A study from UC Berkeley suggests that oxytocin—the hormone often associated with romance and bonding—also plays a key role in how friendships form, particularly in the earliest stages.
Oxytocin is released in the brain during sex, childbirth, breastfeeding, and many kinds of social interaction. It is linked to feelings such as trust, closeness, and attachment, and it is sometimes promoted as a “cuddle hormone” that people can boost through touch, music, or exercise. At the same time, researchers have also connected oxytocin to aggression and social division, showing that its effects are more complex than simple warmth and comfort.
Much of what scientists know about oxytocin and bonding comes from studies of prairie voles. These small rodents are a useful model because they form stable, selective relationships—something relatively uncommon among mammals and, in some respects, unusually similar to humans.
While many prairie vole studies have focused on mating bonds, the UC Berkeley team, led by integrative biology and neuroscience professor Annaliese Beery, focused on selective peer bonds that resemble friendships. Understanding how such bonds form and persist could help clarify social difficulties seen in conditions such as autism and schizophrenia.
In the new work, Beery and graduate student Alexis Black examined prairie voles that were genetically engineered to lack oxytocin receptors. Typical prairie vole “friends” show strong affiliative behavior, including huddling, grooming, and sitting close together. The researchers found that voles without oxytocin receptors could still eventually form these peer bonds, but it took them much longer than usual.
According to the team, oxytocin appeared to matter most during the early formation of a relationship and in social selectivity—essentially, the ability to distinguish “my partner” from “a stranger.” When researchers rearranged animals into new social groups, voles with disrupted oxytocin signaling were quicker to lose track of their original partners and showed weaker long-term selectivity.
The study also found that these receptor-deficient voles did not display the same social “reward” patterns seen in typical animals. They were less motivated to seek contact with familiar peers and were also less avoidant of unfamiliar animals. In addition, they tended to be less aggressive toward strangers, suggesting that oxytocin can contribute to both sides of social selectivity: drawing individuals toward familiar partners while helping them reject unfamiliar ones.
The findings build on earlier research that complicated the idea of oxytocin as a universal bonding chemical. Prior studies using the same genetically modified voles suggested that oxytocin signaling is not strictly required for long-term mate bonding or parenting behavior—though it can affect how quickly such bonds form. Other work showed that voles lacking oxytocin receptors took roughly twice as long as normal animals to establish a mating bond.
To understand friendship bonding more directly, the Berkeley group ran three main experiments. First, they measured how long it took an animal to form a preference for a specific peer. Normal voles typically develop a strong partner preference after about 24 hours of living in close proximity. The voles without oxytocin receptors showed no clear preference after a day and, in some cases, needed up to a week to develop one.
Next, researchers tested long-term bonded voles in a mixed-group, “party-like” enclosure with multiple rooms connected by tubes. In that setting, typical voles stayed near familiar partners at first and only later began interacting more broadly. The oxytocin receptor–deficient animals mixed with others immediately, behaving as if they had not formed a particular attachment to any individual.
In a third test, the team measured motivation by requiring voles to press a lever to gain access to either a partner or a stranger. Female voles with normal oxytocin signaling typically worked harder to reach a partner than a stranger in both mate and peer contexts. Voles lacking oxytocin receptors still showed higher motivation for a mating partner, but not for a peer—suggesting that friendship-like bonds may rely more heavily on oxytocin-dependent mechanisms than mating bonds do, at least in this model.
The researchers also explored what happens to oxytocin release in the brain when its receptor is missing. Using a newly developed oxytocin nanosensor created in Markita Landry’s UC Berkeley lab, the team measured oxytocin activity in the nucleus accumbens, a brain region important for social reward across many species.
One possibility was that oxytocin release might increase to compensate for the missing receptor, potentially interacting with other related signaling systems. Instead, the sensors indicated the opposite: oxytocin was released in lower amounts and from fewer sites in the nucleus accumbens.
The study was published on Aug. 8 in Current Biology. The researchers argue that oxytocin’s most critical role may not be determining how social an animal is in general, but rather shaping who it chooses to be social with—and how quickly and reliably those selective bonds are formed and maintained.
