Social isolation and loneliness in later life: A parallel convergent mixed-methods case study of older adults and their residential contexts in the Minneapolis metropolitan area, USA

Highlights

• Social isolation and loneliness are influenced by personal and neighborhood factors.

• Qualitative findings explain variations in trends identified in quantitative models.

• Poor health can have both negative and positive effects on social isolation.

• Some socially isolated older adults seek solitude and are content, not lonely.

• Results are relevant for strategies to address social isolation and loneliness.

Abstract

Social isolation and loneliness are increasingly prevalent among older adults in the United States, with implications for morbidity and mortality risk. Little research to date has examined the complex person-place transactions that contribute to social well-being in later life. This study aimed to characterize personal and neighborhood contextual influences on social isolation and loneliness among older adults. Interviews were conducted with independent-dwelling men and women (n = 124; mean age 71 years) in the Minneapolis metropolitan area (USA) from June to October, 2015. A convergent mixed-methods design was applied, whereby quantitative and qualitative approaches were used in parallel to gain simultaneous insights into statistical associations and in-depth individual perspectives. Logistic regression models predicted self-reported social isolation and loneliness, adjusted for age, gender, past occupation, race/ethnicity, living alone, street type, residential location, and residential density. Qualitative thematic analyses of interview transcripts probed individual experiences with social isolation and loneliness. The quantitative results suggested that African American adults, those with a higher socioeconomic status, those who did not live alone, and those who lived closer to the city center were less likely to report feeling socially isolated or lonely. The qualitative results identified and explained variation in outcomes within each of these factors. They provided insight on those who lived alone but did not report feeling lonely, finding that solitude was sought after and enjoyed by a portion of participants. Poor physical and mental health often resulted in reporting social isolation, particularly when coupled with poor weather or low-density neighborhoods. At the same time, poor health sometimes provided opportunities for valued social engagement with caregivers, family, and friends. The combination of group-level risk factors and in-depth personal insights provided by this mixed-methodology may be useful to develop strategies that address social isolation and loneliness in older communities.

Introduction

Social isolation and loneliness are increasingly recognized in academic literature and popular media discourse as risks to physical health and well-being among older adults in the United States. Approximately one-third of Americans aged ≥60 years are estimated to feel lonely (Wilson and Moulton, 2010), and one-quarter of those aged ≥65 years are estimated to live alone (Stepler, 2016). Social isolation is defined as a measurable lack of social relationships, while loneliness is an affective state reflecting the subjective experience of feeling alone or lonely (Hawkley and Cacioppo, 2007; Klinenberg, 2016; Steptoe et al., 2013). Both constructs have been associated in large population-based studies of older adults with a range of health outcomes, including risks for dementia, cardiovascular disease and stroke, loss of physical mobility, and all-cause mortality (Pantell et al., 2013; Steptoe et al., 2013; Holt-Lunstad et al., 2015; Shankar et al., 2017; Rafnsson et al., 2017; Valtorta et al., 2016). Loneliness and social isolation have often been attributed to factors such as marital and family circumstances, economic status, and health status. Struggles with both cut across gender, race/ethnicity, social class, and geographic locations among older adults; and may be connected to broader trends in declining social integration, civic engagement, and social capital in American communities in recent decades (Goll et al., 2015; Fokkema et al., 2012; Cornwell et al., 2008; Berkman et al., 2000; Putnam, 1995; McPherson et al., 2006).

Feelings of loneliness may be caused by social isolation, but this is not always the case among older adults. Loneliness and social isolation have only a weak-to-moderate positive correlation within individuals (Steptoe et al., 2013; Cornwell and Waite, 2009). Although the prevalence of living alone increases with age in the United States, loneliness in fact decreases with age from 43% of those aged 45–49 years to 25% of those aged ≥70 years (Stepler, 2016). Social isolation and loneliness may be increasingly ‘decoupled’ as older adults expect and prepare for diminished social networks as their peers begin to decline in the physical and mental capacities needed for engagement (Cornwell and Waite, 2009; Achenbaum and Bengston, 1994).

The personal and neighborhood factors that may influence isolation and loneliness as separate outcomes for older adults have not yet been investigated in a single study. Existing research has investigated relationships between personal characteristics or broader social/neighborhood environmental characteristics and isolation or loneliness, but rarely have considered personal and contextual influences together (Garoon et al., 2016; Fokkema et al., 2012; Bromell and Cagney, 2014; Wu and Chan, 2012). Carpiano's (2006) framework of neighborhood social capital, which postulates that social capital is produced by interacting neighborhood socioeconomic processes and personal sociodemographic characteristics, is an exception to this trend. Social capital, defined by Bourdieu (1986) as the aggregate of actual or potential resources linked to possession of a durable network of institutionalized relationships, is important to recognize here as a potential determinant of social isolation and loneliness within neighborhoods. The small body of literature attending to geographic influences on social well-being in later life tends to apply either a purely qualitative (e.g., Rowles, 1978; Gardner, 2011) or quantitative (e.g., Cloutier-Fisher and Kobayashi, 2009) approach, and none have considered social isolation or loneliness as simultaneous outcomes. Further, the distinctions that individual older adults themselves make between isolation and loneliness, and their perspectives on the causes, contexts, and experiences of each, require investigation so as to help develop strategies to address these challenges in older communities.

In order to address these gaps, we aimed to identify interrelated personal and neighborhood influences on social isolation and loneliness in a community-based study of older men and women in the Minneapolis metropolitan area. Situated in the midwestern US, Minneapolis is known for its cold winters, abundant lakes, extensive park system, and cultural arts scene. The metropolitan area is home to approximately 3.5 million people (76% White, 8% Black, 6% Asian, and 6% Hispanic). The average age is 36.9 years, with 13% of the population aged 65 and over. Median household income is $73,231, and 8.8% of the total population (7% of those aged 65 and over) live below the poverty line (U.S. Census Bureau, 2016).

Three research questions guided the secondary analyses of an existing dataset (Finlay, 2017; Finlay and Bowman, 2017). First, how are social isolation and loneliness defined and experienced by older adults? Second, what personal factors contribute to or undermine reporting each of social isolation and loneliness? Third, what neighborhood contextual factors contribute to or undermine reporting each of social isolation and loneliness? We considered loneliness and social isolation separately, as their lived experiences as well as statistical neighborhood and personal correlates may differ. A parallel convergent mixed-methods design simultaneously incorporated quantitative and qualitative data. Our theoretical framework from the discipline of health geography involved a ‘relational’ approach to space and place. Informed by Cummins et al. (2007), we considered geographic neighborhood context as flexible and relational – an operational living construct that can shape lives and opportunities while being uniquely navigated by individuals. Applying a relational theory of space enabled deeper understanding of reciprocal and mutually reinforcing relationships between well-being and place, wherein older adults' neighborhoods and health statuses were inextricably linked.