Consider someone who attends two book clubs, hosts dinner parties monthly, maintains a group chat with college friends, and schedules weekly coffee dates with neighbors. By any conventional measure of social engagement, this person is doing everything right. Yet they report persistent loneliness — a nagging sense that something is missing despite a calendar full of social commitments.
The standard advice for such a person has long been some variant of "get out more." Join another group. Say yes to more invitations. Volunteer. The assumption embedded in most loneliness interventions is that social connection operates on a linear scale: more contact equals less loneliness, always.
But what if that assumption is wrong? What if each person has a unique threshold for social connection — a point below which loneliness increases predictably, but above which additional socializing provides no further benefit? A 2025 study published in Scientific Reports suggests precisely this, and its implications challenge some of the most common approaches to addressing the loneliness epidemic.
The Set Point Discovery
In 2025, Caterina Mauri published a study in Scientific Reports (Nature) that introduced a concept borrowed from physiology into loneliness research: the homeostatic set point. The study, titled "Asymmetric loneliness response to social connectedness around individual set points," analyzed data from 25,646 adults across 27 European Union countries, drawn from the European Union Loneliness Survey conducted in 2022.
The central finding is deceptively simple but carries profound implications. Loneliness decreases linearly as social connectedness increases — but only up to an individual's personal "set point." Once that threshold is reached, additional social contact provides no further reduction in loneliness. The response curve is asymmetric: below the set point, every increment of social connection helps. Above it, more connection is essentially inert.
The homeostatic control model underlying this finding draws from well-established physiology. Just as body temperature, blood glucose, and weight appear to have regulatory set points — levels the body defends through feedback mechanisms — social connection may operate similarly. The organism has a target level. Falling below that target triggers a corrective signal (loneliness, the social equivalent of hunger). But exceeding the target does not produce an opposite reward in proportion.
This is not merely a statistical curiosity. It suggests that loneliness is not the simple inverse of social contact. Two people with identical social calendars may experience radically different levels of loneliness if their set points differ. The person with a low set point for social connection may feel perfectly content with two close friends and occasional gatherings. The person with a higher set point may need more frequent and more intimate contact to reach their threshold.
What "Asymmetric Response" Means in Practice
The term "asymmetric response" is the key to understanding Mauri's finding. In a symmetric model, loneliness would decrease steadily as you add social connections — and perhaps even reverse into some form of "social satisfaction surplus" with very high levels of contact. This is the implicit model behind most intervention design.
In the asymmetric model Mauri's data support, the relationship looks different. Below your set point, the system behaves predictably: add connection, reduce loneliness. Above your set point, the system flatlines. You cannot "stockpile" social connection against future loneliness, any more than you can eat extra meals today to prevent hunger next week.
This asymmetry also means that the variance in loneliness among socially active people is not random noise. It reflects genuine individual differences in set points. Some people are lonely not because they lack social contact, but because their particular threshold requires a quality or type of connection they are not receiving — even amid high quantity.
Why "Just Get Out More" Fails for Many
If set points exist, they help explain a pattern that has puzzled loneliness researchers for years: why increasing social opportunities often fails to reduce loneliness for a substantial proportion of participants.
The most comprehensive evidence on this question comes from Mathias Lasgaard and colleagues, whose 2025 meta-analysis published in American Psychologist — "Are Loneliness Interventions Effective?" — synthesized findings from 280 studies encompassing more than 30,000 participants. The results are striking. Psychological interventions, particularly cognitive behavioral therapy targeting maladaptive social cognition, were the most effective approaches. Interventions that simply created opportunities for social contact — the "just get out more" approach operationalized into program form — were consistently less effective.
This finding has typically been interpreted through John Cacioppo's framework of hypervigilance to social threat: lonely people develop cognitive biases that cause them to perceive social interactions more negatively, making additional contact counterproductive unless those biases are addressed first.
But the set point model offers a complementary explanation. If a person is already above their set point for social contact quantity — yet still lonely — then providing more contact cannot help by definition. The problem is not insufficient input but rather a mismatch between what they are receiving and what their particular threshold requires. Cognitive approaches may work better precisely because they help people recalibrate their perception of their social needs, or help them identify what type of connection they are missing rather than simply seeking more of the same.
