Friendship in older adulthood is a voluntary, personally meaningful relationship that can provide companionship, trust, shared history, practical assistance, mutual recognition, and opportunities for new experiences. Social networks are the broader configurations of people and relationships surrounding an individual, including friends, relatives, neighbors, former colleagues, community members, and other significant contacts. Across later life, these networks may change in size, composition, frequency of interaction, accessibility, and emotional quality. Their development combines continuity in valued ties with selective investment, loss, renewal, and the formation of new relationships. Population research finds an average reduction in peripheral social connections across adulthood while many emotionally close relationships remain comparatively stable. This average pattern describes neither an inevitable loss of friendship nor a single ideal network size. It reflects interacting motivations, life events, health, opportunity, historical circumstances, and social inequalities. Wrzus et al. (2013); English and Carstensen (2014); Antonucci et al. (2014).
The central scientific question is how friendships are sustained, reorganized, and created as people move through later adulthood. An accurate answer attends to relationship quality alongside network structure; distinguishes the choice to prioritize trusted companions from unwanted exclusion; and follows individuals over time instead of assuming that differences between age groups are produced by age alone. It also recognizes that friendship opportunities depend on transportation, housing, economic security, accessible public spaces, communication technologies, and the social welcome people receive. Such conditions can expand or restrict the capacity to maintain connection at any age. Ayalon and Levkovich (2019); Baeriswyl and Oris (2023).
What Counts as Friendship and a Social Network in Later Life
Friendship is ordinarily characterized by mutual choice and recognition, even where a friendship has lasted for decades and become embedded in family history. Its activities may range from regular visits, shared meals, religious or cultural participation, and neighborhood encounters to brief telephone conversations and sustained long-distance communication. Friends need not be the same age, live nearby, or share every interest. A person may describe a cousin, former coworker, caregiver-turned-peer, or neighbor as a friend when the relationship is experienced as reciprocal and voluntarily maintained. Research questionnaires often force respondents into categories that are more rigid than lived experience, making explicit definitions essential when comparing studies. Antonucci et al. (2014); Siette et al. (2021).
A social network is a pattern of ties rather than a single subjective feeling. Researchers characterize it by size, composition, proximity, frequency of contact, mutuality, stability, density, diversity, and the resources available through its members. Strong ties often involve intimacy and reliable support; weaker ties can connect a person to information, services, unfamiliar perspectives, and social settings beyond a close circle. A large network can contain conflict or superficial connections, while a smaller network can contain exceptionally trusted companions. Distinct network configurations may suit different lives, and their usefulness depends on the functions they perform and the preferences of the person concerned. Ayalon and Levkovich (2019); Siette et al. (2021).
Friendship should also be distinguished from social support. Support describes emotional reassurance, useful information, practical help, or resources exchanged through relationships. Receiving help may strengthen a friendship, but sustained one-sided assistance can also produce strain, perceived obligation, or concerns about independence. Reciprocity need not mean exchanging identical favors at identical times; it can involve listening, humor, advice, shared memory, and recognition of one another's competence. Social participation is another related construct: attendance at an event is an opportunity for contact, and participation becomes friendship only when interactions develop into mutually recognized personal ties. Antonucci et al. (2014); WHO (2025).
Friendship, loneliness, isolation, and chosen solitude
Loneliness refers to the distressing perception that one's social relationships fall short of what is wanted; social isolation generally refers to objectively limited social contact or network integration. A person with many acquaintances can experience loneliness, and another person with very few contacts can enjoy an intentionally quiet life. Voluntary solitude, privacy, and a preference for intimate rather than frequent interaction deserve respect. Difficulty arises when the desired form of connection is unavailable, when remaining contacts are unsafe or burdensome, or when loss and practical barriers prevent meaningful choice. The empirical literature therefore measures network structure and subjective satisfaction separately. National Academies of Sciences, Engineering, and Medicine (2020); WHO (2025).
How Friendship Networks Change Across Adulthood
The most informative overview comes from a meta-analysis by Wrzus and colleagues that integrated 277 studies involving 177,635 participants across adolescence and old age. Their synthesis found that overall networks expanded into young adulthood and then became smaller on average; personal and friendship networks declined across adulthood, whereas family-network size was comparatively stable. Transitions and changes in social roles also shaped network patterns. Importantly, average reductions do not tell us whether individual ties were willingly de-emphasized, became inaccessible, ended through conflict or death, or were replaced by different relationships. Wrzus et al. (2013).
English and Carstensen examined social networks longitudinally across a broad adult age range. Their results indicated that peripheral ties accounted for much of the decrease in network size, with close partners showing greater continuity. Network members also differed in the emotions they elicited, and the emotional tone of social encounters was associated with daily experience. This evidence supports the possibility that social selectivity contributes to emotional adaptation. It does not establish that everyone consciously chooses to reduce their network, that involuntary losses are beneficial, or that a smaller network will improve well-being for every person. English and Carstensen (2014).
