Glucose Metabolism after Sleep Restriction, Total Sleep Deprivation, and Recovery

Kurzfassung

Objective Sleep deficiency which is prevalent in shift work has been associated with an increased metabolic disease risk. Experimentally induced sleep restriction has been shown to impair glucose metabolism and has been linked to reduced slow wave sleep (SWS) as a possible causal factor. This study examined (i) whether total sleep deprivation exhibits similar effects on glucose tolerance and insulin sensitivity as sleep restriction, (ii) whether one recovery night after sleep restriction is sufficient to restore impaired glucose metabolism, and (iii) whether the combination of total sleep deprivation with prior sleep restriction shows cumulative effects. Methods Thirty-six healthy volunteers participated in a 12-day study. After one adaptation night and two baseline nights with 8 h of scheduled sleep each, sleep opportunities were restricted for 5 nights either to 5 h (21 participants, 9 females, mean ± SD, age 26 ± 4 yrs, BMI 23.1 ± 1.9) or maintained at 8 h (control, 15 participants, 5 females, age 28 ± 6 yrs, BMI 23.6 ± 2.9). Then, both groups underwent a single 8-h night of recovery sleep, a 38-h period of wakefulness, and a final recovery night. Oral glucose tolerance tests (OGTT) were conducted in the morning following lights on (>10 h fasting) after the second baseline night, after 5 nights of sleep restriction, after the first recovery night, and after 24 h of sleep deprivation. Blood was sampled immediately prior to the OGTT and then at 30-min intervals for 2 h. Polysomnograms were recorded. SWS per night and areas under the curve (AUC) for glucose, insulin, and HOMA were analyzed in each of the two groups with mixed ANOVAs with ‘sleep condition’ (4x) and ‘sex’ (2x) as factors (post-hoc Bonferroni-Holm adjustment). Results Glucose tolerance and insulin sensitivity decreased after sleep restriction (mean ± SEM, glucose Δ 32.5 ± 7.0 mg*h/dl, p<0.0001; insulin Δ 44.9 ± 9.2 mU*h/dl, p<0.0001; HOMA Δ 20.7 ± 3.9, p<0.0001) and remained low after recovery sleep (glucose Δ 17.3 ± 6.8 mg*h/dl, p=0.0139; insulin Δ 24.7 ± 9.2 mU*h/dl, p=0.0102; HOMA Δ 11.3 ± 3.8, p=0.0053) compared to baseline. After 24 h awake, these parameters were not different from baseline in both groups. The amount of SWS in the final sleep restriction night (Δ 0.6 ± 4.7 min, p=0.8949) and the recovery night (Δ 9.2 ± 4.7 min, p=0.0534) did not significantly differ from baseline. Conclusion Sleep restriction for 5 nights negatively impacts glucose metabolism. This impairment appears to occur independently of SWS, and to outlast a single night of recovery sleep. In contrast, prolonged wakefulness does neither acutely affect glucose metabolism nor exhibit cumulative effects with prior sleep restriction. Chronic sleep loss and acutely extended wake duration appear to activate different regulatory responses in glucose metabolism.