Sleep and the Social Brain

While human culture and technologies have undoubtedly contributed to the need for greater resources being funneled into brain development among humans, it also appears that large brains and prolonged development are required to succeed in complex societies. Dunbar (1998) has shown that the association between neocortical volume and social group size among primates is very large (Pearson r=0.76). He (and many other distinguished neuroscientists) has articulated the “social brain hypothesis” which states that we have evolved big brains because of the social demands we face each day. Primates have larger than expected neocortical volumes (given body size) because they need to deal with the complexities of group or social life. Given neocortical volume in humans Dunbar theorized that the average human social network should be about 150 individuals. Ethnographic and social evidence supports this prediction. For example, it is the typical size of a hunter-gatherer group; of a company in a military organization, of a personal network (number of individuals a person knows directly); of a church congregation; of a small business company and so on. Within this large personal network there are hierarchically organized subgroups; alliances, coalitions and cliques etc that reflect differing degrees of familiarity with the individual at the center of the network.

Consider the cognitive demands such a hierarchically organized network places on individuals. Human societies succeed to the extent that problems can be solved cooperatively. Cooperation is based on trust and trust can only be earned over time and after many social interactions have taken place between the parties involved. Those social interactions have to be remembered, archived and recalled repeatedly when evaluating current trustworthiness of one’s potential partner in an enterprise. Maintaining the stability of relationships over time requires constant renegotiation of the terms of the cooperative agreement. People are capable of deception and thus that capacity needs to be taken into account when weighing evidence of trustworthiness. It requires that individuals learn how to read the intentions or minds of others in the group so as to manage conflict; anticipate strategic moves of the other and repair strained relationships and so on.

The capacity to appreciate that another individual has a mind like one’s own, capable of cooperation but also deception etc is called the “theory of mind” or ToM capacity. Dunbar points out that this capacity involves several layers of cognitive complexity. First order ToM involves having knowledge of one’s own mental states (“I believe that…”). Second order intentionality or ToM involves knowledge of another person’s mental states (“I believe that you understand that…”). Third order intentionality involves individual A thinking about what individual B is thinking about A’s thinking (“I intend that you think that I think that we are going to…”) etc. Most scholars working in the area of social cognition think that human beings are capable of perhaps 4-5 orders of intentionality but no more. The computational demands on the brain for this kind of social cognition must be considerable.

The areas of the brain that handle these computational demands around social cognition have been identified as the “social brain network”. The amygdala is important for evaluation of the emotions of self and others, particularly negative emotions. The fusiform gyrus supports rapid recognition and processing of faces. The face, of course, is crucial for social interactions as it emits all kinds of signals concerning the intentions and emotions of the individual. The ventromedial and dorsomedial prefrontal regions are known to support processing of Self-related information as well as understanding the mental states of others (i.e. ToM tasks). The frontopolar region (BA 10) evidences a uniquely complex structure in humans and is one of the evolutionarily most recent regions of the brain in primates. It is involved in multi-tasking, working memory and cognitive branching and therefore may support processing of 3rd and 4th etc orders of intentionality. The superior temporal sulcus contains mirror neurons that support social imitation behaviors and possibly emotional empathy. The temporal – parietal junction supports ToM tasks and language processing. The insula supports empathetic responses as well as moral emotions and the precuneus is involved in a range of activities from mental simulation to self-awareness. Finally the hippocampus is involved in memory functions. These are the major structures involved in social cognition but there are other regions as well that are sometimes included in the social brain and sometimes not—regions such as the ventral striatum and meso-cortical dopaminergic reward tracts and so on.

This “social brain” network handles all of the thinking and emotional work we have to do to keep track of and regulate our interactions with others. These interconnected brain structures allow us to fluently and more or less expertly, daily process huge amounts of strategic social information that is vital to our well-being and the well-being of those closest to us. It is no wonder that a network of interconnected structures specialize in handling all of this social information. What has all of this to do with sleep?

I propose that the sleeping brain uniquely involves this social brain network—or preferentially functions to maintain and repair this particular network. NREM takes the social brain off-line for repair and maintenance and then REM sleep reconnects its key structures and reactivate it in time for waking life. Sleep obviously facilitates repair and maintenance of other brain networks and regions but I am arguing that it is particularly important for functioning of the social brain.

Two decades of neuroimaging studies of the sleeping brain have suggested that roughly speaking the set of structures comprising the social brain are gradually taken off-line after sleep onset and throughout NREM sleep (basically the first half of the night) and then gradually put back together or re-connected and reactivated during each subsequent episode of REM until it fully comes back online after waking (see references list including Dang Vu et al. 2005; Maquet 2000; Maquet et al. 2005; Muzur et al. 2002).

Brain structures such as the dorsolateral prefrontal cortex that help to regulate structures within the social brain are the first that are taken off-line during sleep onset and then the structures of the social brain itself are shut down in piecemeal fashion with each progressive episode of NREM (N1, N2 and N3 slow wave sleep). For example, in the first NREM episode of the night, there is relatively greater delta indexed slow wave activity (SWA) in frontal than in parietal and occipital regions (Werth et al. 1996, 1997; Finelli et al. 2001; Werth et al. 1996). Synchronization of slow wave activity then spreads progressively to posterior cortical zones. Meanwhile delta waves propagate up through subcortical nodes of the social brain network as NREM sleep progresses through the first half of the night.

Progressive deactivation of the brain within NREM sleep is centered on regions in the social brain network including the medial prefrontal cortex, anterior medial areas in BA 9 and 10, orbitofrontal cortices (BA 11) including caudal orbital basal forebrain, anterior cingulate (BA24), bilateral anterior insula, basal forebrain/anterior hypothalamus, bilateral putamen and left precuneus (Dang Vu et al. 2005; Kaufmann et al. 2006).

Sleep Essential Reads

6 Ways to Improve Your Sleep Without Medication

The Brain Benefits of Being a Night Owl (or a Drinker)

All of this de-activation of brain sites is reversed with the onset of REM during the second half of the night. REM involves the reconnection and activation of a huge swath of midline and limbic and paralimbic regions including especially the ventromedial prefrontal cortex and other key nodes of the social brain network (see review of REM neuroimaging studies in the Dang Vu and the Maquet et al references below).

If this speculation is correct then we have the remarkable fact that sleep involves the deactivation, dis-connection/dismantling of social brain structures during the NREM phases of the first half of the night and then the reactivation and reconnection of these structures during the REM phases in the latter half of the night. Presumably this dis- and re-assembly of connectivity between nodes in the social brain network serves the functional purpose of efficiently processing vitally important social information acquired during the day but left unprocessed or unevaluated. If this scenario is even partially correct then sleep is not only important for bodily repair, or brain repair or memory functions in general--it is vitally important for social cognition in particular.