A person's relational history matters as much as current network size. An older adult who has maintained two close friendships for forty years may experience continuity despite rare in-person meetings. Someone who has recently moved may maintain former relationships while gradually gaining new neighbors and companions. Another person may face a long accumulation of bereavements or changes in hearing, mobility, income, and residence that reduce contact even when the desire for friendship remains high. In each situation the relevant developmental question concerns changing access, meaning, continuity, and choice rather than whether the person conforms to a preferred number of friends. Antonucci et al. (2014); Baeriswyl and Oris (2023).
Continuity in close ties and the significance of shared history
Long-standing friendships preserve a common record of family transitions, work, migration, community participation, cultural events, and everyday routines. Shared memories can make conversation efficient and emotionally meaningful because each friend understands the other's biography without extensive explanation. This history can promote trust and continuity of identity through retirement, relocation, or changing physical capacities. Yet duration alone does not guarantee quality. Familiar relationships may involve criticism, exclusion, unequal obligations, or unresolved disputes, and older adults retain the ability to set boundaries or withdraw from harmful interactions. A balanced account considers emotional closeness, support, autonomy, and the actual experience of each tie. Antonucci et al. (2014).
Continuity can survive altered forms of contact. A friendship once organized around weekly exercise may evolve into calls, home visits, accessible outings, or shared digital activities. Friends can maintain a sense of mutual commitment when illness changes who travels, who initiates contact, or who can take part in a familiar activity. Maintaining the relationship often requires flexibility from both people and support from accessible transport, communication devices, affordable services, and welcoming settings. The continuity of emotional commitment and the continuity of identical behavior are separate aspects of relational development. WHO (2025); National Academies of Sciences, Engineering, and Medicine (2020).
New friendships remain possible
Older adulthood also creates circumstances in which new friendships can form: retirement frees or redistributes time for some people; community classes bring together people with shared interests; moving to a new neighborhood can increase contact; and volunteering, cultural organizations, accessible sports, peer groups, and mutual-aid activities can introduce potential companions. Repeated encounters offer opportunities to learn about one another, establish trust, and discover reciprocity. The likelihood of a friendship emerging depends partly on whether encounters continue beyond an initial organized event. Availability, compatible interests, social norms, accessibility, and willingness to initiate contact all matter. The evidence supports possibility and variation rather than an age at which friendship formation ends. Baeriswyl and Oris (2023); WHO (2025).
Some friendships are formed through reactivation. A former schoolmate or work colleague may become important again following retirement, relocation, or reconnection through a mutual friend. Others begin online and later develop in person, or remain satisfying across distance. Friendship formation should not be reduced to a single social prescription. People differ in comfort with new acquaintances, sensory needs, cultural expectations, safety concerns, interests, and desired intimacy. A realistic opportunity makes repeated, low-pressure, reciprocal contact possible and leaves the person's preferences in control. National Academies of Sciences, Engineering, and Medicine (2020); WHO (2025).
Theories Explaining Stability, Selectivity, and Change
Socioemotional selectivity theory
Socioemotional selectivity theory proposes that perceptions of future time influence social goals. When the future feels relatively open-ended, acquiring new information and meeting unfamiliar people may have particular value; when time is perceived as more limited, emotionally meaningful encounters can receive greater priority. Chronological age correlates with time perspective in many settings but does not determine it. Serious illness, war, migration, an unexpectedly expanded opportunity, or a major life event may also transform perceived time horizons. The theory offers a motivational explanation for some network selectivity, distinct from the practical loss of accessible social opportunities. Carstensen, Isaacowitz, and Charles (1999); English and Carstensen (2014).
A person may invest more in a few trusted companions because these relationships produce familiarity, mutual support, and rewarding interaction. At the same time, someone may wish to maintain a broad, varied circle throughout late adulthood, particularly when intellectual curiosity, neighborhood engagement, or collective work remains central. The theory predicts changing priorities under particular motivational conditions and should be evaluated against other explanations such as retirement, health constraints, transport access, partner loss, and demographic change. Research comparing cultures and historical periods must also examine the different social meanings of friendship and obligation. Carstensen et al. (1999); Fung, Carstensen, and Lang (2001).
The social convoy model
The convoy model conceptualizes a person's social world as a changing set of relationships that accompanies the individual through life. People often arrange important others into circles that differ in felt closeness and commitment. Some contacts stay in central positions for long periods, while others move toward or away from the center as life circumstances, personal needs, and relational quality change. Convoys are shaped by personal characteristics, situational demands, and broader social structures. This framework connects the stability of enduring friendships with network change due to employment, marriage, parenthood, illness, migration, community life, and institutional conditions. Antonucci, Ajrouch, and Birditt (2014).
A convoy perspective is particularly useful for recognizing support beyond conventional kinship structures. Friends, neighbors, chosen family, community companions, and extended relatives may perform different functions. A network can be reorganized when one member becomes unavailable without losing every source of support. Conversely, a seemingly large convoy may conceal dependence on a single person for transport, interpretation, emergency help, or emotional confidence. Researchers and practitioners therefore need to know which kinds of support exist, whether they are reciprocal, and how reliable they remain as circumstances change. Antonucci et al. (2014); Ayalon and Levkovich (2019).
