Every REM period is associated with penile tumesecence. Fisher, Gross, and Zuch (1965) first reported the cyclic occurrence of penile erections with REM sleep periods. Later, Karacan (1966) reported that these REM-related erections could be observed in boys as young as 3 years old. These REM-related erections may even occur in infants. They persist throughout the lifespan but are not reliably associated with erotic desire. Sleep-related erections (SREs) occur in all men, regardless of recent sexual activities. Nor are erections related to erotic content of dreams, previous sexual activity, or a full or empty bladder.
Some of the neural circuitry underlying REM-related penile erections has been uncovered. Sites implicated in regulation of SREs include the septal region (in the rat) and the basal forebrain and the lateral preoptic area of the hypothalamus (LPOA). these are regions previously implicated in sexual function and in sleep so no surprizes here.
There is some evidence that REM-related sexual activation may also occur in women as uterine contractions and pelvic thrusting, appearing with REM onset, but too few studies have been done on this topic to draw any firm conclusions. In some men REM-related erections can become painful, such that medical attention is required. These latter cases can be treated with various medications that enhance central nervous system inhibitory activity.
No-one knows why REM-related erections occur. Interestingly theoreticians of REM have not attempted to explain REM related erections even though it would seem that they invite evolutionary approaches given the extreme significance of the penis for theories of sexual selection.
REM-related erections have been observed in all mammalian species where they were looked for, with the interesting exception of the nine-banded armadillo (Affani, Cervino, & Marcos, 2001). In this armadillo, penile erections occur but not in association with REM. It is not clear why SREs are NREM-related rather than REM-related in this species. One possibility is that the armadillo differs from most other mammals in terms of its reproductive behaviors. Armadillos exhibit a rare reproductive phenomenon or strategy called polyembryony, which results in offspring that are genetic clones of one another. Among other things, polyembryony should reduce genetic conflict between siblings.
This latter fact suggests one potential function of REM related erections: they function as a signal of fitness. Many newborns and juveniles will be killed by their parents unless they signal their fitness. This can be done via behavioral displays like crying, begging, smiling, vigorous muscular activity etc or via physiologic displays like coloring, hormonal levels or possibly erections. If genetic conflict is reduced then the need to show signs of fitness or viability will be reduced and thus the need to display erections (a sign of fitness) in tandem with REM would be reduced.
What more potent sign of fitness can exist in a male juvenile than the ability to display an erection? While the fitness hypothesis would explain the existence of nocturnal erections in juveniles it does not explain why they occur in tandem with REM or why they persist into adulthood. One possibility is that they occur in tandem with REM and persist into adulthood because the function of REM itself involves a continuing need to produce behavioral displays of fitness during sleep. Every animal is vulnerable to predators and to conspecifics during sleep. Many male animals display erections during aggressive interactions with conspecifics in order to display dominance or to protect territory. It would not be surprising if the primary site of sexual selection in males was recruited in the never-ending battle to signal fitness in order to maximize reproductive opportunities and to avoid battles over sites and mates.
REM is composed of a suite of behavioral and physiologic reactions that make the animal appear to an onlooker as paradoxically awake and ready for agression. Perhaps this is indeed one of its functions!
Affani, J. M., Cervino, C. O., & Marcos, H. J. A. (2001). Absence of penile erections during paradoxical sleep. Peculiar penile events during wakefulness and slow wave sleep in the armadillo. J Sleep Res, 10, 219-228.
Karacan, I. (1966). Erection cycle during sleep in relation to dream anxiety. Archives of General Psychaitry, 15, 183-189.