Sweden’s younger generations appear to be more socially connected than those before them. Yet more than half a century of data reveals a striking paradox: however active people’s friendships are in adulthood, the generational advantage has so far largely disappeared in old age.
A study by researchers at Karolinska Institutet and Stockholm University examined Swedish survey data collected between 1968 and 2021. Covering people aged from 15 to 97, it provides an unusually long view of how contact with friends changes across both the individual life course and successive generations. Its conclusion is at once encouraging and unsettling. Later-born Swedes generally report seeing their friends more frequently than earlier generations did at comparable ages. But among older adults, these differences become much smaller. Stronger friendship cultures in youth and middle age have not yet translated into equally strong social contact during the final decades of life.
Fifty-three years of Swedish social life
Published in the journal Archives of Gerontology and Geriatrics in 2025, the study — Friends and Trends: Friendship Across Life Phases and Cohorts — was conducted by Erika Augustsson, Roger Keller Celeste, Stefan Fors, Johan Rehnberg, Carin Lennartsson and Neda Agahi. The researchers combined ten waves from two major Swedish research programmes. The Level of Living Survey, known as LNU, interviewed nationally representative samples of the Swedish population in 1968, 1974, 1981, 1991, 2000 and 2010. The Swedish Panel Study of Living Conditions of the Oldest Old, or SWEOLD, extended the picture to people aged 75 and above through surveys conducted in 1992, 2002, 2011 and 2021. Together, the surveys produced 32,136 observations covering birth cohorts from 1905 to 1984. Importantly, this does not mean that the researchers followed every participant continuously for 53 years. Instead, the combined surveys allowed them to compare age groups, historical periods and birth cohorts across an exceptionally long span of Swedish social history.
Friendship follows a three-stage pattern
The results reveal a recognisable life-course trajectory. Frequent contact with friends declines from young adulthood, reaches a relative plateau during middle age and then falls again in later life. The pattern is not entirely surprising: education, work, parenthood, retirement, health and mobility all shape the amount of time and opportunity available for friendship. What makes the Swedish findings particularly significant is the comparison between generations. Later-born cohorts were consistently more likely to report frequent contact with friends. After accounting for factors including education, income, health and marital status, people born between 1975 and 1984 were approximately 23 percentage points more likely to report frequent contact than those born four decades earlier. This complicates the familiar claim that modern societies are necessarily becoming less sociable. Swedish friendship has not simply withered under the pressures of individualism, technology and changing family structures. At many ages, recent generations appear to be more connected to friends than their predecessors. The problem arrives later.
The old-age convergence
When the researchers examined older age groups, the differences between birth cohorts narrowed considerably. The social advantage enjoyed by later-born generations in adulthood was no longer clearly visible among the oldest participants. There are several possible explanations. Retirement can remove a large part of a person’s everyday social infrastructure. Colleagues may not be close friends, but workplaces create routines, encounters and reasons to remain in contact. Once that structure disappears, maintaining a relationship requires a more deliberate effort. Friendship networks also tend to age together. The same period of life may therefore bring illness, reduced mobility, caring responsibilities and bereavement to several members of a group simultaneously. A person may become less able to travel just as their friends face similar limitations. Geography matters as well. Younger and middle-aged adults may sustain friendships through workplaces, cafés, cultural activities and regular travel. In later life, the loss of a driving licence, inaccessible public transport or simply the physical difficulty of leaving home can turn an ordinary meeting into a complicated undertaking. Digital communication may reduce some of these barriers, but the study cannot yet establish whether today’s highly connected cohorts will retain their advantage when they reach their eighties and nineties. The youngest people in the dataset had not reached old age by 2021. The research therefore identifies a historical pattern, not an unavoidable destiny.
Loneliness is also a question of expectations
The findings raise another concern: loneliness depends not only on how much contact people have, but also on how much contact they believe they should have. A generation accustomed to busy social lives, constant messaging and readily available communication may experience the later-life decline more acutely than people whose expectations were formed in a less connected era. If social contact converges towards similarly low levels in old age, future retirees could face a wider gulf between expectation and reality. They may have more friendships on paper, more former colleagues and more digital connections, yet still experience the loss of regular, embodied companionship. This distinction is essential. A message, reaction or group-chat notification can maintain awareness of another person, but it may not replace a shared meal, a walk or an unplanned conversation. Digital connection and meaningful social presence overlap, but they are not necessarily the same.
A Nordic welfare issue
The study has implications beyond private behaviour. Friendship is often treated as a personal matter, while public policy concentrates on pensions, healthcare, housing and formal care. Yet the places and systems that make friendship possible are partly collective. Reliable local transport, accessible meeting places, affordable cafés, libraries, community centres, cultural associations and walkable neighbourhoods all influence whether older people can remain socially active. The Nordic welfare model is well equipped to keep people physically safe, but safety alone does not guarantee belonging. A technically efficient care system may still leave someone with long periods of unstructured solitude. Sweden’s extensive tradition of voluntary associations offers an important resource. Sports clubs, study circles, choirs, allotment societies and local cultural organisations can create recurring encounters without requiring people to initiate every meeting themselves. The challenge is to keep such environments accessible when health, income or mobility begins to change. Age-friendly urban planning should consequently be understood as social planning too. A bench, a nearby bus stop or a communal room in a housing development may appear modest compared with a hospital or care facility, but it can help preserve the routines through which relationships remain alive.
Friendship cannot simply be saved for retirement
Perhaps the study’s most useful lesson is that friendship does not behave like a financial asset accumulated in middle age and withdrawn later. Relationships depend on repeated contact, shared environments and mutual effort. A large network at 50 does not automatically remain available at 80, particularly when the circumstances supporting it have disappeared. That does not mean the decline is inevitable. It means that friendship in later life needs infrastructure and attention before a crisis of isolation begins. Maintaining connections across different age groups, developing social routines outside employment and choosing housing with access to public life may all make networks more resilient. Public institutions can support this resilience, but they cannot manufacture intimacy. The task belongs simultaneously to individuals, communities and policymakers. Sweden’s 53 years of data do not show that friendship inevitably fails with age. They show that every generation studied so far has encountered pressures strong enough to weaken it. The open question is whether the generations now entering later life — healthier, more mobile and more digitally connected than their predecessors — can finally change the pattern.
Suggested Online Sources
Archives of Gerontology and Geriatrics – Friends and Trends: Friendship Across Life Phases and Cohorts
Karolinska Institutet Open Archive – Anchored or Adrift? Social Connection, Health, and Loneliness in Later Life
PubMed – Friendship across life phases and cohorts
Scandinavia Standard – Fifty-three years of Swedish data show that contact with friends drops sharply in later life
