Effects of Hormones on Sleep
Horm Res 1998;49:125–130
Neuroactive Steroids
As early as 1942, Selye [55] reported hypnotic and anesthetic effects of certain steroid hormones. Previous preclinical research had indicated that apart from classic genomic mechanisms of action, certain steroids are capable of acting through membrane sites such as the GABAA-receptor complex [56]. These hormones are called "neuroactive steroids' and can accumulate in the CNS independent from peripheral sources. A series of studies demonstrated that several neuroactive steroids exert specific effects on the sleep EEG which may be explained by a modulation of the GABAA receptor.
A single oral dose of pregnenolone in young normal men increased the amount of SWS and depressed EEG sigma power. Nocturnal secretion of cortisol and GH remained unchanged [57]. In the rat, subcutaneous injection of pregnenolone enhanced slow wave activity during nonREM sleep and tended to increase the amount of REM sleep [58]. The sleep-EEG changes after pregnenolone in humans and rats are the opposite to some of those induced by GABAA receptor agonists such as benzodiazepine hypnotics. Therefore, these findings suggest an inverse agonistic interaction of pregnenolone or its metabolites at the GABAA receptor. A single oral dose of dehydroepiandrosterone (DHEA) prompted an increase in REM sleep whereas all other sleep variables remained unchanged. EEG spectral analysis showed enhanced EEG activity in the sigma frequency range during REM sleep in the first REM sleep period after DHEA administration. In contrast, the EEG power spectrum of nonREM sleep was not affected nor was the nocturnal secretion of cortisol, GH or testosterone. These findings suggest that DHEA or its metabolites exert a mixed GABAA agonistic/antagonistic effect [59]. In contrast, sleep-EEG changes after progesterone in humans and in rats point to an agonistic modulation of the GABAA receptor. After oral progesterone administration to normal male controls the amount of nonREM sleep increased significantly. The EEG spectral power during nonREM sleep showed a significant decrease in the slow wave-frequency range, whereas spectral power in the higher frequency range tended to be elevated [60]. In rats, progesterone was given intraperitoneally at light onset: Progesterone dose dependently shortened nonREM-sleep latency, lengthened REM latency, decreased the amount of wakefulness and REM sleep and markedly increased pre-REM sleep, an intermediate state between nonREM and REM sleep. Furthermore, progesterone induced dose-related changes in EEG power densities. NonREM-sleep EEG activity was diminished in the lower frequencies and was enhanced in the higher frequencies. In addition, REM-sleep EEG activity was markedly enhanced in the higher frequencies [61]. The sleep-EEG changes after progesterone appear to be mediated via the conversion of progesterone into its major metabolite allopregnanolone [60, 61].