As we have previously covered, the Lasgaard meta-analysis found that the most effective interventions addressed the cognitive and perceptual dimensions of loneliness rather than the structural ones. The set point framework helps explain why: for many lonely people, the structural dimension — raw amount of social contact — is not actually deficient. What is deficient is the match between their contact and their individual needs.
Thresholds Appear Elsewhere Too
The set point concept gains additional credibility from research showing threshold effects in other domains of social life. If social connection operated on a simple linear continuum, we would expect dose-response relationships to be consistent: each additional unit of social input producing a proportional unit of output. Instead, researchers are finding inflection points — levels below which outcomes deteriorate and above which additional input provides no further benefit.
One striking example comes from research on remote work and loneliness. In a 2026 study published in the Journal of Affective Disorders, Tun He, Lijing Wei, Melody S. Goodman, and colleagues at NYU's School of Global Public Health analyzed a nationally representative sample of employed US adults drawn from the Census Bureau's Household Pulse Survey in 2024. Their question was straightforward: does working remotely increase loneliness?
The answer turned out to be non-linear. Working remotely one to two days per week showed no significant association with loneliness. But at three or more days per week, the relationship emerged: the adjusted odds ratio was 1.16 for those working remotely three to four days, and 1.09 for those working remotely five or more days per week. The relationship was not a smooth gradient but a step function — a threshold effect.
This pattern mirrors Mauri's set point finding in a different register. There appears to be a minimum level of in-person workplace contact — roughly three days per week — below which loneliness risk increases. Above that threshold, additional in-person time provides no further protective benefit. The social world, it seems, is full of non-linear relationships that we have been modeling as linear.
The Implication for Policy
These threshold effects have direct policy relevance. If social contact operates on a linear model, the optimal intervention is always "more." More community programs. More social prescribing referrals. More mandatory office days. But if thresholds exist, the optimal intervention is different: identify where individuals sit relative to their threshold, and target resources at those who are below it. People already above their threshold gain nothing from additional programming — and may even experience the added social demands as burdensome.
The remote work finding is particularly instructive for organizations debating return-to-office policies. The data suggest that requiring three days per week of in-person presence may be sufficient to prevent loneliness-related harms — while additional mandatory days provide no measurable benefit to social wellbeing and may impose costs in commuting time, flexibility, and autonomy.
The Implications for a Lonelier World
The set point model becomes especially significant when placed against the backdrop of rising global isolation. In 2025, Thomas E. Fuller-Rowell, Samia Sultana, and Ichiro Kawachi published a study in JAMA Network Open analyzing social isolation trends across 159 countries using repeated cross-sections from 2009 to 2024. Their finding was stark: global social isolation rose 13.4 percentage points over this period, from 19.2% to 21.8%, with the entire increase concentrated after 2019. Low-income groups were disproportionately affected.
If set points exist, this global trend takes on a specific character. The crisis may not be primarily about people who need more connection failing to seek it — the narrative that underlies most individual-level interventions. Instead, it may be about structural changes — remote work expansion, urban sprawl, economic barriers to participation, the decline of third places — pushing millions of people below their individual thresholds simultaneously.
The Gen Z paradox illustrates this most acutely. According to the Cigna/Evernorth Research Institute's 2025 "Loneliness in America" report, which surveyed more than 7,500 US adults, 57% of Americans now report loneliness (up from 52% in 2020), with Gen Z at 67% — the highest of any generation. This is the most digitally "connected" generation in history. They have more communication tools, more platforms, more ways to reach other people than any previous cohort. Yet they are the loneliest.
The set point model suggests an explanation. Digital communication may not count toward set point fulfillment in the same way that in-person contact does. A person may have hundreds of online interactions daily yet remain below their threshold for the type of connection that actually registers with their social homeostatic system. The quantity of contact is high; the quality relative to their set point requirements is low.
This aligns with what we have documented in our overview of global loneliness data: the rise in isolation is not occurring because people have stopped communicating. It is occurring because the nature of communication has shifted in ways that may not satisfy the underlying biological and psychological mechanisms that regulate social homeostasis.
