Overnight growth hormone secretion in short children: independence of the sleep pattern

to analyze the interrelationships between GH secretion and pattern of sleep.

18 children (10 male, 8 female; mean age 9.1 yr, range 5.1-14.3 yr), with short stature (mean height standard deviation score (SDS) -2.52, range -3.86-(-)1.88; mean height velocity SDS -1.1, range -2.40-(-)0.08), including 9 children with genetic short stature and 9 with idiopathic short stature.

blood samples were taken every 15 min from 2000 h-0800 h, and GH profiles were analyzed by the PULSAR computerized peak identification algorithm; simultaneous sleep was analyzed by electroencephalogram recording.

no significant correlation was noted between GH secretion parameters and any of the electroencephalogram parameters evaluated: stage 1 (S1) percent, stage 2 (S2) percent, slow-wave sleep (SWS) percent, rapid eye movement sleep percent, wakefulness (W) percent, and sleep efficiency (EFF); there was no significant difference in GH secretion between children with EFF less than 76% and those with EFF more than 76% (P > 0.5). Maximal GH peak coincided 9 times (50%) with SWS, 3 times (17%) with S2, 3 times with W, twice (11%) with S1, and once (6%) with rapid eye movement sleep. First GH peak coincided 12 times (67%) with W, 3 times with S2, twice with SWS, and once with S1. There was no significant difference comparing the percentage of sleep stages occurring in the 15 min of maximal GH increment, in the 15 min preceding it, and in those following it; there was no significant difference comparing the percentage of sleep stages occurring in the 15 min preceding the onset of a GH peak and in those following it.

GH secretion in short children seems independent of the sleep stage and efficiency; in children it is possible that GH secretion relates with sleep per se and with neurohormonal changes occurring at nighttime rather than with a specific sleep stage or sleep stage sequence.