in Adolescents at Risk of Type 2 Diabetes (SMART2D) Study

Abstract

Background: Sleep duration is a critical determinant of metabolic and cognitive health, yet the effects of modifying sleep in a population of adolescents with severe obesity and at high-risk for type 2 diabetes (T2D) remain poorly understood. Foundational work I conducted - including two published systematic reviews and a cross-sectional study that piloted a motor-preparation task within the context of sleep duration - identified key gaps in youth sleep, health, and brain function. These findings directly informed the design of the Sleep Manipulation in Adolescents at Risk of Type 2 Diabetes (SMART2D) randomized crossover study.

Methods: SMART2D used a free-living sleep manipulation protocol with continuous monitoring (Actiwatch2®) in adolescents aged 13–18y with risk factors for T2D. After a baseline week (average sleep duration=7h31min/night), participants completed two one-week sleep conditions in which time-in-bed was manipulated by ±1.5h/night: sleep extension (average sleep increase of +1h02min/night) and sleep restriction (average sleep decrease of -1h19min/night). Conditions were delivered in randomized, sex-stratified, counterbalanced order, separated by a washout week (average sleep duration=7h33min/night). Outcome measures included insulin sensitivity (Matsuda index; 2-hour oral glucose tolerance test), glucose response curve (30-minute intervals), serum brain-derived neurotrophic factor (BDNF; fasted sample), food-cue behavioural inhibition (go/no-go task), motor preparation, and postural stability (dual-task paradigm).

Results: Of 43 participants, 36 (84%) completed all phases (53% female; mean age=15.1y; BMI >99.9th percentile; >55% with dyslipidemia). Adherence exceeded 80%, confirming feasibility (2h21min/night difference between sleep weeks). Compared to both baseline and restricted sleep, only sleep extension significantly improved insulin sensitivity (~20%) and increased biphasic glucose responses (~30%), suggesting enhanced insulin secretion. BDNF levels decreased (~14%) during sleep extension, coinciding with peak insulin sensitivity, supporting a modulatory relationship between sleep and metabolic regulation. Sleep restriction impaired response inhibition in food-cue tasks, while sleep extension enhanced control and eliminated food-related biases. Motor preparation was unaffected, but voluntary reaction time and postural stability were impaired following sleep restriction, suggesting reduced neuromotor responsiveness under cognitive load.

Conclusion: Sleep extension offers a simple, non-invasive, scalable intervention to improve insulin sensitivity, inhibitory control, and neuromotor function - supporting its inclusion in preventive health strategies for youth at metabolic risk.