Verified by 
Aristotle, when asked what a friend is: "A single soul dwelling in two bodies."
When my best friend from college was diagnosed with metastatic breast cancer, I was awed by the outpouring of practical support that flooded her and her family. As she was an extremely active participant in multiple communities, people came out of the woodwork to run elaborate fundraising drives, cook her meals, and chauffeur her daughter to her many activities.
Typically, when we hear about social connections being linked to better emotional and physical health, it is this sort of practical support we imagine explains the link - many hands making light work when times get tough. We also might think about how our spouses and good friends encourage healthy behaviors - cheering us on in our fitness goals or raising a discouraging eyebrow when we reach for one more drink.
But intriguingly, research has shown that these forms of practical support - called enacted support - are much less important for the link to health than is perceived support, or the perception that your loved ones would provide support if you needed it.
At first blush, this seems nonsensical. How could imagined support be more important for emotional and physical health than actual support?
New research by social neuroscientist Jim Coan provides some important clues.
Your Brain + Threat
When we are feeling threatened or under stress, a number of brain regions are recruited to process these threats. One structure critically implicated in the processing of stressful circumstances is the hypothalamus, which among other things helps govern our body's hormonal stress response. The hypothalamus is the H of the HPA (hypothalamic-pituitary-adrenal) axis, which mediates the release of the stress hormone cortisol, excessive levels of which has been linked to poor health outcomes such as depression, cardiovascular disease, and at least in rats, the death of neurons.
Your Brain + Threat + Friendship
In a now famous study, Coan wondered how the presence of loved ones might alter the brain's response to threatening situations. To introduce threat, he used a threat-of-shock paradigm. While the participants lay in the neuroimaging scanner, they viewed a series of Xs and Os on a screen. When a blue O appeared, they knew they were safe from shock. When a red X appeared, they knew there was a 20% chance of experiencing a slight electric shock to their ankle.
Critically, in some of these trials, the participants' hands were held either by their spouse or by a stranger. Coan found that when he compared the threatening trials with and without hand-holding, the neural regions associated with threat processing were significantly less active in the hand-holding condition, and particularly less so for spouse hand-holding than stranger hand-holding.
People who know me well know that I'm wondering what the effect of replacing spouses with dogs would be.
Coan has since replicated this essential finding (hand-holding of loved ones linked to reduced threat processing) in several subsequent studies, and in some of them the social partners are not touching the participant, but merely present. Together, this research suggests that when our loved ones are near, we react less to stress – and critically, we are less likely to activate structures in the brain that govern our hormonal response to stress*.
Imagine two people walking through life, facing all the usual sorts of stressors small (slighted by a coworker) to large (your elderly parents move in with you). One has few friends, and the perception that s/he must face these stressors alone. The other knows his friends have his back. At each and every threat, the unfriended person may experience elevated stress reactions compared to the befriended person** - a chronically activated autonomic nervous system, greater levels of circulating stress hormones.
As Jim shared in an invigorating talk at the annual meeting of the Association for Psychological Science (APS): "Alone, the world is a lot more demanding".
The paradox of perceived support being more important than enacted support is looking a lot less paradoxical.
Your Friend = You?
So we’ve (partially) solved the perceived/enacted support puzzle. But Jim’s not done with us. In a very recent study, he decided to go one step further and investigate how our brains respond when it is our friend (versus a stranger) who is threatened.
In this paper, he and his co-authors discovered that participants’ brains responded to threats to self and threats to a close friend in a remarkably similar fashion, and that the closer the self-reported closeness in self-identity they experienced with their friend (using a scale similar to the circles at left), the stronger this correlation.
These data suggest something revolutionary about friendship – that friendship may involve a “breach of individual separateness” and that you can see the evidence for this blurring of self and friend in how the brain processes threats to both.
I think we have all felt this on some level, and not just when under stress. After spending a good deal of time with a friend, we find ourselves mirroring their mannerisms, echoing their patterns of speech, finishing their sentences. We can carry on whole arguments with our partners that exist solely in our own heads, parrying each side effortlessly.
enjoined.
Jim concluded his talk with a story about a little girl who was separated from her caretakers. Upon their emotional reunion, she exclaimed, “I feel taller! Do you feel taller?”.
To our brains, the line between us and our beloved others may not be absolute – we become single souls dwelling in multiple bodies, hacking through life’s brambles together.
FURTHER READING
For more on how our social partners may serve as resources to help us regulate our emotions, check out Heather Urry, James Gross, and Phil Opitz's papers here and here.
For more on the importance of social interaction and the dangers of loneliness, consult John Caccioppo's academic work here and his book geared to a public audience here.
For more on the importance of perceived support, check out Shelly Gable's work here.
For more on the health risks of chronic stress, read Robert Sapolsky's highly entertaining Why Zebras Don't Get Ulcers.
REFERENCES
Beckes, L., Coan, J. A., & Hasselmo, K. (2012). Familiarity promotes the blurring of self and other in the neural representation of threat. Social Cognitive and Affective Neuroscience. doi:10.1093/scan/nss046
Coan, J. A., Schaefer, H., & Davidson, R. (2006). Lending a hand. Psychological Science, 17(12), 1032.
Singer, T., Seymour, B., O'Doherty, J., Kaube, H., Dolan, R., & Frith, C. D. (2004). Empathy for pain involves the affective but not sensory components of pain, Science, 1157–1161.
*Caveat: these threat-processing reductions don't always involve the hypothalamus, which seems to be highly sensitive to the quality and type of relationship.
** Of course, the unfriended people probably differ from the befriended people on a lot more than just the presence of friends - reasons WHY they have fewer friends, for instance (personality, physical isolation from family, moving frequently, etc.). To really know whether this is a main effect of friendship, we need some studies with sophisticated controls, and a lot of them.
