A new wave of vole research is reframing oxytocin’s role in social life: the hormone is less a universal “cuddle chemical” and more a fine-tuner of selectivity that helps animals invest in specific relationships while turning away outsiders. In female prairie voles lacking oxytocin receptors, friendships form late, wobble easily, and fail to trump contact with strangers, according to new findings reported by University of California, Berkeley neuroscientists and collaborators and summarized by The Transmitter as a study just out in Current Biology. The work suggests oxytocin receptors are not essential for general sociability or even romantic pair bonds—but are crucial for maintaining loyal, selective friendships that endure distractions in a crowd. Those insights, scientists say, could sharpen how we think about human friendship, loneliness, and the design of social environments in Thailand and beyond.
Prairie voles, long the darling of attachment research for their rare monogamy, have helped science map the neurochemistry of love. But recent genetic experiments shook that story. In 2023, CRISPR-edited voles with no oxytocin receptors still paired up and raised pups, hinting the “love hormone” may not be the indispensable glue once believed. The new study flips the lens from romance to platonic ties, showing the oxytocin system may be more about who we choose—and keep choosing—than whether we connect at all. As principal investigator Annaliese Beery told The Transmitter, oxytocin appears to support “selectivity,” a core human friendship feature: spending more time with a particular person and resisting the pull of newcomers when it matters most (The Transmitter).
Prairie voles matter to Thai readers because their unusual social lives offer biological clues relevant to everyday health questions—how friendships buffer stress, why some bonds endure, and why loneliness bites. In Thailand’s post-pandemic reality, the World Health Organization country office notes that social isolation affects about one-quarter of older adults, and 5–15% of adolescents report chronic loneliness, problems linked to poorer mental and physical health (WHO Thailand). Understanding how the brain supports selective, stable friendships can inform social programs in schools, workplaces, and community centers—without overselling hormone sprays or quick fixes that haven’t panned out in clinical trials.
The new work grew from a paradox. In 2023, a team led by psychiatrists and neuroscientists at the University of California, San Francisco reported that oxytocin receptor knockouts still form pair bonds and care for offspring, contradicting decades of pharmacological studies in prairie voles (Neuron, 2023; see also full text at Cell). That group, which includes co-author Devanand Manoli, also observed a behavioral split: proper bonding seemed to involve two distinct modules—building a relationship and rejecting alternatives. In The Transmitter’s account, Manoli says oxytocin receptors are especially important for that second module: keeping loyalty by rejecting outsiders. Without receptors, engineered voles sometimes “huddle” with strangers more than with partners, a hint that oxytocin’s job extends beyond warm feelings to the social calculus of allegiance.
Beery’s team pressed that idea into the terrain of friendships. They raised oxytocin-receptor-null females from weaning to adulthood with same-sex peers and observed how friendships formed, endured, or faltered. Compared with wild-type animals, the engineered voles were slower to cement bonds; when new animals arrived in a “cocktail party” test, they lost track of their friends and drifted. When given the chance to press a lever to visit either a cage mate or a stranger, they showed little selectivity—pressing similarly for both—unlike typical voles, which work harder to reunite with friends. Those patterns match a world most humans recognize: friendship is more than liking people; it’s choosing “us” and, when needed, saying no to tempting “them.” As Beery put it, oxytocin signaling can promote “rejection of outsiders” in the service of friendship stability (The Transmitter).
The team also peeked under the hood of the brain’s reward hub, the nucleus accumbens. Using an emerging nanosensor technology on stimulated brain slices, the researchers reported lower oxytocin release and fewer release sites in receptor-null voles compared to unedited animals—evidence against the idea that brains simply “turn up” oxytocin when receptors are missing. That tools-based insight tracks with a broader push to directly measure neurochemicals during social behavior, including UC Berkeley’s recent nanosensor efforts across vole social circuits (UC Berkeley QB3). Yet, as University of Colorado Boulder neuroscientist Zoe Donaldson cautioned, these techniques are “at the technological cutting edge,” and slice-based measures don’t yet tell us whether the same oxytocin dips occur during live social exchanges; whole-animal readouts and region-specific manipulations will be needed to seal the story (The Transmitter).
The new picture is consistent with, rather than contradictory to, the 2023 Neuron surprise. That study showed pair bonding and parental care can survive without oxytocin receptors; it did not rule out oxytocin’s importance for the nuance of selective social life—choosing a partner over a stranger, or a friend over a newcomer—where the latest results now place oxytocin’s thumb on the scale (Neuron, 2023). Other work reinforces that selectivity lives in the brain’s reward circuits. In 2024, researchers reported that in bonded voles, dopamine release in the nucleus accumbens reflects the “selective nature” of pair bonds, underscoring the idea that the reward system encodes “who” matters, not just “that” someone matters (PubMed, 2024). Reviews of vole bonding likewise emphasize oxytocin–dopamine crosstalk in the nucleus accumbens as a key driver of partner preference (PMC review, 2024).