Lifespan and life-course perspectives
A lifespan approach examines how gains, losses, and adaptations unfold across interacting areas of life. Friendships can gain depth while an individual's opportunities for travel decrease. Practical support may become more salient when a health condition changes mobility, while intellectual and emotional contributions remain active. Life-course research adds the historical timing of education, employment, migration, family formation, discrimination, retirement, and public crises. These conditions influence whom people encounter, which friendships are institutionally supported, and which ties are likely to survive transitions. The resulting network reflects accumulated biography and unequal opportunity as well as present-day preferences. Wrzus et al. (2013); Baeriswyl and Oris (2023).
The theories are complementary at some points and generate distinct predictions at others. Socioemotional selectivity emphasizes changes in motivation and perceived time; the convoy model emphasizes relationship position, functions, and structural contexts; life-course reasoning emphasizes historically situated transitions and linked lives. Comparing their explanatory value requires identifying the pathway under investigation. A decrease in contact after relocation may primarily reflect distance, whereas a preference for fewer emotionally satisfying encounters in an unchanged environment may reveal selection. A single descriptive statistic about the size of an older person's network cannot discriminate these mechanisms. English and Carstensen (2014); Antonucci et al. (2014).
Why Friendships Matter for Psychological Well-Being
Friendships can offer companionship, emotional expression, humor, recognition, continuity of identity, and a sense of belonging. They may provide opportunities for novel activities and different perspectives from those available within family relationships. Some friends share practical information about services, hobbies, or accessible travel; others help a person process difficult transitions. Relationships can also involve disagreement, ambivalence, competition, pressure, or emotional labor. Research must therefore examine relationship quality and the individual's preferences rather than assume every interaction is helpful. Huxhold, Miche, and Schüz (2014); Antonucci et al. (2014).
Huxhold and colleagues examined informal social activities and well-being in middle-aged and older adults, including activities with friends. Their findings support the relevance of friendship and informal interactions to well-being, while observational designs require caution about causal direction. People who feel healthier or more energetic may find it easier to participate in enjoyable social activities; engaging in rewarding relationships may also support well-being; and both may be affected by resources, personality, and community conditions. The value of friendship is evident in lived experience, but population associations cannot specify how much extra interaction a particular individual needs. Huxhold et al. (2014).
Quality, autonomy, and reciprocity
The benefits attributed to friendship often depend on trust, being understood, respect for autonomy, and mutual responsiveness. A friend who listens without overriding decisions may provide a different form of support from a relative who performs essential tasks but exerts unwanted control. Older people can remain active providers of advice, comfort, humor, experience, and material support. Recognizing this reciprocity prevents friendships from being represented exclusively through the needs of someone receiving care. It also makes it possible to distinguish help that sustains independence from practices that undermine it. Antonucci et al. (2014); National Academies of Sciences, Engineering, and Medicine (2020).
Conflict and ambivalence deserve equal attention. Long friendship histories can make expectations difficult to renegotiate when one person's health or financial circumstances change. A friend may simultaneously feel affection, responsibility, frustration, and fatigue. Setting boundaries, arranging practical supports outside the friendship, and allowing relationships to change form can help preserve choice. These are reasonable relational practices, although their effectiveness as formal interventions requires study. It is unscientific to assign an intrinsic health benefit to the mere presence of a relationship without examining whether that tie is safe and valued. Antonucci et al. (2014).
Health associations and the limits of causal claims
Research has linked broader social integration with health outcomes. A well-known meta-analysis covering 148 studies and 308,849 participants found an association between stronger social relationships and better survival; it addressed multiple kinds of social relationships, not exclusively friendships among older people. The pooled odds ratio reported was 1.50, with observational evidence predominating. This result does not mean that obtaining a new friend causes a fixed individual increase in survival or that family, friendship, and community effects are interchangeable. Health and disability can influence social contact, and social connections may also affect behavior, stress, and access to assistance. Holt-Lunstad, Smith, and Layton (2010).
A review of social relationships and cognitive function in healthy older adults found a mixed body of evidence involving networks, social activity, support, and cognitive outcomes. Such findings justify research on how social environments relate to cognitive health without identifying friendship as an established treatment or proving that expanding a network prevents dementia. The clinical interpretation of memory symptoms, neurological conditions, and dementia requires its own diagnostic evidence. Population-level friendships research contributes one part of the broader picture of well-being. Kelly et al. (2017); National Academies of Sciences, Engineering, and Medicine (2020).
Life Events That Reorganize Friendships
Retirement and the disappearance of routine encounters
Paid work can supply daily contact, repeated conversation, weak ties, and opportunities for friendship. Retirement may remove some of these encounters, especially where the workplace has been the person's principal social setting. Other adults gain time for relationships previously constrained by employment, caregiving, or commuting. Whether retirement expands or contracts friendship depends on the voluntariness and timing of retirement, financial security, health, family responsibilities, transportation, and access to suitable settings. A workplace acquaintance may remain a close friend, fade when shared routines end, or become more important as additional time becomes available. Wrzus et al. (2013); Baeriswyl and Oris (2023).
