If one wished to discover the function of sleep a particularly important place to look would be infant sleep. In the first months of life, the human infant spends most of its 24-hour day asleep (15-18 hrs.)-this despite the fact that most species of primate infants, including humans, evolved in environments where predation on the young was intense.
You would think therefore that Mother Nature would produce an infant that was able to do something (like run or produce some poison etc) to protect itself right after birth given the immense dangers it faced right from birth. Instead the primate infant is born utterly defenseless and produces nothing remotely approaching something that could be called defensive behaviors. It does however produce abundant amounts of sleep.
But its sleep does not occur in isolation. Most primate infants sleep with their mothers. Up until the modern era this co-sleeping patterns was true of human infants as well. While co-sleeping practices have partially declined in the West, it is still a very common practice. Many parental interactions therefore necessarily occur in the context of a sleeping or just awakened infant. Many nursing episodes, for example, occur right after an awakening. When maternal night-time interventions eventuate in a nursing episode the infant returns to sleep but may continue to nurse while still asleep. This sleep-related nursing appears to selectively activate AS/REM sleep.
To some extent then the sleep-wake patterns of the infant depend upon or contribute to the quality its interactions with its caretaker or mother. A series of studies back in the 1970s and 80s on effects of maternal-infant separation showed that sleep was very dramatically altered by that trauma (e.g., Reite & Short, 1978) These studies have conclusively shown that measures of AS/REM sleep (but not NREM or slow wave sleep -SWS) were selectively influenced after maternal separation. There is an initial increase in AS/REM times and then a dramatic reduction after separation. In one of the Reite et al. studies, 10 pigtailed infant monkeys were separated from their mothers at 26 weeks. Sleep measures were taken at baseline (pre-separation) and then prospectively during a 4 days separation period and a 4 day reunion period with the mother. REM duration decreased by almost 60% during separation. Upon reunion, however, REM returned to pre-separation levels.
These old studies suggest that REM sleep has something to do with enhancing maternal-infant bonds. Scher (2001) studied sleep patterns in non-securely attached human infants. Scher (2001) reported that 55% of secure and 60% of insecure (almost exclusively resistant) infants were described by their mothers as night wakers, though this difference was not confirmed by actigraphic data. Whereas only 6% of the mothers of insecurely attached infants thought that their child had settling difficulties when going to sleep, 43% of ‘dependent secure' (i.e., B4: secure infants who nevertheless exhibit some ambivalence as well) and 23% of the mothers of other secure infants reported that their infants had settling difficulties. Infants with frequent night awakenings scored higher on contact-maintenance in the Strange Situation than non-night waking infants. Unfortunately, Scher (2001) did not report percent active (the infant precursor to REM) sleep as a function of attachment status.
As every new parent knows infant night-wakings are typically associated with signaling (e.g. crying and vocalizations) designed presumably to elicit a maternal intervention. In this way night-wakings whether with or without signaling could influence both formation of attachment bonds and developing sleep states and their patterning. Given the fact that night wakings, particularly those with associated signaling, most commonly emerge from a REM episode, REM times may be influenced by pressures to form the attachment tie.
Suppression of active/REM sleep in infant animals prevents normal limbic brain development and emotional functioning (Mirmiran et al., 1983; Vogel, Feng, & Kinney, 2000) but we do not know if it prevents formation of bonds.
In any case these considerations concerning the ways in which infant sleep interacts with maternal behaviors suggests that one function of infant sleep is to manipulate maternal behaviors. But it seems Mother Nature could have accomplished this goal much more effectively by having the infant be born in a more mature state that it is. I conclude that infant sleep does indeed influence attachment with the mother 9and the father too!) but the functions of infant sleep remain a mystery.
Mirmiran, M., Scholtens, J., van de Poll, N. E., Uylings, H. B., van der Gugten, J., & Boer, G. J. (1983). Effects of experimental suppression of active (REM) sleep during early development upon adult brain and behavior in the rat. Brain Research, 283, 277-286.
Reite, M., & Short, R. A. (1978). Nocturnal sleep in separated monkey infants. Arch Gen Psychiatry, 35(10), 1247-1253.
Scher, A. (2001). Attachment and sleep: a study of night waking in 12-month-old infants. Dev Psychobiol, 38(4), 274-285.
Scher, A. (2002). Mother-infant relationship as a modulator of night waking. In P. Salzarulo & G. Ficca (Eds.), Awakening and Sleep-Wake Cycle Across Development (pp. 187-198). Amsterdam/Philadelphia: John Benjamins Publishing.
Vogel, G., Feng, P., & Kinney, G. (2000). Ontogeny of REM sleep in rats: Possible implications for endogenous depression. Physiol Behav, 68(4), 453-461