Some research suggests that different kinds of social ties perform specific functions (i.e., relationship specialization), leading to the idea that there are “customary” sources of support and companionship (Weiss, 1974). In later life, losing a social tie results in losing that support or companionship, unless older adults are able to compensate by finding an alternative (or non-customary) source (Rook & Charles, 2017). Little research has examined the extent of relationship specialization or its implications for well-being, such as loneliness, among married versus unmarried older adults. The current study accordingly examined three questions: To what extent do patterns of support and companionship reflect relationship specialization? To what extent is loneliness elevated when customary sources of support and companionship are absent? Can non-customary sources of companionship and support compensate for the lack of customary sources of companionship and support? Data from wave 1 in-person interviews of a nationally representative, longitudinal study of 916 older adults were analyzed using 3 x 2 x 2 Repeated Measures Multivariate Analysis and Multiple Regression Analyses. Married and unmarried older adults experienced similar patterns of relationship specialization, except with regard to emotional support, which was provided equally across types of relationships among the unmarried, and more often by family members among the married. Spouses performed more support and companionship functions than did friends or children. Additionally, loneliness was elevated among participants who lacked companionship and instrumental support from friends, as well as among participants who lacked emotional support from either friends or family members. Among unmarried participants, however, loneliness was less elevated if they received instrumental from non-customary sources. Receiving companionship from alternative sources buffered feelings of loneliness, regardless of marital status. Loneliness was also buffered among unmarried participants when they received emotional support from alternative sources. Married participants, contrary to expectations, experienced less loneliness when support was provided not only by non-customary sources, but also by customary sources. The results of this study provided partial evidence for relationship specialization in older adults’ social networks, and also suggested that relationship specialization does not constrain effective compensation to some extent.

Background: Invisible social support (i.e., support not recognized by recipients) has been posited to avoid the costs of receiving support, which include distress and reduced feelings of self-efficacy. The effect of invisible support has primarily been studied in the context of acute stressors. Moreover, little research has examined invisible versus visible social control (i.e., others’ efforts to improve a health behavior). Purpose: The current study examined the implications of invisible and visible social support and control for negative affect and self-efficacy in the context of a chronic stressor, namely the management of type 2 diabetes. Methods: Daily diary data was collected from 129 older couples in which one spouse had type 2 diabetes. Results: Invisible support and invisible control were not associated with negative affect, and invisible control was associated with lower, rather than greater, patient self-efficacy. Conclusion: The results suggest that the benefits of invisible social support and invisible social control may not extend to chronic illness.

In later life, the loss of social ties is particularly consequential for emotional health. Widowhood becomes increasingly common as people age, and divorce rates have been on the rise among older adult populations. Such losses, particularly in the absence of remarriage, reduce older adults’ access to support and companionship. Research on spousal loss has yet to consider variations in perceived need for support and companionship. It is also possible that having additional support and companionship beyond a minimum threshold reduces need for support and companionship and contributes little to older adults’ emotional health. A new and growing form of social contact, remote interactions, is becoming increasingly important, however, research has yet to examine whether older adults use remote interactions to garner support and companionship following widowhood and divorce. To address these gaps in the literature, the current dissertation used two waves of data sourced from the UC Berkeley Social Networks Study (UCNets). A subsample of community-dwelling older adults aged 60-70 years (N = 411) were interviewed about their social networks (i.e., support and companionship providers and marital status), remote interactions (e.g., text, email, or other online methods), and emotional health (i.e., days felt isolated, days felt lonely, and psychological distress). Analyses revealed that widowed/divorced older adults experience a greater need for support and companionship than do married older adults. Current support and companionship did not change widowed/divorced older adults’ need for support and companionship. Contrary to expectations, experiencing a greater need for support and companionship was associated with decreased support and companionship providers over time. Even when considering current support and companionship, the duration of widowhood/divorce was not associated with a need for support and companionship. Remote interactions were significantly associated with emotional health, but only for older adults whose number of support and companionship providers was insufficient (i.e., below a threshold). Marital status (widowed/divorced vs. married) did not influence the association between remote interactions and emotional health. These findings convey the importance of examining need for support and companionship as an antecedent to substitution and compensation following spousal loss. Future research examining remote interactions should take into consideration the joint effects of in-person support and companionship.

A prevailing view exists in the literature that having a variety of relationship types (i.e., diverse social ties) is more consequential for health and activity than the sheer number of social ties. Empirical tests of that claim are relatively scarce, however, and relationship quality is rarely taken into account. The current study contrasted the implications of number and diversity of positive, as well as ambivalent, social ties for functional limitations and leisure activity in later life. Participants (N = 874) were part of the Later Life Study of Social Exchanges (LLSSE), a longitudinal study of a nationally representative sample of older adults ages 65 to 91 (M = 74.12 years old, SD = 6.64). Baseline data from in-person interviews were examined in a series of hierarchical regression models. Number of ties was a more consistent predictor of functional limitations than was diversity of ties (albeit in opposite directions for positive and ambivalent ties), but they did not significantly differ in terms of magnitude. Number of positive ties was a significantly stronger predictor of leisure activity compared to diversity of positive ties, whereas ambivalent ties were not related to leisure activity. Contrary to prevailing views, these findings suggest that variety is not necessarily more important than quantity. Rather, a greater number of positive social ties may be most important for older adults’ health and activity.

Relationships that are sources of both positive and negative experiences (i.e., ambivalent ties) are understudied, and findings are mixed regarding their influence on health. This dissertation, accordingly, examined whether exposure to ambivalent ties differentially relates to health outcomes depending on how it is operationalized (Study 1), whether the link between ambivalent ties and cognitive functioning is moderated by interpersonal coping (Study 2), and whether the daily coupling of ambivalent ties and health limitations is moderated by affect valuation (Study 3). Community-dwelling older adults in the greater Austin, Texas area (N = 333 at baseline, ages 65-92 years old) completed an in-person interview at baseline, followed by a self-administered questionnaire (returned by mail), and 5-6 days of ecological momentary assessment (EMA) surveys. Findings revealed that the proportion of ambivalent ties in one’s total network was the best operationalization of exposure to ambivalent ties in the current dataset (Study 1); ambivalent ties were related to poor cognitive functioning, a link not readily buffered by coping strategies (Study 2); and the daily coupling of ambivalent ties and health limitations was buffered by affect valuation (Study 3). Implications for models of ambivalent ties and health, as well as practical implications for older adults’ health and well-being, are discussed.

Unlabelled

Learning outcomes