Retirement also changes the practical organization of time. Contact that formerly occurred by chance may require planning. Friends may have different schedules, responsibilities, and abilities, leading to asymmetry even where affection persists. Activities that produce recurring informal encounters can help maintain a social rhythm, provided that they are accessible, affordable, and personally meaningful. Programs should create conditions for freely chosen friendship rather than prescribe intensive social participation as a moral duty. WHO (2025).
Bereavement, widowhood, and friendship networks
The death of a partner or close friend can change a network's composition, routines, and emotional meaning. The effects on surviving relationships vary: mutual friends may provide continuity, reminders of the deceased may bring both comfort and distress, and friendships may strengthen or weaken after practical routines change. Recently published European panel research followed adults around widowhood and examined changes in close network composition, satisfaction, emotional closeness, and community participation, with attention to gender differences. Its within-person design adds information beyond simple comparisons between widowed and married groups, while conclusions remain contextual to the cohorts and settings studied. Life after loss: gendered changes in social networks and community participation around widowhood (2026).
Grief is a distinct psychological process and should be addressed according to the person's experience rather than inferred from network size. Some people seek companionship immediately after a death; others need intervals of privacy, practical assistance, or different forms of contact. Friends can offer presence without attempting to impose a timetable for adjustment. A person who appears socially active may still experience profound grief. The scientific ownership of clinical bereavement and prolonged grief belongs to grief research, while the present discussion concerns how social ties evolve after loss. National Academies of Sciences, Engineering, and Medicine (2020).
Relocation, housing, and neighborhood change
Moving can interrupt casual neighborhood encounters while creating opportunities to meet people in another place. Voluntary relocation, displacement, downsizing, admission to residential care, and migration occur under very different conditions. Familiar communities may sustain friendship through recognizable streets, services, and shared routines. A new setting may offer contact while also requiring the difficult work of learning unwritten social expectations and establishing trust. For residents with limited mobility or transportation, proximity and the design of gathering spaces can be especially consequential. WHO (2025); National Academies of Sciences, Engineering, and Medicine (2020).
Residential care should not be equated with the absence of friendship. Residents may form meaningful peer relationships, retain external companions, or experience barriers related to privacy, staff practices, scheduling, cognitive or sensory accessibility, and visitor policies. A residence that offers activities may still leave little room for voluntary, one-to-one relationships. Assessing friendship requires asking about relationships residents recognize as their own, their opportunities to meet privately or socialize comfortably, and the continuity of ties beyond the institution. WHO (2025).
Health, disability, sensory access, and caregiving
Changes in health can affect the practical forms of friendship without determining its meaning or quality. Hearing difficulties may make noisy restaurants tiring; pain or fatigue may limit spontaneous outings; visual, mobility, or cognitive changes may require accessible formats, adapted meeting places, or additional planning. In all these situations environmental accommodation can affect whether a social tie remains active. Family members and service providers should ask what contact the person wants and what specific barriers can be reduced instead of equating disability with disinterest. National Academies of Sciences, Engineering, and Medicine (2020); WHO (2025).
Some older adults also devote substantial time to caring for a spouse, relative, or friend. Caregiving responsibilities can reduce opportunities to maintain other relationships, especially where respite and accessible transportation are limited. Friends themselves may become part of practical care networks, but emotional bonds can suffer when one relationship becomes the sole source of essential assistance. Addressing external support needs may help protect space for companionship, mutuality, and ordinary life beyond care tasks. Antonucci et al. (2014).
Inequalities, Culture, and Different Forms of Social Belonging
Friendship opportunities are distributed unevenly across economic resources, neighborhoods, education, health, and social institutions. People with financial security can often afford transportation, accessible venues, and the time needed to sustain varied relationships. Others may face precarious housing, discrimination, rural distance, violence, or exclusion from public places. Swiss survey comparisons analyzed by Baeriswyl and Oris found an increased prevalence of close friendships across historical periods while social inequalities in friendship opportunities persisted. This is an important reminder that historical improvements in the average prevalence of friendship do not guarantee equitable access for everyone. Baeriswyl and Oris (2023).
Friendship takes different forms across cultural and familial contexts. In some settings, daily life centers on interdependent families and neighborhood networks; in others, individually chosen social circles have a particularly salient public role. Cultural practices shape who is called a friend, how obligations are expressed, and whether revealing loneliness carries stigma. A review of older-adult social network typologies across countries found recurring distinctions in family-focused, diverse, and restricted networks, while calling attention to the need for culturally grounded interpretation. A network measure devised in one setting may miss the roles of communal, religious, migrant, or multigenerational relationships in another. Social network typologies of older people (2021); WHO (2025).
Gender, partnership status, and uneven social expectations
Gendered labor histories, differences in paid employment, caregiving expectations, widowhood, and unequal financial resources can shape patterns of friendship. Researchers sometimes observe group differences in the composition or maintenance of networks, but these should not be presented as fixed traits of all women or men. A comparison between two gender categories often combines multiple structural and historical pathways, while nonbinary and transgender older adults may be underrepresented in conventional surveys. More reliable explanations trace actual social opportunities, relationships, and life histories. Erosheva et al. (2016); Life after loss (2026).