At the molecular level, oxytocin’s story is also tangled up with its chemical cousin vasopressin: the two differ by only two amino acids and can bind to each other’s receptors. That crosstalk may partly explain why social bonds survive in some oxytocin receptor knockouts—other receptors may pick up slack—but also why behavioral effects vary across brain regions and species. In rhesus macaques, for example, a 2024 study found that nebulized vasopressin likely enters the central nervous system and selectively improved social interactions in low-social individuals without inducing aggression (PNAS, 2024). Meanwhile, oxytocin manipulations in different vole species can pull behavior in opposite directions depending on target regions; The Transmitter notes unpublished work from Beery’s lab in seasonally social meadow voles where amygdala-targeted oxytocin injections made animals less likely to snub strangers—an effect that contrasts with the “fortify your friend” pattern seen in prairie voles’ nucleus accumbens pathways (The Transmitter). That regional push–pull aligns with earlier syntheses on oxytocin in mammalian sociality: the same molecule can support maternal care, pair bonding, or even social vigilance depending on context and circuit (Frontiers review).
The social machinery of friendship may also involve memory circuits. The Transmitter highlights ongoing, unpublished experiments from RIKEN’s Center for Brain Science showing that activating neurons projecting from the ventral hippocampus to the nucleus accumbens can cause a vole to shift allegiances—favoring a housemate it previously disliked over a former friend. As one scientist on that project put it, “Change of friendship is also a type of memory formation,” hinting that loyalty and betrayal live partly in the brain’s storage of social history, not just its warmth circuits (The Transmitter).
For Thailand’s health and education sectors, the vole story offers three practical takeaways. First, stable friendships rely on selectivity—consistent choices that privilege specific peers. That has implications for school climates, university dorms, and youth clubs: programs that encourage repeated, meaningful interactions with the same peers may strengthen bonds more than one-off mass events. In Thai classrooms, established “phi–nong” mentor–mentee structures and peer buddy schemes already reflect this insight; formalizing and evaluating them could further reduce loneliness, absenteeism, and bullying. Second, oxytocin is not a shortcut. Large clinical trials of intranasal oxytocin have not delivered broad social benefits in humans; a 24‑week, placebo-controlled phase 2 trial in autistic children and adolescents found no overall effect on social behavior (NEJM, 2021). Any excitement about “hormone sprays” should yield to interventions that change contexts—routines, spaces, and expectations—where friendships grow. Third, for older Thais, the WHO’s warning about isolation and loneliness supports investing in community-level selectivity: regular groups (e.g., weekly “ชมรมผู้สูงอายุ” activities) that encourage ongoing attendance foster the repeated, reliable contact that brains seem to reward (WHO Thailand).
The path to these conclusions has been anything but linear. For years, pharmacological studies suggested blocking oxytocin receptors in the nucleus accumbens impairs pair bonding, while infusing oxytocin there accelerates pair formation. The 2023 CRISPR knockout paper forced a rethink by showing that, during development, brains may compensate when an entire receptor pathway is missing—possibly by recruiting vasopressin receptors or reshaping downstream circuits. The new friendship-focused data do not deny that compensation; they reveal what gets lost even after the brain adjusts: the robust, selective preference that shields friendships from the allure of strangers (Neuron, 2023; The Transmitter).
That nuance also reframes negative results in human oxytocin trials. If oxytocin’s key job is not to make us “more social” in general but to sharpen selectivity—reinforcing specific bonds and discouraging outside options—then giving the hormone broadly, without targeting particular circuits or social contexts, may not yield the expected prosocial boost. Indeed, some human and animal studies suggest oxytocin can increase in‑group favoritism or out‑group wariness depending on context, aligning with a “social salience” model rather than a straightforward “empathy enhancer.” In other words, oxytocin might fortify the relationships we already value, not create new ones from scratch—valuable in friendship maintenance, trickier as a generic therapy.
Thailand’s cultural fabric—rooted in bunkhun (reciprocal obligation), kreng jai (considerate restraint), and neighborhood-level mutual aid—already supports selectivity in healthy forms: long-term networks of “phuen” (friends) who show up for each other in times of need. That embedded selectivity can be a public health asset. Community health volunteers, village temples, schools, and workplaces are natural nodes where repeated interactions accumulate into trust. Policies that strengthen these rhythms—for example, fixed weekly activity slots at community centers, scheduled small-group homeroom circles in schools, or ongoing peer-support cohorts for diabetic or hypertensive patients—may prove more impactful than short campaigns. The neuroscience now suggests why: the nucleus accumbens and related circuits encode specific social identities as rewarding; regular contact keeps that code strong (PMC review, 2024; PubMed, 2024).
