An elaboration on the interesting facts about REM sleep in the write-up above:
- In sleep studies in cat brains, the definitive brain event of REM sleep is a series of synchronized firings called pontine-geniculate-occipital waves, or PGO spikes. The neurons that start the waves are located in the pons, (around the locus coeruleus,) and synapse on the neurons of the lateral geniculate bodies of the thalamus. From there, the waves are propagated to the occipital cortex. Why is this neat? The lateral geniculate and occipital cortex are parts of the visual system in the brain. Not only that, but the PGO waves are also propagated in to the limbic cortex, an area of the brain responsible for emotional responses and memory among other things. The PGO waves are essentially tickling the emotional and visual parts of the brain, and this non-specific stimulation is considered to be the substrate of the emotional and visual nature of dreaming. In other words, it is probable that PGO waves make dreams what they are.
- Not only are you paralyzed during REM, but your body also fails to carry out thermoregulation. Although we don't exactly know why we sleep (see Sleep: the metanode, why do we need to sleep?, the teleological argument for sleep, and sleep and memory for some perspectives,) it is most likely that it has something to do with the evolutionary advantages of body energy conservation and thermoregulation. However, during REM, your body cannot regulate its temperature. Since infants spend nearly 25% of their early life in REM sleep, this failure to thermoregulate may be one cause of SIDS. What could the evolutionary advantage of REM sleep be, if it puts the body in danger?
- Some neuroscientists have theorized that REM sleep is directly related to memory encoding and storage - and this makes sense considering that the hippocampus is also stimulated by PGO waves, and the hippocampus has been associated with memory functions since the famous case of HM. While studies of selective sleep deprivation (i.e. only REM, not non-REM deprivation) have not shown anything that confirms this, recent work by Matthew Wilson, a neuroscience researcher at MIT has done some work showing that lab rats dream about the mazes they run. Thus, REM sleep might be advantageous as a mechanism for memory consolidation.