Living alone likewise has several meanings. It may follow widowhood or separation, reflect financial and housing constraints, or express an established preference for privacy and control. Partnership status reveals only one aspect of someone's relationships. Friends and chosen kin may form the core of an individual's support network whether or not the person has a spouse or adult children. A sound assessment asks which specific ties exist and whether the person considers them reliable and satisfactory. National Academies of Sciences, Engineering, and Medicine (2020); Antonucci et al. (2014).
LGBTQ+ older adults and chosen families
For some LGBTQ+ older adults, friendships and chosen-family relationships have long provided affirmation and practical support in the context of discrimination, family estrangement, or exclusion from institutions. Erosheva and colleagues examined network size and diversity among lesbian, gay, bisexual, and transgender older adults; their findings show substantial heterogeneity and associations with social resources, religious participation, employment, service use, and identity disclosure. Related work identified varied network types, including friend-centered arrangements with limited child or biological-family ties. These patterns indicate that assessing a person's network exclusively through spouse-and-children questions can obscure essential relationships. Erosheva et al. (2016); Social Network Types and Mental Health Among LGBT Older Adults (2017).
An inclusive approach asks each person whom they recognize as important, who may be contacted in an emergency, and who should be included in decisions with the person's consent. Community acceptance, nondiscriminatory services, and safe meeting spaces influence the practical continuity of these networks. There is no single LGBTQ+ aging trajectory, and variations by ethnicity, gender identity, income, rural residence, disability, and cohort remain significant. Erosheva et al. (2016); WHO (2025).
Face-to-Face Friendship, Digital Contact, and Hybrid Networks
Telephone calls, messaging, video conversations, online interest communities, and social media can help sustain friendships across distance or restricted mobility. Digital communication may be especially useful for contacting former coworkers, migrating relatives, distant school friends, and people who share uncommon interests. It can also introduce barriers involving device costs, inaccessible interfaces, connectivity, privacy, fraud, digital literacy, or the preference for in-person interaction. Digital contact is a mode of relationship maintenance; the presence of an account, follower, or frequent message does not establish intimacy or reliable support. National Academies of Sciences, Engineering, and Medicine (2020); WHO (2025).
A hybrid friendship may combine text messages with visits, telephone calls with shared activities, or video meetings with future in-person contact. The relevant outcomes include the person's experience of connection, ability to participate, sense of safety, and continuity of mutual exchange. Reviews of interventions involving technology find possible benefits for loneliness and support in some settings, with heterogeneous results and varying study quality. These data do not show that a digital format is superior for all older adults or that a device purchase itself produces friendship. Hoang et al. (2022); Interventions against loneliness and social isolation in older adults (2026).
Technology-related companionship requires conceptual precision. Communication with another person through a digital tool is an ordinary human friendship mediated by technology; interaction with an automated conversational system raises distinct psychological and ethical questions. The latter belongs to the separate research domain of AI and aging. This article examines relationships between people and the settings in which their interactions occur. Digital inclusion matters insofar as it changes access to communication and participation. WHO (2025).
What Longitudinal and Cross-Sectional Studies Can Establish
Cross-sectional surveys compare individuals of different ages at a given point in time. They can estimate how reported network size or friendship satisfaction differs between age groups, but those differences can reflect birth cohort, historical opportunity, education, survival, or survey participation. Longitudinal research follows people over time and can describe changes within individuals, including the persistence or disappearance of ties across retirement, widowhood, or relocation. Repeated measurements improve developmental inference but do not by themselves eliminate reverse causation, unmeasured confounding, or selective attrition. Older participants who leave a study because of illness, displacement, or declining access may differ systematically from those who remain. Wrzus et al. (2013); English and Carstensen (2014).
Network surveys often rely on a respondent naming close contacts. The number reported depends on the question's wording, recall period, cap on nominations, and whether spouses and family are included. A measure of people available for emergency help captures a different feature from a measure of daily companionship. Some researchers seek information from both respondents and the people they nominate, but such designs raise recruitment and confidentiality challenges. Ayalon and Levkovich's systematic review of social-network studies using information from both parties highlighted the relatively specialized evidence base. A later systematic review of measures identified many partially overlapping dimensions and instruments, demonstrating why apparent agreement on “network size” can conceal substantive differences in measurement. Ayalon and Levkovich (2019); Siette et al. (2021).
Scientific analyses should distinguish changes in individual ties from changes in totals. A stable count of three friends may conceal the loss of two lifelong companions and the formation of two new connections; a smaller count can occur when low-quality peripheral ties end while close relationships endure. The identity, duration, quality, location, and functions of ties provide information that counts alone cannot capture. Researchers should also include people who live with disability, in institutions, in rural communities, or without digital access. Cultural and language validation, gender-inclusive categories, and sensitivity to migration and historical disruption all affect whether the resulting science represents the people it describes. Antonucci et al. (2014); Siette et al. (2021).
