The role of social relationships in predicting loneliness: the National Social Life, Health, and Aging Project.

Author:Shiovitz-Ezra, Sharon

The authors explore associations between objective and subjective social network characteristics and loneliness in later life, using data from the National Social Life, Health, and Aging Project, a nationally representative sample of individuals ages 57 to 85 in the United States. Hierarchical linear regression was used to examine the associations between measures of objective and subjective social network characteristics and their importance in predicting loneliness across marital status. For the entire sample, as well as the married-cohabitating subsample, objective indicators such as frequency of contact with social network members were negatively associated with feelings of loneliness, net of background characteristics. However, the authors' analysis highlights the contribution of subjective perceptions of social ties, the quality of marriage in later life for those engaged in marital or cohabitating relationships, and the quality of familial ties for the nonmarried older adults. In a married-cohabitating subsample, the subjective perceptions of one's relationship with the partner explained, by itself, 7% of the variance in loneliness, whereas the quality of family ties explained an additional 6% of the variance in loneliness in the nonmarried sample. Based on the present findings, practical implications for social workers are discussed.

KEY WORDS: later life; loneliness; social networks; subjective perceptions


In research, multiple definitions of loneliness have been proposed. Yet almost uniformly, a distinction has been made between loneliness and aloneness (Andersson, 1998). Loneliness is approached as a discrete, subjective construct associated with the objective social situation but not synonymous with the actual circumstances. Thus, people can feel lonely in the company of many others or be alone without feeling lonely. It also has been agreed that loneliness is a painful or unpleasant experience (Peplau & Perlman, 1982).

Cognitive theory, which is a major theoretical approach guiding loneliness research, focuses on one's perception and evaluation of social ties. Loneliness, according to this theoretical point of view, results from the perceived discrepancy between desired and actual social relationships or the subjective gap between one's optimal levels of social relationships and achieved levels of same. A perceived deficit in one's social interactions is crucial in creating a sense of loneliness. Past experience and experience of other people in the social environment shape this evaluation process. Contrary to other important theoretical views of loneliness, such as the social needs approach (represented by Weiss, 1973, 1987), cognitive theory suggests an indirect relation between objective deficits in one's social network and feelings of loneliness. Cognitive processes of perception and evaluation serve as mediators (de Jong Gierveld, 1998; Peplau & Perlman, 1982).

In recent years, interest in aging and loneliness has grown for two primary reasons. First, loneliness is a socially prevalent phenomenon among elderly people. For example, in a representative sample of British community-dwelling older people, almost 40% experienced loneliness to some degree (Victor, Scambler, Bowling, & Bond, 2005). Similar prevalence rates were recently found in Finland among a random sample of people age 75 and over (Savikko, Routasalo, Tilvis, Strandberg, & Pitkala, 2005). In the United States, Theeke (2007) reported that approximately 17% of people age 50 and above reported loneliness.

Second, loneliness can have deleterious effects in multiple domains of elders' lives, including both physical and mental health. In the Chicago Health, Aging, and Social Relations Study, loneliness was associated with elevated systolic blood pressure, even after controlling for demographics, health behaviors, and various psychosocial factors. Moreover, loneliness was a unique predictor of age-related increases in systolic blood pressure (Hawkley, Masi, Berry, & Cacioppo, 2006). Loneliness was also found to affect cardiovascular activity in everyday life by predicting higher basal total peripheral resistance and lower cardiac output (Hawkley, Burleson, Berntson, & Cacioppo, 2003). Moreover, loneliness has predicted mortality in elderly people in a longitudinal study (Penninx et al., 1997).

Along with effects on physical health, loneliness is associated with poor mental health. In both cross-sectional and longitudinal studies, greater loneliness was associated with higher levels of depression. Moreover, loneliness remained a risk factor for depression after central demographic and psychosocial factors were taken into account (Cacioppo, Hughes, Waite, Hawkley, & Thisted, 2006). Loneliness was also found to be negatively associated with emotional well-being (Lee & Ishii-Kuntz, 1987) and positively associated with serious thoughts of suicide and parasuicide (Stravynski & Boyer, 2001).

According to the cognitive model for conceptualizing loneliness, suggested by de Jong Gierveld (1987), two domains of the social network form the most likely determinants of loneliness among older adults. The first, termed "the descriptive characteristics of the social network," represents the objective aspect of social contacts, as measured by network size and frequency of contact. However, study of objective determinants yields inconsistent results as to their contribution to the prediction of loneliness in later life. In a recent study, levels of contact with family or friends and also proximity and number of family or friends did not demonstrate statistically significant association with self-reported loneliness (Victor et al., 2005). However, among 20-year survivors of the Bangor Longitudinal Study on Ageing, married status and children living nearby were protective factors against loneliness (Wenger & Burholt, 2004). Also, a study of 1,071 participants in the Senior Citizens Nutrition and Activities Program in Florida found that subjects with no children and no friends were lonelier than their counterparts with children and friends (Mullins & Elston, 1996). The latter is consistent with a Swedish study of elderly people age 75 and over, which demonstrated that 50% of those who reported no friends experienced loneliness (Holmen, Ericsson, Andersson, & Winblad, 1992). However, marital status had no impact on loneliness (Mullins & Elston, 1996).

The second domain emphasizes the quality of social ties, as measured by perceived availability of support and satisfaction with interpersonal relationships, and is termed the "subjective evaluations of the social network." Although both the objective and subjective network aspects were included in de Jong Gierveld's framework, she underscored the importance of the latter one, which is argued to be more closely linked to loneliness than the objective domain. Many other scholars have emphasized the importance of subjective assessments of social interactions over objective social circumstances in predicting well-being outcomes (for example, Antonucci & Akiyama, 1991; Bowling, 1994), particularly for loneliness (Cutrona, 1982; Peplau & Perlman, 1982).

Nevertheless, studies most frequently use objective measures of social networks as determinates of loneliness; the subjective domain is less often tested (Pinquart & Sorensen, 2001). Several recent studies that incorporated subjective assessments of the network--measured by perceived social support--found statistically significant inverse associations with loneliness (Kahn, Hessling, & Russell, 2003; Kim, 1999; Tiikkainen & Heikkinen, 2005). However, these studies' findings are not generalizable because of their select samples. For example, Kahn et al. (2003) used a sample of 100 older adults attending a senior community center, whereas Tiikkainen and Heikkinen's (2005) sample comprised 207 residents of central Finland at age 80.

Moreover, the literature suggests that social ties are not always beneficent and supportive, but can also be hurtful (Rook & Pietromonaco, 1987). Beneficent and negativity characteristics can co-occur in the same relationships (Ruehlman & Wolchik, 1988). Thus, there is a serious need to explore the positive and negative aspects of social interactions. This arena was explored previously in relation to depressed mood (Schuster, Kessler, & Aseltine, 1990), life satisfaction, and positive and negative mood (Walen & Lachman, 2000), but not in regard to loneliness in later life.

In sum, there seems to be no prevailing consensus about the contribution of the objective domain of social relationships in relation to loneliness. There is also insufficient knowledge concerning the contribution of subjective quality of the same ties. The current study, using contemporary data from the recently implemented National Social Life, Health, and Aging Project (NSHAP), contributes to loneliness research in two ways. It explored both the objective and the subjective characteristics of older adults' social networks. The latter aspect was tested in a comprehensive manner (perceived support and strain) while using a large, generalizable sample. This latter aspect is particularly important, because much of loneliness research has been criticized as stemming from locally based samples, thus limiting the generalization of the various findings (Victor et al., 2005).


The current study used data from the first wave of NSHAP that is focused on older...

To continue reading