What We Do Not Know Yet
The set point concept is compelling, but several important limitations and open questions must be acknowledged.
First, the anchor study is cross-sectional. Mauri's analysis of 25,646 EU adults captures a single point in time. It can demonstrate that the asymmetric pattern exists in population data, but it cannot prove that set points are stable features of individuals over time. It is possible that what appears as a set point in cross-sectional data reflects transient states rather than enduring traits. Longitudinal replication — tracking the same individuals over years to see whether their thresholds remain consistent — is essential before the concept can be considered established.
Second, set points may shift. Even in physiology, set points are not always fixed. The body weight set point can be altered by sustained dietary changes, medication, or metabolic shifts. Similarly, a social connection set point might change following major life transitions — retirement, bereavement, relocation, parenthood — or through sustained psychological intervention. If set points are malleable, the practical implications change: rather than simply helping people reach their threshold, interventions might aim to adjust the threshold itself.
Third, the sample is Western and European. The European Union Loneliness Survey covers 27 EU countries, which provides meaningful diversity but excludes most of the global population. Social norms around connection, solitude, and community vary enormously across cultures. A collectivist society in East Asia may produce different set point distributions than an individualist Northern European one. The concept requires cross-cultural validation before it can be considered universal.
Fourth, the biological mechanism is hypothetical. The homeostatic model borrows from physiology — specifically from set point theories of body weight regulation — but the neural and endocrine mechanisms that would implement a social set point have not been identified. Unlike temperature regulation, where the hypothalamus provides a clear substrate, we do not yet know where or how a social homeostatic system would operate. The analogy to weight regulation is suggestive but remains just that — an analogy, with limits.
Fifth, the practical measurement problem. If each person has a unique set point, how would a clinician or intervention designer identify it? Self-report measures of loneliness capture whether someone is below their threshold (they feel lonely) but do not directly measure where the threshold sits. Developing reliable methods to assess individual set points would be necessary for the concept to translate into personalized intervention design.
Toward a More Nuanced Understanding
The loneliness set point concept does not invalidate existing approaches to addressing social isolation. Social prescribing, community programs, group interventions, and digital platforms that facilitate connection all remain valuable — particularly for the many people who genuinely are below their threshold due to structural barriers. The Fuller-Rowell data showing a 13.4% rise in global isolation suggests that the pool of people below their threshold has grown substantially.
What the set point model does is add nuance to a conversation that has often been too simplistic. The dominant narrative — loneliness is rising, therefore people need more social contact, therefore any intervention that increases contact is good — misses a critical variable: individual variation in what constitutes "enough."
The practical implication is a shift from optimizing for more social contact to optimizing for the right amount and type. For someone below their threshold due to geographic isolation, economic barriers, or life transitions, the answer may indeed be more contact — and structural interventions that remove barriers to connection are appropriate. But for someone who is lonely despite an active social life, the answer is more likely qualitative than quantitative: deeper rather than broader, more intimate rather than more frequent, better matched to their particular needs rather than generically "more."
This distinction aligns with the design philosophy behind platforms like YaraCircle, which emphasizes quality shared experiences over quantity of connections. The research suggests this instinct is well-founded: what matters for friendship formation is not the raw number of people you encounter but the depth, reciprocity, and sustained nature of the interactions you have.
The set point concept also reframes loneliness as something other than a personal failure. If your threshold is higher than average — if you require more intimate connection, more frequent contact, or more specific types of social engagement to feel satisfied — that is not a deficit in social skills or a sign of ingratitude for the connections you have. It is a feature of your individual social homeostasis, as morally neutral as a higher or lower body temperature.
Understanding this may itself be therapeutic. The person who feels lonely despite a full social calendar is not broken or unreasonable. They may simply have a set point that their current social configuration, however impressive it looks from the outside, does not meet. The solution is not to dismiss their experience or to prescribe even more of what is already not working — but to help them identify what specific dimension of connection their system is seeking.
See the Research in Action
YaraCircle is built on the friendship science we cover here.