Supporting Friendship: Evidence and Practical Applications
Meaningful support begins with the person's preferred relationships and specific obstacles. One older adult may want help reaching an existing friend; another may seek companions for an unfamiliar activity; a third may value solitude while needing a reliable emergency contact. These objectives require different responses. Arranging accessible transport, reducing hearing-related communication barriers, offering language-inclusive community programs, improving public seating and safety, and sustaining affordable gathering places can increase opportunity. A program that merely records attendance cannot establish that its participants have formed mutually satisfying friendships. WHO (2025); National Academies of Sciences, Engineering, and Medicine (2020).
Practical strategies for preserving existing friendships
Continuity is more achievable when the form of contact can adapt. People can agree on feasible ways to communicate, arrange recurring but flexible opportunities for contact, recognize that care responsibilities sometimes interrupt plans, and choose settings compatible with mobility or sensory needs. A phone conversation can preserve a valued relationship during temporary illness; visits may become shorter but more regular; community transport may make occasional shared activities possible. Reciprocity can be maintained through emotional attention and mutual choice even when one person provides more physical assistance. These examples are practical applications of the relational framework rather than individually validated clinical prescriptions. Antonucci et al. (2014); WHO (2025).
Conditions that support new friendship formation
Programs are more likely to create conditions for friendship when they enable repeated voluntary contact, shared interests, accessible communication, a manageable cost of participation, and freedom to develop relationships outside the program. A class may introduce potential friends; small-group activity can allow people to become familiar; informal conversation after an event offers an opportunity to continue interacting. Mentoring, community organizing, arts, neighborhood projects, and peer initiatives can support a sense of contribution and equality. For many participants the most meaningful outcome is a relationship that remains satisfying after the organized activity ends. Whether any particular program achieves that outcome should be tested, not assumed. Baeriswyl and Oris (2023); WHO (2025).
What intervention studies actually show
Randomized evidence is stronger for selected interventions targeting loneliness and social isolation than for an intervention specifically guaranteeing durable new friendships. A 2022 systematic review and meta-analysis included 70 intervention studies and found that several approaches were associated with lower loneliness, though study designs and results were heterogeneous. A 2024 meta-analysis synthesized 67 studies and estimated an overall improvement in loneliness for some interventions while identifying persistent heterogeneity and uncertain mechanisms. A 2026 systematic review of studies published during and after the pandemic reported comparatively promising findings for community-based and combined analog-digital approaches, with a need for tailoring and longer-term evaluation. These outcomes are related to, but not identical with, friendship quality or formation. Hoang et al. (2022); Facing the Next “Geriatric Giant” (2024); Interventions against loneliness and social isolation in older adults (2026).
A sound assessment asks whether an intervention increased desired contact, improved relationship satisfaction, facilitated new reciprocated ties, or merely changed a short-term loneliness score. It also asks who could not participate, whether positive effects persisted, which resources were required, and whether measured benefits were offset by burden or cost. The combination of accessible environments, individual preferences, and sustained opportunity is more credible than the claim that a single activity will eliminate social difficulties. WHO (2025); Patil and Braun (2024).
Roles for families, friends, communities, and services
Families can help preserve a person's existing friendships by respecting privacy, facilitating requested communication, avoiding unnecessary gatekeeping, and recognizing chosen companions as meaningful. Health and social-care services can ask about desired support, assess hearing and mobility barriers, provide information about accessible community resources, and refer for appropriate clinical evaluation when distress or functional change requires it. Municipalities and community organizations can influence opportunity through public transport, libraries, safe walking routes, affordable social spaces, inclusive cultural events, and stable community services. These roles are complementary and should be organized around autonomy and measurable access. National Academies of Sciences, Engineering, and Medicine (2020); WHO (2025).
Evidence Strength and Open Scientific Questions
Established evidence indicates that social networks change over adulthood, that peripheral ties often decline more than close ties, and that friendship patterns vary widely between individuals. Established observational research also documents associations between meaningful social relationships and several well-being or health outcomes. Evidence for socioemotional selectivity and convoy processes is substantial, although their relative explanatory contributions depend on the population and research design. Wrzus et al. (2013); English and Carstensen (2014); Holt-Lunstad et al. (2010).
Preliminary or limited evidence concerns the comparative effectiveness of particular friendship-building programs, optimal digital formats, and the durability of changes after organized interventions end. Findings from trials primarily measuring loneliness cannot be transferred automatically to friendship formation. Some interventions are promising, but heterogeneity and differences in measurement remain substantial. The best program for a particular person depends on access, interests, cultural context, functional needs, and willingness to participate. Hoang et al. (2022); Patil and Braun (2024); Interventions against loneliness and social isolation in older adults (2026).
Contested interpretations include whether decreases in network size generally reflect adaptive selection, structural exclusion, or some combination; whether one preferred network composition exists across cultures; and how strongly friendship independently affects health once selection and confounding are considered. Historical changes in the prevalence of friendship demonstrate why a universal model based on one cohort is insufficient. Better evidence will require repeated tie-level measurements, more diverse longitudinal panels, the inclusion of residential-care and disabled populations, and evaluations that ask older adults which relationships matter to them. Baeriswyl and Oris (2023); Ayalon and Levkovich (2019); Siette et al. (2021).
