Verified by 
In an interesting study published recently in Current Biology (van der Helm E, Yao J, Dutt S, Rao V, Saletin JM, Walker MP. REM Sleep Depotentiates Amygdala Activity to Previous Emotional Experiences. Curr Biol. (2011) Dec 6;21(23):2029-32) investigators from the Sleep and Neuroimaging Laboratory, Department of Psychology, University of California, Berkeley, CA argued and presented evidence consistent with the claim that REM sleep functions in part to facilitate emotional regulation.
The authors point out that REM sleep is associated with a massive reduction in noradrenergic tone in forebrain centers including the amygdala. (The amygdala is known to be involved in processing of emotions—especially negative emotions like fear and stress). In addition, processing of emotional memories via amygdalar-hippocampal interactions takes place during REM. Thus two events take place in REM that are crucial to daytime emotional regulation: 1) reactivity of the amygdala is down-regulated due to suppression of central noradrengic tone and 2) emotional memories are re-activated in amygdalar-hippocampal networks during REM.
The latter process involves the processing of memories in absence of norepinephrine such that they are stripped of their stress-related arousing capacities before being stored in long term networks. In short, REM is hypothesized to depotentiate amygdalar reactivity and re-process emotional memories in a state where noradrenergic activity is suppressed thereby decreasing the overall intensity of negative emotional memories.
In the study van der Helm et al. performed two repeat fMRI tests on 34 volunteers (test 1, test 2), separated by 12 hour intervals containing either a night of EEG-recorded sleep or a waking day. During each test, participants viewed and rated the subjective emotional intensity of 150 standardized emotional pictures. Importantly, participants viewed the same stimuli at both test sessions, affording a measure of change in emotional reactivity to previously experienced affective stimuli (test 2 - test 1) following wake or sleep.
Results showed that those who slept between the image viewings reported a significant decrease in their ratings of the intensity of the emotional images as well as a decrease in amygdala reactivity while the participants who were awake between viewings/ratings demonstrated increases in both ratings and amygdala reactivity.
Interestingly the extent of overnight decrease in both amygdala reactivity and ratings was significantly correlated with the extent of reduced prefrontal EEG gamma activity (a biomarker of arousal and possibly central noradrenergic activity) during REM such that those with the lowest levels of REM-gamma expressed the largest overnight decrease in emotion reactivity.
Thus the authors have shown that amygdalar activity is decreased in REM and that this is associated with both a reduction in behavioral ratings of intensity of emotional pictures and a reduction in EEG gamma power in PFC.
Could it be (as the authors seem to argue) that REM might function in part to reduce intensity of negative emotions via REM related reductions in amygdalar activity (and the associated reduction in central adrenergic tone)?
While I find that evidence strong and convincing, the data the authors present on prefrontal activity during and after REM sleep suggests that the reduction in emotional reactivity associated with sleep may be a by-product of other processes occurring in REM.
For example, the authors also reported a significant group (wake, sleep) by test (test 1, test 2) amygdala connectivity interaction with the ventral-medial prefrontal cortex (vmPFC), that indicated an overnight increase in functional connectivity in the sleep group and a corresponding decrease in functional connectivity across the day in the wake group. These data suggest that the overnight reduction in amygdala activity was due to or related to an increase in vmPFC connectivity.
So, does REM function to increase functional connectivity of vmPFC with other parts of the brain (including but not restricted to the amygdala)? If so a by-product would be better overall regulation of all parts of the brain connected to PFC including the amygdala.
Decades ago sleep scientists (e.g., Ernest Hartmann) noted that REM was associated with an overall reduction in forebrain catecholaminergic activity and that one function of REM could therefore be metabolic repair and retuning of these circuits. Catecholaminergic circuits project from limbic to prefrontal sites so the end result would be a better functioning PFC.
PFC functioning is crucial for a huge array of higher cognitive functions and its dysfunction is the factor most commonly implicated in uniquely human disorders like schizophrenia. Perhaps the secret of the function of REM will be solved when we learn more about the evolutionary functions of the PFC.
