Age-dependent suppression of nocturnal growth hormone levels during sleep deprivation

Although the major daily peak in plasma growth hormone (GH) level normally occurs during the early part of nocturnal sleep, it is known that in about one quarter of young healthy men, peaks in circulating human GH occur before sleep onset. Possible factors associated with these nocturnal peaks were investigated in the absence of sleep, including subjectively defined sleepiness, electroencephalographically defined drowsiness and short lapses into sleep, measures of cortisol and temperature. Healthy men between the ages of 20 and 34 years were studied in a between-subjects design, 16 assigned to a sleep deprivation group, and 16 age matched and assigned to a group permitted to sleep. The average GH peak level in the sleeping subjects was 19.9 mu g/l (+/- 8.4 SD). In the sleep-deprivation group there was a wide range of nocturnal GH patterns, from no detectable rise to normal nocturnal levels (average peak of 10.5 +/- 10.0 mu g/l). Within the narrow age range sampled, age, body mass index, temperature and cortisol failed to predict nocturnal GH peak in the sleeping group, but regression analysis found age to be a significant negative predictor of nocturnal GH peak level in sleep-deprived subjects. All sleep-deprived subjects who had peak GH levels over 7 mu g/l were 24 years of age or less. Subjective sleepiness and electroencephalographically defined drowsiness failed to predict either the peak level of GH during sleep deprivation or the cumulative amount measured across the night. These results suggest that the well-known suppressive effect of sleep deprivation on GH secretion is an age-dependent phenomenon which evolves during early adulthood.