Purpose of review
This review describes evolving concepts and recent data on the relationship between serum testosterone levels and normal and disordered sleep.
Sex-related differences in circadian rhythms and sleep physiology are in part due to organizational and activational effects of sex steroids. Testosterone affects the organization of circadian rhythms and the timing, but not the duration, of sleep. Increasing testosterone during puberty leads to later bedtimes. The diurnal variation in testosterone depends on sleep rather than circadian rhythm or season. Pubertal onset is heralded, well before virilization, by a luteinizing hormone level at least 3.7 U/l during sleep. Total sleep deprivation lowers testosterone, but sleep restriction only does so if it occurs in the first half of the night. The recovery of testosterone from sleep disruption is impaired in old as compared with young rodents. In men with obstructive sleep apnoea (OSA), low testosterone is related to obesity rather than the OSA itself, and improves with weight loss but inconsistently with continuous positive airway pressure (CPAP). Testosterone treatment only transiently worsens severity of OSA, which need not be considered a contraindication to its use.
Testosterone treatment is unlikely to benefit sleep in men with secondary hypogonadism, for example due to obesity or depression, in contrast to the management of the underlying abnormality.