Frequently Asked Questions
Do older adults usually have fewer friends?
Meta-analytic and longitudinal research finds smaller friendship or personal-network sizes on average with increasing adult age, particularly through reductions in peripheral contacts. Close companions often show more continuity. People vary substantially, and a smaller circle may reflect a preference for familiar relationships, lost access to social settings, life events, or several interacting causes. Network size does not measure relationship satisfaction. Wrzus et al. (2013); English and Carstensen (2014).
Can someone make close new friends after retirement or in very late life?
Yes. New friendships can develop through repeated contact, shared activities, mutual trust, and appropriate opportunities. Retirement may expand time for some relationships while removing routine workplace encounters. Disability, relocation, economic barriers, and cultural expectations influence opportunities but do not create a universal upper age limit for friendship. Baeriswyl and Oris (2023); Antonucci et al. (2014).
Is a small circle of close friends sufficient?
A small circle can be deeply satisfactory when relationships meet the person's wishes and the network supplies appropriate practical support. There is no universally established ideal number of friends. The useful questions concern trust, reciprocity, reliability, desired contact, and the availability of additional help if an important companion becomes unavailable. Antonucci et al. (2014); WHO (2025).
Is loneliness the same as having few friends?
No. Loneliness is a distressing mismatch between desired and experienced connection, whereas a small network describes an aspect of social structure. A person may have a large network and feel lonely, or prefer a limited number of relationships and feel satisfied. Assessments should measure both experience and actual access. National Academies of Sciences, Engineering, and Medicine (2020); WHO (2025).
Do friendships protect against dementia or guarantee longer life?
Studies associate social relationships with a range of health outcomes, including survival, but observational findings cannot guarantee protection or establish that forming a particular friendship prevents a clinical disorder. Cognitive symptoms and suspected dementia require specialized assessment. Friendship is a meaningful component of social life, and health claims should remain proportional to the evidence. Holt-Lunstad et al. (2010); Kelly et al. (2017).
Are online friendships as meaningful as in-person friendships?
Their value depends on the relationship, the person's preferences, accessibility, privacy, and the possibility of sustained reciprocal contact. Online communication can preserve distant friendships or introduce compatible people, while digital exclusion can limit access. Intervention reviews show varying outcomes and do not justify a universal ranking of digital and in-person formats. Hoang et al. (2022); WHO (2025).
How should someone respond when an older relative seems to have lost touch with friends?
Begin by asking what relationships the person values and whether the change is distressing. Explore practical barriers such as transport, hearing, safety, cost, scheduling, or reluctance to use a particular technology. Offer assistance that the person accepts and preserve choice about who to contact. When withdrawal is sudden, accompanied by major changes in functioning or mood, or raises safety concerns, appropriate health evaluation may be warranted. National Academies of Sciences, Engineering, and Medicine (2020).
Conclusion: Relational Continuity Within Changing Lives
Friendships in older adulthood are living relationships shaped by duration, choice, history, environment, and continuing development. Population averages show selective changes in the number and type of social ties, but individual trajectories include enduring companionship, reconnection, new friendships, altered forms of contact, involuntary loss, and chosen privacy. Explaining these patterns requires combining social motivation with life-course transitions and the institutional conditions that make contact possible. Sound research places quality, reciprocity, satisfaction, and personal autonomy alongside network counts. Wrzus et al. (2013); Antonucci et al. (2014).
A humane and empirically responsible approach to later-life friendship protects both the right to meaningful connection and the right to choose its form. Community design, accessible communication, inclusive services, and sustainable social opportunities can support those choices, while continued research determines which interventions create lasting relationships for whom. The history of a friendship can be a source of continuity, and the possibility of future friendship remains part of adult development. WHO (2025); Baeriswyl and Oris (2023).
Related Articles
Cluster Adult Development, Aging and Later Life — The overarching lifespan research chapter on adult development and psychological aging.
Loneliness and Social Isolation in Older Adults: Differences, Risks, and Support — Perceived loneliness, objective isolation, risks, and age-specific approaches to support.
Healthy Aging and Psychological Well-Being: Functional Ability and Adaptation — Functional ability, environments, and heterogeneous trajectories of well-being.
Ageism Against Older Adults: Stereotypes, Discrimination, and Prevention — Social exclusion and age-based barriers to connection and participation.
