Effects of sleep restriction periods on serum cortisol levels in healthy men

Abstract

Objectives To clarify effects of partial sleep deprivation (SD) on morning (07:00) serum cortisol concentrations in two protocols that restricted sleep to 3 h/day in healthy adult men. The study was also designed to delineate the relationship between anxiety levels in the morning and slow wave sleep (SWS) periods at night. Methods Ten young adult Han Chinese males were recruited to participate in an ‘earlier-night’ sleep restriction (SR) period (sleep from 00:00 to 03:00) and then a ‘later-night’ SR period (sleep from 03:00 to 06:00). The duration of each SR period was 4 days, followed by a recovery night. The SR periods were separated by 10 days of normal sleep. Blood samples of serum cortisol were drawn at 07:00 during each of two SR periods for six consecutive mornings (for a total of 12 measurements per subject), and anxiety levels were also assessed over the same period by the State portion of the State-Trait Anxiety Inventory (STAI). Sleeping processes were monitored by polysomnogram. Results Serum cortisol levels decreased after SR ( P < 0.05) in both paradigms, with greater decreases evident after later-night sleep loss than after earlier-night sleep loss. Cortisol levels were significantly, negatively correlated to the number of days of earlier-night SR, but not to later-night SR. Anxiety scores increased gradually in both conditions. The time of SWS changed indiscriminately in both paradigms. Cortisol levels returned to baseline after one night of recovery sleep. Conclusions Cortisol decreased in both SR conditions, especially in the earlier-night SR protocol, even though SWS time and anxiety levels changed roughly in the same manner in both conditions. Data suggested that sleep loss at different times of the night affects the hypothalamic–pituitary–adrenal axis (HPA-axis) differentially.