At the same time, the vole data carry a caution. “Reject outsiders” is adaptive in the animal’s world of territory and survival; among humans, it can translate into cliques, exclusion, or stigma—problems Thai schools and workplaces must actively counter. A balanced approach fosters selectivity without sectarianism. In practice, that means designing ecosystems where people keep strong ties but also have structured, positive contact with newcomers: rotating project teams around stable mentor anchors; “open house” days in community groups; and school exchange activities that mix classes while preserving core homerooms. Neuroscience points to how: the ventral hippocampus (memory) and nucleus accumbens (reward) likely interact to update friend lists as reputations change; predictable opportunities for cooperation with “new faces” can shift preferences over time, as hinted by the RIKEN work reported by The Transmitter (The Transmitter).
Technologically, the field is sprinting. Fluorescent nanosensors and fiber photometry promise real-time chemical readouts; CRISPRa/i tools in adult vole brains allow circuit tweaks without developmental compensation; and comparative studies across monogamous species, from California mice to titi monkeys, are testing which friendship and bonding mechanisms generalize to primates and, eventually, humans (PMC review, 2024). Yet one thread remains consistent: dopamine and oxytocin coordinate in the nucleus accumbens to render particular partners and peers especially rewarding, and that selectivity erodes with separation and loss—timelines that eerily mirror human grief and adaptation.
For educators and health planners in Thailand, three applications stand out:
Make friendship maintenance routine. In schools, schedule weekly small-group circles with the same peers for a term; in universities, build cohort-based advising; in workplaces, create steady buddy systems for new hires that last at least three months. Consistency feeds selectivity.
Design for repeated encounters. Community centers, temples, and clinics should prioritize recurring activities over one-off fairs. Stable meeting times (same day, same hour) help people return to, and deepen, specific ties.
Combat harmful exclusion. Pair stable groups with periodic, structured inter-group collaboration. Teach and model inclusion alongside loyalty—Thai classrooms have long used “senior–junior” mentorship; add cross-class service projects to widen circles without dissolving core bonds.
Clinically, the take-home is restraint on pharmacological shortcuts. Intranasal oxytocin has not shown broad social benefits in large, well-controlled human trials, including in autism (NEJM, 2021). Research on vasopressin and other neuromodulators continues, with some early nonhuman primate successes (PNAS, 2024), but translation to Thai clinics should await replicable benefits and clear safety profiles. In the meantime, investing in social infrastructure—in schools, elder clubs, and patient groups—remains the most evidence-aligned path to better mental and physical health.
As for the science, more work is coming. Beery’s group plans to map oxytocin’s opposing effects across brain regions, distinguishing circuits for friend-building from those for outsider rejection. Preprints continue to probe how oxytocin receptor loss affects peer selectivity and neurochemical release dynamics in prairie voles, with hints that sex and development stage matter (eLife reviewed preprint, 2025). And memory–reward interactions, flagged by the RIKEN team, may prove central to understanding how friendships change. Donaldson’s caution about developmental knockouts also points the field toward adult-onset manipulations to isolate oxytocin’s moment-to-moment roles in choice and loyalty (The Transmitter).
Thailand’s society has a head start: cultural norms of care, respect, and reciprocity already support the kind of selective, enduring ties the vole brain appears built to reward. The policy challenge is to protect those strengths while keeping doors open to new members. Neuroscience won’t replace the art of friendship, but it can nudge us toward designs that fit how brains actually choose—and keep—each other.
Actionable steps for Thai readers today:
If you work in education: assign long-term peer groups or “buddy pairs” and meet weekly; rotate tasks across groups every 6–8 weeks to balance loyalty and inclusion.
If you manage a workplace: set up three-month onboarding cohorts that meet biweekly; match each newcomer with a consistent senior buddy.
If you run a community program: standardize weekly activity times; encourage members to pair up and check in between meetings.
If you’re feeling isolated: pick one or two regular group activities (e.g., a weekly exercise class at the district health center or a temple volunteer slot) and attend for at least two months—consistency helps the brain tag new people as “your” people.
For health providers: screen routinely for loneliness among older adults and adolescents; refer to stable, recurring group programs rather than single-session events.
Behind these simple moves sits a sophisticated biology: oxytocin and dopamine weaving selectivity into social reward, hippocampal circuits updating friend lists, and the nucleus accumbens signaling that certain people are worth the effort. The latest prairie vole findings don’t make humans voles. But they do suggest a durable theme: strong friendships are built on repeated choice. Thailand’s communities can make those choices easier.
Sources used in this report include The Transmitter’s coverage of the new friendship study (The Transmitter); the 2023 CRISPR findings that pair bonds can persist without oxytocin receptors (Neuron via PubMed); evidence that accumbens dopamine encodes social selectivity (PubMed, 2024); a review of reward remodeling after bond loss (PMC review, 2024); cautionary human trial data on intranasal oxytocin (NEJM, 2021); vasopressin’s selective prosocial effects in macaques (PNAS, 2024); an overview of oxytocin’s diverse roles across mammalian sociality (Frontiers review); early-stage nanosensor work on social neurochemistry (UC Berkeley QB3); and WHO Thailand’s summary of social connection and mental health needs in the country (WHO Thailand).