Grandparenthood and Psychological Well-Being: Roles and Relationships
References
Antonucci, T. C., Ajrouch, K. J., and Birditt, K. S. (2014). The convoy model: Explaining social relations from a multidisciplinary perspective. The Gerontologist, 54(1), 82–92. https://doi.org/10.1093/geront/gnt118
Ayalon, L., and Levkovich, I. (2019). A systematic review of research on social networks of older adults. The Gerontologist, 59(3), e164–e176. https://doi.org/10.1093/geront/gnx218
Baeriswyl, M., and Oris, M. (2023). Friendship in later life: Thirty years of progress and inequalities. The International Journal of Aging and Human Development, 96(4), 420–446. https://doi.org/10.1177/00914150221092991
Bergsträsser, J., Schmahl, T., Steinhäuser, J., and Goetz, K. (2026). Interventions against loneliness and social isolation in older adults: A systematic review. BMC Public Health, 26, 1562. https://doi.org/10.1186/s12889-026-27683-9
Carstensen, L. L., Isaacowitz, D. M., and Charles, S. T. (1999). Taking time seriously: A theory of socioemotional selectivity. American Psychologist, 54(3), 165–181. https://pubmed.ncbi.nlm.nih.gov/10199217/
Duffner, L. A., Janssen, N., Deckers, K., Schroyen, S., de Vugt, M. E., Köhler, S., Adam, S., Verhey, F. R. J., and Veenstra, M. Y. (2024). Facing the next “geriatric giant”: A systematic literature review and meta-analysis of interventions tackling loneliness and social isolation among older adults. Journal of the American Medical Directors Association, 25(9), 105110. https://doi.org/10.1016/j.jamda.2024.105110
English, T., and Carstensen, L. L. (2014). Selective narrowing of social networks across adulthood is associated with improved emotional experience in daily life. International Journal of Behavioral Development, 38(2), 195–202. https://doi.org/10.1177/0165025413515404
Erosheva, E. A., Kim, H.-J., Emlet, C., and Fredriksen-Goldsen, K. I. (2016). Social networks of lesbian, gay, bisexual, and transgender older adults. Research on Aging, 38(1), 98–123. https://doi.org/10.1177/0164027515581859
Fung, H. H., Carstensen, L. L., and Lang, F. R. (2001). Age-related patterns in social networks among European Americans and African Americans: Implications for socioemotional selectivity across the life span. The International Journal of Aging and Human Development, 52(3), 185–206. https://doi.org/10.2190/1ABL-9BE5-M0X2-LR9V
Guadalupe, S., and Vicente, H. T. (2021). Social network typologies of older people: A cross-national literature review. Ciência and Saúde Coletiva, 26(Suppl. 3), 5133–5148. https://doi.org/10.1590/1413-812320212611.3.23072019
Hoang, P., King, J. A., Moore, S., et al. (2022). Interventions associated with reduced loneliness and social isolation in older adults: A systematic review and meta-analysis. JAMA Network Open, 5(10), e2236676. https://doi.org/10.1001/jamanetworkopen.2022.36676
Holt-Lunstad, J., Smith, T. B., and Layton, J. B. (2010). Social relationships and mortality risk: A meta-analytic review. PLOS Medicine, 7(7), e1000316. https://doi.org/10.1371/journal.pmed.1000316
Huxhold, O., Miche, M., and Schüz, B. (2014). Benefits of having friends in older ages: Differential effects of informal social activities on well-being in middle-aged and older adults. The Journals of Gerontology: Series B, 69(3), 366–375. https://doi.org/10.1093/geronb/gbt029
Kelly, M. E., Duff, H., Kelly, S., McHugh Power, J. E., Brennan, S., Lawlor, B. A., and Loughrey, D. G. (2017). The impact of social activities, social networks, social support and social relationships on the cognitive functioning of healthy older adults: A systematic review. Systematic Reviews, 6, 259. https://doi.org/10.1186/s13643-017-0632-2
Kim, H.-J., Fredriksen-Goldsen, K. I., Bryan, A. E. B., and Muraco, A. (2017). Social network types and mental health among LGBT older adults. The Gerontologist, 57(Suppl. 1), S84–S94. https://doi.org/10.1093/geront/gnw169
National Academies of Sciences, Engineering, and Medicine. (2020). Social isolation and loneliness in older adults: Opportunities for the health care system. National Academies Press. https://doi.org/10.17226/25663
Patil, U., and Braun, K. L. (2024). Interventions for loneliness in older adults: A systematic review of reviews. Frontiers in Public Health, 12, 1427605. https://doi.org/10.3389/fpubh.2024.1427605
Siette, J., Pomare, C., Dodds, L., Jorgensen, M., Harrigan, N., and Georgiou, A. (2021). A comprehensive overview of social network measures for older adults: A systematic review. Archives of Gerontology and Geriatrics, 97, 104525. https://doi.org/10.1016/j.archger.2021.104525
Tambellini, E., Danielsbacka, M., Tanskanen, A. O., and Rotkirch, A. (2026). Life after loss: Gendered changes in social networks and community participation around widowhood. The Journals of Gerontology: Series B, 81(10), gbag141. https://doi.org/10.1093/geronb/gbag141
World Health Organization. (2025). From loneliness to social connection: Charting a path to healthier societies. Report of the WHO Commission on Social Connection. https://www.who.int/publications/i/item/978240112360
Wrzus, C., Hänel, M., Wagner, J., and Neyer, F. J. (2013). Social network changes and life events across the life span: A meta-analysis. Psychological Bulletin, 139(1), 53–80. https://doi.org/10.1037/a0028601