Crew members who spent the most time in close physical proximity to their colleagues during a ten-month Antarctic winter were the ones more likely to report conflict, growing mistrust, and reduced performance.
That result runs against the intuition governing most thinking about isolated groups, which treats loneliness as the problem and companionship as the remedy.
Twelve people wintering at Concordia Station, where temperatures reach minus 80 degrees Celsius, wore sensors that recorded when and for how long they were near one another, and completed psychological questionnaires at four points across the mission. The findings were published in the Proceedings of the National Academy of Sciences.
The Measurements Taken Across Ten Months
The study's design is what distinguishes it from earlier work on isolated teams.
Wearable proximity sensors, developed by co-authors at the ISI Foundation in Turin, recorded social contact automatically at a range of roughly one to one and a half meters. They were worn during several two-week phases rather than continuously across the full mission, producing high-resolution objective data on who spent time near whom and for how long. Additional room sensors recorded where those interactions took place.
That removes the recall problems and self-presentation effects that affect questionnaires about social behavior. Alongside it, crew members completed repeated psychological assessments measuring social contact, cohesion, conflict, loneliness and mistrust at four time points, allowing researchers to track how each changed across the mission rather than capturing a single endpoint.
"The combination of personal self-assessments with objective sensor data provides an unprecedentedly detailed picture of how social networks and psychological strain evolve over time in extreme teams," said Andrea Cantisani of the University of Bern, co-first author with Jan Schmutz of the University of Zurich.
Concordia Station sits above 3,000 meters in the Antarctic interior and is cut off from the outside world for roughly nine months of the year. It is considered one of the best terrestrial analogs for long-duration missions to the Moon or Mars, because during that period no one can arrive or leave.
The researchers reported three patterns: emerging interpersonal strain, subgroup formation, and increased suspiciousness, all occurring despite close physical proximity. Cohesion and perceived performance declined while loneliness, mistrust, and conflict rose. The word despite is doing work in that description, since proximity would ordinarily be expected to reduce all three.
The Subgroup Pattern
Beyond the proximity finding, the crew divided over time along cultural and language lines.
Subgroup formation in a team of twelve is consequential in a way it would not be in a larger organization. When a group splits, the smaller units are very small, and someone outside the dominant grouping may have few alternatives for social contact.
The pattern has practical implications for how crews are assembled. An international crew brings language diversity that may function as a fault line under sustained confinement, which is a different consideration from the individual psychological screening that dominates selection.
It also complicates the proximity finding. If crews divide, then measured contact frequency may partly reflect who was included in which subgroup rather than a straightforward relationship between exposure and strain, and the two findings may be describing one process from two angles.
The researchers also demonstrated that wearable proximity sensors function reliably in extreme conditions, which is a methodological result with implications beyond this study. It allows tracking of changes in a team's daily routine without significantly interfering with activities.
The Causal Question the Design Cannot Answer
The University of Zurich, which co-led the work, states the limitation directly: because the analyses are correlational, no conclusions can be drawn about causality.
The alternative explanation the researchers themselves raise is worth stating. It is possible that lonely individuals sought more contact but that those interactions were not sufficiently rewarding. In that reading, distress drove contact-seeking rather than contact producing distress.
A third possibility is that both reflect something else entirely. A crew member in conflict with others may find themselves drawn into more forced interactions, or someone struggling with the mission generally may both seek company and report more problems, without either causing the other.
"In small teams under extreme conditions, more contact doesn't automatically equate to social support, but can actually increase tensions," said Schmutz, a professor in the Department of Psychology at Zurich, who led the study with Cantisani.
Twelve participants is also a very small sample, appropriate to the setting and limiting for statistical inference. A single crew in a single winter cannot establish that the pattern generalizes. MedicalDaily has covered the aftermath of high-stress service environments in other populations.
The Limits of Extending This to Ordinary Life
The setting is the point of the study, and it is also the reason the findings should not be generalized casually to ordinary living arrangements.
Concordia during winter is a closed system. Nobody leaves, nobody arrives, the same twelve people occupy the same rooms for ten months, and there is no possibility of taking a break from a difficult colleague. Almost no ordinary human situation replicates that.
The findings do not indicate that social contact is bad for wellbeing generally. The evidence on loneliness and health in ordinary populations is substantial and points the other way, associating social isolation with worse physical and mental health outcomes.
What the study suggests is narrower: that in genuinely confined settings without exit, contact volume alone is not a useful proxy for social support, and that quality and voluntariness may matter more than quantity. The research team also frames the sensors as a tool that could detect such developments at an early stage, which is arguably the more actionable contribution for mission planners.
For readers, the relevant applications are limited to broadly comparable situations, including long military deployments, extended caregiving within a shared household, or confined work environments such as offshore or remote postings. Even there, the finding is a hypothesis worth considering rather than guidance to act on.
Future research will examine which social interactions reduce stress and which create additional strain, which is the question this study raises rather than answers. Anyone finding a confined or high-stress living situation is affecting their mental health should speak with a clinician. MedicalDaily has covered expanded access to community mental health services for people seeking support.
Key Questions Answered
What did the study measure? Wearable sensors recorded when and how long 12 crew members were within about one to one and a half meters of each other during several two-week phases across a ten-month Antarctic winter, alongside questionnaires at four time points measuring cohesion, conflict, loneliness and mistrust.
What was the main finding? Crew members with more frequent contact were more likely to report conflict, growing mistrust and reduced performance. Researchers also observed subgroup formation along cultural and language lines and increased suspiciousness, with cohesion and perceived performance declining.
Does this prove contact caused the strain? No. The analyses are correlational. The researchers note it is possible that lonely individuals sought more contact and that those interactions were not sufficiently rewarding, which would reverse the direction.
Why was Antarctica used? Concordia Station sits above 3,000 meters in the Antarctic interior, with winter temperatures reaching minus 80 degrees Celsius and roughly nine months cut off from the outside world. It is considered one of the best terrestrial analogs for Moon or Mars missions.
How large was the study? Twelve participants across one winter. That is appropriate to the setting and small for statistical inference, and a single crew cannot establish that the pattern generalizes.
Does this mean social contact is bad for wellbeing? No. Evidence on loneliness in ordinary populations associates social isolation with worse physical and mental health. The finding is specific to confined settings where no one can leave.
What did it add methodologically? It demonstrated that wearable proximity sensors work reliably in extreme conditions, which researchers suggest could allow early detection of deteriorating team dynamics without interfering with crew activities.
