Sleep timing may modulate the effect of sleep loss on testosterone.

BACKGROUND: Sleep loss has been shown to reduce secretory activity of the pituitary-gonadal axis in men, but the determinants of this effect are unknown.

OBJECTIVE: To discriminate the effects of sleep duration and sleep timing on serum concentrations of luteinizing hormone (LH), testosterone (T) and prolactin (PRL).

METHODS: Fifteen young, healthy men (27·1 ± 1·3 years; BMI, 22·9 ± 0·3 kg/m(2) ) were examined in a condition of sleep time restriction to 4 h (bedtime, 02:45 -07:00 h) for two consecutive nights and in a control condition of 8 h regular sleep (bedtime, 22:45-07:00 h). After the second night, serum concentrations of LH, T and PRL were monitored over a 15-h period. In addition, these hormones were measured in serum samples obtained in a further experiment in eight healthy men (24·5 ± 1·1 years; BMI, 23·7 ± 0·6 kg/m(2) ) in the morning after one night of total sleep deprivation, of 4·5 h sleep (bedtime, 22:30-03:30 h), and of regular 7 h sleep (bedtime, 22:30-06:00 h).

RESULTS: Serum LH, T and PRL concentrations showed characteristic diurnal variations across the 15-h period without any differences between the 4- and 8-h sleep conditions. However, total sleep deprivation and 4·5 h of sleep restricted to the first night-half markedly decreased morning T and PRL concentrations (both P ≤ 0·05).

CONCLUSION: Collectively, our data suggest that the effect of sleep restriction on pituitary-gonadal secretory activity may be modulated by sleep timing. While sleep loss in the early part of the night does not affect T and PRL, early awakening and wakefulness during the second part of the night reduces morning circulating T and PRL concentrations.